ML20249A525
| ML20249A525 | |
| Person / Time | |
|---|---|
| Issue date: | 06/11/1998 |
| From: | NRC ADVISORY COMMITTEE ON NUCLEAR WASTE (ACNW) |
| To: | |
| References | |
| NACNUCLE-T-0123, NACNUCLE-T-123, NUDOCS 9806170010 | |
| Download: ML20249A525 (300) | |
Text
a . ;. - .
. : 4
-a . . -
..... ...L .
- ._ _ ..'. . ~ . . . . . -
'l.[.dki.k@hyhfNV '. ~ <M ~ Y ' W
.s....-,..
4 :% '
? v ! :._g;:f. * ;g; ,,,p;,7:p.g::.: ? n :- i;. .
$^ ;
" \hiftl l. '.f _^
.Y o '
- *~. . ?
' i ? . '_ . '
' '_. . f ' ~ ;h -
y:
_', ~.:< . . . ' - :L : 1
.l. :,;A*f*.@;( t. ..
8 .
- y ,
,.: l $m d.
.. n- 4 :- Y -
,, - :[,.',A p4.;$ '. .:. . ,
o .~,
4 4 . ', ' [.9 ' .. g '
, k '..-
2 qq. .
f,.'.
? ',. S
_b .e,
.I , .
. .v s. g. , .
_.g
' ~
- - V . .> -
?. n: . M ,y -
.' . . ~
[,,
e.
s W& : h,:' Q'. y e f. 'y" -
E ,
f,,.
.J4gs. .A .,m. . J.] - > g
+ . . .- ..
n.. *
.g..g,g . ., ,
[ 1 } . ,.3 ..- -
' ).)] . %. y ! . ':; m
. - [ ]2i I [ ~ . ' Y ' '. ..'l' .
p c.s
. : . w,
'>.<. ... t - s 4 vs ,;; '
g: 1 *, ,
- h .,
4,* [ '...,,', - ' , ' . ., .-
, . . . ) 3 ' ., ; . .'.4..
- m. p ' ;. .
.~ .
~
g m
- . : n. .. .. , r . *
$ W l q: 1- l . . $ .. .y; $.c \ ..
Av1 ._~f._# 3.4 r .
. f.-
~
'b.
3
[' mep - :
he- .
. j y . a,. >
.s
'l'"
b h 'g- '. . !; n.k: :i ' ' " '
~~'\,'
~
yf(tra
~
f r:Elg [:;Bh? * ] [Qy'
~
W W l.A R" . j' g%
$ ~.ffd ,, s ; p. . S. .-
- - f.-I .-sm
' ;) '
w %. -.. t : ., : -
..yp;gn. q.]',,...e b .
. * [-
.s..'
- ., .7 g ; . 3,.--
u :L - -
a d.. -
c ~.. '
' N-0 : :* '
'4 '. _lfo .'l
']y .&:;,,.,.<;.(.a L -
~ . -
. y.
lQ:.Y. ; ;*.l
'f;q;;5. *;4 .gn e
- t w ~ '
n..: - .. :. - -
.r...,. .1.
n
- r. c. w . .,.t,.. .. . . . . , - % m4?.3 .e
. ~. 'c ^ s , -
- 4: ,"
} y , & . .
- , $. . ... .. g4)
.. .. ., , .3$ ~
^:.t ,;M : :;
.& .n s .; .:- . - -
k:fn. , P...." ). ?';
a s .
9l g. . 2- , 7
. - 'J y&,,M:iV f !WW 'Q'fb; P.
. & b '
i::4+#je ~ - +
/e4v 141 ,( - .4 ..
p ;%',',. ,
- f.&:\'%.m * * '
N;:<v u;) W,.
"- ~
- v. -
a .
.t:
<,b. A Mi,
...~
.. M:w.:.:, .v
,g gr . .s . .
Or . .
.M;;;,w:j,'$.v.
,if. f.
^
y gf ai y .
- - - . , e i 33 . ...
q a.;y _ A ' g
- : ~ -
g-oSQ A. .
./ g$ ,, -- '
h.n, Milh I - -
.O ^ .- . 1
- s &:q; Y= '.,A i . , -
} tu.M. t
.W, .a . .. 4 .3, s , a.w w.c., .
. <. g, . ..;
y (.
l k,.:p. -.' a 6: ' . .
- l J =~ p .
' p :)
$ kn '
'0l, -h v ', '.ll-fQ v-
' ';jf -
, :_ t e.3: . .: t ..; . . . .
^
MM%RW%w a h%
..;-. JS .
.. 4 . . . y.
....m t.
.n
- .:.. c .m: . : t:y', q..gy. y.-:1Q ::... ;..,. p ,;,-
- y .y.
f
- w. . .
^'
, Y.^:b'k
k :. ;
c . . s. . u. v . ,.. . ::;; h. :.:.n;6.. y..q. .s. ..I;- ;ha v.':
i~ . # q. . . , y: O.i ac
.7 ,, :. r. %. -
t .s.'
".r. ~ : _ ; ._.;.g.9 .. v, ,%. i. ? .- .. .; .:
.. : : . ..:. .;; .. :;+ . ; u.::,e:$:. :'1 7;.;; 9.f ;;;;;4.?;M; L.g:n ;y ,g:[ 7 '
. ;.. . .,e:.:; L .. v. ;;:}.p. + - :4. d;. .9; .c.y.,. :g.ff=:,-dp4,.n; ... w,.
. . , . . . . . .. ,. .. ;,w .
- .p.
u ,
1- . j . . l .~ . jl
., . . .a ., . .
. ,. - . .. v
=
< OR 3 hA_ 4MdEd/23 OFFICIAL TRANSCRIPT OF PROCEEDINGS Ov i
NUCLEAR REGULATORY COMMISSION
!- l ADVISORY COMMITTEE ON NUCLEAR WASTE L i
Title:
101ST ADVISORY COMMITTEE ON <
NUCLEAR WASTE (ACNW) MEETING Docket No.:
TR08 (ACNW)
RETURN ORIGINAL TO BJWHITE, ACRS T-2E26 7 Work Order No.: ASB-300-330 415-7130 t, THANKS!
- LOCATION: Rockville, Maryland DATE: Thursday, June 11,1998 PAGES: 305 - 568 9806170010 980611 DVCM NACN E
.-[vVU~ ANN RILEY & ASSOCIATES, LTD.
q 1250 I Street, NW, Suite 300 O
,f 'i , Washington, D.C. 20005 /'
'ly,J(a, (202) 842-0034 i 1
ACNWOFFICECOPY-RETAINFOR l q THE LIFE OFTHE COMMITTEE'2 6 .
y 3 ,. t
! I N, f DISCLAIMER UNITED STATES NUCLEAR REGULATORY COMMISSION'S ADVISORY COMMITTEE ON NUCLEAR WASTE JUNE 11, 1998 The contents of this transcript of the proceeding of the United States Nuclear Regulatory Commission Advisory f%
i J Committee on Nuclear Waste, taken on June 11, 1998, as
%/
reported herein is a record of the discussions recorded at the meeting held on the above date.
This ',ranscript had not been reviewed, corrected and edited and it may contain inaccuracies.
l 1
l I
. /~A l s .._/
r
- k. - ._ .
305 1 UNITED STATES NUCLEAR REGULATORY COMMISSION
( 2 ADVISORY COMMITTEE.ON NUCLEAR WASTE 3 ***
l 4 101ST ADVISORY COMMITTEE ON l
5 NUCLEAR WASTE (ACNW) MEETING l
6 7 U.S. Nuclear Regulatory Commission 8 Two White Flint North, Room T2B-3 9 11545 Rockville Pike 10 Rockville, Maryland 20852-2738 11 12 Thursday, June 11, 1998 13 14 The Committee met pursuant to notice at 8:02 a.m.
( 15 16 MEMBERS PRESENT: 4 17 B. JOHN GARRICK, Chairman, ACNW 18 GEORGE HORNBERGER, Member, ACNW 19 E. CHAPLES FAIRHURST, Member, ACNW 20 RAYMOND G. WYMER, Member, ACNW 21 22 23 24 25-l'\#3 ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
306' 1 - STAFF AND PRESENTERS SEATED AT THE COMMISSION. TABLE:
2' DR. ROGER STAEHLE
'3' DR. JOSEPH B. PAYER 4^ DR.-CHRIS WHIPPLE 5 MARTIN STEINDLER 6 DR. MICHAEL APTED 7L DR. WILLIAM MURPHY 8- DR. JOONHONG AHN 9 DR. DAVID W. SHOESMITH 10 ANDREW C. CAMPBELL 11 DR. SHIRLEY ANN JACKSON 12' DR. BRETT LESLIE
- 13 14 15
-16 17 18 19 20 21 22 23
- 24 25 ANN RILEY:& ASSOCIATES, LTD.
[A/ -(
Court Reporters
, .1025. Connecticut' Avenue, NW, Suite 1014 L'
-Washington, D.C.' 20036
-(202) 842-0034
~
mLi , m -m.
- _ _ J
307 1 PROCEEDINGS 2 [8:02 a.m.]
3 CHAIRMAN GARRICK: Good morning. The meeting will 4 now come to order.
5 This is the second day of the 101st meeting of the 6 Advisory Committee on Nuclear Waste. My name is John
'7 Garrick, Chairman of the ACNW.
8 Other members include George Hornberger, Raymond 9 Wymer and Charles Fairhurst. And we have with us 10 consultant, Marty Steindler.
11 Today, the Committee will continue the Working 12 Group presentations on the near-field environment.
13 As yesterday, Andy Campbell is a designated 14 Federal official for the initial portion of today's meeting,
() 15 and this meeting is being conducted in accordance with the 16 provisions of the Federal Advisory Commission Committee Act.
17 We have received no written statements or requests 18 to make oral statements from members of the public regarding 19 today's session, and should someone wish to address the 20 Committee, please make your wishes known to one of the 21 Committee's staff.
22 And it is usual -- it is requested that each 23 speaker use one of the microphones, identify themselves, and 24 speak with clarity and volume.
25 We have made some minor adjustments to the agenda I
/ y ANN RILEY & ASSOCIATES, LTD.
i_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
l L
1 I 308 l
l 1 on the basis of yesterday's meeting to include a little bit
,(qj 2 of discussion this morning on fabrication and welding, and 3 some of the activities tht?. are going on in connection-with 4 the manufacturing of the waste package. The way we've 5 chosen to do that is to squeeze that in just before the 6 break, about 15 minutes before the break, and the speakers 7 preceding that have agreed to accommodate that extension of 8 the agenda.
l 9 Actually, it's not a new agenda item, it's just 10 extending the discussion of what came up yesterday.
11 And, as yesterday, the Chairman for the Working 12 Group will be Dr. Ray Wymer, and he will introduce the 13 speakers, as well as the speaker on the fabrication at the 14 appropriate time.
(
(3,) 15 So, Ray, why don't you take over.
16 DR. WYMER: Okay. If you want to mark up your 17 agenda, your schedule a little bit, the changes we've made 18 are that Dr. Shoesmith will talk from 8:15 until 9:00, and 19 then Dr. Ahn will talk from nine to 9:45, still with the 45 20 minutes that he originally had scheduled. And Dave Staehle 21 will talk from 8:45 to 10:15, at which time, we'll take a 22 break. So that's -- those are the changes that we've made 23 in the agenda.
24 And the first speaker this morning is Dr. David 25 Shoesmith from AECL Canada, and he's going to talk on the l ANN RILEY & ASSOCIATES, LTD.
\_/
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
_ -________-____________-_---_-____w
309 1 chemistry considerations.for' release and transport of
( 2 radionuclides from spent fuel.
)
3 Dave?
4 DR. SHOESMITH: Good morning. In looking at TV's 5 here, I'm just a little bit disappointed that I'm not on 6 television'in Las Vegas. This was going to be the highlight 7 of my career, so -- I may never have been on TV in Las Vegas 8 before, but I certainly been'on' police radar on Highway 211 9 in' Manitoba a couple of times, but it's not quite the same 10 thrill for me as it is for the policemen that collects the 11 fine, but 12' [ Discussion off'the record.]
13 DR. SHOESMITH: Okay. There they are.
14 Okay. What I'd like to talk about this morning is
() 15 the chemistry considerations for release and transpct:t of 16 radionuclides from spent fuel. And you will notice that my 17 address is-somewhat ambiguous. I've left out.AECL because I 18 no longer actually work for AECL.
19 Within another week or so, I'll be working at the 20: Department of Chemistry, University of Western Ontario, 21' ;which means that I don't have to offer you the caveats that 22; Joe' Payer offered you yesterday that says that I don't speak 23' for atomic energy, and I~ don't speak for the University of 24 Western Ontario. Right now, I speak for the unemployed, but 25- -I'think that's only going to last another week or so.
) ANN RILEY & ASSOCIATES, LTD.
A K/L Court Reporters 1025-. Connecticut Avenue, NW,, Suite 1014
' Washington, D.C. 20036..
.(202)-842-0034
310 1 I found yesterday's discussion very stimulating, 2 stimulating as much in a philosophical sense as in a j\.)D 3 technical sense, especially the issue of should we engineer 4 this thing. Should we get this thing done? How much do you 5 need to know? Should we keep it simple?
6 And I wanted to quickly introduce you to what we 7 call a theory of theories, which is any theory that you 8 have, you start off with a preconception of what you think 9 the answer is. Inevitably, it always look simple in the 10 beginning, but the more you do, the greater the confusion 11 until such point as you do enough and you change your mind 12 many times down this tail end, as you get down to 13 understanding.
14 And the issue is, where should you start your 1, 0) 15 modeling. This, of course, is where we would like to start 16 our engineering. Engineers would like to believe everything 17 is relatively simple.
18 And in many cases, we know a lot about the stuff 19 that we're going to use anyway, so it is, indeed, 20 appropriate to start engineering here. But this is not 21 where you should be starting to build a model. This is 22 where you should be starting to build a model.
23 Unfortunately, the time restrictions or dictates 24 of our programs inevitably make us do what we don't want to 25 do, and we're in this sphere here, where we're not sure
,/ \ ANN RILEY & ASSOCIATES, LTD.
() _ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
311 1 'what's going on,Lwhere'it looks like we have an incredibly.
i :2' confused situation.
I's / -
3 ~So.what happens, of course, is the engineering 4 .which is,.fi nevitably will' counteract the problems that-5 you've got,,will do the job, but you entrust or encapsulate
- 6. engineering.which is expensive, not necessarily the best in
- 7 :the long term.
8 This is not an argument _that you shouldn't answer 9 this problem today. It's a suggestion;that what looks like 10 the perfect solution today at a million bucks, may not be 11 the perfect' solution down here.for half a million bucks.
12 But what I'd like to do.is take you through this a 13 little bit.with the spent fuel issue and show you, 1-4 inevitably, that as myself and Bill Halsey agreed, when you l
) 15 .
get down to it, you don't need a particularly complex model,. .
'16 . providing you know what's happened in between.
- 17. That's a little presumptuous of me to speak for 18- Bill. He may contradict everything I'm going to say.
19 Okay. This I stole from Bob Andrews at a recent 20 meeting in Albuquerque. The box -- you don't have to look 21 at all these boxes. I just want to show the number of boxes 22 and the number of important features. This is the. box that 23 I am going to spend my time talking about, what's going on 24 in waste: form' degradation, a little bit more than what's 25 .go'ng i on in transport.
L l
r -
!'. ; \m[ 61-ANN RILEY & ASSOCIATES, LTD.
Court Reporters _
1025 Connecticut Avenue, NW, Suite 1014 c . Washington, D.C.' 20036' l 3 (202) 842-0034 m
E - . . _ _ _ _ _ . _ . _ . - - - - - _ _ - - - . - _ - - _ _ _ _ - - - - . -
'312
~1 But I'd like you to notice two things, one of
[~)
v 2 which separates this program from most other people's waste 3 . management programs, and that is the number of things that 4 feed into the waste form model. Yours is an open system,'
5 which means you have to calculate how much water is going to 6 get in, how much oxygen is going to get in. You need a 7 wnole series of models in order to set this one up. That 8 means that there are many, many inputs into this, and it 9 doesn't fit what Mick said yesterday, which is it is nice in 10 a multi-barrier system if models are decoupled, which means 11 that the model you have is not necessarily sensitive to all 12 the unknowns of the models that went before it.
13 Well, this particular model has inputs from many, 14 many other models, and clearly is not decoupled. This is y_) 15 not an avoidable problem, necessarily, and may have to be 16 something that this program lives in.
17 In all the other programs, because they start with 18 a sealed vault, they don't pay very much attention to what 19 happens before this. They're only interested in what 20 happens to it afterwards. So they start somewhere around 21 the waste package in their modeling issues and go from there 22 on out.
23 Here, you have to start determining what's coming 24 in before you can think about what's going out, so there's 25' an extra degree of complexity here, and this is not a
[]
,/
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NWi Suite 1014 Washington, D.C. 20036 (202) 842-0034
313 1 decouple ' model, . which means ' it 's going to 1x3 very difficult
() 2 3
to'make it robust in terms of the definition that Mick Apted gave you yesterday.
4 Okay. .What I would like to do now is to show you.
5 some of the parameters that go into defining where you are 6 for that model. This is a little presumptuous of me. Many 7 people thought about this much more than I have, but-when I 8 look at'these input parameters, there are many -- and some 9 of which, you'll never actually know.
10 So, for instance, on the waste package, you may be
. 11 ). able to statistically determine tPose juvenile failures, 12 containers which are not properly made.
13 The distribution of failures will come to the 14 model. It's reliability, of course, is another matter. The
() 15 location of the failure is something that you will never 16 really know. It's an intuitive, an intelligent guess, and
. 17 it's probably going to be on the top or on the bottom, but 18 .you'll.never'actually know where it is.
19 The dimension of the failure is something that 20 most~ corrosion scientists -- in fact, I think a very few who 21 ;have actually attempted to define what the dimension of a 22 failure is.
' 23 Is it a single -- a deep pit?
24~ Is some wide crevice?
25 'Is it some wide crack?
j .S y-~I; . ANN.RILEY & ASSOCIATES,'LTD.
"3x2 Court Reporters 1025-Connecticut Avenue, NW, Suite 1014 E s Washingtoni D.C. 20036 (202) 842-0034
_ _ . - _ _ _ . _ _ . _ . _ _ _ _ _ _ - - . . _ _ _ _ . . _ _ _ _ _ _ . _ _ _ m _ _ -
314 L
l 1 'How wide is it?
21 These'are not issues which are predictable within
{)
3- ' corrosion science, and yet, .this is a feature, the pit.or
- 4. patchipenetration'of the waste container model which must 5' feed into the next one.
6 So1those are unknown parameters, and I. don't'think
- 7. there is much chance of actually knowing them ever.
8 Clodding is another issue. The immediate failures 9 are fairly easy. There are defective bundles that go in.
10 We'know the number.
11 Stainless steel clod, you can assume will rapidly.
12 Creep corrosion, I put a question mark. I think 13 if you work hard enough at it, you'll get it.
14' The long-term corrosion failures are not 15 necessarily known, but I think probably.they'could be, so I 16- give them question mark, not red crosses.
17 When you get to the waste form, having gone 18 through those two barriers, you need to specify the water 19 seepage volume. I think this is probably difficult.
20 You can specify the chemistry. I think this can 21 be-calculated. Assured is another matter.
22 Fraction of area exposed is a parameter. This is 23 not something which you'11'ever get a real measure on. We L 24 may-be able to.make intelligent guesses.
~25 The fraction of the surface wetted comes in under b ANN RILEY & ASSOCIATES, LTD.
!VL Court Reporters
. 1025 Connecticut Avenue, NW, Suite'1014 _
Washington, D.C. 20036,
.(202)- 842-0034
l 315 11 .the sameLcategory as does t.he fraction contacting seepage 2 water. .These are difficult parameters -- pre-parameters to 3- predict, in order.to. set up the waste form model.
4- Intrinsic' corrosion' rate, which is what I wil1 5 deal with a-little bit this morning, I think can be known 6 fairly well. And although'we don't presently know what the 7 influence of secondary phase formation is, I think it is a 8 tractable problem.
L The radionuclides source terms, people have 9
10 fractions for instant -- have numbers i for instant release 11 fractions. .You can use-them, if you wish. They are a 12 difficulty to use, if they're not a difficulty to everybody 13 else because they have slower transport.
14 The release from corrosion product solubility.
15' limits retention factors are extremely important issues, but 16 very difficult ones to get a hold of, and I'm generous in 17 giving colloid formation, and I should have added 18 " transport," a question mark. It might have been 19 appropriate to give it a red cross. I don't know quite how 20 you're going to handle that problem.
21 I would suggest that you need to think in terms of 22 what Mick said yesterday which is, if you don't feel you can
'23 -predict something, you've got to look for an engineering way 24 .to stop it.
25 Okay. The things I'd like to deal with are the l
O ' ANN RILEY & ASSOCIATES, LTD.
- b Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 I Washington,.D.C. 20036 l (202) 842-0034- !
____________-_______-____L
t
/ 1 316 I
' l~. instantLrelease fraction. I'll tell.you that we do know, 2-l' what~the mechanism fuel' corrosion is. . Deal with the
.N 43 intrinsic corrosion rate of one or two of the parameters
! 4' .that we feel-are-important, like burn-up radiolysis, ground 5 . water' composition oxygen.
6 The accumulation of corrosion products, which, to 71 me, is-the great unknown, and probably the -- one of the 8 ~potentially'-- aisaving feature for fuel, a little bit on
~9' the formation of colloids.
10 Okay. Here's a chemist's view, an 11 electrochemist's view, because a. chemist might have a better 12 Lview, of the release of radionuclides in the categories that 13 we can separate them into.
14- There.are thole which sit and have separated from
() 115 'the fuel at the clodding' gap. They are a part of the 16 instant ~ release fraction. They're highly soluble and 17 expected to be released as soon as water gets in there.
18 For conservative reasons, anything sitting at the 19 . grain boundaries is also assumed to be instantly released.
20 This is a' conservative assumption you might want to think 2 11 about, though that would you set you at odds with the rest
- 22. 'of the world who don't need,to think about it.
'2 3 . And there are those which are part of the matrix
' 24- of the fuel which are released at a rate dictated by the 25; degradation rate of the fuel, something like.90 to 95 h 'AMN RILEY'& ASSOCIATES, LTD.
1s._/ - . Court. Reporters 1025. Connecticut' Avenue,.NW, Suite 1014 Washington,-D.C. 20036
' 7
, _( 202).842-0034
317
- 1. percent of the radionuclides are in this category. -
Some of
(/v. ') 2 the dangerous ones are in the other categories.
3 You don't have to spend much time thinking about 4 instant release fractions, if you don't wish. I'm not 5 asking you to read this table. I've just reproduced the 6 table from a report by Johnson and Tate, from Atomic Energy 7 of Canada, which' reviews was paid for by the Swedes, but 8 which reviewed instant release fractions for a whole variety 9 of programs.
10 So this is not just Canadian data. This is a
'11 whole series of data. And they specify their best estimate 12 of the percentage, which should be considered as an instant 13 release fraction.
14 So you could use this table, if you wish, and it
() 15 does, indeed, include American data, as well as everybody 16 else's.
17 A quick comment on what the fuel -- what the fuel 18 corrosion process is. I apologize for this smudging at the 19 top, but things have te got to the state where you make your
- 20. own overheads,_you copy your own, and they don't clean the 21 copying machine in between, so it's all a personally done 22 thing, which explains some of the confusion.
23 The fuel mechanism oxidation process is well 24 known. It will oxidize. This is a corrosion potential.
25 We'll deal with the relevance of this to the redox potential l-L[ }
l 's /
ANN RILEY & ASSOCIATES, LTD.
Court Reporters l 1025 Connecticut Avenue,'NW, Suite 1014 L Washington, D.C. 20036 (202) 842-0034 L __ _ - - - - - _ _ . _ - - _ - _ _ - - _ . _ - - _ _
318 1 in a minute.
/N
- 2 The corrosion -- as the corrosion potential goes (x-}
- 3. frvm reducing to oxidizing, you will go through a
- 4. nonstochiometric zone. You will approach the point where 5 dissolution starts somewhere around this corrosion 6 potential. It will be accelerated, if you complex it. The 7- . harder you drive it, you more likely you are to form these 8 secondary phases, which block the dissolution.
9 If you were to measure the corrosion potentials in 10 aerated solutions, you would sit here within this band for a 11 fairly innocuous pH 9 solution, with or without the 12 complexing agent.
13 If you deoxygenate, you sit somewhere down here.
14 If you use -- measure the corrosion. potential where you --
() 15 in a gamma c71 where you'd expect beta gamma doses in this 16 range over a thousand years, that's where the corrosion 17 potential would sit, telling you that you will be oxidizing.
18 For Canadian fuel, if you're to use the alpha dose 19 rates, we expect you will be down here in a nonoxydizing 20 region. So at first hint we don't think alpha radiolysis is 21 important to us. This is putting us at odds with other 22 people that thinh it's very important, and there are a few 23 caveats. Just to make sure that you know what I am talking 24 about, the corrosion potential is the potential established 25 by the coupling of oxygen reduction or oxidant reduction to L[ '
)- ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
~
(202) 842-0034 i _ ___ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ ___
319 1 the fuel dissolution process,uit has no thermodynamic l( f 2- ' meaning.. 2The EH:that most people talk about, which-is the .
3- redox: potential to the environment, is very high in' oxygen,
~4 it's way.over a volt, or about .6 volts,-I.think of pH 9.
S- The. corrosion potential is somewhere in between. So it's a 6 bit of'an~ ambiguous potential, but to corrosion scientists,
^
7 it is the measure.
8- Okay. What would we expect to happen? Well, 9' let'sEstart out with a situation where there's no seepage; 10 . drips. We'll assume that:there's a failure in the waste Lil package and there's a failure in the cladding and all that
- 12- gets.in here is aerated vapor. If this happens early, while 13 the. temperatures are high, and the cladding is ruptured, 14 then the pellet will-break up, strip open the cladding as
- q, 15 you oxidize up to 2.67. You will end up with maximizing the i 7 16' surface area in fragments. The surface will go up.about 150 17 times, according to Walt Gray. But this should only occur 18 'for juvenile failures where the conditions are very hot.
19 A more.likely scenario, assuming that you don't 20- fail to early is that you admit humid vapor into the 21 cladding. You will get a grain boundary oxidation process'.
22 It.will go up somewhere to 2.4, 2.33. You go to 2.4, we go 23 to 2.33,-the difference is the burnup characteristics of the 2'4 fuel. It turns out-not to be important. So will get this
'25 . oxidation. pattern running through the grains. As Il ANN RILEY'& ASSOCIATES, LTD.
A /' Court Reporters 1025~ Connecticut. Avenue,- NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
320 1 dissolution starts, you will get the formation of secondary.
2 phases, which should block off this grain boundary process.
3 Solubility limits should apply in this case and, assuming 4 you don't have convective flow, you should have diffusive 5 flux, not diffuse flux.
6 If you allow a seepage drip, in other words, the 7 fuel actually sees ground water which is being carried in 8 there, not just humid vapor, again, you would come through 9 these two barriers, and you can see a clear advant3ge, if 10 you wish, to the cladding in protecting you against this 11 seepage drip.
12 If the environment is high in carbonate, low in 13 calcium and silicon, then you will get dissolution without 14 necessarily early phase formation to block the dissolution n)
(,, 15 because you will complex the carbonate. If we take this 16 relative dissolution rate to be about one, in order to 17 compare it to this case, you will penetrate about four to 18 ten grains, according to Walt Gray, which will accentuate 19 the surface area by a factor of 15. And this should only 20 occur, again, for a package with a juvenile failure.
21 Assuming that, the carbonate is potentially high 22 in the beginning but eventually low. This was the reason 23 for my question to Dave Sassani yesterday. To me, it is --
l 24 the only features of the ground water chemistry which matter 25 to the fuel are the ratio of calcium silicon to carbonate.
[~ ANN RILEY & ASSOCIATES, LTD.
lN~-
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
! Washington, D.C. 20036-( (202) 842-0034 L J
L.
L 321-1- .This complexes it-.this precipitates it.
h/
^
2' So if you go'down the other path, and in this case
)
3' ~you have carbonate whichJis low,.and here is-a criterion, 4 -theseltwo are:high,. then you it,uld' start to build up the 5' secondary phases and you would perhaps start with uranophane
'6 and'go.through a sequence. The dissolution rate wouldLdrop-
' 7- by a factor of two ordersoof magnitude- according to the ,
t L
'8- -tests of Walt. Gray and John Tate-in Canada. .This secondary 9 ' phase formation should start blocking off. You would not
'10 increase the surface area. Radionuclides solubility limits
- 11' 'should apply because the secondary phase is there, and you 12= should have diffusive convection flow depending on what 13 ' environment you are in.
14 Okay. I am just going to skip one overhead and go
) 15- fonto -- since my time is a little shorter than it was 16 originally. What about the surface area? Okay. Under
-17 humid vapor conditions, no seepage drips, if you just look 18 at one pellet within the failed cladding, then the 19 ; amplifying factors for~the geometric surface area are the
- 20 surface roughness factor. This is not a smooth solid. We
' 21- think'this is about 3 and a maximum of 5, based on 22' electro-chemical. measures. Times the cracking factor, in 23- other words, the normal geometric dimensions of the pellet
. 24: and: surface area is increased by a factor of 10 to 20, 25 ~ calculated by! Lawrence Johnson,.due to the degree of f
ANN RILEY & ASSOCIATES, LTD.
): . Court Reporters
' ~. -
1025 Connecticut: Avenue, NW,LSuite 1014 Washington, :: D. C. 2 003 6
( 202)-842-0034'
322 1 cracking. And, of course, that is dependent a little bit on
(~ \ 2 burnup. So we can multiply the geometric area to get the
\~sl 3 surface area by a roughness factor and a cracking factor.
4 This has got -- you may precede this by something 5 that says cladding will protect 90 percent of your surface 6 from wetting. That's another factor altogether. This is 7 just assuming that the fuel is wetted.
8 Under aqueous corrosion conditions, you would 9 include the factor of 15 for the grain boundary factor which 10 Walt Gray measured. So there are at least rational grounds 11 to calculate the exposed surface area of the fuel, assuming, 12 of course, that it is all wetted.
13 What should the influence of corrosion product 14 . deposits be? We are in the process of trying to put
.( %
( ,) 15 together a simple model to describe this, and we describe it 16 simply in terms of, right now, the attenuation of the 17 available surface area by the porosity of the deposit that 18 goes down on the surface. So you would -- it turns out that 19 the available surface area is directly related to the 20 porosity.
21 Now, I have drawn this as a simple linear 22 porosity. Of course, it isn't, it's porosity attenuated by 23 a tortuosity factor. What it effectively does is decrease 24 the diffusion coefficients through this layer. So you get l 25 an effective diffusion coefficient which is lower than the l ) ANN RILEY & ASSOCIATES, LTD.
N_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
l L .. _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
323 1 value that you would get in free solution.
[)
v 2 I chose not to talk about cladding failures. You
. 3 are welcome to discuss it if you wish. It would dominate l
l 4 the whole discussion if we did.
1 5 Okay. A little bit of chemistry, and I apologize 6- to those people that don't necessarily like a lot of I
7 chemistry, but if any of you ever read John-Paul Sartre, he 8 is of the philosophy that if you wish to get the maximum 9 experience out of every moment of your life, then suffering 10 is far better for you than happiness because time passes too 11 quick when you are happy. So on that basis, here's a bit of 12 chemistry.
13 (Laughter.)
14 DR. SHOESMITH: The other maxim is there's always in
( ,) 15 an infinite supply of bullshit wherever you are. The 16 reduction of oxygen or 02, is a catalyzed process. Now this 17 may not appear important but I hope to show you that it is, 18 because what it does is introduce a huge complexity, which 19 you don't want to think about but which if you do enough 20 , experiments you can actually eliminate, and then the person 21 .who gave you the money to do it says why the hell did you 22 waste your time doing that?
23 It is coupled'to what are known as donor accepted 24 states, different oxidation states within the oxide, and 25 this is the same on magnetite or anything else -- IN2, IN3,
^
/ ANN RILEY & ASSOCMTES, LTD.
m.- / Court Reporters ;
1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 l (202) 842-0034 I
324 1 Uranium-4, Uranium-5 or 46, if you with. The electron is
.(x_/
m) 2 effectively relayed -- it doesn't matter about the chemical 3 details -- between these~two, so that they change oxidation 4 state and give'off the electron to the oxygen -- to couple 5 to the surface which allows you to destroy the oxygen-oxygen 6 bond which makes oxygen so stable.
7 So the oxygen reduction and its ability to be the 8 cathodic reaction is directly tied to the structure of the 9 fuel and its degree of oxidation. That is the basic 10 message.
11 But what happens when you put this stuff in a 12 repository or in reactor? Well, when it oxidizes of course 13 the oxygen effectively ejects 02 minus, converts 4 to 5.
14 You increase the number density so I see degree of oxidation n
(,) 15 goes up, the catalytic ability of the surface to support the 16 cathodic reaction goes up.
17 In reactor you form rare earths by the fission 18 process. This is the equivalent of replacing Uranium-4 plus 19 by rare earth 3 plus. The structure will accommodate to 20 maintain neutrality by putting 4 aside. Again you push up 21 the catalytic properties of the surface to reduce oxygen.
22 Then finally by forming this noble metal into 23 metallic, the epsilon phase, you introduce the perfect 24 dispersed cathode, because now you have put in a catalytic 25 noble metal particle which is very good for catalyzing
[ ANN RILEY & ASSOCIATES, LTD.
's.- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 e _ _ _ _ - _ _ _ _ _ - _ _ _ _ _ - _
325 1 oxygen-reduction, which should theoretically drive the 2 anodic dissolution of the rest of the surface, so you
( .
3 anticipate the spent fuel will be way worse than UO2 based 4 on this kind of discovery.
5 These were done on experiments on synfuel where we 6 could simulate these kind of structures.
7 So does that in fact kill you? The answer is no, 8 it doesn't under oxidizing conditions because as you go from 9 reducing conditions here to oxidizing conditions here, you 10 may increase the catalytic ability of this surface but you Ell actually passivate these intermetallic particles, so the 12 perfect cathodes that you put in won't actually function 13 under your conditions. They will under our conditions where 14 we are not so oxidizing.
() 15 Now in a way this is kind of a red herring, but it 16 is a scientific red herring which could have been missed.
17 It may not have worked. This may have been a catalytic 18 . particle. You may have introduced the possibility of 19 totally destroying the fuel base on the introduction of 20 these small cathodes because they would sit directly at the 21 grain boundaries, which means they would dissolve the grain 22 boundaries. The fuel would have fallen apart very rapidly.
23 It doesn't look like they are important.
24 Another feature of the formation of an insulating 25 corrosion product is that it will block these catalytic ANN RILEY &-ASSOCIATES, LTD.
LDs-
[/') Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L i
^
l _
1 sites'and.this-may'be.the most important, feature of'the
) 2 corrosion product.
3 Ifithe oxygen requires:that it find these donor L 4- accepted states, as a physicist will call'them, in the l
5 . surface, and you put a deposit on here, it is not going to
-6 find them or it is only going to find them at a 7 transport-controlled rate. depending on the properties of 8 :this film, so the corrosion process could be cut off under 9~ oxidizing conditions simply because you block one-half of 110 the corrosion process and therefore the other half can't 11 take place.
12 Now that is speculation but we have some evidence 13; to back up that that will be the case.
14 So there's a whole pile of complex chemistry which I) 15 actually boils down to saying we are on good grounds and 16 most of it we can ignore in a model.
17 The influence of groundwater composition -- again 18 ' don't spend your time trying'to follow everything in this 19' plot. Just notice two things, that in carbonate solution 20 'the rate is up here at 100 but if you introduce calcium and 1
21 silicon, even in the presence of carbonate, you drop the 22 rate by.two orders of magnitude and if you continue to leave 23 this in, leave the calcium and silicon in and you push the 24 _ temperature up or'you decrease it, the rate continuen to go 125 down. I would suggest that is a clear indication that here t ANN RILEY & ASSOCIATES,--LTD.
"A 2 - Court Reporters.
1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034'
327 1 you'had clean intrinsic dissolution. Here you have
() 2 3
dissolution where you are collecting or blocking secondary product.
4 So there is good evidence that if calcium and 5 silicon dominate, that in fact you will get a much lower 6 corrosion rate.
7 Then I took a table from Dave Sassani. This is 8 presented at one of the expert elicitation that we went 9 through and again don't spend a lot of time, but broke up 10 the period into a series of -- he broke up the timescale 11 into a series of periods where he calculated the 12 compositions of the water or he hinted yesterday that he has 13 changed his mind so I am not sure what the new values are 14 but what I took from this is that somewhere out here around rh
( ,) 15 10,000 years this table is saying that the calcium 16 concentration would be up around 10 to the minus 2 molar.
17 The carbonate concentrate will be down around 10 to minus 4 18 molar. If this is the case, then this is the scenario, not 19 this. But again, those numbers would need to be firmed up.
20 A couple more issues. The effect of carbonate --
21 there is a sense around that you can do a whole series of 22 experiments, vary'the carbonate concentration over six 23 orders of magnitude and then just fit the data. I don't 24 .like that approach for the following reason. If the 25 carbonate concentration is very low, then you block the 1
. \_[' ') ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
328 1 dissolution and we can see this in our. experiments, this h
v 2 precipitate will, block it.
3 If you push up the concentration just a little 4 bit, this will be 10 to the minus 4 molar or less, somewhere 5 around between 10 to the minus 4 and 10 to the minus 3 you 6 will take this deposit off. There will be a huge increase 7 of two orders of magnitude in the dissolution rate because 8 of that simple taking off of this corrosion product, and 9 again we can see it electrochemically.
10 If you are interested, we did the assignment of 11 mechanism based on an impedance experiment but I am not 12 going to discuss that unless anybody wants to know the 13 details. That explains that peculiar little figure to the 14 right.
) 15 If you get into an intermediate concentration, 16 somewhere between 10 to the minus 3 and 10 to the minus 1, 17 you see a surface absorbed species which is now instead of 18 just removing that deposit it is kinetically involved in the 19 dissolution process. You can see it in the impedance as a 20 second time response of stored charge, so it is there.
21 There is no doubting it is there.
22 If you get to very high carbonate concentration, 23 then you start to put down a uranyl carbonate surface phase.
.24 The rate is still higher -- these circles are smaller -- and 25 you can see this deposit in terms of a second response, but
.m i ). ANN RILEY & ASSOCIATES, LTD.
' s' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 7 (202) 842-0034
329
.1 the behavior is not simple. There's not too many linear
(~)
(_
2 -regions of behavior. It's not that simple to specify a 3 reaction order, if you wish. You can do it for a limited 4 concentration range.
5 'The effect of temperature is a little strange 6 .because it is different depending on what solution you 7 measure. If you measure the temperature dependence when 8 there is no complexing agents present, this will be the 9 equivalent of 7 to 8 kilocalories per mole,uso this is 10 fairly low if you prefer the other unit. You get almost no 11 temperature dependence, and that is because you have the 12 counter-balancing effect of accelerating the dissolution but 13 also accelerating its blockage by increasing the film 14 thickness.
/~
( ,S) 15 If you do it in carbonate or acid you get the 16 activation energies that you would anticipate, which are 17 much higher, and that is because you don't have the 18 formation of this blocking deposit, so the evidence for its 19 formation is all over the place.
20 Okay. I'm going to leave out the discussion of 21 the beta gamma, its short-term effect, and just quickly 22 mention the alpha radiolysis.
23 DR. WYMER: -You have 25 minutes; you're not 24 pressed.
25 DR. SHOESMITH: Oh, okay. But I think it's more a
- (,-)
's>
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
. e -
I 330 ' ;
1 1 fine point of detail than an essential issue.
() 2 3
Okay. Here are our measurements of the corrosion rate of fuel as a function of alpha source strength. The 4 line doesn't mean anything; all it did was encompass most of 5 the points and tells us that these two, perhaps, are things 6 that we should take a second look at.
7 So the line doesn't mean anything. What's 8 important is the rate goes down quite steeply as the source 9 strength goes down. For fuel, the source strength at ten 10 years will be somewhere around here, and after a thousand 11 years, will be somewhere around here.
12 If we were to do our experiment in aerated 13 solution, we would have got a value here in the absence of 14 an alpha source. If we were to do our experiment without an
() 15 alpha source but with no oxygen present, we would have got a 16 rate down here.
17 So we can see that the alpha source strength 18 effect is there, providing the source strength is extremely 19 large, but disappears quite quickly and should be 20 insignificant at these particular source strengths, and that l
l 21 if your systam is aerated, the rate you will get by reaction 22 with oxygen is at least as great as the rate you will get 23 from alpha radiolysis, therefore, why bother about the alpha 24 radiolysis? Your system is already sufficiently oxidizing.
25 It is necessary to add a caveat to that. Those O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
331 1 measurements were made in a thin layer cell where the UO2
() 2 3
electrode is brought up close to a goldplated alpha source.
And this is about a 25 micron gap, which is about the 4 distance of effectiveness for the alpha particle.
5 You produce the oxidants in here, but there are 6 significant diffusive losses outside. So this empirical 7 correlation between corrosion rate and source strength may 8 not be the same as a good correlation between dissolution 9 rate and radiolysis product concentration. But we are 10 attempting to do that by calculation, and we think we're on 11 fairly reasonable ground.
12 Okay. So let me try and answer the question is 13 there a burn-up effect. I got a report from Walt Gray in my 14 mail a week ago telling me that his experiments say there is
() 15 no burn-up effect, and I certainly agree with that for the 16 following reasons.
17 There is a burn-up effect in the sense that it 18 will dictate the inventory of the instant release fraction, 19 and that it will change the surface area by changing the 20 degree of cracking. So in those physical senses, there is 21 an effect.
l 22 The radiolysis effect will be small. I didn't 23 bother to deal with the beta gamma, but you can see the 24 decay markedly somewhere around 3 to 400 years, providing 25 you keep containment for that period of time, they're O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW ., Suite .'014 i
l l
-Washington, D.C. 20036 (202) 842-0034 l
l
1 I l
332 l l 1 insignificant. We just have lots of data to show that
[\ 2- they're insignificant, and the data is of some value if you w]
! 3 wish to deal with the accident scenario of instant failure.
4 The alpha, of course, looks like it's a little 5 more dangerous, but our evidence says it's not that l
6 important. I would like to add one more caveat to that in a 7 few minutes.
8 So-it doesn't look like, other than a little bit 9 of alpha radiolysis, the radiolysis effect'from burn-up is 10 significant, and if you attempt, as Walt did, to measure the 11 dissolution rate of UO2 spent fuel, oxidized UO2, highly 12 oxidized UO2, the only difference that you can detect, 13 significant difference, is the surface area effect due to 14 break-up. All these, within reason, are within the same
() 15 rate. So there is no reason to think of the burn-up effect.
16 There is no intrinsic burn-up effect, apparently, with the 17 following caveat. This is more of a concern to the 18 Europeans, who would like very much to take credit for the 19 production of hydrogen blocking off any radiolytic 20 dissolution of the fuel. I think that's a little premature.
21 There is a problem if you have two mixed 22 radiolysis sources, beta and alpha, because the alpha will 23 produce a molecular species in high concentration, 24 especially if it's confined in this gap. So this could be 25 one side of a piece of fuel and this could be the other side b\/ ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, .D.C. 20036 (202) 842-0034
lI
'333 1 within'a crack, for instance.
ji '2 You~will produce a really'relatively high 13: concentration-of molecular oxidants. If you'also have a-4 >short-range high-energy form of radiation'like beta,.;then, 5 in fact, you could reconvert-this back-into: radicals,.which
, l6 are:way more reactive than.the molecular oxidants. 1So the
- 7. mix: of the ' beta-alpha is~ dangerous. I think this shows up 8 in John Bates' fuel experiments where he' sees aggressive 9: boundary-attack.
- 10 - This is not necessarily something that should last
- 11' v'ery long, butwif, as the~ Europeans-wish, they would like to 12' take credit for producing hydrogen this way, then this is a' 13 critical. issue.
' 14 : 'How do we know that? Well, we know it from the
) 15 following experiment. This is just a simple corrosion
' 16 . potential measurement as a function of time done in added
=17 hydrogen peroxide. So we simulated the alpha, it's ten to 18 the minus six from all the hydrogen peroxide, the corrosion
.19 potential will rise, and if we have not added the gamma at 20 this' point, which stay. constant -- and that would be the 21 . molecular effect, the hydrogen peroxide effect.
L22 If you now switch on the gamma source, you can see
,s 23:- the additional oxidizing' capacity by having!the gamma plus E24 ' the. alpha-together.
4 25 .
LSa'I don't think it's a. major issue for this h_ b ANN RILEY'&~ ASSOCIATES, LTD.
Af ~ Court Reporters-11025 Connecticut Avenue,' NW,-Suite 1014 mWashington,'D.C. 20036 t
'(202)-842-0034
334 l
1 program, but it's a' big issue for those.that would like very l 3p}
L 2 much to say that radiolysis works to your favor, works in L 3 your favor.
4 Okay. What kind of model could you use? Well, I 5 think the model that's presently used is a parametric fit to 6 a series of experiments. It turns out that when you look at 7 this one and the one down here, we're pretty close together.
8 So I'm not suggesting that this should not be used; I just 9 have reservations about this approach to things.
10 They have a burn-up term which I think don't think 11 is significant, should be taken out. I'm not quite sure how 12 you take something out of a parametric fit, but --
13 [ Laughter.]
14 DR. SHOESMITH: Whatever.
f\
() 15 These are the constants and they have a pH term 16 which is almost insignificant, which agrees with us, an 17 oxygen term, a carbonate term, a temperature term.
18 The way that I would have done it personally is I 19 would have worked out the dissolution rate based on an 20 intrinsic dissolution rate, surface area factor, an 21 activation energy term, if you wish, the ground water, 22 frankly, that comes from the composition between carbonate 23 and two concentration' dependencies.
24 But for your particular waste vault, you can say,
'25 well, if I know the carbonate, we can lump it into the
[ ') '
'\'
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite.1014 Washington, D.C. 20036 (202) 842-0034
335 1 constant and.we can assume that the oxygen concentration is
-2 the'same;.therefore, you can take an intrinsic dissolution 3 rate, which is just.affunction of.the area' exposed, the 4 temperature.and the ground water composition.
5 So in the end, all that chemistry that I just
-6 described becomes relatively irrelevant; all~you n'eed to 7 know is area exposed, temperature, ground water composition, 8 the exposure environment.
9' 'That's a perfect example of how you got from 10 something simple, to total confusion, to something which I 11 think is understandable.
12 Okay. A questian you might ask yourself is, what 13 rate of intrinsic dissolution -- what intrinsic dissolution 14 rate should be used?. Information that I got from Bill, Bill C\ 15 Halsey, tells me that the rate they use is somewhere around
( ,/
16 this value. I think this is the average from single pass 17 flow-through experiments, because the average from this 18 table is about 8.6, it's very close to that one. This works 19 out at a fractional dissolution rate per day of something 20 like 9 times 10 to the minus six.
21 Bill Murphy and English Percy estimated from 22 natural analogs that the fractional inventory per day,.which
- 23 is?the same as the fractional dissolution inventory per. day,.
l 2
.4 would be about ten to the minus seven baced on a l.
25 conservative estimate.from natural analogs. So they are two l T f;) ANN RILEY & ASSOCIATES, LTD.
' 5s/ . Court Reporters 1025 Connecticut Avenue, NW ,' Suite 1014
. Washington,LD.C. 20036 (202)' 842-0034
336 1 orders of magnitude apart.
(v; 2 I think it's an intrinsic need when you're dealing 3 with fuel to make sure that what we do to say that this will 4 be the behavior in our repository should match not just 5 qualitatively,.but as. quantitative as you can with what the 6 natural record tells you.
7 So there are two orders of magnitude discrepancy 8 here, and these guys will tell you that they think this is 9 conservative and, in fact, may have, I think based on 10 discussion we had yesterday, reduced that estimate even
.11 further.
12 The issue, then, becomes the tctal fuel alteration 13 time. If you take the -- the top value is 300 to 1000 141 years, which says the waste form is irrelevant. As soon as l'h 15 it gets wet, it dissolves.
( ,/ If you take the total alteration 16 time, it says maybe there's some credit for the waste form.
17 I would suggest it's worth working very hard to find some 18 credit for the waste form so that you're not totally )
1 19 dependent upon credit from the cladding. ;
20 Okay. Let's move on -- I think I've still got 21 time -- and deal a little bit with some of the issues of 22 forming these deposits.
23 This one we've dealt with. You will attenuate the 24: surface area just because of the porosity and tortuosity of ,
25 the surface layer. Determining what those are may be very, 1 i
'(
i ANN RILEY & ASSOCIATES, LTD.
k/ Court Reporters 1025 Connecticut Avenue,'NW, Suite 1014 Washington, D.C. 20036 L (202) 842-0034 I -
i L '
337
'l very difficult.
- [ ~y) 2 The oxygen will be transport controlled and it
'Q/
3 . will be suffering from a lack of sites at which it can
.4 react. You may have some problem'with confining alpha 5 radiolysis to the surface, and there's a great unknown --
6 well, not unknown; there's lots of data here,'but 7 'potentially the ability to trap many of these radionuclides 8 in these secondary phases'could give a huge advantage.
9 Okay. I'm just going to skip one overhead'and go 10 ..on quickly.to the alpha.radiolysis.
l 11 - Why do I think the alpha radiolysis -- even though 12 I've tried my best to throw it out and say this is 13 unimportant, why do I keep coming back to it? Well, because 14 it's a short-range radiolysis effect. So you could end up (I 15 with a fuel covered with a corrosion product with a certain 16 effective diffusivity based on tortuosity and porosity. You 17 could incorporate the actinides into the corrosion product, 18 which is what you would like to do, but somewhere deep in 19 these pores, you'have a short-range radiolytic deposition of 20 energy which could give you an' oxidizing redux front 21 .directly at.the fuel surface, even for a relatively low dose 22 rate, relatively low source strength of alpha radiolysis.
'23 So while you may be limited in getting the oxygen
- 24. there, it's possible that if the alpha dose rate stays up 25 sufficiently, you will get a radiolytic production of
); ANN RILEY & ASSOCIATES, LTD.
\/ ,
Court Reporters
,1025 Connecticut Avenue, lui, Suite 1014 Washington,.D.C. 20036 (202) 842-0034
L 338
! 1 oxygens right at the center phase, you will get a moving
<~%
( 2 redux froi.t. This deposit will be irrelevant.
i x -)
3 So there is a residual fear for the alpha 4 radiolysis which I don't think anybody has cleared up.
5 We're trying. I'll just show you quickly how we're trying.
6 We're doing some calculations based on a thin 7 layer cell, where we calculate the production of molecular 8 oxygen as a function of time for various source strengths.
9- These are fairly high source strengths. And you can see you 10- can get the high values. These are ten to the minus one 11 molar for these really high source strengths.
12 Even for low source strengths, if you allow no 13' diffusive losses -- that's the important feature in this 14 calculation -- nothing ever moves out of the reaction layer.
15 The concentration is all confined.
16 Even for these relatively small alpha dose rates, 17 the concentration is getting up toward ten to the minus two 18 molar. This is aggressive hydrogen peroxide. Down this 19 right-hand side, I have put a quick hint of how you might 20 get rid of the full radiolysis model and do a simple 21 calculation which tells you the steady state concentration 22 that you should get is directly related to source strength.
23 In the bottom half, again I don't want you to pay 24 a lot of attention to the bulk of the curves, but what we're 25 trying to calculate is what the effect of confinement of the ANN RILEY & ASSOCIATES, LTD.
^~/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 .
Washington, D.C. 20036 (202) 842-0034
339 1 radiolysis product would be in a deep crevice -- in other
() 2 3
words, which we have' simulated here very simply by having.a-
. diffusion coefficient somewhere around ten to the minus 4 five, ten'to the minus six -- aid a shallow crevice. You 5 can see that it makes a little difference.
6 Diffusion from the shallow site is much more 7 rapid. The hydrogen peroxide concentrations are lower. The 8 deeper the site, the less effect the diffusive -- decreasing 9 the diffusive loss has on pulling down the concentration.
10 So as you go into a deep wet site, the potential for a 11' radiolysis effect becomes higher.
12 Okay. Let's take a look at the fate of released 13 radionuclides. I think you can summarize them-all 14 effectively according to one or other of the following
() 15 steps. There are those which are unretarded, things like 16 iodine. There's very little to stop it moving once it gets 17 into the water. That's the same for us as it is for you.
18 These are anines. They don't care about the cathine 19 exchange capacities at play; it just goes zipping right 20 through. Also, they're part of the instant release 21 fraction. There are others.
22 There are things which could be retarded.
23 Technetium, for instance, in an oxidized state is also an 12 4 : anine. But if you had reduced ion species present, it could
[ 25 be retarded by precipitation. Less likely in yourl scenario l .
ff.k ; ANN RILEY & ASSOCIATES, LTD.
Court Reporters-1025 Connecticut-Avenue, NW,' Suite 1014 Washington, D.C. 20036~
p (202) 842-0034
, , = - _ _ _ _ _ _ - _ _
1 df 340 1 than it is in'ours'because ours does go reducing.
() 2 3
Under those which we' feel will be' retarded by limited solubility or incorporation into-the' secondary
~
4 phases and.then re-released at a rate which is dictated'by 5 effectively.the solubility or the kinetics.of dissolution, 6- whichever you want for those second. deposits.
- 7. However, there is a fear that some, instead of 8 going through this incorporation route, will escape by 9 attachment or formation of colloids.
'10 DR. APTED: -Do you have some.new isotopes?
Ell DR. SHOESMITH: Is there a new one there?
12- DR. APTED: There are two of them.
13 DR. SHOESMITH: Are there? Oh, yeah. Well, I did 14 make these myself. Yes, things move slow up there. Those
() 15 neutrons just ain't as fast as they are down here.
16 In fact, there is a rumor nuw going in Canada that 17 we're relying on the next Ice Age, not for this nuclear 18 waste repository, but because we think it's the only way 19 that we'll ever become dominant in the ice hockey world
-20 again.
21 [ Laughter.)
L 22 DR. WYMER: It's about another ten minutes, Dave.
23 DR. SHOESMITH: Okay. That should do it. That
=24 should-do it.
25 Again, this is an overhead I took from Bob p-.
f-l .
ANN RILEY & ASSCCIATES, LTD.
.s_ J Court Reporters 1025 Connecticut Avenne, NW, _ Suite- 1014 L . Washington, D.C. 20036 f, (202) 842-0034
341
-1 Andrews. 'It looks fairly complex. The reason I took.this.
'2- one -- I think this is for expected dose rates out at_the i-- ( )
3 - periphery of the waste vault around 100,000 years. The.
4 ' reason I took this one is I think it encapsulates a whole
'5 series of. problems and issues. The-dark line, which is
~
6 untouched'by myself, in color, is the total ~ dose rate. .The 7- - red and the green are those species which are released 8' 'quickly and not likely to be retarded. So that's the 9 Iodine-129, Technetium-99. They dominate fairly early, and
- 10. I'm told that.this noise that you see is effectively'the
'11 waste package failure rate -- failure features,'which is 12 telling.you that there's almost no retardation of those 13 . processes, and that things like juvenile failures and early 14 failure of waste packages is very critical to the control of
) 15 those radionuclides, and eventually, you start seeing things 16 like neptunium take over, and things like uranium, plutonium 17 are never really the major feature in dictating.
18 But they all have different transport or 19- retardation mechanisms, and you can see the advantages of 20 ' retaining say the neptunium and the plutonium compared to 21 the non-retainment of technetium and the iodine.
22 Okay. I'll. skip an overhead and go to soluble
-23 radionuclides.
24 Okay. What do we do about the soluble 25 ~ radionuclides? .Well, nobody else in the world does anything
) ' ANN RILEY & ASSOCIATES,'LTD.
Ne . Court Reporters ..
1025 ' Connecticut - Avenue, - NW, Suite-1014 Washington, D.C. 20036' (202) 842-0034-k 4
____m__.m.._ ._ _ _ U i
l l 342 l
1 about them. They just admit that they're going to be
' v[) 2 release and transported very quickly.
3 You could attempt to relax the instant release l 4 assumption for grain boundaries. You know, if you could 5 make the argument that the formation of secondary phases 6 leads to transport control, then the sensitivity of the 7 grain boundary goes down, you could make that kind of 8 argument, you could use Walt Gray's argument that you only 9 go.in four or five or six grain boundaries of depths. You 10 don't actually release the grain bo'indaries inventory 11 instantly. But this is a hard task because you'll be going 12 against a trend in the rest of the world.
13 And it's likely to be you would only get this kind 14 of information from long-term fuel leaching studies. You'd
,a
( ,) 15 have to demonstrate that there is a retarded' release from 16 the grain boundary. It's turning up in those kind of fuel 17 leaching experiments, but it's not an easy argument to make, 18 and they're difficult and long experiments to get the 19 information from.
20 You could investigate techniques of interaction 21 redox active minerals, such as the ion corrosion products.
22 Available experimental evidence is sparse, and I don't know 23 of too many people that have done a lot in that area, and 24 you should address this problem of the rapid waste form 25 alteration rate.
O
\._s' ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
-343 1 How is.this really a grossly' conservative number, k) %/
2 -because_if you reduce'that, you will reduce a number of 3 other: things, I think, though the models.may confound my 4 ' opinion.
5 Colloidal' transport. This-is all I'm going to.say 6- about colloids, for the simple reason that I'm as confused
- 7. :as everybody.else about colloids. I know they form. I know 8 they're properly -- they're poorly' characterized. I would 9- suggest that you get a hold of Mick Apted's suggestion.
- 10 - Look for=some way to block it.
11 I don't know whether there is such a thing as a
.1:2 ' colloidal filter. I did read one paper by the name --
13 somebody.in.this program, whose name I can't remember --
14 -that suggested that there are such things. They should be s
[q ,) ;15 looked at. Because I don't see a rational-chemical way of 16 modeling this or demonstrating that you can control them.
17 One question I'd like to ask is, is a colloid 18 formation an early high-reaction rate scenario?
.19 The-fear of colloids is coming from John Bates' 20 experiments where you dripping fuel and you're flowing over 21' .the: edge, and you've got fairly rapid reactivity on the 22 spent. fuel, UO2, in that case. And he sees colloids, 23 Well, I would say that doesn't surprise me 24< because, you add a-limit to the amount of water. It's-125: oxidizing. There'slprobably radiolysis.. You react very ANN RILEY & ASSOCIATES, LTD.
s- -
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036~
(202)1842-0034 i'
L-=-_-_-__-_-__-_______. - - _ _ .]
344 1 quickly. You'll form some colloids, and once they're r~x 2 formed, you're not necessarily going to get rid of them.
(v) 3 But it's not going to be that way once the-4 secondary product is sitting on there. You're going to be 5 drifting, if you wish, on the secondary product.
6 The release of the radionuclides is going to be at 7 .the deep site somewhere within that deposit. It's not 8 necessarily going to see those fast processes.
9 Would you get colloid formation under those
- 10 circumstances? My intuition says, "no," but intuition, of
-11 course, is not the basis for performance assessment. So I'm 12 suggesting that their probability of formation will decrease 13 with time as you form the alteration products.
14 Okay. I'm just going to get this in. We're going
'g 15 to go to that.
16 Retention of radionuclides by co-precipitation.
17 There's lots of evidence for this from Europe, for some of 18 these species.
19 Retention factors from John Bates' experiments 20 drew these conclusions. Many things incorporate into the 21 secondary phases.
22 The issue is, can we put a number on it? These ;
23 retention factors will be very good. If you could take a 24 retention factor of 3,333 for plutonium, you're going to say 25 it's going to stick like glue to your fuel forever. Great.
1 h _ ,-.s
/ ') ANN RILEY & ASSOCIATES, LTD.
\/ Court Reporters 1025. Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
L ,
L' z__. .___ _ ___ _ _ _ __ U
345
'11 Butican you quality. assure that number?
-2 The. issue-is, it's not very. simple to do that.
-3 ' Again,: I stole an overhead I think f rom. Bob Andrews, so it.
'4 . encapsulates.a whole set of works, of data.
5- The issue is, if you attempt to measure the 6' hold-up of, say,. neptunium, according to.the phases that it 7 might form.-- the pure phases it'might form, then you.might 8- get solubilities up here. 'You might argue-that they're not 9 real solubilities because this is an oversaturated 10 -situation. You never actually got to thermodynamic
~11- : stability. That's quite correct.
12 If.you take the retention numbers from some of~the 13 drip tests, and.I think you get values which are down here, 14: Land the program is guessing and taking a range somewhere in 15 the middle.
16: I would do exactly the same if I was in this 17 particular. predicament, but.I feel very leery about doing.
18- this. I would work very hard-to try and demonstrate that 19' these. retention factors are, in fact, real. I think they-20; ~ probably are. I just don't see it as a short-term program
.21~ to get hold of assurance that those numbers are good.
~
22 So I will be presumptuous enough to suggest how 23 - you might go about that. ~You notice I've invented a new-24 :
phase.here, as well. These are=very self-serving, phases.
~25: ~ There,should,.of" course,-be a'"C" in.there. Schoepite, not
,[A})
L
, .s.
ANN RILEY:&~ ASSOCIATES, LTD.
Court Reporters 1025: Connecticut. Avenue,-NW, Suite 1014- I
[.
Washington,*D.C. 20036
-(202) 842-0034' '
I W' -
_---,-.-..b.
_. -_ _ _N
346 1 Shoepite.
(m)
LJ 2 1 Continue the drip tests. Try -- why not try a 3 drip test where you spike it with neptunium. 'What you're 4 looking for is a hold up of this species in the. phases that 5 form, and yet you're relying on only a small concentration 6 coming out the fuel.
7 Try spiking it and see if, in fact, you can 8 demonstrate that large amounts get incorporated into the 9 fuel. Bridge the gap between spent fuel and natural onlogs.
10 There's some rather nice review work by Burns and Ewing and 11 others in the Journal of Nuclear Materials.
12 With respect to Roger sitting here, I would 13 suggest, elegant as this is, in demonstrating that many of 14 these secondary phases will, in fact, incorporate most of
( I 15 these actinide, lactinide species, that they're perfect 16 layered structures, and new aligned sites where they should 17 all fit rather nicely.
18 This is the easy bit. The hard bit is actually 19 getting the experimental data to demonstrate that their 20 chemical arguments are -- structural arguments are correct.
21 And I noticed a few years ago, this piece of 22 information of the oxidizing -- the secondary phase L 23 formation series of products follows what nature does. I L
F 24' thought it was a superbly interesting piece of evidence.
l 25 I've been. hanging onto that for the last five or six years,
,m .
h-( l ANN RILEY & ASSOCIATES, LTD.
's M
. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 L (202); 842-0034 L
L .
347 1E ' hoping that-it would.go somewhere. I have yet to see that
() -2:
3' information truly utilized, and it's certainly not part_of
.the'PA, as far as I know.
4 I recognize how' difficult it is to get in there, 5 but I seriously push this attempt to bridge the gap between
~6 what we're trying to do and what nature tells us.
7- Characterization of radionuclides products is 8 l clearly - radionuclides retaining corrosion products is 9 clearly important.
10 -_ I'd just add the note at the bottom that the Lil carbonate is always a dangerous' factor because not caly does
~ 12 it accelerate the UO2 dissolution, it has a nasty habit of 13 complexing a lot of these other neptunium, plutonium-type
.14 species'which'we would-like very much to retain in the
' 15 environment.
16 CHAIRMAN GARRICK: Just a couple of minutes.
17 DR. SHOESMITH: Yes.
'18 Summary.of key requirements I'm sure that you'll 19 notice are very similar to what you heard, I think from Joel 20- Payer and others, perhaps earlier in this meeting. But we
-21 :need'to have a knowledge of seepage water chemistry and mode 22 of seepage' water ingress. Either that, or you've got to do 23 something about them in an engineering sense because these I ~ 24 are teo factors which continue to dominate, not only the 125 ' waste package, b'ut I'would suggest, . waste form.
// JANN'RILEY & ASSOCIATES, LTD.
l \ s' ~ .
_ Court Reporters-
- 1025 Connecticut 1 Avenue, NW,' Suite 1014 Washington, D.C. 20036-p' (202) 842-0034
___._____m.f ___ i .__..m'E..-.._
348 1 Function of. corrosion products deposits. Both in 2' terms of its ability to block the fuel-corrosion or block 3 the source. term, or its ability to incorporate the 4 ~ radionuclides and control of source term would be something-
-5 that should be done.
6- I still think the alpha radiolysis effect is a red l
~
l 7 ' herring, but it's not something that should be dismissed j
-l
.8 lightly unless experiments and calculation can show that l 1
'9 potential redox interface effect is unimportant. j
'10 I would say block it, if you can, because I don't 11 know what else to do about it. Reinvestigate that rapid j 12 corrosion alteration rate. And perhaps Bill will confound
'13 me by standing up and telling me he already has. I think j 14 it's grossly conservative in the present assessment, and
() 15 therefore, adding problems to other things.
16 And I'm just going to leave cladding credit with a l
17 question mark.
18 CHAIRMAN GARRICK: Thank you. Speaking as a 19 chemist, I want to thank you for the very interesting talk.
20 I'm just sort of waiting for the chemistry to emerge loud 21 ~ and strong.
-22 Were there any' questions around the table on this 23- presentation?
24' Are there any chemists around the table?
- 25. We have a chemist over here.
ANN RILEY &' ASSOCIATES, LTD.
\- Court. Reporters 1025 Connecticut Avenue, NW, Suite,1014 Washington, D.C. 20036 (202) 842-0034
349 1 DR. WYMER: I suppose I feel obligated to say b}- Q_ .
2' .
3 something.
-Your' focus has been on carbonate as the complexing 4 agent. There are a number of other complexing agents in .
'S that water.
1 6 Do you consider their concentration to be below 7 something of interest or why have they been ignored?
8 DR. SHOESMITH: Okay.
9 DR. WYMER: Fluoride, for example, is --
10 DR. SHOESMITH: Yeah. Well, the reason I
!11 concentrated on carbonate, it's the most likely. Plus, it 12 .was something.that we studied within the Canadian program.
13- So it's not like we have ignored it. Fluoride, to us, is 14 not-important. It doesn't turn up in --
15 DR. WYMER: Sure.
16 DR. SHOESMITH: -- our ground water, so I haven't 17 ' thought about it.
18- But we have done a little bit of work on fluoride.
19 Fluoride will-act on -- yeah. Fluoride could be dangerous 20 if.it's in concentrated form. I couldn't put a number on E 21 how dangerous.
22 The f.luoride will actually complex the uranium 4, 23 as well as, it will complex the uranium 6, which means it 24' . will.changeithe nature of-the surface before the oxidation 25 process, and that's potentially.more dangerous than x.
Y 'C ANN RILEY & ASSOCIATES,. LTD.
~
d- // Court Reporters
?1025 Connecticut Avenue, NW,' Suite 1014
-Washington,. D;.C. 20036 (202). 842-0034' E___ _ .__ _
350 1 complexing after the oxidation process. But I couldn't tell
('
Q,)/ 2 you how effective it might be.
3 DR. WYMER: It will do the same thing to 4 plutonium?
i 5 DR. SHOESMITH: Yes. Yeah. That's why the --
6 DR. WYMER: Very effective --
7 DR. SHOESMITH: -- processing environment. Yeah.
8 DR. STAEHLE: Well, fluoride will have a bigger 9 effect on corroding zirconium.
10 DR. WYMER: Right. I mean and you don't -- it 11 would have at least two effects, the complexing you just 12 mentioned, plus the corrosion of the zirconium.
13 CHAIRMAN GARRICK: Chris?
14 DR. WHIPPLE: Dave, you commented on the retention (h 15
( ). of radionuclides by corrosion products. And it strikes me 16 that the potential for retardation that I've seen in your 17 list and elsewhere tends to suggest that the radionuclides 18 that are well absorbed in transport outside of the near 19 field tend to be those most likely to be retained within the 20 corrosion products.
21 DR. SHOESMITH: Yeah.
22 DR. WHIPPLE: And the quick ones, like iodine, are 23 least likely to be retained in both cases.
24 Is that -- is that correct, and if so --
- 25 DR. SHOESMITH
- Yeah. I think as a general I
i
[ l ANN RILEY & ASSOCIATES, LTD.
Es/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 !
Washington, D.C. 20036 l (202) 842-0034
351 1 perception, that is correct. Yes.
['T
- 1 2 DR. WHIPPLE
- So the relative benefits from x_/
3 accounting for that, if you've got adequate sourcing 4 elsewhere is not so great?
5 DR. SHOESMITH: Well, that's true. Yeah, if you 6 can show it, you don't need to retain.
7 I-- I don't know what -- too much about the 8 details of what the transport processes are in this model, 9 but --
so -- but I keep hearing they're turning into 10 something which dominates the dose rate release, which says 11 to me that, once it's out, the fuel -- whatever retardation 12 mechanism there is, after it has left the fuel surface are 13 not good enough, but I don't know what they are. So I -- so 14 I would say that, in that case, retention at the surface of
() 15 the dissolving surface is quite critical.
16 Now plutonium, unless it hits the colloids and 17 gets transported that way, I don't think is a big issue.
18 And most people find plutonium just sticks. It doesn't want 19 to go anywhere.
20 DR. WHIPPLE: Do you -- well, let me -- do you of 21 much data that exists on technetium and how it might do, 22 because, typically, no credit is taken for anything that 23 retains it or retards it.
24 DR. SHOESMITH: Yeah. Well, there aren't too many 25 that do. It's a bad actor in our reducing system, except
^
[ _J~^) ANN RILEY & ASSOCIATES, LTD.
Court Reporters
.1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
352 l
1 insofar as it meets iron -- reducing ion -- oxidizing ion
/ 2 minerals. So if it comes up against IN2 in a soluble state, l 3 it will, in fact, be reduced, and it's much more insoluble 4 when it gets to TCO2.
5 So in very sort of crude preliminary attempts to l
6 look at the reaction of technetium 7 with magnetite or ion 7 oxide surfaces shoulder, this will occur very rapidly, but 8 it's so localized that the TCO2 then blocks most of the 9 surface.
10 So most of the ion that's in the system is not 11 available. Unfortunately, you don't have that. Most of 12 your ion is going to already be oxidized. It's not going to 13 be reduced.
14 I find it hard to think that there is a real (Oj v
- 15. retardation mechanism for technetium, but I still think it's 16 worth chasing because one of those early release species, 17 which is coming out early -- but I -- I think you're 18 probably in rough shape for technetium because it's 19 oxidizing in your environment.
20 We didn't have a lot of luck, as I say, on the ion
)
21 oxide. It will do it, but it blocks the surface pretty 22 fast.
23 CHAIRMAN GARRICK: Bill?
24 DR. PAYER: Dave, you focused on solid phases that 25 form during the dissolution of fuel. What's happening to
(}
N,/
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 !
353
'l the aqueous environment that's in contact with that?
~/ D 2 And I guess the interest is, how might that G-
'3' 'nteract i with the remaining waste package structure?
4 Is it going to become more corrosive or_less 5- corrosive to the --
6 DR. SHOESMITH: Sorry. Which is this, . Joe?
7 DR. PAYER: When you envision the dripping onto 8' the fuel-and the formation'of these products, what's 5f happening to that aqueous phase? Is it becoming highly.
10 ~ acidic?
11- DR. SHOESMITH: Oh, okay.
12 DR. PAYER: Its corrosiveness on cladding and' 13 other materials --
14 DR. SHOESMITH: Okay. I get the point.
I L 15 DR. PAYER: -- the remaining waste package itself?
16 DR. SHOESMITH: The answer is, I don't think so.
17 None of the rates.that you would anticipate in this
- 18. environment.
1 19 And you could pull the pH down on a reacting UO2 20 surface to between three and four, if the rates are very 21 rapid, because the rate has to be rapid so the hydrolysis
-22 process can do what it does in crevices and pits. So if it l 23 gets going very quickly, say, where there's a grain missing, l 24- then you will start to pit at that site because it will go 25- acidic, and the transport losses will be slow as soon-as you L, .
hN ANN-RILEY & ASSOCIATES, LTD.
\/ Court Reporters 1025 Connecticut Avenue,.NW, Suite 1014 y Washington,.D.C. 20036 (202) 842-0034 o
,_ !_: - ._ .__ - ..- - - _ - _ - . - . . . - = - - - - - --L_------=--------
,--__.7 - - - - - _ - , - - - , - , - - . _ _ - . . _ - . . . _ , - . _ _ - . _ _ _ _ _ - _ _ _ _ . - - , _ _ _ _ _ , _ _ . _ _ . _ . _ . - - - . - _ - - _ _ _ . _ _ _ _ , _ _ _ _ _ _ . _ _ - _ - _ _ _ _ - - _ _ _ , - . _ _ _ _ , _ - - _ - - _ _ _ _ - , . , _ _
354
- 1' - ' start accumulating; product. 'So it will; pit. It's pitting 2 .' lby.a peculiar. mechanism..
3 'But1that: requires 1that the rate be' fast,.and-I c4 -think:you'couldfonly see that -- I didn't bring-the data, 5 -but I think you can only see that'effect -- you can see it
-6 -on fresh fuel. So if the radiolysis; fields are very high;.
.7:
~
you get rapid dissolution acidification'and you' eat your way 8 -through the grain boundaries.
l 9 Outside -- it's a bit of a sideline to your 110 discussion, but~what I'm trying to demonstrate here is I
' ll- don't think the UO2 will be reacting rapidly.enough to.give r
12 Lyou'a bulk acidified solution. And,~therefore, it will not 13' have a. major effect-on the corrosion of other parts of the 14- fsystem. But if you fail early, then it's a nasty business.
G-( ,/ 15' DR. WYMER: One more question from' Roger.
16 DR. STAEHLE: Dave,~ technetium ~-- you're probably 77- aware of this, butLthe chemistry of technetium was studied L 18' in the 1950's with respect to it's being an inhibitor, and 19 its role seemed to be at serving on the surface and doing 20 something.
1 21' DR. SHOESMITH: Yeah.
~
22 DR. STAEHLE: And I don't know whether that's a l
23- .usefulLinput, your thinking.about'that.
24: DR. SHOESMITH: .Yes.
l, '25 DR. STAEHLE: TheLsecond thing I wanted to mention L
L 4
E ANN RILEY & ASSOCIATES,-LTD.
Jd%-h Court > Reporters 4~' c1025 Connecticut Avenue,~NW, Suite 1014 H .
~
Washington, D.C. 20036-Li c
4(202) 842-0034
.~
~
355 1 is that, .you know, most colloids are charged. And it seems 4
l') 2 to me that they're going to end up accumulating someplace as G'
3 opposed to being transported. And this is, for example, it i t
4 happens at steam generators. You get charged particles, and 5 they. accumulate at various interfaces. And I don't know 6 what -- it seems to me that sort of a natural filtering 7 process is going to result --
8 DR. SHOESMITH: -Yes.
9 DR. STAEHLE: -- as a result of their being 10 charged.
11 DR. SHOESMITH: Well, that's a -- that's a good --
12 a good point. I mean I wouldn't claim to know very much 13 about colloids, but if, indeed, there is some kind of 14 charged neutralization process which will force them to
( )\ 15 accumulate somewhere, and there is a natural filter in the 16 system, then that's just great. But I haven't had anybody 17 talk to me about it, not that they should have to talk to me 18 about it, but the idea is a good one.
19 From the point of view of the technetium, two 20 factors. Yeah, it acted as an inhibitor, but I think it 21 acted as an inhibitor on things like stainless steel for the 22 very same reason that we found that it blocked magnetite l 23 dissolution. So if it finds a corrodible ion surface, 24 TC042- is a great cathodic reagent. So it acts like a 25 cathodic reagent, but its product, TCO2, is a highly
/ ANN RILEY'& ASSOCIATES, LTD.
\_/- Court Reporters 1025~ Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036 (202) 842-0034
356 11 insoluble phase.
[~'J 2' So I think it probably achieved its corrosion v.
3 inhibition by reacting initially with the carbons -- with 4 the stainless steel, which is where thora experiments were 5 done, forming a TCO2 layer at the active site and just 6 blocking the corrosion.
7 DR. STAEHLE: Just like the chromium?
8 DR. SHOESMITH: Yeah. It's -- but, here, I don't 9 think it's enough technetium for that to be an effective-10 inhibitor. As a dose factor, it's important, but it's not 11 enough of it around to be a real inhibitor.
12 DR. WYMER: We have another chemistry presentation 13' by Dr. Yoonhang Ahn from the University of California at 14 . Berkeley. He will talk about the chemical issues and (n) 15 considerations for the use of backfill in an unsaturated 16 zone repository, and I'm sure he goes beyond backfill in 17 what he's going to say.
18 CHAIRMAN GARRICK: Just a word of information 19 here. For the people on the other end of the video, they 20 are seeing or attempt to use that camera to project what is 21 on the screen, so we'll continue with the overhead but Lynn 22 -is going to turn the viewgraphs on this camera.for the
- 23. people in San Antonio and Las Vegas whenever they come up,
-24 because'they are just not seeing the viewgraphs because of
'25 the shadows.
[]
\ L ANN RILEY &' ASSOCIATES, LTD.
s- .
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
_a ___
357 1 .DR. AHN: Good morning. My talk today is about
() 2 3-the chemical issues and considerations for the use of
. backfill in a repository. As.you may know, I have been 4 working _or I was working in Japan since 1995 and I have been 5 and I am working for the Japanese repository program, so
'6 what I am going to.present today is deeply influenced by 7 .that program and I am trying to extract the analog to the 8 Yucca Mountain Project as much as possible, so you will see 9 the word " glass"'or something like that, relevant to the L10 Japanese case, but I will try to interpret that in the 11 context'of the Yucca Mountain repository.
12 The study I am going to show today was done with-13 various co-authors and I am also pointing out that when the 14 . place comes.
() '15 As far as I understand, at Yucca Mountain no 16 backfill is considered because of these reasons. Yucca L
17 Mountain is, as you know, located in a partially saturated 18 Topopah Spring with a tough layer about 200 meter below the
~19 surface and the infiltration rate is very low but we can 20 expect that the current dry period will end within 10,000 21 years or so. That is corresponding to the waste package 22 lifetime, since the spent fuel generates heat significantly 23 during the first 1000 or couple of thousand years.
24 So for this time period, the very slow corrosion n 25 .of container isLexpected so practically no release of
/ b ANN RILEY & ASSOCIATES, LTD.
hw / , Court' Reporters
'1025 Connecticut Avenue, NW,' Suite 1014 Washington,fD.C. 20036 (202) 842-0034 L
. _ _ _ _ _ _ . _ __ - A
358 1 radionuclides is expected and so as a conseque..ce the
() 2 3
backfill is not that important.
However, I think with no backfill uncertainties 4 will increase because the redistribution of water moisture 5 and the partially-saturated conditions following waste 6 emplacement has emerged as an important yet contentious 7 issue.
8 In a pluvial period, precipitation could be more 9 than one order of mc3nitude greater than now and however 10 several different conceptual models leads to rapid episodic 11 condensate drainage and/or saturated flow of significant 12 volume of water through fractures and the first pluvial 13 period will probably come after the heat generation ceases, 14 and so we can expect that significant volumes of water
() 15 flowing into the emplacement drifts may concentrate on some 16 waste package and rubble of waste forms may be generated, l
17 which is very difficult to model and the colloid facilitated 18 radionuclides transport can occur.
19 Well, this has been mentioned many times during 20 this meeting. The Department of Nuclear Engineering at U.C.
21 Berkeley was involved in the criticality analysis for the 22 last couple of years and we found that this is pretty -- or 23 very much difficult issue when we consider the scenarios for 24 the criticality and if that is removed from the scenario, 25 that would be very helpful for the confidence-building.
O .,
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
359 i
1 I would like to propose backfilling of the I
() 2 3
excavation hole ~by bentonite that might -- the reason for my choice is very simple because I have been working on that 4 for a long time.
5 (Laughter.]
6 DR. AHN: And there was no material in the tunnels 7 so why not? So I made a little, simple experiment -- what 8 if we fill the space with backfill? What will happen for 9 the Yucca Mountain repository?
10 For those who are not familiar with bentonite, I 11 summarine basic feature of bentonite. Bentonite, the name 12 of bentonite comes from the name of a place, Benton, in 13 Wyoming, and so we have lots of bentonite in the United 14 States. Well, the bentonite is a mixture of several
() 15 minerals. About half of the bentonite is smectite and 16 smectite is depicted here. Others are called feldspar, 17 pyrite; and cosite.
18 As you wil) see later, pyrite has a very important 19 role for the chemical buffering by bentonite.
20 Smectite has a sheet layer like this. We have l 21 octahedral sheet and tetrahedral sheet and these three 22 sheets consist of one layer of bentonite and these black 23 dots represent silica or aluminum and due to the changing 24 balance these cathode ions will be included in between two 25 layers.
i O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1C25 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
360 1 Okay. The material properties are summarized
() 2 3
here. The porosity is usually can be as high as 2 to 2.5 grams per cubic centimeter if
.3, 30 percent. The density 4 conducted. The hydraulic conductivity is very low. If you 5 compare with that.of granite, there is a couple of orders of 6 magnitude difference here.
7 Cation exchange capacity is also very important.
8 If you compare with natural zeolite, the bentonite has a 9 very good ion exchange capacity. It also swells very well.
10 By imbibing water, bentonite can swell. I will show some 11 pictures later. The reason for swelling can be summarized 12 these two: one is the repulsive force caused by electric 13 double layer of charged particles, shown here, and the 14 expansion of inter-layer of smectite caused by hydration of
/~N
() 15 inter-layer cations.
16 You may think that since -- by imbibing water 17 bentonite becomes very soft and so the waste package may 18 sink. There is an analysis done by P&C Japan, a 19 mathematical analysis, and also some experimental work. The 20 waste package displacements over 1000 year period was 21 estimated to be 1 millimeter or less, so with bentonite we 22 can settle the waste package very well for a long time 23 period.
24 With bentonite, we can expect these four 25 functionalities. One is the low hydraulic conductivity; the
[ ) ANN RILEY & ASSOCIATES, LTD.
\/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
361 1 .second one is swelling and self-sealing of fractures;
[)
V 2L chemical buffering.and colloid filtration.
3 I am going _to talk about these three points but 4- for this I would like to point out that.also-P&C did a.
5 . well-controlled experiment for that-and with benton
~6 . practically no colloid was observed at the surface of the 7- . bentonite, and'so with some well-controlled. abstracted 8 2 experiment the colloid filtration has.been demonstrated 9 .actually by experiment.
'10 The low hydraulic conductivity -- by this, as Dr.
11 Apted pointed out yesterday, release of radionuclides from i
12 the waste package will always be limited by readily 13' predictable molecular diffusion regardless of uncertainties 14 in future hydrological flow conditions of the geosphere, so
- 3) 15 we can decouple the near-field environment and far-field 16 environment.
17 Swelling and the self-sealing -- the intersecting-18 fractures will be sealed by expansion of bentonite by uptake 19 of water. Well, this has a very important effect on the 20 performance of engineered barrier. By swelling -- well, by 21 filling fractures by bentonite, hydrology around the bore 22 hole may change. Also geochemistry of water may change and 23 that is also a subject of this point.
E2 4. The chemical buffering, the geochemical conditions 25 intbentonite will be stabilized in reducing conditions for a
[ ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut' Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
s 362 1- wide variety of compositions of groundwater flowing into-Cp L' 2 bentonite.
N/
3 '
I'm going to'show analysis on these three points 4 in my presentation today. So for the first point, in this 5 picture'I showed the water low around an in the EBS. In 6 this picture this circle represents the boundary of EBS and 7 surrounding near-field rock. This outer circle represents.
8 the fictitious boundary of near-field rock. So if we have 9 another neighboring canister here, then a tunnel, this'is 10 the meet-plane between two tunnels or two waste canisters.
11 So.this is rock, and here I have waste package. And so we 12 have bentonite in this small doughnut like region, and we 13 have rock here.
14 Okay. Here I generated fracture network,
) 15 depending on some fracture statistics. 'Here I consider 16 primary fractures and secondary fractures. Primary 17 fractures exist prior to the excavation of the hole, so this 18 :is.the fractures existing before the excavation, and the 19 next, the other one is the. secondary fractures. Secondary 20 fractures will be introduced as a result of the excavation.
21 So here I assumed that the distribution of 22 secondary fractures are high around this interface and low 23 around this perimeter. Okay. Based on some assumed
-24 statistics of fractures, we can generate fracture network or 25 fracture clusters. Depending on the lengths of the 1
( ANN RILEY & ASSOCIATES, LTD.
\~s/ Court Reporters.
1025 Connecticut Avenue,. NW, Suite 1014 P ' Washington,. D.C. 20036 (202L 842-0034
363 1 fracture, or orientation of,the fracture, we can identify
/ i- 2 which fractures are directly connected with this region, the V
3 -EBS. region and-the surrounding ~ global water flow.
4' In this case we have fairly long. fracture segments 5- and so the network actually can cover almost all the entire
'6 region of the.near-field rock. And here I.show the water 7 flow in the near-field region. As you can see, the water 8 flow inside the buffer is very, very slow. On the righthand 9 side, I show the simplified version of this picture. I show 10 just the fast flow part only. As you can see, the water
.11 flow-in the EBS is quite-negligible.
12 If I change the fracture statistics, we can have 13 this kind of configuration. I assume that the fracture 14 . segment is a little bit shorter than before, then the 15 connectivity becomes worse. And so the cluster, fracture 16 cluster can cover small fraction of the domain. And the 17 water flow, as you can see here, is very limited, very 18 limited around the boundary. And the entire near-field 19- region is characterized as stationary water practically.
20 If I assume very short fractures, we have 21 something like this. So water flows in to this region and 22 flows out and the water here is actually stationary. So 23- from here, from this analysis, I can summarize that, 24 depending on the fracture statistics, water supply to the 25 EBS' differs significantly and water is considered to be
- ( ANN RILEY &' ASSOCIATES, LTD.
N-r Court: Reporters
, . 1025 Connecticut Avenue, NW, Suite.1014 Washington, D.C. 20036
.(202) 842-0034
364 1- stationary'in the EBS and in the near-field rock. Mass l' h 2 transport in the EBS'and_in the near-field rock will be V-3 -dominated by molecular diffusion.
4- AndLthe second point, swelling and self-sealing.
5 This is the experiment done by PNC more than seven years ago 6 for bentonite expansion in a planar fracture.. Fracture is 7- simulated by aqueal resin and here you can see the 8 apparatus. The bentonite specimen is here, simulated 9 fracture intersect the bentonite specimen. Water is filled 10- and -- actually, they did two kinds of experiments, static ill water experiment and flowing water experiment, but I would 12' like to show only the static water experiment here, since 13 the previous analysis shows that the water is probably 14 ' static in that region.
() 15 As you can see, _as these pictures show, taken from 16 this, from top side, bentonite expands into the fracture, 17 and as time goes on, the tip location of the expansion 18 increases. The' expansion of the bentonite can be analyzed
- 19. based on the consolidation theory of clay mineral.
20 With that, we can reproduce the experimental 21- results fairly well. And here the tip position is expressed 22 as.a function of a square root of time in hours. The 23 e-limit here is the volt ratio at the tip of the expansion.
24 And with these numbers we can have very good agreement with
- 25. the experiment.
U ANN RILEY.& ASSOCIATES, LTD.
} -'
- 3. ^- . Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C.-20036 (202) 842-0034
____.___m__- _ _ _ _ _ _ _ . - . _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _
365 1 But the time scale here is very short, it's like
() 2' 3
-2500 hours, around 100 days.
-context of repository performance?
So what does that mean in the Well, first concern here 4 Lis how much bentonite will escape from the EBS region. This
.5 is very rough estimate of that. I assumed that the total 6 fracture aperture of fractures intersecting the EBS'is one 7 centimeter. That's very hypothetical, but let's use that 8 number as a hypothesis for a while. Then the bentonite.
9 volume flux out of the bentonite region is calculated like 10 that, and the. integration of this curve will give the 11 cumulative bentonite loss.
12 At around 900 years the bentonite, about .055 13 -cubic meter of bentonite, is lost from the EBS region. What 14 that means is that since the total volume of bentonite
() 15' ' allocated for one canister is around three or four cubic 16 meters, only a few percent of bentonite is actually lost 17 into the fractures. So I think the bentonite loss does not 18 affect the mechanical stability.of the repository.
19 So here I summarize this part like this.
20 Bentonite lost from the EBS region would be negligibly 21 small. That negligibly small amount of bentonite will fill 22 the fractures around the EBS. That will affect the water
- l. 23 flow and the geochemistry. Actually, this, even though the 24 amount'of bentonite lost from the EBS is small, this has 25 very strong effect on the radionuclides transport from the l
3-LANN RILEY & ASSOCIATES, LTD.
- l. Court Reporters L
-1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
366 1 ' waste canister to the-outside water flow system.
'2'
-Also, this has -- so the_ region where bentonite 3 fills fracture will play as an intermediate region for the 4; geochemistry inside the EBS. So-we have to first consider
~
5 'the geochemistry around the disturbed region and then we 6 -have to consider the geochemistry of'the EBS, pore-water in 7s EBS.
8 However, in a realistic situation, fracture 9 network is formed, not that simple fracture, planar 10 fracture. So maybe we have to consider the fracture network 11- . geometry in the future analysis.
12 But the same kind of experiment is-being done in a 13 different geometry and the time law is exactly the same, the 14 square root of time law. So probably in different (A) .15 geometries, the bentonite will expand into the fractures 16 with square root of time behavior.
17 Now, the chemical buffering, for that part I 18 worked with a colleague in Japan, Professor Ohe of Tokai 19 University. He considered that equilibrium will be
- 20. established-between pyrite, which is originally included in 21 the bentonite, and magnetite. As an evidence, he considered 22 this picture. Snellman made an experiment in late '80s, I 23 think, for the pe and pH of bentonite pore water and he did 24; an1 experiment for two years. At the end of two year period,
'25 the pH and pe converges to this' level.
j L
h) k/-
iANN RILEY & ASSOCIATES, LTD.
. .. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
~
Washington, D.C. 20036
'{202).842-0034 g;___a_,_
L 367 1 Based on the thermodynamics model, or equilibrium I l
l 2 model, Professor Ohe claimed that there are two l
}
3 possibilities for the combination of iron minerals that 4 established this pH-pe combination. The possibilities are 5 pyrite magnetite combination and pyrite hematite 6 combination. However, no hematite is included in the 7 bentonite, so he concluded that the only combination 8 conceivable here is pyrite magnetite combination.
9 Based on that., we tried to identify the pH and pe 10 of the pore waters in the bentonite. However, the bentonite 11 pore water ratio in the repository condition is rather high.
12 That means in the EBS region, almost all the volume is 13 occupied by the solid particle and very little water is 14 included. So the measurement of pH and Eh, direct
() 15 measurement, is very difficult.
16 So what we did is the application of PHREEQE 17 simulation. And for that, we first checked the validation 18 or the applicability of the PHREEQE code for such a system.
19 In the JSS project, Japan, Sweden and Switzerland, I think, 20 in 1987, shows that the bentonite dissolution experiment for 21 365 days with nitrogen purge, closed system at 90 degrees C 22 with de-ionized water, these points for pore water 23 chemistry.
24 Well, with the PHREEQUE code, we could reproduce 25 this behavior, except for aluminum, quite well. So from
~
(N~')' ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 6 (202) 842-0034 1
368 1 this comparison we concluded that PHREEQUE can actually
/ 2' If that is so, we simulate the pore water reasonably well.
ks)-
3 can use that code to estimate the pH and Eh of the water.
4 And we did quite a broad extrapolation to 7,000 gram'per 5 liter region based on the very dispersed system experimental 6 . data. And the' extrapolation shows that the pH of repository 7 -- I mean pore water will be somewhere around 9.
8 And about the chemical buffering by the bentonite.
9 Pyrite can exiL+ only at extremely low oxygen concentration.
10 and magnetite can exist for a wide range of oxygen 11 concentration. If oxygen intrudes into that region the pore 12 water recovers reducing condition by dissolution of pyrite.
13 Dissolved pyrite will reprecipitate as magnetite. And so by 14 consuming pyrite, pore water will be buffered for a certain A) s_, 15 time. And depending on the amount of pyrite and the amount t
16 of oxygen entering, we can estimate the time for which the 17 buffering is achieved.
18 As you observed before, water flow into that 19 region is very, very small, so in a very -- not realistic, 20 but kind of wishful thinking, but in some case, more than 21 millions of years time, the buffering can be achieved.
22 So for the summary for that part, equilibrium is 23 established between pyrite which exists in bentonite 24 originally and magnetite. Consequently, pH around 9 and the 25 reducing environment would be established. Considering the
( !n 2 ANN RILEY & ASSOCIATES, LTD.
\- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
- c. (202) 842-0034
)
]
369 I
1 very small water flow rate, the buffering by bentonite may 2 last for a sufficiently long time' period. But, of course,
(\_)h 3 for the -- we have to do very detailed analysis for that in 4 the future if we apply bentonite for the project.
5 Well, so far we have been observing the background 6 conditions of repository performance. Now we are going to 7 consider how that -- what is the implication of that to the 8 repository performance. Here I considered with Dr. Nakayama 9 of Japan Atomic Energy Research Institute the migration'of 10 neptunium through the overpack and through the backfill.
11 I learned a lot yesterday about the corrosion l I
12 that's actually very helpful. But one thing I could not get i i
t 13 was what would be the final stage of corrosion of overpack 14 or canister. Many analyses seem to be done for determining A
( ,) 15 when the first hole will be created or when the first 16 release of radionuclides, the pitting hole will be created, 17 was done, I think, very intensively. But for the 18 radionuclides transport objective, we need to know what kind 19 of corrosion products will be created with, with what 1 20 amount, in what form, will that be mixed with buffering j i
l 21 material. What is the hydraulic conductivity? What is the )
I 22 pore structure and so on?
{
23 I don't think the current corrosion simulation or 24 studies have not shown that kind of information. Well, if I 25 am wrong, I am very pleased to learn and modify this model.
s 1 l t ANN RILEY & ASSOCIATES, LTD. 1
(/ Court Reporters {
1025 Connecticut Avenue, NW, Suite 1014 I Washington, D.C. 20036 I (202) 842-0034
)
__ __ 1
370 1 So I considered here that overpack'is completely
'2 corroded and is porous medium with porosity of 10 percent.
.3 Some may.ask how do you determine that number.
Well, there 4 is'no rationale here, but I set somewhat lower number than 5 'the porosity of bentonite.
6 Here I considered that the neptunium.is released 7 in neptunium'4 form if there is no oxygen contact arriving 8 this point. The neptunium taay oxidate -- oxidized by oxygen 9 coming from this side, or reduced the oxidized neptunium 4,
- 10. which is neptunium 5, will be reduced once again by ion 2 in 11 the overpack region or some ion diffusing into the buffer.
12 And so we have to consider first the diffusion of iron and 13- oxygen in this region. And after we establish the oxygen 14 ion distributions in this region, we are going to consider
() 15 . neptunium transport through this media.
16 The reason for the coupling of two things is the 17 following. Because oxygen diffusion may be much faster than 18 the neptunium movement since neptunium is expected to be 19 retarded by sorption. So it is expected, or it is ase.imed, 20 actually, that the oxygen ion concentration profiles 21 established and reached some kind of steady state before 22 neptunium starts transport through this region. The 23 chemical.reacticas I assume here are another point for 24 discussions.
-25 Dr. Nakayama and I tried to find a good chemical 9 l ANN RILEY & ASSOCIATES, LTD.
k^- .
Court Reporters jfj .1025 Connecticut Avenue, NW,' Suite 1014
-Washington, D.C. 20036 y_ :(202)842-0034
[ 371
'1 . basis for.the reactions of neptunium, which-is iron and O-
.O 2- oxygen.. We could find chemical reactions information in the 3 very acidic = condition. The reason for that is we did much 4i research on'the reprocessing, and'so in the high acidic
'S condition, we have very much information, .but in this 6 particular near neutral pH conditions, we do not have much' 7' information, or actually very little.
l 8- So we assumed these chemical reactions for this 9 study and established a mathematical model for distributions
-10 of neptunium into the various systems.
11 These two equations represent the diffusion of 12- neptunium in two different species. As you can see, there 13 are four. terms representing the chemical reactions in there.
14- Interestingly, with pH of 9, reactions with iron 15 and oxygen is very fast, and so practically all the oxygen, 16- including into the vessel from this outside boundary, will 17_ react with iron and precipitate here as iron three. -
18- So practically no oxygen will arrive at this
.19 surface, and so neptunium will be released in neptunium 4 20 form and diffused through the region and then with much_
21- oxygen and iron 'three here, neptunium is oxidized and 22- diffuses as neptunium 5 and released.
23; With pH of 6.5, reaction of iron with oxygen is 24 very slow,-so some oxygen can survive up to this point, and 25' so neptunium may be-released-as neptunium 5. So since there
/ I ANN RILEY & ASSOCIATES, LTD.
)/ . _
1025 Connecticut Avenue,.~NW, Suite 1014 Court Reporters-
-i Washington, D.C. :20036- i (202):842-0034 j
372 1 are both reductant and oxidant in these regions, some
[~)
'O -
2 neptunium 5 may be reduced but again oxidized, and so we 3 have two species at the same time in both regions.
4 After many calculations, we summarized the 5 calculations like this. The quantity quoted here is 6 neptunium mass flux a the outer surface of EBS, and the 7 horizontal axis shows the time from the beginning of the 8 release of neptunium from the waste form.
9 If we assume only neptunium 5 is released and 10 transported through the buffer, we have this low release 11- rate. If we do not have chemical buffering or reducing 12 mechanisms, then we have a high release rate of neptunium.
13 But if we have iron overpack, that is a major source of 14 reductant, then the neptunium is released as neptunium 4 and
) 15 then oxidized in the buffer and released at this mass flux.
16 If I summarize, if pH or pore water in the buffer 17 is at nine, oxygen entering the EBS from the outside could 18 be consumed by pyrite and iron from the container and the 19 waste form surface can be kept reducing. Neptunium 4 at the 20- waste surface is released and for neptunium 4, low 21 solubility and high retardation factors are assumed, so a 22 smaller release rate can be expected.
23 If pH is lower, oxygen consumption becomes slower, 24 and the environment around the waste surface becomes 25 oxidizing. Neptunium 5 will be released and the release
[ -ANN RILEY & ASSOCIATES, LTD.
- . _ ' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l l-l 373 L
.1- rate becomes greater.
/ w) 2 So far, we have considered the effect of redux kJ 3 reactions of neptunium. I would like to come back to this 4 swelling and self-sealing, the interaction of these two t
5 points, f 6- In the disturbed region, as I mentioned earlier, 7 more fractures are introduced in the surrounding holes.
8 Water is imbibed into the bentonite in the EBS, bentonite 9 swells and extrudes into the fractures, radionuclides 10 released from the waste solid diffuse in the EBS and then
.' 11 eventually diffuse into the extruding bentonite in the 12 fractures. Radionuclides are released at the tip of 13 extruding bentonite into the groundwater. Geochemical 14 conditions are determined by EBS and this host rock in the 15 disturbed region.
16 MR. HORNBERGER: Dr. Ahn, could I ask a question 17 of clarification? Is there enough water at the unsaturated 18 Yucca Mountain site to have the bentonite swell?
19 DR. AHN: Yes and no.
20 [ Laughter.)
21 DR. AHN: If at the -- the swelling -- the 22 ' complete swelling will not occur at the current 23 precipitation rate, but if it becomes greater, there is a 24 possibility that it will -- the major part of the EBS is 25 filled with water.
[). ANN RILEY & ASSOCIATES, LTD.
1~/' Court Reporters 1025' Connecticut Avenue, FM, Suite.1014 Washington, D.C. 20036
'r (202) 842-0034
___=__
374 1 Okay. To study the effect of the disturbed
.( %
2 region, I simplified the situation to a great extent. I 3 consider buffer and the fracture filled with bentonite, and 4, here, I assumed the radionuclides diffuse.is released from 5 the waste form and diffused through the bentonite and 6 eventually to the fracture and released at the tip of the 7 bentonite.
8 I consider diffusion into the rock matrix, and so 9 we have constant -- two governing equations for diffusion in lLO the bentonite and diffusion in the rock.
11 I assume that a certain tmount of radionuclides, a 12 unit amount of radionuclides is injected into this mouth for 13 simplicity, and I'm going to observe how much will be 14 released from the tip. For that, we need to know the A
1 ji 15 retardation factor of neptunium in this case in this if bentonite filled fracture region.
17 For that, Professor Nagasaki of Tokyo University 18 did the experiment for me with this simulated condition. We
- 19. 'have rock samples on top and bottom of the fracture and the 20 fracture is filled with bentonite. He controlled the 21 environment with this wide range of redux potential. He did 22 .the absorption distribution measurement for neptunium and
- 23. americium from the aerobic to anaerobic condition or from 24 anaerobic'to aerobic condition.
l 25 In either case, americium shows very -- quite i
\
. ANN RILE & ASSOCIATES, LTD. I Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ~
Washington, D.C. 20036 (202) 842-0034 L
375 1 constant behavior. For neptunium, the absorption j
,- ' \
'l 2 distribution coefficient is as high as around 70,000 3 milliliters per gram if the redux potential is negative, but 4 if we have an oxidizing condition, that is as low as around 5 30 or so.
6 We incorporated these results into the model I
.7 just showed before, and we obtained this kind of 8 concentration profile. We have a fracture here and 9 diffusion into the rock.
10 For americium, since this is relatively 11 short-lived, we don't see so much penetration into the rock, 12 but with the neptunium 237 in an aerobic condition, 13 diffusion is relatively fast, but at most up to the ten --
14 ten percent of the thickness of spacing between two (Oj 15 neighboring fractures.
16 To quantify the effect of the distribution region, 17 I calculated the reduction factor. Reduction factor is 18 defined as the ratio of total amount released at the tip of 19 the bentonite to the total amount of radionuclides flowing 20 into the bentonite -- the fracture at the mouth as a 21 function of the Sherwood number.
22 Sherwood number is the ratio of two mass transfer 23 processes in the fracture, but basically it represents the 24 rate of mass transfer at the tip, and the typical Sherwood 25 number range is around somewhere between .01 to 1 in the
. V_/
[ ') ' ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
376 1- ' repository condition.
2 In.this: region,.1as we can see; if:the -- for
.(A.)
35 neptunium.237 in reducing' condition, reduction factor-is I
'4 faround'.lito .008. So what that means isLonly a few percent
-5 'ofLneptunium 237 can be released from the tip:of,the 6' Lbentonite, and-the rest will decay out within that disturbed 7 region.
8 For. americium 243, even though, if we inject 9; Leverything, all the americium 243 into that region, we'can
- 10' -expect almost all the americium will decay there.
ll- .DR. WYMER: You have a couple of minutes.
12 DR. AHN: Sure.
13 For that part, diffusion in the fracture is 14 cretarded by, absorption and matrix diffusion, americium,
)
~
15 plutonium,-and if the environment is reducing, neptunium 16, decays to negligible' levels after transport through the
.17 disturbed zone. Disturbed region around the EBS can also be 118 regarded-as a barrier which has characteristics different 19: from the EBS or-from the natural barrier.
20- So far, I have been talking about the good things, 12 1 good side of the buffer, but there is one thing we have to-
- 22 consider -- the degradation of bentonite material. Here,.I
~23 considered the' utilization of smectite. Smectite could
~
24 change to illite in the presence of. iron, and the illite has 25 .two or three orders of magnitude larger hydraulic N ' ') ANN RILEYl& ASSOCIATES, LTD.
\/ Court Reporters. .
1025 ConnecticutiAvenue, NW, Suite 1014 g, Washington,;D.C. 20036 L (202) 842-0034.-
u_ l
377 1 conductivity, and as you can see here, lower distribution,
('v ) 2 absorption distribution coefficient.
3 This was also measured by Professor Nagasaki of 4 Tokyo University.
5 How does the illitization affect EBS performance?
6 As more illites are generated, hydraulic conductivity 1
7 increases, and so mass transfer rate in the EBS increases, 8 and so faster waste form dissolution. So we release more 9 radionuclides in shorter time period.
10 With absorption distribution coefficient 11 decreasing, we have faster transport of released 12 radionuclides, and so all together, we have very unfavorable 13 effect on the repository performance.
14 I will skip all these pictures. Please take a (n,) 15 look at that. But let me jump to the concluding picture 16 here.
17 As we can see, the effect is quite negligible.
18 Actually, these differences occur more because the 19 difference in temperature, and difference in temperature 20 --so for -- especially for neptunium 237, the release rate 21 doesn't change so much, maybe within factor of two or less.
22 So the summary I already mentioned, but I would 23 like to point out that the hydraulic conductivity of illite >
24 is the key parameter in this analysis. So to finalize this l
l 25 study, we have to measure this one very carefully.
[d
- ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington,'D.C. 20036 (202) 842-0034 L: _ _ - _ . . - . _ _ _ . _ _ _
?
378 1 _ Summarizing these conditions so far, water in the
'2 EBS andlin the near. field rock is practically stationary.
- 3. Bentonite loss ~from the EBS region would be small, but there 4 are the fractures around the EBS affecting water flow and 5 geochemistry.
6 Equilibrium is established between pyrite and 7 magnetite, resulting in pH around nine reducing environment.
8- If.pH of pore water in the buffer is nine, we can expect a 9 very slow release of neptunium as four, 10 The disturbed region around the EBS can also be 11 regarded as a barrier which has characteristics different 12 from"the EBS and from natural barrier. Illitization has 13 negligible' effects on radionuclides release.
14 ~ So I think I can -- I would recommend bentonite
(
' 15 for the Yucca Mountain repository, but, of course, we need
-16 to do many things to conform that. But the good point.here 17 is that we can consider a very robust model for mass 18 transport in the EBS. Diffusion is well known and we can L 19 characterize -- we can quantify that very well.
20 So I think in that regard, bentonite is good.
- 21' Also, bentonite will eliminate the possibility of plutonium 22 colloid. transport, and also this may be secondary, but
-23 knowledge and concepts can be shared with other repository 24 . programs. Already many studies have been done for that, and 25 I think the analysis method is well established for that, so ANN RILEY & ASSOCIATES, LTD.
E -O4 . Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036' (202) 842-0034
-379 il we can' share that_information. ,
'[f 2 The final point I would like to make is'that we
-k/
3 .need to~ develop an' integrated model that includes all these 4 ' concepts or considerations. I think yesterday,'I was very 5 ~ happy to know-all these corrosions and so on, because I
-6 :always wanted to know, but it's very useful if we could
~
7 interpret that kind of result in the context of performance 8 assessment.
9 For example, if the corrosion rate becomes -- or 10 decreases from such number to such number, what does.that Lil mean in the context of radiological safety?
12 DR. WYMER: I~think we probably have to saw.it off 13 'at this point.
14- DR. AHN: And in such a -- for such an objective, 15 I'think we should develop an integrated model. And whenever E 16 we present the effect'of corrosion or the effect of 17 bentonite or something like that, I'think we should show at 18' the same-time the effect on the overall performance.
1 19 ~Thank you very much.
20 DR. WYMER: Thank you.
21 Let's take one question.
22 Charles?
23' LMR..FAIRHURST: , I just - .first of all, I find 24 your paper very interesting Is PNC involved with the Febex 25 experiment in. Switzerland?
<' %g . .
-3 1 -MG9 RILEY & ASSOCIATES, LTD.
p h b: [ Court Reporters P , , 1025 Connecticut Avenue, NW, Suite 1014 1
.a Washington, D.C.-20036.
f- -
'(202)f 842-0034 -
p b
u
$ r
0 380 1 DR. .AHN: Yes. PNC is actually a very large
[\_/ }
2 sponsor for --
3 MR. FAIRHURST: Participant. Yes. Because the l 4 modeling of the ir.gress of water and so on is being done by l' 5 'the Spanish --
6 DR. AHN: That's right. That's right.
L 7 MR. FAIRHURST: --
and by the Swiss.
8 DR. AHN: That's right.
l l 9 MR. FAIRHURST: With I think a more realistic 10 network of fracture and flow than you have described.
.11 DR. AHN: Yes.
12 MR. FAIRHURST: So I think some of the 13 recommendations you will be making, your colleagues are 14 actually participating in an international experiment (O_j 15 full-scale. So it's very interesting. I don't want to say 16 any more than that.
17 DR. WYMER: Okay. Well, thank you very much. I 18 think we are coming up on break time. I think that Dave 19 Stahl is -- is he available now?
20 Why don't you make your announcement?
21 MR. CAMPBELL: What we are going to do is regroup 22 a little bit and Dave Stahl will be -- you are going to be 23 doing that from Las Vegas? You are going to have somebody 24 in Las Vegas give a presentation, so what we are going to
)
25 do -- and they won't be ready until just before Noon or
/ T AITN RILEY & ASSOCIATES, LTD.
\-.I Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 ,
(202) 842-0034 l l
i
-_ A
381 l 1 around 11:00, so at 11:45, we will be getting a feed-in from
- 2. Las Vegas on the testing program, is that right?
1 N~/'[
3 DR. STAHL: Engineering development and 4 fabrication work.
5 MR. CAMPBELL: The engineering development and 6 fabrication work -- so we will break now till 10:15 and then 7 we will start in with the DOE presentation by Bill Halsey, 8 okay? -- so that will run from 10:15 to 11:00 and then from 9 11:00 to 11:45 Dr. Leslie of the NRC will give his 10 : presentation and then Dave Stahl.
11 In addition, for the panelists, we are going to 12 distribute pens and blank viewgraphs during the break to 13 each of you so that you can use those to make up comments 14 and so on.
). 15 If you want us to make a viewgraph from an 16 overhead or something like that, contact me and we can have 17 that done as well.
-18 DR. WYMER: Well, thank you to the speakers for 19 these really interesting chemistry discussions. I think 20 there's a lot to chew on there, as a matter of fact.
21 Let's take a break for 15 minutes and come back at 22 about a quarter after 10:00 -- break for 11 minutes.
23 (ReceFs.)
24 DR. WYMER: Take your seats, if you will. ,
J 25 Following all these interesting chemistry discussions, now j/ )-
s -
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
_ _ - _ -_ _ _ a
it ,
y
'382 1 -- .we have'the model' treatment of the chemistry details, two
. 2 presentations.
3 The first presentation is by.Dr. Bill Hal'sey
~4. ~ representing DOE. He is from Lawrence Livermore -- model 5~ treatment:of chemistry.
6- DR. HALSEY: Well, I guess I have to thank Dave 7: Shoesmith for setting the stage this morning and explaining.
'8 all of the details of what does and doesn't happen and more
- , 9. importantly, what we know, and even more important_than-l 10- .that, what we don't know and what we have done wrong --
211 because that can cut my talking time down quite a bit.
- 12 I am going to talk about what we have done in the 13' TSPA-VA, which as Abe said yesterday is still a work in 14 progress.
Oj
- _t .15 We have the base case-done. .We are still working 16 on some of the sensitivity analyses and we.are still writing 17 ~ 'it up. _I'll touch on some of~these topics -- the-waste form 18 representations'in the total system performance assessment.
19 what the waste forms, briefly touch on the expert 20 elicitation that we have held.
-21 I will talk briefly about cladding and spent fuel 22 .and defense high level waste dissolution and a few of.the 23 sensitivity studies that are going on.
- 24 - They have also.used this -- this is one of_our 25L Eboxology charts,-as we call it, showing-once again we are I[
ANN RILEY & ASSOCIATES, LTD.
"A--f . .
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
.(202) 842-0034'
383 1 talking in this portion here in the TSPA code RIP waste form
[V) 3 2 degradation. We are also going to touch on the cladding degradation portion here.
4 The new version of the RIP code has a new and 5 improved EBS module in here which I won't dwell on, but it 6 allows us to do more in the way of sensitivity studies and 7 it gives us additional capability for including a lot of the 8 processes that we have been discussing at this meeting in 9 that it represents the engineered system as a series of 10 cells, actually you could think of each cell as batch 11 reactor and then there are both advective and diffusive 12 connections from one cell to the next. That is a capability 13 we hadn't had in the past.
14 A few assumptions in the area of waste form (O
, ,) 15 degradation and some of these are different than in previous 16 performance analysis.
17- One important assumption is that the waste forms 18 are assumed to be exposed to the drift environment upon 19 failure of the waste package and upon failure of cladding 20 when in those cases where we are taking credit for the 21 cladding, and that water films adsorb onto the porous 22 alternation product layers assumed to provide aqueous 23 conditions. That is an assumption that we had not used in 24 the past and this is partially based on observations of low 25 flow rate and water vapor exposure of both defense high-
,m
.{ } ANN RILEY & ASSOCIATES, LTD.
N Court Reporters 1025 Connecticut Avenue, NW, Suite.1014 Washington, D.C. 20036 (202) 842-0034
384 1 level waste glass and spent fuel -- the experiments for
() 2 3
example at Argonne where the waste forms generate a porous alteration product when you don't have a high water flow 4 rate, as Dave was discussing this morning, and this 5 alteration layer, whenever we can observe it, appears to be 6 pretty well saturated with water.
7 So this is an assumption that we have taken which 8 has changed some of the way that our models end up working 9 and it leads in part to the problem that Dave was 10 discussing, that our waste forms tend to dissolve rather 11 quickly.
12 Waste form degradation is represented by an 13 intrinsic dissolution rate equation. I will get to this-a 14 little bit later.
/\
! ,) 15 The radionuclides are considered potentially 16 available for mobilization at that rate but they are not 17 necessarily mobilized at that rate.
18 The highly-soluble radionuclides are mobilized at 19 that dissolution rate into either a diffusive or an 20 advective transport pathway. Those radionuclides that are 21 not highly soluble are mobilized at an aqueous solubility 22 limit which we have a distribution that has been developed.
23 We sample stochastically over that distribution 24 within the analysis and that range represents some of the l
25 unknowns in chemistries and dominant solid phases which
[)
' /
ANN RILEY & ASSOCIATES, LTD. !
Court Reporters 1025 Connecticut Avenue,.NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
I L___________ _
j
__- _-_-_ _ __ _ ___ _________ _ _____-________ _ _y 5
-385 1- . control 1the solubilities.
l
.2 Afpreliminary' representation-of aqueous l, 3_ concentrations limited;by the secondary.phasas;has been-p '
K '4- prepared. We did not.use it in.the base case. We:a're using 5 Lit-in some of the sensitivity' analyses. .I will. touch on 6 Esome of thatta little bit later.
7 As part of the process, a year and a half ago we 8 held'a workshop on where was it that we wanted to go with-
~9 the models forfthe TSPA-VA. If we had had a succinct.
L10- ' discussion -- if'.we had known everything that Dave was able ._
211? .to explain this morning'a year and a half ago, we could have
.12 avoided a few of.the pitfalls, but then again the year and a 13 . half's worth of work'is how we learned some of those things. ,
14L As you said,. Dave, you do allJof this work and you
' O( j
' 15 L come up that this is not important, so they say, well, why 16 did you do it?
17: [ Laughter.]
'18 DR. HALSEY: Because we didn't know it at the 19 time.
.20 I would.like to borrow your. curve, by the way, of
- 21 " : confidence or_ uncertainties as a function of' time. I think 22: we.are just on the back side of this into that rapid, jagged
~
- 23) -section:of your theory.of theories.
24 DR. SHOESMITH: That is'usually where.they cut
- .25! lyourtfunding.: Be careful.
, j~s (l U ANN;RILEY & ASSOCIATES, LTD.
Ms/[
'?
. _ Court Reporters;
'1025-Connecticut Avenue,x NW,JSuite 1014 f ' Washington, D.C'. 20036 1(202)-842-0034 s
1 386 1 [ Laughter.]
() 2 3
DR. HALSEY: I understand.
More recently, as we developed the models we held 4 a waste form expert elicitation to provide a second or 5 another outside sanity check and provide input on the 6 parameter ranges, the uncertainties and the alternative 7 conceptual representations. We have got some 8 representatives from that in the room.
9 Let's see -- the report is in review. The results 10 have been used to guide some of the sensitivity studies.
11 This is a list of some of the issues that were elicited.
12 The ones not in -- a few of them here we will not talk about 13 today but some of the rest of them I will touch on. That is i 14 the range of the dominant processes -- that is what we have
() 15 been talking about most of the morning -- the exposed surface area; the wetted surface area and active surface 16 17 area, how we get at those; the cladding degradation; the 18 spent fuel dissolution rates.
19 We will not discuss solubilities, secondary phau 20 retention of radionuclides. I was not going to go into the l 21 rapid release fraction, the defense high level waste 22 dissolution I will touch on. I was not going to get into 23 the colloid formation and rransport. Some of these would be 24 a good topic for an afternoon meeting all by themselves.
25 Let's launch right into a few of them.
O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, ') . C . 20036 (202) 842-0034
387 1 I am not going to go into the details of the
() 2 3
cladding degradation modes. That again would be a whole 45 minute talk, but, yes, the base case for TSPA-VA does use 4 cladding credit. This is something that we developed over 5 the last year and decided to -- the decision was made to put 6 it into the base case.
7 We also have done analysis with different levels 8 of credit and with no credit for comparison.
9 It is built of a number of parts. We have an 10 assumption of a certain amount of juvenile failure in the 11 couple of percent range. Stainless steel cladding is 12 assumed to have all failed by the time the waste packages 13 fail. There has been a preliminary mechanical disruption 14 model developed which allows rocks to fall from the top of
() 15 the drift through to the waste package and through a 16 degraded waste package, impact the cladding or the fuel 17 pins, and this gives us failures in the range from 30,000 to 18 500,000 years in the few percent to tens of percent range.
19 We looked at creep failure. This is the creep 20 rupture and it looks like that is going to be a fairly small 21 contributor to failures and they are all in the early ;
22 timeframe because as temperatures drop the driving force for 23 that increases rather quickly. And under the conditions 24 that we were anticipating on the cladding, we did not find 25 any real strong evidence for stress corrosion cracking.
- h. ..
O ANN RILEY & LSSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
388 1 I'll have to agree with some of the earlier j 2 discussions that at very long times, are there processes 3 that might change that?
4 We'll have -- those are some things we have to 5 look into. I'll get to that, down to localized corrosion, 6 also.
7 Oxidation and hydrogen embrittlement. We looked 8 at the rate of oxidation at the temperature histories for 9 the hottest waste packages, with early failuras of the waste 10 package. And for the worst cases, there may be some 11 contribution to zirconium oxidation driven failure and 12 pickup of additional hydrogen.
13 But with the information that's available, it did 14 not look like a significant contributor, either. I also I
(j 15 agree with Dave. We need to look at very long times. If 16 we're looking at a million years, we need to look at where 17 might additional hydrogen pickup come from, and what might 18 it do inside the zirc alloy. But the data we have to date 19 did not look like a significant problem.
20 Localized corrosion is one of the issues that came 21 up from the expert elicitation. The zirconium oxide layer 22 is extremely passive, as long as it's intact and you don't L 23 have something interrupting it.
I l 24 There has been some pitting corrosion observed at 25 low pH, particularly with the ferric ions. Ferric chloride a
I ANN RILEY & ASSOCIATES, LTD.
[ms)#
\
Court Reporters 1025 Connecticut' Avenue, NW, Suite 1014 Washington, D.C. 20036 l
( (202) 842-0034 I
L_.. __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _
389 1 ~at low pH can attack zirc alloy.
l3)
L_J 2 We have considered crevice corrosion potential 3 with oxic waters in contact with a more active material.
4 For example, if the zirc alloy is up against another 5 material which could be subject to crevice corrosion, that 6 can.give you low pH concentrated ionic strength conditions 7 down in the crevice which might then pit the zirc alloy.
8 We don't have good data on that, and we have not 9 had the time to develop detailed models. What we did was 10 leverage off of some of the other corrosion resistant 11 modeling' work that was discussed yesterday by Joe Farmer to 12 try and take a similar type approach in terms of time 13 history of failure for the zirc alloy by localized 14 corrosion. And this has given us a fairly uncertain free
( ). 15 parameter that we've used in some of the sensitivity studies 16 at how aggressively do we allow localized corrosion over a 17 long time period to break or fail all of the zirc alloy 18 which did not fail by one of the previous methods. Use that 19- as a way of ramping down the zirc alloy credit after all of 20 those processes that we had any mechanistic basis for have 21 left some of it still. intact.
22 That's about all I was going to say about the zirc 23 alloy corrosion model. I will show an example of the 24 sensitivity calculation toward the end.
25 I'm going to move on to spent fuel dissolution.
> ; ANN RILEY & ASSOCIATES, LTD.
' ./ Court Reporters l 1025 Connecticut Avenue, NW,-Suite 1014 Washington, D.C. 20036 (202) 842-0034 u
l l:
390 l 1 The objective of the effort over the last year and a half l
l f'}
2' was to update our bounding intrinsic dissolution rate model, v
l' 3 and we wanted to develop a model that was a function of t
l 4 temperature burnup, pH carbonate, and the amount of oxygen i 5 in the vapor phase. This'was done. As it was discussed l
6 earlier, we ended up with a multi-parameter equation, fitted 7 against the available data.
8 I'm going to back up a little bit and discuss the 9 -- this is -- again, it's a -- it's called a " bounding 10 intrinsic dissolution rate model" for the UO2 matrix. And 11 this is not intended to represent the evolution of secondary 12 phases'. It's. intended to look at what is the upper bound 13 for the aqueous driven dissolution of the UO2 matrix. So it 14 is fitted to high flow rate tests where the water's flow i
q 15 rate is presumed to be high enough that you do not get j
16 secondary phases. 'You do not get these protective layers on 17 the surface of the UO2. And from that perspective, again, 18 it's an upper bound on the dissolution rate.
19 All right. So we ended up with this fairly 20 complex fit where we have an intrinsic dissolution rate, 21 which is what we're after, and then we have temperature 22 terms, some chemistry terms, a burnup term which lumps 23 together a variety of sins, including the radiation effects 24 that were discussed earlier, and cross terms, where we 25 thought- that there was some evidence in experimental dat - nr
.; T ANN RILEY & ASSOCIATES, LTD.
',/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
391 1 in our understanding of the mechanisms why you should have 2 the potential for.a cross term between these parameters.
(
3 We took all of the available data for both UO2 and 4 spent fuel'within the' program, the available data, and did a Then you go in and see how important some
~
5 regression fit.
6 oof the terms are, and some of them are terribly important.
7- Then in the analysis for TSPAVA - this is the 8- fifth that we had asked for. We asked for these parameters 9' as a starting point, and we got them. Then in the TSPA 10 analysis, we ended up holding burnup constant, holding oxygen constant.
~12 You heard some discussion yesterday that we have 13 some variation in carbonate an pH in a few steps. It's a --
14L 'it's not a continuous distribution in chemistry, but we have
() 15 several steps of water chemistry coming into the waste form 16 cell that Abe Van Luik discussed yesterday. And we have the 17 temperature histories, which are continuous.
18 So, as you keep several of these fixed, very 19 rapidly, this collapses into just a few effective 20 parameters.
i 21 Dave, you asked earlier how do you simplify this f 22 Land remove some of these from a parametric fit, and you can 23' do it a couple of; ways.
241 One isJt o fix these and recalculate, or you can
- ;25 remove them and then redo the regression. We've tried
$[l' . ANN RILEY'& ASSOCIATES, LTD.
?s /- '
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 o
- y (202) 842-0034 1
)
wi i a. i im, r-i. . s e,i ip- . . ..,..,2.
.,,,..i. , . . . .
,. ,,,,,,i, ,,,,,i
392 1 looking at that a couple of different ways, now that we've ja j 2 fixed a few of these parameters, and the difference is 3 typically less than a factor of two, whether you take our 4 average numbers and do the calculation, or if you re-regress 5 using a subset that removes some of the data points that are 6 .not relevant to the reduced parametric fit.
7 DR. SHOESMITH: Could I ask a point of 8 clarification?
9 MR. HALSEY: Sure.
10 DR. SHOESMITH: Is that process then telling you 11 that those particular terms were within the 12 irreproduceability of the data set? They don't change it 13 very much?
14- MR. HALSEY: I can't tell you whether they're in G
-( ) 15 the irreproduceability of the data set or whether the 16 parametric dependence was small enough that it's -- that the 17 dependence itself is within that irreproduceability.
18 There's a slight difference in those --
19 [ Discussion off the record.)
20 MR. HALSEY: At least one of *. hem did. I'm not 21 sure if they all.did. I'd have to go back and look.
22 So the way we have it implemented right now is we 23 have this equation left in-there, but we have many of these 24 terms fixed at this time. Okay. In other words, the base l 25 case was evaluated with an average burnup, fixed 02, and l
O ANN RILEY & ASSOCIATES, LTD.
jk / Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 S4 - b3 L
L _ _ _ _ _ _ _ . _ _ - _ - - - _ - _ - _ _ _ _ _
393 1 some limited variation in the water chemistry.
/ 2 I guess this is a good point to point out that
\s) 3 with the analyses going out to a million years, the time 4 steps are typically in the range of a thousand years. They 5 can be 300 years or a thousand years in the early times, and 6 with the assumption of aqueous processes, as'soon as we are 7 exposed to the drift environment, this bounding dissolution 8 rate, which is for high flow rate experiments, does give us 9 rapid dissolution of the fuel in less than a thousand years.
-10 What we -- we do know that at low flow rates and 11 with just vapor phase exposure, you will get the processes 12' that Dave was talking about. You will get secondary phases.
13 You will get some protection of the surface, some limitation 14 of the surface area. And we do not have those details in at
'O
) 15 the moment.
16 I'll talk about some of the approaches to gain 17 some of that credit back a little bit later, but, indeed, 18 the combination of those assumptions and this type of 19 bounding high flow rate model have given us, you were 20 saying, it's about a two order of magnitude difference in 21 how rapidly -- you were saying how rapidly radionucli6es are 22 dissolved. It's actually how rapidly radionuclides are 23 . mobilized.
24 We believe that, under aqueous conditions, the 25 fuel could still react as fast, but most of it stays in h
LA- l ANN RILEY & ASSOCIATES, LTD.
Court Reporters
.)
1025 Connecticut Avenue,'NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 4
394 1 place at secondary phases. It would react that fast if'the
] )' 2 full surface area'is active that we're assuming.
3 I would.also point out that we applied this rate 4 onto a surface area which is assumed to be the one that Dave 5: was talking-about, of the geometric surface area, plus the 6- surface area from cracking of the fuel pellets, and opening l
7 up of the top several grain layers, which is the 15 times i
l 8 the' geometric surface area, evidenced by experimental data, 9 but, again, that's high flow rate experimental data.
10 If, indeed, at the low flow rates or in just vapor 11 hydration, if you do get the protected layer, as was l 12 suggested, that you do not get that large a surface area 13- active. -Right there is more than an order of magnitude that 14 we're not getting credit for.
() 15 Once we have made radionuclides potentially 11-6 available for mobilization, we then mobilize them, all of 17 those, except the highly _ soluble ones, at a solubility 18 limit. So most radionuclides are released into EDS 19 transport processes at a solubility limit. They're sampled 20 over a range with a minimum, maximum, average and a 21 probability distribution function.
22 In the current base case, the solubilities, except 23- for neptunium, are the same that were used in TSPA 90, 1995.
24_ They are not all the same distribution function, and these 25; were developed from a mix of experimental data and expert i
I ANN RILEY & ASSOCIATES, LTD.
L '} Court Reporters 1025 Connecticut Avenue,.NW, Suite 1014 Washington, D.C. 20036-(202) 842-0034 ,
l l
- + .
.b
395 1 judgment in a expert elicitation, and have been published in f
L 2 '95.
3 Several of those solubilities during the 4 development of TSPAVA were identified and set up in a 5 priority list for review and possible revision. Neptunium 6 was top on the priority list, and it has been reviewed, and 7 the solubility ranges from 95 have been reduced.by two 8 orders of magnitude.
9 That was one of the charts that Dave showed.
10 All right. We did have an activity to look into 11 this and try and develop models because we knew that it 12 occurs, and we knew that it would probably allow us to 13' reduce what is a unnecessary conservativism.in our current 14 representation, particularly once we went to the assumption r
~( h) 15 of aqueous conditions on exposure to the drift environment.
16 So what we were trying to do is look at the 17 experimental work, primarily on low drift, high drift and 18 vapor only tests.
19 The flow rates in these tests are many orders of 20 magnitudes less than those in the high flow rate tests the 21 dissolution rate model was developed from and are more in 22 the range of the water flow conditions that we would expect j 23 in the repository. They are, however, very complex 24 experiments to understand.
25 And what we have looked at is the aqueous t
\ ANN RILEY & ASSOCIATES, LTD.
h"L-!~'/
Court Reporters !
1025 Connecticut' Avenue, NW, Suite 1014 Washington, D.C. 20036
.(202).842-0034 L
r j
L .396 l
l' . concentrations _that come off of this thin, slow-moving water 2 film that drips off of these samples. We've-taken two 3 approaches. One is by looking at the radionuclides 4 . concentrations in the catch and trying to calculate back to 5' what that meant in terms of aqueous concentrations in the
'6 film,'and we have also done some reactive transport modeling 7- of the secondary phase evolution to see if -- how close we .
j 8- can come to calculating some of.these observed values, and 9 these tests are ongoing. So what we're looking at is-f10 . interim test results. I'll show you a few of those things.
- 11 The current status of it is we are not using the 12 secondary phase retention in the base case. We have done 13 some preliminary sensitivity studies.
14 We'll dwell on this a little bit because I do r
( . 15 believe that it's a direction that we are going to continue 16 to pursue, and hopefully, in the future, we will have an 17 approved basis for using these experiments, some modeling to 18 get a more realistic mobilization rate for the 19 radionuclides.
20 Okay. Has provided some initial radionuclides 21 concentrations that are in balance with the thin surface 22- films. These concentrations combine both dissolved and 23 transportable particulate species, such as colloids And
- 24. .one.of the1 encouraging things is'where we have a low flow 25 rate experiment, which is showing us a combination of .
iO t ANN RILEY'& ASSOCIATES, LTD.
's-) . Court Reporters-1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036-
~
(202) 842-0034
397 1~ dissolved' radionuclides and colloidal radionuclides, is that
]f 2- they are still below the solubility numbers that we have 3 been assuming for some of the key radionuclides.
4 ,The concentrations are-independent of the flow 5 rate and the surface area,-the way that they're extracted.
6 They are.not necessarily chemical equilibrium values, as the 7 : radionuclides can be~ held up in substitutional solid 8 solution and the secondary phases.
9 The' goal is that these observed concentrations 10' might be usable in place of aqueous solubilities to
- Li' represent the transfer functions, and.through the complex 12: processes that occur on the surface of the spent fuel, and 13 ' additional modeling is needed to understand what happens as 14 you go -- as the original UO2 matrix disappears, and as the
() 15 -- some of the phases that form progress through the 16 ferrogenic sequence of tertiary and further phases.
17 The bottom line is, why we're doing it is for 18 several important radionuclides. The values that are 19' observed are significantly lower than the current solubility 20 ranges.
21 I would also point out that we, a couple of slides
.22 back, were reevaluating the solubilities as we have the time
- 23. to get to them, and the data base to get to them. And one 24- of the things.that we're trying to do is reevaluate 25 solubilities controlled by,the secondary; phases that are
"['I ' ANN.RILEY & ASSOCIATES, LTD.
Th > .
. Court' Reporters m -1025 Connecticut Avenue,.NW, Suite 1014
? Washington,-D.C.,20036 (202) 842-0034 J.
iliL__G -
+
398 1 seen and/or are expected with both experimental data and l
- 2) some modeling efforts.
.J
'3' When we have sufficient understanding,.this effort 4 _to predict what comes out of the secondary phases and the-
!5 ~ experimental evidence for observation of the aqueous 6; concentrations coming off them should converge with the 7 calculatio.s and measurements of solubilities -- aqueous 8 solubilities -- controlled by the secondary phases. And SL .when those finally converge together on the same numbers, we
'10 should have realistic mobilization rates that.ke'believe
'11 will be-several orders of magnitude lower than what we're
-12 currently using.
13: Preliminary data from interim results off of the 14 ilow drip rate tests that are gone for six radionuclides.
I "15 This is this chart. A little complex at first, but for each 16 radionuclides, the three bars on the right-hand side are the l 17 311 cited solubility range used in TSPAVA. This is the 18 solubility minimum, maximum and average.
19 And then these are two of the different fuels.
20 .This is two fuels, ATM 103 and 106 in low drip rate tests.
-21 And thf.s is the observed aqueous concentrations.
22 Over here, for neptunium, the observed 235 concentration coming out of those experiments is five to six 24- orders of. magnitude below the solubility range that.we're l
- 25: ; currently.using in the analysis. .That's one of the reasons ij[ h : ANN RILEY & ASSOCIATES, LTD.
i M (c Court Reporters
-1025 Connecticut Avenue,' NW,: Suite 1014 Washington, D.C. 20036
'(202) 842-0034 e _=_ _-__ _ _. \
399 1- why.we're. pursuing:this, and as we. understand it better,.we-O y,
2 will try and use these to gain back some of that 3 conservativism that Dave was-accusing us of, quite justly,
'4 earlier,-
5' Technetium also appears to be releasedLat less 6 than what we would say its solubility is. Juni whether that 7 ris -- we need a more precise measurement as=to the actual 8 reaction rate, surface area and reaction rate, to know
- 9. whether o'r.not there is any retention of technetium in the 10 secondary phases in the experiments.
12 One encouraging thing of this is the elicited 13 range, even for uranium solubility, add almost six orders of 14 magnitude variance in it -- variation in it, which, if we
-f) 15 know the solubility of anything, it should be uranium, I 11 6 would hope, off of this list. And the fact that we have six 17 . orders of magnitude range in its solubility is so*t of a 18 testament to what we haven't figured out yet.
19 But the fact that the -- in developing the expert 20 elicitation, they were keeping in mind a variety of 21.- experimental results which had resulted in that range, m: 22 including some spent fuel dissolution studies. And the fact "23- that the observed aqueous concentrations from the two j24 ; experiments bracket the lower end of that range at least 25 :gives=us.some confidence that they were thinking about these i
I'! ANN'RILEY & ASSOCIATES,-LTD.
4%- .
-Court' Reporters e1025. Connecticut Avenue, NW, Suite 1014 L .
Washington, D.C. 20036
'( '(202)- 842-0034 s.
(n- .. ~
l 400 1- types of experiments:when they did the elicitation.
2' I'm not going to dwell on.that anymore. That's.
]}
3 work-in progress, but it is -- does indicate a direction 4 that I think the program will' continue to move.
l 5 Defense high-level-waste glass dissolution. .Over 15 the last year and a half, we wanted to update the glass 7- dissolution model to give us, again, a intrinsic dissolution 8 rate, as a function of temperature, water chemistry, water 9- contact, and the extent of vapor hydration prior to liquid 10 water contact.
- 11 Since we.are now assuming that we get an aqueous 12 -film on the surface, we did not pursue the vapor hydration l-13 as a separate model. We may still want to do that,.but at 14 the moment, we simply assume aqueous conditions on exposure
(') 15' to the environment.
16 We did end up with a revised dissolution rate 17 model, which is an extension of the ones that have been used 18 previously, and it's -- then the radionuclides mobilization 19 is again bounded by the alteration rate and solubility 20 limits, except plutonium, which we are assuming not to be 21 solubility limited, due to the formation of colloits in the 22 glass -- at the glass surface.
23' And in the low flow rate experiments on defense 24 high level waste glass, a significant fraction of the 25 plutonium.is seen in_a colloid form. So we're not giving it j] ANN RILEY & ASSOCIATES, LTD.
M Court Reporters 1025EConnecticutLAvenue, NW, Suite 1014 Washington, D.C. 20036' (202) 842-0034 x_ _ _ _ _ _
V p
p 401-
~
'1 a' solubility limit.-
' ' ~
~
- 2. ' This revised dissolution. rate is currently?in use r3: in.the base" case,.and.like I say, we're not' currently-using 4- lthe:vaporfhydration model. LThat . model is . f airly f amiliar :to 5 .thosefin the glass dissolution business.
6 We have'now a temperature dependent; long-term.
7 : rate,-which is probably the most'significant addition ~to 8- thislmodelLover the. previous.one. It's a function'of'the:
l- 9~ rsurface area, a rate constantlwhich is measured t-
- 10 experimentally. It depends upon'the concentration of-til dissolved silica in the. water, so this can be used as y'ou 12 .
build.up. silica to the content to reduce.the reaction; rate, s
13 and thenLat':long times, it ends up settling into a long-term 14- ' rate.
.h Q 515' I'lm not going:to dwell on that.
~16 Let me touch on what'are some of the residual 17- uncertainties in the waste form models, the active area 18- t continues to be the active area of the waste form. That 19 ,.
gets into some of the things that Dave was discussing this t20 morning.
l21~ To what extent, at low flow rates, or in vapor 22 - hydration experiments, do the secondary. phases protect the 23 fsurface andLkeep it from opening up a large. number of active
' 24- . grains?
- 25: ' A t t h e .. m o m e n t ,~ ,.we're; assuming that we'do get a
.s >
Ijb ANN RILEY.& ASSOCIATES,- LTD.
b./S V Court Reporters.
1025 Connecticut Avenue,:NW, Suite:1014-M, ,
_ L..
LWashington,LD;C.,20036
,m.
+ S
- (202)-842-0034 'l l s ', ,
'-:. , . 4 g. , .'
- 4
..- M. . . . *2 .
m.
. fj
402
-1 fair amount of grain boundary opening, but not -- we do not i
(; 2. assume that it all falls apart to grain boundary areas-for 3 the entire fuel mass-.
4 What is the~ effective surface area of the 5 secondary phase? That is one that we're going to have to 6 address. The cladding performance model, what we~have is a 7 preliminary model based primarily on literature data, and 8- there are several processes at very long times that we need
-9 to continue to address.
10 Spent fuel' dissolution. We need to continue to
- 11. look at the interdependence of some of these process 12 parameters represented by the cross terms, and see if we 13; really don't need them, then we can eliminate them once and 14 'for all.
(G)
~
15 We'd also like to get in a little additional 16 chemistry such as the silica and calcium, but then we're 17 starting to go away from an intrinsic dissolution rate model 18 -into something that represents the evolution of secondary 19 phases and the mobilization of radionuclides from that.
20 We don't believe that the calcium and silicon are 21- intrinsically involved in the dissolution process at the UO2 22 surface as the carbonate can be. I don't know that some of
- 23. the other people agree with that, but to put those l 24 chemistries into this model, it would no longer be an 251 ' intrinsic dissolution rate of the UO2. Then it-would~ start L <
!I ) ANN.RILEY & ASSOCIATES,,LTD.
( -OL / 2 Court Reporters
( , -1025 Connecticut Avenue,r NW, Suite 1014 1 Washington, D.C. 20036-(202) 842-0034-V h
e- .= , -
403 l' to",be a secondary. phase' evolution model,.and we haven.'t
'2 .
gotten.there yet.
3- Rapid release' fractions. We do need to review
.4~ 'these. The .ones. thati were provided by David in that
.5 .' reference to.che recent report that they've done I think are
~6. new and better than the ones that-we've used in the past.
7 Howe.ver, the way we're currently modeling release in the 8 .EBS, that doesn't make a significant difference. We would 9' like.to get=more of.the~ time details in the release process..
~
110 :into the model in the future, at which point this would' E11 become1 import ant again.
12 Solubility ranges as elicited have a wide range, 13 .and we believe that.many of them are conservative. We will 14 continue to address those.
- .15 And as I've dwelled on,.I guess, at least enough,
-16 : .but.the potential for retention of radionuclides in 17 secondary' phases, we want to pursue that.
18 Colloid generation'and mobilization. We have an 19- initial representation that I'm not really going to talk 20': about a lot'where we mobilize plutonium both from spent fuel
.21[ ~ and'from glass as colloids, and then we have some transport
.22 processes in the EVS which can use either assumptions of
-23' reversibility or. irreversible colloids. We need to expand 24: ,theidata base.and-realism in those models.
, 25 And'onellast.one that I' haven'.t. discussed is N
[h^N ANN RILEY & ASSOCIATES, LTD.
. Court. Reporters-L1025l Connecticut' Avenue,<NW,! Suite 1014 Washington,;D.Cn 20036 4 <
(202) 842-00341 i
"b'. -.sf '- ' -
_-i
404
-1 diffusive transport may significantly delay the releases t
) 2> from the waste' forms. The way we're currently modeling the 3~ ' mobilization and the initial transport is we assume that 4 there is, since.we have this water film on the surface of
- 5. .the waste forms, we also assume that it is in rapid enough 6 diffusive transport to an' advective flow pathway that within 17 a time step we're not limited by'the diffusive couple. We 8 ihave started doing~some sensitivity studies, putting in 9 various amounts of diffusive transport delay, and there is-11 0 'the potential for a couple orders of magnitude reduction in 11 radionuclides concentrations in the advective flow paths from 12 that as well.
13 Those are more details that we can put in using 14 the new capabilities of RIP, and we haven't had the time to 15 do so yet.
16 Some of the sensitivity studies we've done -- wrap 17 this up fairly quickly -- we have done credit versus no 18 credit on the cladding and differing levels of cladding 19 performance. We have ctarted doing some sensitivity studies 20 on the secondary phases. We've got some numbers like I 21 showed'on the chart using observed concentrations from one 22 set of experiments. And we've also done some reactive X 23- transport modeling with a couple of different sets of 24 assumptions regarding retention of radionuclides in the 25' ' secondary and. tertiary phases.
N ANN RILEY & ASSOCIATES, LTD.
b/ _ (Court Reporters
-1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 i j (202) 84290034 F
1 I
o_ __ u - - __ m _ _ -
l 405l l 1 Let's see, I brought one of these sensitivity (p,) 2 studies and this one, and one of these. That's the v
3 diffusive versus advective release, and that's using the i
4 capabilities in the RIP multicell EBS to represent diffusion l
5 through the altered waste forms.
l 6 Some of those sensitivity studies that we haven't 7 'done yet, some more things on the dissolution rate to look 8 at the larger ranges in the EBS environment and look at some 9 of these differences in simplification of that model. When 10 I said we've tried simplifying that model a couple of 11 different ways, what we haven't done then is gone in and 12 plugged those numbers in to the TSPA to find out if they 13 have any significant contribution. The dissolution rates 14 themselves were all fairly close, so we haven't pursued
<^x
( ) 15 that. We do want to look at some different effective 16 surface area, particularly with the secondary phase 17 evolution.
18 And then we're also going to be looking at or we 19 hope to look at some design alternatives in terms of 20 backfill, alternative designs, thermal loadings, and 21 inventories.
22 All right. Here's the example, a couple of 23 examples of some of the sensitivity studies. This is one of 24 cladding. Here is the base case with the cladding credit, 25 and then here's a couple of curves with differing amounts of
) ANN RILEY & ASSOCIATES, LTD.
s/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
406 1 reduction in that credit, including no cladding credit, T 2 which is the top curve, and then using the localized
-[Q 3 corrosion model to arbitrarily force failure of all of the 4 cladding by 100,000 or 1 mf.llion years giving curves here 5 and another one that's right on top of this one. So we're 6 looking at order-of-magnitude type variations in the results 7 with or without cladding. We may also do some with a little 8 more aggressive cladding credit in some of the early times, 9 which might give us a little more separation.
10 DR. APTED: Bill, point of clarification. When
- 11. you say failed by that time, does that mean it's failing all 12 the way up to that time? Not at that time.
13 DR ., HALSEY: Yes, it's not a step failure.
14 DR. APTED: Okay. Thanks.
U .
15 DR. HALSEY: It's using the preliminary models 16 that we've got, but forcing more aggressive localized 17 . corrosion which results in all of the cladding failing by 18 certain times.
19 The fluctuations in this, for those of you that 20 haven't looked at TSPA results, are due to climate change 21 that's put into the analysis. This would be -- these would 22 be pluvial events and these would be -- this I believe is 23 the super pluvial, which is cimilar to a new Ice Age. So 24 those glitches are significant changes in climate and water 25~ flow.
ANN RILEY & ASSOCIATES, LTD.
f[-\~/') Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
,. l 407 i
1 i Secondary phase sensitivity. This'is using the n-2l reactive _ transport modeling, retention of neptunium in the
~
3: I secondary phases,'trying to use a model that represents the I !
4 l ~ numbers that are seen in the Argonne tests, and turn that i
5l into a concentration history -- and-this'is just neptuniur i
6 3 dose, this is the base case is this line here, and when we
- 7. put in some credit for the secondary phases, we get one to 8 two orders of magnitude reduction in neptunium out to 9_ several hundred thousand years.
10 This is a different type of sensitivity analysis.
i
- ll ! This is not a total dose up to a population. This is 12 - release rate from the EBS, and this is some of the initial 13 work on advective versus diffusive transport through the 14 secondary phases and the altering waste forms of just 15 technetium and neptunium. So there's one highly soluble 16 - radionuclides, this is technetium, assuming that you're 17 rapidly coupled into an advective flow path, and this is 18 with a diffusive transport cell put in representing the 19- altered phases or the secondary phases. And this is 20- neptunium with all advective transport, and this is with a
- 21 - diffusive delay. Sc we're looking at about an order of 22 - magnitude reduction in radionuclides concentrations from the 23 l diffusive representation there.
24 So in summary:
'25 We-have had a variety of. improvements in the waste ANN RILEY & ASSOCIATES, LTD.
> :) Court Reporters 1025. Connecticut Avenue, NW,' Suite 1014
' Washington, D.C. 20036
'(202): 842-0034
_ _ _ . _ _ _ _ - - - - - _ i
408 1 form degradation and mobilization models.
'f).
V
- 2 We have a preliminary cladding credit model that 3' we are'using.
4' Evolutions above the spent fuel and glass 5 dissolution.
6 We have an initial representation of colloid 7 mobilization at transport through'the EBS.
8 And we have some new alternative conceptual 9 approaches, secondary phases, and also the diffusive 10 transport limits.
11 That's sort of the status of what we're doing at' 12 the moment.
13 DR. WYMER: Are there questions, comments?
14 Charles.
(
( ),) 15 MR. FAIRHURST: Yes. This may be a little bit --
16 off the center. You mentioned you had a mechanical 17 disruption, failure model. Would you say a little bit about 18 that, because I'm interested in whether the impact is really i 1
-19 significant.
- 20. DR. HALSEY: I could say something about it.
21 There was a long list of different cladding failure 22 processes, and that's one that we put together a few months )
23 ago. Kevin McCoy from the waste package design group was 24 responsible for most of that, taking rockfall -- there's two 25 primary areas that we thought of for mechanical disruption.
.[\- , ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025, Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
409 1 One would be things falling on top of a degraded waste t i 2 package, and the other would be the slow settling of as the L.J .
3 waste package and its internal support-structure is 4 degraded, so you'll get bending and disru,"cion of the nice SL array of spent fuel elements that we always see in the 6 pictures.
7 The first one that we modeled was the impacts. We 8 haven't gotten to the bending moments and things.
9 MR. FAIRHURST: We could just assume some impact.
10 DR. HALSEY: Right.
11 MR. FAIRHURST: You didn't go to the mechanics of 12 how the impact would occur.
13 .DR. HALSEY: Well, they went through -- there are 14 some estimates as to the size of rocks, rockfalls that are
((~% ,) 15 credible for the drifts, and you assume that at long times 16 the waste package has degraded to the point where a rock can ,
I 17 fall and hit some of the spent fuel, and that was basically {
18 taking different areas of rock as a punch with different 19 amounts of kinetic energy behind it, and then taking a set 20 of assemblies and allowing this punch to hit it and how much 21 of it breaks, and how far down can a credible rockfall 22 damage the fuel.
22 MR. FAIRHURST: I don't want to go into too much, 24 but did it take account of stress concentrations at welds :
I 25- and places'like that?'
i l
, ~') ANN RILEY & ASSOCIATES, LTD. !
N_/ .
Court Reporters:
1025 Connecticut Avenue, NW, Suite 1014
-Washington, D.C. 20036 '
(202) 842-0034 l
l j 410 1 DR. HALSEY: I don't believe so.
[V3- 2 MR. FAIRHURST: 'All right. Thank you.
3 DR. HALSEY: I would refer that to Kevin.
4 DR. WYMER: One more question.
5 Joe?
6 DR. PAYER: David did a nice job of showing all 7 the inputs that go into the EVS and transport it and how 8 that, you know, couples this together. Can you in a simple 9 way, and maybe not a simple way, I still can't -- how does 10 the configuration of the drift, if it's collapsed, if 11 there's rubble all over the canisters, for example, at these 12 various long time frames, how does that condition work into 13 the modeling? And then this other issue of what are the 14 penetrations, shapes, sizes, distributions look like? How g
(],/ 15 are those difficult issues dealt with?
16 DR. HALSEY: If you're talking primarily for the 17 waste package, I could take the easy way out and defer that 18 to Joon Lee, but we assume that there is rubblization of the 19 drift, the concrete liner. This will put debris onto the 20 waste packages, giving us localized corrosion sites and 21 water wicking sites. That's factored into how aggressive 22 are the environmental conditions.
23 We also have water trickling down through this.
24 The corrosion model has -- you get fairly large corrosion 25 areas for the carbon steel, and a pitting factor which gives
,\
ANN RILEY & ASSOCIATES, LTD.
)
(
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
411 1 us fairly large, low-aspect ratio areas of exposure of the
[^')
V 2 corrosion-resistant material underneath. Then you get 3 localized penetration of that as the -- as you get 4 sufficient number of localized penetrations, the waste 5 package is divided up into patches, and when you get 6 sufficient penetration, numbers of penetrations of the 7 corrosion-resistant material, then a patch is assumed to be 8 open, and you transition from small area openings to a 9 larger area opening, several square centimeters.
10 Dave, do you want to elaborate on that at all?
11 DR. PAYER: That's a level of detail beyond what 12 I'm really trying to get at. I guess what I'm trying to get 13 a feel for is I believe you said that you're dealing with 14 all advective flow in films. You haven't plugged in the
( ) 15 diffusive flow yet. And there's a, you know, realistically 26 that could slow things down.
17 Is the fact you're doing that then, does that also 18 say that it's.not important in the current TSPA as far as 19 how these are getting out, that the tortuous path, does it 20 make it longer if it has to go through rubble, or are you --
21 DR. HALSEY: Yes, it would, and that's --
22- DR. PAYER: And it's already built in, or that's 23 something that --
24 DR. HALSEY: No , that's something we have to put 25 in. Right now in the -- I didn't even bring slides on the fO. ANN RILEY & ASSOCIATES, LTD.
E/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036 l (202) 842-0034
412 1 EBS transport portion of it, but there is a cell in there 2 for the invert, for example, because the radionuclides have j%s}.
3- to cransport out through the invert, and we assume a 4 porosity and tortuosity in that. But it's primarily an 5 advective flow through that.
6 We can put in a diffusive transport portion to 7 that when we have a basis for it.
8 DR. WYMER: You have probably heard enough 9 chemistry. Thank you very much.
~10 The next speaker represents the NRC modeling 11 treatment of chemistries, Dr. Brett Leslie.
12 Before we get started, can we get any indication 13 if Las Vegas is on the line? Can you talk to us?
14 Apparently not.
. f3 15 If Vegas is there, please talk to us.
()
16 SPEAKER: We're here.
17 DR. WYMER: Okay. Thank you.
18 Go ahead, Brett.
19 DR. LESLIE: Good morning, my name is Brett Leslie 20 and I am a geoscientist and the team leader for the 21 evolution of the near-field geochemical environment team
.22 here at the NRC. And although I am.the presenter today, I ;
23 .really need to point out that the work that I am presenter 24 is'a result.of'an extremely close collaboration between 25 .several teams and team members. Representing the near-field 2('
s_/
I: ANN RILEY & ASSOCIATES, LTD. I Court Reporters 1025 Connecticut Avenua, NW, Suite 1014
' Washington,'D.C. 20036 4 (202) 842-0034 l l
i
__._____-.____-_-Q
E. ,
l-413 1 :atLthe Center is Bill Murphy, who you heard speak O')'
%j 2- previously. 'We also are working very closely with the l- 3- container life and source term key technical issue team and 4 we have Tae Ahn'from the NRC staff and Dr. Sridhar from the
- 5 4 Center representing that. And last,'and certainly not least 6 importantly is performance assessment. Dick Codell from the 7 NRC staff is a member of that the near-field-team and.the a performance assessment team, and we have Sidocam Mohanty l ~
9 down at the Center helping us out in terms of release 10 models.
F11 Se'cond slide. I'll provide a general interview of 12 what I really want to talk about today. I want to spend at
,13 least a little bit of time talking about the NRC's overall 14: approach and after that I-will spend a lot of time our total-( .- 15 = system performance assessment waste form models.
16 I will next spend a little bit of time talking 17 about some of the solubility considerations that have been 18 brought in Bill Halsey's talk and David Shoesmith's talk.
19 I'll be talking about the engineered barrier systems 20 interactions. This is a new module that has been added 21 since we last met in April to address their EBS transport 22 module that they didn't really discuss in Bill Halsey's 23 talk,.but it was in response to Joe Payer's, where we 24 . evaluating the impact of potential chemical interactions in
'25: invert; I -
. ANN RILEY & ASSOCIATES, LTD.
\_- -- .-Court Reporters 1025 Connecticut-Avenue, NW, Suite.1014 Washington, D.C. 20036 (202) .842-0034
4 J
414 1 I would spend some time talking about a comparison
() 2 3
between both the DOE and the NRC's approach and, finally, provide some conclusions.
4 Before I move on, I would say that I have added a 5 few slides in response to Dr. Wymer, Garrick's and 6 Steindler's responses yesterday on where does this meet 7 performance. And I think I will try to tie in some of the 8 previous presentations on container life, but also on our 9 waste form models. Why are we studying what are doing? Why 10 are we putting in the models that we are putting in? I 11 think you got a good idea from Dr. Shoesmith's presentation 12 where two orders of magnitude difference in release rate.
13 What I hope to get across to you by the end of 14 this talk is that the NRC's approach I think follows Dr.
() 15 Jackson's exhortations yesterday of being timely, efficient 16 and effective. In addition, I think we are also very 17 integrated, integrated and iterative. Again, the three 18 teams are constantly falling over each other trying to help 19 each other out, which is good. But I also want to bring out 20 in this talk that there is a distinct interplay between the 21 process level modelings and the TPA level sensitivity 22 studies and the overall system sensitivity studies. Without 23 the process level models we don't have a very good basis for 24 determining what are the ranges, and so I want to bring that 25 out.
O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
415 1 I think you will realize as I go on that our PA
() 2 3
approach is continually being enhanced, and we are modifying the code, and some of the information I am presenting today 4 will show that.
5 The majority of the talk will be talking about how 6 the NRC envisions radionuclides release from the EBS, and not 7 just the waste form. Like DOE, we are assuming a congruent 8 dissolution. We are only assuming spent fuel waste form.
9 And we use what is called the bathtub model. I think this 10 is an example of a flexible model, this is just a brief 11 schematic, where we have a hole at the top and a hole in the 12 side. These can change variable. If the hole was right at 13 the bottom you would have a flow-through model. So it's a 14 function of how we might envision the corrosion. If we have (h 15 pinhole leaks we might have the bathtub. If we have uniform 16 corrosion and broad -- we might just have a flow-through.
17 So this is an example of an approach to the corrosion that 18 allows us some flexibility in evaluating performance.
19 Again, we have a bathtub or flow-through 20 configuration. Like the DOE, we also have some sort of 21 solubility constraints applied on the radionuclides once 22 they are released from the waste form.
23 And, finally, I'll spend some time, a couple of 24 slides, on the transfer function approach that we have just 25 recently incorporated into TPA 3.2 to abstract potential O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
416 l' other EBS material--interactions.
2 A' couple of. cautions in notes. First, use of a
<3? particular approach or model.or parameter in TPA 3.2'should 7-4 not be construed.as regulatory acceptance. Part'of.that is
,5 =because'our insights and assertions are preliminary. We are 6' -- just like DOE is constantly working.a parameter,.a model 7' of development,'it's ongoing and continuing. And'some of 8 the. outputs I am going to talk about are based upon very 9-- limited analyses.
10- So let's get to the main meat of the talk.
11 . Radionuclides release from waste form in TPA 3.2. Again, 12 -reemphasizing we are only using the spent fuel form. We are 13 treating release.in four options as alternative conceptual I
- 14 models. And this fuel dissolution in the presence of l
(Gj :15 carbonate is what DOE is considering for their base case j i
16 We have three other options that we are evaluating .l i
17 and will be evaluating. Fuel dissolution in the presence of
- 18. calcium and silica-which Dr. Shoesmith suggested was 100 19 times lower. We have derived an equation to use in the 20 performance assessment to evaluate the impact on dose. We ;
21 have also added something that you heard about for the first 22 time in April, which is the constant rate of release. I'll 23 go into a little bit more about that, what sort of insights 24' that might provide us.
L25 And, finally,'since April we have added another L.
ANN RILEY.& ASSOCIATES, LTD.
[ ('s ~
Court Reporters l
1025. Connecticut Avenue, NW,' Suite 1014 p' Washington, D.C. 20036 (202)- 842-0034' 3 i
r__ . _ _ _ _ - _ _ _ . _ _ _ _ _ _ _
417 1 one, which is a secondary mineral dissolution model based
()' 2 3
upon schoepite, and this is primarily the work, collaborative work, of Bill Murphy and Dick Codell. All of l
4 these, in terms of their impact on performance will be l
5 evaluated and are being presented in two MRS papers where we 6 are looking at what is the response to the overall system l
7 performance assessment.
8 Before I get into each of the different conceptual 9 models I want to list a few of the common assumptions for !
t 10 each of these models. No colloidal release, tima-in variant l l
11 chemistry. What I mean is that, conceptually, the chemistry ;
12 doesn't change. As we saw earlier, the pH and carbonate 13 could change as a function of time and temperature, but what l
14 I am saying is that when we choose a conceptual model, we !
r
( 15 say calcium and silica are always absent, that's the only 16 way we can evaluate these.
37 We have a bathtub model that allows for 18 flow-through calculations. We assume that the radionuclides 19 are released at the rate of fuel dissolution. We do have 20 implied solubility constraints on the radionuclides once 21 they are released from the fuel. We have included an option 22 for evaluating credit for cladding since the DOE is going 23 forth in the VA space evaluating cladding credit. We have 24 that option in our code to evaluate the potential impact on 25 performance.
( ANN RILEY & ASSOCIATES, LTD.
\s / Court' Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034-
418 1 Again,rwe release the: radionuclides to the 2' ' interior of the waste: package. We assume a vective. release-3' 'from the waste package, and we assume an oxidizing 4 environment. .And I think the information presented by many-5 Jof-the; presenters would really emphasize that'it is an-E6 . oxidizing environment.
7 The first model is what we consider our carbonate 8l ~ dissolution model, but.it is not just carbonate dissolution.
.9 .It'is in the absence of calcium and silica-and J-13 water (10 and'all the waters indicate, even in the thermal single
-~ ~ ~
heater test there Y abundant calcium and silica in the
~
11 12 water, so this model is really dependent upon not having 13 calciuu and silica present, otherwise the release would be
- 14. Emuch lower.
l 15 For each of the models I am going to.be talking 16 about I will talk about what is the basis for forming these.
17- model,'the conceptual model, how is it incorporated into 18 performance assessment, some of the additional assumptions 19 that we have in addition to the previous list that I gave, 20 and what I think are some of the reactions that would impact choice of that alternate conceptual model and what could 22- control the variables in the equation.
i 23 Now for the carbonate dissolution model, Bill 24: Halsey and David Shoesmith did a very good job of explaining L25 .that these are based upon the flow-through tests of Wilson, i
l' ;
j Lc ,
ANN RILEY & ASSOCIATES, LTD.
D - //
Court Reporters-L 1025; Connecticut' Avenue, NW, Suite-1014-L ,
Washington, D.C. 20036 (202): 842-0034 L ,
419 1 Gray and Wilson,-1995. 'The way the'NRC Staff has
() -3 2' incorporated this into a PA code is that we have it on the
' total carbonate concentration P02 temperature and pH.
4 Some of the assumptions of.this model are.that no 5- calcium and silica are present and no secondary. mineral
.6 ' formation and a constant surface area as we have now 7 envisioned it in our PA model.
8' I will, wait to~ discuss the actual performance-9 after'I have describedLeach of the four models, but in L10 general, as Dr. Shoesmith suggested, carbonate dissolution 11 model is -- as Bill Halsey suggested as well -- bounding,
- 12. overly conservative.
13L I would point out that this is not the model that 14 we assumed for our base case.
() 15 For the near-field geochemical reactions,
- 16. Lcertainly CO2'and the gas chemistry is. going to be affected 17 by volatilization and the thermal pulse. We should derive, 18 in fact we see high CO2 concentrations in the gas in the 19 drift scale tests, also indicative of the single heater 20 tests where.they had low pHs and they argued that was a 21 -reaction. That is much more likely that's it's simple l
22 'devolatilization of CO2.
23 So this will be variable. It is a function of 24 coupled processes, thermo, hydro, chemical' processes.
-25 One of the big things you need to consider is the
/ ; ANN RILEY'& ASSOCIATES, LTD.
4 LL A . .
Court Reporters
? . 1025' Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 #
(202) 842-0034
420 1 . interaction with cementitious-materials. It is going to
[') 2 . affect both carbonate,and pH and I think DOE has attempted
- L/
-3 and has made very good progress in trying to come to grips 4 with how to:take into account the input of a lot of cement 5 into.this repository.
6 -In addition,-the. concentrations and acceptance of 17 this model assumes or could be impacted by reactions with.
8 the metal of the container corrosion products and those 9 could influence both the gas and water chemistry, and we 10 . heard a little bit about how the corrosion processes could 11 affect pH but also the mineralogy could affect it.
12 The next model is our Tae Ahn developed and we 13 have incorporated in our code. This is based upon the 14 laboratory batch tests of Wilson (1990). Very limited data.
f\ 15 In essence, this model assumes that calcium and
( )
16 silica are dissolving in J-13 type of water. It has a 17 slight temperature dependence. This ranges from the 18 uncertainty in the measurements of the experiments.
19 Again,. as an alternative conceptual model, it 20- assumes that calcium and silica are always present at 21 approximately J-13 values. In fact, we see that this rate 22 is a function of calcium and silica concentrations -- at 23 least there is some indication of that in those experiments.
24: Also, this model currently, as we entail it in our 251 performance assessment', is a function of constant surface h
l-ANN'RILEY & ASSOCIATES, LTD.
YN / Court Reporters 1025 Connecticut' Avenue, NW, Suite.1014
.f' Washington, D.C. 20036 L (202) 842-0034 c _ _ _ _ - _ _ _ - _ _ _ _ - _ _ _ _ -
421
~1 area. Again, there are reactions in the near-field as the
' ~
/m
'l ) 2 near-field-evolves that will influence this. We certainly
\. /
3 could have a reaction with fuel that could reduce calcium 4 and silica concentration, leading to a secondary formation 5 of secondary minerals leading to formation of secondary l 6 minerals such as Schoepite, and I will get into this a 7 little bit more when I talk about Model 4, but we see this 8 in the laboratory, in the drift tests.
9 In addition, interaction with cementitious 10 materials is certainly going to affect calcium and silica.
11 The concrete is highly reactive in the tuffacious 12 environment and also reaction with metal could influence the 13 water chemistry. Sometimes corrosion of metal containers 14 can cause formation of iron silicates, so we might be also
() 15 influencing the silica concentration.
16 The third model that we have included in our TPA 17 Version 3.2 is a fixed-release rate model, and the advantage 18 of this is you can evaluate how the performance of the 19 repository would go given any release rate. We have chosen 20 to use this based upon either empirical or natural analog 21 data. You were presented some tentative results in April on
-22 the natural analog data. Bill Murphy again is the primary
'23 mover of this and we have worked hard on incorporating this 24 into TPA 3.2 but Tae Ahn is also evaluating whether we can 25 use the drip test results to fix our release rate, so it is l
ANN RILEY & ASSOCIATES, LTD.
f [\ /)
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
, (202) 842-0034
422 1; not Just a. release rate based upon natural analog. It'is a
../)
-L fixed release rate. .If you have'some basis for evaluating
.%)
3, it then you can go forward'.
4 'Some of.the results that I will present in'a 5 'little-bit are on' Pena Blanca natural analog that Bill 6' Murphy presented. The' constraints at that site are based 7 .upon the mass of the ore body, radiometric dating of the 8 . period of oxidation and alteration, and that the release is 9~ controlled by-secondary uranium mineral solubility controls 10 the overall. release.
11 We can use those constraints in determining what 12 the maximum average value for uraninite alteration at Pena 13 Blanca. You can calculate that. We can scale that to what 1-4 'the Yucca Mountain repository is of spent fuel and it yields 15 '- .a. lower limit on the release from the proposed repository.
16 Some of the assumptions that go along with this 17 model;and incorporation of the data to support this model is 18 that the.long-term release is controlled by the secondary 19 uranium.
20 This model is not dependent upon surface area. It 21 -is based upon bulk observation and an important constraint 22' is the release'of other radionuclides proportional to 23 . uranium release, and both this model and the fourth model _
L l24 are dependent upon a lack of information. Right now we just 25
- incorporateieverything in it. What Bill Halsey suggested l
a, J : ANN.RILEY & ASSOCIATES, LTD.
I Court-Reporters 11025' Connecticut' Avenue, NW, Suite 1014
' Washington, D.C. 20036
- (202), 842-0034' o
i__.-.---_.. _
L 423 l
1 and Dr. Shoesmith suggested is there is a lot of information In\ 2 that could be attained to better support these approaches,
. O l
3 i.e., how much of the radionuclides are taken up into the l
4 secondary phases and how stable are those phases, and the 5 release of the other radionuclides are proportional to 6 uranium release.
7 Finally, again if you use a fixed rate based upon 8 a natural analog you can always argue, well how analogous is 9 Pena Blanca to the repository? We may not have had quite as
.10 much iron or cement at Pena Blanca -- certainly no cement, 11 but there is a fair bit of iron at Pena Blanca.
12 The last model, which has been developed over the
- 13. last six weeks and now was incorporated just last week our 14 TPA 3.2 model is a Schoepite model, and it is based upon the D)
(_ 15 short-term vapor phase and drip test observations, 16 information from our studies and natural analogs at Pena ,
l 17 Blanca, and thermodynamic studies that Bill Murphy 18 summarized in 1997 in an MRS paper, 19 In essence, it assumes that solubility of uranium 20 is controlled by Schoepite. The release, as we have put it 21 into the PA code, is a function of temperature, pH and total 22 carbonate. This is because Schoepite exhibits some 23 retrograde solubility.
24 The mineralogy of Schoepite is this. This is 25 important because what it is - .it's just hydration of the
. ,-) '
! ANN'RILEY & ASSOCIATES, LTD.
\/ Court Reporters 1025' Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034 .
I
_ -_ _ __ L
424 1 UO2.with absence of cations and so.let me spend a brief r~ l 2 moment on why -- what happens.
(v) 3 In the vapor tests, what they see are the complex 4- calcium silicate uranal minerals such as uranopbane forming 5 on the top of the surface of the fuel, but as you go down 6 the reaction path, you only get Schoepite, and so in effect 7 you are scavenging your cations. You are left with water.
8 The water reacts with the fuel, forming Schoepite.
9 Some of the assumptions are that Schoepite would
-10 remain as the dominant secondary mineral throughout the 11 period of your assessment. Again, dissolution is not 12 dependent upon the surface area. It is solubility-limited.
13 Again the way we have done it so far is that we 14 have rapid uptake of other radionuclides by Schoepite. This
/"
(%) 15 can be relaxed somewhat because we have an instantaneous or 16 gap pulse release, so we could adjust this as we go along in l
17 the next iteration of performance assessment.
l l 18 Finally, again release of all the radionuclides 19 are proportional to the uranium release. Again the 20 short-term tests and the Pena Blanca stuff may not be L 21 exactly analogous to the Yucca Mountain facility, the large l
! 22 amount of metal and cementitious may affect whether this is 23 the stable phase.
24 To address some of the concerns on how and why we 25 are doing these, let's look at the impact on performance.
l l
ANN RILEY & ASSOCIATES, LTD.
[)/-
\_ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 j i (202) 842-0034 !
p l
425 II This is aLslide that Di'ck Codell presented-to you back in L() 2:
3 April, and basically I want you to focus on base 625 and Natan down here.
E L 4' 'This'is.again an earlier version-of our TPA-Code 5 3.1 point something, at which time we were assuming 625 as 6 the container base' case. This describes what the base case 7 is -- no backfill, no matrix diffusion, the. bathtub model, 8 no cladding, but the point is that'the natural analog is
~
- 9. lower than the base case, assuming -- and again our base 10 case was not what DOE assumes as their base case, which we-di - ' feel is overly conservative perhaps, but is based upon the
.12 calcium and silica model with those present.
L 13 A better look at this results is a slide that Bill 14 Murphy did not present but had in his package. This is j
) 15- again preliminary output based upon limited results. These 16 are the CCDF of the base in bs and ns for the natural i
17 analog. Again, alternate conceptual models -- the model i 18 stays the same throughout the period of performance, 50,000 19 year max, 10,000 year max.
20 This suggests that even the natural analog or a 21 model three is maybe a factor of 10 lower than the calcium 22 :and silicate model. This might change, as Bill Halsey l
23 suggested, because the surface area might change, so our ]
24 ' calcium silicate model might in fact more closely reflect
.25-' what.the natural analog model looks like.
)
[/.
L ANN RILEY & ASSOCIATES, LTD.
js- Court Reporters 10251 Connecticut: Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) '842-0034
r 426 1 Very, preliminary work, not taking it out to total
() 2 3
system performance, because we just put the coding. changes intlast. week, the'schoepite:model' suggests-that the release-4 might be, oh ,1 100 cn 300 times _ lower.than the net trianalog 5 release. rate.
'6 So these are the. reasons why we are evaluating.
7 There are' orders of magnitude difference in these different
- 8. mosels.
9 One of the questions that was brought up 10 yesterday, and I'll switch gears just slightly for a second, 111. but I think it bears.on why we are evaluating what we are 12 evaluating. I have heard a lot about how much water drips.
13 This is based upon our system levels, this is also a Dick 14 Codell slide. As you see, 10,000 years, 400 vectors, 3.1 15 base case with seismicity. These are the factors that are 16 most important to performance. And these top three -- the 17 top three are an effect of how much water comes in, 18 FowFactor FMult.
19 This is -- we really need to understand the 20 dripping. Okay. How much water gets into this. The spent 21' fuel wetted fraction is -- how high is the bathtub if it's a 22 bathtub configuration, it plays a big impact on that.
23 And, finally, Sridhar was explaining from a 24 material science perspective, and a process level 25- . engineering perspective, why we are concerned about ANN RILEY & ASSOCIATES, LTD.
[~
v- ) Court Reporters 1025' Connecticut Avenue, NW, Suite 11014 Washington, D.C.'20036-
-(202) 842-0034-K-.-_-L-_ _
427.
1 chloride ~. Well, chloride multiplication factor. How much 2- evaporation actually occurs strongly. influences the overall-
)
-3 performance.
.4 To' switch gears a little bit and talk about how.we 5 are approaching radionuclides solubilities. We have some
~6: basic assumptions. 1The solubilities are chosen for-7, oxidative Yucca Mountain conditions. We have assumed the
- 8 solubilities are controlled by pure mineral phases. I would 9- indicate that these are derived from a poorly documented' 10 expert elicitation and there may be ways around, bolstering
' 11' the values chosen or evaluating those values chosen.
12 We have done what we consider some process level 13 modeling to evaluate what solubilities are reasonable. Bill 14 Murphy, in his evaluation of TSPA-95 solubilities, looked at
, 15 a range of chemistries at Yucca Mountain from the pore water 16 saturated zone and evaluated what the solubilities would be 17 using'EQ3, and his preliminary conclusion was that the 18 values used in TSPA-95 for a range of chemistries were 19 conservative.
20 We right now have process level studies going on.
21 .Th'ese are similar to the types of studies that were done for 22- the: site 94-Swiss program, Arthur and Apted did, looking at 23 reacted waters. What if we have waters reacted with
'24 1 container, how does that change the chemistry? Does that.
25' affectfour solubilities? Those are ongoing right now. They '
b'
- Dw/
ANN RILEY &~ ASSOCIATES,'LTD.
. Court Reporters .
<1025 Connecticut Avenue,,NW, Suite 1014
' Washington,' D.C. 20036
-(202) 842-0034
428 1 'are very early. I-don't know if:I have any results. I j } 2' .would have.to talk to Bill after this talk.
- 3l 'Also, one of the things we were.looking at, and, 4 :again, Bill Murphy is one of the people doing this, and is 5 ' working with' Dick Codell, is evaluating a correlated
- 6. ' solubility approach. The chemistry is involved in here and 7 the solubility shouldn't necessarily be independent, but 8 could be dependent.
-9 Why are we concerned about this for solubilities?
10 The doses are sensitive to solubilities, especially the 11 neptunium solubility.
12 And I would reiterate what Bill Halsey suggested 13 'and, more importantly, Dave Shoesmith suggested, is that how 14 we approach solubilities in our release models needs to be
() 15 considered. This range of solubilities has to be consistent 16 with the model you are using for release. And let me 17 explain that a little bit.
18 This is a slide that you have seen a couple of 19 times, once in Dr. Shoesmith's talk. Again, the TSPA-95 20 range is way up here, primarily at the upper end by Heino 21 Nitsche's experiments. What DOE is doing for VA is down 22 here. They drop the top in two orders of magnitude. We 23 have dropped our upper limit by one order of magnitude. I l
24 think we have a reason for this. The reason is that this
[ .25 upper limit is~ based upon.a FARDS study in 1997 where they
- h. L ANN RILEY & ASSOCIATES, LTD..
H4- Court Reporters h 1025 Connecticut Avenue, NW, Suite 1014
. Washington, D.C. 20036
'(202) 842-0034-L C z_ _
429 1- : reversed the solubilities and it's, in fact, very tightly l
() f2'
- f
. constrained. So if'you are going"to say that the
- solubilitiesLare controlled by secondary -- by. pure 4 . minerals, not the secondary minerals, then.the limit, the-
- 5. upper limit should probably.be up here.
6 -What you see, again, is that the lower-7 solubilities or mobilities being released from these drip i
8 tests are quite low. If you are going to use a secondary l 9 mineral release model, then it doesn't really make sense to
- 10 have solubility limits way up here because it's counter to I' 11 your assumption of what the model is.
12 So in choosing the appropriate range for 13 solubilities, you have to keep in mind what are your release 14 models. You know, are you going against what your j' 15 assumptions are?
- 16. Again, a new thing since April. Because we need 17 to be in the position to evaluate the DOE TSPA-VA, and they 18 have a separate module which you didn't really hear much 19 about, which'is the EBS transport module, which is, in 20 effect, the invert. So if they had a way of evaluating 21- . ret'ardation in the invert, if they had some magical material 22 down there, we would now have a way to evaluate this.
123 The approach we have taken, again, is to have 24' Etlexibility, because we are likely to see a. multitude of 25 designs and changefwith time. It-should allow us to b)
A K/ .
ANN RILEY & ASSOCIATES, LTD.
. Court Reporters
~1025. Connecticut Avenue,::NW,: Suite 1014 Washington, D.C. 20036
.(202) 842-0034
__ _ _ _ _ =
430.
1- , evaluate sorption or co-precipitation.
2 We have not -- we don't a design-specific approach
)
13' -in TPA 3.2, but it.should be sufficient to evaluate.the base L.
[
l 4 case. And here, .again, is where we are evaluating this both
- 5 at the system level and in the-process level. We are
~
6 looking.at multi-flow in terms of interactions. What might 7' it do to porosity and permeabil'ity, absorption?. And also, 8 we'will evaluate what those ranges are in-our sensitivity-l 9' studies._
10 We might look at what are some metal' specific, 11 i.e., if you had that large corrosion fluff of the container 12 there, and DOE attempts to take any credit for that in one
'13 .of their sensitivity studies, we would be able to evaluate-14 that by changing.the KD. Or cementitious-specific. Cement
(~%
.( ,7 15 .isn't necessarily bad to performance overall, it just 16 depends upon where you are in the flowpath. And we would be L 17 checking out alkaline solubilities via alternate solubility 18- limits,.as I suggested Bill Murphy is doing.
19 An example of this, again, this is Dick Codell's 20 work coming in. This is just a simple example, this isn't 21 in-the performance, overall performance. This is a transfer 22 . function. Basically, just a pulse of iodine. We can assign 23 properties to this transfer function to mimic either the
. 24' thickness and diffusivity or KD of that invert and it will
-25 just allow us evaluate what DOE is doing.
c; - '1001 RILEY & ASSOCIATES, LTD.
Ek- . .
Court Reporters 1025 Connecticut, Avenue,'NW, Suite'1014 Washington, D.C. 20036 (202) 842-0034
431
~1. ~Some'more insights on what our EBS transfer
'2
~
function' approach.is. We are using it to model mass 3 transport in the near-fieldfoutside of the container, 4- because, again, our release is inside'the container,
- S advective out of the container. We have, again, recently s6 added this to allow some flexibility. It is an option that 7 is user chosen, it can be turned off or on. One of the 8 restrictions ~is it requires a time invariant' flux of water.
9 And whether we need to make this more complicated, because 10 our overall flux of water is not time-invariant, will be
.11 based upon just doing some sensitivity studies.
12 We have adopted a very simple approach to try to 13 capture the main conceptual details, and if we need to 14 increase the complexity of this model, we will. If 15 performance suggests it. But it will require is appropriate 16 radionuclides specific input values. And DOE is doing some 17 experiments on corrosion products and sorption where they 18 'have suggested that-they might take some credit, at least in 19 sensitivity studies.
20 Now, as the last spea:cer, I get one job, which is 21 to summarize a comparison of the approaches. First off, the 22 NRC is only evaluated in the spent fuel wante form. As you 23 heard from Bill Halsey, we are doing both the spent' fuel --
24 DOE is doing both the spent fuel and the glass waste form.
L 25 Currently', in;3.2~.we will not have colloids. ,
We are 1.
[ Ic\
E ANN RILEr &. ASSOCIATES, LTD, NM Court Reporters 1025 Connecticut. Avenue, NW, Suite 1014-2 84 b3 L
432 1 anticipating incorporating that into the next version of the 2 PA code. We are already starting to think about how we 3' could do this. For the VA, DOE will have colloids.
4 One aspect that we need to keep in mind is that we 5 are primarily all fracture flow with no retardation, so we 6 might not be too bad in terms of plutonium, but we will 7 evaluate that more in the future.
8 I think in terms of the release models they are 9 relying on the intrinsic dissolution rate in the presence of 10 carbonate with no calcium and silica. They also have done 11 some preliminary work on secondary mineral release. We have 12 a nice presentation where they have done some reactive 13 transport to bolster some of the net flow analog stuff.
14 We have four alternate conceptual models. The
() 15 ones -- our base case is the calcium and silica. I would 16 like to point out again that we have taken an overly l
l 17 conservative approach. Overly conservative might be up 18 here.
l 19 We have a constant release rate and schoepite 20 solubility rate dissolution model. Our solubilities for the 21 base case we think are conservative. But we think they are 22 flexible for the type of model you are going to choose for 23 analysis. You should keep that in mind, you shouldn't just 24 keep a constant solubility if you are going to choose a 25 model that is not consistent with your assumptions.
!/ ANN RILEY & ASSOCIATES, LTD.
'ss Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l.e (202) 842-0034
___1 _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ __
l_
433 1 Again, neptunium is ten times lower than the NRC's, and it may affect dose. Our model is a transfer.
(A)
N-- -
2 3 function and we will be doing sensitivity studies that 4 assumes advection only. The EBS approach of the DOE allows 5 . advective diffusion and retardation.
6 So, I'll finish up early I think. I would like to 7 conclude, I think I have shown that our program is 8 integrative, iterative. The modeling with radionuclides 9 release and container life, and release from the EBS, we are 10 using process level analysis to help guide some of our 11 sensitivity studies and to support some of the abstractions 12 that we have done. Continuous controlled modification of 13 the TPA code, it is under NQA. As we go along, we can 14 carefully document how we have modified the code.
r~n
-(,) 15 Timely, I think. Two months turn-around, the 16 different model. I think maybe if the ACNW had us in here 17 every month we might cut our time down to three weeks for 18 every model, I don't know. Well, I didn't -- just kidding, 19 Dr. Garrick, 20 The assumption of time-invariant requires analysis 21 of radionuclides release using alternate conceptual models.
22 We are really looking at what is the expected range of what 23 we might see. Support for these alternate conceptual models 12-4 are limited right now. I think that we have little data to 25 support all the radionuclides are taken up in the secondary T ANN RILEY & ASSOCIATES, LTD.
.- [
k- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
, 434
~
-1 minerals and released at'the rate of schoepite. I
~
2.- ,
There is'some preliminary work, Edgar Burt and Pat i3; . Finn, show,'. yeah,-neptunium and we'see these retention 4' factors that they say. But we need a lot more information 5 to be"able to' support that in the quantitative performance.
6 . assessment. Some short duration laboratory tests, as Dr.
7 Shoesmith suggested, we'will need longer term tests to 8 evaluate the potential thing. I would say that there are 9 very few experiments to evaluate EBS interactions and their 10- impact on performance. Perhaps some of this information 11 will come from heated drift where they have a fair bit of-12 -studies going on on'the impact of cementitious materials 13 they will be pouring later on. They are also doing some 14 -laboratory experiments.
O 5 ,j 15- So1with.that, I would conclude and be happy to
? 16; answer any questions you might have at all.
17 DR. WYMER: Thank you. And thank you for getting 18 us back on our new schedule.
19 Mick?
, -20 DR. APTED: Nick Apted. Two questions about your 121 second to the last line, your presentations of different 22 assumptions and models, and first as a comment.
23 Certainly, there are other groups out there, such
- 24 as the group at Berkeley and Epry and other people developed 125 -these type of models for this,-and'it might be prudent to at
) ANN RILEY'& ASSOCIATES, LTD.
's Court' Reporters
'1025 Connecticut Avenue, NW, Suite 1014 (Washington,-D.C. 20036 (202) 842-0034'
< --s
435 1- -least look at what'other people have done in the same area,
() 2 3
because'there may be some advantages and' disadvantages,~and
.if nothing more,_ comparing alterative conceptual models.
'4' .The second is a question, which is, are you now 5 engaged in some way of comparing -- what I worry about is 1 6 that there'is so much difference in both data bases, l
7 conceptua.' models, solution methods between your codes.
8 -When you come to compare results, how do you focus 9 Lon the differences in the data base and the difference in 10 the boundary conditions, or how are you going to come to
- 11 that resclution?
'12 DR. LESLIE: A part of that is how our PA code has 13 been constructed, is that we don't just get those out. We 14: get a lot of intermediate outputs. And it's, you know, we
( 15- can look at what the release is from the unsaturated zone.
16 We can look at the release from the EBS-transfer function.
17 We can'look at the re3 ease from the waste camp.
18 There are a lot of intermediate outputs that will 19 allow us to more directly compare some of the results that 20 the DOE have, and others.
21' I think your point on looking at other approaches 22 to the waste form or other conceptual models is a good one.
23 My primary focus for this talk was just looking -- comparing 24 DOE and'NRC.
25 DR. WYMER: . George?
.l ANN RILEY & ASSOCIATES, LTD.
L. . Court Reporters 4 1025 Connecticut-Avenue, NW,. Suite 1014 .)
Washington,,D.C. 20036
.(202) _842-0034
436 1 MR. HORNBERGER': Yeah. I donft know if this 2-( }( question is for you,. Brad, or for perhaps equally;for 3 several of the other people, but'since you're on the spot, 4 I'll address.it to you.
5' We listened to Bill Murphy yesterday present this 6 wonderful' array of complexity and things we don't know. And 7.. Rod Ewing,:at break, pointed out to me that just for
~
8 : u ranium,' we had.the thermodynamic data base for perhaps five 9 out of several hundred of potential minerals.
10 But listening to Dave and to you and to Bill this 11- morning, I get the distinct impression that there is a good 12 deal'of confidence that we can do very sensible bounding 13 analysis in the face of the ignorance that Bill pointed out 14 to us.
() 15 Is that a safe assumption?
16 DR. LESLIE: I think it's a factor of two things.
17 It depends.on which level you're looking at it.
18 For the detail process level, as you get closer 19 and closer to the actual' reactions and kinetics, the 20 information appears to be limited. I think as you go to an 21 ' abstracted model, you need less data to try to support that.
22- But I think it's really the integration of the 23 overall long-term, you know, looking at laboratory tests, 24 looking at natural analog and discreet modeling, I think
- 25. there is a conversion. And so,.you know, I don't think it's fh K /'
ANN RILEY &' ASSOCIATES, LTD.
Court Reporters L -10251 Connecticut Avenue, NW, Suit'e.1014 l Washington, D;C. 20036.
l (202) 842-0034
=_ _ _ _ _ _ . --
437
- 1 as bad as. Bill! presented, but I-think Bill was presenting it
.;\ 2' -atfthe process level.
3 And I think Peter Lichtner said the same thing. I 4 mean we're.never' going to know this at the process level, I
'5' mean all the details. But, again, the question is whether 6 ~it's important at the process level or at the system level.
7 DR. SHOESMITH: Of course, I think it's separable 8 'into two-areas. I think we understand the fuel. I think we 9 understand it fairly well, and I think we can place very 10 good - .put down very good numbers to how we think it will
-11 dissolve, and which particular parameters affect it.
12- The next step of what kind of phases you will get 13 is not.so certain, and we are relying on empirical numbers 14 and a natural record -- geological record is, indeed,
( ) 15' extremely complex. I think it's resolvable, but not at the 16 present time, and it;will never be. resolvable in the same 17 detail'that nature has reproduced, but I still think it's 18- reasonable to expect that something will happen there. But
'19 .that's fairly weak right now.
- 20. DR. LESLIE: Yeah.
21 DR'. SHOESMITH: Fuel, we know. Secondary phases, !
22 we don't.
/ 23 DR. LESLIE: Yeah. I think I tried to point that l
24 out, as well. .Thank'you.
25 MR. HALSEY: I'd address that a little bit, also, L
j ANN'RILEY'& ASSOCIATES, LTD.
's Court Reporters
- ! 1 0 2 5 C o n n e'c t i c u t A v e n u e , IDf , Suite 1014 h Washington, D.C. 20036 (202) 842-0034-p p
-438 l~ :1; plus:,a'little_ bit of mixed' question'.
~2. .It's encouraging that'different organizations are f-3 takingtslightly different approaches, but coming to'some of 4' the same conclusions. One of-the things.that the panel 5 ' asked' yesterday is,..what have we learned that we data need V
h 6_ to pursue further, versus where are we getting guidance of 7- what is the high priority or high leverage
~
8 And we should probably not look1at.the' bottom-line
-5L dose rates from TSPA,~from the two organizations, places 10 where they're very similar is probably fruituous, because 11' ' when you-look at the details of all the submodels that went 12 into that, they're quite a bit different.
.13 -What you should look at is where they're both 14 : agreeing on the sensitivities. Where is the high leverage?
15 And I think.we are starting to see some of that.
16 And it's the sensitivity analyses coming from 17 Tdifferent approaches that are giving the same answers as to 18 whether something is important or is not important that I 19 think are one of the great values here.
'20 DR. WYMER: Well, in order to -- thank you. In 21 order to stay reasonably on our new schedule, we probably.
22 ought to proceed.
2 31 We're going to shift gears rather dramatically 24! "here:and'go into the engineering and construction 25 . considerations a little bit. And to do this, we're going to
' ANN RILEY & ASSOCIATES, LTD.
~ ~-[
Court. Reporters 1025, Connecticut Avenue, NW,. Suite 1014 Washington, D.C. 20036-
- (202).842-0034.
h_._i_1 ii__...'____Z--_
.439.
1- have a presentation on the engineering, development, 2 fabrication and materials selection for Yucca Mountain by
([
3 Dave Staehle, who:is -- represents the -- the contractor at 4 .. Yucca Mountain. -And if we're -- so you want to introduce 5 .this, or do you want to --
6' DR. LESLIE: Yeah. I walked away with it.
7 DR. WYMER: There will be a short delay.
8' [ Discussion off the record.]
9 DR. WYMER: I'd like the opportunity.to talk a 10 little bit about material selection and have Jerry Cogar of
- 11 our Engineering Development lead -- address the issue of our 12 .w elding and fabrication development work, and he'll do that 13 .from Las Vegas, 14 I wanted to reiterate some of the history of 1 15! materials testing. Yesterday, we talked about the fact that 16- we had a design evolution. We initially started, as you 17 know, with.a thin wall waste package design that was a bored 18 hole in place.
19- We did corrosion studies at Lawrence Livermore Lab 20 in the 1980's, and culminated in a material selection that 21 was peer reviewed externally, and actually, Bill Halsey was
~
22 the lead in that peer review.
23 As a result of that peer review, we selected three (
'24 materials.
25' I don't need that on, now.
ANN RILEY.& ASSOCIATES, LTD. 1 L [~'
i N .
. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C '20036 (202) 842-0034:
L
'440 l' 'One'of the materials was the alloy ~825. -The other l'
() .' 2 3
two materials was - .weretalloy.C4.and titanium grade-12,.
which is, interesting because'we're still considering those
-4 materials'in the program.
5 At any rate, alloy 825 was chosen on the. basis of 6 its wide:use and'its lower cost. Unfortunately, materials 7_ testing was put'on the back burner during that period of 8 time,. and the project emphasis, as probably most of you 9~ know, shifted to surface and surface based studies.
10 It wasn't until the M and O came on board in 1992 L :11~ that we reexamined the waste package design.and restarted-l 12 the waste package materials testing and modeling programs.
13 We had a shift in emphasis, as many of you know, from hoist 14 to_ lamps, from small to large waste packages, from j( k 15 consideration of a thin wall to a multi-barrier design.
16 Dimensions were roughly established by myself and 17_ Tom Dorr'ing. We came up with the design that we have now, 18 .which is roughly a ten centimeter carbon steel outer 19 barrier, and the two centimeter corrosion resistant 20 material.
- 21. Now, if you want to put that chart up. I don't 22 necessarily need to show it here.
23 But just to reiterate some of the points that:Dr.
24 Farmer made yesterday in regard to the material, the outer
,25 ' barrier certainly is amenable to handling. We felt that
( . ANN RILEY & ASSOCIATES,.LTD.
A - Court Reporters
- 1025-Connecticut Avenue, NW, Suite 1014
. Washington, D.C.,20036
~(202) : 842-0034 m
441 1 this was an appropriate way to handle the package in the
(
- 3 2 repository, loading it in the surface facility, and placing
%. J' 3 -it in the drift. It does give you that shielding 4 requirement, at least partially shielded. The rest of the 5 shielding is taken care of by the transporter.
6 As you can see, it also has the electrochemical 7 potential aspects that Joe talked about.
8 As far as the thin wall corrosion resistant 9 material, as I indicated, we started out with 825. We then 10 moved, as we learned more about the corrosion resistant of 11- those materials, and as shown yesterday in the pitting 12 studies, we moved from the 825 to 625 in the 1996, 1997 time 13 frame, and then more recently, we baselined in January, 1998 14 the use of alloy C22.
- (/'hj e 15 We always carried the alloy C22 in the corrosion 16 test program, but we didn't consider that material in the 17 earlier period because we felt that there were some concerns 18 in regard to weldability of that material. Those have been 19 resolved through the effort of Jerry Cogar and his staff.
20 So let me move on. Currently -- concurrently, I 21 should say, with the corrosion testing effort, we've had an 22 engineering development effort looking at welding and 23 fabrication. And, hopefully, we can get Jerry to pick up 24 and talk about some of those efforts.
25 Jerry?-
IL j -- ANN RILEY & ASSOCIATES, LTD.
l ;- / Court Reporters
! 1025 Connecticut Avenue, NW, Suite 1014
' Washington, D.C. 20036 (202) 842-0034l l
ll N _ _ _
442 1 -MR. COGAR: We.are having a little difficulty with
. s~
( 2 the machine, Dave.
3 DR. WYMER: You're on', now.
4 MR. COGAR: Can you hear me?
5 DR. WYMER: Yes.
.6. MR.-COGAR: Okay. We can't seem to get the slides 7 to come up out.here. Sorry. But we can basically talk to 8 .the' fabrication process, if that will do.
9 DR. WYMER: Go ahead,;please.
10 MR. COGAR: Can you hear it?
'11 DR. WYMER: Yes.
'12'
~
MR. COGAR: Okay'. We can't get theLslides to come 13- up to be able to show you any of the pictures. .
But, 14 ' basically, the fabrication process amounts to deforming the I) 15 outer barrier out of carbon steel, approximately 100 16- millimeters thick It's a rolled and welded plate, with a 17 long seam in-the middle, formed in two pieces, and a 18 circular seam in the middle. This rolled and welded carbon
.19 steel plate then sees a stress relief, radiographic 20 inspection of all the wall seams, zul ultrasonic inspection 21 of all the wall seams, and a magnetic particle inspection of
- 22) .all'the' wall seams.
23- .The inner barrier is formed in basically the same 24' way. 'It's currently alloy C22 plate, rolled up in two 25 ' pieces, with'a long-seam weld in each one of those, and then L
h3 ik--C ANN RILEY & ASSOCIATES, LTD.
Court Reporters L 1025 Connecticut: Avenue, NW, Suite 1014' l
iWashington,-D.C.-20036-(202) 842-0034 e
l:
F >o . _ _ _ _ . _ _ . _ _ _ _ _ . . _ _ _ _ _ _ _ _ _ _
443 e
1: joined with;a circular seam.-
[ : 2' All of these welds the ultrasonic, radiographic hb)-
3 and di-penetrant inspection.
4- Once.these two cylinders are formed, the outside 5 'ofLthe inner barrier and the'inside of the outer barrier.are 6 then. machined with approximately a 70,000ths interference 7 -dimension.
4 8 The outer barrier is heated to approxinstely six 9 to 700 degrees. fahrenheit, and the inner barrier is then J10 ' dropped in, and both pieces are allowed to cool.
J 11 L 'Once these are put together, the lower plates or lids are put on. The outer barrier lid is 110 millimeters-
.131 thick. The inner barrier is 25 millimeters thick.
14 The inner barrier lid is put on first, welded, 15- UT'd, di-penetrant' inspected. Then the outer barrier lid is
- 16 .put on, welded, ultrasonically inspected, and magnetic 17 particle inspected. Then the whole vessel, at this point in -!
18 time, is slated to be stress released.
19 After that, there's a final machining sequence, as
- 20. well'as any final NDE's to the maximum extent possible. And 21 then the vessel and the upper lids are shipped to the test
?22 site.in the Yucca Mountain.
23 Our. engineering development program in '96 did a
> 24 selection of the processes'for-the-closure. weld The -- we
'25- did a; matrix'of all possible' welding processes. An I , " ANN RILEY & ASSOCIATES, LTD.
k[ .
. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
. Washington, D.C. 20036
.(202) 842-0034-v .s
_____---____-_b
, 444 1 automatic gas tungsten arc weld was picked as a process that 2 floats a waste' package. We did a number of PQ's,and testing 3 on those.
4 In '97, we did a mock up, which simulated the l .5 diameter of the waste package. It was approximately 44 6 inches long, or give or take one-quarter of the waste 7 package.
8 The mock up at the time was carbon steel 8516, and 9 the inner barrier was alloy 625.
10 We warmed the barriers by the shrink-fit method.
11 It'was actually done by Mary McTague in Fabricating in 12 Cleveland, Ohio.
13 We then welded the lids on to prove the process 14 could actually be done on a circular cylinder, and we did
) 15 the UT of that weld. Those were all done, essentially, 16 manually. The UT was done manually. The outer pit, 17 obviously, was automatic, but not remote.
l.
l 18 The development program for this year formed a
.19 similar cylinder, only with alloy C22 on the inside. We 20 did, instead of furnacing the piece to do the shrink fit, we 21- used locally a much lower temperature, resulting in very 22 little' oxidation on the inside of the outer barrier. That
=23 piece has now been formed.in a sample, and it's in l.
24 Lynchburg, Virginia'for a welding of the plates.
L 25' The plates this year will be welded totally l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l' <
(202) 842-0034
l-E 445
.1 ' automatic and remote. The UT will be done remote. That's 2 the major difference this year, as well as the material.
(
3- There are two sets of plates this. year, which gives us two 4 . shots at perfecting the weld parameters.
5 There's a number of other things going on within 6 the program. We are' locking at ways to model the defect, 7 the undetected defect probability in the package. There's a 8 modeling effort going on at Livermore. We're looking into a 9 system called, "Siswelds," which models the residual stress 10 ~ on the waste package' as well as some work here at Las Vegas 11 on that.
12 And that's, basically, very quickly, the 13 engineering development side of the program. ;
14 You got questions? !
15 DR. WYMER: Are there questions?
16 DR. PAYER: Jerry, one of the questions that came 17 up in the context here, in these discussions was the J 18 e'/aluation of welds and how they might affect corrosion 19 processes, reliability of the, you know, the welds,
- 20 mechanical properties, things of that sort.
21 Is there a fairly -- how extensive is the
, 22 evaluation once those prototypes are made?
23 Are'those going to be sliced up for metallurgical 24 analysis and so forth? )
l
, : 25 We have sent.for one, and the '97 mock up-is the
./
'N' ANN RILEY & ASSOCIATES, LTD.
A._, , Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l d i
446
-1 : Lawrence Livermore. Now we're testing theni. What. tests,
~'Y 2 you know, whichever tests they desire to run. Certainly,
[O 3 that's their choice.
4 There is some testing going on on'some of the 5- pieces, but at this. time, I'm not sure of the extent of 6 that. And, certainly, we'will send the one this year to 7 Lawrence Livermore, too. So there's plenty of material 8 there available to do all the tests.
9 DR. WYMER: We have another question. Do you want 10 to step up to the microphone? Identify yourself.
-11 MR. SEALEY: My name'is Roger Sealey. I work for 12 Hanes, International.
13 You mentioned in the' fabrication, after the 14- welding and the shrink fit of the inner and outer, that
() 15 there was going to be a stress relief.
-16 What temperature are you going to do that stress 17 . relief at?
~
18 Maybe that was after the welding of the bottom 19 ' lids, as well?
20, MR. COGAR: Yes. That's true. There's a number 21 of issues here. We're going to try to stress on the mock 22 up, but-there is, from the Nickel Development Institute
-23 meetings and a number of_other things, we still are 24 evaluating.what we need to do with the inner barrier, and q
25 even our stress relief standpoint.
[\ ~ ANN RILEY & ASSOCIATES, LTD.
s l' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034-E
~_ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ - _ _ - _ _ ____ -
p 447 i .The intent would be, normally, to' stress relief-l ;2. the carbon steel at a'model level.to.1150 for -- it leaves
- 3. - .my mindinow'-- something like an outer break,'an inch _of 4 thickness, or whatever the -
- says. I don't remember _right.
5 off the top of my head.
6 But the' inner barrier is the area of concern..
.7 We're.not -- I guess we haven't settled on what the needs 8- :are for that, from a corrosion standpoint. Certainly,.if Snh 9! were doing-just normal manufacturing, we would not need to 10- stress relief that at all.
all There is a workshop being planned by the Nickel 12' Development Institute'here in'Las Vegas. We're tentatively _
13: scheduled for' November, where a number of manufacturers, and 14 certainly, material people will be there- . And, hopefully,
() 15 we will get some -- a lot more data on these issues at that 16 time.
17 MR. SMALEY: Thank you.
'18:
DR. WYMER: John?
19' CHAIRMAN GARRICK: I_think maybe this session 20 deserves one off-the-wall question, and I'm afraid to ask f21 it, since I was accused of that at one time.
-22 I'd like to'get your opinion,' Jerry, on what might
- 23 befdifferent if you had been given.the task'of designing ~and 24 starting with-La clean piece of' paper the million year waste
?25 - _ package ~ system' .
is-y '
. ANN RILEYL'd ASSOCIATES, LTD. .
CourtLReporters 2 . .1025 Connecticut Avenue, NW, Suite 1014
'; -; Washington, D.C. 20036
, 1 (202) 842-0034- j
1 448' 1 MR. COGAR: That's certainly anioff-the-wall 21 ' question.
3 CHAIRMAN GARRICK: Well --
4 MR. COGAR: The guide -- and,'certainly, I would 5' like to help with it, but my, I guess, expertise.and my 6 responsibility;is in the manufacture of whatever is l
l 7 : designed.
8 I guess if I were going to design this myself, it
- 9- 'would not look a lot different from what it is here,-because 110 -my' background,-obviously, is Navy nuclear and commercial til nuclear. So I would design it'to be something like that.
12' I don't pretend to understand the intricacies of a 13 million year package and the material and corrosion l'4 ' properties to go into that, so I don't think I would get t f 1 51 into that. LI would. tend to be more -- to design it more 16 from a manufacturing standpoint, and it would look a great 17 deal like what it does now.
18 CHAIRMAN GARRICK: 'Thank you.
19 DR. WYMER: Are there any other questions?
20 If not, I thank -- well, I have one, I guess.
21 As I understand it, then, the outer container,.
'22 inner container are in intimate contact?
231 MR. COGAR: Yes. Yes, they are.
-24 'IMR. WYMER: Which raises the -- a point that was 25 brought up yesterday.-about-the effects of possible corrosion 2,n . . . ..
f( .
, ANN RILEY &-ASSOCIATES, LTD.
LA '
~ Court-Reporters
, .1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202)H 842-0034 g
_ _ _ _ _ _ - _ _ _ _ = - _ _ _ _ - _ _ _
(449
- 11. of the :inside .of the outer. container pushing on' the -inner 2 container.
'3 Did you have any observations about that?
-4. Do'you.think that will be. diminished because:of
-5 the close contact?-
6 MR. COG.1R : Well, there is certainly an~ issue
~
7 'that's been brought up by a number of different boards or _
- 8, experts, or whatever, different venues that wedging stresses 9 :and a number of other things. And certainly those are~under-10' consideration', as are the materials in the. waste package and L
l 11' a number of~other things.
12 -Right now, the requirement'is that we have 13' ' intimate contact because of mechanical properties. People zl4 tell us they need intimate'. contact, so we manufacture to 15- give them that.
.16 Obviously, there has been a number of proposals 17 for waste packages that have various gaps in them and that 18- float free, or that have something that holds them in place 19 that had much larger gaps.
20; We have also, over the time, looked at welds 21'- cladding, welds deposited overlay clad that put the inner
'22 barrier-in, and a number of other weld on and explosive clad -
23{ Jand all'those things to actually put the contact there in a 24 number-~of different; ways.
25; We have not found any yet that is as antsy --
ANN RILEY~& ASSOCIATES, LTD.
r\ // Court-Reporters
~1025 Connecticut-Avenue,?NW,_ Suite 1014 ~
, Washington, D.C. 20036
-(202)-842-0034; T
~__-_.-.=-_.:...-
450 1 antsy is probably not the word -- but as simple to do as the 2 shrink fit.
(G)'
3' Certainly, if wedging stresses become an issue, 4 then there's going to have to be something else, and 5 obviously, we can decide to -- or we can design a 6 manufacturing sequence that would accommodate those.
7 DR. WYMER: Okay. Well, thanks very much.
8 Are there any additional questions or comments or 9 -- if not, thanks for this insight. I think it was 10 important to us to give a more detailed understanding of how
.11 these two containers work together. Thank you very much.
12 MR. COGAR: Thank you.
13 DR. WYMER: Now.that concludes this morning's 14 presentations, We have an hour and a' half break. We should
,~~.
( ): 15 come back -- well, it's an hour and 20 minutes. We should 16 come back at 1:30 for what I think will be a very important 17 and interesting panel discussion of the key issues and 18 concerns. And then probably we'll have some chance for 19 public participation. So let's break for lunch.
20 [Whereupon, at 12:11 p.m., the meeting was 21 recessed to reconvene at 1:30 p.m., this same day.]
22 23 24 !
I 25 i-. ANN RILEY & ASSOCIATES, LTD.
\2 Court Reporters 1025_ Connecticut Avenue, NW,_ Suite 1014 Washington, D.C. 20036 (202) 842-0034 E
f .'
p E2__ __ __ -----__ _ _ _ - _ - _ _ - -
I -:
t-451
.1 . AFTERNOON SESSION
() 2
- 3. DR. WYMER:
.[1:32 p.m.]
Let's commence-the afternoon part_of
- 4' 'this working group. -This is the fun part where we have the
_5 panel discussion'and then we open up later on to the
- 6 audience for participation, public_ comment.
7 We have_one panel member who has not made a 8 ' presentation, so'we're going to give him a chance to talk
- 9. first. That's Chris Whipple.
10 Oh , let me -- I've been reminded here by John that
-11 I should set the stage, and the stage -- this may: preempt 12 'your, talk, Chris.
- 13' DR. WHIPPLE: I was hoping to start before you 14- told me I had ten minutes.
) 15 DR. WYMER: What we would like you to do, you 16 panel members, is if you will, to the extent possible, focus
~17 your comments on the status of the knowledge base and the
-18 _ credibility, if you will, of the engineered barrier systems 19 and its performance. That's the first point. And the 20 second point is what.else do we need to know in order to
-21 license a repository Preferably what do we need to know in 22 rank order.
23' CHAIRMAN GARRICK: And that's with respect to the 24 engineered barriers.
25L DR. WYMER: So that's a challenge.
ANN RILEY & ASSOCIATES, LTD.
5- . Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
-Washington, D.C. 20036
.(202)- 842-0034'
L 452 1 With that guidance, Chris.
[} 2 DR. WHIPPLE: Am I on? Okay, you can hear me now.
- %/
3 By the way, I've always felt that bad handwriting 4 is important if you can't spell. 3 5 [ Laughter.)
6 DR. WHIPPLE: It masks bad spelling.
7 Okay. To dive right in and talk about the 8 near-field environment and the engineered barrier system, as 9 luck has it, Ray's question was what do we need to know for 10 licensing, and the first point I tried to write down are --
11 there's a few topics that address information needs. I'm 12 going to have to keep looking at this to remind me what I 13 wrote, because I can't read it off the overhead.
14 What I have to admit that not just from
) 15 presentations here but perhaps more from reading several 16 iterations of the near-field environment chapters in various 17 versions of the TSPA, I come away with a sense that they 18 were written by people who were very good at details but 19 that nobody ever used the word " priority setting" in front 20 of them.
21 It's a laundry list of things that we would like 22 to do if we'had time and budget. It's not what are the one 23 or two things that we absolutely have to get done in order 24 for us to know enough to do this analysis.
25 And what I heard from the customers for the (m
\ -). ANN RILEY & ASSOCIATES, L7D .
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
453 ll near-field environment was different'from what I heard from
() 3 the near-field environment people.
. Roger by and large when you start working with C-22 you I heard from Joe and 4 : pretty much don't care what the pH is as long as it's, you 5- know,-not boiling picric acid or something that you just 6 aren't going toLget'in a mountain.
7 So I think that's the kind of issue. Some 8 interest things on chlorine content and perhaps an 9 offsetting positive effect-from other compounds that I'm not
~
-10 sure are on the program's agenda too well right now. But' 11 again-to look at the waste package corrosion people as 12 setting the priorities for what the near-field environment 13 work'should be I think is kind of a structural change that 14 would help a lot. All right?
p/-
(,, -15 Similar things, for behavior of the fuel and its 16 alteration products, as we heard this morning, there's 17 really not at this time a detailed set of models for
-18 alteration products and how they might hold on to 19 radionuclides. And without even'a pH 7 base case, it's hard 20- to know how one would do sensitivities on that to account
, 21 for-other factors in the near-field environment. But again 22 I think that that is probably an area where the pH of the f
"23 water and'some of the ionic content-is probably important.
24 LBut.to pick up1on Dave Shoesmith's slide of 25- confusion,L rising?and falling, you need to have some sort of
'l-s D ANN RILEY & ASSOCIATES, LTD.
' '- E Court Reporters
'1025 Connecticut Avenue,-NW, Suite 1014 LWashington, D.C. 20036-a: ( 202) 842-0034
-i c .
j
454 f I
1 a model, if: only a mental model, before you .know what data 2 you might get, before-you know what parameters of the l
-3[ near-field environment might make a big difference. Okay.
4 All:right. Some waste package issues. One is 5 that we sort of heard 2 rom a number of people,at this l 6 meeting,that the so-called corrosion allowance material is 7 really not a corrosion allowance material. .It:s a transport 8 task that provides shielding and mechanical strength to be 9 handled by big, heavy _ canister-moving equipment and to help
~
10 the. thing survive rockfalls that might occur. Okay?
11 The durability of the material from corrosion
- 12. appears.to be very short. I did my own little amuse 13 yourself back of the envelope, and I got from the rates I've 14 heard thrown around'here that the whole 10 centimeters of
- 15. steel outer pack has a corrosion lifetime equivalent to less 16 than half a. millimeter of C-22. So I just told Joe, all you 17 have to do is make a slight fabrication error and make the
. '18 C-22 too thick, and you're there.
19 So I think back in the time that this analysis got
-20 framed, the outer pack was seen as being a substantial 21- barrier and a slow corrosion process. The analysis just 22 doesn't show that now. And I think the analysis is still 23 organized arounc the idea that the steel layer is an
-24. important barrier to. release, when in fact what the analysis 25 would probably show is that the corrosion products:that l ANN RILEY-& ASSOCIATES, LTD.
. _4[ ' -
l Court Reporters 1025 Connecticut Avenue, NW,. Suite 1014 Washington, D.C. 20036 (202) 842-0034
455 1 result from that outer barrier have an effect on the system (c')
x_/
- 2 that's more important than before it's rusted out.
1 3 I don't know what those effects are completely, 4 but-again it seems to be that, you know, anything that's 5 only good for a couple hundred years you shouldn't spend too 6- much time on in trying to understand the performance. And 7 that may be where we are with those cans.
8 It also leads to a question of the tradeoff 9 analysis. We heard several times about the'possible denting 10 problem with the expansion of corrosion products in the 11' outer can occurring in a crevice around the inner can, and 12 compressing it or denting it. I'm not a corrosion chemist.
13 I don't know if that's a whimsical worry or a real one. I 14 do know I wouldn't like to be given the job of proving that ts
( ) 15 it can't happen. Proving things that can't happen is 16 inherently difficult, and, you know, some of the obvious 17 questions are what happens if you put C-22 on the outside?
18 Can you do that and get away with it?
19 Yes, the other tradeoff we heard mentioned, and it 20 came up in Joon Ahn's talk about tradeoffs between cladding 21 and backfill.
22 Okay, other analytical issues. The addition of a 1
23 cladding credit in the program was a fairly recent one, and 24 I guess the question I ask myself, if I were sent to go mine 25 the analysis and find some more margin, would I have picked
[ '
- ANN RILEY & ASSOCIATES, LTD.
'N Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 j
y 456
~
J IL cladding. And I' guess the people I talk to would have 2 encouraged me to pick -- to look at the secondary phases 3- that;might come in with much lower solubilities for some of-42 Jthe actinides. .Bret mentioned the ability -- the extremely 5 insoluble uranium complexes that presumably-could contain l 6 neptunium as well and could serve as a'much smaller source l
7 term.
l f 8' An issue that got raised briefly was that the 9 current analysis I think by both NRC and DOE assumes a 10 constant' wetting of a certain number of waste packages, 11 depending on the infiltration rate. As Joe mentioned, the 12 corrosion rate is the time integral of the wet periods, but 13- that if you're switching it on and off, you can get.some 14 bu'ildup of minerals on the surface of those packages that
() 15 may make it somewhat. accelerated corrosion when the thing 1 <6 does get wet.
17 I don't know whether that's an important factor or 18 if it's a self-canceling factor, but again it's a feature 19 'that at least cught to be listed as another probably source 20 of conservatism if you can't quantify it. And I don't know J21 ' how well one can quantify that. I think the model to 1
22 !-~ predict seats on the cans is a difficult problem. I can't 23 do better than they've done, but I can't say I have high i
- 24. 'cre'dibility in it. Okay.
'25- Issues of-the-influence of cement, which h
'\ /
ANN.RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
, Washington,fD.C. 20036 (202) 842-0034
457 1 presumably is limited in its analysis to raising the pH of
/)
\_/
2 water. Mae made the comment this morning or yesterday that 3 their sensitivities analysis show that if you do account for 4' the effect of cement on the pH of water in the first few 5 thousand years, it has a profound effect, more than an order 6 of magnitude, on dose rates at 100,000 years. To me that 7 kind of sensitivity analysis is really useful not because I 8 believe it but because it tells me there's something wrong 9 with the model that thinks that something you jiggle at 10 1,000 years is going to make a huge difference at 100,000 11 years.
12 Now you've got to work those sensitivity analyses 13 backwards. All right? It seems rather implausible to me, 14 but again, maybe I'm wrong, maybe it begins accelerating
()
(~%
15 corrosion while things are still hot and it goes much 16 faster. But I think that the take-home lesson from those 17 kind of results is that that's screwy and let's go find out 18 why it says that. Okay?
19 Just one of the questions that comes up, and we 20 heard just a little bit about thermal mechanical chemical 21 coupled processes at the front end. Joe showed a slide --
22 again, I'm just going to kill your whole recap, Joe -- in 23- which he pointed out under one curve where you had open 24 drifts and ventilated them for 100 years, that you barely 25 got above 100 C. A whole lot of unknowns in this near-field 1
(s- )' ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 j Washington, D.C. 20036 (202) 842-0034
458 1- environment analysis get taken off the~ table if you don't
() 2' 3
boil lots-and lots of water for hundreds or thousands of
'yearsfin the'near field, particularly the chemical ones that o.
.4 .have an unknown effect on the fracture surfaces where you 5 expect to get.-- if you have chemica1' retardation.
6 That falls'into the category of finessing 7 uncertainty, which is -- if that's an undoable problem, then c8 .maybe you need a design change that takes that problem off
'9- ;your critical path'.
10- Okay. Just programmatic issues. The first bullet 111- reports what I've already said, which is -- I get a strong
, 12 . sense that priority-setting is needed for the near-field 11 3 environment work, and that the drivers for that have to come 14 from the customers who are the waste package people and the
) 15; waste corrosion and behavior people.
16 With the residual or with even high-priority 17 near-field environment issues, I think we have to decide 18 which ones we can reasonably bound or characterize and model 19 and which we can't, and then go to the question of can we 20 live with the bounds for the ones we can't adequately 21' characterize, and if we can't, then we need to make a design L -22 change. And there seems to have been a lot of convergence 23 in the near-field environment work, as I have followed it,
- 24. but.there.is a~long list of things to do next year kind of
- 25
- things.
. ANN RILEY & ASSOCIATES, LTD.
Court Reporters L .1025 Connecticut Avenue,.NW, Suite 1014 2 84 - b3
(;
I" 459
-1. I think it is -- the timing of the overall program
() 2 3D meansethat this has to b . converged to a'few doable things.
-Other points. This is what comes from hanging 4 around John Garrick for awhile, which is you get this
.5- ' engineering sense that sometimes things are too complicated 6 toLdo from first principles, and you ought to go.out and 7 'look for a way to short-cut a lot of the-things that you 8 would build up if you worked from first principles.
'9 Again,'some of these were suggestions to do n
10 . experiments'that are.as nearly as pcssible replicas of what Y11 is going to happen in the mountain. Joe's comment again on 12 crevice corrosion test measures, packing metal samples'in
'13 wet sand if that is what we thing the situation is really 14 going to-resemble or in ground' tuff, which is clearly
() 15- something-we could do, rather than in water baths 16 - half-submerged, which is probably one step removed from what 17 you really want to know.
18 A minor. digression into yesterday's comments that 19 Mick got started and that Dave commented on on robustness, 20 ' simplicity and so'forth. I think the number of people who
. 21; have said we would like;a much simpl'er analysis, which of 22 course I would too, what they~really want is they want a 2
- 23. deterministic bounding-analysis that shows you are on-the-24 -righti side of the line, and if you are working in a
- 25- -
regulated-. environment -- I should say this to' DOE because
? JANN.RILEY &. ASSOCIATES, LTD.
-Court' Reporters
- 1025iConnecticutiAvenue, NW,' Suite 1014 Washington, D.C.~20036 (202)' 842-0034-Q 1
. t _a __:
460 1- the NRC knows this - the regulators I have worked with have 2 always been more comfortable with a simple model they can
( _
3 ' follow, with numbers they can look up, that'gets you to the 4 safe'part of'the operating space than they are with 5 something that has got 73 Monte Carlo variables.
6 Now I don't know whether you can do that with 7 Yucca Mountain. You might be able to for a 10,000 year 8 case. I think that is -- it's possible. I don't know that 9 someone sat down with one piece of paper and tried to do it, 10 but it may be possible.
~11 In hindsight, at the end of the WIPP study, what 12 they handed to EPA was the big Mcnte Carlo does everything.
13 The EPA staff took that, went back, did the one page 14 simplified bonehead analysis and said okay, these guys have
() 15 a safe repository -- we'll have to figure out how to approve 16 it because we don't believe this inscrutable analysis, even 17 though we have run the models ourselves, we have done some 18 -hand calculations on the side that give us confidence that 19 this place is safe.
20 Unfortunately WIPP was easier than Yucca Mountain.
21 The other point I wanted to make on that is the 22 term " safety case" is one that the Europeans use a lot, 23 .which is the opposite of the risk case. It is tell me why 24 you think.this thing will work -- why will it be protective?
'25' .Right? ---and.a simple analysis that doesn't elaborate on i ANN RILEY & ASSOCIATES,'LTD.
[\ Court Reporters 1025 Connecticut Avenue, NW, Suite'1014
. Washington, D.C. 20036 -
'(202) 842-0034
' 1
)
461 1 things that don't matter is a key way to highlight that.
t,.,) 2 Next point. I have a little sense of uneasiness 3 every time the defense-in-depth term is supplied to a 4 repository. I know the concept of using multiple barriers 5 is in common, but I think the concepts are slightly 6 different in ways that need understanding.
7 In a reactor you have got let's say four layers of 8 defenses that if something goes bad right now any one of the 9 four is adequate to protect a release, all right? What you 10 have in a repository is you have barriers that are set up to 11 fail sequentially, and if you were wrong about one of them, 12 well, that time step towards failure gets taken out of the 13 calculation. Everything gets moved up. That is just 14 different than multiple independent barriers.
s
-( , ) 15 The other thing is I don't think in the mountain 16 it is possible for them to be independent the way you can 17 work very hard in a reactor system to make sure you have got 18 independent power supplies and you don't buy the same pumps 19 from the same vendors all over the plant and so forth.
20 I just think that you are always going to be bound 21 by common chemistry, common infiltration rates and so on, 22 and if I look at how many things in the PA are influenced by 23 infiltrate rate, that is a major driver for the whole 24 system, and it is going to affect canister lifetime. It is 25 going to affect the duration and amount of retardation you
(( )-
- N ~-
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025' Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 I
(202) 842-0034 l-l
462
.1 : get"in! the' UZ ' and. SC andfit is just not what
[ :2 defense-in-depth means to me.
i 3 .And'I m'done.
4 DR WYMER: Thanks, Chris.
~
5 ~We're'not' going to entertain any questions as 6 these speakers make their presentations. We'll hold them 7' till the'end.
8- What we're going to do now is~--
9 DR. STAHL: Do you want comments first or do'you 10 want to-wait until ---
?ll DR.LWYMER: 'We're going to hold'everything till'
. - 12 the end so we.can be sure everybody gets a chance to speak
- 13 his piece,
- since time is limited.
-14. We're going.to proceed now according.to the order
() (15 of'the presentations that were made,'and we'll start with 16 Mick.
17 All we seeLin Las Vegas is a coffee cup.. Is there 18 anybody else -- oh, I'm sorry, that's San Antonio. We don't
'19 see anybody'in Las. Vegas.
20- Is anybody home'in-Las Vegas?
21 SPEAKER: Yes, we're here.
22 DR. WYMER: Okay. Good. .Thanks.
~2 3 < -DR. APTED: How do you turn this on, Chris?
- 24 -Probably with the.on switch.
-2 5 ' Okay. Let'me start with some good news, and In L
1; .
\
[d .
~ ANN RILEY & ASSOCIATES,.LTD.
- Court Reporters
.1025 Connecticut Avenue,-NW; Suite.1014
, ~Washingtoni D.C. 20036.-
(202)'842-0034 t
R '
1^
l-463 l
1 think then maybe go to the medicines going down.
s
,l ) 2 This is a viewgraph Charles gave me just perhaps v
3 to show it, to indicate -- I think right from the beginning 4 I'think Yucca Mountain is a viable place for a repository, l
5 and I think.there are a lot of viable pathways for it.
6 There's not one unique pathway. And a lot of it 7 unfortunately because we all have different disciplines that 8 we start as our starting point, we often end up in a sense 9 working maybe to defeat the other person's argument because 10 it might trump our own. But I think this in a very 11 qualitative way from NAGRA is showing that if we're trying 12 to meet some sort of required safety level.
13 SPEAKER: We're not seeing your viewgraph.
14 DR. APTED: Why not get the camera on like you did 15 yesterday?
16 DR. WYMER: Okay. You will in a minute.
17 We need to get rid of that background noise.
18 -DR. APTED: There we go.
19 DR. WYMER: That's good.
20 DR. APTED: Now we can talk about Las Vegas.
21 -
, [ Laughter.]
22 DR. APTED: All right. And again, these are 23 justi-- would be different independent designs where we're 1
24 put different emphasis on geologic setting, repository l 25 engineering,-by that I would mean backfill, waste package,
[,-~)
\
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
__ _0
464
- 1- ' which I would include' areas such as container.and waste form
/ 2 credit, or: institutional controls.
.31 So I.think yes, the answer is yes, Yucca Mountain 4 is viable,.and there are a number of ways forward, and we've
.5 got to find a way to organize ourselves to get. forward on 6 ^this:rather than sort of.each focusing on just one part of.
7- this~ geologic-setting, site suitability in the past. . Well, 8' we know that didn't-go anywhere, and we can't -
I'm a big 9 advocate of. backfill, but I mean you can't.be successful, J10 just by saying-backfill I'think in this.
cil One'I showed earlier, I don't-know if it -- trying 12 to-get a.. big laugh last. time, it didn't -- but unfortunately
- 13 '- I saw a lot of bottoms-up approach.through all of my
'14 c colleagues here that presented things, and I won't mention
- 15. ' names, Dave, but there's a lot of. focus on -- my part of the 16l ; puzzle or my infonnation was very important, and I don't 17 think we ought to assume that,
- 18 People say for years and years we have a big 19- discussion about glass leaching, 28-day glass leaching.
20 People were just saying don't you understand, Mick, that is
'21' the important thing, how a glass behaves in a repository.
22 But if you.look at a. fuller analysis, if you bring in things 23'. like transport constraints and so on, you see that.it's not, HU . 24 lit's not.'so.important.
125l 'And I.think:the same thing that I!ve'seen in' terms
[! ANN 5RILEY &' ASSOCIATES, LTD.
'\ : Court Reporters h '
.1025 Connecticut Avenue,~NW, Suite 1014 h Washington', D.C. 20036 (202) 842-0034-
.I i
465 of looking at global models on things like UO2 dissolution,
,2 the performance is' robustly insensitive-to the dissolution
[}
3 rate of UO2, 'because what are the key nuclides?
4 Neptunium-237, that's a solubility nuclide. So we don't 51 care how UO2 dissolves. -What'are the other ones? Iodine, 6 technetium. Where do we see'those peaks? Those peaks are-7 the 2-percent instant release fraction. Well, technetium,
- 8. in theLpast they've used 2 percent. Now with your newer 9- values -- well, in technetium they've used 2 percent in the 10 past, like TSPA-95.
11 So no, I could be. wrong, but there's a lot of 12 this.--lagain, this is the basic problem of the program.
13 I've got something I like to do,'and by God it's obvious' 14 it's important. It's not obvious until you put it through a
() ' 15 ' total' system model, do some sensitivity on it, and then look 16 back. And this gets down back to the last point I'll make, 17 and I'll come down to that.
18 So I'm not so sure UO2 dissolution rate or the 19 time of dissolution is that important. I mean, I know 20 there's an article by Tom Pickford in this month's ANS 21- Transactions where he shows that the far-field dose rates 22 .are absolutely independent of UO2 dissolution rates.
23 So I mean again we have to look at the
.24 sensitivity. What are we trying to achieve? Should we be
- 25- putting a spotlight 'cn1 that?- When do we have enough
/ ANN RILEY & ASSOCIATES, LTD.
[E]h Court Reporters 1025 Connecticut Avenue, NW, Suite 1014-l_
Washington, D.C. 20036 (202)' 842-0034 1'
I' i e_ -__: - -
466 1 information?
2 -Containment time and the functions of containment.
(V) 3 I think there's a lot of I think misunderstanding or just 4 not clarity and what is it. Are we trying to achieve just 5- something to get us through the thermal period, and now it 6' gets to what let's say a NAGRA approach would be, let's get 7 past the point where we're thermally driven couple 8 processes, whatever.that period might be at Yucca Mountain.
9 I think we've heard and for the first time I think 10 between Joe and Roger I've understood and maybe even gotten 11 some confidence, there might be a credible way forward on 12 10,000 years, 100,000 years containment time. My view from 13 assessment is that that time itself is not so important, 14 it's that probably that there's distributions that are being (m
! ,) 15 dragged along created by a longer time of failure, and I 16 think we need to keep those two issues separate. Long mean 17 time to failure itself isn't so important over this time 18 scale. However, if there was a sort of a uniform failure
- 19. from 10,000 years to 100,000 years, that's going to buy us a 20 lot of performance with a dose-rate model.
21 Or are we looking at containment as some sort of 22 beat-the-reg type of approach, that we have a 10,000-year 23 maybe cutoff and if we might be successfully done with L24 containment. That's possibly very legitimate, but I think l l
I 25 we need to be clear and up-front about exactly what's m.
I j' ANN RILEY & ASSOCIATES, LTD. 1 N/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 ;
1
467 1 driving our containment function objectives'.
2 There was some talk about whether containment in 3 . terms of time is maybe defeating some of the disruptive
~
4 scenarios that we have to consider. Part of it here is I
~
5J .think the initial conditions that any of us come in with,'
6 assumptions in terms of a design, impact. our final results 7 and conclusions.
8' So if I start with a design, let's say I'm from a 19 Navy reactor background and I have a great belief in
-10 containment and those type of materials, and I start with
- 11 twell, let's get a long-lived container. Pretty soon I'm led 12 to some conclusions like well,'I don't like backfill. That 13 actually creates.some high temperatures for me that I have a 14- hard time dealing with. And there may be other features
() 15 about compatibility or what-I need to know about the 16 environment. So depending where I start really strongly 17 skews my viewpoint of what's important or not.
18' Now if I were to start with a diffusive backfill, 19 I might -- as my main barrier, because that's my approach, I 20 might come to some very different conclusions. This comes 21- -back'that there's no unique way forward, but just be aware 22 that.if we start with a viewpoint that we've talked a lot 23: iabout at this meeting, about a long-lived. container, and 12 4 ~t hen trying to protect that,.we may be throwing out other
- 25. parts _of-this system that.might have very valid ways forward p
$f ANN RILEY & ASSOCIATES, LTD.
lLb 6 Court Reporters 1025= Connecticut Avenue, NW, Suite 11014 Washington,.D.C. 20036 (202)z 842-0034'
. . . _ . _ ._.. _. ___.________m
468 li also.
()
r 2 3
Finally,.I can't think -- I haven't seen a lot of evidence in the last two days of how we've used the multiple 4 past PA studies. We've done a' lot of PA in this business 5 and on this site. How have they been'used to guide, 6 constrain, andLultimately resolve the PA, the safety, or
- 71 design issues that are out there?-
8 I mean, we need to believe what we've.done before.
9- We've done sensitivities,' done sensitivities and dissolution 10 2 rate. What are.they? Let's take a look at them and.then 11 believe them and say sorry, you know, we know you'd like to 12 do some more work on that, but it's just not sensitive, we 13 have enough information. We need to set priorities, and
' 14 we're going to hear I think a lot of that from the other.
Qp . 15 speakers.
16 We also talked around the issue of uncertainty.
- 17. .And,1to me, it is important to keep something separate from 18' uncertainty and variability. Uncertainty is not a property 19 of the site, it is a property of you and I and Chris and so 20 on ',-decisions we make. That's where uncertainty comes from.
21' The site isn't uncertain. You know, there may be stochastic 22 . issues and stochastic variability of material properties 23 Lthat influence. corrosion, and that's important. The 24c uncertainties, however, I think we have to be careful about
- 25 .lumpingLthose together with variability.
i
,[k ^%/
ANN RILEY & ASSOCIATES, LTD.
u Court Reporters
~10251 Connecticut Avenue,- NW, Suite 1014 Washington, D.C. 20036
-(202) 842-0034.
469 1 So PA in the face of uncertainty. Every time we
[)
v 2 introduce a PA parameter, we are going to be dragging along 3 some uncertainty about that and possibly some variability, 4 some variability that is truly innate with that parameter.
5 Now, improving PA and safety that arises from 6 characterizable stochastic variability I think can be 7 justified. I think, for example, the Canadians, looking at 8 this issue of distributed container failure, tried to base 9 that on an understanding of the distribution over time and 10 space of temperature in the repository and came up with 11 different failure functions for their container based on 12 something that can go out and be measured and validated, and 13 was based on true stochastic variability of heat contents in 14 the canisters.
O I,j 15 What can't be done, I think, is improving PA and 16 ' safety arising from uncertainty, ignorance. And so every 1
17 time someone says, well, we are pretty uncertain about that, !
18 let's use a little broader PDF, and if that leads to a lower l
19 dose, well, let's stop working right now. The more ignorant 20 we are, the better off. Clearly, the repositories are very 21 safe.
1 22 And you saw an example of this with SKB in the 23 KBSC-3 analysis. A very wonderful analysis, but they came 24 up with a million year mean-time for containment and they 25 -found that, geez, there's still enough iodine.129 that l 1 lx) ANN RILEY & ASSOCIATES, LTD.
o '% /' ' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i i
L: _ _ _ _ _ _ _. _ _ __ o
470 1 unless_they distributed those failures, they were going to
() -2 3
exceed their safety limit.
well, we are uncertain and so we will assume a uniform So what did they do? They_said, 4 failure time starting at 100,000 years and going forward'a 5 million years.
6 What did they buy? They bought a factor of 7 900,000 dilution in peak dose. That's pretty good. But we 8 base that on ignorance. Is that something we are going-to 9 take to licensing? I don't know.
10 Finally, I think taking the PA and safety credit 11 .for the physical properties of degraded and failed barriers 12 is highly dubious-, I hear a lot of saying, well, the 13 pathway ana2ysis, this inside of the canister, and looking 14 at a diffusion analysis of that, or the size of our pinholes
() 15 or failures or crevices, or the porosity of this 16 information, I think that's extremely dubious and 17 non-robust. I think it's possible that you might want to 18 take a credit for the chemical properties. If something 19 breaks down, what is the sorption of alteration product or 20 something like that?
21 But I hear an awful lot of sort of claims that 22 when it breaks down, it's really going to be pretty 23 favorable. Of course, we didn't engineer it that way, we 24 didn't put it in that way, and we have a hard time ever 25 establishing it is going to look that way when it fails, but Y ANN RILEY &-ASSOCIATES, LTD.
3d Court Reporters-1025 Connecticut' Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
p p 471 l' let's make some favorable assumptions about degraded ,
J
~\ 2 products- I.think that's really very dangerous.
?[J We didn't hear a lot about --
3 4 DR. WYMER: You are going to have to kindLof push
- 5. on through here, Mick.
6- DR, APTED: Yeah, I know. This is the last one.
7 Didn't hear a lot about backfills. This is a l 8 point I made earlier about if we started'with backfills, for 9 examples, there's a lot to be gained, my opinion, from l'
10 backfills that don't obviate containment, don't obviate Lil: waste form studies, but they reduce the effective dose in
~12 terms of.our variabilities or uncertainties to nothing. So.
13 instead of having to deal with all this extra burden of 14- uncertainties and. variabilities every time we have a PA
() 15 16 parameter, there may be some things we can do, some barriers we institute that themselves will obviate and eliminate that 17 uncertainty being passed along into the dose. And that's
- 18. it.
19 DR. WYMER: Thank you.
20 Well, the next speaker was Roger Staehle, so he's 21 the -- I'm sorry. I'm sorry.
22 DR. STAEHLE: No, I think Bill is next according 23 -to your program.
24- DR. WYMER: Oh, okay. Somehow or other you got 25 left.off my list here. Yeah, I'm sorry. Bill, go ahead.
ANN RILEY & ASSOCIATES, LTD.
.i
%/ . Court Reporters-
. 1025 Connecticut Avenue, NW, Suite 1014
. Washington,'D.C. 20036-(202). 842-0034'
___.___Al_. ._- ._d.. -_ _ _ . _ _ - m _ . - -. _ _ _- _ _. - _ _ . _ _ . _ . . _ . . _ _ _ _ _ _ _ _ _ . _ _ _ . ._____..E
472 1- 'DR. MURPHY: Thank you. And thank you, Mick, for 2 setting.me up. Now,.I'll speak from my bottom of the pile.
I 3- I'm sometimes skeptical of the notion of a top-down approach j 4~ 'because_it requires someone'to-decide which way is up.
5 I am going to narrow my discussion to some aspects
- 6. .of_'the near-field environment which was, in part, my role in 7 presenting material at this meeting, and I'll summarize my 8' summary slide here. I think that the near-field environment 9 is-clearly important to performance of the repository. It 10 is complex and heterogeneous and we need to recognize that
-11 complexity and heterogeneity. It's transient, there are big 12 changes, and I'll talk about that a little bit more. It has 13' ta) concern itself with an evolving engineering design. Many 14 data that may or may not be relevant are poorly constrained
() 15 and the coupling that is related to coupled, or the modeling 16 that is related to coupled thermo, hydro, chemical and 17 mechanical properties is really at the state of the art, and 18 I sustain these conclusions of my description of the aspects 19 of the near-field environment.
20 Nevertheless, I had not been asked initially to 21 speak about this, but I think it's important to bring up 22 some aspects of the program, and I am in a delicate position 23 here because I am really speaking on my own behalf, not
'24 . strictly on behalf:of the NRC or the CNWRA, but I am going
.25 to-talk.about the program. In my view, I think ta) a very
[/.
\~-
i- ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 ConnecticutLAvenue, NW, Suite 1014
- Washington, D.C. 20036 (202) 842-0034
_ _ _ - - _ .-_____ _-_- _ - _ =
473 l
1 large degreeLthe NRC and CNWRA programs are highly focused.
l There is a deliberate /and extensive effort at prioritization
( ) 2 3~ of issues through the use of performance assessment, as 4' illustrated in our issue resolution-status reports.. And 5 there is an emphasis on resolution of these issues within 6 the NRC' program.
7 One aspect of our prioritization lately has f
8' involved extensive use of sensitivity studies to test the 9 sensitivity of performance assessment models to'various 10 parameters that are associated with the near-field
- 11. environment. Brett Leslie talked about some of these 12 earlier today.
=13 There are necessarily a great many simplifications l'4 adopted readily in this approach in conducting sensitivity
( 15: analyses and in attempting to-establish priorities for 16 performance assessments.
17 I would like to speak briefly on the issue of 18 bounds on environmental conditions, and I think is a very 19 important aspect of dealing with a problem as complex as the 20 Yucca Mountain repository is. And I think that much of the 21- . field and laboratory and modeling studies that have been 22 conducted by NRC and its contractors, and by DOE, has really 23 been devoted to establishing what these bounds are. And
- h .unless one takes into consideration some degree of the 25 -mmplexities, one has a' difficult time justifying what those I
.[ )
's /
ANN RILEY'& ASSOCIATES, LTD.
Court Reporters . !
1025 Connecticut Avenue,.NW, Suite 1014 l Washington, D.C. 20036 !
(202)842-0034 '
a I
- p. 474-L 1l bounds may be.
I 2' .I'll' speak a: moment about simplifications and' 3' ~ share my concept of geologic-disposal of nuclear-waste,,
4 'which, in some' senses, is a-fairly simple concept. We have 5- .a problem-of nuclear. waste 1that.will persist for a'very long 6 period:of time relative to our experience ~and relative to 7 human societ'y or human civilizations and, nevertheless, we 8: have to deal with this,.and so-it is very straightforward'to 9 turn to systems that we understand fairly well that have 10 longevity on that time scale. These are geologic systems.
- 11 It's a relatively simple. concept.
12 We know that are effectively closed geologic 13 Esystems that can persist for geologic time scales. .That's 14 what we need for a geologic repository, and so that's my
() 15 view of why we are interested in geologic disposal of high 16 level nuclear waste.
17 Many of'the complexities'arise in the repository 18- system because of its transience. Mick Apted, in his talk 19 _ yesterday, talked about the philosophy of designing a 20 geologic repository in which gradients should be small. The 21 environment should be such that there are compatibilities 22 'between the various components of.the repository system. We 23' are not necessarily vored with such.a circumstance. In 24 the case of Yucca Mountain we have very steep gradients in 2 5 .- temperature,. very steep gradients in chemical. potential and
-[ ANN RILEY & ASSOCIATES, LTD.
N' Court Reporters
.1025-Connecticut' Avenue, NW, Suite 1014-Washington, D.C, 20036 (202) 842-0034
475 these things need to'be recognized.
~
l (l' But they are transient, 2 they won't last necessarily a very long time.
- 3. I was impressed by the performance assessment 4.' ' presentations which showed that many of.the aspects of the
'5 transients'of the Yucca Mountain repository die out in a few 6 thousand years. That's a short period of time relative to i
L 7 the: necessity of geologic disposal.
8 And as a consequence of that, and this -- I don't 9 want this next point to seem too simple. But for the long 10 term, and I really believe that releases from waste from
-11 Yucca Mountain as a' repository will indeed occur after a 12 very long time. For those conditions, the ambient geologic 13 . conditions are the pertinent conditions, and an analysis of 14 those ambient conditions is quite relevant for-long-term 15- performance of this geologic repository.
16 -- Another aspect of simplicity is that in many
- 17. regards the natural system at Yucca Mountain really is quite 18 simple. It is thoroughly oxidizing. You don't have to 19 worry about redox reactions whatsoever in the natural 20 system. It is essentially free of organic matter for all )
21_ practical purposes.
22 The bulk chemistry of the rocks can be very neatly 23- described by a two component system of an-intermediate j
-24 feldspar composition plus a' silica phase, which not 25: coincidentally;are the predominant minerals that occur at' ANN'RILEY'& ASSOCIATES, LTD.
A_
~
. Court Reporters 10251 Connecticut Avenue,-NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 ~
z- _
_ _ _---__j
l 476 1 Yucca Mountain, but you can describe roughly 99 percent of n
2 .the chemistry by those two components. There are some (v) 3 relatively simple aspects of the geologic setting that 4 enable us to abstract from the complexity and consider these 5 things for performance assessment, for example.
6 Finally, I will make my pitch for natural analog 7 studies. These studies I feel are compelled by the very 8 long timescales that we are concerned about. There is no 9 way that laboratory studies can approach these long 10 timescales. They are very informative. I will mention a 11 couple examples.
12 As I mentioned previously, the Nopal 1 repository 13 is an uncanny analog of the Yucca Mountain repository and 14 one of the very key observations made in our studies at
() 15 Nopal and one that struck me quite from the beginning of our 16' studies there was that the oxidation rate of UO2, which 17 occurred massively at the Nopal 1 deposit at Pena Blanca, 18 was rapid relative to removal of uranium from the system.
19 There is a huge mass of oxidized uranium that is still there 20 that was once -- that was derived from a primary reduced 21 uraninite deposit.
22 This gives us an idea of the relative rates of 23 processes on a geologic timescale.
24 Also, I would like to refer again to the natural 25 alteration processes at Yucca Mountain and two cases in n
l ) ANN RILEY & ASSOCIATES, LTD.
' _/ Court Reporters ~
1025 Connecticut Avenue, NW,. Suite 1014 Washington, D.C. 20036 (202) 842-0034
l.
[ 477 1 .particular. The diagenesis of-the' vitric tuffs illustrates
.2-' zto us how the rock may vary as a' consequence of the
- 3 repository thermal impact. This has been well-studied.and I 4 -think'that.any model that purports to predict the evolution-
- 5. of the near-field environment should at a minimum be capable l
L 6. of representing what we.see to be the natural diagenetic 7 alteration of the same materials.
8 Furthermore, in the vicinity of the Yucca Mountain 9 in Paint Brush tuff, in fact we see or it's been reported 10 that there's matrix sealing which could clearly have a big
'11 . impact on repository performance if this does indeed occur, 12 and I think in addition to the waste package and waste form 13 customers, the near-field transport customers are also out 14 there.
( 15 I will stop at that. Thank you.
16 DR. WYMER: Thanks, Bill. I didn't mean to cut
- 17. you out of the program.
18 I think now we.are to Roger Staehle.
19 DR. STAEHLE: Wasn't there one more person? Abe 20 and Peter Lichtman.
21 DR. WYMER: These are just the panel members.
22 CHAIRMAN GARRICK: Fortunately, we get to hear 23 .from Abe a lot. !
24 DR. STAEHLE: I have some specific thoughts and 25 some general thoughts. I will start with the specific ones
/ ' ANN RILEY & ASSOCIATES, LTD.
- \m- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington,' D.C. 20036 (202).'842-0034
478 1 first.
-[) '2 It1 occurred to'me that there is another mode of
.V 13 . failure here that we hadn't' thought about.
4' If we start with this iron.over C22,-and we have a 5 localized penetration like a high anodic or cathodic --
6 anodic area ratio, and we poke.a hole to some extent the 7- forces due to that expansion will cause the C22 to deform 8 and fairly readily to break, and so you could have~a 9 perforation of the C22 in a matter of a' couple hundred years 11 0 as a result of a process like this, and I think that is a 11 very credible idea.
12 It would be geometrically affected. A round, 13 small hole probably wouldn't fail because of the geometric 14 constraints but several holes in a row probably would, so I
() 15 think that is a failure mode that we need to think about as
, 16 we think about using the iron and C22.
17 A different kind of a failure mode occurred to me 18 in thinking about just simply C22. We assume that the 19 corrosion is. dominated by a general corrosion because it is 20 so corrosion-resistent and if we assume the average rate 21 we'll say is a 10th of a mill per year, and that is based on
'22 'what I understand to be a reasonable average of .05 microns 23 per year but assuming there is a factor of 2 range which is 24' not a'very precise probabilistic statement but it's not bad.
25 'for a few minutes of effort here, and assume that the r
h[
LN'
- ANN RILEY & ASSOCIATES, LTD.
-Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) . 842-0034 ~-
L - -- - ---_--_----_---
479 1 failure occurs at a loss of 1. centimeter. That is, if you iT L 2' have ai2 centimeter' wall, let's assume.that'you.have failure
- U 3 of'1 centimeter; Then the life from'these assumptions is' 4 100,000
- years.
'5 Now if you have a uniform corrosion and uniform 6 weakening to.this 1 centimeter level, what you.have 7 initially is'this, that after.100,000 years the thing just 8: opens up,=and-so the' thing just opens up like this and you 9 have all the fuel elements just sitting'out-open, assuming 10 there was'no constraint on them.
~11' The possibility of such a failure occurring I 12- think is reasonable in terms of the fact the final wall will 13 be only 5 percent of the total diameter, which seems to me a 14 reasonable-failure criterion. So those are two approaches
-( ) 15 to thinking ab'out failure that I think we ought to consider, 16 Now I have a few. specific comments on some 17 processes mentioned by others. There is a question of the.
~18 availability of ferric ions in solution. We need to note 19 that the ferric' ion is much less soluble than the ferrous 20 . ion, furthermore that in other words as you increase the --
'21 the higher the valance, usually the less solubility of any
- 22. kind of'an ion, and as you increase the potential, the
! ~23 ~ solubility drops according to the general pattern here.
24 So it seems-to me fairly unreasonable that there
~
25 4will be'very much~' ferric iron available to affect the
' ANN RILEY & ASSOCIATES, LTD;
! N'% Court-Reporters i 1025 Connecticut
- Avenue, NW, Suite 1014 l
, . Washington,' D.C.' 20036 -
.(202)~842-0034
~
- a. ,
480
- 1. ' zirconium inside the waste' package. There are some other 2' arguments there, but those are the central ideas.
(
.Now welhad-this very interesting-discussion aboutL 3
'4 bentonite-this morning, and I want to make a couple 5 observations about the-bentonite.
6 First of all, if we use bentonite and actually add 7 a'little bit of calcium to it, we could easily adjust the pH
- 8. to the minimum solubility point for steel, and therefore 9 make steel to be a very, very corrosion-resistent material.
10 I don't know if there is a lot of data in this area but I-
-11 know that work that had been done many years ago by Morris 12 . Cohen at the' Canadian NRC shows that these kind of corrosion 13 rates are on the same order of. corrosion rates of the C-22.
14 Now on the other hand, the bentonite might
() 15 decrease the stability of the C-22. First of all, the ]
16 tungsten will be absolutely soluble. The molybdenum will be 17 absolutely soluble and the solubility of chromium goes up at 18- around pH 10 or 11. You can see directly from the Pourbaix 19 diagrams, and furthermore at this range of pH you are 20 beginning to get into the kind of range of instabilities 21- that give you stress corrosion cracking and also zirconium 22 becomes much less stable as you increase the pH, and that
.23 comes right off thermodynamic stability, so those are things
~24 that need to be considered. On the one hand, it favors L25 steel. ' On the other hand, it may not favor the'C-22 and the V
i ANN RILEY & ASSOCIATES, LTD.
O' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 c 4 Washington, D.C. 20036 (202)'842-0034
l-E
481 5 1- : zirconium ~.
[h l2~ 'Now I want to'make a' comment or two about the
- () :
J3 Lwelding issue. -I: thought that was.quite' properly brought up L4 yesterday. I'have been. associated with the nuclear business l^
5 .since.1957'and I:think I have been associate'd with virtually .
- 6 everyl major failure that has occurred in all the commercial 7 . plants. I didn't.cause them, i
l <
- 8. [ Laughter.]-
9- DR. STAEHLE: But I have been involved one way or 110- 'another in. virtually all of them. Not all of them have been i11- ' welding failures, incidentally, but welding has always 12 -caused ~ difficulties'in heavy sections. In virtually every
- 13- engineering component, welding ~has always been a problem,
, 14. >all the way from the thin welds on cladding to the thick
' 15 : walls on pressure vessels and in pressurizer and on and on
-16T and on, and wel' ding is not a trivial issue.
17 The problems are inevitable but they are also not 4 1
'18 - insurmountable. -
19 Welding produces three effects and these are I am
- 20. sure well-known to all of you, but just to put this out in.
2 11 front of everybody, welding produces very high stresses, "22 ' always virtually at.the breaking point, multiples'of the 23 yield usually, but there is a complex chemistry-and
'2 41 -
structure in the weld itself because of the dendrites.
25: The h' eat-affected zone.has.a chemistry change on a
-t lANNLRILEY & ASSOCIATES,"LTD.
Ad 4 Court: Reporters-1025' Connecticut l Avenue, NW, Suite 1014-Washington,_D.C. 20036-
- (202)'842-0034'-
c
- k. ' .
r i a_ L - - - _ _ . . _ _ _ - - . _ ~ - -
r- - - - _ - _ - _ - - - _ _ _ - _ - _ - _ - - - - - _ _ - _ _ _ _ _ - _ _ _ - --- - - _ _ _ _ _ _ _ _
.. 482 L.
L 1 'the' micro-level that is in the grain boundary kind of levels 2L and these are. frequently important and.-I won't read you the
.3 l'itany which Itam sure you are all aware of, and so we need 4- to consider the optimization of the welds,.to lower the.
5 stresses-and to reduce the chemistry uncertainty or _
6 variability'and we'need to also consider major-stress 7 rel'ief.
8~ The term " global stress relief" comes from the 9~ -nuclear industry, where for example B&W has routinely stress 110 relieved their entire steam generators and the final points i
11 is that weld development can't be left to the last minute 12 because it will certainly produces failures of types that 13 all of us'could imagine and some of us predict.
14 I want to' mention a point on the hydrogen issue
() 15 and zirconium, because I am concerned about that as a 16 failure mode. I would like to point out that the reactions 17 for the reduction that lead to -- are significant here. One 18 is the reduction due to water. That is not a big reaction 19 at-room temperature but you get some hydrogen and it is more 20 likely to have an oxygen reduction than the water reduction 21 and then of course hydrogen is not a problem at all.
22 If you1take a comparison between the oxygen 23 reduction or between the oxygen in normal water and the
-24 oxygen in the BWR where this is an issue, the oxygen in
'25 . normal water inside the Yucca' Mountain facility would be 10
-ANN RILEY &' ASSOCIATES', LTD.
' ^-
Court Reporters 1025 Connecticut Avenue, Inf, Suite 1014 Washington, D.C. 20036
.(202) 842-0034-
483
[)-
\m s 2 .about 10 the PPM, and there the fraction of hydrogen 3 entering the zirconium is about 10 percent, but the Yucca 4 Mountain chemistry, I think it'would be a lot less than 5 probably one percent and you could mitigate the entry of 6 hydrogen by catalyzing the combination reaction and if it i
7 was really a bother to somebody, you could maybe plate some 8 nickel on the surface and that would probably stop it 9 altogether.
10 Now we have had some discussion about container 11 development and I would like to ask some questions.
12 First of all, it is not obvious to me why the 13 first design has to be perfect. I think we are making an 14 implicit assumption that whatever the design we are going to (A) 15 come up with, ultimately if enough of us work long enough, 16 we will have a perfect container with the ultimately low 17 cost, and when you look at all the other big engineering 18 developments around the world, nothing ever works that way.
19 You always start with the first one. It is the most 20 expensive. It has the shortest life. But it works.
21 I think we need to question this implicit 22 assumption, at least what I can see, that the first thing we 23 build has to be the best and lowest cost. It's just not so.
24 The second point is why do we have to have just 25 one design first? We might have two designs, two
[\/ ') ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
l:
484 1- . alternatives, and.go to the field with two alternatives.
() 2 3
They would be. adequate -- the'best reasonable.
Then1why don't-we continue testing'and .{
1 4 -development? ~ WeLdon't have to-stop testing and development 5~ once'we put something in place, and then we' develop Mod 2, 6' better performance, cheaper, and then we' develop Mod 3, 1 7'- which is better performance and cheaper.
j 8 This is'the way that all of engineering proceeds 9 and I don'tisee any reason we don't proceed the same way L 10 here,.and so if we are going to multiply the costs of the l
11 first container over all containers, sure,-it's going to be j '12- expensive, but it assumes that the first container is the l 13 one that applies in all cases, and I think if we built the 14 first one percent of containers and Mod 2 might be the next
() 15. l .5 percent and Mod 3 is the next 10 percent and Mod 4 is the l 16 rest of them, I think we would have a healthier approach.
17 I'd like to mention something a little bit off.the 18 wall here, but I am on a review panel for the Army Research 19 Laboratory in Aberdeen, and we are looking at their approach 20 to modeling the battlefield.
21 Now what we have here is not really a battlefield, 22 although sometimes you kind of wonder, and they have 23 approached the battlefield by modeling the whole thing. I 24 mean helicopters -- they even model the grass and the trees-25- andethe bullets 1and the this and the that, and it's really ;
i I
l 1
-I \ ANN RILEY & ASSOCIATES, LTD.
A-sb LCourt Reporters
.3025 Connecticut' Avenue, NW, Suite 1014 Washington,' D.C. 20036 (202) 842-0034
__-____w. :
I 485 1 impressive. They do it by a massively parallel system, and l
l
[)
\_/:
2 I watched them as they ganged together most of the big l 3 computers on the East Coast and showed us how this worked.
l l 4 Now what they showed us is pretty simple but 5 nonetheless what they did. I don't see any reason why this 6 program here can't build a model of what this thing looks 7 like in detail all the way down to the interface between the 8 C-22 and the iron and build a model that works.
9 I would like to mention that the people who have 10 built this model, there are only about five people involved 11 in this -- a lot of computers but just about five people, 1:2 and they are very clever, smart people, very energetic, and 13 I think this would have enormous benefit to this problem, 14 and it would assist us with the interdisciplinarity. I mean A
( ,) 15 we all feed off each other. .You know, Dave makes a set of 16 comments from his point of view. They interest me and so on 17 and so on and so on, but if I could see better what someone 18 else meant'in some kind of visualization, it would 19 substantially improve my understanding, your understanding 20 and so on. It would help in sensitivity analysis, public 21 understanding. ;
i 22 I think we ought to give some thought to a 23 three-dimensional model -- develop this Army approach -- a 24 visualization model.
25 Finally, this is just pure fantasy in terms of a
('}
N-ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
l
_-_____________a
486 1 sketch but as I have mentioned the last couple of days, I 2 think it is time'to get down to engineering elements and
[G~)
3 these are just some general thoughts, and I think Mick 4 started here too.
5 We have to define an engineering life. It's not-6 clear to me that we have one that we all believe in and we 7 need to define failure -- what is failure? Is it 50 percent 8 gone of the C-22 or is it some amount of radioactivity 10 9- miles away? What is failure?
10 And we need to develop a concept of bounding 11 parameters. We mentioned this yesterday. I saw almost no 12 probabilistic involvement in any discussions we have had for 13 the last couple days, and to me that is part and parcel of 14 design is a probabilistic definition; almost no discussion
()
,~
15 about accelerated testing. We absolutely have to understand 16 how this is going to behave at long term, and we need a 17 rigorous, credible accelerated testing program.
18 I mentioned the thing about integrated computer 19 models and the concept of Mod 1, Mod 2, Mod 3. Dave told us 20 about a fabrication development program. It looks like it 21 makes sense to me, but fabrication development is always 22 bootstrapped. It's never adequate. It's never thoroughly 23 done and it is never done in concert with the people doing 24 the design, and having been involved and seeing a lot of l 25 fabrication developments in the past, they all have these l
i-I
/~%
/j. ANN RILEY & ASSOCIATES, LTD.
1/ Court Reporters 1025 Connecticut Avenue, NW,. Suite 1014 Washington, D.C. 20036 ,
i' (202) 842-0034
487 1 'same flaws in them in the sense of lack of integration into j'.v) 2 the rest of the system -- and so on.
-3' So what I have here, isn't the perfect set. It 4 just makes the point that I think it's time to put this in 5 the framework of an engineering program. There are 6 well-known elements to such programs and they can be 7 implemented and, as I said yesterday, it's time to get on 8 with it.
9 DR. WYMER: Thank you, Roger. Joe Payer is next
-10 on the agenda.
11 DR. PAYER: Roger lets me borrow his laser pointer 12 but now 'is A pointer. This is his loaner.
13 DR. STAEHLE: You can have it.
14 DR. PAYER: No, I wouldn't trust myself with that.
'O 15 The timing.on it and everything.
V 16 Before I start the handouts that are being shown, 17 I just had a couple comments that I wanted to mention 18 regarding sensitivity analysis and PA, and personal bias, 19 but if you run sensitivity analysis with PA and you find 20 something's not sensitive, it could mean that in fact the <
l 21 system is not sensitive to what you've just run, or it could 22 mean the PA model doesn't capture reality. And one of the
~23 dangers with PA is that we start to believe it all the time.
l 24 And I think Chris gave a caution there where if 25 something gives a result, either a very positive result or l l
\ l
- )
N/
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
488 1 no result, the onus is on us to'look at that and see if it t'%
( ,) 2 makes sense working it backwards and not assuming that the 3 model captures, you know, reality. These are tough things 4 to do.
5 So that's one issue.
6 The other is that by experience in this just over 7 the time period that we've been involved in this peer panel, 8 there's been in some cases a confounding of a process 9 function and an engineering component in doing PA 10 sensitivity analysis. So for example if we do a sensitivity 11 analysis on what's the benefit of a drip shield and we 12 assume that the drip shield stops all dripping water on the 13 package, so we go in and we turn off the dripping on the 14 packages, guess what? The packages last a lot longer. They n
15 don't pit.
16 Anything that would stop dripping on the packages 17 would do that. That assumes that a drip shield would in 18 fact be able to be put in place and do that. And so 19 there's, you know, there's this confounding between the 20 actual engineering device in some cases, particularly when 21 we get down to doing sensitivity studies within the EBS and 22 the waste packages and waste formats we're talking about 23 today.
24 I've played the game we normally play at this, and 25 that is you come to these, you know, you prepare outlines, j1j ANN RILEY & ASSOCIATES, LTD.
i Court Reporters L 1025 Connecticut Avenue, NW, Suite 1014
( Washington, D.C. 20036
- l. (202) 842-0034 I
p t ______
a
(._
489
'1 and'you think about it and you prepare all us biases and 2 then you sit here and you wait for people to say things that
.3 . agree--with you and:you' remember that and'you disregard 1 4 -anything anybody else says is-not relevant or -- and so'I'm 5 going back to some of my early slides here, but I'm-going to l
l ~
l 6: try to focus on a few of the things that just. pick up on it,-
7- because.I do think there were some common themes.here that 8 are worthwhile.
9 We said the waste packages have'to last a long.
10 time in order to get long-term isolation. They're not the 11 only thing that we need, but they're a very important thing 12 we need.
13 We focused in,.and I think it's very clear that
- 14. the primary design mode to be defensed against in the 15- . packages is crevice corrosion,'and we focused on there's a 16 lot-known about crevice corrosion,:but the specifics to the 17 Yucca Mountain. application and the Yucca Mountain conditions 18 have to be developed.
19 From the environmental standpoint, the modulations 20 within-the EBS are the crucial items. What's the water 21 composition in contact with the waste packages? What 22 controls that? Local modulation'in and on the waste package L23 are going to overwhelm changes outside that area. We need 24 increased -- the. processes are increased concentration and-25 (deposits, metal-nonmetal crevices, and metal-metal crevices.
I I -ANN RILEY & ASSOCIATES, LTD.
- b. Court Reporters 1025. Connecticut Avenue, NW, Suite'1014'
^
l Washington, D.C. 20036 l
-(202)1842-0034 1
490 l' Personally I'd-like to see some of the j )
~
'2 intell'ectual power resources,fprimarily people, and it takes 3 money and time to get those people active and~ involved, .that 4: if it's spent on-some of the near-field environment outside 5 lof the-EBS looking at what's going to happen within the EBS.
6 And I think there's a disconnect, I'm not so clear in the 7 NRC, but'I sense it both in the NRC and DOE programs where 8 there's this gap, the near-field environment people see
.9 their job stopping'when they define the. water that drips ICL onto the package, and then the package-corrosion people in 11 that sort of pick it up at'that point. .Okay.
12- I mentioned this concept, and I'm not going to-go 13: through this in' detail, but the idea that if we're designing 14 for corrosion resistance, crevice corrosion of corrosion-15' resistant material, C-22 and titanium, there's this concept
- 16. of having the highest temperature -- excuse me, the lowest 17 temperature at which an aqueous phase will form, so as the 18- repository is cooling down, we get to a temperature limit at 19 which we get condensed phases, aqueous water, on the 20 packages. That's T aqueous.
21 There's a temperature below which crevice 22 corrosion will not be sustained. That's a property of the 23- ' material and the corrosive environment. If we can select a 24 . material that has a crevice temperature higher than the 25= ' aqueous temperature, there is no temperature vulnerability.
ANN RILEY.& ASSOCIATES, LTD.
[s / _
Court Reporters
'1025 Connecticut Avenue,'NW, Suite 1014 L Washington, D.C. 20036
.(202) 842-0034
491 1 The material from crevice corrosion is not susceptible 2 within the repository.
[\_)T 3 If crevice temperature is below the aqueous, that 4 defines a temperature range when it's below the aqueous and 5 above the crevice, we're in a vulnerable rimo, we could look 6 at the time-temperature phenomena and translate that to a 7 time period inlwhich the crevice -- in which the packages 8 are susceptible. And I just repeat this with a little color 9 added to it, this is hypothetical, okay? If hypothetically
'10 - the aqueous temperature is 112 C, how does it get there, we 11 get a little boiling point elevation, we get some more 12 concentrated solutions, maybe things are happening in tight 13 capillaries where some moisture forms at 112 C, and we have 14 a material that has a crevice temperature of 80 centigrade.
. /m
( ) 15 This range between 112 and 80 is the susceptible range, 16- these are six outputs of the base case for six different 17 regions, six different packages, this package A starts at 18 80, so it's within the vulnerable range. It heats up, it 19 goes out of the vulnerable range, over this period of time, 20 it comes back into the vulnerable range as the repository 21 cools down, and it comes out of it here.
22 That gives a small period early on when it's in 23 that potential range, and from about 100 years to 2,000 l
24 Lyearc we're in this temperature range where crevice 25 corrosion'could occur. It's only going to occur if it's
[s/)- ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 )
492 1 wet. It's'only going to occur in an environment that will
[J
) 2 3
~ cause breakdown at 80 centigrade.
The point to be made here is the advantage of a 4 more corrosion-resistant material. A more 5 corrosion-resistant material has a higher crevice 6 temperature. I just knocked this aqueous temperature down 7 just to show it wasn't -- we don't know either of these is 8 the worry. I mean, if you look at the bottom line, we don't 9 know the aqueous temperature very well, although we can 10 guess at that much better, we don't know the crevice 11 temperature much at all for the Yucca Mountain environment.
12 In this case if we upped that you see here the 13 effect on this cooler temperature, we have a slight range 14 here, it's higher than the range, it's in the critical range t' h
( ,)
15 from this point to this point, and then from this point on 16 that material will not crevice corrode. So for A it goes 17 into the range at 100 years, comes out at 600 years. B, the 18 hotter package, comes down, enters here, leaves there, 700 19 to 2,500. These numbers and analysis aren't carved in 20 stone, but they give us an idea of two important things.
21 One, how long is the critical temperature range.
22 Plus the condition of the drifts and the drift walls, and 23 the surroundings of the package, in this case, from 100 to 24 600 years. For this case, it's what the condition of the
- 25. concrete and the liners and anything else at this particular l
/'N l ) ANN RILEY & ASSOCIATES, LTD.
^x / Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 u_- __
493
~1 : time period.
() 2
- 3: you this,when you showed these figures earlier.
MR. HORNBERGER: Joe, I didn't get a chance to ask
~4. What's wrong with me concluding that I should only 5- dispose of cool-waste?
6' DR'.' . PAYER : If you dispose of cool. waste, then 71 you're below that temperature all the time, and you can take 8' crevice corrosion off the platter. That's a design option, 9 .okay?
10' It's not what's in the' current base case or, you L11 know, but that's a fact. If it was cooled down below the
'12' crevice temperature prior to emplacement, then crevice 13- corrosion comes off the platter.
l'4 What are the key technical issues? They focus on
) 15 determining realistic extreme boundaries of the water.
~
16, chemistry, and it focuses on determining critical
~17- temperature for crevice corrosion in those realistic extreme 18 environments. Those are the two, in my mind, most important 11 9 ~ design parameters, or a' rationale for material selection and 20 - design.of the packages.
21 So.what_are the experimental material that's 22 needed?
23- I added this.one just because Bill Murphy reminded 124 me of the importance of and the benefit of getting 12 5 information.from service experience and natural analogs.
'rm t . .
ANN RILEY:& ASSOCIATES, LTD.-
'l
^-b Court Reporters 1025 ConnecticutiAvenuev'NW,- Suite 1014
- Washington, D C. 120036
-(202)'842-0034 i
_ _ _ _ _ _ _ _ _ _ - _ A
f 494 1 There's~'really a lot to be. mined,-if not in'a robust
() 2 3
modeling type mode, at least in a public and a expert feeling of comfort.
'4 What we-need is experimental data in many cases.
5 'And in'the stuff I've heard today and in past reviewing of l 6 .this, there are many locations where, in my mind, the 7 program is overmodeled and undertested. And I don't mean-8 that as a~-- in any way denigrating the people that are 9 doing'the modeling. They're doing a marvelous job.
10 But the validation and verification of those 11 models have to be tied back to some good data base. And, 12 certainly, in the corrosion area, that does not exist today.
13 There is some good data that's being generated, but we 14 aren't there yet.
() 15' For the waste package design, base case and 16 alternatives, we've heard about the steel on C22. This is a 17 large canister. It's got its freckles and scars, and it's 18 got;it's benefits and pros and cons.
19 We've heard about the alternative of staying with 20 'large packages, but going with titanium C22 and steel, in 21 multiple orders, in multiple thicknesses, and we've talked a 122 little bit or heard a little bit about going to. smaller
'23 packages that would be made of only the corrosion-resistant
,24 materials.
~25- It's my opinion that we have insufficient 1
s,
[
A
-ANN RILEY~& ASSOCIATES, LTD.
s Court Reporters
! 1025_ Connecticut. Avenue, NW,: Suite 1014 i Washington,-D.C. 20036 j (202) 842-0034' '
p L _ _ ________________J
495 1 engineering data and analysis to support, at this time, Ih 2 which of these is-the best idea. Laboratory data -- the G
3 point is, though, from a corrosion perspective, and that's 4 what I'm here primarily to talk about, the experiments, the 5 data needs for any of these are the same sets of 6 experiments. You want to know how do these materials behave 7 in crevices, with rocks and corrosion products. You want to 8 know about C22 to C22 crevices and so forth.
9 So the corrosion tests, the corrosion needs don't 10 change with those selections.
11 What's the rationale for the multi-layers? What's 12 the order they ought to be in, the function they ought to 13 be, how thick they ought to be, and then how do I prove 14 those performances?
/T
(_j 15 From a corrosion concept, if the steel is going to 16 corrode pretty rapidly, it provides structural, then one 17 important thing to consider would be to put 18 corrosion-resistant material on the outside. The mechanical 19 people thinking about that, how do you pick it up, how do 20 you lift it, how do you move it, tell me that that's a 21 difficult thing to do.
22 I think it's worth seeing about what the options 23 are of that, how to carry out that difficulty.
24 So this is just, again, another list of materials.
25 We need realistic extreme environments. We need crevice
[~ ) ANN RILEY & ASSOCIATES, LTD.
N_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
- Washington, D.C. 20036 l' (202) 842-0034
(,
- l. 496 1 corrosion performance-test data, 2 One:way to approach that:would be test packets and fV) 3 moist soil and soil that's saturated with these extreme l-L 4 environments. We want to determine the critical temperature 5 for crevice corrosion of C22 -- C22, theLvarious combination 6 : materials. That's going to have to be supplemented'with 7: -some shorter-term, or short-term electrochemical chemical l.
8 . test to develop the rationale ~for why this critical 9 temperature is a good thing'to do.
.10 Stress corrosion cracking has_to be addressed, and 11 the way you address that is it's length in many ways to this 12 item 4, the fabrication welding, shrink fit and other 13 issues.
- 14 Stress corrosion cracking, you know,-the failure
- p.
3 15' analysis files are filled with a lot of stress corrosion-16- cracking failures associated with welding and-other
' 1'7 . fabrication-type methods.
11 8 The focus for this I ought to think -- think ought
- 19. to be much more in the shorter term. I'm just suggesting 20 50,-100,=300 years because that's when the surprises are 21 going to occur. Rather than spending a lot of time on 22 what's going to happen on 100,000 and a million years, I 23 would_suggest that we weight the program in the analysis and 24 the-testing 90 percent in the first several hundreds of
.25 ' years, and 10 percent,-looking at these very far out issues, h ANN RILEY & ASSOCIATES, LTD.
b-? CourtiReporters 1025 Connecticut Avenue,-NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 m, ,
h m___m_ _ _ - - _ _
l.! 497 1- because.if you get this wrong, this is when it's going to 2' : occur.
~
- 3. .The' failures from a corrosion standpoint,' stress 4 _ corrosion cracking standpoint, if they don't--happen in a l-51 period"of wetness in 10, 50,_ 100 years, the' likelihood-of-6-
them.then kicking in and happening are much, much less.
7 Okay. So if-you survive that early time, you i
L 8 survive all the time.
9 There was another issue here of well, what kind of-10 corrosion -- you know, how long would a waste package last
, "11' the entire -- you know, if you did this, if you could keep 12 the passive metal passive -- if you keep the passive metals
- 13. passive, as Roger said, you can extrapolate. You can 14 predict, based on that life, but you can get package 15 lifetimes of 100,000, a million years.
- 16. Do I think a package'is going to last 100,000, a 17 million years?
18 No. What I think it does is it moves the crevice 19 corrosion failure mode off the platter in that time frame.
20 And.then you've got to think about, well, what's 21 the next thing that might start limiting life. But right
'22 now, the critical thing to get us up into the several 23 thousands of years I think is crevice corrosion.
24 .Thank you, i
25 DR. WYMER: Okay. The next speaker is Dave e
[_ - ANN RILEY I & ASSOCIATES, LTD. ;
n ? ~_' Court Reporters-
~ '1025.: Connecticut: Avenue,.NW, Suite 1014
. Washington,,D.C. 20036-(202). 842-0034 l_. 5
498 1 Shoesmith.
,~
2 DR. SHOESMITH: Well, first of all, I think I need
]v) .
3 to respond to the challenge that my presentation was purely 4 self-serving.
5 It's not cles* to me what the difference between 6 " bottoms up" and " tops'down" is. I think what tops down 7 means is that you go and review the literature. You accept 8 what is known. You put it into the form of a PA model.
9 From then, it is an implicit certainty that this is going to 10 be correct.
11' .And I think bottoms up means that you start 12 without paying any attention to-what anybody else has done.
13 You create for yourself a little niche within which you 14 develop something, and then you go and test it against the (h
(_,/ 15 -reality.
16 Well, I can't say it any better than Joe about 17 believing a PA model that you put together from the point of 18 view of pragmatism without testing it against a good 19 experimental data base.
20 The second point I'd like to make is you could do 21 that for carbon steel. You could go out and find that a lot 22 f people know a lot about carbon steel. But I would i
23 ~ suggest to you that a few years ago, you could not have gone L
24 out and found that a lot of-people knew very much about 25 nuclear fuel from the waste disposal point of view, because f 'h ANN RILEY & ASSOCIATES, LTD.
A/ Court Reporters 1025 Connecticut Avenue, NW,. Suite 1014 Washington, D.C. 20036
-(202) 842-0034
i 499 l
1 why the hell would anybody bother with it unless they wished
. ,~
2 to dispose of it?
L(J,)
1 3 So.the burden to find that kind of information is 4 directly on these kind of programs.
5 Just to show that this was not self-serving, let 6 me give you a little bit of history. The history of the way 7 the fuel modeling went was, if people developed 8 solubility-based models on the understanding that UO2 would 9 not oxidize very much, and then they said, "Look, we don't 10 release anything. We can forget about it."
11 Then the Swedes said, "Well, this is a very --
12 this is an easy thing to do. What we'll do is we'll assume 13 that all the alpha radiolysis products react with the fuel, 14 and we'll just do a mass balance.
((S,) 15 So they did a mass balance calculation, and holy 16 smoke, it's eight magnitude difference in the dissolution 17 rate between the fuels by those two concepts. So you can't 18 have it either way. There was no option but to come into 19 the middle of the game and try to understand what fuel would 20 do.
21 Okay. Just to go back now to the main theme of 22' what'we're supposed to address here, rather than personal 23 challenges, I would like to amplify on-the point that Roger 24 Staehle made which is I attended a lot of meetings recently 25 and heard a new design or somebody has a new patent design
.,~
i ANN RILEY & ASSOCIATES, LTD.
'k Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
la ,
500 p
1 for-how this waste package might be changed. I don't think 2 ,the one we've got~is too bad, except for a few problems with
.3 the carbon steel, and even those I think potentially are not I 4 show stoppers.
5' I agree with Roger that there is a good' design on 6 the table. It makee sense to think about alternatives, but 7 from the licensing point of view, it doesn't make' sense that 8 the first thing licensed put waste'into this_ repository in 9 'the first' year will be the same thing that's licensed to put 10 waste into this on the last year. So there will be an-ill'. evolution in the container design, waste package design, 12 -which would make it better and cheaper, more easily 13 licensable, perhaps,'in the longer term.
14 So I think second guessing what the materials
( 15 should be all the time is not a particularly good idea.
16 That is -- I think there are a number of issues on the waste 17 package which still bother me, one of which is the need-to 18 predict the dimensions of failure in order to deal with the
.19 next model, which is what is the water going to do when it 20 gets inside the waste package.
-21 These are not predictable things.. Predicting the 22- size of patch failure, or even the dimensions of the pit 23 . failure'or a crack. failure is something that corrosion 24 scientists cannot do. This, to me, boils down to the issue 25' -I tried to point out at the beginning of my' presentation O ANN RILEY & ASSOCIATES, LTD.
A[. . Court Reporters 1025 Connecticut Avenue, NW,' Suite 1014 Washington, D.C. 20036.
-(202)- 842-0034
501
'l- which is that.this model, the waste package is not l () '2 sufficiently decoupled -- I beg your pardon. The waste form 3- is not sufficiently decoupled from those systems that go l
L 4 before it.
5' So it's not robust in the sense'that if you did 6 open up a huge space in the waste package, then you have a 7 big impact-on the next model down the line.
18 I'm not quite suro.how you readdress that balance, 9 but it'would be a lot better if we did not have to rely on trying to specify patch failures and pit failures.
l l 11 There is a clear need, obviously, to understand 12 how much water comes in, where it comes in, what frequency 1
13 it comes in with and what chemistry it comes in with. I 14 feel the chemistry is tractable, but I think the other may
() 15 probably be beyond the capabilities of this particular
, 16 program.
-17 The~ fabrication welding issue, to me, is a 18 critical one, and I feel it would be the one that's most 19 important to licensing. It's a justification that you can 20 close this and assure yourself that it will stay closed for l
21 a reasonable length of time. That doesn't fit into a PA L .22 analysis. It's almost-separate in the sense'that the l
L 23r fabrication and welding issues are not things that you can l-p 2
24 assess by performance assessment.
25' I would suggest that a lot of work needs to go
) ANN RILEY &' ASSOCIATES, LTD.
' Ad 1 Court Reporters . .
_1025 Connecticut. Avenue, NW, Suite 1014 Washington, D.C. 20036. '
(202) 842-0034~
y Y:
502 1 into-that area, even though it doesn't show up necessarily
[ ) 2 in performance assessment.
V 3 It shows up in the sense that, if you lock at 4 juvenile failures, they are the ones that critically control 5 failure release fractions.
l 6 From the point of view of the waste form, I don't' 7 think the game is over for the waste form. In fact, I think 8 the -- one of the problems with uncoupled models is the 9 degree of conservatism that you put into one component of 10 the model. It's often compensated for by what you haven't 11 put in somewhere else or what you've done somewhere else.
12 So the fact that the model may be insensitive to 13 something is not necessarily a proof that it's not 14 important.
(q ,) 15 I think a lot of attention should be paid to the 16 waste form because the feeling I get from this particular 17 waste repository is that it's front-end loaded; that you 18 need to stop the waste getting out, because once it gets 19 out, there are a number of critical pathways which are 20 relatively fast, compared to people who are working in a 21 compacted bentonite environments.
22 So I would say that the waste form is important, 23 as I would still say that the cladding is important. My 24 point about the cladding is, I don't think that the critical 25 scenario thinking has gone into cladding in a complete ANN RILEY & ASSOCIATES, LTD.
(
A '). Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington,.D.C. 20036 (202) 842-0034
503 1 sense, It's happening, if the number of phone calls and the 2 number of time Tae Ahn spends berating me about cladding is (Jn)
~
3 anything to go by, then this is clearly happening.
4 But there are a number of critical things about 5 the cladding. One, you do have to get a ferric chloride 6 environment, so you have to concentrate. You can only 7 concentrate on a heat source. The cladding will be a heat 8 source.
9 You can only concentrate at sites where water 10 resides for a sufficient length of time that you will 11 concentrate, thus likely to be low sites, not high sites.
12 There is not necessarily a correlation between the 13 failure site and the cladding and the drip site on the fuel, 14 because if it fails low, it's not going to drip on the fuel
() .15 that's above it.
16 These kind of arguments and scenario thinking will 17 place probability limits on cladding failure, as opposed to 18 absolute limits. That, again, increases the way Roger would 29 like to view this.
20 I think it's probably a rare event that cladding 21 will fail by a localized corrosion, but it's one for which 22 no real thinking is being done yet. So I would say the 23 cladding is still an open issue.
24 From the point of view of the fuel, it's quite 25 clear that the two biggest factors, and in this regard, I Dx
-l ANN RILEY & ASSOCIATES, LTD.
'- ^
i.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
[ (202) 842-0034 I
t _ - _ _ _ _ _ _ _ - _ - _ - _ _ -
504 1 agree with Mike, now that we feel that we know what the fuel 2 is going to do, are what will the retention factors for
( )
3 radionuclides be in the secondary products, and what will 4 the diffusive advective pathways be?
5 From the sensitivity analysis that Bill Halsey 6 presented, those were the two factors which appeared to 7 exert the biggest influence on the sensitivity, especially 8 the -- if you could go to some kind of diffusive pathway.
9 So, clearly, the emphasis on the fuel, based on 10 what we've heard at this meeting, should be on the diffusive 11 advective pathway analyses, and on the impact of the 12 secondary phase formation, both in blocking the fuel 13 dissolution.
14 If you block the dissolution, you don't have to i%
( ,) 15 worry about anything else, but also in returning the 16 secondary product. So it is a post-dissolution process 17 which is most important now for the fuel. I would certainly 18 agree with Mick on that regard. Yes, with regard to the 19 retention of radionuclides in secondary phases, I don't 20 think tliis is something which is amenable to thermodynamic 21 calculation, because you are never going to -- these are 22 -hosts in a matrix where the thermodynamics is controlled by l
23 something else. So it's the stability of the uranium phases '
24 which are going to be very -- which are important. So you l 25 don't have worry.about what the stability of a small l i
p
, ANN RILEY & ASSOCIATES, LTD.
L \ J- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 p
L_i___________-_
i l
l 1
- j. 505 1 concentration of neptunium is in a uranal based secondary s
2 phase. You just have to worry what the thermodynamic (v )
3 stability of the secondary phase is, and then worry about
. 4 how much -- what the retention factor will be.
I f 5 It will be an empirical analysis, I don't seen how l
6 you can do it from-any first principles. In that regard, I 7 suspect it is top-down, not bottom-up. So you will rely on l
8 empirical results. But all empirical results require that j 9 you have done a bottom-up scientific analysis so that you 10 can explain why you accept the empirical results.
11 Yeah, I think that's all I had to say. Thank you.
12 DR. WYMER: Fine. Thank you. Now, we have Dr.
13 Joonhong Ahn.
14 DR. AHN: I guess all these being said, what I can A() 15 add is very small. But I would like to extend my thought to 16 the issue of process model and integrated model 17 relationship. This is in a very simplified cartoon, but I 18 would like to discuss the relationship or comparison between 19 the conventional artifacts and the repository.
20 Well, I can start whichever -- whichever. But let 21 me start with the airplane. Here, this is -- here is NRC, 22 maybe we can take an example of reactor. Well, we first set 23 the functionality of the artifact and design for that based 24 on some principle. And usually prototype is created and 25 tested, and if it is lucky, the artifact actually functions.
'l ) ANN RILEY & ASSOCIATES, LTD.
\/% Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
506 {
1 In the case of airplane, it actually flies. It occurred in
'N
[d 2 3
the beginning of this century, and until the beginning of this century, people repeated this cycle again and again.
4 Well, at some point Wright Brothers succeeded and 5 since public felt that is good, so the airplane prevailed-in 6 the society, but at some point accidents happened and more 7 safety, more safer system was wanted, so safety regulation 8 and improvement occurred. Not necessarily in this order, 9 almost at the same time. But, gradually, the system was 10 accepted by the society.
'11 However, in the case of repository, this is not 12 so. We set the functionality of the repository, maybe 13 confining the radioactivity for certain time period, for 14 certain level of radioactivity leak. Well, even though we
( ) 15 can design, the prototyping is very hard. We can not 16 usually make -- build a prototype repository. It's under 17 --under the current circumstances, this is very difficult.
18 So we do performance assessment. This is very conceptual, 19 and we have to build everything in our mind, and there is no 20 grantee that everybody has the same idea, the same --
21 exactly the same idea.
22 But in case of airplane, because we have some real 23 things and so we can have a very coherent or good agreement 24 among the idea.
25 Well, as a result of performance assessment, we
/ ANN'RILEY & ASSOCIATES, LTD.
A- -
1025 Connecticut Court Avenue, Reporters NW,~ Suite 1014 Washington, D.C. 20036 (202) 842-0034
507 1 can. judge if.that is safe or not. If no, we go back to the l J) 2 3
design or go~back to the performance assessment.
'are in this stage. Hopefully, sometime in the future, we I think we 4 . reach this point and we'go to the public and if public say-l 5 yes, we start the construction of the repository. I think l
l 6 this stage corresponding to this stage.
7 So what is done after the realization of something 8 is being done before the realization. That is I think the 9 most difficult part and why the performance assessment is so 10 important.
11 Well, since, as I mentioned earlier, the 12 performance assessment is -- the major part of the 13 performance assessment is being done conceptually, we have 14 usually difficulties. But for performance assessment, we
( 15 develop process models to understand the details of the 16 processes and for the safety assessment we tried to abstract 17 the model. Unfortunately, I think there is a very big gap 18 still.
19 In some case, the safety assessment model picks up 20 very favorable or easy to use models among the process 21 models and construct the safety assessment models. And so 22 the safety assessment model is trying to establish in such a 23 way that the result will give the bounding results,. or the
. 24 - robust'results. But whenever we are asked if that'is really
-25 bounding, we have'to go back to the process models. And so
- [ ANN RILEY & ASSOCIATES, LTD.
EA _
Court Reporters 1025 Connecticut _ Avenue, . NW , Suite 1014 Wachington, D.C.-20036 (202)-842-0034
E 508
!=
L 1- the robustness'or the bo'undingness is very hard to prove,~I 2 think, in many cases.
3 .So I'think since we have very good computer
~
4' technologies these days, as~ Roger said, we had better try to
- 5. develop the methodology to abstract the model. But at this 6 point we have better synthesize what we know, or integrate 7 'what we know into a very realistic model. That can be used 8 later when someone asks if that model, safety assessment 9 model is really bounding or conservative, we can compare.
10 that model with the very realistic model and we can say that
' 11' this is really conservative and the abstraction or 12 simplification occurred, that occurred in this process was.
~13 justified.
14 So I think the performance assessment so far has
() 15 been focusing on the' analysis, but I.think now, since we 16 have so much information and knowledge, I think we had 17 better, at least once, try to integrate, synthesize all the 18' 'information together. And what we can see from the scratch, 19 from the beginning of the emplacement to the, say, 10,000, 20 100,000 years, in very visualistic way, or very realistic 21 way.
22- That's my comments.
23 DR. WYMER: Thank you.
2241 We have-just about used up our time, but.I want to 25' ask one other person;if:he has any comments to make. Marty?
[
JN /
ANN RILEY'& ASSOCIATES, LTD, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
' W a s h i n g t o n ,- D . C .. 20036 (2C2) 842-0034'
509
-1 MR. STEINDLER: No, I don't think so. I don't h
J 2 ~ think there is anything that I can add that hasn't already
- 3. been said. I would, you know, simply reiterate some of the 4 issues that have already.been touched on.
5 Coming from an experimental domain, I would echo 6 and second and third and do whatever to emphasize the 7 . general notion that it's okay to guess up to a point, but 8 eventually, folks,. nothing is better than a good set of l
9 experiments, repository' irrelevant, carried out in a focused 10 sort of.way to give.you an idea of whether you were 11 somewhere near the right track.
12 DR. WYMER: Even though we're according to the 13 schedule out of time, I think we should invite some
-14 questions of the panelists. I know that there was at least 15 one comment'that was to be made here a little earlier, so 16 let's please proceed.
17 DR. STAHL: David Stahl, M and O. I'll try not to
.18 be too defensive in regard to some of my comments on what 19 has been said by the panelists.
20 Firstly, with Chris Whipple, he talked about C-22 21 over carbon steel. Certainly that's one option that has 22 been evaluated and basically been rejected not only by us
'23 but by others because it does not give you any defense in 24 depth. If you have a fabrication error or some kind of 25 _
problem with.the C-22, you have no barrier at all, because b -
ANN RILEY &.-ASSOCIATES, LTD.
V-- Court Reporters L1025: Connecticut Avenue;_NW, Suite 1014 Washington,-D.C. 20036-
,(202) 842-0034 m= =
510 1 the carbon steel's not going to last very long thereafter. .
I
() 2 3
You mentioned, and I agree entirely with the need to establish priorities, and a comment that was made by Mick I
l 4 Apted is that those priorities should be done in concert 5 with the TSPA folk. That's exactly what we're doing in 1
6 regard to the testing that's done in my department on 7 container materials and waste forms.
8 We have a very limited budget. I mentioned some 9' of the budget problems we've had in the past. This year in 10 my department we have about $12 million in R&D. We hope to 11 -increase that in FY '99 to perhaps 16 or 17 million. But'I L 12 have requests for funds for over $20 million of work, so I'm 13 hoping that we'll work closer with the TSPA folk and the 14 design folk to make certain that we prioritize the work that
() 15 we will do. It's essential, and we certainly have in the 16 next year, perhaps 15 months, to provide the technical basis 17 for the license application. So it's going to be a fairly 18 busy year for us to try to gather the information that we're 19 going to need.
20 In regard to the CAM, Chris identified some of the 21 functions, and also provides radiolytic protection for any 22 corrosion processes that could occur on the surface. So you 23 do need some mass of material to prevent that. Certainly in 24l the 2 CRM design you can get by with a much thinner wall, 25 but then:you bring in the potential for some radiolytically b)
U-.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1.
( 511
~
1 ' enhanced corrosion, and that's the reason that we were l.
( ) "2 looking for a structural material to help knock'down that
-3 dose'and' deal.with that issue.
4 But then that brings in other negative r 5 interactions. For example, if you put; stainless steel on I .-
- 6. the outside of the titanium and C-22, you do-have some
.7 perhaps hydrogen embrittlement and other localized corrosion 8- problems that.you have to address. We can put the 9 . structural. material on the inside; but then we introduce 10- significant handling problems for the surface and subsurface
.11 : people.
12 As far as cladding credit, certainly we think 13 that's important, not so much as a containment barrier, but 14 more importantly for the surface area that is exposed for
() 15 alteration and release. We assume, for example, in TSPA-95 16 that once the containers fail, all of the surface area of 17 the waste form is available for release. And this is very, 18 very conservative. So we do have an analysis -- we agree 19 it's very preliminary -- that we've utilized for TSPA VA.
20 .And we're going-to have to work harder to justify that for 21 the license application.
22 As far as Mick's comment in regard to targets,
~ 23 goals, functions, I would disagree. We have a whole listLof-
'24 those. We can't perform'any; design function without having 25 an idea:of-what those requirements are. And we trace those ib Yy N ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue,.NW, Suite 1014
. Washington, D.C. 20036 (202) 842-0034-
_ _ _ . - _ - _ _ _ _ _ - - . . . _ . - - _ . . . - ~ - .
512 1 requirements from 10 CFR Part 60 through our engineered h 2 barrier design requirements documents, our control design
[O 3 assumptions, waste package interface documents, there's a 4 whole host of documents that trace those requirements down 5 to then what we utilize as design inputs. Some of those 6 were talked about by myself, some Jerry Cogar, I've 7 discussed briefly.
8 In regard to Roger Staehle, Dr. Staehle's comments 9 in regard to the high aspect ratio pit as a rew phenomenon, 10 and suggesting a failure of the CRM, I personally don't 11 believe that we're going to have localized high-aspect ratio 12 pitting for the carbon steel. Probably be more of a general 13 corrosion with a very wide pit, wide area where you have 14 corrosion products, and also I would suspect that you'd have r
( )h 15 corrosion in the interface, which would not lock in the CRM 16 to the container, the outer container. But that's something 17 we'll certainly look at. But my opinion off the top of my 18 head is I didn't think that was possible.
19 Lastly I did want to cover some of the things that 20 Dr. Shoesmith said in regard to cladding corrosion. We have 21 come tests under way that I didn't have a chance to discuss 22 in regard to Argonne National Laboratory both in the vapor 23 phase and in unsaturated drip testing. We're looking at the 24 cladding integrity issue, whether indeed under those 25 conditions you will rupture the cladding and provide j ANN RILEY & ASSOCIATES, LTD.
'A- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 i (202) 842-0034
'11 additional exposure area.
2i
~
I IOne'of the things that we were hoping to start 3~ .this year:was some crevice corrosion tests for unirradiated 4- . cladding, and unfortunately dollars were not available at 5 . midyear, so that will be part of the hopefully high-priority 6 : work that we'llLdo in FY '99. In parallel with that, not-7 unfortunately sequentially, we want to do some irradiated ~
8 1 cladding ' tests using similar geometries where we.would. pack
- corrosion. products and use modified chemistry such as.high I 10 'ferrochloride to see whether the irradiated cladding would ill . .be. susceptible as well.
12~ > JAt any' rate, that in a nutshell are my comments on 13< the panelists. And I thank you for the opportunity to 14 present-'them.
15 DR. WYMER: Well, thank you for those additional 16- insights into what you're planning.
17 And let-me say that when we take our break, which 18 we aren't going to take for a minute or so, we'll take a 19 15-minute break, then we'll have some additional time for 20 ~ comments after the break before we go into the public
.21- ' comment-period at'four o' clock.
122 Mickey, you:had a comment you wanted to make.
'23- DR. APTED: .Yes, I just want to offer a very l
241 sincere; personal apology'to Dave Shoesmith. I feel badly 4 25' there's no'none in this business:I have a higher regard both I
ANN RILEY: & ASSOCIATES, LTD.
? 'b/ Court Reporters' 1025 Connecticut' Avenue, NW,. Suite 1014
-: Washington,'D.C. 20036-.
- (202) 842-0034-
+
n.n, ,
514
.1 ' technically and for --
L , im Ll( ) 2 DR. SHOESMITH: Please, Mike, don't.
3 DR. APTED: His' personal integrity, and.if my I
L' 4- joking comments 1were misconstrued, I think I was intending 5 not to single'out Dave or particularly UO2 but just try to 6 iterate the point' of.taking that type of. study, and in this
- 7. Ilthink everyone. agrees an iterative process of PA and l'
8- . testing ~and so on. But I sincerely apologize.
9- DR. SHOESMITH: Well, the enthusiasm of my l> .1'O response should not have been thought of as anything other
- 11. than I normally lock heads with you, Mick, and we usually
- l. .12 argueflike this.
13' DR. WYMER: Well, that's probably enough of that.
14 Could you come back at 20 minutes after. Then O
-(,) :15 we'll have time for additional comments from the audience 16 and discussion. Thanks.
17' [ Recess.]
18 CHAIRMAN GARRICK: We're going to continue the 19 panel discussion, and the questions and comments on the 20 ' panel discussion until 4:00 or thereabouts, depending on how 21 things work out.
(
- 22. If anybody in the audience here wants to make a
- 23 L comment,-it would be helpful if you'd come up kind of 241 ' quietly
- and tell Andy'who you are and what your affiliation 251 iis,;and then.we'llDgiveLyou a chance to make a presentation.
2 ANN'RILEY'& ASSOCIATES, LTD.
A- .
Court Reporters ,
c1025 Connecticut Avenue, NW, Suite-1014 1
- Washington,'D.C.'20036 l
,_ 1(202) 842-0034- i 1
i 515 1- I know there is one person already who -- where l( 2 -did he go?
3 Mr. Englebert wants to make a -- would you come up 4 here and expose yourself to public view?
5: [ Laughter. ]
6 MR. ENGLEBERT: If I had known this, I wouldn't 7 have volunteered.
8 CHAIRMAN GARRICK: That's the plan.
9 MR. ENGLEBERT: This will be my first and last 10 time here.
11 CHAIRMAN GARRICK: Oh,. don't say that.
12 MR. ENGLEBERT: Thank you for the opportunity.
13 It's hard to find a new issue to be raised after listening 14 to this panel, but I'd just like to reinforce an issue
() 15 that's of concern to Clark County, and that's, basically, 16 the fabrication issue which Dr. Staehle and Dr. Shoesmith 17 have also addressed. And I will just keep my comments to 18 the present design because I don't know what the future is 19 going to look like.
20 There will be approximately 200,000 sheets of weld 21' in the C-22 that will not-be able to be annealed or stress 22 relieved because I can't see of any.way that you can stress
- 23. relieve C-22 at 1500 degrees centigrade with a steel 24 overpack around-it.
.25 This is going to probably cause a lot of weld E[-
[t ANN RILEY &' ASSOCIATES, LTD.
l.N* . .
. Court Reporters 1025~ Connecticut Avenue, NW, Suite 1014 Washington, D'.C. 20036 p _
(202)-842'0034 L.
. = _:__-_ ___ :
)
l 516 l' defects will not be found, will not be mitigated.
(). 2 There is some literature, and I haven't had a chance to look at a lot of C-22 corrosion issues, but we, 3
4 back in Coomb, from -- did a paper on the corrosion o'i C-22, 5 and they did find some pitcing corrosion in the weld that 6 was exposed to ASTMG 28dB solution, which was not something 7 I'd expect in the repository, but it didn't attack the C-22 8 in other areas, just in the weld. So I think there's some 9 history that this is a concern.
10 Another concern that we have has to do with the 11 annals in DOE's quality assurance program. There have been 12 -a lot of instances recently where lack of compliance to the 13 QA program has caused the issuance of corrective action 14 reports and DR's. In fact, they're on the increase. Part
(~%
( ,) 15 of the data that is the current DOE data base is under 16 question because of these quality concerns.
17 And it is hard for me to project from an inability 18 to follow written procedures in writing to manufacturing 19 something that is very complex to written procedures. I 20 think we're going to have the current issue of a continuous 21 -- a lot of problems, and this will lead to juvenile 22 failures.
23 And juvenile failures, if you look at the PA, have 24 a significant impact on the quality performance. And that's 25 basically all I have to say.
I ANN RILEY & ASSOCIATES, LTD.
< k/ ) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 I
517 1 DR. WYMER: Thank you.
<^s 2 DR. STAEHLE: Can we ask you a question, or is
(
3 that appropriate?
4 DR. WYMER: No. I think we have enough time to 5 get a little give and take.
6 DR. STAEHLE: This is -- this is -- you've opened 7 up a thing where we could probably all talk for quite a long 8 time.
9 DR. WYMER: Since we do have one or two more 10 people, let's keep it brief, but go ahead.
11 DR. STAEHLE: But I was curious. You're talking 12 essentially about welding C-22. Is that your final --
13 MR. ENGLEBERT: Yeah.
14 DR. STAEHLE: What's the possibility, in your I
(D,) 15 judgment, about using a welding approach like electron beam 16 welding, or a narrow -- equivalently narrow weld technology?
17 MR. ENGLEBERT: I'm certain that would be 18 beneficial, but, you know, I'm not that familiar with --
19 DR. STAEHLE: Okay. If that would get away from
'20 all the problems.
21 DR. WYMER: Okay.
22 MR. ENGLEBERT: What's the temperature for C-22?
23 I think it's 1100 centigrade.
24 DR. STAEHLE: Yeah.
25 MR. ENGLEBERT: By the time you have the steel
.c1 t
A/
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
! Washington, D.C. 20036 L (202) 842-0034
518 1 :overpack around it.
3.
'2 DR. STAEHLE: Okay. Well, I~was just curious to l[% )
3; know whether or not it would txa possible, because it would-
- 4. substantially reduce'the residual stresses and reduce the 5 kind of things-that.come from the key input, and I didn't
'6 have any sense of whether or.not --
75 MR. ENGLEBERT: Electron beam welding I'm sure 8~ fwould be beneficial. How feasible it:is, I --
9' DR. STAEHLE: Okay.
- 10 MR. ENGLEBERT: It should be feasible.
11 DR..STAEHLE: You're quite right on this idea of 12- . globally heat treating. I wasn't thinking about that, about 13 they have to heat.the fuel up to --
14 MR. ENGLEBERT: Yeah.
- - q'.
s ). 15 DR. STAEHLE: -- 800 centigrade or something.
16 That's crazy.
17 MR. ENGLEBERT: Actually, at 800 --
18 [ Laughter.]
19 DR. WYMER: Thank you. Did you want to comment on 20 that?
21 DR. SHOESMITH: Yeah.
22- DR. WYMER: Okay.
23~ DR. SHOESMITH: Excuse me. In a way, that's part L24: .of it, a failure.to some degree perform this assessment
=25' process because~it has to make -- I'.m not' suggesting you'can ys
'f b . ANN RILEY & ASSOCIATES, LTD.
b.1 '
-Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington,;D.C. 20036-(202)' 842-0034
_ . - _ - - _ _-.-n_--..-._ . . - - _ - _ . - _ - - - _
519 1 lose this problem in a probability argument. That's always 2, a' dangerous ~ thing to do. But -- so the assumption here is
])
3- that the failed weld will then be the site for instant 4 ~ ingressive dripping water, and all the other consequences 5 that-come after it.
6 .The PA mix -- I'm very -- I'm not sure why I'm-7- defending it, but I think because we went through the same-8 argument with our container in Canada, the probability of l 9 .that particular failure site being the ingress site for
- 10 water.is quite low. So I'm not suggesting the failed
~11 container should go in there, but it's not necessarily the 12 catastrophe that the PA predicts.
13 This is a perfect example of believing the PA 14 assumptions. 'We all do'this, and I think it's a little 15 dangerous sometimes. I think that's a special case for what 16 I would call, you know, a special analysis. Like we tried 17 to analyze a container failure where we assume we put it in 18 with a pinhole in it. We assume it floods instantly, and 19 then we try and calculate walking it out and see how far 20 away we are from meeting, in our case, a specified dose 21 rate.
22 ,
I think that's a case for a special analysis, if 23 .there is a problem of that kind.
124 MR. ENGLEBERT: And I don't necessarily agree with 25 -that'- ~ disagree with that.. My concern is also the quality l
ANN RILEY'& ASSOCIATES, LTD.
TN_/ Court Reporters
'1025 -Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034'
520 1 assurance' issue facing the global issue and not just for
[
- i'
). 2 that particular final weld.
3 DR. WYMER: Okay. Andy, do we have some other 4 people who want to comment?
i 5 MR. CAMPBELL: I believe Tae Ahn.
i l 6 DR. WYMER: Yeah. Okay. Just if you would i
7 identify him.
8 MR. CAMPBELL: Tae Ahn. Go ahead.
9 DR. AHN: Tae Ahn of Division of Waste Management 10 of NRC.
11 I have one question to Dr. Apted. I wonder --
12 vour statement of the contribution of instantaneous release 13 to pick those is from the slow dissolution rate of the spent 14 fuel, or do you consider all values of dissolution rate of
(_j
/
15 spent fuel dissolutions?
16 I believe for relative high dissolution rate. I 17 don't think we could see the pit dose from instantaneous 18 release.
19 The second question is, again, it is related to 20 the electro magnitude of solubility limits and dissolution 21 rate. In other words, the release of neptunium could be 22 controlled by dissolution rate, depending on the electro 23 magnitude of dissolution rate and solubility limits.
24 That's my first question and comment. Therefore, 25 I don't'think we could eliminate the spent fuel dissolution
( ) ANN RILEY & ASSOCIATES, LTD.
' > Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 !
(202) 842-0034
521 1~ studies at all.
, n
- ( ) 2 Second comment is regarding the cladding crevice, l~V-( 3 current and also evaluating the cladding crevice. I reread 4 it -- cladding crevice is more for getting the surface area l 5 of spent fuel.
6 We are considering three failure model right now.
7 One is rock fall induced mechanical failure and hydrogen 8 embrittlement and localized corrosions. We believe rock 9 fall induced mechanical failure could be the major damage to 10 crevice.
11 As long as hydrogen embrittlement is concerned, 12 even at significant corrosion of cladding, still you will 13 not have sufficient hydrogen -- amount of hydrogen to form 14? the radial hydride, which is called embrittlement.
fm
'q ) 15 We don't think the solubility limit of iron oxide 16' release the ferric concentration of ferric ions in a 17 -significant amount, even it won't reach one ppm range.
18 In Canadian studies, ferric ion can change the 19 corrosion potential with at least 10 ppm of ferric ions.
20- That's my comment.
21 DR. APTED: I agree I think exactly with your 22 comments. The instant release fraction is particularly 23 dominant when dissolution times are longer.
24 Unfortunately, I think they have to be on the 25 . order of'about.1,000 years or less for the major
[
/
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202): 842-0034
522 lL l' ' distribution itself to be contributing; And, so, yeah, 21 there is certain environmental space conditions where it
- 3~ would be fuel dissolution 4 I think I was widely misconstrued in terms-of the 5" need-for UO2 dissolution studies. They certainly-needed the 6- -summary to hear where we are. I just.hadn't seen any sort 7 of sensitivities.of.saying, "Okay. Where do we go from 8 here.," based on_that, and that's exactly what you pointed' 9- out.
10- Neptunium ~237, I agree with you that, for example, Lil -the distribution of values used in TSPA 95, the upper 12 portion of that distribution would lead.to conditions where
- 13, neptunium was reaction rate controlled, rather than 14 solubility controlled, so'you're exactly.right on that, i
l5 DR. AHN: Thank you.
16 DR. WYMER: Do you have any other --
17- MR. CAMPBELL: Brett can announce himself.
18 DR. LESLIE: I'm Brett Leslie from the NRC staff.
19 You heard from me earlier.
20 I appreciate the panel's wrapping it up, but one 21- of the concerns I had was that a lot of people talked about
- 22~ the program,- and there are different programmatic a 23 Responsibilities. So.when Mick or Mr. Shoesmith said, 24 -"Well,..we need to_be doing this," the ACNW is hearing, "We'
- 25. - need'to be doing'this." And.they're. thinking of it in terms N\ '
~A NN RILEY.&. ASSOCIATES, LTD.
Court Reporters-
. 1025 Connecticut Avenue, -- NW, - Suite 1014 Washington, D;C.-20036'
'(202)"842-0034 e:.s
/
___._.__.____._____.____________E___
m
1 l
523 1 of the NRC, and I think we need to be very clear on who's s
(q)) 2 responsible for doing what, and what are the impacts in 3 terms of licensing.
l 4 If the answer is that we need the data for this, 5 we will evaluate what we provide. You know, if we think 6 this is a licensing issue and the ACNW thinks that staff 7 needs to be really aggressively pushing this, that's okay.
8 But just be aware that that's -- we have -- we are concerned y 9 with the NRC's responsibilities, as well as, you know, to 10 the program.
11 I think I tried to touch on it just peripherally 12 in my talk, and I think the representative from Clark County 13 also hit it.
14 You know, this is an integrated program at the NRC C\
( ,) 15 which includes -- we talked a lot about the performance 16 assessment process level models.
17 We are incorporating all of this information in 18 the issue resolution status report. We have acceptance 19 criteria which clearly state what we would find acceptable 20 in a licensing arena. These acceptance criterias are going 21 to form the basis of our standard review plan.
22 Included in those are two programmatic issues or 23 - acceptance criteria. One is that expert elicitation are ;
24 done according to the new rate that we put out on expert 25 elicitation. It concerns a little bit about how l
/
(m)
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025. Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 u
I
524 1: ' solubility,: quote, " expert elicitation" have been done.
j Have to-take a'look at that to see whether that would be
) 2 3 acceptable.
~4 The other!is that the data that is used to support 5- it has been collected under a quality assurance program.
~
6- -And so we are -- we are concerned with quality assurance, as 7 Lwell.
8 I think in the end that ifiyou want to understand 9' -how the staff is looking at some of these issues, you also 10 really~need to look at what our issue' resolution status 11 reports'are. They're tied to performance. They're telling 12 us what the staff would accept.
13 DR. WYMER: Joel, maybe you should --
14 DR. PAYER: No. That's okay.
I 15' DR. WYMER: Okay.
16' DR. PAYER: It's just a comment.
17 It's hard enough to_try, from a technical 18 standpoint, understand how the mountain is going to respond, 19- let=alone the relationship between the various government 20' entities and staffs that are involved in this. So I
-21 personally apologize when I say, "the program," because I 22 ' don't know who else to -- you know, I think mostaof us 23 Taround'the table'are trying to say, "What do we have to
.24 .know?" And then, you know, it's your folks' responsibility.
25 tolhave-to-decide who's bailiwick that's in.
[ ANN RILEY & ASSOCIATES,'LTD.
b/
~
Court Reporters 1025 Connecticut Avenue,- NW,-Suite 1014 Washington, D.C. 20036 y -
(202) 842-0034-
[
V F" ,
F v '
~
j
1 525 1 DR. WYMER: Okay. Do we have any other requests
~
[') 2 or -- if not, I' challenge --
G' 3 Oh, do we have somebody?
l'
- 4 CHAIRMAN GARRICK
- San Antonio or --
5 DR. WYMER: Based on the panelists' presentations, 6 now, rather than on any other wider ranging comment, does 7 anybody in San Antonio or Las Vegas who wants to comment?
8 Okay. Yesterday, I challenged the two near-field 9 environmental modelers to tell us today if they, insofar as 10 they can, what the 800 pound gorillas are. That's what --
11 as I call them. What are the -- when you go through all 12 this modeling, what really pop out?
13 Where is the attention needed?
14 What are the really -- we've got some insights (m) 15 into this from some of the panelists' speeches and other 16 presentations, but I wonder if Abe or Peter would be willing 17 to tackle --
18 DR. APTED: Are they the right people?
19 CHAIRMAN GARRICK: Well --
20 DR. APTED: The consumers of that information 21 should be --
22 CHAIRMAN GARRICK: As a consumer, you'll have a 23 chance.
24 DR. APTED: No, no. I meant the container people.
25 Shouldn't they be --
~ - -
/D ANN RILEY & ASSOCIATES, LTD.
' J' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
.(202) 842-0034
l l 526 1 CHAIRMAN GARRICK: Well, I think the modelers are
[N_/-) '2. the ones who have probably the best grip on the --
l 3 DR. VAN LUIK: Let the modelers talk first.
l l' 4 CHAIRMAN GARRICK: Do we have a taker?
5 DR.. VAN LUIK: Yeah. We have a taker. This is 6 Abe Manager of Modelers. It's like being a den mother. You 7 think you're in control, but you're not.
8 The thousand pound gorilla -- I'll go better than 9 800 pounds -- is -- and if you look at the total system 10 implications of the near-field, the biggest item, I think, 11 is -- on total system performance is, does the waste package 12 see dripping water or not, and the chemistry of that water 13 becomes a second order issue after that.
14 Right now, we have a range of probably from 30 to 15 60 percent of the packages that see water, and therefore, 16 you know, subtract that from 100 to see the packages that 17 don't see water.
18 The ones that don't see water, even with our 19 current conservative modeling, last hundreds of thousands of 20 years. And so a change in the amount of packages that see 21 water is -- almost totally overshadows the nuances of waste 22 package failure dictated by the near-field environment that 23 we've been talking about. So, you know, that's kind of 24 undercutting what I had said yesterday about the importance 25 of looking at the formation of salts, looking at the
.(n) t ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
527 1 likelihood of early attacks by concretes modified waters.
.r (m) 2 If~there is no dripping water, none of these things really 3 matter, and the environments that dictates corrosion is 4 basically 80 to 90 to 100 percent relative humidity 5 environment in which you just have normal layers of water 6 slowly eating away.
7 So that would be my contribution to the biggest l
8 item in the system.
l 9 DR. WYMER: Well, let me ask. Does that imply 10 that a change in the basic philosophy of the approach that 11 there is no problem would be one that prevented water from 12 hitting the package like a Richards barrier or like a 13 backfill material?
14 DR. VAN LUIK: Well, you're leading me into a f
i 15 trap, now. You know, in fact, this is why one of the 16 alternates -- is that the correct word -- alternate designs 17 being evaluated in the viability assessment is, what does it 18 buy us to put a very long duration, strong ceramic around 19 the waste package to keep the metal from seeing water? !
l 20 And what we see there is that we don't have any 21 failures for many hundreds of thousands of years.
I 12 2 I h: .e not seen the curve that's going to be in i l
23 the VA, but it may be a flat line to a million years. So, 24 but then the problem is, we - .you knew, if you're going to l 25 do that particular design, we've already heard some i
[\ ')/- ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
(_
p 528 1 skepticism, even in this meeting, about the viability of 2 that material.
(vl
[ 3 And.so it just opens up another can of worms. We l 4 have to do some defense in depths. Yer have to look at some l 5 non-optimistic failure modes, et cetera. Premature
! 6 failures, juvenile failures, and model accordingly.
7 DR. WYMER: Thank you.
8 DR. STAEHLE: I don't know whether what we are 9 doing is appropriate to what you guys need or what the M&O 10 needs or what. This is all kind of a mushed-up discussion, 11 possibly, but one of the problems in thinking about, at 12 least from my point of view, about performance, is that how 13 much material you need depends on just the strength 14 requirements.
px
( ,) " 15 I don't know how thick this thing has to be, just 16 purely from a strength point of view, and that really kind 17 of affects how you think about the thickness from a 18 corrosion point of view. j 19 When we were having dinner last night, a number of 20 us including Dave Stahl, and we were talking about the 21 possibility of a graphite or a pyrilitic carbon container.
22 Well, the first question-you ask is brittleness. Well, 23 again in your early days of the Navy fuels they used metal ,
24 fuels because people worried about brittleness, but later 25 .they went to UO2 and it was just fine and it has been just j
/ ~s l' ' ANN RILEY &' ASSOCIATES, LTD.
'/
Court Report _ers i 1025 Connecticut Avenue, NW, Suite 1014 !
Washington, D,C 20036 )
(202) 842-0034
L 529 l-1 fine ever since.
2 It seems to me that it is kind of the same thing,
?
-(L,)
3 but I mean brittleness isn't necessarily bad, but we don't 4 have.a perspective on it, and so there is a whole thing that 5 is missing from this discussion in terms of corrosion and 6 that is the corrosion story sort of presumes on certain 7 mechanical integrities that we don't have available for this 8 discussion.
9 I have trouble kind of thinking of a total problem
<10 without knowing what the mechanical constraints are and the 11 corrosion constraints.
12 DR. WYMER: We certainly didn't go into that much.
13 DR. STAEHLE: Well, no. That wasn't meant to be 14 . critical. I am just saying that the kind of response p
( ,) 15 someone like me has, there is -- I think.about the corrosion 16 problem, I think about the mechanical problem, I think about 17 the other things, and if you concentrate only on the 18 corrosion problem it may be giving you guys the wrong 19 perspective -- I mean not that you couldn't have figured it 20 out yourselves, but --
21 DR. WYMER: Yes?
22 DR. APTED: Maybe Abe or I don't know if Bill is 23 still here.
24 This question about water entry I think is exactly 25 a key part. The cement liner -- I don't know what is in the l
- s->) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue,.NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 X
530 1 VA. Is there backfill in there? But if there is not, when l
-) l'
'/ 2 the liner fails -- I mean the liner is there to keep rock V
3 spalling and rock collapse so what happens when the crown of 4 this long drifts no longer have the support of this? I mean 5 do they crack and do they become sort of focuses of future 6 flow or do we know that or -- it seems like suddenly we lose 7 this nice, ideal geometry and isn't there a potential for 8 someone having a much larger set of flow packages contacted 9 or --
10 DR. STAHL: David Stahl, M&O. Bill Halsey is not 11 present but certainly we have looked at the degradation of 12 the system as a function of time and the PA folk working 13 with Jim Blink have come up with a scenario that I think has 14 been presented to at least many of the panelists, committee gg
(_) 15 members. Basically it describes the concrete liner going in 16 a few thousand years at most.
I 17 It would have aged presumably at that point. You I 18 would have some granular material on the surface as well as 19 large chunks. The remainder of the material would be on the 1
20 drift floor. That would be followed by rock from the area j 21 above the crown falling down and I guess in 10 centimeter 22 and larger pieces. There is some distribution of sizes that 23 they are predicting based on rock-falls in mines, so there
- 4 2 is some technical basis for it. I 25 There will be some local crevicing certainly that l
,- m
/ \ ANN RILEY & ASSOCIATES, LTD.
\# -
Cou"t Reporters 1025 Connectf , Avenue, NW, Suite 1014
' Washington, D.C. 20036 (202) 842-0034 l
531 -1
~
1 would'be created, but we don't believe that is of concern to 2- the' corrosion allowance mat'erial. It may give you some l 3 difficulties with.the corrosion resistant material but as 4 you know,Lthere is already crevice there. Certainly with
~
5 the. rocks orLthe concrete material, the pH at least 6' initially would be much higher so crevice corrosion would l 7' not be likely.
8 -I am not sure if.I have addressed..the point.
l' 9 DR. APTED:. Abe was talking about water entry
.10 being.the key factor and it just.seems --
11 DR. STAHL: Yes, thank you. We have looked at the 12 . water pathways,through that rubblized' bed and through the 13 region that has the' concrete and that has been used by TSPA
-14 in their VA analysis to determine the release of 15 radionuclides from the waste form.
, 16 Abe had shown a picture of the various spots where 17 they-'look at the water chemistry and that has -- I think it 18 justi showed- the non-degraded. case, but there was also 'an 19 ' analysis of the degraded case'and I believe that was input 20 to the VA, so,there is some at least preliminary 21' representation of what is happening there.
22 DR. APTED: I think maybe understands my point of
.23' getting.at it a little differently.
24- DR. STAHL: -Yes, go ahead. ,
'25 DR. VAN: LUIK: I think your point'is something I ' ANN RILEY'& ASSOCIATES, LTD.
N .
-Court. Reporters:
- 1025 Connec_ticut Avenue,.NW, Suite.1014
, Washington,cD.C. 20036-
-(202)1842-0034
'l
______._.._...mm..-_._-i_._ _
532 11 that we:should look at how this concrete-is implaced. It-2 will befimplaced in sections that will be leaning on top.of
])
.L : each othe'rowith notches and it.will be implaced in L
4 horizontal sections just like -- have you seen the'way that 5 the invert is implaced? -- so there will be plenty of 6: opportunity, every few meters, for water to come in through
.7 those cracks and so.the. idea that the liner is giving you a~
j 8 preventive against influx is wrong. ,
9 I think there is plenty of opportunity in the. .
10 cracks in the liner for water to ingress into the repository 11 even before the scenario that you are talking.about when 12- things start falling in.
13 DR. WHIPPLE: Maybe I could take a stab, because I 14 understood Mick's question differently, and he has this look i) 15 of you'didn't answer what.I meant.
16 What I think the VA does, and please correct me if 17 I am wrong, in its base case the only role of the liner is 18 to; affect the chemistry of water entering the drift. It is
~
19 not explicitly accounted for as a physical barrier that does 20 anything.
- 21. External events in the VA are treated as outside 22 of the mainstream side calculations -- volcanism, seismic,
- 23. human intrusion-are all separate calculations off to the 24 side,'and it is in the seismic analysis that you get rocks 25- of various size' distributions falling from the ceiling as a k..N-- <
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
__-__x-_ .__:_
533 1 function of ground acceleration, and there is a return
((
ss,~)
2 frequency for different acceleration levels.
3 The seeps model, which calculates the likelihood 4 of water dripping on a waste package conditional on 5 infiltration is, as I understand it, for an intact circular 6 drift. The criterion for deciding whether or not you have a 7 seep is whether the rock directly at the center line of the 8 top of the drift is saturated, and if it is you get drips 9 into the package, but there is no accounting for surface 10 roughness there, for oddities caused by the fact that you 11 have lost-rock and the shape has changed. It is just not in 12 that model.
13 DR. APTED: That's exactly it.
14 DR. VAN LUIK: Now I understand your question, b)
(, 15 That is exactly the way it is being modeled now. However, 16 we are working on a more sophisticated model that takes that 17 roughness into account in looking at the likelihood of 18 creating locally saturated patches on the top of the drifts.
19 DR. WYMER: David?
20 DR. SHOESMITH: Yes. I don't think we should lose 21 site of Joe Payer's description of cutting off the ring on 22 what you need from these barriers.
23 Depending on what you choose in material from the
, 24 corrosion point of view, there is a relatively short window I
-25 of opportunity for water to-do any damage and it is not that i
1
{ ) ANN RILEY & ASSOCIATES, LTD.
Court' Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l (202) 842-0034 i
?
L
,534 if long whenLyou think about it. .It is h'undreds to thousands 2- Lof years.
3 If the temperature gets_ low'enough then, you'know 4- ~there are very good arguments to say that'the material'is
- 5 not
- susceptible to drips. There are other engineered. ways:
6 of getting around the drip scenario, so although the PA is-7 l presently showing.that this is highly sensitive, it is not 8 an insoluble issue with a change of design or incorporation 9 of something else. It shouldn't be seen as -- I don't think 10- this is the thousand pound gorilla. I think this is the 15 11 pound baby baboon maybe.
12 [ Laughter.)
-13 DR. WYMER: Is Peter Lichtner here? Does he want 14 to comment on'the --
/) .
( j. 15 VOICE: He has already left.
16 DR. WYMER: Well, he was clever.
17 [ Laughter.)
18 DR. STAEHLE: There's something that we didn't 1
19 discuss here that maybe you didn't intend to discuss.
20- There's only so many things you can do in two days, but the
-21 one thing -- this is a corrosion meeting, so we didn't 22 discuss volcanism, we didn't discuss far-field kinds of 23 things.
24 DR. WYMER: No , that's right.
25' DR. STAEHLE: This was kind of a corrosion
. , .([ ANN RILEY &' ASSOCIATES, LTD.
LV Court Reporters l 1025 Connecticut Avenue,. NW, Suite 1014 L Washington, D.C. 20036 i(202) 842-0034
l l
535 t.
1- meeting -- what's the environment, what is the material.
~
2' In the framework of a corrosion. meeting the thing
}
3 youLdidn't discuss is let's suppose that we have two 4 materials-here, steel and C-22, of great interest.
5 What we didn't discuss is the data on C-22, how 6 good is:it,Lwhat.are the uncertainties in it,.what.is the 7 .probabildstic' nature of it. We-didn't discuss the four 8 modes'of corrosion in C-22, where they do occur. We didn't 9 discuss the range of -- the effective chemical composition 10' of real materials on the corrosion behavior, so there is a 11 certain class of things with respect to using this as an 12 engineering material.
13 For example, one of the first things I showed in 14 my talk the other day was how 10 different people using a O
i ,j.
s 15 very well-controlled environment and set of materials, got 16 four orders of magnitude difference in data.
17 Now I think we have to face up to that kind of a 18 thing here too, and so we didn't really discuss the things 19 that have to do with the corrosion of C-22. It was really 20 known implicitly by certain people to have certain patterns, 21 but-in terms of a design review or a design consideration of
.22 what are the best numbers or what is the dispersion of data 23 . or whatever:-- we didn't-discuss that.
<24 .Now, I don't know whether that's importanL at this 25- time:because, as I say, in two' days you can only do so many h(
N/
ANN RILEY & ASSOCIATES, LTD.
Court Reporters
~
1025' Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202).842-0034
1 536 1 things. But it would seem to me somehow, coming to the NRC, rN
- 2 there's going to be a set of data on C-22 and we are going 3 to have ask the question -- How good are those data?
4 Now, it's the same thing with the -- you can map 5 it on the carbon steel, you can map it on titanium, 6 zirconium cladding and ask the same set of questions. Only 7 that's something that I don't think we did. There's not --
8 there are people who have some of this information. But I 9 also know that there's a lot of information that we don't 10 have and some of these holes really are there, and I think 11 it is important to establish those holes and lack of 12 certainty and so on, and so on, and so on.
13 DR. STAHL: David -- David Stahl, M&O. I just 14 wanted to give you a partial answer. Dr. Farmer presented n
k 15 some of the data that we have for the C-22 tests from the 16 long-term corrosion test facility, that was basically one 17 year data. Conditions range anywhere from acidic in bulk at 18 2.7 to basic about 10 pH. And you could see that we have 19 seen some surface modification and we need to further study 20 that.
21 The literature out there is pretty sparse, 22 frankly, on C-22. As was noted, it is a new material from 23 about 1982, so there isn't a lot of long-term data. Most of 24 the data that is out there is very aggressive testing type, 25 either green death type solutions or ferrochloride by
[E ')
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
537 1 itself.- So it is useful.in the sense that it helps define a s
.! i- 2 mechanistic model that Dr. Farmer briefly described, where
\/s 3 it has in it concentrations of chloride and some of the 4 other inhibiting species. But we have a long way to go to 5 fully validate.that model and that approach.
6 DR. STAEHLE: I mean I wasn't trying to be picky 7 here, what I was trying to identify is that there is a class 8 of things we actually didn't do when it comes down to how do 9 you think about corrosion.
10 Now, for example, another kind of thing here, we 11 don't really know what the long-term behavior of C-22 is.
12 Let me give you some examples why that's important. You 13 look at zirconium and one of the classics in zirconium 14 corrosion, it was discovered early, it was known as early as
/
(,/ 15 1963, was that the zirconium had a transition and rate 16 processes. So you had cubic law up to some point, then 17 after that point, it vent linear. That was a very important 18 observation because there are certain mechanistic things 19 associated with it.
20 Now, we don't know whether C-22 has a similar 21 transition where it's real terrific or something sometime, 22 then because of the mismatch of the oxide and whatever else, 23 it goes and does something else. Or, on the other hand, 24 maybe rather than taking the linear assumption, which is the 25 assumption I just took with my simple-minded analysis -- not c x
('s > ) ANN RILEY & ASSOCIATES, LTD.
l Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
[ (202) 842-0034
~
I
, 538-
.1 necessarily simple-minded, but at~1 east it was simple -- but 2 I assume the. linear thing achieved. If it is a cubic'or a (f
3- ' parabolic or something, it makes a big change in terms of
-4 ' predicting-life.
5 So, you know, I think that there are some issues here which are very important, especially in terms of this 6
7 long-term predictability, that I would think somehow we need 8 to have -- again, from the point of view of the kinds of 9 questions I understand you have to ask, we didn't-ask those 10 questions today. Notwithstanding the fact that Joe Farmer
.- 11 has done quite wonderful work and some quite wonderful 12 analysis, I think those questions still have to be asked and 13 answered, and they weren't.
14 DR. WYMER: Joe.
' O/
.\ - 15 DR. PAYER: I agree fully with what Roger said.
16 But, again, I think the priorities are, in my mind, for the 17 program to be what are the extreme environments that we are
- 18. going to encounter in this place. There is no question that 19 where it drips and how often it drips is a very critical 20 issue, and the response is sensitive to that, and that's 21 being tackled by another group of folks.
22 It's prudent, I think, and nobody has said any
-23 different here, that it's good sound engineering and 24 performance assessment to assume that some of these packages
-25' will'get' dripped on. We don't know how many, we don't know i
[
n ' \ e["
ANN RILEY & ASSOCIATES, LTD.
Court Reporters
-1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036.
~(202) 842-0034
539 11 ' exactly where, but some of them are. -And so how-is the I'I 2 material going to respond and what is the best design of a V
3 package material to'put in place that is-going.to get 4- dripped on?.
.5 And so the highest priority is to determine the c 6 crevice corrosion behavior of these materials-in those 7 realistic environments. And then there's another whole 8 layer of needs we have, but I think it's another layer. One 9 of the things you asked is priorities, and I would still g10 like to just chime in with I think those are where the 11 priorities are.
12 DR. WYMER- David.
13 DR. SHOESMITH: We'll continue the corrosion
-14 love-in here. I agree with you, Joe, that those are the j f 15- ' priority issues. But I also agree with Roger that -- this 16 is'my point on the long-term cladding. If you look at the 17 expected-rate, you say this is going to be incredibly slow, 18 -but I don't think you have any option but to assume that
'19 even though the temperature is low, you will effectively go 20 to the linear rate that you see, because you can not -- you 21 can not' rule out anything else. So you have to accept that 22 there is a linear rate and that the buildup of the passive 23 film will not reduce the rate to zero, that it will only 24 reduce it to a certain standard constant level, which you L 25 probably will never be able to totally assess, so you will
[ ANN'RILBY & ASSOCIATES, LTD.
03-<* Court Reporters 1025. Connecticut Avenue, NW, Suite 1014
' Washington,:D.C. 20036 (202) J 842-0034 1 7
540-l- have toicover it-in conservative assumptions'and make some
- 2 'long-term' predictions,- .long-term calculation that says.we 3- feel certain with the'following bounds.
4- hit's okay to use a very low passive general
.5 ~ corrosion. rate if you want 1,000 years'of,lif'-ime. But-if
'6- you are: going to~say the viability of this repository relies 7- upon getting a million years out of C-22 in the passive 8 state, then it becomes a little bit of an issue as to-how 9 you can justify that that film won't change over that period l'0 of-time.
~ 11- So although the critical issue is rule out 12 localized 1 corrosion, which I feel'will be done, I feel the 11 3 . experiments will do that, and tracer material will do that, 14 there still is this long-term niggly issue of are we going
() 15 to stake all our marbles on 10 to the 6 years ofLsomething
-16 'we think is.not going to change. So it is an issue.
17 CHAIRMAN GARRICK: Marty, I think you had a 18 comment.
19 MR.~STEINDLER: Yes. All this I think simply 20 indicates that somebody has made choice, and it was an 21 interesting choice. There's an alloy called C-22 whose 22- .' experience level is, you know, perhaps 15 years, give and
-23 take,=whose database is so sparse that certain industries t - 24' find'it useful, but they don't have to do any long-term 25 projections and disa'sters in that' area are largely economic
<[I ANN RILEY &-ASSOCIATES, LTD.
- Ow- -
' Court' Reporters; 1025 Connecticut Avenue,-NW,' Suite 1014 Washington,;D.C.;20036 (202) 842-0034-
- - l_ __-_---_ - ----_i_L----
541
- 1 rather_than.'anything else.
L 2 . Compare that to an alloy that may not be;-- that 31 may notLlook'quite as good at the surface'just-now, but 4 whose information base, because of work that'has been done, 5- -either in this, program, and I don't know.yet what I mean'by 6 program, but somewhere within the DOE and NRC complex, or 7 from other sources, produces a much'better database, gives 8 you much higher reliability and allows you to defend much 9 more easily, come a bunch of cynical inquisitors, who will
- 00 come.
11 Those kind of trade-offs are being made on the 12 basis of -- I guess what Roger might call practical ,
13 engineering. There are not being made, I don't think, on 14 the basis of defensibility in the regulatory domain. And I l
N).
q_ 15 think ultimately this may give you some serious difficulty.
'16 .Unless it is possible, and it certainly is, considering the 17 vagaries of the political scene, that the schedules for
- 18. having to come to closure, to wit, license application and 19 all the other things where you have to finally decide, you 20 know, this is the data I want to stand on, unless those 21._ schedules are drastically changed.
22 What I think is a threat in this little exercise 23 is that because you are not ready to defend at the level 24 .that you have to,'somebody!may change those schedules for
- 25 'youi which may not be, in_the overall system, the best way
'[Y "i '
ANN RILEY & ASSOCIATES, LTD.
s . .
Court. Reporters-
- 1025.' Connecticut Avenue, NW,~ Suite 1014
- Washington,.D.C. 20036
~
, 1 (202) 842-0034-C..R___.n___'._ ._ __:
542 1- tolgo.
2- -So those trade-offs, I think, need.to be made with
(
3 a little:more thought of the consequences not to corrosion 4 alone, but to other factors.
- 5. DR WYMER: I think we probably ought to listen to 6 the rebuttal. Keep it short because we are infringing on 7 the public comment' period. But I think this is an-important 8 discussion.
9' DR. STAHL: This is David Stahl again.from the 10 M&O. I agree with you completely. 'One of the things that 11 we have looked at very recently is the corrosion data for
> 12 related alloys in the C system, an6 experience base there
'13 goes back about 40 years. So we have increased 14 substantially but not enough, I think, to give us a warm and
() 15 fuzzy for licensing, but it's certainly going to help, and 16 we found some data that's out there that is useful. 1 17 DR. WYMER: Roger, you can fire the last shot.
18 DR. STAEHLE: I would like to just try your 19 patience with something.
-20 A number of years ago, there was the development 21- of this alloy called Inconel 600. Some of you know this 22 alloy well and some of you know it less well, but it was the 23 alloy to be used for tubing for steam generators and 24 pressurized water reactors, and it was said that this was
[ 25- actually created by God himself, whose initials were HGR.
l I .. .
(- ANN RILEY & ASSOCIATES, LTD.
%6 . .
Court Reporters
'1025 Connecticut Avenue,.NW,-Suite 1014 Washington, D.C. 20036
- h. . -
(202) L 842-0034 w_
543
- 1. In 1959, an obscure French corrosion person l <s j } 2 ' discovered this alloy would stress corrosion crack despite 3 the enormous momentum of data which at the time showed that
'4 this alloy would not crack in chloride environments, and he 5 had the temerity to demonstrate this would crack in 6 absolutely pure water.
7 Well, it took ten years of going back and forth to 8 determine that, in fact, his results were so, because 9 everybody had a lot of vested interest, and it became an 10 alloy that, after a lot of testing, was shown to be just 11 enormously prone to failure of all kinds, and still later, 12 was shown to actually crack in the chloride environments 13 that were predicted not to by International Nickel years 14 before.
.A
(_) 15 There are some important lessons in this story 16 that are like the C-22 issue, that many of the things that I 17 can see in C-22 that could happen are exactly like the 18 Inconel problem starting in 1959, and I think that it would 19 be helpful for somebody just to kind of give a 45-minute or 20 a half an hour or a 15-minute, something, to say that the 21 analogs of that are right in front of us, so we're looking 22 at the analog and it looks good because there's no data.
23 .That's an overstatement, but I mean -- that's not fair to a 24 lot of really competent people, but I think you all know 25 what I mean, fx
- ANN RILEY & ASSOCIATES, LTD.
-- J'- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
544 l' I think that there are some lessons here that
() 2 3
'might be part of the discussion we've been having here about how we should be looking at these materials that we're going 4 ito commit-to that we really didn't -- we didn't answer 5- today. .I think what we did these last two days is worked
~
6 out.a structure in which they need to work, but I don't 7 .think we have addressed the' kind of things that are implicit 8 in this analog that I laid out, and I think that's something 9 that needs to be done.
~10 DR. WYMER: I think at this point, we will turn to
'11 the public discussion. I only have one request to make with
'12 respect to the public comment period, and that is that the 13 people who make comments adhere to the context of the 14 meeting, which is a discussion of the near field and
) 15 engineered barriers in the near field, and not get too far 16 ' afield from-that with respect to other types of political or 17 social or other aspects of the problem.
18 So with that, I think we'll commence the public 19 comment period. I think because they got up at five o' clock 20 in order to tune in on this hearing, we ought to give the 21 people in Las Vegas the first chance to make public 22 comments. So if there are any comments from that. corridor, 23- we would like to hear them now.
24 Are you there, Vegas?
'25 AUDIENCE PARTICIPANT: No comments here. Yes, I'm 1
! ' ANN RILEY & ASSOCIATES, LTD.
4~J . Court Reporters 1025_ Connecticut Avenue, NW, Suite 1014 lashington, D.C. 20036 (202)' 842-0034 4
545 1 'the only one left.
.q)[m) 2 DR. WYMER: Everybody is home asleep.
3 MR. CAMPBELL: Thank you. That's a comment.
4 DR. WYMER: Okay. Are there any other comments 5 you would care to make? Okay. No comments from Las Vegas.
6 Okay. Well, we'll turn inward, then. How about 7 -- are there any comments from the Center, Nuclear Waste?
8 San Antonio? How about San Antonio? Comments from there?
9 AUDIENCE PARTICIPANT: We don't have any comments.
10 DR. WYMER: Okay. Well, we've got an hour, give I 11 or take, for the people who are left in the room.
12 [ Laughter . ]
13 DR. PAYER: The problem in the room is fatigue and 14 not corrosion.
l
( ,
, 15 [ Laughter.]
16 DR. WYMER: Well, we can go back to this very 17 interesting more specific technical type discussion based on 18 the presentations, if you would like.
19 DR. MURPHY: This is Bill Murphy, and I would like 20 to make one additional comment that probably reflects my own 21 biases and interests, which I admit freely, but I think it 22 may broaden the perspective about the longevity of 23 materials. There are three-million-year-old uranophane 24' crystals at Nopal-1, and given the importance of neptunium 25 in people's performance assessments and given, in my view, U, [' ANN RILEY & ASSOCIATES, LTD.
[
Court Reporters 1025' Connecticut Avenue, NW, Suite 1014 -
Washington, D.C. 20036 h
(202) 842-0034 L
L.
546 l
1 the likelihood that uraneal phases will eventually form and j
,m .
) 2 persist for relevant long periods of time, I think that an (V
3 important issue is, for. example, the solubility of neptunium 4 in uranophane in solid solution, which was something that 5 David Shoesmith mentioned was unknown and would' rely on 6 empirical data.
7 DR. WYMER: Dave, did you want to comment on that?
8 DR. SHOESMITH: Yes. Sure. I mean, that kind of 9 data is quite clearly going to have a big effect on overall 10 performance assessment if you can show that it's going to be 11 trapped.
12 Unfortunately, though, the geological record is 13 something that exists, and you can study, Bill, and be 14 certain about what you see. The real critical issue is, can p) q, 15 anybody -- since we can't rely on the thermodynamics or a 16 kinetic model, we have to rely on empirical data. Somebody 17 has to do the experiment and that's going to be slow and 18 tedious. It's going to be like spent fuel. leaching.
- 19. Somebody is going to have to incorporate neptunium into 20 these phases and give a measure of how much you can 21 incorporate. You know, the retention factor must be based 22 on a measured number.
23 So I agree with you, there's a great scenario out 24 there for trapping neptunium. There isn't a good database, 25 not a really good one.
(9 '/
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
' Washington, D.C. 20036 (202) 842-0034
547 i
1 DR. STAHL: David, if I could respond briefly.
~D
[d 2 3
This is David Stahl. Those are just the kind of experiments that we have planned for later on in this fiscal year and
! 4' carrying over into FY '99.
l 5 DR. SHOESMITH: Go for it. That's good.
6 MR. STEINDLER: I was just going to make the 7 comment that it shouldn't take a rocket scientist to design 8 a reasonably --
9 DR. SHOESMITH: No, no. I just think it's going 10 to be a difficult experiment.
11 MR. STEINDLER: Somebody's got to do it.
12 DR. SHOESMITH: Yes. But it could be difficult.
13 DR. MURPHY: I volunteer.
'14 [ Laughter.]
I ) 15 CHAIRMAN GARRICK: I would like to switch the 16 gears a little bit here and maybe reflect a little bit on 17 why this session came about, and I think that something has 18 been said about that already, and that was basically that we 19 had heard a lot more about the far field than we had heard 20 about the near. field, and that there was an increasing 21 dependence on the basis of increased filtration rate, 22 infiltration rates and percolation fluxes and what have you 23 on the performance of the near field.
24 So I would like to kind of put my hat on as a part 25 of the citizenry and ask the following question. Since I j ANN RILEY & ASSOCIATES, LTD.
[/
'x- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
548 1 ~think the public has been pretty much conditioned that they
^2 don't have to depend on the waste package, they don't have to' depend on'anything but basically the geology, because of 4 .the emphasis in the program'and because of the population of 5 ~ the different disciplines and generally in the path, kind of l 6 the absence of what I would call the engineering l
l 7 perspective, at-least the kind of engineering that you think i
8 of when you think of a major project, research engineering, l 9 analysis engineering, in addition to design and project 10 engineering.
11 But I would like to ask this august panel is it 12 reasonable, is it reasonable for the public to expect as 13 much protection from the engineered systems associated with 14 this repository as they will receive from the natural 15 setting? I think that's an entirely different view that 16 --you know, we've been talking here about designing a waste 17 package, a few ifs are in line that it can retain the waste 18 100,000' to a million years, well beyond what is expected to 19 be any compliance time. So that would suggest that it 20 certainly is reasonable to expect, for the citizenry to 21 expect that we can engineer into this repository that level 22 of protection.
23 What I'm getting at is, if we're coming to a point 24 in this-project where we're kind of counting on essentially as-much protection'from the engineered' systems as we are 25-
-[/
\-
ANN RILEY.&~ ASSOCIATES, LTD.
Court Reporters
~1025 Connecticut' Avenue, NW, Suite 1014 Washington, D.C. 20036
-(202) 842-0034 i
549 1 from the geologic setting, I think that's a real turn in the
,'[V
~h 2 corner of philosophy and thought about this whole approach L -3. to radioactive waste management, and, you know, I would like L 4 to hear some comments about what we'can expect reasonably, L
l 5 what we can reasonably expect.
6 The finding that!the NRC has to make is, in fact, 7 a reasonable assurance finding, not an absolute assurance 8 finding. It's a reasonable assurance finding. Can we 9 reasonably expect to get a comparable amount of protection
.10 from engineered systems as we might get from the geologic 11 setting?
12 DR. WYMER: If you geologists want to deal with 13 -that question.
14 MR. STEINDLER: As long as we're relaxing a little O
-( ,j 15 and getting away from hard science --
16 CHAIRMAN GARRICK: I don't think it is. You see 17 --
18 MR.'STEINDLER: You're asking the wrong question, 19 ~ because it isn't the comparison between what the geology 20 will do for you -- for the guy at the critical group 21 downstream who's pumping water out of the aquifer versus the 22 engineered' barrier. What the licensing folks and the NRC 23 orientation is, and you have lectured from time to time 24 about that, is, in fact, the speed limit.
They're going to look at the system and say, well, 1s ANN RILEY & ASSOCIATES, LTD.
L b[ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036-p (202). 842-0034 l
1
550 1 .we've got two possibly separable issues. One is what 2 geology will do and'the other one is what engineered systems fw )/
3 can do, and the sum of the effect of both of those have to 4 meet one criterion. That is, they have to be compatible 5 with the regulations.
6 Now, if it turns out that you now know enough, 7 which is I think the issue, perhaps, that you now know 8 enough about the geology so that you can focus your 9 attention on the engineered barrier, that's an economic
'10 argument. You can quit spending X hundred thousand million 11 dollars a year drilling yet one more hole to refine the 12 fourth decimal point on a -- sorry about that --
13 [ Laughter.)
14 MR. STEINDLER: -- and you can focus your (O) 15 attention on just the kind of experiment we're talking about 16 here, incorporating neptunium and uranophane to see what 17 kind of solubility you can expect and, therefore, get better 18 confidence on the behavior of that system, or at least bring 19 that confidence up to the same level that you already have 20 in the case of geology.
21 CHAIRMAN GARRICK: Yes. The point I'm trying to 22 make, though, I'm a great believer in the integrated 23 approach and the multiple barrier and the concept of 24 conservative design, but I'm also one that's pushing, 25 because I think we can do better than making the concept of i
ANN RILEY & ASSOCIATES, LTD.
1_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
I
L i
! 551 L '
1 a defense in-depth a mystery. I think it needs to be (p) 2 accountable.
3 I think we need to be responsible in what we put 4 into this repository in terms of how much benefit we're 5 receiving from it with respect to protecting the health and-6 safety of the public.
7 All I'm suggesting is that if, in fact, we're 8 going to end up depending on engineered systems for a 9 substantial amount of the protection, then we're sure slow
'10 in getting around to it in terms of the population of 11 distinguished geoscientists versus, for example, the 12 population of distinguished engineers.
13 MR. STEINDLER: That was Roger's point --
14 CHAIRMAN GARRICK: Right. Right.
,o i s ,) 15' MR. STEINDLER: You know, we can get on with it.
16 CHAIRMAN GARRICK: Yeah.
17 MR. STEINDLER: But you already have those data.
18 You've got a dose versus time curve, which is, you know, 19 every TSPA you've ever seen which basically has a sample 20 point where the regulators want a sample point.
21- '
CHAIRMAN GARRICK: But the problem I have with --
22 MR. STEINDLER: But you can make that --
23 CHAIRMAN GARRICK: Yes.
24 MR. STEINDLER: You can sample that whole system 25 anywhere along the line you want becau'e you've got a whole
,s l 't ANN RILEY & ASSOCIATES, LTD.
k/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202):842-0034
l 552 !
1 composite of codes that will give you that kind of answer.
,~
- 2. You can draw the very same curve when your sampling is at
(% ))
3 the edge of the engineered barrier as a function of time.
4 CHAIRMAN GARRICK: I understand that, but my 5 problem is that in much of the work and in much of the 6 performance assessment, the engineering was basically an 7 assumption that if you get a drip and a certain corrosion 8 rate, that you're going to start mobilizing the waste and i
9 creating a source term, and that whole aspect of the l
10 repository program has been a kind of a vague and mysterious 11 arena in terms of the level of sophistication that's 12 involved, the amount of thinking that has gone into it and 13 planning and what have you, especially compared to the 14 characterization aspects of the mountain, (q,j 15 All I'm suggesting is if we took a little 16 different perspective here, maybe -- and turned up the 17 microscope a little better on the near field as this working i 1
18 group session is attempting to do, maybe the benefits would 19 be enormous, and we would be in a position to say, we've i 20 done enough characterization, or, we don't need to spend any 21 more money to do that. I 22 DR. WYMER: Let's give Joe a chance to jump in 23 here and give another view.
24 DR. PAYER: I think the concern I have is that if !
25 these projections are pushed out to the hundred thousand
[m't ANN RILEY & ASSOCIATES, LTD.
k) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
. (2 02 ) 842-0034
_--_____________ -__-_- O
l 553
(
l' million year framework, it seems to me that amongst the
() 3 2 engineering community.and the general public, they're pretty aware that we're being ludicrous. The fact that'we think we I
4 can predict what's going to happen a hundred thousand years 5 down the road, a million years'down the road -- the mountain 6 is_not going'to be there, all right? This is new stuff.
7 It's going to erode away. Whatever is there, if the 8 engineered barrier is still there, it's going to be laying 9 on top of the surface, or whatever else. I mean, who knows 10 v5at's going to go on,.I mean, when the glacier comes 11 through Cleveland, Ohio.
12 I'm serious. I took a cab ride with one of our 13 peer panels in Las Vegas, and the cabbie is asking, you 14 know, what are you working on? Ah, this stuff out at Yucca l' 3
( ,) 15 Mountain. He goes, oh, that's a big deal. I've heard about 16 that. You know, what are you doing? I said, well, you 17 know, we're looking at design packages and trying to predict 18 how safe this stuff is for several hundred years and 10,000
- 19. years. And, 10,000 years, he said. He said, you tell those 20- folks that there's a cabbie in Las Vegas that said this old-21 world ain't going to be around in 10,000 years.
22 But that's not the point. The worrisome thing I 23, 'have is that.if the focus becomes 100,000 to a million, we 24- lose sight at looking very carefully'at what can go in in
.25 ten, 50, 100 and 500 years, because--- and that's when the .
- ANN RILEY & ASSOCIATES, LTD.
1
\- -
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
. Washington, D.C. 20036
.(202) 842-0034
554 t^
1 engineering has to;be done correct, to get you through that
() 2' period, and then you can get into a standpoint and say, okay, if it survived that time period, it's now returned to 4 basically an ambient condition out there -- certainly you've 5 had some eff3ct on the surrounding rock with fracture, C things of that sort, but it's down to its ambient 7_ temperature. The waters are now the ambient waters, and I 8 think you could make_a case that we can see no processes by 9 which this is going to, you know, degrade.
Yes.
-10 CHAIRMAN GARRICK:
11 DR. PAYER: But, you.know, to go from 300 years is 12 --
13- CHAIRMAN GARRICK: I agree with you.
14 DR. PAYER: I'm concerned that if you try to give
) 15 the public or anybody else a warm feeling that we've 16 analyzed this for 100,000 and a million years and say that 17 with a straight face, that you sort of lose credibility.
18 CHAIRMAN GARRICK: But Joe, all you're saying is 19 how ridiculous the licensing process is.
20 DR. PAYER: No, I'm saying the focus of it. Yeah, 21 if the-focus becomes 100,000, a million, then I think it's 22 if not ridiculous, perhaps, but maybe the focus is in the 23' wrong area'. :It really ought to be, can we credibly get 24 through 300 and 1,000 years. By that time, things are 25 . cooled down and, you know, they may last a long time.
[dY ANN'RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut' Avenue, NW, Suite 1014 Washington, D.C. 20036 (202): 842-0034
sss 1 CHAIRMAN GARRICK: Yes. I --
m j i 2 DR. PAYER: I don't have a problem, I guess, with
%)
3 trying to push the engineering barriers as far as we can 1
4 push them and credibly defend and present those, because I 5 think it's the more analyzable or the more testable, 6 experimentally testable part of it. It's hard to do an 7 experiment on a mountain.
8 So then we say, okay, we can take credit for 9 geology and retention. We don't know how much is going to 10 be held by the uranol, but we know it's not zero. I think 11 we're just trying to focus where we are and saying that, you 12 know -- and the TSPA currently now I think is not taking 13 credit for some things that many of the folks that are a lot 14 brighter than I am in those aspects say there's some credit l
() 15 there. We don't know how to convince people, we don't know 16 how to do that in a licensing arena, so we're not going to 17 take any credit for it, but we know it's there.
18 CHAIRMAN GARRICK: Yes. Well, what you're talking 19 about and describing is a very natural engineering process 20 and it's probably the way we should do it. But we have ;
21 heard presentation after presentation and had meeting after 22 meeting of a material that is trying to demonstrate and 23 present evidence that this thing will sustain its integrity
-24 for tens of thousands of years.
25 So it brings up an interesting question of
,m -.
I i ANN RILEY & ASSOCIATES, LTD.
$M Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
-_____________D
i 556 j 1^ ' repository safety strategy. You know, if'we could kind-of
- 2 switch that strategy from the standpoint of incrementally
-( )
( 3j approaching the long-term solution of the radioactive waste 4 management problem,'you know, that would be awfully logical 5 and that would,: as some of my colleagues have said, that 6 would be solving the problem, and we're not here to solve 7: the problem, we're here to figure out how to get this thing 8 -licensed.
9 DR.'WYMER: Do you have another comment?
10 DR. STAHL: This is David Stahl again.
11' 'I just wanted.to comment on what Dr. Payer said in
' 12 ' regard to long-term behavior.. We do have a. performance-
-13 confirmation program which will last upwards of 100 years.
14 .and perhaps longer.
r 15 Blake Barrett from DOE has expressed the desire to
- 16. keep the repository open as long as necessary and to design 17 the subsurface facility so that it is capable of perhaps a 18 300-year lifetime. I'm not sure we'll run the confirmation 19 period that long, but certainly it gives us an opportunity 20 to look at the behavior of those materials. We'll be doing
.21 in situ and laboratory studies.
22 I do want to have an opportunity to correct the 23 -statement I made in regard to the neptunium and schoepite.
24i The. initial studies uill be -- excuse me -- and uranophane.
-25 The initial' studies'will be on schoepite because we can make f
E Sd. :
ANN RILEY'& ASSOCIATES,.LTD.
Court Reporters
'1025 Connecticut-Avenue,-NW, Suite 1014
. Washington, D.C. 20036
- (202).-842-0034-
557 1 =the schoepite-fairly easily. Uranophane is a little tougher 2 to make. We will'.be doing some studies of neptunium and 3 boltwoodite and hopefully trying to make some uranophane at 4 the same time. So I wanted to correct that before I. leave, 5' which i's going to be shortly.
6 I don't know if there's anyone else from.the DOE 7 .or the Yucca Mountain. project that will remain to try to 8 respond to questions.
- 9. [ Laughter.]
10 DR. WYMER: Roger, you might want to catch him
'11 - quick.
12- DR. STAEHLE: Well, John raised I think a pretty 13 interesting question and draws a lot of fire because there's 14 some bit in it. But, you know, it seems to me that it's not 15 unreasonable to build an engineered barrier storage for 16 almost however long you wanted it. You say, I have a design 17 objective --
18 CHAIRMAN GARRICK: Yes.
19 DR. STAEHLE: -- to build something for this long; 20L is it-inherently possible that something like that could 21 ' occur? And the answer is probably yeah, I mean, providing 22 you give-me a little bit of flexibility in maybe controlling 23- the water and not starting it at a hot phase, but starting 24 at a cooler-phase,;and a few things like that, and 25- notwithstanding the fact that I think there's probably some. -
1 ANN RILEY & ASSOCIATES,'~LTD.
A q; ._
Court. Reporters
'1025-Connecticut Avenu'e, NW, Suite 1014 Washington,-D.C. 20036
^(202) 842-0034 !
. e
$ f 558 l' - failure: modes we' haven't even lookediat But,l[ mean, the 2- . mountain-has,some~ uncertainties, too.
.3 So,'you.know,-'I. don't think that what you're:
4 ' suggesting.is unreasonable.. I don't think this thing;is 5 fgoing to dissolve tomorrow afternoon. But that is not 6 saying that ws don't have some homework we have to,do --
L' ~7= ' CHAIRMAN GARRICK: Right.
- 8. 'DR.-STAEHLE: L-- and some' paying attention we've 9 got to-do. But.there'is an inherent thing there that I.
l 10 think makes some kind of sense'at the level at which'we're.
~'*
11 . looking at it.'
12 DR. WYMER: -David?
13 DR. SHOESMITH: Your perception'that things have 14 ch'anged from a reliance on the geosphere to a' reliance on
() '15 engineered barriers is spot-on. It's also'a transition that 16 has occurred in all.other waste management programs.
17: Lars Vermer, a guy in Sweden that we know well on p 18 the1 fuel side, says he has a. rock in his backyard; according 19 to the geologists, every time he steps out there, it gets 20 - more porous and diffuses stuff through it a lot faster,
~
l ' 21 ' because every time-they look.at it,-they get less certain j 22 about.the performance of their geology. So they have gone 1
.23- - to' relying almost completely on their engineered barrier-f24 '
Iri' our ' review process . that we went through,7 we
'25 - < presented?a case basedion studies at.the Underground t ,
-ANN RILEY & ASSOCIATES, LTD.
L\ , ..
. Court Reporters-21025 Connecticut Avenue,-NW, Suite'1014
- ,sWashington; D.C. 20036
-(202).842-0034;
+ ,
'q ':. t '
4
... _:__._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _______._1________--__________a
559 1: Research Lab. We said that we felt that ana could find 50 meters of intact rock and that this was a tremendous barrier
) 2 3 ' between, you know, the vault and the first fault.
But then 4 .the. geologists argued like crazy, and right in the middle of
'S 'the review process, we had to present the second case for a 6_ poorer rock,_ assuming the engineered barrier system was 7 something different to show that even if the rock wasn't 8 good, you were going to still be able to contain~this waste.
9- So everybody else has gone through exactly.this 10 transition from a reliance on the geosphere to a reliance on 11- the engineer <_1 barriers. The thing that hasn't gone with it 12- is_the' funding.
13 DR. WYMER: Yes.
14 MR. FAIRHURST: Can I make a -- I've been trying 15'- to make a comment for a long' time.
16 Before the corrosion people -- it~seems to me that 17 there are two issues to think about. One is how wet is this 18 region that is' going to be dripping on the canister, and the 19' second one appears to be something related to temperature, 20 what level, wl:2ther you're achieving that temperature by 21 leaving the.~ region open or what. I would like to hear from 22- the corrosion people whether, indeed -- you know, Yucca 23 : Mountain is different in two ways. One is:that it is an 24 unsaturated environment; the second one was that there was
- 25 the" deliberate choice to'make it a high. temperature I ANN RILEY & ASSOCIATES, LTD.
p.[b-L '
Court Reporters
- 10251 Connecticut Avenue, NW,' Suite'1014 Washington, D.C. 20036
~
(202)'842-0034 s
r E
t ig 560
[
(
l :1- repository. I don't think those are immutable, and maybe if L:
j} 2 they need changes, they should be changed.
3'- ~MR. HORNBERGER: ' Ale f act that it is . unsaturated D4 is immutable.
5 MR. FAIRHURST: The unsaturated is, but the 16 ~ question of whether or not you keep it dry,-et cetera.
7 But the next thing that I might say is with regard 8 Lt o this apparent reoving away from a natural underground 9- environment and so on, there has been a moving away in the
'10 sense, ILthink, that Joe Payer has talked about, and that is L ill that you.cannot perhaps give the absolute' assurances that L
12 were thought at one time, although I might even quarrel with
,13 that having been involved with wet recently. There are 14 certain characteristics of salt which actually give you a
() 15 great deal of feeling of security over the long term because 16 salt is essentially a fluid which flows to an equilibrium 17- .which you can predict for a long time reasonably well, 18 'providing there's not a lot of heat involved.
19 But if.you want to say that the geological setting 20 is not . making la major contribution -- I don' t think anybody 21, is.saying'that -- then you ask why don't you put it on.the 22 surface. There'are fundamental ~ feelings and fundamental 23 . realities about keeping this at a certain depth. Like Joe 224 is talking about a glaciation period 50,000 years from now, 25 Lwell obviously it would be' ludicrous to stick _that on the
)[^)
s /~
ANN RILEY & ASSOCIATES,-LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014-Washington,.JD.C. 20036:
'(202) 842-0034-i
- p. ' '
561 b ~
-1 Lsurface and have'to deal with that. issue.
, '2 There are a lot.of issues that-do go away,.there 3' are a lot ~of reasons why. geoscientists do~believe that there 4- is a major' barrier of the natura1' setting. The point'is 5- that they found out because the state of the art.of the
' 6 L. predictive sciences.-- nobody had ever asked geoscientists-7; lbefore about making these kinds of predictions. They' wanted 8 .. -- they-would.tell you how old the rock was within plus'or 9' minus' half a million years. Good engineers could' design it
.10' for'about 100 years and it was a no man's land in between.
- Lil- .So therefore, the geoscientists had to play catch-up as fast 12 as;they could to try to somehow make predictions in this 13- uncertain range,.as somebody - I forget who it was; it may-14 'have been Nick Apted who said, geology is not uncertain,
'15 [ it's our understanding of it that's uncertain.
'16 So I think that -- and I'll finish now the point
.17 ' -
I think it is absolutely incumbent on engineers if, by a
-18 moderate effort, .they 'can, in fact, show that the engineered 19 ' barrier can be.a major addition to that,.to;do it. I don't 20 . think one has a moral' choice to do otherwise if it's a 21- moderate amount of work. Obviously if it.goes into a 22 multi-billion-dollar-project,. then there's another aspect to
- 23. !thatLmorality.
124 So I don't-think.it's a question of'either/or.
- 25' One we know
- we've gct a big-benefit'from; the other one, j
ANN:RILEY.& ASSOCIATES,.LTD.
'3, ,_/ .
~
Court Reporters 1025 Connecticut. Avenue, NW, Suite 1014
. ', ' ~ Washington, D.C. 200361
,b
~
1 ,
- . (202)1842-00344 _3 n
u
l 562 1? we're looking'at.
() .3 2 CHAIRMAN GARRICK: ' I think you give them the same budget that you gave the geoscientists-and you'll get it.
~4 'MR.. FAIRHURST: But I would like to know'~if, 5 indeed, th'ere is merit,.as appears to be with most other 6 ' countries, to keeping the maximum temperature of the 7_ repository low from the corrosion point of view. Is it 8 something worth worrying'about?
9 DR..SHOESMITH: Yes. I'think that's a debateable L 10 point. .There are. clear advantages to having it low. There
, 11 1 are clear advantages to keeping the-system' dry. So it's 12 something that you can debate back and forth.
13 So to me, the critical issue here is still water, 14 not temperature.
/
(G ,) 15 MR. FAIRHURST: Thank you. That helps.
16 IM1. - SHOESMITH: Yes. Just as in everybody else's 17 case, the critical issue is lack of oxygen. That's why they
'18 all feel pretty comfortable with, you know, metals and 19 things of that kind. So I think that the temperature time 20 profile is something that one could play with,-but I don't
.- - -21 think it's critical. I still think water is the critical I
I22- ~ issue.
23- MR. FAIRHURST: Thank'you'.
B i
, 24 MR. .HORNBERGER: Dave, .then, just to follow up on 1 25: that, you didn't take. exception with Joe's curve time of .
' :[ ANN RILEY & ASSOCIATES, LTD.
~L? '
Court Reporters
- L 10251 Connecticut Avenue,'NW, Suite 1014
^ ~~
Washington,'D.C. 20036-
'(202). 842-0034 m ~
h > >
r I 1
563 1 vulnerability.
(,,,) 2 DR. SHOESMITH: Oh. I think that's exactly the L./
3 kind of process which should be undertaken. It does what 4 Roger is asking, which is it defines limits. It also gives 5 -- if it can be put into the right frame work and explained 6 .to the layperson, it gives a sense of where you feel you're 7' in danger, and it shortens up the period. So it makes a 8 million-year extrapolation look like a thousand-year 9 extrapolation, which is what -- it defines a critical 10 period, which is what Joe would like to do, and I think 11 that's the essential way to do it.
'12 MR. HORNBERGER: Yes, but as I understood Joe's 13 cucce, it was contradicting what you just said. You said 14 that water was important and temperature wasn't.
f)
Ts,) 15 DR. SHOESMITH: Oh , okay.
16 MR. HORNBERGER: He was saying that water wasn't 17 important as long as temperature wasn't high.
18 Did I misread that?
19 DR. SHOESMITH: Yes. In fact, you've picked out 20' an-inconsistency in what I say, and I would have to draw 21 back a little bit and say that the combination of the two is
- 22. important, yes.
23 MR. FAIRHURST: -If you could get rid of the water, 241 you needn't worry ~about temperature.
25 DR. SHOESMITH: Exactly. I still'think water is-v'~x
-ANN RILEY & ASSOCIATES, LTD.~
)
'm/-
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C.'20036 (202) 842-0034 l .'.
I-
564 1 -primary. If it remained dry, then you wouldn't worry about
(~3 2 the temperature.
4v; -
-3 MR. HORNBERGER: Okay. That's obvious, but is the 4 converse true? If you can keep the temperature lower than 5 the T-crevice I think is what Joe used, is the converse j 6 always true?
7 DR. SHOESMITH: That if you keep the temperature 8 down --
9 MR. HORNBERGER: If you keep the temperature down, 10 then it doesn't water if water comes into contact.
11 CHAIRMAN GARRICK: As far as crevice corrosion is 12 concerned.
13 DR. PAYER: Well, almost all corrosion processes 14 get less severe as you go to room temperature. I mean, the f^%
i ,) 15 accelerated tests are all done in boiling and that sort of 16 thing. But I think from a practical standpoint, a little 17 -bit of pre-aging isn't going to do you a lot of good. It's 18 this issue from when it goes from on the order of 105, 110, 19 120 centigrade down to on the order of 90, 80, 70.
20 So, I mean, if you could really get the stuff to 21 room temperature, then there's a lot of things that go away.
22 I mean, even sensitized stainless steel doesn't stress 23 corrosion crack very well at room temperature, but near 24 boiling 1*. goes through it, you know, in days sort of 25 thing..
' 7m
- ANN RILEY & ASSOCIATES, LTD.
s J Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
' Washington, D.C. 20036 (202) 842-0034
565
- 1 DR. SHOESMITH: No argument with that point of 2 view.
(
3 DR. PAYER: I think when people start saying, 4- well,'let's cool it a little bit or ventilation, you're 5 talking about a long-term, I think, or I don't know enough 6 about the loadings and how you get the temperature. I just 7' don't know.
8 DR. SHOESMITH: I think the way to view water is 9 that there is no transport medium whatsoever in this waste H10' . vault unless you do have water. So, you know, you might ill corrode your -- if you never put any water in there, it's 12 just going to sit there. So perhaps I oversimplified the 13' corrosion issue, and I draw back a little bit on saying that 14 it's only water that matters. I agree it's water and
( 15 temperature. But water is still the critical feature 16 because it's a transport medium. So for a corrosion 17 process, it can move material rather than leave it in place,
- 18. and for all the radionuclides, it can move it. So I still 19 'think water is a critical factor for the waste. How many 20 ' turns can you make in the same sentence?
21: MR. HORNBERGER: But again, .not to argue this 22 interminably, but if it doesn't corrode, then it doesn't
. :2 3 - matter that water l flows past it.
I-24, DR. SHOESMITH: Exactly. Yes.
- 25. DR..WYMER: Have we run dry?
l 99 A/
ANN RILEY &-ASSOCIATES, LTD.
s Court Reporters
'1025 Connecticut Avenue, 1&i, Suite 1014 Washington, D.C. 20036-(202)-842-0034
3 w
566
'l -DR. STAEHLE: We're also at low temperature.
b(? 2 [ Laughter.]
j 1
3 DR.'WYMER: Maybe this is -- we'll ask for one 4 more round of comments and then we'll saw this off.
5 DR. PAYER: The focus of this meeting turned out 6- to be an awful lot on corrosion and C-22 and its behavior, 7 and'we have'had some very interesting talks on backfill and
-8 a lot of comments about that. But I think we' haven't 9 talked, or at least I haven't heard the analysis of that 10 backfill, and again, particularly, how do you get it in 11 place, how do you keep a barrier if you were able to put one 12 in place over this massive distance and all, and those -- I 13 mean, those are the issues I have.
. 14. There is no question, if you could do something
() 15 that would divert water,~that would retain things and all 16 that, it's an admirable goal, it's a marvelous thing, but I 17 .seeffrom an engineering structural putting it in place, 18 analyzing, making sure it's in place -- I guess I didn't 19- hear much of that here, and I understand David has left or 20 he would get up and tell me that they are analyzing it, L21 evidently.
22 [ Laughter.]
23 DR. PAYER: And they are', I'm sure, and NRC is
- 241 .probably looking at it. But I haven't seen --
25' DR. WYMER: If-they had more money, they would be.
[
A-. . '
ANNLRILEY & ASSOCIATES, LTD.
~
Court Reporters-
-1025 Connecticut Avenue,?NW- Suite 1014 ,
- Washington,cD.C.J20036
~
-(202) 842-0034-i
- n m.
[ 567
'l .DR.-PAYER: I haven't seen a lot of that really
() 2.
3' T being done .cuf put _ into place, and it's not going to be a simple _-- you know,' one of the things here was to-identify 4 research needs and experimental needs, and I don't know how
-you do that very well over long times. You can demonstrate.
6- :they-work if you put them in this way and they're in place.
l7 MR. FAIRHURST: There are, as far as understanding 8' 'the characteristics of backfill, full-scale experiments l
l- 9 going on in other. parts of the world, but are you asking how
~
/10 you put this' stuff in?
l11 DR. PAYER: 'I'm just bringing it up.as a topic cf
- 12 something on the list that I think we' haven'tfdiscussed very.
13 much here at this meeting. I'm not saying nothing hasn't 14 ibeen done.or it.isn't in perspective, but we didn't, I 15; sense, probe into what needs to be done and what are the 16' priorities in that area nearly as much as we did for 17 corrosion.
' 18! ,MR. FAIRHURST: No, because that was a deliberate 19- choice, too.
20 DR. WYMER: Yes,.'it was.
. 21' 'DR. PAYER: Okay.
22f .DR. WYMER: We' simply had to truncate what we were going to'get'into; but it does seem like it's an area that 2
24 shows promise, at11 east philosophically if not practically.
254 MR, FAIRHURST: 1The-impact problem goesiaway.
5 ANN RILEY: & ASSOCIATES, LTD.
' A Court' Reporters . . .
T L1025 Connecticut Avenue,LNW, Suite 1014
, LWashington, D.CL 20036
- (202)' 842-0034.
l 568 1 DR. WYMER: Yes.
/3
( j 2 MR. FAIRHURST: A few other things that it could 3 help, too.
4 DR. WYMER: Okay. I think at this point, we'll 5 shift gears again.
6 If you notice, at five o' clock, and we're going to 7 declare it to be five o' clock -- that's sort of the 8 political' approach -- that we go into ACNW discussions, and 9 I think that that's what we'll do now. You're welcome to 10 sit here and endure this, if you like.
11 At this point, John Garrick takes over.
12 CHAIRMAN GARRICK: The Chair thanks you.
13 DR. WYMER: Yes. Thank you all for coming.
14 [Whereupon, at 4:46 p.m., the meeting was
'(O
_ ,/ 15 recessed, to reconvene at 7:30 a.m., Friday, June 12, 1998.]
16 17 18 19 20 21
~22 23 24 25 l
/m) ANN RILEY & ASSOCIATES, LTD.
\/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 0- _ _ _ _ -
REPORTER'S-CERTIFICATE
.This'is to certify that the attached proceedings before-the United States Nuclear Regulatory Commission in the matter of:
NAME OF PROCEEDING: 101ST ADVISORY COMMITTEE ON NUCLEAR WASTE '(ACNW)
DOCKET NUMBER:
' PLACE OF PROCEEDING: Rockville, 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. .
AJA Mark Mahoney Official Reporter Ann Riley & Associates, Ltd.
O CHEMISTRY CONSIDERATIONS FOR RELEASE AND TRANSPORT OF RADIONUCLIDES FROM SPENT FUEL D.W. SHOESMITH*
Whitesbell Laboratories Pinawa, Manitoba Canada ROE ILO O
- As of June 17'",1998 Department of Chemistry
- University of Western Ontario London, Ontario Presented to the ACNW Working Group on the Near-Field Environment and Performance of Engineered Barriers in the Yucca Mountain Repository 1998 June 10-11 Washington, DC
p e im o o 5 e i c
s n
o imtno ear t
a ta r
i s
r 1
a u e h it n it l t
n v 8 it a r
o a r e
f s e n 9 ir t
n a as n c o 3 ity " n w tn ta e
n ai e e c e 2 e n a mo me e r n d io s s
c t r c 4 u P ne te c is 1 n h n p B e
aW t
- co e i e de tr d o i t
' ca m'r tg e-ph l
ts e o a r R m u hu "* sf r s n oo f c n c u d i o lc p re g
ms u T gfrM u e t t d f l c W a t
- r a
a r ev
- s m l f
i n t
a n
r ip a e d u u s e e
ts p ta e n n n h ra N ta o oe p i " mfaiu d u e -
a p r a o ops id o o r s o r 6
! t l t
r i o c i af l
l 2
di t
ml a t t
- 0 l b i i o ca r s qf ic r n t r t.
a it e s a a Zi e a e
9' iAGl a o in nP e
' i. n r n T t
T R ' ,pCl I P i R R S B f l P I I I P P n re F, S o!
]
! 8
- C O J l
. , , -, , U p w,i mD L
C i
i 8 X q q qf ypIC I # *' A A MC B . T T R t t t
. S t o O a
r o .
P u
C n E g o m^
n g
t 1, D_iddai J Dd aar l g i O A. c e l
C C d f
n ,
l 1,
ip o
.l! {8 'l illl.' l8-A,
, ]
~
,l E C T n e
g n *r " mn, t
r o
t r Ut r o -
it -
d tn G pia o I
~
r o p op L o a -
e I E *u a Osp lu 6
I mI c t ~
oi a sn M M s y C t
e a e t c
/
n P D a pd a I f' Pf a I l n yV Nanr f P lac I -
3 o NP I e- d r I l
E a i R
Odo Qr -
r Rt sa r t r -
r A e g E ena t
i t a gS F Z OtZ e -
v n
e Wad s e
" wdeB E U C_ S s
o w -
Z D -
S T y '
kra m' -
". a8 llI*
s!:l ri ; ;l,L{5 C
, I I g Q- S 4
a f,
o X
A g F. P p' g C D
I R
V B
i n
h
- g,
)
g y
g ka t A
i o
, lo lelor 2 lea w 2 c ,t S er w o Hs Mwr I
e Gs clo o o d
r Ta cd I G % p l
i h n F s y g Vf lE f ns l
Np s P u y f
- h U t h Ut -
Z U a t
F Z a Eo
, o m
Ni o m f r Oi Td U r O Tu S r t GB i
R d
S r r e e m ht ht T
D ,..
/ - .'
- , s
/ -
6
[- q'. -
p m 2 I
n l
, ei c
s I w i_ n.
i Fi a a t
u '
-Glota i f OZ - H bi l
N nl i I uf n 1 i
- T1 ac.
7 I
t O t
O SOME OF THE PARAMETERS REQUIRED IN MODELLING OF WASTE FORM DEGRADATION WASTE PACKAGE
- Juvenile Failures @
- Distribution of Failures /
Location of Failures X Dimensions of Failures X
0 -
Pit penetrations Patch penetrations V CLADDING
- Immediate Failures Defective bundles /
- Stainless steel clad /
- Creep Corrosion f Corrosion Failures Long-term embrittlement f
- Localized corrosion failures ?
continued...
4
)
'\
O WASTE FORM DEGRADATION
- Specification of Water Seepage Volume f
- Specification of Water Seepage Chemistry /
- Fraction of Area Exposed p
- Reoxidation
- Grain boundaries
- Fraction Wetted [
- Fraction Contacting Seepage Water f Intrinsic Corrosion Rate /
Parameter influences Sec ndary Phase Formation V 0 -
- Blocking of dissolution
- a-radiolysis effect RADIONUCLIDES SOURCE TERMS Instant Release Fraction f
- Cladding gap Grain boundaries l
l
- Release from Corrosion Product Deposits
- Solubility limits 7 Retention factors
- Colloid Formation 7 l
l O g. ,.-
~
O FUEL CORROSION AND RADIONUCLIDES SOURCE TERMS
- INSTANT RELEASE FRACTION MECHANISM OF FUEL CORROSION O - INTRINSIC CORROSION RATE Burnup
- Radiolysis
- Groundwater composition Oxygen
- ACCUMULATION OF CORROSION PRODUCTS Influence on fuel corrosion
- Influence on radionuclides transport Incorporation of radionuclides
- FORMATION OF COLLOIDS O
Q Ridnase cd Wh from used Fuel Eta 1=====i *M en exposure to youndwater
=
3 O 7r
.m i - -
m ,,, ,
. sh sm V "
produces .b tTe. nu, M., N1 t: bubbles 1 3 .
L '
',- . t -
+ %r -
q
}\
\
Fission ,
h mp C. .5JI y , i sp oi s
- o in md . '
soeueton nl.Am,udoz
_/
\ -
I Phdeseed instantly on enposure to groundwater Role slowly at a rate controlled by the dissolution rate of 9mm fuel ametrix a
... i .: r. . .. . . . . .
-::'^3;yL6 LAO'O.ING'. GAP lNVENTORY
,v . . .,. .
- J r[Unn> sounoenv m,v.,w.w A,envony
, 4 . , -pe. *
- 5"
(-
Table'3. IRF Values fatKey n=&aanrih IRF(%) IRF(%)
Nuclide T %(year) best. pruimictic Rationale estimate ' estimate C-147 .5730 ' '~ r " 1D' -see. text Cl,36 3 LO* 6- 12 see text Co-60 53 -
- acavation product in Zircaloy e3=dding Ni-59 7.5 10' '- - acavanan product in Zutaloy eladding I Ni-63 100 - - activation product in Zircaloy claddmg
.Se49 6.5.10' 3 6 Cs:Se assumed to be comparable to Csl in volatility in fuel during irradiation (Cubicciotti
' ~
and S=naeM 1978).
.~ '
Kr-85 10.8 2 :4 see text
$r-90 28.5 - 0.25- -1 see text 2:r-93 1.5 10' - - mamly dissolved in fuel mams .
Nb-94 2.0 10' . - -
mamh dunolved in fuel numx IEc-99 2.1 10' O.2 I see test P.d-T07 6.5 10' O.2 1 eBoyed wish Tc in metal =*-aac Kg-198m 127 3 6 present as metal, sonnewhat veistile dunng inadission(ch and Senecki1978)
O ;
cd-t'23- '46 3 6 i. -
- amassaism(ch and Sanecki 1978)
> 'ii a=r-s Sn ~1.26 1.0 10* 2 4 zelarusty mouvoistne dmung nradaation (ch ad hnerM 1978)
I.129 L610' 3 6 see text Cs-E5 2.0 10' 3 6 see text i Es-137 30.2 3 6 see text Ben-151 93 - - y.a in solid solution in fuel mamx Bu-l54 SJ - - present in solid solution in fuel matrix
.Ho-1_66sn 1,200 - - pitsent in solid solution in fuel matnx hotandts 1
- - present in solid solution in fuel marnx ToHNSON; .l . S. Angf '$. C. TMIT
,qq.N
= . ,' - [ .f'.f.y9akd suc6We fem
.sged$wt . ~
[
6ka YiiChicil 'Rejaort 97-l8
.h -
o
Di .colution end tho chsico of cn cppr ppte mad:l l
l O s o o -----------------------
Dissolution blocked by 200 $2EE"0*9- h_a_s_e _o3m_a_tyn_
A Accelerated by Model 0 ---------- --- g h 100 - - - - - - - - - - - -
Li it Mo el
-300 -- -- --- -----------
Oxidation of Fuel Surface
-400 -- - - - - - - - - - - - - - - - - - -
l
-500-------------------------
Corrosion Potential O (mV vs. SCE)
Solid interface E
h<
Environment M
l Eoxmeo a
' Change in E as i Oxhssnt Conc'n i Decreases I
. . t Ecom 4, ,
Change in E I ;
as interfacial I
[ CORR '
]
s ;
Cone'n of U0l+ 8 !
increases O se :
I --
w.
l i
l - - . . .. _
..______._________________________________________j
l l
I
- NO SEEPAGE DRIPS O Failed Waste Package ma
~
-j m v~.<.;m;y7;;,b;;rg m al t?? 4 yr; 7 , ~7 s;y. p.-;.
. m, ;T; .
5.
4
. . c: 9 Tf8, f.ity r e .' m.
w-Humid Aerated Vapour High temperature oxidization l Grain boundary oxidation leading to pellet breakup, i by aerated vapour cladding rupture r UO2 UO 2.4 (UOz.33)
UO2 U O z.s7 m
j F
g er we mse....ee G
V d4484#+# 4 na4 6# 46a ea Y T L'1 *
~Q
- ,*.g54*sse5:1.*u: N ,
Formation of secondary phases FSA - 150 v
Should only occur for package U O2.4 r UO3.2H2O with juvenile failure a -
- ~ . ; f. * .;*p xz ggh v.ap g .~='
w- .
D u m a ..
8 459 Q 3 .- ' 7-y A
2 . A E D .a
- e t w issa c 3 D las Solubility limits apply Diffuse flux
SEEPAGE DRIPS O
Failed Waste Package .o.m
~' [
$7I@ikf[
'1:1. . g . . .. h
. .,u.%m.
A.
Seepage Drips
[CO3 ] 510 4 mol L 1
[CO3 ] high [ Cal, [Si] low [Ca], ISi] high v
Dissolution without Dissolution with secondary secondary phase formation phase formation UO2 r UO2 (CO3 )I UO2 : Uranophane (DR),w = 1 (DR),.i = 10-2 FsA-15 FA*1 S
Should only occur for package Radionuclides solubility with juvenile failure limits apply Diffusive / convective flux Diffusive / convective flux O -
a I 4
L - _ - _ _ _ _-________ _ - _- _______ _
Und:r oxidizing canditions tha Ancase vissoluusn1 of Uranium Dioxids csupies Elactroch micrily to
% tci.sction of availablaexidants gJimmimasumisme:d:_4 x*N mtsk m? we y m
' )
l $$ff .)$)
.g :p a q wa, .
bIkwhe-u h v ,d) h%;; .
h opw E bm sS.~a':;r
- 7;st?p agttS.:>
xa:ww% , n gj:yeRW gp pibuy waeg%w&g% s an g -
eQp=gggw;naf 6m as y, ~: ,. [ essounTs]
eamy % . ..
(i&Q%a@J.%Q2 - :~ Je$W.,
3y ~
S~ h s@cesw @ 'Jy $ f M ; M Environmental l Enrmm E $ ewn s.EiWFP; oxidants n .g h'4 m n ,wA;;oz sun #:RCE l 7
on FACPER.T(ES *
. {j&YR55"d&f % 52.1GTN b. n:
p;
,a w%%:tkiv;p-f : T># :y 2,"
- OMS *
%? 1
. ,Q. MkDjf2M~L ~
pp,g .. .3 CSC'L,::v: X .* V '..- O. m&G.m Ew These two reactions couple at the Corrosion j Potential, Ecosa Over extended periods of time,the evolution ,
in fuel dissolution / secondary phase formation processes should be similar to the Paragenetic sequence exhibited by natural ore deposits.
- s. l .,,n..., ,l
.,g- , .
Environment
@, .fi$M9seci ingjg;Q
$ W M @ @ plemoedF phsiss? %
ppuwpea# 4g&@f2;; M w;4 M gn
+1ong&a+%gyF p.mS p a' y WRW ,a vm M kjfk74 ,3 Wi 4 (RNieds (RN) ppt ca2 +, slot 888" * ""
f coHPos! Tion j
W I (UOf).
y ad ;- SiO 2OH12 5H2O) wng)g,;; . R;q rm gy
. .n u%< :,~nse
- g r
. RD N UO *22 m
kb/Gg44 $.pu %q _ 5% .-
d!ijd % h k% pf_ d )$ Nil @_ _ h dE % $
1.m
- - - ==
w- -_ __ _
SURFACE AREA OF FUEL SUBJECT TO DEGRADATION PROCESSES 1 HUMID VAPOR CONDITONS (No Seepage Drips)
For one pellet within a failed cladding .
Grain Boundary g,p Fuel Grain .
Grain Boundary ' - .,
Release ,
F .. e
.Ihl[
[ ! Zlrealoy Cladding "o.
}ii
?!Y r;'
s ~
g.'
'J hl.~.'
c- Pellet
> Interfacial Gap g ., *g l
06 -
Surface Area = Geometric Area x Surface Roughness Factor (3-5') x Cracking Factor (~ 10 to 20*)(assuming pellet breakup prevented by residual cladding)
AOUEOUS CORROSION CONDITIONS (Seepage Drips)
Surface Area x Grain Boundary Factor (~ 15")
(assuming penetration of 4 to 10 grain depths during dissolution)
From electrochemical measurements From spent fuel particle size measurements From surface area measurements l
O -
L____
1 0 '
INFLUENCE OF CORROSION PRODUCT DEPOSITS Fraction of fuel surface exposed at the base of pores is given by the deposit porosity:
.2r; x x. n n. ..
Fsec,non or usn n>
. $Fi. 3: .%x l:i17...: elf ,Tl.: ..@ *
- 55= " T*s anse I.?!.(.:$,6illi'j$1:!
@i
- .. : . 9;.. f.. s,:s
&N.!!} 1
$ "j"e 'c"mN osiryy oEFoerr y
. . . n a_.c.u;.:.1 A[= L N Q*
l l
INFLUENCE OF CLADDING FAILURES S
i l
l l
i d l i
l l
l 1
~___-~.u w7 e - m uy
~7 O, Rsouc.rton.. IS VERY D6PEHOSNT.'ON~ THE. PROPERTIEE -
~.J .
~
0F THE 'l FUEL -sonft9cE C PREcoN/NkMT4.Y, NON- GTQ/CH(OHETRYb R-B+R+
. . .e
. somos IVM H Y "
. 4WSES /N- R EAC TO R* 'AND /N - REkosJTDRY WILL. INFt., !EME WE KI N E TICG of THIG CRrMOD/C. '
REACTIO N \
\
.A g C OW W O
E5PosORE TO RnME ERATH FORMMTION 0F THE AMitMYdo * .SoCUYtONS. DOPIN& DUE TO [ ~PM AY 'N ~
..u..n:rs r.. . _
p, p, y
."' sWibb. Y M WM
); *
. 1 CAiHbNNS* ' ;
N .
'Skr$bYYibbYST ....~-.. .
.; 7.;;.. .
' *'i ,. .g.; y.:w ??" #
(-Te$t) 0 .
i
l TN2: MESENCE. ef ' THE E- Phrf5 /Nrnocct.Es TIYE.
~ ~ . . ~ . )%$$1gluyy 'OF A spODR[ D/995tENG IN ' TNG l AELMT1VC QCMinostok RATES or ud, RND SMNr i' FU4 .
okr
~~
.q .
OW 02 -OW '02
.OW %
. East. Slow (bidanna J -
1 M-- ;,. k..... . +
~
Q. 2% pf ,, -
-[.
h +f.y.<..'i Mj>#73 , fp MTh(4(I'kf:g.'m,y.r, y .
y','
...;'.h* . %9;.; t . .--.. ; .h' y,',,.,fd,if;
.- u.
, a W~ tu og
.~{-
.. 7.'. .s ..
- 9. n ' * ' f'ra f'?gjg ...../p.4,;z4 GC2$. ,l ,.
'.0.,,,.**'A:.*f.
- . . < , ~;f,jgi;g'
' ' -- 4~t y, -, . wp
' T'
. .k%s? l 15 e ..
L 0' r.~*$: . a59 h,n t';Rh?f".?.Q..
.. . . , .A, pg]E .N 'l,b.i, m __.
.gR 5 Q) . -;~ .:,$,*hN:$Qy .. p. -% . . .
s .ag Hc w2VER , THIS 16 AVDIDED G Y Po95Siv$7/0N OF THis NOGLE HETiet PHMSS UNDER 01-to12/NC CONDITtDNS
- *THE FORNMT/DN of /NSO/ ATIN6 CDRROSAON PRODUCT DEPCS/TS ALDCh5 5t)$tfRC.g dMTMLYT/C S/TL*S .
02 Og h \AD2 g \
/ /0 1
/ / /'
s // '
/
/ / Qo n H,o .
Y~ ~
j'/ r >.
D,
- e. t. . e .
g ..
- m at:.1 -~j y (* r . ::.g. -
- ~
- _ ....f- ,..
!, . w . , .. .-s
,.. s...
- p *
- j..=,,p
. m;, -~? .f' .-ca..
?
p n . g, g.= s.,
.i- n. ka Lv.. .*
bW 11lt PtawWrrow of caeecsmo ~
Proott r neposITs .
ZNFL UENCE ' QF $ROUND WMTER COMPOSITION
_ soooi Y
! SYu8to#8
- #18
'q. basemos Team Dessesse MC easc Toeparehoe Dohmesed Water Ce + 8 l wowr
)- 10 memo, wano a l' ! 4 10 j w::
I 1 1
0 /
g munoo .ca, o.1 ! a 5 - - - - - - - -
0.01 -
0 20 40 00 80 100 120 Days FIGURE 1: Uranium Concentrations Measured in a Flow-Tluough Emma usst With Unirradiated UO2 Pellet Fr; =-- = (Wilson and Gray 1990). Air-equilibrated solution; 4-.nue 25'C (excqx when noted otherwise).
Ca. j' Si snouMD WATEa CONTEN Y SOA* MESSES rua rum. asnaasiou anre.
Boiling penod l O 95 c eerio 4 irhoe. years
'"" " " W A I0-200 M7a B l200-1000 C j 1000-2000 [eo ] f Cooling -
70*C Period (Time, years ! ,.
A 2000-4000 -% E I 1
)
B T4000-11000 j l EtPecT50 $6EPRGE WATER 30*C Period Tsoe, years weo r.o a e- Aow enn sexnTs -
(1100 4 100,000) mse ence,uu
( 6 ASS ANI WFDEE 4.:0 GalW12 sere (su 6. PGAGE Aias u. PGROS F, Jus u PenOE G (W6. PERES A (W h. PWEBE 5 airw. .
pH 11.9444 11.uins i 19.stiv IG.suss IEwams T.45e 1 e G. Z
- T W W wnusufR. m 5.57Ea04 T.M 05 4.74E6 1.1134E 1 TWW ARNUWRfR. m 1.EG) L13E6 5.73E 6 LEC LE = '- _ _ - -
O ;
summu -
s wu. . . _ _
. _ _, m
_,m
?. -6 4 3.GO 1-6 8' 05 1.51E6 A57Em
- 1. Wet 95 1.COl
- 1. wee 45 1.112 45 1.TE C 5.TEG i
. puuus "_ _ _ fR T.E m1 T.51E41 '"6 5.ZTE 6 g I l ?_^""aOS
, TWW - .m LITE 64 L31R 6 1.435 6 1.1W6 .11E 6 1.1' E6
1 1
FUEL DISSOLUTION IN CARBONATE SOLUTIONS l
O 1
m cambonnee
..- :y.. . a '
- $!Ap.k,;pw;p
- :;
-. .n . ,,
% uO 22+
e-: mmsv. .
.,_2 0 .
- ~; ~7 1:,4 g g.
il.; &g Ob S es
, u."
won os Corrosion product deposit accumulates - dissolution rate suppressed.
Low Carbonate (< 10-3 neoH1')
._u O "
l ,. &
.c L +
L ., l Vi n ,c;; 2' UOz -.2
- --> UO (CO 2 3)2 32 3
', .;,o
~y - cs; 9 65 kHz
' .x. : s. .%
y,- , %t ?
- 4,. ., pq
,+ s l}
- c l
'."I~ u.<,:Aae%'"
..uu l
Corrosion product deposition avoided; solubility of Oj+ increased; dissolution
~
~
l unimpeded. _
i l
95 0 . . . . . ..
.e #
FUEL DISSOLUTION IN CARBONATE SOLUTIONS Intermediate Carbonate (10-3 to 10-1 mol L-')
O
^
U02
[ ::,,
(UO2CO3) >UO(CO)I 2 3
, , *N!
4 - ',
19 w
., g: +
. 3 q;i.,1 7 ", h n x 4
a[ e yg@f
- mm. s mam g CO[directly (kineticaNy) hved in the dissolution reaction via the formation of a surface adsorbed reaction "mtermediate.
O High Carbonate (>14-1 moHJ)
, -,y.,,,,.y
'+c -;g ;,
rW g..
e Ma.:a '
. , , ' . U0.2m%_.
1,.,,, nr O2 CO3 )2--> UO (CO 2 3)3
, . ,5) V t
, ; u~ - n qy.:
. m s,g fh
. ; ,, g a
. .a -
Zn Uranyl carbonate layer present. Chemical dissolution of this layer controls the rate O ,
4 j
1 l
EFFECT OF TEMPERATURE j O
The value of activation energy (EA=/)for UO2 dissolution varies with environment.
Non-complexing neutral solutions 29 - 34 kJ.mol~'
Carbonate solutions 42 - 63 kJ mol
Acidic solutions 50 - 67 kJ.mol
l Non-Complexing Solutions 9842307 -
_ -v 7
. . ~ r a=;5zh
{. 2:y&$$
UOj' > U UOj+
UOt -> U'O ' Increasing T'C $(UO t J , :;g g->?JI"
" @mt@
, .. e Q l(VQQdi
. r s _, r
- s v m as.;pu
%%s. -a axar
, l y Q::;f7...f ~
- , . ~ _ . . -
m ,_
UO3 .2H2O Thickening of Surface oxides prevents expected increase in dissolution rate.
Carbonate Solutions
-- n ni-
>;; w, .g o .nm,
.e .
.t. ..g;tu 2p .
w.*; .a. v,sg ,g : q .w m l UOr---> UO > UO2 (CO 3 )!'-> juOz.gUO UO2(CO3)!"
i o- h renMy wen.c q
, . .. :n e., c ,
. .. . eww-- . , ,
ner
- m. ;,,. 2q
- i. m%ug:q.,
unst.e x-No influence of increasing film thickness.
~ - - _ - - .
g 7,
~ _ _ -
10 3 r = 13.0
- DRast hLow yNI TWoo06H g3 102 , NTS ]
s'> - +
3
= 1.5 ' O Ra
.. . .(. .D. ..R..)-y..=. 0.. .. ... . . ..........r st.......................................... g)(2.).....
g ,o, , .........
(*) o r = 0.22
- DR*" {
E
! e **..
si li 100, O {
O"
. r = 0.33
- DR"88 CD R), - o 3...*****,_ e m cenpmig 5 "l*WNNHfH'h?.be67HH//N/W/HL/f4/H/H]/HH/H/HH/.H/.f.).'y,h.
1!i 10 1 , (a) e _
= _ _ _ _
5 Et.ecTn ocH nM tcAt E
b MEASUREnattis
.9 s 5 10 ,
a, y g y gg gfE4(T FU58.
$ 'g :
---O .
snosamen ano sunser(21(HcooNa) snoemman ano sunaw(2)(i.enannon IOp h
M h 08HH,SoJ 103_- = coa a.popco j r = 5.6E 07
- DR"77
+ cen n u nen e .em Tantwomen ar aunsuHcop @)
to" . > m 0 2 8 102 10 10 10' 10 10 10' 10 5 Dose Rate, DR (Gy/h)
O Fig. 3. Dissolution rates (r) of uranium from UO 2in irradiated, aerated / oxygenated solutions as a function of the gamma dose rate (DR):
(a) predicted using the electrochemical model of Shoesmith and Sunder [2], in aerated 0.1 mol L NaClO4 (pH = 9.5) solutions with 0.01 mol L HCOONa (e )
or with 0.01 mol L t. butanol (O);
(b) measured by Gromov [19] in acidic sulphate solution (pH ~1) (D), and alkaline carbonate solution (pH -10)(m);
(c) measured by Tait (unpublished results) in 0.1 mol L Nacl and 0.01 M NaHCO3 (pH = -8.5) solution (A);
(d) measured by Christensen et al. [18] in oxygenated solution (pH = ~8.2) (+).
(Note: 1 Gy/h = 100 R/h). l The horizontal dashed lines show the rate predicted for unirradiated aeratr.d soluuon ;
(1) (from Fig.1) and the rate measured in flow-through experiments in bicarbonate solution (2)(from Fig. 2)
AFT 6A ~
300 years CF WR$Tiii PRck-196E G
INrsCatry , (3l g- ts N e s t.tS t B L E .
-- ~ ~ ~ _ - - - --- _ _ _ -
ht M 1 i e o.1md. 4~'Na Cfo4(pH =Ss)
I 0 -
b v
-7
,i -1 - I n
- E e (DRh =0 DERATED}
N hWN*- //////$//O///8Y///////////////
Is '* e ee P
i
- c
,o -3 -
e
. o g o
I / .
en 4 . es o
" ...............%...........e,,,,,,,,,,,},,,,,,,,,,,,,,,,,,,,,,
4 -
d :
1 Ioy 10y i i 4
3 2 1 0 a - Source Strength Log (microcurie)
O THESE RATES PRE GAsSO ON ELECTROCHEMICAL.
HERSOREMENTS IN R THIN L RYE A CELL
.,. .,(
, .e ." ,
o L ; ,2r y dp Jf j;..
.t .e;-
\ QAyQ__ _ -
l t f
n N Juo,,*
D****p**#*a [
Or T De=seunen Joo,
/ i N* UO'* #
HO 2 2 Dilution into OH- - Large Solution M lh = w Hr0 .02 Y"**
p Jn,o,.Jo, SS.^JIJu,o,,Jo, u / % ,
+g "ging}; ;a a.w,%. \ )
/ .
M.%.9h h .# '#"
qi
-~
g 1.: -W:&ae J .. h^A.TY-%_ i . .. .~g' !
-, .D DCr i.
$:,2_fb 4; w-en,
~. :. c -p. .- *r.p4.
. < -u _
l l
O itHOS/2 doRRO5 f0N PAOD UCT DEPOS/TS.o %IHERE O fffts/UE Lo3SES RAE MtAt/M/2.SO, THE PRTE t GOULD las HifHSR 1
~
- S THEAY R RSRL (30RNUP Erfsc.T 'f O r m Grain ~^r^i=ri gep r ~
- CNSTRNT fR E L E R G B.
Fuel Grain p gggggy Grain Coundary ( Shor8 '/t!!fW)
Release Zircaloy Cladding l
ZNCRERSEO SURPRCE Pellet RRER
_, interfacial Gap l7 108 7, 107 :- 'g '
L :, ,
r- "'
x '-
( )18 a A/9O/O AYS/5 5FFSCT.
y 'd b ..,.,s,i,, ( swg gy Ij Mf i 'p\
a . .. \
git' :- ', \
5 108 :- 7 'i \____....--- ^^ -
1 '
., M'fPSRIMENTS ON E to -
' ....,_____..... -~~~~~~~'
SPENT PUEL DOMINRTED 10e . 84 SHORT TERH EFFCc7;5 .
1F 10 1 02 1 08 104 1 08 1 08 -
- va l
Time (a)
Run ~ R SF ~ R 4 R Ms 0g g tro,,y Peu.er ansmeup
@ S I
A 60H 8/NR T/CN Of rad' Cd ARD/RT/QN @uLO 3E P n R T ic u t. M R G Y A66RES GIVE ow M **1 - a + e.n OW W
,g-- ,
ow I 4 di
, . : . : _-. - .u.
d 'F f4,:M_Qp"i.J gg;.3. g g .p 4E .g
- , .-r ~ e; , ww. u . , . . ..
. La_,, m.
. lL. ,* ' ' ~^
- * ~I . u. . , & j.sh ' Y a, .
M AR'N ENNMNCED m4 R1901cA4 coMC'ns AccRLERRTE i
AA M ! $
c- _I W+r i A
O $
s l
l
? Y otl 0
StHti L R T E D a
5 E- .
M l u n l 0 l
' i e \
2 - . I E 5 i i
. I t i c e A Je c.: c-e
' l A l 1 I I 20 40 O l Time /h Figure 3. Corrosion potential of a UO2electrode in 0.1 mol de-3 NaC10, solution, pH-9.5, containing 1x10-6 mol of H302 , argon purge, gamma field of strength 35.1 Gy/h was introduced af ter 16.83 h of corrosion without any radiation field.
O 20 L'LO KYPLR/N VERy R&GRECstyE cuppoSjoy QBSEAVED IN SPENT POEL DR/P RESTS
L i
MODELS FOR FUEL CORROSION O
i (1) l [CO ]3 + m3 log [O 2] + a4pH + as log [BUj+ E a,x,xj log R, = a, + ai T + m2 og 8
[BU] - Barsup a,to a3 - constants to parametric fit of data from single-pass flow-through experiments
- interaction terms describing interactions between E
parameters (2)
Ro = ko f (A) f(T) f(GW)[CO ]P 3 [O 21' O u, - i.t e dissa .tio. ra P, q - experissentally deter = mined reaction orders f(A) - factor to usodify fuel surface area f(T) - Arrhenius temperature dependence f(GW) -
1 in the absence of seepage drips 10-zin the presence of seepage drips (3)
Ro = ko' f(A)/(T)f(GW)
Assuming [CO3 ] and [0 2] constant O
l O
INFLUENCE OF CORROSION PRODUCT DEPOSITS
- Attenuate the Avaienkie Surface Area of Exposed Fuel
- Retard 0 Transport 2 to Surface Sites and Inhibit the Kinetics ofits Reduction O - Confine a-Radiolysis at the Fuel Deposit Interface
- Trap Radionuclides in Insoluble Secondary Phases ]
l l
l l
l O l
CORROSION PRODUCT DEPOSITS SHOULD BLOCK FUEL DISSOLUTION AND RETARD KADIONUCLIDE TRANSPORT OA Aerated Atmosphere Fuel Water Film Jo, High UO j' Ja= Dade + vc dx
+.
t
++.&,
y 2
L-Dn 5 c m s-1 B
[ ? Fuel Deposit 9 Water Film Aerated Atmosphere
~
O n)) f N ;;' ,.-
Jo2 Depends on Film Permeability and s; .' 13 0 2 -
D Degree of Saturation j lb '
UO j+ l R N h ,,,
- Jn = Dn [- + uc
.a.. ..
_f y?: , ;
l f-llh ~y .' ,.. ,), ; ,>
i
>- S 9 l v .>: t e - ~
- V L (
mom - (
2 Dn 7 c m s-' 104 cm 2s-If Ro large and film porosity / tortuosity high, then (Jn)A - (Jn)s O If Ro small and film porosity / tortuosity low, then
- (Jn)A >> (Jnis C
.- - ~ .- . -
_g - ~.~ -_-_. ,
ZNTERPACE.
- =~ -
Corrosion Product- AqueousS lution
_w O .
u.. ,--
p.- ' <;, l;- .. :e.~
, ,,, e .. , ,. .A: . ., .,--
. 'p.F. .
' E <-,- U l *'
. 3.' Ei.7- ~
, .2 a .. 2.%
'NM _ -f?QQ a
~
.P I ... U
( . .. . c. t~ M t: ,rgt 1- e . _
_ E$ . . WA l
{5kyy, "
7
,fQ.
. . - - .nw w
. t Y .. , ,
? . . -
2 -
^ I'jp ,
y.. -
, m -s:.
I J
O O( - DOSE RATE RTTSNUA TED BY l*DARCS/ON P/2CD UC Y DE PCG tT.
So w7tCN IN P0RES IPR R D IAT50 THREE - OtNEN:lbNALLY ARDIOLYTIC O YIO RNT5 RETAINED RY THE FUEL SURA *it'E 005 TO RETMRDED O/FPU.s /vE LOSS PC)5 GIG iL ITY C F r:AINTRIN!N(- A RBD C Y FR&NT RT THE fcEL / DE PSStr INTER.cR CE .
4 O
4
ZH -rHE' A6SENCE Cf D/FfuGIVE 40GGES LOCffL OY tD AN*r 6;WC*NG DVB ~1"O }?LpftenF Aff0/CLYLIS CCM.O GE Hl6/4 1.00E41 rw.
.p+ *. .-.- '
. 'a 5, ' 4 ,4
- % y y< . ~ '
'e '- if .
'% , A' l !. . . -
-j + a [jypg.IP(60 Cf4LC'M OE42 .: ' n.4 ,hr', : - .,. ' . .e. d*; - . . . ~p
>m
?
e 00E42 .U 4.- ' . y ~' , :,- .;q p,_:- g
.. .- .- y 7.coE-02 - - 'E r+Mfw
, , . f(%' & Q, .] L')
- * * ' % ,. ? ~ }h ', ; f.~ -.;>- f'i'-
~
7 --, . is,./
-k '
6.00E42
\.i,
?p; y , 9 :_ .
< >' '>' -/
- / :
f 9 5.00E42 2,ht 5,
[' ; , :[.e# ~
Rc **
o H'
..,- 1- ..': . , .:
4 00E-02 t.N 7T: .f /
y =.c- T a h < . v#~
f~.} e ^' , ,_-. j , ,, *J . ' .
ss _
^
h*'** i ' ~ , $ .l . - - '$t,
^
, <. )
,'-A.~.>.
. ..' f .t
- 9 .' *, $ .4 . ,
,: .. w (,(g(g C 1.00E-02
, D. . Z$. @t
' . J. I[ N.ii
+.,,g_
L 3 - mc /. I ' ,
O 00E+00 ' _
0.0E+00 5.0E +04 1.0E+05 1.5E+05 2.0E+05 um t.)
2.5E+05 3.0E+05 3.5E+05 4.0E+05 4.5E+05 h~/ U. "
O 1.00E 01 250 mero-Ci re H202 d.composnion o.
,_ nog.n2 c \0~}t f am
- Ca >0s. . . * * , , . . * * * * *4*- * *) )
OIPPUS80N FRCH s t 1,T,,g.* : . a e * *** * . % . . to-G a<~ j g oggpc,ggygcy
,a e o ir7== 10-7 ) ,,
{ , .a* 8 o, is (gi E / .=**,.. . . e+ea. e ..ee .ee+ ee+ i J
! 1.00E.03 ,a
- D I FFOStON FA0N 8
E -- _ _ _ _ - - -.,go.W.m-~_____ p syrpLL.OW SITE lC o . ... - .s 1.00E 04 /
open ermboes, m mern 5 I
, soiki symbods; eWnern 1 0DE+00 2.0E+04 4 OE+0d 6 OE+04 s0E+04 1.0E+05 1.2E+05 1.4E+05
{ h'-. -u r] v 1 0Hc$,n. '
- - ~ , . (
,+ e.
. w
O FATE OF RELEASED RADIONUCUDES umC93 00 Dissolving ' Unretarded N 8I
, Fuel . l::.
^ - M atri,x? ,. .
._4 .: m4:. ,
'i
- ~ g's.
v, _ ue r:
, Tr Partially
' ,' t TcH Tc0 " A 83Tc
^
-M;,. .. 9L . Retarded d
- ;Wf D FeX fee l
m- ygxple l
$;!JYQ[fl.y nii @ .
Id interaction with Carbon Steel Corrosion Products y ,, , ;,,; 3 y.3
^ '
.7
, y.
"+
, ; f'.y{ Retarded By ls. a .. , , .- i ,;' .. - Umited Solubility u.. .w .. ..
Colloidal
%s#;nJGjd, 9 w4u 12sPu pj " "' '
1a .; Transport
< e.,. :p
-... f &, .
- a (;.3 , :.
.. . }
J' 12sNp
, Q
- a. } 12sPu
~
i
\
~U L
? m 8
? h 9 3
C 2 m
4 hWP W
0 E R
k y nI Um 0 T W
r 0, N 0 o
!i d d . ,y% 1 '
0 S 2
2 t r d_ _.
0 1
T s
s, ofa c- 1 f' P
ur a c t f e S a i H
e f O
L C
l e ' d x. re k T S
a o n " ~
0 0
O P
u t a pc 0, miu@j-N d
i R- '
o n r . t
- 0 v e t s d. c @
^Wf-8 i
s n -
d n o km're2 okms r d a -
0 e D 0 I z$ o j m -
J 0 )
s s e2 '
jh(
y d d1 o
.- , Q!j' .
0, r a
r c-
&_f6 "e a usy ' b, t .
0 e
C a a h< by-w'ek l
g (
c y a
r e
s V h h
(
i d M *a e-
-- e e a dta '
~ '
0 B teP !dbMM p- 0 m O v eW'4 0, i A T cl l
R l* " -
0
" eA p
b 1 -
t o
e E
x 7# "
r p J'0 0 s y '
gN up 0 g' '32* ,
o 0
- 7'
- " i2 0
2 D
0 r
y 0,
0
~ -
I
- 0 '
0 . -
1 I ri r rg ! r!
i 0
0 _ - - -
2 3 4 2 o 0, '0 '0 1 ~
0 0 g 0 0
- 0 1 1 3 1 1 1 1 0
1
$E
- eTot emOQ i O
l
TRANSPORT / RETENTION PROCESSES FOR RADIONUCLIDES 96 @ 3 01 .
O ro.i .
l (1)
RX"'
(R v 'R 1, .
.r;
- o 2 i (3)
+
~
w
.{
. d y,[ (R)
, , (4)
(R)
~
- py .
O 9 05:9s
, ,wg l$R'
' c$
m.
R
.. M wo # N
~ fle34 a, ., 'f ,
w ,.gr u .e4:.
(1) Solution Transport (C, Cs, Sr, Tc,1)
(2) Colloid Transport (Pu, other actinides)
(3) Absorption on Geological Media (actinides, rare earths, Tc (?))
(4) Incorporation into Secondary Phases (Np, actinides, rare earths)
O L.
-v O
SOLUBLE RADIONUCLIDES
- Instant Release Fraction
- Cladding gap (Cs,I)
- Grain boundaries (Sr,Cs,I)
(1) Relax the Instant Release Assumption for the Grain Boundary Inventory
- Fuelleaching studies O - Take credit for corrosion product deposition (2) Investigate Tc Interacties With Redox Active Minerals, Corrosion Products
- Available experimental evidence is sparse (3) The Problem with the Present PA Analysis is the Rapid Wasteform Alteration Rate (~1998 years) l i
.O u - .- . - -- -- --
O 4
Fuel ,
Deposit ~ \
l a Slow transport leads to " anoxic" u O2 conditions at fuel / deposit interface
. s h
c:i. J
~
,, - ;u w i .
- - g -
- R Slow transport controlled release.
i'
, ;; . : -- G g'
_ , , , ** ~~,
h '*
9 W I
rc=a rco, O :yg; a w
rer een se - .
l l
l 3
O )
l l
l
)
3 O
COLLOIDAL TRANSPORT (Pu, Actinides)
- Known to te Formed
- Properties Poorly Characterized c"" 'd ' '" (')
0 -
- Is colloid foranation an early high re ate r scenario?
- Will the probability of therr formistion decrease with time as fuel reaction rates slow and alterstica products accumulate?
o a
I l
0 l RETENTION OF RADIONUCLIDES BY COPRECIPITATION WITH CORROSION PRODUCTS
- Am"', Cm'", Eu"' Concentrations Controlled by Coprecipitation Processes With Uranium from Dissolving Spent Fuel (QuiBones et al 1996, J. Nucl. Mater. 218,38).
- Measured Retention Factors for Various Radionuclides in Drip Tests With Spent Fuel (3.7 years)
O u ic 1 (W) !
I
Cs 6 12'I 1.5 23sU 333 237Np 15 23'Pu 3333 Bates and Finn (WFDEE #1)
- Alteration Products Incorporate Cs, Ba, Ru, Np
- Pu Appears to Concentrate on the Spent Fuel O
\
]
( Can these results,(i.e., retarded release rates in spent fuel tests) be used in PA calculations?
Comparison of Np Solubilities
-1 . . .
Range of Previous Distribution 4
2
~
/ g jpH = 5.9}
pH = 7.0
\ ti
- ' N H = 8.5 1 o Rany of e Revised Distribution e O - a a.i = ~ 2>
g . e [-] -m a,im r.e>
0 c d ea a i.es(pH.s.s)
T 0 : O m snm e O g ,
gg -2 g %i U+ M 'a8 * **
4 .
ma. ==
3
.mi l
m3 : 4 - a. im i., - >
w m a a. ms t.g im i G $) , , ,
, , A 1 % , ,, , ,,, ,, , , ,
V -10 20 40 60 80 100 y maa.imia nimi Temperature, *C Oversaturation tests probably do not yield equilibrium solubilities and are measured on phases which are not expected in the repository.
Concentrations from spent fuel tests are empirical and may be system specific. '
As proved to be the case for fuel dissolution and leaching studies, many years of data collection (in drip tests) could be required for an l , acceptable consensus to develop.
l l
- c. >
(1) CONTINUE DRIP TESTS O - Np shown to incorporate in dehydrated shoepite (UO3 . O.8H2 O) under humid vapour conditions. ,
- Not so readily incorporated in fast drip tests.
(2) DRIP TESTS ON UO,(NOT SF) USING No-SPIKED J-13
' - Fate of Np During Paragenetic Sequence
- Limits on Np Incorporation
- Dependence of Retention Factor on Flow Conditions (3) BRIDGE THE GAP BETWEEN SPENT FUEL AND NATURAL ANALOG STUDIES
- Incorporation of transuranic into alteration products has been predicted on crystal chemistry considerations (Burns et al 1997, J. Nuct Mater. 2.0,1).
O - It has been noted that the sequence e(akeration (corrosion) product foranaties in amaded (drip) tests fenews the paragenetic sequence observed in natural deposits, but this indernastion remains unutilized. ;
I CHARACTERIZE RADIONUCLIDES-RETAINING ALTERATION l (4)
(CORROSION) PRODUCTS f Is Np (say) incorporated in solid solution?
- Is it surface absorbed on precipitated particles?
i (5) Continue slow accurate unessarement of key thermodynamic parameters for use in chemical code calculations.
NOTE:
' Carbonate not only prevents formation of cormeiou pmdset deposits it complexes, and potentially enhances the transport ofactinides'.
l
9 l l
O i
SUMMARY
OF KEY REQUIREMENTS (1) KNOWLEDGE OF SEEPAGE WATER CHEMISTRY
.- CO3, Ca, Si content
- Modification by waste package corrosion products (2) MODE OF SEEPAGE WATER INGRESS Drip frequency
- Volume flow rate (3) FUNCTION OF CORROSION PRODUCT DEPOSITS O
Incorporation of radionnelides (4) ALPHA RADIOLYSIS
- Significance at fuel / corrosion deposit interface (5) FORMATION OF COLLOIDS
- Efficiency of radionuclides mobilization (6) REINVESTIGATE RAPID CORROSION / ALTERATION RATE (7) CLADDING CREDIT (?)
O
..