|
|
| Line 16: |
Line 16: |
|
| |
|
| =Text= | | =Text= |
| {{#Wiki_filter:.- .. . g | | {{#Wiki_filter:}} |
| >..4 ..o - .. .. ; . .. a .~ .r. . ..: , . .> ~.. ? .
| |
| . i...
| |
| .. ; ~ 7. vi..:
| |
| % m 3..- + +~.w.
| |
| %p:.g. 2,: ,' ;;;g*f.y a " ;. . . .., , ; ;, .v.
| |
| . . . m. ; . < l v; . _e. .. _.
| |
| ,.,.. .q:,..q. ...Q, .j. ; .
| |
| ~ % ~ ' I :. . . n. .* . {. .;. .; . . .. . ,. . ., ;. x. + .'. a. V,' & ' ::6 Q.W . ".. W2 2: i:..Y:
| |
| . . . . ~_ . ,. -
| |
| ,'W:., F. ..+ .w&.&m{ .;;.
| |
| . *, . .~ . . ..:
| |
| , . .L . \ _ . 'L
| |
| .w . . ;9 :za N..su . :c.~
| |
| w m'. ; . . . . . . .:s
| |
| . . .: . -,. L . .
| |
| . <f.a..
| |
| y.
| |
| g
| |
| . .; . . 4. i ',.y 7,,..y
| |
| - 7
| |
| ,) .. e. ,,
| |
| 4 4,' * - '. ,' ; . ;_ ' , ' . . .. .s. 9; : ., _j
| |
| . .- ' .. ' g . . . - . ' ' 'f . ' '* 'O f u_.i,,;.d;'M." ]..
| |
| '.
| |
| * l ,
| |
| ,, y ~.
| |
| .-,,.,:<.,.- . ;. )$... .-. !; ,::'q,; .. .:.: .;.<l L - E.&'.yi ,q ., .\ [ .
| |
| g g
| |
| ,. . ' =-
| |
| f
| |
| : h. "e
| |
| . - . . ' . ' , /:. .F,'''.
| |
| " , .".7.
| |
| [
| |
| *. s,; .
| |
| m p.
| |
| ;f.. n. ,?.
| |
| gi.N .). 4-f;. f? .:
| |
| .,A . - . ;-
| |
| yq
| |
| ;. ,j j.,_
| |
| ,p ' :
| |
| 3 .
| |
| . .. ,, l
| |
| _ -_e_ ,;, .
| |
| ._. . y' ;. .
| |
| .s., ,, l ", . , - Q ; . , . ;;f) 3. !."
| |
| ? --
| |
| .,,,-,'*j'..e.
| |
| '^ y, . y- q q '.p, . . ] .{ ,. . ; -
| |
| .. .. c, , . 4 .
| |
| g/, . , '. , ;;
| |
| ,4. . . ,
| |
| .., ,q . q ". c , ., ;,.
| |
| . .j..
| |
| 1
| |
| , m
| |
| . . c ;. :q
| |
| .,,t, n .a . g. ,
| |
| o %
| |
| ') , ,, _
| |
| . _ , _. - ,. - . . :* .,:..'f''..,*:
| |
| . . . ' , . . , . .' .' w .
| |
| . . 4 9
| |
| )'
| |
| .w- y; ,V q.. _[ . .. f.f
| |
| - -i F, : ; . y h l~ ;z
| |
| . ' .' . l'. . . .l ' ' '_'?.ll'.(- ,
| |
| . ., . ,. 3 g - f
| |
| ; , , . * .i t '.' . l ~ ' l iq ., (*
| |
| 'b v :' . ,, m ... , . . - ,.
| |
| O.Q
| |
| :' ;j}.-
| |
| l ,
| |
| e 's . , -
| |
| s',-
| |
| . . ' . . ' 'i .. ,',. . . ' .
| |
| . Y(, .: ' J '8 '. '.. ;, B '
| |
| - ,F
| |
| .T '; ; ^* .
| |
| }.j,,'^ ~l' 'f':;,1. .f. .". f ],->
| |
| , lj p "f ' " (. c 3 'MQ ; .
| |
| g- , - ; - - ,, . .
| |
| ,:y E' - u .-. ' - '
| |
| : .\, q, *. .. .. ..' ,l..
| |
| .k.
| |
| *r . <- * ,
| |
| k.'
| |
| ;f}
| |
| s i' ~gs 1 : , . . .. ., . . * . . . *
| |
| ''*)." x. -
| |
| . ; ._ . _.. , _ j. ..j' ' ' , . . g. --'}.
| |
| .[
| |
| . 4. ,. _,' .,l. ', '._,,,^^. _?.. G
| |
| . ,. l l ,
| |
| b.
| |
| e ,,, ,
| |
| , g. i,. '' #.
| |
| s 3
| |
| - .- .- . - ^ , . .. . ' .- . ,. _. . -
| |
| ..~....a.-', ,:
| |
| , . . ..,.; ' g ,
| |
| g, .;.,-+..- '' ~. ;
| |
| ' 1, ; ,.; u. .- . , , . , - -
| |
| 3.. /ik t
| |
| n ' '.,h ;'
| |
| '~ '
| |
| 9;
| |
| -6 . . .
| |
| .' l- ' , - - ' } $ . ,.? { ' ~ .
| |
| 'l
| |
| :.,,... -.,l.~ ';rls j '. .E f.-
| |
| ., ; . , -v. :>
| |
| *' . . , .- .F .;, -
| |
| ~. 3.; , b. ,t. .. . .
| |
| ..a: . a: 2
| |
| , .. . :*3 ..g , Q: . .-
| |
| -- p- - n;.a : w. ./.. m
| |
| ^
| |
| .:.y. ; . ; 4.y
| |
| :. l - .
| |
| :. ,- .. .- . ;^.. . . , . . . , ,;< ,lM.
| |
| : f. , ,%. , . .. ' . . :~ L '' bl.__, '. _ , . .*i:...' - V '. ' ; , '' N " .' '
| |
| fs l . e W' I '
| |
| -- . L....
| |
| =%''. . ..
| |
| '... , ''-[!
| |
| , ; .e - a . , , . . .
| |
| j . g y. ..,; . . l'.'- [, s. W 4. O, '' ' $
| |
| g' ',
| |
| ' 'j, l ?l ;.^ ..'* ' '(hY.)Qla.7.}.%' :.' *ve, . : ;r :[:],f;
| |
| .'; ^.[!' [.h.)
| |
| %er * " :
| |
| .l l r* y '
| |
| *:J ' '
| |
| .'.,'(
| |
| \ .
| |
| ~_l 1.% ,..;
| |
| n' W
| |
| T: . ,? .
| |
| 'S
| |
| ... ': :v ' n$
| |
| k- ;'.k
| |
| ; j. :
| |
| N'
| |
| . . . . .} . ,.
| |
| .A l . . .
| |
| p-g . :. < . '.,3 .
| |
| ? 'c-4-?.
| |
| - ,4 , .
| |
| L <
| |
| 3,
| |
| ,*q _ , . , .
| |
| _ ', . . ' . - ;_ .. - .T ." _: .q. y9 . '
| |
| '.y .?; . 7 :
| |
| ,Q;
| |
| .: . . . .. }. ; .;; _ ., .. ._:v. .;
| |
| . 3 *:.
| |
| ,y
| |
| ..p 4,'g. .
| |
| '., f. . , .; .
| |
| E '
| |
| s; ". ,(N ', 13N
| |
| .f L;;.;.
| |
| y ;
| |
| i j*L _ .aw:. -.*>>.l.
| |
| ,.#. - . u.. _.s >':,2 ;,,
| |
| , .y "> - - -
| |
| . .. ,e ..
| |
| . . _? _. . . :: 3:
| |
| , W.
| |
| .... - . *.; y : v s
| |
| y . ['* b b .
| |
| J,
| |
| '.l ' Y:l ; 1. )
| |
| ,c r;
| |
| ' .g ' .
| |
| ._ . . : W' .;; ("..,'_ '' .;, s_. y
| |
| : 7. 3 ,. .
| |
| Q. . ,
| |
| g- .,t ;. Q., . ' . , , ff
| |
| '.'! s ,h'- .k
| |
| .''"., )E
| |
| ~
| |
| q ,, .
| |
| '.e. 'E.
| |
| [<..- -
| |
| ' ' ),
| |
| g t
| |
| . lk p /n, 's [ ' f y
| |
| - . - I.' 1 -
| |
| 2'
| |
| ;3D y . . ' ' ;y <; ..'91 , i
| |
| , .1 : ,, .s. ~
| |
| . . ; . . '. .w .
| |
| }.:. 5, q:, .,
| |
| - .. . - '. ^~ ;I . : _ .
| |
| . - 6 y
| |
| ;_;}'-[.... Q.
| |
| k -
| |
| , . $ \'.' % , .(
| |
| : g. , - ( ' .- ,,.. , , . ('' ,
| |
| ; . . . - 1. , . ' ''i*.' (f''
| |
| i
| |
| - ?. . . : l ' [. . .l
| |
| ,j ' ,ip$b ~ i. . e ;. 4. .
| |
| .~ . .- .' . , , t .
| |
| . " ; , .. + : . ! 5 ,, ;, u iy g g 3 + ., *j ,.;.: - (. .
| |
| ' ;f --.' ;; : 3,; .. W; ;. 1 f i, e, '
| |
| ,[J ' '. '. , f* .;J/ .-* # ;-''' ' ' , ,f '. '. ', i . d . .f .3. -
| |
| 7 y. . se j'.. ; . . .
| |
| ' - ). '
| |
| i .
| |
| , ;q ,
| |
| .. sD , .. ,
| |
| ..[ .,S. j,.*'. fg,( ;'L ,'A. L ' k
| |
| ..,o ; . _.- -
| |
| (.s ',, e s
| |
| , f* cap 'i' J ' ' -
| |
| _. . , ., ,. . ,y., . .
| |
| . . ' . ., f.3y ;,,
| |
| .< e . . ' * . ' ' ' '- -
| |
| , ,*s ~
| |
| ,n ...y .
| |
| e 3.[ - - a ?
| |
| e
| |
| [. -,.,, p[. ,;.j..g ** . ..../ , .;. i.,...,
| |
| . )
| |
| ., , ., i -
| |
| ., s s . ,
| |
| f
| |
| .' _). .T..
| |
| '^
| |
| .... 4'~.,',',., ;..,
| |
| .'.) \ , ~.. .h ;1.;V.
| |
| . ;,',. ' ',_ .: ,y.
| |
| - . * (q ', ,g.
| |
| , y.a p.y .
| |
| P : V. '' s
| |
| .% gy. " , _ ' '. , , . ', ... ., ',
| |
| .t 4( ,+ - -
| |
| "< ; . +
| |
| ;p:{g ., g. . ,q ,3
| |
| : 9. ;.C : ,;.0" j f ? < .'J :. . P t 44 '.!
| |
| .< - ; - . . . '.41 . . ,e s ' c. . \ ' . ' ; c * ?,7 I , , , . , , , - .
| |
| ,' '' '. . ~.
| |
| , + } . . . . ,Y ' % '(, [ '. [.. , '' } .3 . $ q' :s.
| |
| k
| |
| ^
| |
| _; .l. [. _ a bl .'_.' ,
| |
| 3 ; _.: l[ & ..Q.;..? ; I I .[ll[}i \L.-
| |
| (
| |
| 'y :- :
| |
| w
| |
| ". ',# n.
| |
| - .. . .- ( '...:, . .
| |
| ' * ' .'i
| |
| .'=.;..', l
| |
| - -T, 7
| |
| ,y ?;1 ' -'
| |
| a ,,. '
| |
| a f'.n .;..p
| |
| _a. 1..,.
| |
| :.r. y , :N ... &q:c: _ ?e
| |
| * ._ , ! q.j ? . . l. p-
| |
| * b.' I,
| |
| ^
| |
| ~
| |
| l . -
| |
| '. l '
| |
| l.'
| |
| ;.t
| |
| / - ,
| |
| i q.. .;_._. ''.~.. 3 , _a,[ 6..'j 9'r _ , [ s ' . ;
| |
| :q%.t.
| |
| app; , .. .. .,
| |
| e
| |
| ...~.,.- -" J:" -'.*7, Te}n':tp 4*%
| |
| 5 ! . ; - ,- - . .~3.~".,*.." . .
| |
| . .. ,; s,
| |
| ,q;g C r; * ;,.< ; -
| |
| -.;,.o A. p . S.e. b..y . . , -
| |
| ., n;
| |
| ...p, , ,
| |
| , ,, ; : . . ; , , . g . . , . . , ., q ;e,,7 c.. ?. -,; ~ ,J
| |
| ,- . - , . . t .::- p
| |
| ',v, ,, . '- ;, y 4
| |
| ^
| |
| ,M ~,
| |
| a, . ,1 :.
| |
| ;; - . = :.+ 4 [ t
| |
| ^
| |
| ._ m a . . .. .
| |
| .o .
| |
| : 'J ; ,
| |
| . n
| |
| --4 pg w .
| |
| . . . . . . . , -~
| |
| ,; , ~ . ~; ,
| |
| .a . : q. ,A .;.q q 7.g g . g.g g, g men y
| |
| n- _ {(y ;. ; p3;g.yg.g: [ :;,y f es .- , ? t. ..q. . . - -
| |
| } .; ; ; ((y5.,g_;,.
| |
| : ygj g e
| |
| : s. , -+ .....;.,. ..';. .
| |
| ; , - i.;9 p . ; f,3 f.!;.3- 7,9y.Q;;
| |
| -s . .; .
| |
| T.M kM .$g c.,,,
| |
| : 7. y.., .,i e@ :' . nsHy "/
| |
| g-f.
| |
| e ." N dy -Qf
| |
| [
| |
| {lk
| |
| $;lY. " .e%
| |
| ,0.:;. .l,. f. ," 4 19. .'l.f.[i l y;;9? f: QWQf. .. r..z;
| |
| : 3.j.',. -;
| |
| _ 7 -: ..j .(,- , 'r
| |
| .'- % ~ . .*Q T -
| |
| y -p
| |
| .g
| |
| ; . .. - (-
| |
| a -
| |
| e .
| |
| y,7; g . ;;d. y; :.; .::. ,p
| |
| . . . , p.g.
| |
| , . w . . ..
| |
| ; ; +
| |
| : 3. .,. . > '~.m+ . . wy.n; .: >.m) : yy:; ;, ss. .,. .:
| |
| ,t,.
| |
| a ., * . .. - ..:,. ;.. . .
| |
| : 4. .:
| |
| ' , .y.
| |
| .M .; . -
| |
| . . - . c' ' . i
| |
| .'. ....,..',',...';*i.3m - ~.. - ' . .V , . . .#: q, $h' ,:m; sc ,'. >;"..Ayg.'. M> .f.;,.t
| |
| \..;y. ,
| |
| ,.. q.% Nj.) ,. , .
| |
| .. . . . .. . ., , .- _7 l'.' .
| |
| .y .
| |
| .a.w n j ht ' ' ' ;
| |
| ~
| |
| ~*
| |
| ,, ..', .! l
| |
| . O%.k. y ;, . Qk%; m .
| |
| ,,N,.l / : $.f .p . J_ 4*.; .',Q \
| |
| * * . y . . . . _. ::/ ;. . ;
| |
| . , : . .b s y: . n.:n. .:w , .n n:;.; y w:y;.,7 m .
| |
| | |
| n,
| |
| ~ ' '''.ln.-(c
| |
| .- . / . :}r ^' ,. ., .. .., ---7
| |
| '.'?h')yhje[ ** Q.g _ T a~~
| |
| '5,., .O,Q y', , fl ! l. '
| |
| y c-l!;hd ,; ,: M-M /
| |
| : tg
| |
| --/)
| |
| _y3 p.
| |
| ; ;V.
| |
| ?- j.g ' .
| |
| , -11 Nb r. T, . .
| |
| e MG NA; . ,
| |
| ACB3F-W/
| |
| (-lQ' V 'N, , . t ' rp N' --
| |
| f +
| |
| .l , . '
| |
| F,lTff f)lf re ., ' . 4 -4 ;,
| |
| n n R MOF,FICIAU5TRANSCRIPTL OF L PROCEEDIN'GS s s ,
| |
| hisM :)'$ w , ''N t -
| |
| ,..) ,
| |
| f_# .. , , ,'' ).
| |
| Mal ay-
| |
| " Ap
| |
| % m no s. 1 NUCLEAR (REGULATORYLCOMMISSION' ry._. r m
| |
| g, [bC/a i- 6f Op, v-j;y
| |
| @$cDVISOR . JCO$1NilTTEE ON. REACTOR SAFEGUARDS m" .: L cM, .+n.q.;c ~' ~ < n v -~, ,o '
| |
| mr 3.e :p i;<r !% ; s 3
| |
| f.v- t ys.. .s, .- :.
| |
| * ./f; 7 O l1 e i *' S;g''M^^N .. pj-
| |
| '; Q*%,]'
| |
| 3 y )[r f yJy . .
| |
| w+! w:inW::h ? kL. k* n,5 w.
| |
| sn~g. . g- n, a ' '
| |
| 1' 3 1 b. a -
| |
| v@; <u ' Mr W ~ s.i> ,
| |
| 8 '
| |
| f, xp 'i ~ ' iO y @n,p: m[s-y '
| |
| wm , ,, ,~ ,
| |
| b n.m $ 'b9 MT,itle:! -
| |
| MINNTING ONlTHE'RMAL HYDRAULIC m% .
| |
| n
| |
| . '.#.- } .
| |
| /:. ,+,: ~ 'X
| |
| 'i
| |
| &m ~8 1
| |
| ~ ., t(W ' '@ f, N W,N' 1,
| |
| % s 4'
| |
| e U
| |
| .c g' PHENOMENAL -
| |
| > .9 .\ ?^6, , y Q f Ms ,
| |
| " ' ^
| |
| ; rao4 Acas; Q Q: p ,g ge g;y%s ,
| |
| m i ,
| |
| l aEruRx carc:NAL ii A ! TO BNHITE
| |
| ,,%;> e & W ,*++ - ; m-if M 6 "'c%n , ,, , ,. I
| |
| 'j
| |
| +
| |
| n''
| |
| 4, 5, 1 tEtt u,ihmw%ms; -p.
| |
| 4 f,Gs $.,d.'~a W .
| |
| 4
| |
| +
| |
| ,. - 4 415-7130 >
| |
| mn
| |
| .hY ?y. . +.
| |
| o n. , , , _ a;, Ta m s:
| |
| w f;Y y #
| |
| ^
| |
| .?
| |
| x ... .
| |
| q m' v) 5 y&%w %a.;- nnc o ^
| |
| p's''
| |
| i Wl1.}'' yf ;',,,y.,%.; < DocketiNOF
| |
| [,
| |
| yy ^^
| |
| ?7~
| |
| 4 r;
| |
| ,c
| |
| _f f .)% - l\h, _$
| |
| y
| |
| (,p,,
| |
| 1
| |
| + 3- , <
| |
| : p. , .
| |
| I ;[ ff 4 ,'- '. -
| |
| ~"
| |
| :.: m A[ i.. u' +
| |
| : e. 's Qc
| |
| ?
| |
| Q y::;p f' ^<
| |
| > hk,' f.Q.,E, .}kk '
| |
| - JG :Y ,
| |
| wmq-g s 'q, <
| |
| p,, ,
| |
| ! ) y 5 J 5, i ,, , . , .
| |
| TMi%
| |
| . 1 EWork;O'rder
| |
| ; e' '
| |
| No.:c 'ASB-300:596:
| |
| N;'e< m g[My; i
| |
| .. . .d ." i
| |
| .m : n W;;c M a [xk'~i ~s 3 ,
| |
| ,. ,...g.
| |
| 2 s v>-
| |
| ' > , ' (t [ .
| |
| M., }~
| |
| o
| |
| ' T A.f' f[Q' -
| |
| gm 'y[J
| |
| 'r. ) .y.
| |
| m ,a wm.. >
| |
| Q .y .. +^".
| |
| Mq C
| |
| ny 'y n />; <
| |
| 7, i LOCATIONi/ .
| |
| iRockville,MD'- ) 4 rm, m EA.
| |
| c ,
| |
| , w-
| |
| ,m a n ,m.% - ,h ..
| |
| R.p . ,MW Ss 1.DATEU m hVednesday, December 16,1998 w s-. PAGES:1 - 361
| |
| ' %:3,.Stl % , '< ..
| |
| 4 ~..
| |
| ::f > x 44 - ;
| |
| $<s W MY W.O 9812280204 981216
| |
| -Qf%: y T .
| |
| Mf,J ' " PDR ACRS T-3061 PDR
| |
| , %mw4 %.c ,. a, , , ug; a . '
| |
| i yhw
| |
| / GnM w$ 4 j9 LANN RILEY & ASSOCIATES, LTD.
| |
| , 4 ,
| |
| 1025 Connecticut Ave.,NW, Suite 1014 p@f :~g "c " g3%
| |
| Osyu .
| |
| p/. ,4
| |
| _4
| |
| "/] ' - j j .
| |
| Washington,D.C. 20036 G02) 2-0034 %
| |
| ; i M. CCe s0Dy- Re&_
| |
| _ nh _j
| |
| ^
| |
| _ - . ~
| |
| | |
| . - - - . -- . ~ . -- . ..
| |
| ~ (D v
| |
| DISCLAIMER UNITED STATES NUCLEAR REGULATORY COMMISSION'S ADVISORY COMMITTEE ON REACTOR SAFEGUARDS DECEMBER 16, 1998 The contents of this transcript of the proceeding of the United States Nuclear Regulatory Commission Advisory
| |
| () Committee on Reactor Safeguards, taken on December 16, 1998, as reported herein, is a record of the discussions recorded at the meeting held on the above date.
| |
| This transcript had not been reviewed, corrected and edited and it may contain inaccuracies.
| |
| '- k.
| |
| | |
| _. .. . -. _ _ _ _ . . _ _ . _. . . . _ . _ _ . _ _ . . . _ _ . _ . . _. ._..__.m.. . . . . _
| |
| , 1 1- UNITED STATES NUCLEAR REGULATORY COMMISSION
| |
| ( 2' ADVISORY COMMIT *EE ON REACTOR SAFEGUARDS 3
| |
| '4 ***
| |
| u 1 5 MEETING ON THERMAL-HYDRAULIC PHENOMENA
| |
| -6 7
| |
| 8
| |
| -9
| |
| '10 USNRC, ACRS/ACNW 11 11545 Rockville Pike, Room T-2B1 12 Rockville, Maryland J l
| |
| 13 I l
| |
| 14 Wednesday, December 16, 1998 l
| |
| '(j%
| |
| _( 15' 16- The subcommittee met pursuant to notice, at 8:30 17 a.m.
| |
| 18 19' MEMBERS PRESENT:
| |
| 20 GRAHAM B. WALLIS, Chairman, ACRS 21 MARIO H. FONTANA, Member, ACRS
| |
| .22 : THOMAS S. KRESS, Member, ACRS 23' '
| |
| 24 25 ;
| |
| ,O . ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| . - , , m ., .. - .c..p. y
| |
| | |
| 2 1 PROCEEDINGS
| |
| (' )
| |
| AJ 2
| |
| [8:30 a.m.)
| |
| 3 DR. WALLIS: Good morning. The meeting will now 4 come to order. This is a meeting of the ACRS Subcommittee 5 on Thermal-Hydraulic Phenomena.
| |
| 6 I am Graham Wallis, the chairman of the 7 subcommittee.
| |
| 8 The ACRS members in attendance are Mario Fontana 9 and Thomas Kress. The ACRS consultant in attendance is 10 Virgil Schrock, and we hope that Novak Zuber will join us.
| |
| 1 The purpose of this meeting is for the 12 subcommittee to discuss the application of the Westinghouse 13 Electric Company's WCOBRA/ TRAC best-estimate large-break 14 LOCA code to nuclear power plants with upper head plenum (A) s./
| |
| 15 injection; the NRC's Thermal-Hydraulic Code Review Action 16 Plan; and the status of the NRC thermal-hydraulic research 17 program. The subcommittee will gather information, analyze 18 relevant issues and facts, and formulate proposed positions 19 and actions as appropriate for deliberation by the full 20 committee.
| |
| 21 Amarjit Singh is the cognizant ACRS staff engineer 22 for this meeting.
| |
| 23 The rules for participation in today's meeting 24 have been announced as part of the notice of this meeting 25 previously published in the Federal Register on November 30, i
| |
| (9
| |
| \_ /
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 j (202) 842-0034 l
| |
| l
| |
| | |
| 3 1 1998.
| |
| /'' 2 V} A transcript of this meeting is being kept. It is 3 requested that speakers first identify themselves and speak 4 with sufficient clarity and volume so that they can be 5 readily heard.
| |
| 6 We have received no written comments or requests 7 for time to make oral statements from members of the public.
| |
| 8 Now I'd like to proceed with the meeting, and I 9 call Mitch Nissley of Westinghouse Electric Company to 10 begin.
| |
| I 11 I think our emphasis is going to be on how well '
| |
| 12 this code represents the important physical phenomena. The l 13 history and all that is interesting, but we really want to {
| |
| l 14 know does it work well.
| |
| [)
| |
| 15 MR. NISSLEY: Thank you. My name is Mitch 16 Nissley, and I'm with Westinghouse Electric. The agenda for 17 this morning has been distributed, and the time allotments 18 we will try to follow very carefully.
| |
| 19 I will be giving the introduction, which we expect 20 to take about 15 minutes. That will be followed by a 21 description of the upper plenum injection plant geometry and 22 hardware of interest, given by Sue Dederer, which will also 23 last about 15 to 20 minutes. And then we will review the 24 PIRT results for upper plenum injection phenomena and show 25 some PWR sensitivities that further investigate some of l-i /~' ANN RILEY & ASSOCIATES, LTD.
| |
| ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 l (202) 842-0034 l
| |
| l l ,
| |
| | |
| 4 1 these behaviors.
| |
| 2 After the break, Kenji Takeuchi will spend about 3 an hour on each of the next two topics, which are the 4 details of the UPI code physics, and the second one being 5 code validation and sensitivity of consequences to the 6 model.
| |
| 7 We hope to wrap up with about a five- to 8 ten-minute summary statement and a review of the schedule of 9 what we hope will occur after this meeting.
| |
| 10 In this introduction, I want to first start off by 11 quickly summarizing the documentation as it exists today, g 12 and also future actions to complete the documentation.
| |
| 13 DR. WALLIS: Will that focus better on the bottom,
| |
| -14 or are we stuck with it?
| |
| () 15 MR. NISSLEY: No , I'll.--
| |
| 16 DR. WALLIS: That focus, are we stuck with it?
| |
| 17 DR. KRESS: I think we're stuck with it.
| |
| 18 DR. WALLIS: We're stuck with it, distorted on the 19 bottom. It won't focus better?
| |
| 20 That's better. Thank you.*
| |
| 21 MR. NISSLEY: And in the second part we will 22 discuss a comparison of the CSA applications for plants with 23 ECC injection into the upper plenum, contrast that with the 24 application for three- and four-loop plants with cold-leg 1 25 . injection with an emphasis on the changes.
| |
| (~}
| |
| \m ,(
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| (. Washington, D.C. 20036 L
| |
| (202) 842-0034
| |
| | |
| 5 1 Our documentation relies quite a bit on the
| |
| ''% original approved application for three- and four-loop (J
| |
| 2 3 plants with cold-leg injection. The topical report number 4 is given on the slide. It reflects the approved code 5 version and methodology for three- and four-loop plants.
| |
| l 6 Those code models are directly applicable to upper '
| |
| 7 plenum injection plants. We have not changed any of the 8 models. What you will see is we have madF some adjustments 9 in which physical models are ranged in the application of 10 CSAU, the only change being that we have added multipliers 11 to those physical models which are triggered via input. So 12 there --
| |
| 13 DR. SCHROCK: Is that date correct? I 14 MR. NISSLEY: Yes, it is. If you'll recall, there
| |
| ( 15 was an SER requirement to overhaul the documentation and 16 issue it in a form that reflected all of the final 17 methodology changes. That was accomplished March of this 18 year. i 19 DR. SCHROCK: Do we have that revision?
| |
| 20 MR. NISSLEY: Yes. We sent a copy to Dr. Larkin 21 at Dr. Larkin's request in either May or June.
| |
| 22 MR. SINGH: May or June of this year?
| |
| 23 MR. NISSLEY: Pardon me?
| |
| 24 MR. SINGH: Has a copy of it?
| |
| l 25 MR. NISSLEY: Yes.
| |
| I
| |
| /"N ANN RILEY & ASSOCIATES, LTD.
| |
| * (,,) Court Reporters
| |
| . 1025 Connecticut Avenue, NW, Suite 1014
| |
| ! Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . . - . - . , - . - ~ - . . . . . . . - . . - - - - . . . - . - . . . . - . . . . . .
| |
| 6 1 AUDIENCE RESPONSE: There's one in your ACRS
| |
| ( 2' library. '
| |
| 3- MR. SINGH: Okay. I
| |
| ~
| |
| 4 DR. SCHROCK: I. guess it's not been distributed.
| |
| 5 Graham, is that right? I haven't seen it.
| |
| 6 MR. SINGH: You should have gotten it in the 7 package that Paul --
| |
| -8 DR. SCHROCK: This package?
| |
| r j 9' MR. NISSLEY: No, I'm pretty sure Paul had not 10 planned to distribute that. It's about 8,000 to 10,000 11 . pages . -
| |
| 12 DR. SCHROCK: I'd just like to say I found this to 13 be a dismal paper to'look at, to. review in preparation for a 14 meeting, and if what you're telling me is I was supposed to
| |
| '( ) 15 ' learn what this document is by reading this pile of stuff, 16 .I'll tell you this is substandard engineering. You ought to 17 be ashamed of it.
| |
| l-18 DR. WALLIS: Well, this is in the ACRS. library.
| |
| l R19 It's on this floor,
| |
| : l 20 MR. SINGH: So, Virgil, you could look at it.
| |
| 21 DR. WALLIS: You could stay late tonight and look i-
| |
| :22 at it.
| |
| 23 DR. KRESS: It's only got what, 10,000 pages?
| |
| L, 24 MR. NISSLEY: We had hoped to have made it clear l
| |
| L '25 that we have not changed the code.
| |
| l, i
| |
| [''' -
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1
| |
| | |
| . . _ - ._.m._.._..- _ . _ . - . . . _ . . , _ . _ . . _ _._ _ _ _ _ _ _ _ . . _ _
| |
| 1E 7-l 1 The second topical report, which is indicated p
| |
| !( )
| |
| ~2 here, was submitted-in August of''95. It was the original 3 ' submittal on our application to upper plenum injection 4 plants. That particular submittal predated the final code 5 revisions and methodology for three- and four-loop plants, 6 and for purposes of this meeting, and I think we've-7- indicated this in our documentation to'you, this really 8- 'should be considered as obsolete.
| |
| 9, In August of this year we issued a report which I-10 believe you have definitely received. It started out with 11- about a 40-page executive summary in the CSAU format that 12 walked through the methodology. It also included all the 13 RAI responses with cross-referencing in the executive
| |
| '14 summary back to the RAIs that had additional detail on major 15 issues.
| |
| '16 -This document does reflect the code and 17 methodology changes. We had believed that it was scrutable 18 and that someone could use the executive summary to find 19 their way through the RAIs, recognizing that it was a large 20 document. I believe from Dr. Schrock's comment there might 21 b2 some disagreement there.
| |
| 22 Several weeks ago we issued a letter to the ACRS 23 which summarized what parts of that'large document we were 24 going to use in today's presentation, with the hopes that 25 that would allow you to focus on only 100 pages or so, ANN RILEY & ASSOCIATES, LTD.
| |
| O1 Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| ~> _ _ _. ~ - . . _--
| |
| t i
| |
| i 8 1 rather than the full 600 or 700 pages.
| |
| 2 The final document I have listed here, which I
| |
| ' f~N x m 3 believe you've also received, is the draft technical 4- evaluation report which was prepared by Idaho National 5 Engineering Lab.
| |
| l 6 Future actions regarding the documentation are to
| |
| ; 7 issue an approved version of that obsolete topical report 8 that has been overhauled to reflect the final methodology 9 changes and the use of CSAU. By updating the documentation 10 to reflect the final product, there's numerous benefits from
| |
| ! 11 that. One is internally in Westinghouse it reduces the 12 reliance on the people who were involved in the development.
| |
| j 13 With regard to licensees, it allows them to better 14 understand what our methodology is in a more straightforward
| |
| ( ) 15 manner. And it also will enhance future dialogue on this 16 topic with the NRC and ACRS.
| |
| 17 The next several slides are arranged in the CSAU 18 format. CSAU consists of three major elements with 14 19 steps. What I will try and do here is focus on the areas 20 where we have either done additional code assessments or 21 revised the PIRT or done changes to the parameter ranging l 22 for PWR applications.
| |
| 23 In element I, the two areas where we are doing 24 something different, obviously item number 2, select a l
| |
| l 25 nuclear power plant, instead of looking at plants with
| |
| --(''N ANN RILEY & ASSOCIATES, LTD.
| |
| (,,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 9 l 1 cold-leg injection, we are looking at plants where the
| |
| , ('''
| |
| -2 low-head safety injection is into the upper plenum. That ;
| |
| l 3 will be covered in the next topic.
| |
| 4 We also have reviewed and modified the PIRT to 5 reflect phenomena which .'re of more ii..portance for upper 6 plenum injection plants, and that will be re.iewed in the 7 third topic.
| |
| 8 The second element of CSAU is the assessment and 9 ranging of parameters. Based on the results of the PIRT, we 10 identified need for additional code assessment work. That l 11 information will be presented in topic 5 today, which will 12 be after the break. '
| |
| 13 Nodalization for the PWR calculations, consistent 14 with previous experience we have first set the nodalization
| |
| [/)
| |
| 15 for the PWR and then used it for the code assessments to 16 assure consistency. That will be summarized in the next 17 topic.
| |
| 18 In terms of determining the code and experiment 19 accuracy consistent with the previous application, we have 20 used separate effects tests, in this case some CCFL tests 21 performed by General Electric to establish the parameter 22 ranging on the physical models.
| |
| 23 We have also used those results and applied that 24 ranging to the integral effects tests performed in the UPTF 25 and CCTF facilities, and we will discuss that later.
| |
| . /''\ ANN RILEY & ASSOCIATES, LTD.
| |
| (_) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| ?
| |
| l 10
| |
| ( 1 We also have several slides where we looked at f')N s,
| |
| 2 data observations of parameter trends with scale as reported 3 by MPR, and what we've done is compare how COBRA / TRAC 4 predicts those trends with scale. And that will be shown 5 later today.
| |
| 6 Element III, which is the sensitivity and 7 uncertainty analysis, is where the results of the PIRT and 8 code assessment work will be translated into the methodology 9 to be used for analysis of PWRs with upper plenum injection.
| |
| 10 Step number 11, although we have not changed this, 11 I wanted to comment on this, that in the Westinghouse 12 methodology, in addition to the code uncertainty which was 13 presented in the original CSAU methodology, we also put 14 emphasis on important PWR initial and boundary conditions, N
| |
| V) 15 and also included those in the uncertainty treatment, and we 16 continue to do that here.
| |
| 17 In step 12, where we've actually performed 18 plant-specific analyses, model uncertainties are propagated 19 in the same manner by ranging global and local models. In 20 this case global models refer to those that affect the 21 overall system thermal hydraulics and the overall hot 22 assembly response. Local models are those that appear 23 locally on the hot spot on the hot rod.
| |
| 24 One change we have made here is that the global 25 models that are ranged have been revised in terms of the
| |
| ,Q ANN RILEY & ASSOCIATES, LTD.
| |
| (_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| )
| |
| i 11 1 code-physical models to reflect the UPI phenomena. !
| |
| /'')
| |
| (
| |
| 2 Another difference here is that due to the i
| |
| %d 3 difference in size of the plants, the split break is more l 4 limiting than.the double-ended guillotine break, and that's 5 an easy adaptation to the methodology, but it's a point 6 worth making.
| |
| 7 The overall uncertainty methodology where we use a 8 combination of Monte Carlo and response surfaces to develop 9 PCT uncertainty distribution is unchanged, and is shown on 10 the flow chart on the next.two pages.
| |
| 11 The plant uncertainties are reflected on the first 12 page of the flow chart. This is separate from the code 13 model uncertainties. Items 1 and 3 deal with uncertainties 14 in the power-related parameters at the beginning of the
| |
| ![v ) 15 transient: What are the peaking factors that exist? What 16 are the axial pcwer distributions that exist?
| |
| 17 Items 5 and 6 refer to et;her initial conditions l 18 such as the accumulator initial water volume, accumulator 19 temperatures, SI temperatures, et cetera. Also 20 uncertainties in initial RCS temperatures and pressures are 21 included here. These items have not been changed.
| |
| 22 DR. SCHROCK: What do you do in power distribution 23 response surface? What are you varying?
| |
| l 24 MR. NISSLEY: We have several parameters that are l 25 considered there, the main ones being axial power l
| |
| l l
| |
| i /'' ANN RILEY & ASSOCIATES, LTD.
| |
| '(,,)
| |
| /- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| : Washington, D.C. 20036 l (202) 842-0034
| |
| | |
| 12 1 distribution, total peaking factor, which is frequently
| |
| ("N 2 referred to as FQ, the enthalpy rise peaking factor, which
| |
| '0 3 is the integraJ over the rod, which is typically referred to 4 as F delta 3.
| |
| 5 Also in this part of the uncertainty treatment we 6 de.11 with decay heat uncertainties, gamma redistribution 7 Uncertainties, and local peaking factor uncertainties.
| |
| 8 DR. SCHROCK: Do you deal with the effect of the 9 composition variations, spatial composition variaticas on 10 the uncertainty in the decay power?
| |
| 11 MR. NISSLEY: Yes. The hot assembly, we have 12 done -- we use the limiting burnup in terms of the hot 0 13 assembly, and that is a fresh fuel assembly. We use decay 14 heat that corresponds to approximately 2,000 megawatt-days
| |
| () 15 of burnup for the hot assembly for the terms of decay heat.
| |
| 16 For the other assemblies we use an average value 17 of 10,000 megawatt-days.
| |
| 18 DR. SCHROCK: What I'm referring to is a spatial 19 effect that arises due to the fact that the composition of 20 different pellets is different, and the decay power function 21 depends upon that. And so the shape of the decay power 22 distribution within the reactor is not identical to the 23 shape of the power during the operation just prior to the 24 accident. So I'm asking if you are taking that into 25 account, and if so, how.
| |
| [D ANN RILEY & ASSOCIATES, LTD.
| |
| \_,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 13 1 MR. NISSLEY: We do not have a -- the axial -- the j''} 2 decay heat does reflect --
| |
| the burnups of the rod do vary
| |
| \/
| |
| 3 the decay heat products that exist. In the fresher fuel you 4- have more decay products from the U235. In the more burned 5 assemblies you have more fission products from plutonium.
| |
| 6 And that is captured in the modeling.
| |
| 7 DR. SCHROCK: Is there a reference you could cite 8 where that's described?
| |
| 9 MR. NISSLEY: That is described in section 8 of 10 that large WCAP that we mentioned was in the library down 11 here. That would be volume 1, section 8.
| |
| 12 DR. SCHROCK: Thank you.
| |
| 13 MR. NISSLEY: The second page of the flow chart 14 indicates where the code uncertainties and the model e-(g) 15 uncertainties are factored into the overall uncertainty 16 assessment. As I mentioned before, the overall Monte Carlo 17 process and response surface techniques are the same as 18 previously. I also indicated though that we range different 19 physical models to capture the important parameters for 20 upper plenum injection plants. Box number 12 there is the 21 only one that's affected out of this process, and the 22 comparisons with test data that we have later today will 23 show how we derived those changes.
| |
| 24 AUDIENCE RESPONSE: I don't think we have it.
| |
| 25 DR. WALLIS: Are you speaking for the record,
| |
| [~N ANN RILEY & ASSOCIATES, LTD.
| |
| i
| |
| (_s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| ..m_ _ . . _ . . _ _ _ _ _ _ _ _ . _ . _ _ - _ _ . _ _ - _ . _ _ . . _ _ _ _ . _ _ . _ .
| |
| l' 14 i
| |
| \
| |
| 1 Norm? You have-to speak so we can hear you.
| |
| l 2 That concludes the introduction. If there are no i 3 'further questions at this time, I'd like to turn it over to
| |
| \
| |
| .4' Sue Dederer.
| |
| )
| |
| :5 MS. DEDERER: I'm Sue Dederer, I am also from 6 Westinghouse. I am going to be discussing the UPI plant l 71 geometry.
| |
| 8 .The Westinghouse UPI plants are very similar in 9 overall design to the three and four loop plants that have H10 already been analyzed and approved in the three and four 11 loop methodology that was recently approved.
| |
| 1:2 The main differences between the UPI plants and L -13 the three and four. loop plants are'the size of the vessel 14- and the loops. We are talking about a two loop plant with a f()- '15 smaller vessel. It has 121 fuel assemblies versus the 157 i 16 that you saw in'the three loop ~ plants.
| |
| : 17. The fuel type is a 14x14 lattice versus the 17x17 l-18 or 15x15 lattice that was analyzed in the three and four 19- loop best estimate methodology, but the main difference is 20 the location, the safety injection.
| |
| 21 As Mitch pointed out before, the Low Head safety
| |
| ;22 . injection injects into the Upper Plenum and then the High 23 Head and the Accumulators inject into the cold leg, the same
| |
| .24 Ras the three and four loop plants.
| |
| -25 This next slide shows a cross-section of the Upper l
| |
| -[# ANN RILEY & ASSOCIATES, LTD.
| |
| \ Court Reporters 1025 Connecticut Avenue,HNW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
| |
| t' . _ - - . . _ . . - . _ .
| |
| | |
| 1 l
| |
| l 15 1 Plenum at the location of the hot legs and you can see the
| |
| ! /~' 2 hot legs and then the UPI locations are the smaller I
| |
| s 3 locations shown there.
| |
| l 4 DR. SCHROCK: What is the elevation of that 5 injection with respect to the tie plates?
| |
| 6 MS. DEDERER: With respect to the tie plates?
| |
| 7 DR. SCHROCK: Yes. What holds the core -- l 8 MS. DEDERER: You mean the tie plates in the Lower i
| |
| 9 Plenum? 1 10 DR. SCHROCK: Yes. l 11 MS. DEDERER: Well, this is the Upper Plenum. Let 12 me show you a noding diagram.
| |
| 13 DR. SCHROCK: One of my reasons for making the 14 comment earlier is that we were given a diagram with a. title ,
| |
| [)\
| |
| R.
| |
| 15 something like Schematic of a UPI System --
| |
| 16 MS, DEDERER: Yes.
| |
| 17 DR. SCHROCK: It is a terrible reproduction.
| |
| 18 MS. DEDERER: Yes. 1 19 DR. SCHROCK: Which even in its original form had 20 inadequate detail for anybody to judge what is the geometry 21 that we are dealing with here, so if we are going to talk 22 about this problem we need to understand what the geometry 23 is.
| |
| 24 MS. DEDERER: I agree. That is why we left that 25 slide out in the final package, because it was a very poor l
| |
| ' ("]
| |
| (_/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| I 16 1 reproduction.
| |
| /~'s 2 DR. WALLIS: Could we go back to your --
| |
| U 3 MS. DEDERER: Sure. Let me just answer his 4 question.
| |
| 5 DR. WALLIS: Okay.
| |
| 6 MS. DEDERER: The UPI is up here at the same 7 elevation of the hot legs. The tie plates I believe you are 8 -talking about are down here in the Lower Plenum.
| |
| 9 DR. SCHROCK: Well, I was really referring to this 10 other hardware --
| |
| 11 MR. NISSLEY: The upper core plate?
| |
| 12 DR. SCHROCK: Yes.
| |
| 13 MS. DEDERER: Oh, yes, the upper core plate would 14 be right here and then UPI injection is here.
| |
| ,m
| |
| # 1 15 MR. NISSLEY: About three feet.
| |
| O 16 DR. SCHROCK: And do you determine that your 17 flooding limitation occurs at those upper restrictions? j 1
| |
| 18 MS. DEDERER: Yes. We are going to discuss that a 19 little bit later, if you want to hold off, and then if you 20 have further questions --
| |
| 21 DR. SCHROCK: No, I would like to understand the 22 geometry we are talking about when we get to those 23 discussions and at this stage it is still unclear.
| |
| 2
| |
| .4 MS. DEDERER: Yes. I am going to go through that 25 in just a minute.
| |
| l I
| |
| ( /7 ls ,/ ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| : 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . -- . - - . . . ._._.....__m. . . . _ . _ _ _ . _ . _ . _ _ . _ _ _ . _ . ~ . _ . . _ . _ . _ . . _ . . .
| |
| 17 1, DR. WALLIS: Could you go back to the previous.
| |
| /~ . -
| |
| 2 picture,-so we can understand that?
| |
| 3- Are these -- presumably those holes go right .
| |
| 4 through those UPI injection'--
| |
| 5' MS. DEDERER: These?
| |
| '6 :DR. WALLIS: 'Yes. Right?
| |
| l 7 MS. DEDERER: Yes. I am not sure why this drawing' l 8 shows them closed that way, but, yes, they do go into the 9 . Upper Plenum.
| |
| I 10 DR. WALLIS: What are we looking at? We are L 11 looking'at a cross section at what level here?
| |
| ! 12- MS. DEDERER: This is at the hot leg elevation.
| |
| 13 DR. WALLIS: But that box around everything isn't l 14 'there, that cross-hatched box?
| |
| 15' MS. DEDERER: Yes. i i
| |
| 16 DR. WALLIS: That would be in the way of the. '
| |
| 17 injections, so what is'that?
| |
| 18 MS. DEDERER: Well, yes. This is a composite '
| |
| [ 11 9 drawing. You are looking down at the --
| |
| '20 DR. WALLIS: That is below something.
| |
| 21 MS. DEDERER: -- t he core plate and these are the 22 baffle plates that are down in the core.
| |
| [ 23 DR. WALLIS: Way down below.
| |
| 24 MS. DEDERER: Way down below --
| |
| 25 DR. WALLIS: What are the other things, those t
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| )
| |
| Court Reporters
| |
| ; 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036 l-(202) 842-0034 1
| |
| | |
| 18 1 square and.round things? Those are present at the UPI k Llevel?
| |
| jx__,'} - 2 i
| |
| 3 MS. DEDERER: Yes, those are the structures in the t '. 4 Upper. Plenum. We are going to discuss those next.
| |
| l^
| |
| ! 5 DR. WALLIS: So the injection is coming in and L -6 impinging immediately on this forest of structures?
| |
| 1 t
| |
| 7 MS. DEDERER: That is correct. The injection
| |
| ~8 comes in here -- these vertical structures -- they would be l L9 ' coming out-this way. I will show you a picture of.that in a I i
| |
| ,. 10 minute. !
| |
| l 11 DR. FONTANA: What again are the small, very small I
| |
| : 12. ' dots that you have? i 13 MS. DEDERER: This here? This is the UPI l g 14 injections --
| |
| ~15' DR. FONTANA:
| |
| l
| |
| ( No -- those things.
| |
| L 16 MS. DEDERER: Oh,-those are just bolt holes,-. ;
| |
| ' 17c DR. FONTANA: Oh, okay.
| |
| 18 DR. WALLIS: Bolt holes? ,
| |
| i 19~ MS. DEDERER: That is where the structure bolts ;
| |
| j 20 down onto the. upper core plate. '
| |
| 21 DR. WALL 16. So that is all down below too.
| |
| 22' MS. DEDERER: That is on the top of the upper core i
| |
| ; -23 plate.
| |
| L 4 24 DR. WALLIS: That is all down below. {
| |
| 1 25' [ Discussion off the record.]
| |
| 1
| |
| * l i
| |
| i )
| |
| 'f ANN RILEY & ASSOCIATES, LTD. l
| |
| \ l Court Reporters j
| |
| ! 1025 Connecticut Avenue, NW, Suite 1014 '
| |
| Washington, D.C. 20036 j.
| |
| i.
| |
| 1
| |
| ; (202) 842-0034 l
| |
| l-
| |
| | |
| I i l
| |
| 19 1 -lMS. DEDERER: Mitch was going-to discuss this 2 figure later but we will just put it up since you brought up 3 .the question. l 4 Right here is showing'where the UPI comes in and 5 -these are the structures that you are impinging upon, the 6 support columns and the guide tubes so you see -- and we 7 have a discussion of this in one of the REIs, 49 I believe
| |
| : 8. it was, where we. talk about as the water comes in how it is 1 i
| |
| 9 impinging on the structures and how much of it comes !
| |
| 10 through. Thank you. l 11 We're going to discuss that further in the later i
| |
| 12 discussion.
| |
| 13 DR. WALLIS: It makes a pool. The water makes a 14 pool up there.
| |
| l
| |
| () 15
| |
| : 16. plate --
| |
| MS. DEDERER: Yes, it does -- on the upper core 17 DR. WALLIS: Righc.
| |
| 18 MS. DEDERER: We are going to talk about that 19 later too.
| |
| 20 DR. WALLIS: But then we have to understand can 21 the pool drain out over the outside of this --
| |
| 22 MS. DEDERER: If you will do me the favor of 23 allowing me to continue, I think I will answer a lot of your 24 questions.
| |
| 25 DR. WALLIS: So you are going to describe the l
| |
| , O ANN RILEY & ASSOCIATES, LTD.
| |
| i \m/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 t (202) 842-0034 i
| |
| | |
| . . . . .. . .. _.-_.- -. .. . ~ . - . . _ - . . - .. .-. . .
| |
| ~
| |
| l 20 l i
| |
| 1- system better so we can understand it? '
| |
| i 2- MS. DEDERER: Yes. I am.
| |
| 3 DR. WALLIS: Okay. Thank you.
| |
| l 4 MS. DEDERER: Just to continue on with the slides, g
| |
| 5 this shows the ID component modelling of the UPI plants and l
| |
| 6- the only thing -- this is very similar to what the three and 7 four loop plants look like.
| |
| 8' You have in this case only two loops, an intact 9 loop and a broken loop, and then the additional item we have l.
| |
| : 10 in ID components is the addition of the UPI injection, which !
| |
| 11 connects directly to.the Upper Plenum, and that is really 12 .the only difference between the two loop UPI model in the 13 loops and the three and four loop plant loop model.
| |
| l 14 Somehow I lost a page here.
| |
| l l
| |
| . ;( ) 15 MS. DEDERER: Okay. The other thing I wanted to 16 point out -- this is a cross-section of the upper plenum.
| |
| : 17. This is the vessel model that we have built for the UPI 18 Plant, and here again we see the hot legs, channel 59 and l
| |
| L 19 channel 60, and then here is the upper plenum injection
| |
| ! 20 locations, and we have this model set up slightly different 21 than the three and four loop plants because of the safety 22 injection being in the upper plenum.
| |
| 23- One of the things that we wanted to make sure we f 24 could model is the possibility of asymmetric flow, because 25 of the safety injection, single failure assumption.
| |
| I i
| |
| - f' ANN RILEY & ASSOCIATES, LTD.
| |
| 'T Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036 (202) 842-0034
| |
| | |
| t.
| |
| 21 l
| |
| 1 The way this plant is set up, when you assume a l 2 single failure for safety injecticn there is only safety
| |
| (~)}
| |
| 3 injection in one UPI port and it happens to be this port in i 4 our model. The other port does not inject and so there is a 5 possibility that you would have flow that would be 6 asymmetrically distributed in the upper plenum.
| |
| ! 7 We did a more detailed upper plenum model in order l
| |
| 8 to capture that possibility if it was predicted to occur, 9 and what we have is an inner global model which is the same 10 kind of a design or same kind of a model as we had in three 11 and four loop plants, but then we took the outer global 12 model ring and we split it into four channels instead of 13 one. That allowed us to have a separate channel for the UPI 14 injections and then there's a separate connection for the l[)
| |
| x_/
| |
| 15 hot leg connections.
| |
| 16 We felt this detail was important and we wanted to 17 make sure that we could model the possibility of the change 18 in flow.
| |
| l 19 Now as you just brought up earlier, there's a lot 20 of details you want to be careful to model when you are 21 talking about an upper plenum injection plant. There are 22 several locations where there is a possibility of CCFL
| |
| ! 23 occurring and we need to make sure that the model could 24 calculate that if it was predicted to occur.
| |
| l 25 One of those regions where you can get CCFL is at
| |
| ' (~T ANN RILEY & ASSOCIATES, LTD.
| |
| ('') Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l
| |
| (202) G42-0034 l
| |
| | |
| l l l 22 1
| |
| 1 the top nozzle for the fuel assembly.
| |
| /''g 2 Right here I have a detail, a little bit clearer l
| |
| . \_/
| |
| 3 detail, of what the vessel looks like. This is the top of 4 the fuel ~ assemblies here. There are the top nozzles and 5 then this is the upper core plate that Dr. Schrock was i
| |
| 6 talking about, and these are the structures in the upper 7 plenum. i i
| |
| 8 Now the section four of our vessel model is what 9 we call the CCFL region. It goes from the top of the active I 10 fuel to the bottom of the upper core plate, and in that 11 region we have two locations where you could have CCFL 12 occur.
| |
| 13 The bottom of that section we wanted to model the l 14 flow restrictions that you see in the top nozzle, and we n
| |
| ) 15 brought along -- thank you -- our top nozzle so you could (G
| |
| 16 see what we are talking about physically. It is much better 17 than the pictures, but this is a 14x14 top nozzle and this 18 bottom is the adaptor plate which we are talking about with 19 the restricted flow.
| |
| 20 You can see obviously the flow area is reduced and 21 also it has a large wetted perimeter. This is important for l
| |
| 22 us to model this, so that we can calculate CCFL. You get a 23 very small hydraulic diameter because of the amount of 24 wetted perimeter here, so that is what is modelled at the
| |
| , 25 bottom of section four.
| |
| l 1
| |
| ((~j}
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| : Washington, D.C. 20036 l (202) 842-0034
| |
| | |
| 23 1 .DR. SCHROCK: What I inquired about, I thought I'd 2 ~ clarify, is that upper plate, and I asked if you were 3 concerned with flooding at that level.
| |
| 4 I thought I hea'd r the answer "yes."
| |
| '5 MS. DEDERER: Right. That's next.
| |
| 6 DR. SCHROCK: That's next?' Okay.
| |
| : 7. DR. WALLIS: What are these slots and these guide 8 tubes or whatever that is up there?
| |
| 9 MS. DEDERER: We will talk about that next too.
| |
| 10 DR. WALLIS: That also comes into it.
| |
| 11 MS. DEDERER: At the top of the section four we're 12 then concerned about the flow area at the upper core plate, 13 and this is where we need to know what structures we have in 14' the upper. plenum and where they are' located.
| |
| () 15 This drawing shows the different structures that 16 are in the upper plenum for this type of plant.
| |
| 17 We have support columns _which have flow slots for 18 lateral flow. They also have a solid portion here, and then 19 there is of course a hole at the bottom for flow down to the 20 core.
| |
| 21 We have guide tubes that extend all the way to the 22 upper' head. We have open holes that have no structure above 23 them, just a hole in the upper core plate, and then these 24 are freestanding mixers which are just short, 13 inch stands 25 with mixing veins inside of them, and those mixing veins are ANN RILEY & ASSOCIATES, LTD.
| |
| O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| i l
| |
| 24 l 1 just'for mixing L.he flow as it comes up through the core.
| |
| 2 l g DR. SCEROCK: Are these tubes that extend-into the l s) 3 upper head open at some place? Those are slots that are l
| |
| l 4 shown there, is that the way this works?
| |
| ~5 MS. DEDERER:
| |
| l Yes. Those are what we call " cards" ;
| |
| 6- and there'are leak paths there, so essentially all the l- 7' liquid in the upper head can drain down through the guide ,
| |
| 8 -tubes during blowdown.
| |
| 9 DR. SCHROCK: Is there another alternate path or 10 does it all drain through that?
| |
| l 11 MS. DEDERER: I believe that for this plant this 12 is the only' drain path.
| |
| L 13 DR. SCHROCK: It's the only drain path.
| |
| L 14 DR. WALLIS: There is no flow-path around the edge 15'
| |
| ( of this whole thing?
| |
| 16 MS. DEDERER: No. There's flow into the downcomer 17' also.
| |
| 18 DR. WALLIS: Where is that path?
| |
| :19 MS. DEDERER: Well, that would be out on the edge l 20 beyond the upper plenum. We were only discussing the flow 21 path to the upper plenum, but there actually a flow path
| |
| .' 22 through the nozzles in the downcomer up to the upper head.
| |
| 23 DR. SCHROCK: It would seem to me you need to 24 address the two things simultaneously. If you are going to 25 assess how much goes down the guide tubes you have to know I
| |
| ' ANN RILEY & ASSOCIATES, LTD.
| |
| :s Court Reporters
| |
| ; 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i;
| |
| i
| |
| | |
| . _ _ . _ . _ . _ _ . . _ _ . . _ _ _ . . _ . _ _ ._._ _ _m._.._. ___ . . . . . _ _ . _ _ . . .
| |
| {. 25 i
| |
| L '1 how much'you are putting in and how much of it is going
| |
| /"' - 2 somewhere else.
| |
| (
| |
| 'm I
| |
| 3 MS.uDEDERER: - Yes, we do model those flow Paths. '
| |
| 4 ,
| |
| We are focusing tcday on the things that are key q
| |
| l 5 to the UPI plant. All of the other things you are l- l 6 mentioning are already considered and they are the same as 7 what were considered in the three and four' loop application, 8 so we were try: ag -- in order to save time --
| |
| 9 DR. SCHROCK: So there is no synergism here -- the i.
| |
| I !
| |
| 10 existence of a new flow path doesn't change -
| |
| . 31, l l 11 certainly the phenomena are not the'same. That is, you !
| |
| 12 didn't have subccoled water going out of the upper head into I l 13 the douncomer previously, so what is it that is the same?
| |
| 14 MS. DEDERER: The guide tubes are the same. There i
| |
| {() 15 is no difference in the guide tubes between.three and four 16 loop plants and two loop plants. The structure might be
| |
| )
| |
| ; 17 different but the modelling is the same.
| |
| L 18 DR. SCHROCK: So far as the path between the upper j 19 head and the downcomer.
| |
| 20 MS. DEDERER: Yes, that is the same in three and 21 four loop plants and two loop plants.
| |
| 22 DR, SCHROCK: But what are you calculating in the 23 four loop plants? Calculating the draining of water that 24 gets carried up through the guide tubes and tends to pool in l
| |
| 25 . the upper head, runs back into the downcomer?
| |
| t I' h ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| ; 1025 Connecticut Avenue, NW, Suite 1014 Washington,.D.C. 20036 I-(202) 842-0034
| |
| . y u .. -fpy w. ..% y g u .e y .m -
| |
| +e+-
| |
| | |
| l l
| |
| 26 i 1 MS. DEDERER: Well, I don't believe we have any 2 pooling la the upper head. We show that the water during
| |
| {J"')
| |
| 3 blowdown is all dispensed from the upper head with maybe the 4 exception of an inverted top half.
| |
| 5 DR. SCHROCK: Well, my problem here -- what I hear 6 is that the flow through the, from the upper head to the 7 downcomer is the same as in three and four loop plants.
| |
| 8 MS. DEDERER: Are you talking about thF actual 9 amount of flow? Is that your question or --
| |
| 10 DR. SCHROCK: I am talking about the phenomena 11 that are involved and what the important flows are, okay, so 12 obviously they can't be the same because you don't inject 13 water into the upper head in those plants, so if there is 14 water up there and it does drain through such a path the
| |
| ()
| |
| p 15 16 circumstances are not the same.
| |
| It would not be subcooled water specifically.
| |
| 17 MS. DEDERER: The upper head water drains during 18 blowdown before the UPI injection even begins and that is 19 the same in the three and four loop plants and also in the 20 two loop plants.
| |
| 21 That particular phenomena is not unique to the UPI 22 plants, so the cooling that we achieve from the upper head 23 draining into the core, we saw that in the three and four 24 loop plants and we still see it in the two loop plants, and 25 it is not really --
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| /~N) q, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l- .. .. -
| |
| | |
| i 27 i 1: DR. SCHROCK: Are we unconcerned with this l
| |
| / 2 draining during upper head injection? ;
| |
| 3 MS. DEDERER: It's not upper head injection. It's 4 upper plenum inj~ection. We are talking about upper plenum 5 . injecting here. . Upper-head injection would be up here and :
| |
| 1
| |
| - 6; we are not discussing that today.
| |
| 7 DR. WALLIS: IJthink the point of the question 8 is -- l
| |
| .9 DR. SCHROCK: That is.where I understood you were 11 0 injecting.it was up ai ve?
| |
| .]
| |
| 11 ~MS. DEDERER: No, that.is -- I am sorry you 12 misunderstood.
| |
| 13 This is.the injection location was roughly 14 half-way up-the upper plenum for UPI plants. Okay? I am
| |
| -15: sorry you had that misunderstanding.
| |
| 16 DR.,WALLIS: When you say things are the same, I 17 -think.the point of our questioning is even though physically 18 it looks the same as in the three and four loop plant, do 19 you need.a better model in some way because of the phenomena 20 with upper plenum injection?
| |
| , 21 MS. DEDERER: We do need a better model. It is 22- .not in the guide tubes.
| |
| 23' DR. WALLIS: That is why we are asking these 24 questions about where does the water go and can it get to 25 the downcomer. I still don't quite understand how it gets ANN RILEY & ASSOCIATES, LTD.
| |
| O. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| , _.- - .. - . -. - . . - - - . . . . ~ . . . . _.- - ~ . --
| |
| 28 i l' to the downcomer, i
| |
| l
| |
| /~' 2 Could you show me how it gets there in the figure?
| |
| D) 3 MR. NISSLEY: This plate is referred to as the 4 upper support plate. The reactor vessel set wall is out ;
| |
| 1 l 5- here. There is a mating surface thatHcomes out here. I !
| |
| 6 forget exactly -- the design looks like, and there's flow 7 paths through here.
| |
| 8 DR. WALLIS: At.that height that it gets to the --. I 1
| |
| 9- MR. NISSLEY: Everything above here is considered 10- the upper head.
| |
| 11 DR. WALLIS: Yes.
| |
| 12 RMR . NISSLEY: All of this water flashes and is 13 removed during blowdown before upper plenum injection even 14 starts. Upper plenum injection occurs about here at C7.
| |
| / ) ' 15 ' DR. WALLIS: So there's no path to the downcomer J. J l16 below the elevation that you've shown us there. I 17 MR. NISSLEY: That's correct.
| |
| ! 18' DR. WALLIS: Okay. Thank you.
| |
| 19' MR. NISSLEY: The hot leg and cold leg nozzles are
| |
| : 20. about this elevation as well.
| |
| 21 DR. SCHROCK: So does any of the upper plenum 22 injection find its way up into the upper head?
| |
| 23- MS. DEDERER: No.
| |
| 24 DR. SCHROCK: No.
| |
| 25 MS. DEDERER: Now you asked on one of the previous l
| |
| p ANN RILEY & ASSOCIATES, LTD.
| |
| ( ,/g Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 j Washington, D.C. 20036 (202) 842-0034 l-l l l l
| |
| | |
| . , . _ . _ _ _ _ . _ . _ . _ _ _ . _ .__._..m.. _._._..m.___._ . ~. .
| |
| 29
| |
| :1 ' slides.what all those squares and holes were that were
| |
| : 2. showing up. This slide details what structures are above 1
| |
| : 3. which fuel assemblies in the core. So what you see here is 4 the structures of the upper plenum superimposed over the 5 fuel assembly cross-section. So that you can see which' type f 6 of structures we have over each fuel assembly.
| |
| 7 And what we do in the vessel model is we separate 8 out low power channels which are on the out --
| |
| there's 24 l 9 assemblies on the' outer ring. They're on the flats of the i
| |
| 10 core. And then we have the inner global channels, which are 1 t-
| |
| [. 11 the remaining assemblies on the inside.
| |
| L 12 And you can see from this diagram that we have i
| |
| 13 different structures scattered around across the upper 14 plenum. And what we tried to do in our vessel modeling is 15
| |
| '( ) maintain the vertical flow integrity of each of these _
| |
| l 16: different types or structures. We're going to discuss these 17 structures individually so you can see what's different 18 about the flow paths and why we model them separately.
| |
| l p 19 The first structure we're going to discuss is the 20' guide tubes that we were just talking about earlier. This 21- is a' picture of.just the lower portion of the guide tubes, l
| |
| i 22 and I wanted to show this to you because it shows you what i 23 the lateral flow paths are at the upper core plate.
| |
| l-j 24 This would be sitting on the upper core plate, l'
| |
| l
| |
| '25 this bottom plate. Okay? Above the upper core plate, this l ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters !
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L i
| |
| | |
| p 30 )
| |
| p t
| |
| .1 is the plate.that sits on the upper core plate, and you see l
| |
| /~N 2 these fingers that are sticking out that hold the control h
| |
| 3- rods as they're inserting into the core.
| |
| This is an open l 4- flow area. This is Just a plate'on the bottom. But then j l I L
| |
| 5 the rest above it is just an open flow area.. These are kind l 6 of fingers that stick down. They're kind of skinny little 7 guides.
| |
| i 8 And then on all four sides there's .een-flow area. !
| |
| 9- So if'we were to have any pooling of water at the upper core 10 plate, you would be able to have crossflow into the guide
| |
| :11 - tube, and then down through the upper core plate into the 12 core.
| |
| 13 DR. WALLIS: So all these unique and fantastic 14 structures have been tested at full scale for their l( ) 15- characteristics?
| |
| 16 MS. DEDERER: When Westinghouse designs them, they i 17 test them for their flow characteristics.
| |
| 18 DR.-WALLIS: Including two-phase characteristics?
| |
| 19 MS. DEDERER: I'm not familiar with the kind of 20 testing that they do. I know we get different kinds of loss
| |
| [ 21- coefficients that we use.
| |
| L
| |
| -22 DR. WALLIS: I see no way to predict without 23 testing this sort of geometry.
| |
| 24 MR. NISSLEY: To the best of my knowledge we have 25 no'two-phase --
| |
| i i -
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters t.'
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| - = _ _ -
| |
| l' 31 1 DR. WALLIS: No two-phase model for these things 2 or no two-phase test of these things?
| |
| [G~')
| |
| 3 MR. NISSLEY: Other than to the extent they're l
| |
| 4 tested in the CCTF and UPTF facilities.
| |
| 5 -DR. WALLIS: So they're tested in an integral G facility.
| |
| 7 MR. NISSLEY: Right. We've done single-phase 8- testing of them, but not two-phase, to the best of my 9 knowledge.
| |
| 10 DR. SCHROCK: There was a test done at INEL in the 11 2D-3D program on actual Westinghouse geometry, but I don't 12 know that it was this geometry. This was roughlyly around 13 '82. Are you familiar with that?
| |
| 14 MR. NISSLEY: What was the name of the facility?
| |
| 15
| |
| ( , DR. SCHROCK: Well, it was Y. Y. Shu's 2D-2D 16 program. Benedetti was the investigator on it.
| |
| 17 MR. NISSLEY: I'm not familiar with that data.
| |
| 18 However, these designs are used in all of the Westinghouse I 19 PWRs except the new AP600 des _ign, and a particular design 20 with what we called upper head injection, which are no 21 longer in use.
| |
| 22 So if Westinghouse hardware was tested, it would 23 have been of this design.
| |
| 24 MS. DEDERER: This next structure that I have up 25 here in figure 5 is a support column. You see again, as I i
| |
| l l
| |
| [~') ANN RILEY & ASSOCIATES, LTD.
| |
| (_ / Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 32 1 mentioned before, the support column has a solid portion 2 which is 13 inches higP. Then it has flow slots which allow 3 lateral flow into the support column. And then the upper 4 portion is solid again, although that doesn't really concern 5 us in the UPI application.
| |
| 6 It is important for the UPI application'that we 7 model this solid cylinder at the bottom, and this part here 8 sits on the upper core plate, so this 13 inches that you see 9 here restricts the lateral flow completely. So there's no 10 flow, a cross-flow, into the support columns until the water 11 in the upper plenum builds up past 13 inches. So that's 12 critical in our model to make sure that we model that and 13 account for that in the channel noding that we've put 14 together.
| |
| [/)
| |
| 15 DR. FONTANA: What's that small tube that runs 16 down there Is that a --
| |
| 17 MS. DEDERER: That's an instrumentation tube.
| |
| 18 DR. FONTANA: Is it a sipper or is it a 19 temperature element?
| |
| 20 MS. DEDERER: I believe it's a temperature sensor.
| |
| 21 DR. FONTANA: Okay.
| |
| 22 DR. SCHROCK: Will your model description include 23- something about the distribution of subcooling in that pool?
| |
| 24 MS. DEDERER: We had an RAI where we discussed the 25 subcooling, and I believe we're going to discuss that later O ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| ~ . _ _ _ . _ - - . ._ .
| |
| L 33 1 on.
| |
| /"N 2 DR. SCHROCK: Okay.
| |
| b 3-l MS. DEDERER: The other structure that's important 4 to the UPI application is the free-standing mixer. That was 5 the short structure I showed you in the upper plenum 6 structures.
| |
| l 7 This portion here is the same 13 inches high as 8 what we were discussing in the support column. It sits 9 right on the upper core plate, and inside of it is a mixer 10 which is strictly for mixing the flows, but we don't have 11 streaming of flow from the core up into the upper plenum.
| |
| 12 The mixer that is inside of this free-standing 13 mixer also appears in some of the support columns, and when 14 we acedally look at the calculation of flow areas, this
| |
| () 15 16 structure becomes the most limiting flow area at the upper core plate because of the mixer. !
| |
| 17 DR. SCHROCK: Could you show me again at what l
| |
| 18 elevation this is mounted?
| |
| 19 MS. DEDERER: This is the freestanding mixer right 20 here. So here's your upper core plate. The freestanding 21 mixer's here. The support column is here. And then the 22 guide tubes are here. So you can see that the support 23 column and the freestanding mixer both have the restricted 24 lateral flow at the upper core plate, which is important to 25 model because of the pooling in the upper core plate from l
| |
| (~)' ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters t
| |
| 1025 Connecticut Avenue, NW, Suite 1014 L
| |
| Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| . _ . . . _ . _ . _ . . . . ~ . _ _ _ . _ _ _ _ _ _ _ _ _ _ . _ . _ . . . .
| |
| 34 1 the.UPI water.
| |
| i !
| |
| 2 So we have separated out these structures to model j 3 them, and in particular when we want to model the hot 4- assembly and have the most limiting flow characteristics, we
| |
| : 5. ended up picking the. freestanding mixer.
| |
| l-6 DR. SCHROCK: Doesn't the water have to go over 7 the top'of that freestanding' mixer?
| |
| t 8 MS. DEDERER: Correct.
| |
| 9 DR. SCHROCK: To get~into its -- have to puddle 10 that level before you can get anything there.
| |
| MS'. DEDERER: 'That's correct
| |
| -12 DR. SCHROCK: But isn't flow through these slots 13 into the tubes?
| |
| 14' MS. DEDERER: Yes.
| |
| ()
| |
| f15 DR '. SCHROCK: At the lower level going to occur 16- prior to that?
| |
| l 17- MS. DEDERER: Yes, that's correct. That's why in '
| |
| 18 our model we have separated out the guide tubes into a
| |
| -19 separate channel by themselves. The freestanding mixer i
| |
| 20 which is modeled over the hot assembly is of course by 21 .itself,.and then in the places where we expect the water to 22 =be going down into the core, we have also separated out the 23 support columns-or mixers from open holes, which of course 24 also have no lateral flow restrictions.
| |
| -25 DR. WALLIS: Earlier you showed upper plenum i
| |
| I l-ANN RILEY.& ASSOCIATES, LTD.
| |
| l .W, k j/ Court Reporters
| |
| '025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l.
| |
| ; 35 1 divided into four quadrants.
| |
| (~}
| |
| , Q]
| |
| l 2 MS. DEDERER: Yes.
| |
| 3 DR. WALLIS: Because you were concerned about lack l
| |
| 4 of uniformity. So the level up there could be higher at the 5 injection point than over the rest of the region.
| |
| )
| |
| 6 MS. DEDERER: Yes. I l
| |
| 7 DR. WALLIS: In which case the water is going to 8 go down these 13-inch-high chimneys or funnels, whatever, )
| |
| l 9 where the water is high, and not at all elsewhere.
| |
| 10 MS. DEDERER: That's correct. j 11 DR. WALLIC: So you could have a very lopsided 12 flow distribution.
| |
| 13 MS. DEDERER: That's correct. We saw that in the 14 UPTF test, and in order to make sure that we could capture
| |
| () 15 16 that if it was predicted to occur, we developed this more detailed upper plenum model.
| |
| j 17 DR. WALLIS: So then you need a more detailed ;
| |
| 18 model of the flow in the structures, and that some of the 19 structures have different flows than the others.
| |
| 20 MS. DEDERER: That's correct.
| |
| 21 DR. WALLIS: And that's also captured here.
| |
| 22 MS. DEDERER: That's what we're trying to capture.
| |
| 23 DR. SCHROCK: Did the UPTF test include injection 24 of subcooled water?
| |
| 25 MS. DEDERER: Yes, we did UPTF test 72 -- no, I'm i
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| , (O)
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . . ~ - ~ - -. - - - . - - ..._--~. - .. ... _._.-.__..- -__ _ _ .. _
| |
| l l 36 1 , mixing those up.
| |
| 2 MR- TAKEUCHI: 20.
| |
| {) .
| |
| L 3 MS. DEDERER: 20 has the UPI injection, and we I 4 simulated that, and Kenji will be discussing that later.
| |
| 5 DR. SCHROCK: Okay. Thank you. :
| |
| f 6 MR. NISSLEY: Can I make a few points here? One l 7 is we will be showing the predicted level distribution l 8 around the upper plenum as well as the subcooling 9 distributions, and a very important point you'll see later-10 on is that during upper plenum injection, we predict L 11 flooding.at all of the interior assembly locations.
| |
| 12 Draining only occurs on the very edge of the core, where 13 there are no guide tubes.
| |
| 14 DR. KRESS: Does that mean in your hot channel
| |
| ' 15 . that there's no flow ever goes down --
| |
| 16 MR. NISSLEY: It's bottomed on the --
| |
| 17 DR. KRESS: It's all bottom up. So this all --
| |
| 18 all this discussion is to say how the water gets down to the 19 bottom, really.
| |
| 20 DR. WALLIS: I think your eventual conclusion is 21 enough gets down that it essentially fills from below, and 22 -all of this detail doesn't matter. Is that the proper 23 -conclusion?
| |
| 24 MR. NISSLEY: After many months, yes, that was our 25 conclusion.
| |
| l i
| |
| (- N ANN RILEY & ASSOCIATES, LTD.
| |
| '(
| |
| l Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 37 1 DR. KRESS: I guess it impacts on how much gets
| |
| [~'} 2 entrained and carried out the hot leg.
| |
| v
| |
| ! 3 MR. NISSLEY: Pardon me?
| |
| 4 DR. KRESS: It would probably impact on how much 5 liquid that's injected gets carried out the hot leg instead 6 of going down.
| |
| 7 MR. NISSLEY: Yes. We'll be showing parametric 8 studies of that.
| |
| 9 MS. DEDERER: Okay. This is a cross-section of 10 cur vessel model right above the upper core plate level, and 11 I want to show you this because this is where we have the 12 detailed modeling, and we separate out the physical nature 13 of the structures so that we can model the different flow 14 restrictions and try to -- well, fairly accurately predict A) t, 15 the flow that comes down into the core.
| |
| 16 The previous drawing or sketch that I showed you 17 showed the four quadrants in the outer ring of the upper 18 plenum. Now what you're seeing in addition here is in those 19 four quadrants we_have separated out the structures that we 20 just discussed.
| |
| 21 In each quadrant we are showing two additional 22 channels. One channel is modeling the open holes so that 23 any water that w1uld get into, for instance, this channel 24 here, which is channel 32, any water that was here could 25 " low through this gap, 36, into channel 36, which is an
| |
| % 5001 RILEY & ASSOCIATES, LTD.
| |
| ['~'/
| |
| s- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| - . .. - , . .- . . . . . - . - . . - . - . - _ _ . - - - . ~ . . . ... - ..-.. -
| |
| (-
| |
| l L 38 l
| |
| 1 open-hole channel. There would be no flow restrictions
| |
| [~'\ ' 2 here. .Any water here could go directly down into that iV l 3 -channel as long as there wasn't a CCFL occurring at that L 4 time.
| |
| L 5 On the other hand, channel 40 is a freestanding 6- mixer, and as you recall, we just discussed there were 13 I l
| |
| 7 inches of solid cylinder there, so we have modeled that with I
| |
| 8 this gap, number 40, the bottom portion of the gap has 9 blocked flow. So there would be no cross-flow right at the 10 upper core plate level for the first 13 inches through this l L
| |
| 11 gap. And until the liquid level builds up above 13 inches, 12 there will be no cross-flow there.
| |
| l 13 DR. WALLIS: Excuse me. You could have a level of l
| |
| ! 14 say 15 inches in one of those quadrants where you're ;
| |
| ()
| |
| p~s 15 16 injecting and then maybe 10 in the next one; is that right?
| |
| j MS. DEDERER: That's true, you could.
| |
| l 17 DR. WALLIS: So the question arises then is four j, 18 quadrants enough? Why four rather than eight or 16 or i
| |
| l' 19 whatever?
| |
| 20 MS. DEDERER: It's just a matter of computation 21 . time, how much did-we see when we looked at the tests. When
| |
| '22 we looked.at UPTF test it seemed like there was a quadrant 23 where the water was coming down versus the other quadrants 24 were not. So it was kind of a judgment call. Plus four is
| |
| .25 a good physical -- you can see the physical relationship.
| |
| i-l ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| ; 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 39 1 For four you have two hot legs, you have two UPI injection
| |
| /''N 2 ports, you know, four is an obvious divisor.
| |
| (1 3 DR. WALLIS: Four is the minimum probably.
| |
| 4 MS. DEDERER: Yes. I think in the four-loop plant 5 we only had one outer plenum, but we didn't have UPI 6 injection, so it wasn't important there. But for this 7 Model, we felt this was an important thing to do.
| |
| 8 DR. SCHROCK: Those larger circles surrounding the 9 square boxes with numbers in them are not indicative of 10 hardware?
| |
| 11 MS. DEDERER: This right here?
| |
| 12 DR. SCHROCK: Yes.
| |
| 13 MS. DEDERER: That's the channel. The little 14 square boxes are chetnnel numbers, okay? This circle is fm 15 channel 36, and this circle is channel 40.
| |
| I, J This is channel 16 45.
| |
| 17 DR. SCHROCK: I understand you're describing the 18 COBRA / TRAC modeling of the reactor, but I don't understand 19 what geometry you are envisioning for the COBRA / TRAC model 20 that involves some imaginary hardware of that scale.
| |
| 21 MS. DEDERER: Well --
| |
| 22 DR. SCHROCK: So what does it mean?
| |
| 23 MS. DEDERER: This circle or this channel that we l
| |
| 24 have set up is modeling a certain number of the structures 25 in the upper plenum, so say that we had ten freestanding i
| |
| /~' ANN RILEY & ASSOCIATES, LTD.
| |
| (_,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 j (202) 842-0034 l t
| |
| I I
| |
| | |
| I 40 1 mixers, it models the flow area of ten holes going down l
| |
| ~
| |
| 2 through the upper core plate, the cross-flow area for the 3- ten, you know, perimeter of that structure. And that would l
| |
| 4 all be contained in the channel modeling for that channel.
| |
| 5 DR. SCHROCK: So am I hearing this right, you 6 l imagine that the flooding _ characteristics of a restriction '
| |
| 7 representing some number of small channels in the actual 8 system will have the same total flow result as that i 9 collection of smaller channels? l 10 MS. DEDERER: We're.modeling the flow area -- we l 11 lump the flow area together for structures that are similar, i
| |
| 12 so we would have the same flow area. We also model the 13 wetted perimeter so that we get the correct hydraulic l 14 diameter.
| |
| () 15 DR. KRESS: As long as you only have that one 16 node, that's equivalent to modeling every one of them. I 17 don't see another way to do it, unless you make more than 18' four quadrants, and then you'could separate them out.
| |
| 19 MS. DEDERER: ~ Well, you know, these four quadrants 20 represent only 24 locations. There's 24 assemblies in the 21 core below them. There's 24 structures, so there are 22 basically six in each of these quadrants. So we're only 23 talking about a --
| |
| 24 DR. KRESS: Since the height of the water is the 25 same in each quadrant, modeling it this way is -- there's no
| |
| , ("N ANN RILEY & ASSOCIATES, LTD.
| |
| (,,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 41 1 finer detail -- you can't do it any differently, as I can
| |
| ("\, 2 see, because it would take a distribution of the level in q,,)
| |
| \
| |
| 3 order to model each one of them separately. !
| |
| 4 DR. SCHROCK: Well, the flooding characteristics
| |
| {
| |
| l 5 might be different. '
| |
| 6 DR. KRESS: In each one of them. Yes, certainly
| |
| : 7. could be in reality, yes. But as long as you only have one I
| |
| 8 quadrant -- four quadrants, then there's no way you can i 9 differentiate that any other way. You'd have to have more 10 quadrants to be able to capture that.
| |
| 11 DR. WALLIS: Are you saying that the flooding 12 characteristic of a parallel channel isn't the same as the 13 flooding characteristic of one big channel?
| |
| 14 DR. SCHROCK: I don't think so.
| |
| [m}
| |
| %.)
| |
| 15 DR. WALLIS: Probably right.
| |
| 16 DR. KRESS: Yes, that --
| |
| 17 DR. WALLIS: So you're forcing all the channels to 18 behave the same way in that particular quadrant.
| |
| 19 MS. DEDERER: Yes, that's correct.
| |
| 20 DR. KRESS: And there would be no other way to 21 really analyze it unless you had more quadrants.
| |
| 22 DR. WALLIS: I guess the question is, is that a 23 representation of reality or not? Is that your question, 24 Virgil?
| |
| 25 DR. SCHROCK: Yes.
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| k'~')
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 42
| |
| : 1. MR. NISSLEY: It's worth noting that we used the 2 same nodalization in CCTF and UPTF, which are a different 3 scale, and we predicted the scale trends quite well, as 4~ you'll see later. I think that is relevant.
| |
| 5 DR. FONTANA: What do LP and GT mean?
| |
| -6 MS. DEDERER: LP stands for low power. That's the 7 lower power region of the core, the 24 assemblies on the 8- flats of the core, so the four quadrants on the outside of 9- the upper plenum are directly over the low power channels.of
| |
| '10 the core.
| |
| , :11 We are doing this so the we can separate out the 12 CCFL effects where you have the highest steam flow at the 13 hottest assemblies in the core and the low power channels at 14- the lower assemblies would have less steam flow, so this
| |
| () 15 modelling allows us to separate out where the downflow would 16 most likely occur over the low power assemblies in the outer 17- ring and the high power assemblies are in the center and the 18 channels coming up from the core there have higher steam 19 flow and would have less downflow, if any.
| |
| 20 All of our tests, all of our runs that we made we 21' see all of the water going down below power region so all 22- the water in going down the outer assemblies, not down the 23 inner assemblies.
| |
| 24 CCFL keeps the water from going down the inner 25 assemblies.
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| (G,/L: Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 43 !
| |
| 1 DR. FONTANA: What does GT mean?
| |
| -2 MS. DEDERER: Guide tubes.
| |
| 3 DR. FONTANA: Right, okay.
| |
| 4 DR.'KRESS: Does that geometry for the model 5' continue all the way down into the lower plenum? !
| |
| 6 MS. DEDERER: No. We start doing some 7' simplifications once we get down below the upper core plate 8 and that is on the next slide.
| |
| 9 DR. SCHROCK: I guess I would like to hear, as the 10 explanations develop further, how you deal with the 11 existence of a hydraulic gradient across these quadrants and 12 in addition to that gradients in the subcooling.
| |
| 13- These are certainly factors that influence the 14 flooding characteristics of the individual channels and bear
| |
| () 15 on the issue of can you lump them together in a way that you 16 are trying to do in your computer model?
| |
| 17 MS. DEDERER: In our test simulations we have done 18 several test discussions. Kenji is going to be discussing 19 those later.
| |
| l 20 DR. WALLIS: In CCFL you have subcooled water in i
| |
| 21 one quadrant going rapidly down. Is it enoug' to prevent i 22 any vapor formation at all in that channel?
| |
| 23 MR. TAKEUCHI: Professor Wallis, I would like to 24 discuss those things.
| |
| 25 DR. WALLIS: You will discuss that later, because ANN.RILEY & ASSOCIATES, LTD.
| |
| (s~' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . ~. . _ _ _. - . _ _ . - _ . __ _ _ _ _ _ _ . . ._ . __ - - _
| |
| 44 1 maybe it is easier if CCFL doesn't occur at all if you have
| |
| () 2 3
| |
| enough downflow of cold water -- essentially single phase flow.
| |
| 4 MR. TAKEUCHI: We have a distribution of the pool 5 , height and also.subcooling distribution in both channels.
| |
| l 6 DR. WALLIS: So we will hear about that later?
| |
| i l
| |
| 7 MR. TAKEUCHI: Yes.
| |
| 8 DR. WALLIS: Thank you.
| |
| ,9 MS. DEDERER: You asked about whether the 10 channeling remained the same as we go down through the core 11 to the lower plenum.
| |
| 12 This slide shows the cross-section right below the 13 upper core plate in what we call the CCFL region. If you 14 recall from one of the earlier slides I showed you, the CCFL
| |
| () 15 region is from the top of the active fuel to the bottom of 16 the upper core plate.
| |
| 17 This is a cross-section of the vessel model at 18 that location.
| |
| 19 Here you see that we have maintained the inner 20 global channel and the separate guide tube, the lumped 21 support columns and the separate hot assembly. Those are 22 maintained the same as what you saw in the upper plenum but 23 we have done a simplification on the outer ring, the lower 24 power' ring. We now have just one outer global channel l 12 5 instead of the four quadrants and we have lumped all of the l
| |
| i
| |
| <O) ANN RILEY & ASSOCIATES, LTD.
| |
| !( /l_. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l' 45 ;
| |
| E 1 open holes together -- the four open hole channels above are -
| |
| /''N ,
| |
| !'O f
| |
| 2 3
| |
| now lumped to'one, and the four support column, freestanding mixers are lumped to one.
| |
| 4 We found through some testing, we made different 5 runs of the different noding schemes and we found that this 6 simplification did not change the results so we went ahead ,
| |
| 7 and did the simplification once you got below the upper core l, 8 plate.
| |
| ! 9 DR. WALLIS: Let me see if I can understand this, 10 because we talked about four quadrants in the upper plenum 11 and we said that the level could be higher in one of them-l 12 and water could be going down in that one. l 1
| |
| L 13 When you get below the plates you somehow mix that 14 water all the-way around?
| |
| .fg !
| |
| L( ) 15 MS. DEDERER: .Yes. This is the second place where 16 you get pooling of water.
| |
| 17 That plate that was on the top nozzle restricts 18 the flow quite a bit so you.actually can get pooling in two.
| |
| 19 locations, one on the upper core plate, where we just 20 discussed how we keep the modelling separate so that you can 21 model the cross flow, and then the second place is at the
| |
| . 22 adaptor plate at-the bottom of section four. Pardon me?
| |
| l l 23 DR. WALLIS: Did you show that in Figure 2 or 24 something? This adaptor plate in section four? )
| |
| 25 MS. DEDERER: The adaptor plate I don't have h I 1
| |
| ./ ANN RILEY & ASSOCIATES, LTD.
| |
| I''
| |
| Court Reporters
| |
| ! 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
| |
| 1 -. ._ . .-
| |
| | |
| 46 1 sketch of. That was the adaptor plate of the model there, L\
| |
| (^ - 2 the top nozzle.
| |
| ! 3 DR. WALLIS: I don't see a place to pool in this 4 cross-section.
| |
| 5 MS. DEDERER: Let me see if we can --
| |
| 6 DR. SCHROCK: Is this piece of hardware the 7 adaptor plate constriction?
| |
| 8 MS. DEDERER: That is the top nozzle and the 9 bottom plate is the adaptor plate where the water would be 10 pooling.
| |
| 11 DR. SCHROCK: It's very close, isn't it?
| |
| 12 DR. WALLIS: Where is the adaptor plate?
| |
| 13 MS. DEDERER: Pardon me?
| |
| 14 DR. WALLIS: Where is the adaptor plate in this
| |
| ( 15 figure?
| |
| 16 MS. DEDERER: The is your top nozzle right here 17 and the adaptor plate, as you can see on the model, is en 18 the bottom of the top nozzle.
| |
| 19 DR. WALLIS: So water comes down in one of these 20 quadrants at the top there, say the one next to you, somehow 21 mixes right across to the other side of the picture?
| |
| 22 MS. DEDERER: No. It would remain in the outer 23 ring.
| |
| 24 DR. WALLIS: But you somehow mix it?
| |
| 25 MS. DEDERER: It could be mixed.
| |
| . (~'g ANN RILEY & ASSOCIATES, LTD.
| |
| ( ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034 l
| |
| l
| |
| | |
| 47 1 DR. WALLIS: In the model for the upper plenum you
| |
| ['] 2 have it going down in one place and then soon as it gets
| |
| \_/
| |
| 3 through it spreads everywhere?
| |
| 4 MS. DEDERER: Well, it has to go -- let's see how 5 I can describe this.
| |
| 6 DR. WALLIS: I didn't quite understand how it does 7 that.
| |
| 8 MS. DEDERER: The water from the upper plenum 9 could be coming down through the upper core plate. It would 10 be coming down through one of these channels, okay?
| |
| 11 If it wants to spread, it then has to go through 12 this gap into the outer global and so then it could spread.
| |
| 13 DR. WALLIS: But your model forces it to spread, 14 doesn't it, by having --
| |
| 15 MS. DEDERER:
| |
| (A) If you have pooling, it could 16 spread. If you have enough steam flow, it could keep the 17 water from going down into the core.
| |
| 18 DR. WALLIS: But your model forces it to spread, 19 because you have lumped everything in one --
| |
| 20 MS. DEDERER: That's right.
| |
| 21 DR. WALLIS: So the question I guess we have is --
| |
| 22 I would have is whether that represents what really happens.
| |
| 23 MS. DEDERER: One of the other things I should 24 point out is you can sort of see it on that nozzle.
| |
| 25 There is a lot of flow restriction laterally for
| |
| -G ANN RILEY & ASSOCIATES, LTD.
| |
| ( ,)
| |
| g Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 48 1 :' the top. nozzles. The springs that you see on that model
| |
| }
| |
| 2 actually compress down once they load the fuel and the 3 crossflow area i;s not very.large so you do have a lot of 4 resistance to flow, but you are right. 'Any water if it.
| |
| .5: comes-down, if it goes out -- if the water is. coming down 6 here, and there's sufficient steam flow so.that it cannot 7 continue down through to the core, then it will crossflow' 8 out into the global channel and then it could go.over to one 9 of the other channels here or it could go into the inner 10 channel.
| |
| -11 DR WALLIS: But you are assuming that it does --
| |
| 12= your model.
| |
| [13 MS. DEDERER: We are modelling it to do that.
| |
| 14 DR. WALLIS: Your model forces it to do that, 15 because unless I am --
| |
| 16 MR. NISSLEY: Could I show a nodalization study?
| |
| 17 DR. WALLIS: Unless I am confused -- I thought you 18 told us that-there's~only one CSFSM in that whole, 19 representing the whole core at this level.
| |
| 20 MS. DEDERER: There is one channel representing --
| |
| >21 DR. WALLIS: Number 23 is representing all of l-L 22- them.
| |
| 12 3 ; MS. DEDERER: No , it only represents the ones on
| |
| -24 the outer --
| |
| 25 .DR. WALLIS: Yes, all of the ones on the outer are ANN RILEY & ASSOCIATES, LTD, O+, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington,-D.C. 20036 (202) 842-0034 1
| |
| | |
| 49 l 1 represented the same. .
| |
| /'I 2 MS. DEDERER: .Right. 'That's correct. At this lv 3 ' elevation we are representing all'of them the same.
| |
| 4 DR. WALLIS: So it doesn't allow you to have water I iS' flowing down one side and not the other? :
| |
| 1 6- MS, DEDERER: Right. '
| |
| .)
| |
| L 7 ', DR. WALLIS: Okay.
| |
| l 8' MS. DEDERER: But.we did'a nodalization study' 9 where we kept the four separate freestanding mixer channels 1
| |
| 10 both.above the upper core plate and below the upper core 11' plate, and we=saw that it really didn't make any difference 12 in the results. That's why we made the simplification. i 13 DR. WALLIS: Are you going to show us any evidence
| |
| '14 that it didn't make any difference?
| |
| pf..
| |
| g 15' MS. DEDERER: Yes. ;
| |
| 16 MR. NISSLEY: Yes. j
| |
| ; 17 DR. WALLIS: Okay. I i
| |
| 1;8 EDR . SCHROCK: That one that you are just taking 19 away might'be better titled the name of this piece of l 20 hardware grid at the bottom -- what is the name of that?
| |
| 21 MS. DEDERER: "The. Top Nozzle" -- actually, we 22 call it the CCFL section because of the type of flow we get.
| |
| 23 DR. SCHROCK: So.that means the cross sectional 24 areas correspond to the cross sectional area of the bottom 25 plate in this piece of hardware?
| |
| i ANN RILEY & ASSOCIATES, LTD. I Court Reporters
| |
| ! 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 1 (202) 842-0034 )
| |
| l
| |
| | |
| . . _ - ~ . - _ _ . _ . . _ _ _ . . _ . . - _ _ _ _ - _ . _ . . _ _ _ _ _ _ . _ . _ _ . . _ _ _ _ . _ _ . . . . . _ _
| |
| 50
| |
| 'l ''3. DEDERER: That's correct.
| |
| l-2- DR. SCHROCK: All right,
| |
| }
| |
| o 3 MS. DEDERER:
| |
| ~
| |
| l- And just to give you some idea of 4' ~ the relative size of the flow areas,-I put together this
| |
| !- 5- table. It gives you the size of the flow areas at-various 6 locations vertically in the vessel.
| |
| l l 7 Starting with the core, you see the fuel assembly i
| |
| 8 flow area is the largest flow area. The top nozzle that we 9 just discussed, the structure sitting on the table there, '
| |
| 1 L 10. then has a reduced flow area.
| |
| i l 11 DR. WALLIS: By area you mean a flow area of each 1
| |
| l 12' one of these things, right? !
| |
| 13 MS. DEDERER: Yes, this is per assembly, per fuel !
| |
| 1 14 assembly per single structure in the upper plenum. This is t
| |
| ' l L 11 5 just one of each.
| |
| I l
| |
| 16' DR. WALLIS: But the characteristic dimension of l l
| |
| L 17 this thing has something to do-with the size of these slots i 18 and holes. l 19 MS. DEDERER: That's correct. I was just going to l 20 say that. We also modelled the wetted perimeter so that we i
| |
| L 21 calculate the correct hydraulic diameter, so actually when i
| |
| 22 you are looking at this table the smallest flow area is the 23 freestanding mixers, but the smallest hydraulic diameter is !
| |
| L 24 actually at the top nozzle, because it has such a high 25- wetted perimeter. ''
| |
| L i
| |
| ['') ANN RILEY & ASSOCIATES, LTD. ;
| |
| 11\m)
| |
| Court Reporters '
| |
| 1025. Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034
| |
| | |
| 51 ,
| |
| 1 DR. WALLIS: This is designed by an artist.
| |
| ( 2 -[ Laughter . )
| |
| )
| |
| 3 MS. DEDERER: I don't have any knowledge of the 4 designer.
| |
| i 5- . DR . SCHROCK: I guess you are contending you can 6 have the same total cross sectional area for flow and
| |
| ; 7 simultaneously the same wetted' perimeter?
| |
| l
| |
| : i. 8 MS, DEDERER: I'm not sure I understood your 9 question.
| |
| 10 DR. SCHROCK: Well, from what I heard you say, it
| |
| '11 sounded as though you can model the actual hardware using 12 - some fictitious larger flow channel representing some number !
| |
| 13 of channels that you lump together and I think I hear you l
| |
| -14 saying that you do that with both the same cross-sectional
| |
| () 15 16-area and the same perimeter.
| |
| MS. DEDERER: The same meaning we.take one l 17 assembly and multiply it by the number of assemblies being j 18 represented, is that what you are saying?
| |
| 19 I am not sure I am getting your point --
| |
| 20 DR. SCHROCK: That's all right. It will come out 21 subsequently when we look at the model.
| |
| 22 MS. DEDERER: All I want to point out is that if 23 we are modelling, say, 10 assemblies, we have 10 times the l' 24' single area, we also have 10 times the wetted perimeter so l
| |
| 25 you get the same hydraulic diameter, which is the important '
| |
| s-ANN RILEY & ASSOCIATES, LTD.
| |
| !C_,)/ .
| |
| Court Reporters
| |
| , 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036
| |
| ! (202) 842-0034 l
| |
| r f.
| |
| l
| |
| | |
| {'
| |
| 52 1 parameter when we are calculating the CCFL.
| |
| 2 So in summary, what we have shown here is the UPI d(~T 3 plant geometry. We are modelling the CCFL region the same as 4 what we did in the three and four loop plants in the way 5 that we put together the model.
| |
| 6 The only thing that is different is the amount of 7
| |
| detail that we use in the number of channels and the number 8 of nodes, but the actual modelling of CCFL is the same as 9 what we did in the three and four loop plants, so we haven't 10 actually changed the approach. All we have done is add 11 additional channels so that we could show the change in 12 asymmetric flow if it was calculated to exist because of the 13 upper plenum injection.
| |
| 14 DR. WALLIS: How could it be asymmetric if you ry
| |
| ( j 15 then lump it all together in the first node on top of the 16 core?
| |
| 17 MS. DEDERER: It's asymmetric in the upper plenum.
| |
| 18 You can have water going down the outer channel and no water 19 going down the inner channels and actually we do show that 20 in our vessel models in our calculations.
| |
| 21 The water goes down the outer channels and not 22 down the inner channels.
| |
| 23 Once you get that water going down the outside, it 24 goes down to the bottom of the core and then we are 25 reflooding the core.
| |
| /~N ANN RILEY & ASSOCIATES, LTD.
| |
| !() Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 a
| |
| | |
| 53 l 1. DR. WALLIS: I guess what I mean by symmetric is l
| |
| ^T 2 the same all around, and you are saying that the inside is
| |
| .(d 3 different than the outside -- that means not symmetric. But 4 'it is still cylindrically symmetric.
| |
| 5 MS. DEDERER: Right. What I meant by asymmetric 6 was in the upper plenum you could have asymmetric flow, but 7 you're right. Right below the upper core plate when we lump 8 it together then it is no longer that way.
| |
| 9 DR. WALLIS: Physically you would think that there 10 are forces tending to even things out in this big upper 11 planum and then this little constriction that adaptor plate,
| |
| ! 12 between the adaptor plate and the upper-core plate there 13 really isn't much to even things out, so physically,
| |
| ! 14 assuming everything is mixed up in there, is somewhat a leap t
| |
| i p) 15 of faith or something.
| |
| 16 MR. TAKEUCHI: Professor Wallis, the thing is the 17 model is injected at the corner of the upper plenum but 18 water is distributed almost evenly with upper core plate.
| |
| l- 19 That is the basic reason why we could use that lumped -- at 20- 'the core plate.
| |
| 21 DR. KRESS: It is already distributed around the
| |
| -22 four areas.
| |
| I 23 MR. TAKEUCHI: Yes, that is what I was just L 24 saying.
| |
| l 25 DR. FONTANA: The injection just comes in from the i-l l
| |
| ! ('N ANN RILE ( & ASSOCIATES, LTD.
| |
| '\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1 . . .-
| |
| | |
| 54 1 side. There is no attempt to distribute it or anything like
| |
| . '2 that?
| |
| 3- MS. DEDERER: That's correct.
| |
| 4 DR. FONTANA: Okay.
| |
| 5 MS. DEDERER: We inject on one side and we put 6 together the model such that if it was predicted that the 7 flow would stay in one quadrant, we would be able to show 8 that, but as Kenji just mentioned, it actually does pool in
| |
| , 9 the upper core plate so all the detail maybe wasn't 10 necessary in hindsight but we wanted to be able to predict 11- it if it was going to occur.
| |
| 12 DR. FONTANA: The question might be -- the next 13 question might not be relevant but since the flow ends up 14 going off the bottom of the fuel assemblies anyway, what is
| |
| () 15 the advantage of upper plenum injection as compared to
| |
| : 16. injecting into the cold leg?
| |
| 17 MR. NISSLEY: Actually the upper plenum injection 18 design predated the cold leg injection design.
| |
| 19 DR. FONTANA: Because it's there -- okay.
| |
| 20 MR. NISSLEY: It's historical.
| |
| 21 DR. FONTANA: Okay.
| |
| 22 MS. DEDERER: Yes. Because we built the two loop 23 plants with upper plenum injection before we did three and 24 four loop plants with strictly cold leg injection.
| |
| 25 Are there any further questions on plant geometry?
| |
| (~ ANN RILEY & ASSOCIATES, LTD.
| |
| (, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| P 55 1- [:No response.]
| |
| [' 2 MR. NISSLEY: We've obviously promised to show you D] 3- a lot of things.later, and I think we're about to start 4 doing_that.
| |
| 5 This next topic _is entitled "PIRT Results for 6 Upper Plenum Injection Phenomena." At the beginning we'll 7 touch on the PIRT changes, and then-we'll do some modeling 8 sensitivit*, studies which are relevant to some of the
| |
| . I 9 previous topics of discussion.
| |
| 10 The use of the PIRT for upper plenum injection 11 best-estimate LOCA application is really the same as in 12 other applications. It allows you to identify highly ranked 13- phenomena which you then need to assess in subsequent steps l 14 of the CSAU process, such as what tests do you need to
| |
| () 15 assess your code against and how do you range the physical 16 models in the PWR calculations.
| |
| 17 After you've developed the PIRT, one thing you L 18 have to keep in mind is it really is a guide, and the i
| |
| 19 subsequent steps of CSAU actually clarify what the 20 importance of the phenomena is. And it can go either way.
| |
| 21 Sometimes you'll find that phenomena that you considered to 22 be highly ranked, once you range them over the expected l
| |
| l 23 range based on experimental data, it may not actually impact l 24 your results that much. This was shown in the original CSAU i
| |
| 25 work, and Westinghouse has had similar experience.
| |
| t [''T ANN RILEY & ASSOCIATES, LTD.
| |
| I \_ /
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 4
| |
| | |
| 56 1 DR. FONTANA: So the final PIRTs are not iterated
| |
| (~h 2 on the basis of what you learned in --
| |
| \vl 3 MR. NISSLEY: The only place where I think there 4 may have been updating of the PIRT throughout the process 5 may have been in AP600. It could be done, but it was meant 6 to be an initial guide, and typically we have not gone back 7 and backfitted that, although there's no reason you can't.
| |
| 8 Conversely, you may find as you're doing your 9 sensitivity studies perhaps even more importantly the things 10 that you thought weren't all that important, your 11 calculations are showing that you may be much more sensitive 12 than you thought, and then you have to rethink your 13 development of the overall uncertainty methodology.
| |
| 14 As one might expect, our conclusion was that upper p)
| |
| (s ,
| |
| 15 plenum injection is expected to primarily affect the upper 16 plenum region and the phenomena there, and the core region 17 as well.
| |
| 18 DR. SCHROCK: The knowledge of the plant response 19 is based on a code prediction. Is that right?
| |
| 20 MR. NISSLEY: Yes. Yes, it is.
| |
| 21 DR. SCHROCK: And in the case of your small break, 22 which is also under discussion, for the other plants, I 23 noticed that the concept of or your knowledge of what the 24 response is is based on evaluation model calculations. By 25 that I mean Appendix K. You seem to use evaluation model in l
| |
| l l CX ANN RILEY & ASSOCIATES, LTD.
| |
| (_,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| _ _ - _ . _ _ _ _ . . - - _ _ _ _ _ - _ - _ . . . . . ~ . _ _ . . . _ .
| |
| 57 1 a.different meaning than I thought was the accepted one.
| |
| /''t 2 That's a separate ~ point. But best estimate versus L ,]
| |
| 3 evaluation model'to me has always meant a best estimate, the 4 best calculation you can do given current technology as 5 opposed to something that follows the mandates of Appendix K
| |
| (
| |
| 6 for the EM calculation. That's a secondary. point.
| |
| 7 The point that'I'm making here has to do with the 8 fact that for your small break, what you're saying is that 9 your PIRT is based on knowledge of the transient acquired 10 from code calculations using the Appendix K restrictions.
| |
| 11 And so I want to clarify what you're doing here. Are-these 12' best-estimate calculations that give you your knowledge of l 13 what the transient is, or are these Appendix K calculations?
| |
| 14- I don't think the difference is trivial, because f() 15 . we've seen examples where the phenomena may be quite 16 different-in certain. parts of the transient based on those 17 two methods of analysis.
| |
| 18 MR. NISSLEY: These are based on calculations with 19 the best-estimate WCOBRA/ TRAC code.
| |
| 20 DR. SCHROCK: So this is not the same as the 21 AP600 -- or excuse me, not the same as your small-break 22 evaluation that we looked at last month.
| |
| L 23 MR. NISSLEY: Okay. If I could comment on that, 24' at this point, you know, if you think of the CSAU process, 25 the PIRT development comes fairly early and guides some of I
| |
| i I
| |
| /"N ANN RILEY & ASSOCIATES, LTD.
| |
| [( ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 e - -
| |
| | |
| 58 1 the later steps, one of the most important being obviously
| |
| './~}
| |
| t
| |
| . N,l 2 the assessment against experimental data. And in your 3 assessments against experimental data, once you start doing 4 that, thet does begin to show some of the weaknesses in what 5 you have referred to as evaluation model, and I'll call them i
| |
| 6 Appendix K just to clarify that. '
| |
| 7 You may find that some of the things, if your 8 knowledge base going in is primarily Appendix K, at that 9 point in time that's the best you know. As you continue 10 through the process, compare with experimental data, and use 11 best-estimate codes and acquire that knowledge base, you may 12 change your rankings.
| |
| 13 DR. SCHROCK: I don't think that this is correct 14 at all. Best-estimate codes have been available since
| |
| [V) 15 before CSAU was a fact of life, and to imagine that you're 16 getting better understanding of the plant performance from 17 the code which was compliant with Appendix K, then whatever 18 your best-estimate code is at any point in time is quite 19 erroneous.
| |
| 20 MR. NISSLEY: Excuse me. I probably should not be 21 speaking to the best-estimate small-break process in that 22 I'm not directly involved in that. I think it would be more 23 beneficial to focus on the upper plenum injection here, and i
| |
| 24 our PIRT assessments were based on best-estimate l 25 calculations with WCOBRA-TRAC.
| |
| /~N ANN RILEY & ASSOCIATES, LTD.
| |
| k' ~ '). Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 59 1 As we continued through the process, we still did 2- discover, this should always be a learning process, that we 3 .have overranked some phenomena and.underranked others.
| |
| 4 This is the'PIRT results for the large-break UPI 5- versus-the three- and four-loop applications. The CSAU
| |
| '6 rankings are also shown for completeness.
| |
| 7 The hot assembly location here we originally 8 ranked relatively high based on some of the considerations I l
| |
| 9 that Sue just discussed. As I've also said, our later work i
| |
| 10 showed that we had flooding at the interior locations in the l 11 . core in any of these cases such that the actual calculated l 12 results indicate this may have been overranked. 3 l
| |
| 13 In terms of entrainment and deentrainment, you l 14 have a cituation where you have a horizontal jet of j l[V ) 15- subcooled water impinging on a variety of structures causing 16 breakup of the jet into' droplets, films, sheets, et cetera.
| |
| )
| |
| 17 Some of that will subsequently deentrain on other ;
| |
| 18 structures, and in addition you do have to consider 19 entrainment of the hot legs.
| |
| 20 For phase separation we ranked this higher because 21 of the pulling phenomena. Obviously countercurrent flow 22 drain and fallback is ranked very high because that's the 23 mechanism to get the water into the core. And condensation 24 is important because you have superheated steam exiting the 25 core at most of the locations, and you have subcooled water i
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| ;g'')N
| |
| !(, Court Reporters
| |
| , 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036
| |
| ! (202) 842-0034 l
| |
| l L
| |
| 1
| |
| | |
| .___y 60 1 being. injected into the upper plenum.
| |
| -2. DR. KRESS: Refresh my memory now. A large number j k means it's more highly ranked?
| |
| l- !
| |
| I 4- MR. NISSLEY: Yes, 9._is the highest.
| |
| l-l_
| |
| 5 DR. KRESS: 9 is the' highest. l l
| |
| [ _6~ DR. WALLIS: I've always wondered about PIRT. You l 7 have this ritual you.go through and you rank things. It' L 8 seems to me that's nice to do, but what does it mean? If I
| |
| 9 something is high, then it should_mean'that you apply some ;
| |
| 10- ' more stringent standards to evaluating your comparison
| |
| -11 between prediction and' reality or something?
| |
| l 12 MR. NISSLEY: I think at the first step in the l
| |
| 1 l' 13 process that is-true. Those items get.the most attention.
| |
| 14 But if you're doing a reasonably complete job, you're also j
| |
| -15 -looking at the lower-ranked items, i
| |
| 16 DR. WALLIS: There isn't some discipline that l'
| |
| ! 17 forces you to apply more stringent standards to the high l-L 18 numbers, it's just a qualitative thing, isn't it?
| |
| L
| |
| [< :19 MR. NISSLEY: Yes. And that was my point about 20 the PIRT being an initial guide that you can't completely l live by. You have to update it as you acquire more 22 knowledge.
| |
| l 23 DR. WALLIS: But you could have something which is I24 ranked high all the way through and yet is modeled in a very 25 poor way, and that survives, unless there's a discipline t
| |
| lj ANN RILEY & ASSOCIATES, LTD.
| |
| ig Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 L
| |
| i (202) 842-0034 l
| |
| 7, . . . . , - - - , ,,, -
| |
| | |
| .. ._ _ . _ . _ . . ~ . _ . _ . _ . . . _ - - _ . - . _ . _ _ . - . _ . _ _ . _ _ _ _ _ _ _ _ . .
| |
| l.
| |
| 61
| |
| -1 that"says'you've'got to somehow apply some yardstick for !
| |
| 2' -measuring quality to the high ones_that you don't apply to
| |
| {' ,
| |
| 3 'the low ones.
| |
| .4- DR. KRESS: I think that discipline comes in when l 5 you do the step that says compare the code with the !
| |
| L 6- experiment. You make sure. your experiment focuses cn1 those I 1
| |
| 7 'high phenomena, i
| |
| 8 DR. WALLIS: But somebody has to do that, make l 9 sure --
| |
| l 10 DR. KRESS: Somebody has to be sure you're 11 ' focusing on the high-ranked phenomena in the experiments.
| |
| 12 MR. NISSLEY: And we'll be seeing evidence of all
| |
| .13 these items in later comparisons with test data.
| |
| l-14 DR. SCHROCK: Are you going to tell us who the L; 15 PIRT. team was?
| |
| l L 16 '. MR. NISSLEY: Yes, I will. These were all current 17 or former Westinghouse employees, and I know that's been a 18 point of. contention before. I can name them, if you'think 19 that's productive. But I think --
| |
| 20 DR. SCHROCK: Well, I think it's essential. The 21 credibility of any PIRT is dependent upon the credentials of 22 the group of experts that create it. That's fundamental in 23 the process.
| |
| 24 MR. NISSLEY: Okay. Let me name some of the 25 people-involved.
| |
| .(''
| |
| ; E\s,e)
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| (- Washington, D.C. 20036
| |
| ! (202) 842-0034 t
| |
| | |
| 62 1 The current Westinghouse employees involved in g/'} 2 this PIRT development were Mike Young, who I believe you're V>
| |
| 3 familiar with from previous presentations, myself, Sue l 4- Dederer and other UPI analysts, several former Westinghouse 5 employees including Dr. Hochreider, who is now at Penn 6 State, and Dr. Steve Pajorek, who's at Kansas State 7 University. These people were all involved in the 8 development of this PIRT.
| |
| 9 Again, once you do the PIRT, you're not done. You 10 have to continue to investigate and reevaluate the 11 importance of the phenomena.
| |
| 12 DR. SCHROCK: That's okay, but the PIRT has to be 13 done in'a structured way, and it needs to be documented. So E14 -I think, at least speaking for myself, it's of no use to A
| |
| (J 15 have you present numbers if you want to be intentionally 16 vague about the credibility of those numbers. What you need 17 is to be right up front about who are the people that 18 determined these numbers and where is the report of their 19 activity that resulted in these numbers.
| |
| 20 When you look at the CASU exercise, you will find 21 such a report.
| |
| 22 MR. NISSLEY: In our original submittal, we had a 23 detailed discussion of the casis for the selection of the
| |
| :24 - . .inerical values .
| |
| That has been updated and would be 25 included in the final document. It does not have people's 4
| |
| [~' . ANN RILEY & ASSOCIATES, LTD. '
| |
| \, Court Reporters {
| |
| 1025 Connecticut Avenue, NW, Suite 1014 ,
| |
| Washington, D.C. 20036 )
| |
| (202) 842-0034 i
| |
| | |
| 63 1
| |
| ~
| |
| resumes or professional histories.
| |
| [' i 2 DR. SCHROCK: I didn't ask for resumes. l l 1 I .
| |
| )
| |
| 3 ~DR.:WALLIS: I think we need to move on. I would i
| |
| 4 note for the record-that our consultant, Dr. Novak Zuber, i
| |
| 5 has joined us.
| |
| L 6 DR. ZUBER: I had a problem with my car. I'm 7 still kicking.
| |
| 8- DR. WALLIS: You're not allowed to ask'all the 9' questions you might have asked if you had been.here. I 10 [ Laughter.)
| |
| 11 DR. KRESS: Before we. move on, can I ask one more 12 question about CSAU?
| |
| 13 DR. WALLIS: Sure.
| |
| 14 It seems to me like the weakness in DR. KRESS:
| |
| j .
| |
| 15 CASU that I have seen is that the experts may not identify 16 some important phenomena as being as important as it should.
| |
| l 17 Therefore, it may not get very well modelled in your code 18 because of that focus, and it may not show up as an 19 important thing to simulate in the experiments, so that in a 20 circular process, you end up with something that may have 21 been important and won't show up in your test and didn't get 22 medeled correctly, and by virtue of recalculating with the 23 code,.it doesnit show up as being important.
| |
| 24 Is that potential trap caught some way in the CSAU 25 process that I'm not aware of?
| |
| , . ANN RILEY & ASSOCIATES, LTD.
| |
| l Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
| |
| | |
| 64 '
| |
| 1 MR. NISSLEY: I don't think it's categorically ,
| |
| (. 2 impossible-that an important item could --
| |
| l 3- DR. KRESS: Could slip through the crack.
| |
| 4 MR. NISSLEY: -- f all throuc h the crack.
| |
| l l 5- DR. KRESS: Yes. It's not likely that, you know, '
| |
| 6- you missed something that was -- well, okay.
| |
| l 7 MR. NISSLEY: Well, we identified several things 8 in the three and four-loop beyond the original CSAU. I
| |
| -9 mean, it's a continuous learning process, l 10 DR. ZUBER: May I -- it was just for this reason i i
| |
| l 11 that the comment that Virgil made is very important. The i 12 capability of the CSAU really rests on the capability and 13 credibility of the papers,'and the numbers don't mean 14 anything unless you have paper. I think it's essential.
| |
| t#'l 15 DR. WALLIS: We have to move on, if we can.
| |
| ,O A6 MR. NISSLEY: Okay.
| |
| 17 DR. WALLIS: I think these questions are 18- important, but --
| |
| 19 MR. NISSLEY: Okay. I will try and pick up.
| |
| 20 DR. WALLIS: I think the real meat of the 21 Westinghouse presentation is yet to come.
| |
| 22 MR. NISSLEY: Okay. The one point I would make i 23 here in the core is that when we developed this PIRT, we had
| |
| ~24 already done the calculations that showed that cooling of i
| |
| 25- the hot assembly is by bottom-up reflood. So there are i
| |
| /~% ANN RILEY & ASSOCIATES, LTD.
| |
| (),
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| l
| |
| | |
| . . -- -. _ . ..._ _ __. _ . m _._m. . . _ _ _ _ _ _ _ - _ . _ .
| |
| { . l 1
| |
| 65 1 really no different rankings here.
| |
| 2 However, in hindsight, given the flow pattern that
| |
| .3 does exist'within the core where you have the draining in 4 the outer low power regions down to the bottom of the core
| |
| ,5 and then the bottom-up reflood process, that in hindsight, 6- :the 3-D flow should have been ranked higher. However, inn
| |
| (
| |
| i 7 did assess it and include it in the uncertainty assessment.
| |
| t 8 Conclusions of the highly ranked phenomena. I'm
| |
| ! 9 going to skip over the early ones here. This is the list we r'
| |
| [ -- 10 came up with for four new plants which was very similar to l
| |
| 11 CSAU. Same parameters-apply with the addition of what we 12 have called upper-plenum drain distribution, and that 13 reflects., you know, how the injected water in the upper 14 plenum is.actually getting down into'the core has to do with
| |
| ' !''' '15 pulling phenomena, CCFL, et cetera.
| |
| ! : V.) '
| |
| 16 I'll now talk about some the PWR sensitivity 1
| |
| 17 studies looking at the sensitivity to various injection !
| |
| l 18 modeling methodologies. We'll talk about injecting the 19 liquid in a continuous liquid field versus drops of -- where
| |
| ( 20 we've done studies with different drop sizes.
| |
| 1 21 The code is only capable of injecting one drop 22 size, so we have done parametric studies using different 23 uniform drop sizes.
| |
| 24' We have also looked at injecting into the outer
| |
| [ 25 region of the upper plenum only versus a spatially I
| |
| l ANN RILEY & ASSOCIATES, LTD.
| |
| El j- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l l Washington, D.C. 20036 L
| |
| (202) 842-0034
| |
| | |
| + - . _ _ . . . - - ~.. . ..- - .-. .- - - . . ..
| |
| I l- 66
| |
| ! 1 distributed flow and combinations of those.
| |
| l'
| |
| -2 We have done some hot assembly location studies 3- which I have a few slides on. We have done nodalization L 4 above the average assemblies.
| |
| .5 In addition, while I don't have slides in here, I 6 have some backups I'll show, and we can make copies. The 7 question about how we go from four quadrants in the outer 8 region of the upper plenum and then basically smear that as 9 we go below the upper core plate, we've done nodalization 10 studies for that and I have those as backups and we will 11 make copies.
| |
| 12 DR. SCHROCK: Could I ask about the drops that 13 you're modelling? Is there a spray into the upper head, or 14 what is the motivation for that part of your study?
| |
| :. 15 MR. NISSLEY: I'll get to that. I'll show some of 16 the background for why we did the ranging studies that we 17 did. There's some experimental evidence that led us.to 18 select the parametric studies that we did.
| |
| 1 19 DR. SCHROCK: It's a simple question.
| |
| 20 MR. NISSLEY: My slides I believe cover it.
| |
| 21 The low head safety injection water enters the 22 upper plenum as a horizontal jet, and after traveling only a 23 short distance, it begins to impact on the structures in the 24- upper plenum.
| |
| 25 At the injection location, some of the structures ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| : i. Os 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 67 1 that are present lower in the upper plenum are no longer
| |
| ('')
| |
| C/
| |
| 2 present at this elevation. The circles are support columns 3 and the squares are guide tubes.
| |
| 4 The jet is injected 80 degrees away from the hot 5 leg location and almost immediately impinges on support 6 column and a guide tube here.
| |
| l 7 That jet impact will cause formation of droplets, '
| |
| 8 falling film, sheets of water, et cetera, some of which will 9 de-entrain on other structures. Some of the injection may 10 also penetrate further within that forest of structures to 11 the inner region of the upper plenum. So the obvious l
| |
| 12 question is, how do you go about modeling this?
| |
| 13 Our nominal modeling technique is to inject the 14 water as a continuous liquid field into the outer region of
| |
| () 15 16 the upper plenum.
| |
| injection ports.
| |
| This particular slide shows both 17 With your limiting single failure assumption, you 18 lose one of these, so what we're doing is injecting in our 19 nominal modeling into this channel a continuous liquid field 20 with no momentum.
| |
| 21 DR. ZUBER: What do you mean by nominal model?
| |
| 22 MR. NISSLEY: Nominal model?
| |
| 23 DR. ZUBER: Yes.
| |
| 24 MR. NISSLEY: That's our base way cf modeling it.
| |
| 25 DR. ZUBER: What do you mean by no momentum.
| |
| l'
| |
| /~'\ ANN RILEY & ASSOCIATES, LTD.
| |
| \s / Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| ! Washington, D.C. 20036 l (202) 842-0034
| |
| | |
| .. . . - _ _ . . . . . - .. - . , ~ ~. - . . . . . . - . . . . - . - .
| |
| 68 1 MR. NISSLEY: No momentum is --
| |
| l-l 2 DR. ZUBER: What is the physics?
| |
| \
| |
| : s. :
| |
| 3 MR. NISSLEY: That the jet comes in --
| |
| l-
| |
| [ 4 DR. ZUBER: It has a momentum when it gets in, ,
| |
| -5 right? l 6 MR. NISSLEY: Sure.
| |
| 7 MR. ZUBER: Then it disintegrates. How do you 8 model that without the momentum?
| |
| 9 MR. NISSLEY: We introduce this as a mass source 10 with no momentum,-okay? This is a sketch of where the jet, 11 which is the outer dash lines here, first impacts on the 12 support column and the guide tube -- very low porosity l
| |
| 13 through to the inner region.
| |
| 14 We're basically.saying all the momentum -- our j ). 15 assumption is all the momentum is lost here, and you have a 16 mass source, you've created a mass source with no horizontal 17- or vert.ical momentum.
| |
| 18 DR. KRESS: You are neglecting the energy compared I 19 -- you just don't use that energy anywhere.
| |
| 20 MR. NISSLEY: Right.
| |
| 21 < DR. SCHROCK: What is the typical velocity of that 22- jet?
| |
| 23 MR. NISSLEY: About 44 feet a second.
| |
| 24 DR. SCHROCK: That's pretty high velocity.
| |
| 25 DR. ZUBER: Don't you think this will have a great b
| |
| l-I /'' ANN RILEY & ASSOCIATES, LTD.
| |
| l L(,)) Court Reporters L
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 69 1 impact how the thing disintegrates?
| |
| (~N 4
| |
| 2 MR. NISSLEY: It certainly will on the breakup.
| |
| Q );
| |
| 3 After the breakup --
| |
| 4 MR. ZUBER: Well, after the breakup, it's over. I 5 mean, then you come and you break it up.
| |
| 6 MR. NISSLEY: Right.
| |
| 7 MR. ZUBER: This is a function of the momentum 8 MR. NISSLEY: Right.
| |
| 9 MR. ZUBER: If it's right, then how do you justify 10 your approach?
| |
| 11 MR. NISSLEY: Well basically, for all intents, the 12 purpose is saying the water shows up here, the -- the impact 13 has already occurred and it'c showing up as a mass source.
| |
| 14 DR. KRESS: When you range your particle size, --
| |
| [j) s_
| |
| 15 MR. NISSLEY: Right.
| |
| 16 DR. KRESS: -- you probably have a range that 17 somehow may be related to that momentum.
| |
| 18 MR. NISSLEY: It's based on experimental data --
| |
| 19 DR. KRESS: Oh, it's based on experimental --
| |
| 20 MR. NISSLEY: -- with different impact velocity --
| |
| 21 DR. WALLIS: So that somehow captures this 22 momentum.
| |
| 23 MR. ZUBER: That's a good question. I mean, then 24 the question which follows is, what experimental base do you 25 have to determine the spectrum of these droplets if you --
| |
| 7"N ANN RILEY & ASSOCIATES, LTD.
| |
| ('~' ) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . _ __ . . _ _ ___ _ _ _ _ _ _ _ , ._ _ _. . ~ .__
| |
| L
| |
| . 70 t
| |
| l' MR. NISSLEY: Can I.get to that in a minute? ,
| |
| l 2 :That's really a lead in to my next set of slides.
| |
| ' .("N
| |
| \.
| |
| )
| |
| - /'
| |
| o 1
| |
| 3 DR. FONTANA: You do assume that the water level l' 4 that you end up with is uniform in that particular quadrant?
| |
| 5 MR. NISSLEY: Yes.
| |
| -6. DR. FONTANA: Yes. I I
| |
| 7 MR. NISSLEY: But it's basically going to fall by 8 gravity.
| |
| 9 ~DR. FONTANA: Okay. l
| |
| : 10. MR. NISSLEY: So the basis for the ranging studies
| |
| .11. is a combination of several things. One is that sketch I i l
| |
| 12 showed where the jet impacts on the support column and the '
| |
| 13 guide. tube.
| |
| 14 When we looked at the porosity, we saw that the N .15 ' porosity is less than 20 percent for the first impact,
| |
| :[&-
| |
| 16' implying that certainly there is going to be some
| |
| '17 penetration of some droplets, et cetera, into the inner 18 region of the upper plenum, but it --
| |
| 19 DR. KRESS: That porosity is a cross section?
| |
| 20 MR. NISSLEY: Yes.
| |
| 21' DR. KRESS: Okay.
| |
| 22- MR. NISSLEY: Horizontal.
| |
| 23 DR. KRESS: A horizontal --
| |
| ~24 MR. NISSLEY: If you're looking at the jet, what 25 does it see its porosity to get through that first impact.
| |
| l (''g ANN RILEY & ASSOCIATES, LTD.
| |
| l ( _j . Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 I
| |
| Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| . .. . ~ . - . . . - - . - - . - - . . . ~ . . . . - . - . .-. ..-.- . . . - _ . ~ - - - . , . . ~
| |
| l 71 i
| |
| 'I
| |
| 'l DR. KRESS: Oh, the first impact porosity. I l
| |
| A 2- MR. NISSLEY: Right.
| |
| ~s I 3' DR. KRESS: Okay. )
| |
| l I see. l l
| |
| 4 MR. NISSLEY: So for the --
| |
| I 5 MR. ZUBER: What's.the diameter of the jet?
| |
| 6 MR. NISSLEY: -Four inches.
| |
| '7 MR. ZUBER: Four inches. And what is the 8 porosity?
| |
| '9 MR. NISSLEY: About 44 feet a second.
| |
| 10 MR. ZUBER: And-what are the open areas? You have 11 a four-inch jet coming.
| |
| 12 MR. NISSLEY: A four inch.
| |
| 13 MR. ZUBER: Yes, four inch. Okay. Then what does 14 it see? I mean, what is the porosity? What free area it i
| |
| : 15. sees?
| |
| {
| |
| 16 MR. NISSLEY: This is a four-inch jet between 17 these dashed lines impacting on the support column and a 18 guide tube. The direct free path is this very small amount 19 here.
| |
| 20 DR. KRESS: That's the 28 percent?
| |
| 21 MR. NISSLEY: Now, there are some slots in here.
| |
| 22 There's a limited porosity through here. The total porosity 23 through here and here is on the -- it's less than 20 percent 24 as the slide indicates.
| |
| 25 DR. SCHROCK: But is your modeling guided in any l
| |
| l l.
| |
| '(g ANN RILEY & ASSOCIATES, LTD.
| |
| ,- Court Reporters l (_f 1025 Connecticut Avenue, NW, Suite 1014 l
| |
| Washington, D.C. 20036 (202) 842-0034
| |
| | |
| _....._..-__m.. . - _ _ _ . .._ .. _ . . - . _ _ . . _ . _ . _ _ _ _ _ __ __ _
| |
| 72 1 way by-experimental data?
| |
| j 2 MR. NISSLEY: Yes.
| |
| 3 DR '. SCHROCK: What experiments?
| |
| 4 MR. NISSLEY: Westinghouse conducted tests in 5 stagnant air, an important point, which showed that upon 6 impact, it broke up into -- the jet broke up into, in 7 addition to films and sheets, drops on the order of a L 8 quarter of an inch to a half an inch.
| |
| !. 9 Now, again, this was in stagnant air. There was a l-10 Dartmouth study performed where they had a simulated UPI .
| |
| 11 design. It had air injection in it, and drop sizes -- this t l
| |
| j 12- is written incorrectly. This should say reported drop sizes l 13 of less than-or equal to .007 feet for air velocities in i
| |
| L 14 excess of 15 feet a second. And in the Dartmouth study,
| |
| () - 15 16 velocities that high were adequate to basically entrain almost all the liquid out the hot legs.
| |
| l This was a very 17 high air velocities. With those high air velocities, where ;
| |
| 18 they could entrain most of the liquid out, they estimated 19 drops of .007 feet or smaller.
| |
| p 20 DR. ZUBER: How did you model -- perform this l 21 experiment, number 2? What was the geometry, what were the 22 flow rates, the diameter? Was it prototypical or not?
| |
| 23- MR. NISSLEY: Yes, it was. We used -- we designed i 24 wooden structures that had the right cylinder size for the i
| |
| 25- support column and the right square enclosure for the guide
| |
| '(C ANN RILEY &-ASSOCIATES, LTD.
| |
| (, ' Court Reporters
| |
| , 1025 Connecticut Avenue, NW, Suite 1014
| |
| [ Washington, D.C. 20036 l (202) 842-0034 1
| |
| | |
| Il 73 i
| |
| l 11' < tube, and put'a' design nozzle size with the right velocity
| |
| ;() 2 on it. I
| |
| ( 3: We had some tapes, I believe, we have-shown to at 4 least.the. staff previously on those tests. But again, that E
| |
| 5 was in stagnant air.
| |
| 6 The last point I-would'like to make from the l
| |
| , 7_ Dartmouth study is --
| |
| I 8 DR.'SCHROCK: When you deal with-droplets, then
| |
| '9 you make the assumption that all of that jet is somehow !
| |
| l 10 converted into droplets-of some size?
| |
| l 11 MR. NISSLEY: We did a mix. Okay. We either said 12 all continuous liquid or all drops. But we did a mix of l l13 spacial distributions. I'll show you the studies we did in i
| |
| 14 a minute. I'm trying to lay the basis for why we did the II
| |
| 's.) 15 studies we did, 1
| |
| 16 Finally, the Dartmouth test -- they had a series I 17 of tests with different-jet velocities. The one that was )
| |
| 18 closest to our design had an injection angle of 20 degrees, 1
| |
| l 19 and this one has about the same target Froude number as our J l 20 mix of structures.
| |
| 21 In this particular case, I think almost 90 percent l- 22 of the water drained down in these positions right here, and i 23 what we've done with this dashed line is overlay the i
| |
| 24 quadrant that that would correspond to in the upper plenum. ;
| |
| j 25 So from this picture, you can see that although L
| |
| I 4
| |
| I /~' ANN RILEY & ASSOCIATES, LTD.
| |
| Ik,)) Court Reporters 1
| |
| ; 1025 Connecticut Avenue, NW, Suite 1014 I Washington, D.C. 20036 (202) 842-0034 i
| |
| | |
| l l
| |
| 74 1 'the structure arrangement is somewhat different than our
| |
| -actual design, that with the comparable target Froude 1
| |
| .. V. O 3' . number, that the drainage was all occurring within that same 4 quarter,-one-quarter of the upper plenum.
| |
| ; .5 DR. KRESS: I see how the Froude number is 6' important to your drainage, but I don't see how it-impacts 7 the. question of particle size. Could you maybe explain that i 8 to me?
| |
| 9; MR. NISSLEY: I don't think that impacted on the 10 -- I'm really mixing things, i 11: DR. KRESS: You're mixing drain -- l 12 MR. NISSLEY: That droplet size was a different i 13 test than this.
| |
| ~14' DR. KRESS: It's a different test.
| |
| () 15 16 MR.'NISSLEY: Yes.
| |
| DR.'KRESS: Okay. So you're on another test now.
| |
| 17 MR. NISSLEY: Yes.
| |
| 18 DR. KRESS: Okay.
| |
| 19- MR. NISSLEY: They ran a series of parametric t
| |
| 20 studies and extracted as much information as they could out 21 of that. The point on the drop size was when they had an l 22 extremely high air velocity that entrained almost all the H2 3 - liquid out of the hot leg, the estimated drop size was no 124 bigger.than .007 feet.
| |
| 25 DR. SCHROCK: You imagined that the existence of l
| |
| ' [] ANN RILEY & ASSOCIATES, LTD.
| |
| (,/. .
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l (202) 842-0034
| |
| \
| |
| : l. 75
| |
| ! 1 subcooling has no. influence ~on that pattern?
| |
| 2' MR. NISSLEY: Well, in terms of -- yes, I think 3 condensation which would occur would'certainly reduce the 4 air -- or'in that case steam velocities. These tests are 5 all on air, unfortunately.
| |
| -6 But in the case of'sub-cooled water in the 7 presence of' steam, the condensation effect on the steam I 8 would certainly reduce the velocities and would tend to '
| |
| l 9 reduce the amount of entrainment out the hot leg.
| |
| 10 DR..SCHROCK: It~seems to me that that's a key 11 feature of the upper head injection plant, and that you 12 need, in constructing modeling, to understand something l 13 about that interaction.
| |
| L 14 MR. TAKEUCHI: Can I.make one comment? The i
| |
| '([ 15 numbers in those -- the numbers in here are a percentage of
| |
| ,s-16 the liquid collected. Those numbers down there are the 17 percentage of the liquid collected down draining of the film 18 on those rods.
| |
| 19 DR. ZUBER: You measured that?
| |
| i
| |
| ; 20 MR. TAKEUCHI: Yes.
| |
| 1 21 DR. SCHROCK: My point was it's air water, not 22 steam water, with subcooled liquid, and they won't behave in l 23 the same way.
| |
| 24 MR. NISSLEY: And my comment was that the 25 condensation of the steam would reduce -- its velocity would i
| |
| l-ANN RILEY & ASSOCIATES, LTD.
| |
| i g_,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 u . _ _ . .
| |
| | |
| m.. . _ .
| |
| 76 l
| |
| 'l reduce the -- would reduce the amount of entrainment out the L (~'i 2' hot-leg relative'to'what you would get with an air wa'ter'--
| |
| .V 3 DR. SCHROCK: I imagine it.has a dramatic effect on flow 4 distributions in this system, and that what you see in the 5 air water system is quite different from what you'll see
| |
| ~
| |
| 6 with subcooled liquid injection.
| |
| 7 So it seems to me you need to be guided by some 8 . experimentally based knowledge of what happens when you
| |
| ! '9 inject subcooled water into this kind of geometry.
| |
| 10 MR. NISSLEY: We have not done that. I will show 11 you what we have done.
| |
| 12 Based on the various -- the major points we had 13 there were that given the limited porosity for the jet to 14 penetrate into the inner' region, our nominal modeling puts
| |
| . I~ \ - 15 all of the water'in the outer global channel. Recognizing V
| |
| 16: that some of it may extend into the inner region, we'have l
| |
| 17- dome some parametric studies where we put 60 percent of the 18 water in the outer global and 40 percent in the inner 19 global, and those numbers reversed.
| |
| 20 For the original studies with the outer global 21 channel, these were the first cases we looked at, we first 22 used the larger of the Westinghouse observed drop sizes, 23' then we used a lower number, and finally we went down to the 24 size that was reported by Barathan when all of the water was 25 entrained out the hot legs, and then using that drop size, A ANN RILEY & ASSOCIATES, LTD.
| |
| f-(~) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 77
| |
| '1 we did the case where we were spatially distributing it and
| |
| / ')
| |
| (/
| |
| 2 looking at the parametric behavior of that. We looked at 3 that minimum drop size versus continuous liquid.
| |
| 4 What you can see from these results -- right here, 5 I've only shown PCTs at the maximum value; I'll show a few 6 time-dependent slides -- is that there's really not a huge 7 effect on the calculated hot assembly PCT. ~
| |
| 8 DR. FONTANA: Let me understand something. The l
| |
| 1 9 .007 diameter drops, that has more of the water coming in on I i
| |
| 10 the upper plenum injection being lost out through the hot 11 leg?
| |
| 12 MR. NISSLEY: Yes. I would expect this case --
| |
| 13 one of these two cases to have the most entrainment of the 14 hot leg.
| |
| 15 DR. FONTANA: But you don't know how much.
| |
| v 16 MR. NISSLEY: I don't have those slides here.
| |
| 17 DR. FONTANA: Okay. Becaus there's not much 18 difference in PCT.
| |
| 19 MR. NISSLEY: No. You're getting enough water 20 down in any of the cases. You're getting -- the water is 21 draining down the outer channels and you're getting the 22 bottom up briefly.
| |
| 23 DR. FONTANA: Okay.
| |
| 24 DR. WALLIS: It's rather peculiar that drops are 25 the most effective in PCT terms and continuous liquid is g ANN RILEY & ASSOCIATES, LTD.
| |
| g-'/
| |
| (, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| I
| |
| | |
| l 78 1T worse. Continuous liquid is about'the same as the smallest 2, drops. So it's rather peculiar. It's unexpected.
| |
| 3- So you're using PCT as a measurement of the
| |
| ;4 adequacy of the model? l
| |
| '5 MR. NISSLEY: That's what we've done in these 6' cases where we've seen -- these are the -- this has that .04 7 foot case as shown on here. When we've looked at the I
| |
| 8 overall hot rod response and have seen them overlay this j l
| |
| 9 closely, we have not continued with exhaustive investigation l
| |
| 10 of other parameters. Where we have seen more of an impact, 11 this is really our figure of merit when it comes to 12 licensing basis analysis.
| |
| 13- When the overlays are this close, we have not 14 really done a real detailed investigation of --
| |
| () 15 16 --
| |
| DR. WALLIS: The actual arrangement of the dashes there seems to be random orientation.
| |
| : 17. MR. ZUBER: Suppose you really could not-model 18 this at all -- assume that -- and you obtain results like
| |
| .19' this and say everything is fine, I don't have to look any
| |
| '20 further -- I mean, this is the base of your argument.
| |
| 21 MR. NISSLEY: We also did drop size studies with 22 UPTF and what we found is that the results --
| |
| 23 MR. ZUBER: Are you going to discuss this?
| |
| 24 MR. NISSLEY: I have them as back-up material.
| |
| 25 They're not in the package.
| |
| 4 r" ANN RILEY & ASSOCIATES, LTD.
| |
| L( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036-(202) 842-0034
| |
| | |
| ~ _ . - . -.. - . . - .-- - -. - - _ _ . . _ . . .
| |
| 79 3 l' .MR. ZUBER: Well, that's important because -- 1 JI' 2 MR. NISSLEY:
| |
| Okay. What we saw was.--
| |
| U 3= MR. ZUBER: There were experimental data on that.
| |
| : 4. MR. NISSLEY: We took UPTF'and did -- went from 5 continuous liquid to various drop sizes. Down to.about .01 6- foot, there was very little difference in the amount of hot 7 leg entrainment, et cetera, and the -- but when you went 8- down below that value, you started to greatly exceed the 9 amount of entrainn. It up hot legs.
| |
| 10' So we had pretty good agreement with experimental 11 data on both steam flows and liquid flows until we got down
| |
| : 12. to drop sizes less than .01 feet, and that's in the package 13 we sent to you earlier.
| |
| 14 The reason we did not -- that was originally what !
| |
| 1
| |
| () 15 we were using as our basis. That was not accepted by INEL 1
| |
| 16 because UPTF has a jet velocity which is low by a factor of '
| |
| 17 2.4, I believe it is. So the fact that the velocity was off 18 by that much was deemed to be insufficient to use UPTF to 19 make a final conclusion. But that study was done.
| |
| 20 DR. SCHROCK: What is the typical steam velocity 1 21 upward during this? l 22 MR. TAKEUCHI: That number is less than three 23 meters per second, and where the steam velocity is less than 24 three meters per second, entrainment is almost zero T
| |
| 4 25 according to the Dartmouth test.
| |
| )
| |
| (''
| |
| s, ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 80
| |
| -1 DR. SCHROCK: -I'm'having trouble reconciling that
| |
| . ['N f2 - "with what'I heard before. I thought you said that liquid U 3 for~small drops is all carried out with the' steam. Did 'I
| |
| ~4 . hear that right?
| |
| 5 MR. NISSLEY: What I said was that for UPTF, when 6 we went:to drop sizes below .01 feet, that we began to 7 predict much more entrainment than the data showed out the 8 . hot. leg.
| |
| 9 MR. ZUBER: .01.
| |
| 10 MR. NISSLEY: .01. And we have figures in there 11 of hot leg flows, hot leg liquid and hot leg vapor flows as 12 a function of drop size that I'll show you.
| |
| 13 DR. WALLIS: That's a pretty big drop.
| |
| 14 MR. NISSLEY: Pardon me?
| |
| () 15 16 DR. WALLIS: That is a pretty big drop.
| |
| DR. SCHROCK: A tenth of an inch.
| |
| 17 MR. NISSLEY: We are representing 100 percent of 18 the liquid as being drops of that size. We cannot capture 19 .the full spectrum of drop sizes.
| |
| 20 DR. SCHROCK: Do you look at what is a possible 21 trajectory for such a drop horizontally, how far could it go
| |
| -22 before it impacts something?
| |
| 23 MR. NISSLEY: We have not done any studies of that 24- type.
| |
| 25 DR .- SCHROCK: Well, isn't it favorable to have a A ANN RILEY & ASSOCIATES, LTD.
| |
| 3% ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington,-D.C. 20036 (202) 842-0034
| |
| | |
| _. .. . . . _ . _ . _ . _ _ _ _ . . _ _ _ - _ . . _ _ _ . _ _ _ _.. . - _. ..~
| |
| 81 1 lot-of condensation on this liquid that's injected,- and 2 doesn't that get enhanced by making this assumption that all 3 of the liquid suddenly turns into drops?
| |
| 4 DR. WALLIS: I think they were going to tell us j 5- that you varied the condensation coefficient in some way, 6' the parameter, to see if it made any difference.
| |
| ~7' MR. NISSLEY: And we'll show CCTF, UPTF and GECCFL 8 with that raised.
| |
| 9 In the interest of time, let me make a proposal 10 here. I had some other figures in here showing some 11- parameters that we showed to not be important. One of them 12 was hot assembly location. The reason for that is we have 13 flooding in the interior assemblies in many of these cases.
| |
| 14 I also had some slides that showed our basis for
| |
| ( ) 15 eliminating ranging of condensation in the downcomer and 16- lower plenum, and those figures clearly show that if you 17' range it as was done for three- and four-loop plant;, that 18 there's really no impact.
| |
| 19 Rather than showing those slides, what I would
| |
| '20 ' propose I do right now is go to some back-up slides which
| |
| ; 21 'showed that nodalization study we talked about where we had l
| |
| 22 more detail in the low power region as we went from the 23 upper plenum down into the core.
| |
| 24 Would that be acceptable?
| |
| H25 DR. WALLIS: How much time are you going to take?
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| ' '(,f'")Y Court Reporters j- ' 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L-L - -
| |
| | |
| 82 1 MR. NISSLEY: I could do that in less than two 2 minutes without questions.
| |
| (V~)
| |
| 3 DR. WALLIS: Okay. That sounds good.
| |
| 4 MR. NISSLEY: Pardon me?
| |
| 5 DR. WALLIS: That sounds good to me.
| |
| 6 MR. NISSLEY: I'll keep this brief.
| |
| 7 The nominal modeling has the four quadrant --
| |
| 8 DR. WALLIS: Is there a focus problem?
| |
| 9 MR. NISSLEY: I'm going to just show the top of 10 this and then the next page.
| |
| 11 The nominal modeling or base nodalization has the 12 four quadrants in the upper plenum. There's the four open 13 hole extension and the four support column flow mixer 14 extensions collapsing down into a single channel in both
| |
| ( ) 15 cases as we move below the upper core plate. Those are both 16 then connected to a single low power assembly.
| |
| 17 One of the reasons we don't continue that level of 18 detail is unfortunately, we have some hardwired logic in the 19 code that restricts the total number of fill rods that we 20 model.
| |
| 21 What we were able to do with our existing 22 limitations was to do a study with double that amount of 23 noding where we have gone from -- where we have basically 24 kept it consistent. We've actually split these up above the 25 upper core plate and continued down with twice the level of
| |
| (~'T ANN RILEY & ASSOCIATES, LTD.
| |
| ( ,) Court Reporters 1025 Connecticut Avopue, NW, Suite 1014 Washington, 7.C. 20036 (202) 842-0034
| |
| | |
| b 83 -
| |
| o [1 detail down into the core. And again, what we have here is i I
| |
| 2 a very low sensitivity of the calculated hot rod response to 3 that more. detailed nodalization. It's slightly favorable, 4 but not enough to be deemed significant.
| |
| 5 DR. KRESS: Do you know why it gave you lower 6 numbers on the temperature?
| |
| l _7 MR. NISSLEY: I believe in this case, you would .
| |
| 8' have -- I think you have colder water coming down.
| |
| .9 .R. KRESS:
| |
| D You'end up with a colder inlet 10: temperature to the hot --
| |
| 11 MR. NISSLEY: Whatever minor asymmetry you had 12 above the upper core plate you've reduced. That will mix 13 together and now it's maintaining into -- one is going to be 14 slightly hotter than that mixed mean and one slightly
| |
| () 15 16 cooler. So I don't think I can comment beyond that.
| |
| DR. WALLIS: Since you've got several slides that 17 all look very similar, like this, could we move to the 18 conclusions, then?
| |
| 19 MR. NISSLEY: Yes. Those other ones, I had H2O intended to cross over.
| |
| 21 Conclusions from this part of the presentation are 22 that our highly ranked upper plenum phenomena were the hot
| |
| .23- assembly location, entrainment, de-entrainment, phase
| |
| : 24. separation. We did several PWR sensitivity studies to look 25 at how we model upper plenum injection, the hot assembly i
| |
| l l r"T ANN RILEY & ASSOCIATES, LTD.
| |
| (,,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| i 84 1 location, several nodalization studies, and, in an unrelated
| |
| ( }
| |
| 2 study, condensation in the downcomer lower plenum.
| |
| : 3. These were all supporting our continued use of'our 4- nominal nodalization and modeling; however, we need to do 5 assessments of. experimental data and separate effects and 6 integral effects tests to identify the physical models to be
| |
| : 7. ranged to address these, and that will be the topic of -- ;
| |
| 1 8 DR. .WALLIS: Yes, I think that's interesting l
| |
| 9 because the fact that you're peak clad temperature is not l l
| |
| 10 sensitive to the modeling really doesn't prove any way that 11 you have modeled the phenomena correctly.
| |
| 12 MR. NISSLEY: I think the assessments with 13 experimental data -- l 14 DR. WALLIS: All this could be completely wrong
| |
| [V) 15 and it just turns out that the clad temperature is 16 insensitive to that.
| |
| 17 MR. NISSLEY: We did some of these same studies 18 with the test facilities. I'll show you the drop studies we 19 did for UPTF. l 20 DR. WALLIS: So we will hear that? I think, if 21 the committee permits me, I would like to declare a recess 122' for a break, and then we'll come back to this.
| |
| 23 DR. KRESS: We were about to --
| |
| 24 DR. WALLIS: You were about to get restless 25 anyway.
| |
| , ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| ' Washington, D.C. 20036 (202) 842-0034 L
| |
| l l -- ,. - - - _ - . -
| |
| | |
| 85
| |
| '"1 DR. KRESS:- Yes.
| |
| /
| |
| 2- DR. WALLIS:
| |
| ;&. )
| |
| So I think we will break until 3 . quarter to eleven, ten-minute break.
| |
| 4 .R, NISSLEY:
| |
| M Thank you.
| |
| 1
| |
| '5; [ Recess.] I 6 DR. SCHROCK: This is different -- it's broken off
| |
| : 7. 'there.
| |
| ._ 8 -MR. TAKEUCHI: Yes, because you can have a 9 communication ~between here,jupstream and downstream over the 10 -break point. That indicates the guillotine break there is 11 no communication.
| |
| 12 DR. SCHROCK: Have you ever done any testing on a 13 pipe that has longitudinal rupture in it?
| |
| 14' MR. TAKEUCHI: There may be but we don't, we are
| |
| () '15 not comparing this one to the test data, therefore we take
| |
| ^
| |
| 16 the worst case. That is the reason for taking, limiting 17 split --
| |
| ~
| |
| 18 DR. SCHROCK: Well, I think your descriptions are 19~ misleading because I don't think that Westinghouse has any 20 knowledge of critical flow through such irregular geometries 21' as longitudinally split pipes that are tending to open up.t
| |
| -22 MR. 'AKEUCHI: Yes=.
| |
| 23 DR. SCHROCK: The shape of it is quite uncertain 24' but more than that if it-is of any lena'. whatsoever, the 25 fluid coming to it from two different iirictions, having O ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| ._ __ _ _ _ _ -- ._ _-- . _ _ . _ . _ ~ .
| |
| l l
| |
| I 86 l 1 different enthalpies can be more or less parallel critical
| |
| : i. ;
| |
| : 2. ' flow problems through that long slit.
| |
| l [N,/} l 3 If it is instead a smaller pipe that is connected, 4' everything is forced to pretty well mix before it arrives at l 1
| |
| 5 the critical section, that will produce one result, and L
| |
| 6 certainly it would be a presumption to expect that it would 7 be the same as in the first description that I gave you.
| |
| !- 8 MR. TAKEUCHI: Yes. This is -- l 9 DR. SCHROCK: So describing it as a split break to
| |
| : 10 me is quite misleading, because I don't believe you have any 11 knowledge of what happens with a split break.
| |
| . 12 MR. TAKEUCHI: When we talk about split break, we 13 model the pipe and the break point like this, and because we 14 don't have the knowledge of the shape of the break and how O)
| |
| ( 15 the flow rate is coming up, therefore we don't take the 16 benefit of uncertainty but we try to build in conservative 17 by taking a bounding --
| |
| 18 DR. ZUBER: Let me ask you, did you run any 19 experiments on anything similar to what you calculate?
| |
| 20 MR. NISSLEY: Westinghouse says no.
| |
| 21 DR. ZUBER: So there is no experimental basis for 22 any code verification on this model?
| |
| 23 DR. WALLIS: This is probably a regulatory world.
| |
| 24 MR. NISSLEY: This is a deterministically selected '
| |
| 25 representation.
| |
| i ("') ANN RILEY & ASSOCIATES, LTD. ;
| |
| I l \m / Court Reporters j 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| - _ _ _ . _ _ . _ . - _ . _ _ . - _ _ . . - . ~ _ . _ _
| |
| , 87 L
| |
| l
| |
| \.
| |
| 1 DR. WALLIS: But this has nothing particularly to .
| |
| 2 -do_witn the upper plenum injection, is it?
| |
| 3- DR._KRESS: That is-the same in both codes.
| |
| 4 DR. WALLIS: The same in both codes, so maybe we 5 should go on to the things that are special about it.
| |
| 6 DR. KRESS: But you know, we-were still looking at 7 the codes' capability to calculate.
| |
| F. 8 DR. WALLIS: It may well be that this should be 9- looked at too, but that's not --
| |
| l l 10 DR. KRESS: Yes, it may not be here.
| |
| 1 11 DR..WALLIS: -- not part of the scope for this l 12 particular study.
| |
| 13 DP. KRESS: Well, one might ask why wasn't it 14 looked at in the previous approval --
| |
| 1--(p) 15 DR. WALLIS: Right. That's right.
| |
| 16 DR. KRESS: -- the three to four loop.
| |
| 17 DR. SCHROCK: It came up at that time.
| |
| 18 MR. TAKEUCHI: The method is the same as the three 19 and the four loop plant that now exist.
| |
| 20 DR. SCHROCK: Well, I wasn't satisfied with it l 21 then and I am not no'.
| |
| 22 MR. TAKEUCHI: It was licensed though.
| |
| 23 DR. SCHROCK: Yes, I understand that, but that 24- doesn't make it the right physics.
| |
| 25 MR. TAKEUCHI: I understand that.
| |
| L T ANN RILEY & ASSOCIATES, LTD.
| |
| t -
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034
| |
| - - . _ . . - , -=.
| |
| | |
| 88 1 DR. SCHROCK: It makes it legally okay.
| |
| ~')-
| |
| . y[ 2 MR. TAKEUCHI: I understand, but we are trying to 3 doLthe'best by taking the worst case because we don't have 4 knowledge, instead of taking advantage of uncertainty.
| |
| 5 DR. ZUBER: So how do you determine the worst 6 case?
| |
| 7 MR. TAKEUCHI: By varying the break area and '
| |
| 8 finding the highest peak, the highest -- the case that 9 produced the. highest PCT. We varied this break area. '
| |
| 10 What we actually do is vary from .6 to 2.4 or so, 11 and this one is a PCT and we see PCT something like that, 12 then we think this one -- this is the limiting - .this is 13 one example.
| |
| 14 Solid curve is split. The coefficient is 0.6 so
| |
| <.(p) 15 we pick the coefficient equals 0.6 as the limiting case.
| |
| 16 DR. ZUBER: Is this -- do we have that in our 17 handout?
| |
| 18- MR. SINGH: I will get you a copy.
| |
| 19 DR. ZUBER: Okay.
| |
| 20' MR. NISSLEY: It's in a thick report.
| |
| 21 DR. ZUBER: But you know I have about two feet 22 of --
| |
| 23 MR. TAKEUCHI: This is the collapsed liquid level 24 inside the core -- inside the hot assembly and collapsed 25 liquid level inside the guide tube and open hope FMSC -- and l
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| I
| |
| ('')i (m, Court Reporters
| |
| , 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 89 1 collapsed liquid level inside the lower power region is 2 shown by this curve and you can see that it never completely
| |
| [v'}
| |
| 3 dried out during the transient inside the lower power 4 region.
| |
| 5 Now this is the cross-sectional view of the upper 6 plenum on the loop elevation and the UPI injection is made 7 at this elevation in the channel 55, and this is the 8 cross-sectional view of our upper plenum just below the loop 9 level down up to the top of our upper core plate and these 10 channels have four axial cells and we are going to 11 investigate the water injected in this channel 55, how it is 12 going to behave -- into the core.
| |
| 13 For this, let me refer, channel 35 below UPI 14 because it is sitting just below UPI injection point.
| |
| r~h
| |
| ( ) 15 In channel 34, next to UPI, and channel 33 lias
| |
| %J It opened to the UPI, and this center region has the inner 17 global.
| |
| 18 This is the void fraction inside the Below UPI and 19 the solid is a void fraction just above the upper core plate 20 and in high elevation void fractions are the rest of it.
| |
| 21 Let me just re-concentrate in the timeframe from 22 50 seconds and 80 seconds for the moment, and as soon as UPI 23 was injected, water is dumped onto -- on top of the upper 24 core plate and this is the void fraction next to UPI 25 channel. A similar thing is happening.
| |
| l l
| |
| l
| |
| (''}
| |
| (_j .
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . . .. _ __ _ . _ _ _ __.m_.-__ _ _ _ _ . _ . _ . _ _ _ . _ . . . ~ . _
| |
| l 90 1 If we go to opposite to the UPI channel, the water !
| |
| i 2 is delivered somewhat later,~however more water is
| |
| } j 3 accumulated in this channel, and this -- for the inner 1
| |
| 1 4 global channel.. A similar thing is happening. However, out l
| |
| 5 of those_four figures, we can figure out that water is 6 pooled on top of the upper core plate rather uniformly, ;
| |
| [ 7 however it is tilted towards the -- below upper core plate, 8 and this is the effect of temperature distribution of the 9 water accumulated, on top of the wattr the effect made on 10 the CCFL control, and I will come back to the temperature 11 later but for the moment let me show you the flow.
| |
| 12 The pool of water over the upper core plate is 13 going to penetrate into the core. This is integrated down --
| |
| 14 DR. WALLIS: Quick question? I'm sorry.
| |
| ,.( ) 15 MR. TAKEUCHI: Yes?
| |
| 16' DR. WALLIS: I am trying to figure this out.
| |
| 17 The region of your figures just before, where the 18 data seems -- the prediction is going all over the place --
| |
| 19 MR. TAKEUCHI: Do you have the void fraction?
| |
| 20 DR. WALLIS: Yes. It's the region where the peak 21 clad temperature is. The peak clad temperature occurs 22 around 70 seconds?
| |
| 23 MR. TAKEUCHI: Peak temperature --
| |
| 24 DR..WALLIS: It stays up there -- why this onset 25- of all these great fluctuations and things there?
| |
| i i
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| lk''N,
| |
| ,,) Court Reporters
| |
| ; 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
| |
| | |
| 91 1 MR. TAKEUCHI: Peak clad temperature or void 7m
| |
| '; i 2 ~ fraction?
| |
| GI 3 DR WALLIS: Well, the peak clad temperature 4 behaves sort of -- this sort of double hump thing and then 5 it comes up and then the second hump --
| |
| '6 MR. TAKEUCHI: Yes, sir.
| |
| 7 DR. WALLIS: -- followed by some oscillations --
| |
| 8 MR. TAKEUCHI: Yes.
| |
| 9 DR. WALLIS: -- is the period where the figures 1
| |
| 10 you just showed us are covered with oscillations all over 11 the place. What is going on?
| |
| 12 DR. ZUBER: The void fraction is jumping all over j i
| |
| 13 the place?
| |
| 14 MR. TAKEUCHI: Ah, the void fraction is jumping j C\
| |
| is ,) 15 all over. That is the --
| |
| 16 DR. WALLIS: What is going on? It can't jump all 17 over physically.
| |
| 18 MR. TAKEUCHI: After 80 second --
| |
| 19 DR. WALLIS: After 70 seconds.
| |
| 20 MR. TAKEUCHI: 75 or 80 seconds.
| |
| 21 DR. WALLIS: Something like that, yes.
| |
| 22 MR. T.'XEUCHI: This one is -- after accumulator is 23 emptied, the water, nitrogen pushed the lower plenum water 24 into the core and a lot of steam is generated because of 25 that, and then because the steam is coming up, water won't
| |
| [h ANN RILEY & ASSOCIATES, LTD.
| |
| \~ / Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| ._ _ __ _ _ _ _ . . _ . . _ _ _ _ . . . .m._ _
| |
| l 92 1 get'down through and more water is accumulated in the upper
| |
| ![E q -
| |
| '2 core plate. That is what you are seeing here.
| |
| 3 DR. ZUBER: You have got CCFL at the core plate.
| |
| .4 MR. TAKEUCHI': Yes.
| |
| 5 DR. ZUBER: How do you calculate that CCFL?
| |
| 6 MR..TAKEUCHI: Let me come up to that a little bit 7 .later. I 1
| |
| : 8. DR. WALLIS: While I'm trying to think, what is -
| |
| 9 GL?
| |
| 10 MR. TAKEUCHI: GL must --
| |
| 11 DR. WALLIS: It says void fraction -- GL.
| |
| 12 MR. TAKEUCHI: Global channel, okay.
| |
| 13 DR. WALLIS: So what does that mean? What is a 114 .GL?
| |
| J
| |
| ,g 15 MR. TAKEUCHI: Inner global. ;
| |
| : 16. DR. WALLIS: That is in the whole section? l 1
| |
| 11'7 MR. TAKEUCHI: The upper plenum is divided in*o I 18 the channels, four quadrant, and this one is inner global.
| |
| : 19. DR. WALLIS: That's inner global.
| |
| 20 MR. TAKEUCHI: That's inner GL -- global.
| |
| 21 DR. WALLIS: Out of GL -- is that thing, okay, so L22 it's the void fraction in that whole region you are 23 plotting?
| |
| 24 MR. TAKEUCHI: Void fraction --
| |
| 25 ~ DR. WALLIS: But this void fraction is jumping all L
| |
| _, ANN RILEY & ASSOCIATES, LTD.
| |
| i \s - Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l
| |
| (202) 842-0034 E
| |
| l l
| |
| | |
| _ - . _ _ . . _ , _ _ _ ~ - _ ~ _ , _ _ _ _ ~ . . ~ _ _ . _ . _ _ _ _ _ . _ . _ _ . _ _ _ _ _ . _ . _ . _ - . _ . _ .
| |
| l>
| |
| i.
| |
| 93 i'
| |
| l' over the place. In reality it cannot.
| |
| 2 MR..TAKEUCHI:
| |
| ( After accumulator is-emptied, after 3 this timeframe, yes, but-let me do this timeframe from 50 to '
| |
| ~4 80 seconds. ,
| |
| 5- DR. ZUBER: Let's do the one that Graham wants.
| |
| I 6 .have seen graphs like~that.from other codes and this is-
| |
| =
| |
| 7 really a plain numerical with condensation effects. You
| |
| , 8 cannot calculate this and it jumps all over the place. -
| |
| Is l 9 this the case you have here?
| |
| 10 MR. TAKEUCHI: That may be'the case because some ;
| |
| 11 amount of the steam'is coming out from the core after the 12 . lower plenum is filled up1and the water gets into the hot i
| |
| '13 core and some amount of the steam is generated.
| |
| 14 DR. WALLIS: But the void fraction is a one minus '
| |
| () 15 a void fraction is a measure of how much liquid is in that
| |
| .16 area --
| |
| 17 MR. TAKEUCHI: Yes.
| |
| l 18 DR. WALLIS: -- and the liquid cannot leave and 19- come back into that area rapidly.
| |
| : 20. DR. ZUBER: But the problem is this is -- it's a 21 numerical thing associated with condensation. Really all 22' our codes are not able to do that and theirs cannot either.
| |
| 23- DR. WALLIS: I am just wondering what are we
| |
| -24. supposed to conclude about this?
| |
| -25 DR. ZUBER: 'Well, it is white noise It shows l-E i l
| |
| L/''h ANN RILEY & ASSOCIATES, LTD.
| |
| , \s,/ Court Reporters i
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 !
| |
| l L i
| |
| | |
| s l 0 94 L 1. lack of ability to calculate this process.
| |
| ]') 2 DR. SCHROCK: I am~still disoriented about the L 3 different traces. You have got lines and AL and then some l
| |
| L 4 numbers and I don't know what they mean.
| |
| 5 MR. TAKEUCHI: Oh, AL -- you are saying about this
| |
| '6 one. That's the --
| |
| l 7 MS. DEDERER: That's just the variable -- it's 8 just from our plot package --
| |
| l 9 DR. SCHROCK: We try to teach engineering students 10 that when they make graphs.they put the parameters that 11 describe -- that are'necessary to make the graph understood 12 and we continuously have to try to interpret stuff which 13 doesn't have that. I don't know what to make of these 14 different traces. What do they mean?
| |
| () 15 MR. NISSLEY: These are channels and those numbers j 16 are a nodalization diagram. For example, up here channel i
| |
| : 17. 33 -- from the bottom up -- and then this is the actual l 18 injection elevation, so this is axially decreasing in the l 19 same region of the upper plenum.
| |
| 20 DR. SCHROCK: So you node axially as well as in l 21 this cross-section.
| |
| 22 MR. NISSLEY: Yes.
| |
| I 23 DR. WALLIS: We can't see any detail here.
| |
| t 24' DR. ZUBER: Does Research -- did you run more 25- calculations on UPTF? Because when I was there with NRC, i
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| O,' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| c 1
| |
| | |
| . _ ~ . . . _ - . _ _ . ~ . . . _ . . . _ . . _. .._. _ _ __. . . . . . . ._ _ _ _ . _ _ _ . _ - _ _
| |
| 95 1 I
| |
| 'l TRAC was not able to do this at'all. l
| |
| 'O. 2- .MR. LAUBEN: Norm Lauben, Research. I am not lQ ~
| |
| 3 aware ofiany UPTF calculations we have run since the early 4- days.
| |
| 5 JDR. ZUBER: Well, the level of capability still is l
| |
| J 6 .. as it was 10. years ago, which is essentially what we see 7 here.
| |
| 8 MR. LAUBEN: Well, I don't know. I don't know 9 what'it is.t i 10- MR. TAKEUCHI: -- the dose rate -- quite a lot. I 11 DR. ZUBER: But this may not be physics. It may 12 be just the way you calculate, the problems you have-in 13- modelling condensation. This was always a problem with 14 -codes.
| |
| () 15 MR. TAKEUCHI: Yes.
| |
| 16 DR. WALLIS: Maybe we need to go on, but you are 17 showing the figure for reason presumably?
| |
| 18 MR. TAKEUCHI: The reason I am trying to you is t
| |
| 19 'that if you look at the timeframe between 50 second and 80 20' second, I picked those timeframe because, first, the columnc 21 are clear and another thing is that this one, this behavior 22 is mixed up with the steam coming up from the bottom, 23 interfering with -- the water on top of the upper core 24 plate. Therefore, I picked this timeframe.and see what 25 ' happened when UPI water was injected how it's accumulated on l
| |
| t,
| |
| ! ("] ANN RILEY & ASSOCIATES, LTD.
| |
| \)
| |
| % Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
| |
| | |
| l 96 1 top of upper core plate.
| |
| [~'') 2 DR. SCHROCK: And the node represented by the
| |
| \s 3- solid curve is very different from all the others. Am I to 4 interpret that there is some important aspect of that?
| |
| j 5 MR. TAKEUCHI: Yes. The water just shoot down to l l
| |
| 6 the top of the upper core plate with the accumulation of -- l 7 this one means that the liquid is not accumulated in the 8 cells between the injection point to the top of the upper 9 core plate. It shooted down.
| |
| 10 As soon as it comes in, it goes through, therefore 11 there is no accumulation of liquid.
| |
| 12 DR. SCHROCK: So you are saying that this 13 calculation tells you all the water goes down through the 14 holes in one node? l
| |
| /~T 15 MR. TAKEUCHI:
| |
| ! j Yes.
| |
| 16 DR. SCHROCK: Well, that helps me to understand 17 what you are showing.
| |
| 18 MR. TAKEUCHI: And if you look at this type of 19 accumulation of the liquid, and this one opens to UPI, this 20 one is an inner global and this accumulation is below UPI 21 channel and this accumulation is the next to UPI channel.
| |
| 22 Therefore it is almost a uniform -- the water is almost 23 uniformly accumulated on top of that upper core plate 24 everywhere.
| |
| 25 That is what I am trying to show you.
| |
| l l
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| (~]
| |
| (_- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| - - .. - .. . ~.. . . - . -- . - -.
| |
| 97 1 DR. ZUBER: Then what happens?
| |
| l.["') 2 MR. TAKEUCHI: The'n what happens -- I am going to
| |
| \~)
| |
| 3 show next.
| |
| 4 DR. ZUBER: You have here graphs --
| |
| 5 MR. TAKEUCHI: Okay. Then what happens is that 6 . water, accumulated water goes through the open hole channels 7 and gets into the low power channel in the core, 8 selectively.
| |
| 9 The other place -- excuse me?
| |
| 10 DR. WALLIS: In the upper plenum?
| |
| 11 MR. TAKEUCHI: Yes.
| |
| 12 DR. WALLIS: This region where the curves go 13 everywhere is showing that there is water all over the upper 14 plenum?
| |
| '.() 15 ; R. TAKEUCHI:
| |
| M You are talking about the timeframe 16- after 80 seconds?
| |
| 17 DR. WALLIS: There is water everywhere.
| |
| 18 MR. TAKEUCHI: Yes. Because a large amount of 19 steam is coming up after this time. Before this time, core 20 is almost empty, so there is not much steam coming out, not 21 a-substantial amount.
| |
| 22 Just enough to support that over that plate. Now 23 after this time, a substantial amount of the water is coming 24 from the lower _ plenum to the core, hot core, and it
| |
| ~25 generates lots of steam. So that the water accumulated just l-I'\ ANN RILEY & ASSOCIATES, LTD.
| |
| k-m Court Reporters L 1025 Connecticut Avenue, NW, Suite 1014 l
| |
| Washington, D.C. 20036 j- (202) 842-0034
| |
| | |
| 98 l
| |
| 1 1 above the core is going to spread out all over top -- higher 2 elevations. That's what it is showing here.
| |
| O(~N Before water entered into the hot core from the 4 lower plenum, steam was coming out, but not substantial 5 enough to make things so radical. And we can see that water 6 has accumulated.
| |
| 7 DR. ZUBER: But still enough to have a CCFL at the 8 break.
| |
| 9 MR. TAKEUCHI: That's right.
| |
| 10 DR. ZUBER: So what you said, you have some steam 11 coming, the water is now uniformly on the plate --
| |
| 12 MR. TAKEUCHI: Yes.
| |
| 13 DR. ZUBER: And then it comes from the bottom and 14 you generate more steam --
| |
| fh g
| |
| v) 15 MR. TAKEUCHI: More steam, so the water is 16 spreading at higher elevation than -- juet above upper core 17 plate. That's what you're seeing, void fraction is coming 18 down from just a void mixture of the liquid phase in there.
| |
| 19 So if we look at the time frame between 50 and 80 20 seconds, we can see how it's behaved without any 21 interference of substantial steam coming out frcm the core.
| |
| 22 DR. SCHROCK: Your conclusion that the liquid is 23 all uniform at this time, 60 seconds, is based on a 24 comparison of four traces, and it's unclear to me what the 25 nodes are that those four traces represent. Those are the
| |
| /9 ANN RILEY & ASSOCIATES, LTD.
| |
| (m / Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 99 1 four quadrants, I guess?
| |
| /''A 2 MR. TAKEUCHI: Yes.
| |
| !V 3 The first figure is in this channel, below UPI.
| |
| 4 The UPI water is injected above this channel. And this one, 5 second figure is for the next two UPI channels, and third 6 figure'is opposite to UPI channel, and the fourth figure of 7 void fraction is this inner global.
| |
| 8 DR. WALLIS: Do you have some gravity-driven flows 9 that exchange water between these regions?
| |
| 10 MR. TAKEUCHI: Yes.
| |
| 11 DR. WALLIS: Because you seem to have some times 12 when the void fractions are very different in these 13 different regions in some instances. If you look at the 14 opposite UPI, there's a big dip that goes down to .15 or
| |
| . (m) 15 something.
| |
| 16 MR. TAKEUCHI: It takes -- there is a time lag 17 appearing in this calculation.
| |
| 18 DR. KRESS: Does the injection flow rate stay 19 about constant?
| |
| 20 MR. TAKEUCHI: It is almost constant all the time.
| |
| 21 DR. SCHROCK: But if you take the specific time of 22 60 seconds, and you compare the void fraction in the ,
| |
| 23 different quadrants, the low number looks like it's .1, 4 24 slightly more than .1, and the high number looks like it's 25 about .6. And yet you conclude they're all the same.
| |
| /O ANN RILEY & ASSOCIATES, LTD.
| |
| (%,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| : Washington, D.C. 20036 l (202) 842-0034 t
| |
| | |
| - _- . ~ - . ~ ~ . . - - . . . - . . . - . . - . . . . - . . - - , - - . . . -. . . . . . . ~ ,
| |
| 100 1 MR. TAKEUCHI: Overall, yes. Overall behavior.
| |
| (2 If you draw a line'like this --
| |
| l 3~ DR..SCHROCK: Well, if you' average each one of
| |
| [ 4 'those things, you diminish the spread, but then the spread
| |
| '5- looks like comething'between .4 and .6.
| |
| 6 MR. TAKEUCHI: If this one is static, it can
| |
| : 7. become'even' eventually. But it takes time. t i
| |
| 8 DR. ZUBER: What is the steam velocity -- steam ^
| |
| 9 velocity at.the plate during the time?
| |
| 10- MR'.:TAKEUCHI: Let me see. I
| |
| ! 11. DR. ZUBER: Let me say, because you would make a 12 more believable case if you really showed that for that 13 1 velocity there is a CCFL at the plate and therefore you
| |
| '14- ' maintain this uniform water on top.
| |
| / I 15- MR. TAKEUCHI: I think this curve might say that.
| |
| \ ,) ,
| |
| \
| |
| ~16 DR. ZUBER: What you could have shown, I mean, you 17 have a correlation for CCFL, then you can say this time the 1
| |
| : 18. velocity of the steam is so much, and therefore we are !
| |
| 19 holding the water on the plate and it's uniform.
| |
| '20 MR. NISSLEY: Kenji, could you show the TSAT I 21 distribution and then --
| |
| 22' MR. TAKEUCHI: This is the curve, I checked how 23 .the behavior at upper core plate is. And I am going to come 24 back'to this one after a while.
| |
| ; 25- DR. WALLIS: I have a question. I'm afraid there 1
| |
| l l'
| |
| \
| |
| l
| |
| ( -~ ANN RILEY & ASSOCIATES, LTD.
| |
| : Court' Reporters .
| |
| L 1025 Connecticut' Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034 3
| |
| i
| |
| | |
| 101 1 are too many questions here, but we heard earlier on about
| |
| (~]
| |
| \J 2 these 13-inch-high, the water didn't get in until it got up 3 to 13 inches. How is that height related to these void 4 fractions that are shown here? Do you assume a water level 5 or something in there?
| |
| 6 MS. DEDERER: There's a water level building up, 7 and that's what is showing in that void fraction.
| |
| 8 DR. WALLIS: Void fraction is not the same thing 9 as water level.
| |
| 10 MR. TAKEUCHI: I think you are talking about 11 freestanding mixture. In the case of a freestanding mixture 12 the metal is like this, and this one is -- flow rates -- I 13 was going to compare, the next two pages is the flow rates 14 through the freestanding mixture, and the other one is open
| |
| ,o 15 hole.
| |
| 16 DR. WALLIS: How do you go from void fraction to 17 level?
| |
| 18 MR. TAKEUCHI: 1 minus alpha times --
| |
| 19 DR. WALLIS: Level? But some of it's droplets.
| |
| 20 MR. TAKEUCHI: Some of them are droplets, but here 21 you can think there is no droplets.
| |
| 22 DR. WALLIS: But if it's level, then the upper 23 node should not have any liquid in them, and you've got 24 liquid -- you've got void fraction dancing around in all the l 25 nodes. So --
| |
| 1 i
| |
| /''i ANN RILEY & ASSOCIATES, LTD.
| |
| (_,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| L .102 , J l'
| |
| 1- MS. DEDERER: These are the global channels, not 2 the: freestanding mixture channels. Maybe that's where the 3' confusion is. He's.showingLthe water that surrounding those l l 4 structures, okay?
| |
| [
| |
| -5 DR. WALLIS: And we heard it has to get to 13
| |
| ; 6 inches'before it overflows.
| |
| 7 but. NISSLEY: And it doesn't.
| |
| l l 8' DR. WALLIS: Never gets to 13 inches?
| |
| i
| |
| : 9. MR. NISSLEY: It only drains down the open' holes, i l 10 and-he has a figure here that shows --
| |
| l 11 DR. WALLIS: So where's the void fraction that R l
| |
| 12 you're plotting there? .What.does it mean?
| |
| i 13 MR. TAKEUCHI: This void fraction is on --
| |
| 14 DR. WALLIS: Is there?
| |
| 15 MR. TAKEUCHI:
| |
| ( There.
| |
| 16' DR. WALLIS: Is the average?
| |
| 17 MR. TAKEUCHI: Average.
| |
| 18 DR. WALLIS: And so you translate that into level? l 19- MR. TAKEUCHI: It can be -- 1 minue alpha and dx.
| |
| l l :20' This one is a delta x. That's the level. And void fraction L-f 21 is plotted here.
| |
| i
| |
| .22. DR. WALLIS: And the other nodes are higher up?
| |
| 23: MR. TAKEUCHI: The other nodes are higher. There
| |
| ~24- are four cells in this channel. And the void fraction in 25 this elevation is very high because as soon as liquid comes i
| |
| e j .. ANN RILEY & ASSOCIATES, LTD.
| |
| l( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| : l. Washington, D.C. 20036
| |
| : l. (202) 842-0034 o
| |
| k'
| |
| | |
| . . . . . . .. . . . _ . . ,m__ . . _ . . - _ . . __m._-_. _ . . . _ . - _ _ _ . _ . _ _ . _ ....._._
| |
| 103 '
| |
| 1 in, same amount goes out. ~So'there is'no accumulation of i
| |
| ~
| |
| 2< .the liquid here. :
| |
| 3 To start-with it's empty. And the leaks come down 4 with the same speed as comes-out.
| |
| 5 DR. WALLIS: Where does it come from? !
| |
| :6 MR. TAKEUCHI: UPI.
| |
| 7 DR. WALLIS: Up there?-
| |
| l
| |
| '8 MR. TAKEUCHI: Yes, up there.
| |
| !' 9 DR. WALLIS: But it's only being put into the one i
| |
| 10 next'to the UPI. ,
| |
| l 11 How does it get ~ to'the channels which are not l
| |
| 12' where'it's injected?' You've got four quadrants, right?
| |
| 13 MR. TAKEUCHI: Four quadrants.
| |
| 14 DR. WALLIS: How does it get from one quadrant to
| |
| [V l 15 the other? How does it get up there? How does it get up 16 high?
| |
| 17 MR. TAKEUCHIi How does it go up:high? Are you 18 talking about this one?
| |
| 19' DR. WALLIS: Yes, how does it get --
| |
| j 20 MR. TAKEUCHI: This one is a substantial amount of l
| |
| L 21. steam coming out from the core.
| |
| l l
| |
| 22- DR. WALLIS: It entrains droplets?
| |
| .23 MR. TAKEUCHI: Yes. This is the below UPI I
| |
| 24 channel. The UPI water'was injected above this channel.
| |
| 25 And this channel has four cells. Here is the UPI water l
| |
| ! /~ ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters L 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l ,
| |
| | |
| _ __ _ _ . . . m.-
| |
| l 104
| |
| ; 1 injected.
| |
| i ('N 2 DR. WALLIS: Up there.
| |
| l %
| |
| (
| |
| 3 MR. TAKEUCHI: Yes. And then water is accumulated 4 on top of the upper core plate. That's --
| |
| 5 DR. WALLIS: It's injected up there, but it only 6 goes in on the region next to its injection.
| |
| 7 MR. TAKEUCHI: Only goes in? No, it goes -- it 8 comes in, but it's spread out everywhere.
| |
| 9 DR. WALLIS: I thought we just heard from Mitch 10 that it all came down in the region near where it's 11 injected.
| |
| 12 MR. TAKEUCHI: No , it comes down here, but after 13 it hits the upper core plate, it's spread out all over, so 14 the water is accumulated, this outer channel, inner global
| |
| -A 15
| |
| ( ; channel, as well as outer global channels. It comes like
| |
| , %/
| |
| 16 this.
| |
| 17 DR. WALLIS: But then it's thrown up by the steam.
| |
| 18 MR. TAKEUCHI: And after 80 seconds --
| |
| 19 DR. WALLIS: It's thrown up by the steam.
| |
| 20 MR. TAKEUCHI: The pressure by the steam is coming 21 up.
| |
| l 22 DR. WALLIS: Okay.
| |
| 23 MR. TAKEUCHI: And it threw it out.
| |
| l 24 DR. WALLIS: Because I thought you were telling us l 25 that the upper plenum injection put it in all the other r~% ANN RILEY & ASSOCIATES, LTD.
| |
| i!''') Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i' Washington, D.C. 20036 L (202) 842-0034 i
| |
| | |
| .___.m. .. ._ _ . _ - _ _ _ . _ _ _ _ _ _ _ . _ _ _ . _ _ . . _ . _ . _ . . ._ _
| |
| l 105 1 channels. Juxi now you're saying it's -- !
| |
| 2 MR. TAKEUCHI: Upper plenum water is injected 3 starting -- 50 seconds, and after 80 seconds, steam is !
| |
| 4 relatively' quiet. So we can see how the water is delivered 5 from UPI water injected at high elevation in the UPI. And 6- we can come up with this kind of a picture.
| |
| 7 DR. ZUBER: Do_you have data to show how that 8 water-spreads from coming down and how it spreads across the 9 global section to the other three quadrants? '
| |
| 10 MR. TAKEUCHI: There is no direct data of that 11- kind.
| |
| 12 DR. ZUBER: -So I mean this conjecture.
| |
| 13 MR. TAKEUCHI: This is computed results. l 14 DR. ZUBER: No , no , no. Do you have --
| |
| () 15 16 MR. TAKEUCHI:
| |
| DR. ZUBER:
| |
| However, we are standing on --
| |
| Hold on, hold on. Let me ask you. Do 17 you have calculations to show how the water comes down and 18 then how it spreads? Do you have numbers to show, or what 19 you're just saying is a conjecture what you think it is?
| |
| 20 MR. TAKEUCHI: I did not look into the flow rate 21 here, if you are questioning that.
| |
| 22 DR. ZUBER: Well, let me say my problem. I don't 23 believe these calculations. I have seen them before. I 24 have seen these calculations on the right-hand side before.
| |
| 25 I think this is more or less incapability of the code to do
| |
| . (~N ANN RILEY & ASSOCIATES, LTD.
| |
| T Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 106 1 calculations. You try to explain based on the physical i
| |
| /'')
| |
| \' ' '
| |
| /
| |
| 2 argument that you have water spreading on the core 3 plate under injection and therefore it's uniform. My 4 question then is that's a physical explanation. Do you have 5 data calculated to show that spreading?
| |
| 6 MR. TAKEUCHI: There is no data for --
| |
| 7 DR. ZUBER: Do you have calculations, numbers, to 8 show --
| |
| 9 MR. TAKEUCHI: This is a calculation.
| |
| 10 DR. ZUBER: But this doesn't show what -- you are 11 trying to explain something which your graphs do not show.
| |
| 12 And what it shows, I don't believe.
| |
| 13 MR. NISSLEY: You are calculating transverse 14 flows --
| |
| /,-) 15 DR. ZUBER: This is what he is explaining. We are
| |
| %.)
| |
| 16 trying to understand what is the difference here and what is 17 here and he says -- you have CCFL. I ask him, do you have 18 calculations to show.
| |
| 19 DR. WALLIS: I think we probably should move on, 20 but I'think what you are going to tell us is all this 21 doesn't matter, so we may need to move on, but if the 22 purpose is to really predict what is happening in the upper 23 plenum, I think you'll have great difficulty, but if you are 24 going to show that it is all stirred up and it doesn't 25 really matter because it gets down somewhere -- then we can
| |
| (~'g ANN RILEY & ASSOCIATES, LTD.
| |
| ('~,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 2 84 - b3 I
| |
| | |
| - . . - . - . . _ _ . . - _ . . . . - . - . - . _ . . - ~ - - _ . . - - . . . - . ~ . ~ . . _ -
| |
| l:
| |
| 107 1- move:on. I don't think this is really important to the
| |
| /~N. 2 conclusion.
| |
| b '. 3: DR. ZUBER: The question is how does thr water get
| |
| =4 down? It says'here fairly immediately -- all over the
| |
| .5 place.-- uno this is something we had problems with-6 calculations 10 years ago.
| |
| ! 7 If it is not important, tell us why~it is not
| |
| [.!
| |
| '8 -important.
| |
| 1 9 MR, TAKEUCHI: It is important because it is the L.
| |
| 10 spreading'out all over the -- on top of the upper core
| |
| .111 . plate.
| |
| '12 DR. ZUBER: This you didn't show, i 1 l 13 MR. TAKEUCHI: This I showed you, l 14 DR. WALLIS: There is some water all over.
| |
| ,( 15' Presumably if there is some water then it's available to 16 come down.
| |
| 17 MR. TAKEUCHI: That one is UPI. water. That one 18 there is no question about it. It's empty before UPI water 19 was injected.
| |
| 20 MR. NISSLEY: Where it is relatively well-behaved i 21 in void fraction you can see that it has spread the whole 22 way around the upper core plate at the lowest node. Where 23 the numerical noise if you will comes into play is after the
| |
| .24 hot rod has turned around and the average rods are quenching 25 and the steam generation goes up dramatically.
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| ..g
| |
| (" Court Reporters
| |
| : l. 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
| |
| .(202) 842-0034 L
| |
| i
| |
| ( ,;, - , - - . . _ . . ,. . - . . . -
| |
| | |
| [ .'
| |
| L 108 1- DR. ZUBER: Okay. You could have made the i
| |
| e 2 presentation much more effective had you shown' calculated s
| |
| 3 value this water is (n1 top uniformly and I didn't see it,
| |
| . . - i l 4 and this is the question.
| |
| 'S MR. NISSLEY: The horizontal line, blue line, that 6- he has on each of these --
| |
| '7 MR. TAKEUCHI: Those are the indications of the
| |
| .8' water is accumulated rather uniformly all over the top of --
| |
| l 9 DR. WALLIS: I think if you just drop the word _
| |
| l-i 11 0 " uniformly" and simply said that there is a substantial
| |
| . 11 -amountzof' water everywhere, therefore it is available to 12 come down.-
| |
| 13 MR. TAKEUCHI: Right.
| |
| 14 DR. SCHROCK: But the calculation of that is
| |
| , I'h :15 dependent upon some assumptions or some modelling that you D
| |
| l 16 have done for cross-flow between the nodes.
| |
| 17 MR. TAKEUCHI: Yes.
| |
| , '18 DR. SCHROCK: I guess whether we believe.it or not 119_ depends on whether we believe in that cross-flow analysis.
| |
| L 20- MR. TAKEUCHI: I am believing because this code L ~ 21' has been verified with. substantial amount of assessment.
| |
| 1 o 22 This is not the first time we are doing it.
| |
| 23 DR. SCHROCK: You assess against pretty global l~
| |
| 24 measures and this cross-flow is a very difficult aspect of
| |
| '25' this model. I mean you haveLdone it in coarse noding.
| |
| 1-
| |
| ;! i
| |
| , ANN'RILEY & ASSOCIATES, LTD.
| |
| \ Court Reporters
| |
| . 1025_ Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| - . - . .. ~ - . _ . . _ . _ . . . . ._
| |
| | |
| -.. - - - - - . - . . _.. .. .. - - . - . - ~ - . - ~ . . - - - -. .. - ...~. ..
| |
| 109 l 1 DR. WALLIS: 'We should probably move on.
| |
| 2 DR. SCHROCK: Yes, I think so.
| |
| l :3 My point is simply you can't make a definite ;
| |
| 4 conclusion-based on calculations that are themselves i
| |
| 'S questionable.
| |
| l 6 MR. TAKEUCHI: Yes. I can understand. Therefore, 7 I thought I'd try to make sure that we went through'a .I
| |
| '8 substantial amount of verifications and because of that
| |
| ; 9 situation I tried to use computer results and to' find what 10 is happening for the water injected at upper plenum.
| |
| 11 DR. FONTANA: If I understand the importance of 12 this, the water that's over the hot channel doesn't really 13 get down there, 14 'MR. TAKEUCHI: It does not get in --
| |
| b
| |
| ;\4j '
| |
| 15 DR. FONTANA: The water that is in a lower power I
| |
| \
| |
| 16 channel does get down and affects the cooling?
| |
| 17 MR. TAKEUCHI: Yes. I 18 DR. FONTANA: So a model in which more water is j 19 over.the lower power channels will end up with better core 20 cooling, is that correct?
| |
| 21 MR. TAKEUCHI: That's right.
| |
| 22 5 DR. FONTANA: Do you have a cross-flow model for 23 the level of the water above the --
| |
| 24 MR. TAKEUCHI: Eventually I come up with this kind 25 of picture you have been describing. The water injected in l-l l'
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| ! Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 110 11 the upper plenum is going -- accumulated over the top of the 2 upper core plate, but the water penetrate into the lower 3 power core region, preferentially through -- and then 4 sometimes it goes down to the lower plenum and joins with 5 the other ECC water and it comes up into the upper region 6 hot channel and some cross over but at the very low 7 elevation, one-third from the bottom, then there is a 8 cross-flow, but there is no cross-flow at the assembly in 9 the top two-thirds elevation.
| |
| 10 DR. FONTANA: Okay. UPI water comes in on the top 11 of your picture there.
| |
| 12 MR. TAKEUCHI: Yes.
| |
| 13 DR. FONTANA: Now if the level of the water is not
| |
| .14 uniform, does it make any difference?
| |
| () 15 MR. TAKEUCHI: If the water is not uniform --
| |
| 16 well, because of the upper core plate and CCFL it's going to 17 be spread out all over there.
| |
| 18 DR. FONTANA: Okay. Does that come from 19 experimental data or does that come from a cross-flow model 20 for the flow of the water across the top?
| |
| 21 MR. TAKEPC9I: That is the cross-flow model, but 22 we have substantial verification for those.
| |
| 23 DR. ZUBER: What verification do you have, because 24 you mentioned it? I mean do you have real experimental data 25 where you showed that you can calculate this?
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l l
| |
| l
| |
| '111: :
| |
| MR. TAKEUCHI: I-have done this kind of 1'
| |
| (('}x_-
| |
| :2 verification. There is a. hot leg and this one is modelled 3- by --'usually hot leg is modelled by ID-components, but use !
| |
| 1 4 the vessel components. Try to -- made to this kind of-a_ l S. channel and inject water from this side and inject the steam !
| |
| 6 from this side and see what happens.
| |
| I 7' DR. ZUBER: Always the proof.is in' pudding, and 8 the pudding is experimental data. Here we have to believe
| |
| -9 calculations. Then.you said you have verified.the' code.
| |
| 10 Do you have experimental data to show that the-1 11 ~ code can really calculate this? _Then you have proven your.
| |
| 12 case. Then you have tested against data, not against 3 I
| |
| 1 13 something which is artificial. You can show how well it l 14- calculates this.
| |
| ()r 115f 16 MR. TAKEUCHI:
| |
| related problem.
| |
| This one is a mainly cross-flow 17 DR. ZUBER: You have experimental data, yes or no?
| |
| 18 MR. TAKEUCHI: Experimental data is there for real
| |
| : 19. flow correlations. Real flow is when the water goes through 20 the : core and it gets into the crypt. How much is the water 21 level here? That's the real flow correlations -- how much 22 water is going to be accumulated in here.
| |
| 23 DR. WALLIS: Does this model of the core here 24 different for upper plenum injection than it was for three j
| |
| : l. 25~ and four loop? Are we looking at something new or something l-
| |
| , ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| __ . .. _ _ - . . _ . . ~ . . _ _ . _ . . . . __ _ . . ._ __ . - _ _
| |
| r ?
| |
| 112 l: 1 that-has been' approved already?
| |
| i~ ;
| |
| /'' ~ 2 MR. TAKEUCHI: You mean his core?
| |
| \
| |
| 3' DR. WALLIS: Yes. '
| |
| 4 14R . TAKEUCHI: .This one, the sketch of UPI -- the 5 core-channels of the UPI plant.
| |
| 6 DR. WALLIS: Are you looking for us to approve or 7 recommend approval or whatever we do of the model here, 8- =because it is something new or is this something that is the j 9- same as it was before?
| |
| l 10' MS. DEDERER: It is the same.
| |
| 11 MR. NISSLEY: The core modelling is identical.
| |
| '12 DR. WALLIS: So we can ask questions but we can't 11 3 do anything about it, is that right?
| |
| 14 [ Laughter.]
| |
| / 15 DR. WALLIS: So we should concentrate on the kJ 16 things that you are asking us to actually evaluate, which is 17 one the specialties of your -- Figure 2.4, which is next to I
| |
| 18 this one. Are you going to show that?
| |
| 19 MR. TAKEUCHI: 2-4.
| |
| =20 DR. WALLIS: Next to this one. I guess again I'm 21 so curious I have to ask about something I can't change.
| |
| 22 DR. ZUBER: But what is important here is --
| |
| 23 DR. WALLIS: In 2-4 you have counter-current 24 horizontal flow carried out of the LP channel. That is what 25 I can understand. Mixing in the core allows counter-current l
| |
| l l
| |
| l' ANN RILEY & ASSOCIATES, LTD.
| |
| l g,,,, ' Court Reporters ll 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 113 1 . flow horizontally.
| |
| 2- MR. TAKEUCHI: Are you talking about this one?
| |
| 3= DR. WALLIS: -Yes.
| |
| :4 MR. TAKEUCHI: Oh. This is --
| |
| +-
| |
| 5 DR. WALLIS: Counter-current flow., horizontally.
| |
| 6 MR. TAKEUCHI: This I don't mean to say that the 7 counter-current flow is important, but the point --
| |
| 8 DR. WALLIS: Sideways flow.
| |
| 9 MR. TAKEUCHI: Sideways flow --
| |
| 10 DR. WALLIS: -- is counter-current.
| |
| 11 MR. TAKEUCHI: The matter I am trying to show here 12 was that a substantial amount of the liquid coming into the 13 low power channel and then how it is going to be distributed 14 and some gets into the lower plenum and about 6 percent of
| |
| ( 15 the drained water gets into the other channels right next to 16 the low power region.
| |
| 17 DR. WALLIS: My question was -- I was just curious 18 that you have got, if you look at, say, the top of the 19 figure there that you were just pointing at, you have 450 20 black going to the right, and 117 white going to the left.
| |
| 21 MR. TAKEUCHI: Yes.
| |
| 22 DR. WALLIS: That is counter-current horizontal 23 flow.
| |
| 24 MR. TAKEUCHI: But I don't think this one is 25 anything.
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| O' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 114 1 DR. WALLIS: What makes that happen?
| |
| (~N g 2 MR. TAKEUCHI: Why does this happen?
| |
| U' 3 DR. WALLIS: Why does it happen?
| |
| 4 MR. TAKEUCHI: Because this channel is cold and 5 outside --
| |
| 6 DR. WALLIS: Is it condensation that has come, 7 causing --
| |
| 8 MR. TAKEUCHI: No. This channel is cold, compared 9 to the other channel, therefore steam is generated in the 10 other channel and it gets into the colder low power channel.
| |
| 11 DR. SCHROCK: As I read the caption on the thing, 12 it is time-integrated mass flow so one explanation could be 13 that part of the time liquid is going one way and part of 14 the time steam is going the other way.
| |
| [ '\ 15 DR. KRESS: I think that is more likely.
| |
| O 16 MR. NISSLEY: Also, each one is a summation of 17 four axial, nodes, so it might have liquid going in that case 18 to the right in the bottom two and vapor going to the 19 left --
| |
| 20 DR. KRESS: I don't think your code is capable of 21 calculating counter-current flows in any one channel there, 22 so I think that explanation is probably -- and Virgil's is 23 probably right.
| |
| 24 MR. TAKEUCHI: And I don't think this one is a 25 limiting counter-current flow.
| |
| /~ ANN RILEY & ASSOCIATES, LTD.
| |
| ('~'N) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i
| |
| Washington, D.C. 20036 l (202) 842-0034 l
| |
| i
| |
| | |
| 115 l l
| |
| i 1 DR. KRESS: I don't think it's an issue. I don't l
| |
| (~~ j 2 think they are going on at the same time.
| |
| \J 3 DR. SCHROCK: It'n just if you give 4 time-integrated flows I guess what is the basis of your 5 selection of the interval over which you are going average?
| |
| 6 I don't know how to interpret numbers that come about in j
| |
| 7 that way.
| |
| 8 DR. WALLIS: Why I am asking the question, I mean 9 all we can do is look at results you present to us and say 10 does this look reasonable.
| |
| l 11 If something looks peculiar, that's what we ask a 12 question about. That is why this came up, and I was saying 13 this looks peculiar, how come?
| |
| 14 We don't really know. Someone is hypothesizing
| |
| () 15 it's because of time integration or something.
| |
| 16 MR. NISSLEY: I think it's a combination of time 17 integration and summation of --
| |
| 18 DR. WALLIS: We don't know. We don't really know 19 the answer.
| |
| 20 MR. NISSLEY: The selection of the time periods, 21 the first one is when you are building up a level in the 22 outer channel sufficient to start driving significant 23 cross-flow into the inside and the second thing then is when 24 you are getting lots of cross-flow.
| |
| 25 DR. SCHROCK: Well, these are sequential, 50 to 75 f}
| |
| (_/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l (202) 842-0034 i
| |
| I t
| |
| | |
| - . . . _ . - . . ~._ . - . . - - . _ . -. _. . . -- . .
| |
| l i
| |
| 116 1 seconds, 75 to 100. seconds, so if it is the time integration
| |
| ' g('') 2 and you start out with liquid flowing early in time, but 3 then-it shifts over to steam flow to give the picture you 4 have on 50 to 75 seconds, then it is cast in doubt when you 5 look at the next time interval and it has to go through the 6- gyration again of introducing additional liquid flow and 7 additional steam flow.
| |
| 8 DR. WALLIS: We shoul'd probably move on.
| |
| 9 All we can do is look at these things and does 10 something look peculiar and ask a question.
| |
| 11 MR.-TAKEUCHI: And the point I am trying to come 12 up with is that we can have this kind'of failure and there 13 is no liquid flow out the top of hot assembly. And just to 1
| |
| 14 . prove it, we have this curve for the liquid flow, total
| |
| () 115 16 liquid flow at the top of hot assembly.
| |
| more detail, this curve is expanded here.
| |
| And just to see l
| |
| 17 DR. KRESS: Could you go back to the. previous l 18 figure, please? What I am interested in is in that hot
| |
| ; 19 channel.
| |
| l 20 MR. TAKEUCHI: Yes.
| |
| l 21 DR. KRESS: Hot assembly.
| |
| l 22 MR. TAKEUCHI: Hot assembly, i
| |
| i 23 DR. KRESS: How is the enthalpy or the temperature 24 at the lower node into that channel calculated, based on the 25 downflows in the peripheral channels?
| |
| , ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| ; Washington, D.C 20036 i (202) 842-0034
| |
| | |
| 117 1 MR. TAKEUCHI: How enthalpy is calculated.
| |
| /''g 2 DR. KRESS: Yes. l l
| |
| 3 MR. TAKEUCHI: Enthalpy in the hot assembly is 4 calculated by the enthalpy carried into the cell from the 5 flow down bottom and the top anywhere, plus heat deposited' !
| |
| 6 from the hot load in the hot assemblies.
| |
| 7 DR. KRESS: Those two down arrows in the bottom of 8 -the core, --
| |
| 9 MR. TAKEUCHI: Yes.
| |
| i 10 DR. KRESS: -- they go into a plenum region? ;
| |
| 1 11 MR. TAKEUCHI: Yes. '
| |
| : 12. DR. KRESS: And they come back up out of that i l
| |
| 13 plenum region into the bottom?
| |
| 14 MR. TAKEUCHI: Come up. Yes.
| |
| . ,O 15 DR. KRESS: What goes on down in that plenum
| |
| ! . . y/
| |
| 16 region? Is that calculated as one, having one enthalpy?
| |
| 17 MR. TAKEUCHI: Yeah. I plotted only those liquid 18' flow --
| |
| 19 DR. KRESS: This is just the behavior --
| |
| j 20 MR. TAKEUCHI: -- UPI water. I did not include an 21 arrow due to the ECC water injected in the cold leg. That 22 one is included.
| |
| l 23 DR. KRESS: Is there still cold leg injection
| |
| !- 24 going on at.this time?
| |
| L 25 MR. TAKEUCHI: Yes. Yes. Accumulator water is i
| |
| , ("% ANN RILEY & ASSOCIATES, LTD.
| |
| 4
| |
| (_) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
| |
| I I
| |
| | |
| , _ _ . . . . _ . . ._ _ __. ~. . . . -. _ _ _ . _.
| |
| 118 1 Lcoming in.
| |
| /''% 2 DR. KRESS: Is still coming in?
| |
| ld 3 MR. TAKEUCHI: Yes. And that one I did not f 4 include in'here. This is the pattern of the water injected
| |
| ! 5 at the top of upper plenum.
| |
| L 6 DR. SCHROCK: So your picture is that you have j 7 cross-flow of liquid getting into the hot channel that --
| |
| -8 MR. TAKEUCHI: Only in the lower one part.
| |
| l 9 DR. SCHROCK: Let me finish my statement, would 10 you? Cross-flow into the hot channel that produces steaming 11 prior to the time that it begins to reflood?
| |
| : 12. MR. TAKEUCHI: Yes. Well, the interesting thing I
| |
| : 13. is that although the UPI water was injected at the top,-in !
| |
| j 14 the higher elevation in the upper plenum, and the water is
| |
| () 15 16 evenly distributed above the upper core plate, still, water is not draining to hot assembly.
| |
| 17 DR. SCHROCK: Well, you made a point of the l 18 importance of this high steam flow coming up.
| |
| 19 MR. TAKEUCHI: Yes.
| |
| I 20 DR. SCHROCK: Until you begin reflood, until ycu 21 get the lower plenum filled up and you begin to reflood, you 22' wouldn't have steam coming up.
| |
| j 23 MR. TAKEUCHI: But still some, I think some coming 24 up. Enough to support.
| |
| i 25 DR. SCHROCK: Coming up from what?
| |
| i
| |
| /* ANN RILEY & ASSOCIATES, LTD.
| |
| i ( ,}f ' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 i (202). 842-0034 i
| |
| L L. _
| |
| | |
| 1 119 1 MR. TAKEUCHI: From where? Maybe -- but that is
| |
| ('')
| |
| %/
| |
| 2 substantial amount of steam --
| |
| 3 MR. NISSLEY: Cross-flow into the average 4 assemblies.
| |
| 5 MR. TAKEUCHI: The thing I tried to come up with l
| |
| 6 is these conclusions. Blowdown through EC -- end of ECC.
| |
| 7 UPI is not injected, that is independent of UPI injection.
| |
| 8 A pool of UPI water was formed on the top of upper core 9 plate.
| |
| 10 DR. ZUBER: This is item 2 and 3?
| |
| 11 MR. TAKEUCHI: Two.
| |
| 12 DR. ZU3ER: Two. Okay. This is something which 13 is new in UPI, that pool of water.
| |
| 14 MR. TAKEUCHI: Yeah, this one is a new. This is a (m
| |
| %.)
| |
| 4 15 start at the very baginning.
| |
| 16 DR. ZUBER: Okay. Good. Okay, okay. Good.
| |
| 17 okay. This is something you really have to show. What 18 experimental data you have to show that the code can 19 calculate this in a reasonable way.
| |
| 20 MR. NISSLEY: The UPTF.
| |
| 21 MR. TAKEUCHI: Yeah.
| |
| 22 DR. ZUBER: Not to tell you you have to really 23 show the capability and how you do that. How do you model 24 the CCFL, how you calculate that particular process? I 25 don't know whether you want to do now, but I think this is i
| |
| i 7JR1 RILEY & ASSOCIATES, LTD.
| |
| l (("N) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l
| |
| Washington, D.C. 20036
| |
| '202) 842-0034
| |
| | |
| 120 1 something you have to --
| |
| (~}
| |
| \,_ /
| |
| 2 MR. TAKEUCHI: Convince you.
| |
| 3 DR. ZUBER: That is new. And this is something --
| |
| 4 DR. WALLIS: The claim, though, is that it is all 5 stirred up and there is water everywhere, so as long as 6 there is some water over the colder channels, it is 7 available to come down.
| |
| 8 DR. ZUBER: Okay. Fine. I 9- MR. TAKEUCHI: Yes.
| |
| 10 DR. ZUBER: Yeah. The point is then how do they 11 -- do they have really experimental data to show that this 12 really happens? 1 13 DR. WALLIS: We are going to hear about j i
| |
| 14 experiments.
| |
| 15 MR. NISSLEY:
| |
| [c )\ UPTF has a single injection nozzle, 16 so it has the same kind of asymmetries and we hope to get to l 17 that.
| |
| 18 DR. WALLIS: The experimental verification part of l 19 Dr. Takeuchi's presentation this morning, are you going to 20 get to it?
| |
| 21 MR. TAKBUCHI: Yes. Second -- in the second 22 presentation.
| |
| 23 DR. WALLIS: Okay.
| |
| 24 - MR. TAKEUCHI: And liquid downflow is controlled l
| |
| 25 by subcooled counter-current flow at upper core plate.
| |
| t ANN RILEY & ASSOCIATES, LTD.
| |
| /^)'
| |
| i s, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 i
| |
| (202) 842-0034 i
| |
| | |
| L 121 i
| |
| 1 DR. ZUBER': Okay. 7 tis is also -- t
| |
| _( 2 MR. TAKEUCHI: It is not shown to'you --
| |
| 3 DR. ZUBER: This is also in your, what you --
| |
| 4 MR. TAKEUCHI: Yes, I showed it.
| |
| l I 5 D:R . ZUSER: Okay. No. For UPI.
| |
| I 6 MR. TAKEUCHI: UPI. This come. All'the UPI, !
| |
| 7 after one. Two, three, four, five are all -- you get it.
| |
| l 8 And the liquid downflow takes place preferentially into the
| |
| : 9 low power core region, not hot assembly, not directly into 10 the hot assembly, not average channel either. Juld hot l 11 assembly, the flow pattern in the hot assembly, there is no 12 ' cross-flow, liquid cross-flow in the two-thirds of hot 13 assembly, but what we see. Therefore, many issues are 14 covered by 3/4 loop front because the behavior the same as
| |
| () 15 3/4 loop plant, as far as hot assembly is concerned.
| |
| 16 DR. SCHROCK: You concluded that the models in the 17 approved code are sufficient to analyze the processes that 18' are described in 2 through 5, and I think that is 19 questionable. I don't think that you have shown us how you 20 can reasonably come to that conclusion.
| |
| 21 MR. TAKEUCHI: Okay, then, how about this? As 22 long as the computer model test, this is the way it happens.
| |
| 23 DR. ZUBER: I don't believe it. The only way to 24 prove it, you showed compared to experimental data.
| |
| 25 MR. TAKEUCHI: That is coming the second, second i
| |
| l' l [~' ANN RILEY & ASSOCIATES, LTD.
| |
| !( . , Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l 122 1 presentation.
| |
| ['')
| |
| %J 2 DR. WALLIS: So these are the conclusions yet to !
| |
| 3 be verified by experiment?
| |
| 4 MR. TAKEUCHI: Yes.
| |
| 5 DR. WALLIS: Okay. So these are sort of 6 hypotheses or hypothetical conclusions?
| |
| 7 MR. TAKEUCHI: This goes around in the SALP code 8 because --
| |
| 9 DR. ZUBER: What do you expect --
| |
| 10 MR. TAKEUCHI: This exercise has tried to pick up 11 the important phenomena related to the UPI. And then take 12 into what physical model we have, where it is going to 13 applied. Then go to the verification. When they come up to 14 support, now we have the verification, so that maybe we can (s) 15 believe this, the results, computer results of UPI plant.
| |
| 16 We go through the scaling verification, we go through aging 17 verification of the codes against each individual test.
| |
| 18 DR. SCHROCK: When you introduce subcooled water 19 into steam, as you do in the reactor, you generate 20 extensive, vigorous mixing flows which are absent in the 21 situation that you have an approved model to analyze.
| |
| 22 MR. TAKEUCHI: Yes.
| |
| 23 DR. SCHROCK: And for you to argue that you find 24 that the approved models are adequate for this case, where 25- you must know that there is a major impact on how the liquid i
| |
| l
| |
| [~h ANN RILEY & ASSOCIATES, LTD.
| |
| (_,/ Court Reporters .
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| l
| |
| | |
| _1 - . _ . . _ _ .. . _ _ . . __ . . _ . _ . _._ . . _ _ . . _. _ _ _ _
| |
| o 123 moves as a consequence of;the non-equilibrium between the I l 1''N 2 _w ater and the steam. l
| |
| .Q 3 MR. TAKEUCHI: Yes. j 4 DR. SCHROCK: You change the volume of steam-flow i 1
| |
| ;5 markedly. You do a lot of things that modify the I 6 ' hydrodynamics.
| |
| 7 MR. TAKEUCHI: Yes.
| |
| l: ,
| |
| 8 DR. SCHROCK: And you are arguing that.the same j 9 models for counter-current flow limiting situations will be
| |
| : 10 applicable.
| |
| 11 MR. TAKEUCHI: Yes. Especially counter-current
| |
| ! 12 flow limit with subcooled water is the major -- going to be j i 13 a major subject of my second presentation.
| |
| 14 DR. WALLIS: So that you have really got two
| |
| ( ) 15 different. parts. You have got the central part, which is i 16 hotter.
| |
| I 17 MR. TAKEUCHI: Yes.
| |
| 18 DR. WALLIS: And the outer part which is colder.
| |
| 19 MR. TAKEUCHI: Yes.
| |
| l '20 DR. WALLIS: And the water goes down the outer l 21 part. But the four quadrants no longer are playing a role l p
| |
| 22 -in the CCFL?
| |
| 23 MR. TAKEUCHI: Right.
| |
| ~24 DR. KRESS: Refresh my memory. What would the 1 25 behavior be if you didn't have any upper plenum injection at EA f u i, ANN RILEY & ASSOCIATES, LTD.
| |
| (,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 l (202) 842-0034 l
| |
| i I
| |
| i- ,
| |
| | |
| . . . . _ . . . - - . - - .. ~ . - - .. .. . . . -- - - . - . . .
| |
| i l 124 l 1 all? Would you still be injecting into the hot leg with the 2 other sources of water? Does that end at the time the tipper
| |
| (
| |
| 3 plenum injection starts?
| |
| 4 MS. DEDERER: No. The cold leg injection 5 continues. If there's only high-head safety injection it's 6 not as great a volume as the upper plenum injection.
| |
| t l 7 DR. KRESS: It does run out then, in other words?
| |
| 8 MR. TAKEUCHI: No, it does not run out.
| |
| l 1
| |
| 9 MS. DEDERER: It will continue to inject. i 10 DR. KRESS: But it's a lot lower volume of I i
| |
| 11, injection than the upper plenum. So by itself it's not l 12 enough to reflood. I l
| |
| 13 MR. NISSLEY: It's not -- !
| |
| 14 DR. KRESS: Okay. I 1
| |
| {f 15. MR. NISSLEY: There's three major ECCS -- the
| |
| [ - 16 accumulator is capable of filling the lower plenum and most 17- of the downcomer by itself and initiating reflood. That 18 goes into the cold leg. The low-head safety injection is 19 the higher of the pumped sis. It goes into the upper 20 ~ plenum. The high head also goes into the cold leg and is I 21 maintained through the balance of the transient.
| |
| 22 Dk. ZUBER: Let me say something just to what 23 Richard was saying. I agree with him. The code always had
| |
| . 24 difficulty calculating effective condensations. I think 25 this goes back for'20 years. And I didn't see any i
| |
| 4.
| |
| A- ANN RILEY & ASSOCIATES, LTD.
| |
| . k ,)
| |
| m Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034 1'
| |
| f-
| |
| | |
| l 125 1 improvement since.
| |
| I [''}
| |
| 'v' 2 And the difference what you have here, and this is
| |
| ! 3 I think what Richard was alluding, you want to really sweep 4 many of these problems there like in three or four loops.
| |
| 5 They are not. Because here you have cold liquid which is a 6 characteristic of this plant coming and you have 1
| |
| 7 condensation on it. And dynamically that's not going to be 8 the same. And then to arm-wave that this is the same, I 9 don't have to address them, is begging the issue.
| |
| I 10 MR. NISSLEY: Then let me clarify. That last 11 comment there is only relevant to item number 5. What we're 12 saying is the hot assembly has a bottom-up reflood, so the 13 reflood heat transfer for the hot assembly --
| |
| 14 DR. ZUBER: Then you should be more specific.
| |
| ! ,-~.,
| |
| 15 DR. KRESS: Yes, but that's why I asked the (v) 16 question about how you calculate the enthalpy going into 17 that, because these things that Virgil's talking about can 18 affect flow patterns and how they mix in that lower node and 19 will affect the enthalpy going into that or could, and 20 that's the reason I asked that question about how do you 21 actually calculate that enthalpy.
| |
| 22 MR. TAKEUCHI: Okay.
| |
| 23 DR. KRESS: And that was the reason, because of 24 these things going on in the upper part may affect what's 25 going on in the lower part.
| |
| I
| |
| \
| |
| ' FN ANN RILEY & ASSOCIATES, LTD.
| |
| ( ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . . _ . . _. . - . ._ _ _ _ . . . _ . . . _ - _ _ _ _ . . _ . . . _ , _ . _ m. _.. ._ . _ _ .
| |
| 126 i
| |
| 1 MR. TAKEUCHI: And this one I just meant'for the
| |
| ("'N " hot spot" in the hot. assembly. That takes place top i
| |
| %I 3- 'two-thirds of hot assembly. And the effect of this one I v4 was going to say this way, that this is the total 5 uncertainty, the parameters is for the best estimate LOCA.
| |
| 6 And the first group is for the initial fluid conditions.
| |
| 7 Second group is the power distribution uncertainties. And 8 the third group is the global model parameter uncertainties.
| |
| 9 And I come back here in a moment, because there 10 are some differences. And this is the uncertainty -- to the 11 . local hot spot. At the local hot spot we have the same 12 situation as three- and four-loop plant, so we can use this 13' uncertainty calculation the same as three- and four-loop 114 plants can be applied to the UPI plant, PWR. That's the n
| |
| 15
| |
| .( thing I was trying.to emphasize. Because of the' flow
| |
| : 16. patterns, that comes from -- the flow pattern of the UPI B
| |
| 17 water from this UPI water is not directly entering in this 18 region. Not directly. Not like here. So that when we come 19 to the uncertainty treatment of the peak clad hot spot in 20 this one, we can use the same imputation as three- and 21 four-loop plants. That's the thing I wanted to come up 22 with.
| |
| 23 DR. WALLIS: Well, I suppose that's all right, 24 except if for this particular method of injection peak clad 25 temperature turns out to be more sensitive to some of these ANN'RILEY & ASSOCIATES, LTD.
| |
| T(
| |
| s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| ., , . - .. _. - - . - - .-. . . . ~ . . _
| |
| - . . _ - ~ . - . .
| |
| 127 1 ' things. I'm not sure you can say that'because it's being 2- used f or three and four it 's therefore okay.
| |
| I'm not quite 3 sure. '
| |
| 4 :MR . TAKEUCHI: Because injection is made at the
| |
| ~5 upper core plate, but it does not get into the top of the 6 hot assembly directly. And it's coming only through the 7 lower one-third of the elevation it comes into. So when 8 this comes to the hot spot, there will be no difference than
| |
| ! 9 the three- and four-loop plants.
| |
| -10 MR. NISSLEY: Let me clarify something 11ere. The 11 point here is these are all the parameters that are arranged i
| |
| 12 on a plant-specific basis. We are not carrying over any
| |
| .13 - numerical results for three- and four-loops. We're carrying 14 over the parameters that are ranged in a UPI calculation.
| |
| :O 15 ,DR. ZUBER: When do you arrange your capability to 16 model, calculate CCFL?
| |
| 17 MR. TAKEUCHI: That comes into the uncertainty 18 parameter, XCONDU and XYDRAG. Those are new uncertainty 19 parameters, did not exist in the three- and four-loop l 20 plants.
| |
| 21 DR. KRESS: But those are just multiplying factors 22 on your original correlation for those.
| |
| L 23 MR. TAKEUCHI: Yes.
| |
| : i. 24- DR. KRESS: So you vary the multiplying factor.
| |
| l 25 DR. WALLIS: I was puzzled by XDRAG and YDRAG l
| |
| L l
| |
| l; ("] ANN RILEY & ASSOCIATES, LTD.
| |
| \/ Court Reporters 1025 Connecticut a'enue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
| |
| | |
| l 128 1 being different. Now there's XYDRAG.
| |
| 2L MR.-TAKEUCHI: XYDRAG is --
| |
| (
| |
| 3 DR.-WALLIS: On XDRAG and YDRAG, doesn't it?-
| |
| '4 MR. TAKEUCHI: Yes. That's a combination 5 abbreviated by XDRAG and YDRAG by XYDRAG.
| |
| 6 DR. WALLIS: But your drag coefficients --
| |
| 7- MR. TAKEUCHI: Drag coefficient for downscale flow 8 and horizontal flow calculated differently. So originally i
| |
| 9 where we don't abbreviate originally we_ call the XDRAG for !
| |
| i 10 the' axial flow drag coefficient. I 11 IMt. WALLIS: Is this dependent on flow regime, and 12 for a droplet you'd expect it to be the same. Maybe for an 13 interface it's different.
| |
| 14 MR. TAKEUCHI: This one is the independent flow
| |
| .f~g l
| |
| () 15 regime. That's the way we -- ;
| |
| 1 16 DR. WALLIS: Coefficients horizontally and 17 vertically?
| |
| le MR. TAKEUCHI: Correlations independent of flow 19 regime, but those uncertainty parameters --
| |
| 20- DR. ZUBER: How can it be independent of flow 21 regime? The correlations are just ti.ed to the flow regime. !
| |
| 22 I mean, each flow regime has a diffe. cent configuration of l*
| |
| 23 interface. Each one has a different drag coefficient.
| |
| 24 DR. WALLIS: I think that was a mistake, because 25 your next part of the handout has different flow regimes.
| |
| t
| |
| [ .rO- ANN RILEY & ASSOCIATES, LTD.
| |
| j (l Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 l -(202) 842-0034
| |
| < - ,, + , , - . , - - - .m-,- -+ --
| |
| | |
| l 129 1 MR. NISSLEY: The multiplier is the same for each i
| |
| ['')
| |
| ~
| |
| %j 2 flow regime, and the multiplier is the same in the axial and l 3 lateral direction.
| |
| l 4 DR. WALLIS: Yes.
| |
| l 5 DR. KRESS: But you vary them differently, don't 6 you? I mean, you will increase one versus the other one l
| |
| 7 instead of increasing both at the same ime or decreasing 8 them. So that you get different ratios, drag is what --
| |
| 9 MR. NISSLEY: We looked at ranging the 10 condensation alone, we looked at ranging the drag alone, and 11 we looked at ranging them in combination, and concluded that 12 ranging them in combination --
| |
| ; 13 DR. KRESS: No, my question was, do you -- when
| |
| (
| |
| 14 you range the drag, does it affect the horizontal drag and
| |
| ,a l (w.-)
| |
| 15 the vertical drag at the same time, or do you vary those so 16 that you get a different movement of the materials in those 17 directions?
| |
| I 18 MR. TAKEUCHI: We did not check that sensitivity 19 study, but we applied both -- same amount of the uncertainty 20 factor. Yes.
| |
| )
| |
| 21 DR. KRESS: Okay. I 22 DR. ZUBER: It is conceivable that if you just had l
| |
| 23 different values, that changing the horizontal one would l l l 24 then preferentially either permit or preclude the flow of 1
| |
| l 25 the liquid.
| |
| [''}
| |
| ( _,/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters I
| |
| 1025 Connecticut Avenue, NW, Suite 1014 t Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| -. ..- .. . - . -. . . - - . . . . - . - ~ . . .. - -
| |
| ' '130 l .
| |
| l L 1 DR. KRESS: Yes. And physically.that could be
| |
| (.
| |
| I
| |
| , (''\ . 2 real, because you.could have different flow regimes in those l- 3 directions.
| |
| 4 MR. TAKEUCHI: That is possible physically, but 5 ' practically we didn't do'it.
| |
| 3 l
| |
| 6~ DR. ZUBER: Practically.you want to get approval !
| |
| 7 for what-you did, and therefore you have -- therefore you 8 have to model the physics.
| |
| 9 MR. TAKEUCHI: With the least amount of 10 uncertainty. That's what we tried.
| |
| 11 DR. WALLIS: Can I ask where you are on your 12 presentation? Do you.have another presentation besides the 13 one you've given us here?
| |
| 14 MR. NISSLEY: We have another one that has --
| |
| 't 15 DR.'WALLIS: I was just wondering-how much we can 16 get through before lunch.
| |
| 17 MR. TAKEUCHl: I hope at.least I will be done in 18 one hour, but it looks like it's going on two.
| |
| 19_ DR. WALLIS: Well, I think we're probably going to 20 have Westinghouse continue after lunch. We're going to have 21 to do that. I was wondering if we could get through the
| |
| '22 experimental or something before and maybe someone could --
| |
| 23 when you have time to think about it over lunch, you can 24 come back with a summary for half an hour which tells us 25 what we real.ly need to know from everything you've said.
| |
| l l
| |
| 6
| |
| /~ ANN RILEY & ASSOCIATES, LTD.
| |
| k ,h) m Court Reporters
| |
| , 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| -__ - . ~ _
| |
| | |
| 1 l
| |
| 131 l l
| |
| 1 Can you do that? !
| |
| [' 2 MR. NISSLEY: Fair enough.
| |
| ()1 3 DR. WALLIS: Can you get through say half an hour, l 4 can you talk about the experiments, or do you want to talk i 5 about these physical models? Are these the same as in the 6 three , four-loop plant, physical models?
| |
| 7 MR. TAKEUCHI: The physical model is the same as 8 three- and four-loop plants.
| |
| 9 DR. WALLIS: So which ones do we need to look at, 10 if any?
| |
| 11 MR. TAKEUCHI: Probably --
| |
| 12 DR. WALLIS: Because we could question every one 13 of these equations.
| |
| 14 MR. TAKEUCHI: Yes, that's already licensed the
| |
| ( ) 15 correlations and the correlation is licensed and its 16 application is also approved or -- some of the correlation 17 I'm familiar with and some of them I'm not. But the kind 18 of -- the type of the correlation, it's being used for the 19 subcooled CCFL can be seen from this example.
| |
| 20 This is the analysis of the GE CCFL test. The 21 test was conducted by GE and the test section is made of 22 BWR, full-scale BWR, 8 by 82 assembly, and that was one is 23 modelled for COBRA / TRAC here and test section fuel bundle is 24 represented by channel four, which tie plate placed at the 25 top of channel five, tie plate.
| |
| l
| |
| /
| |
| '-) ANN RILEY & ASSOCIATES, LTD.
| |
| (,j Court Reporters 1025 Connecticut A'renue, NW, Suite 1014
| |
| ; Washington, D.C. 20036 l (202) 842-0034 l
| |
| | |
| , _ > , . . .s_ . _ . . . _. _ .. - . _ _ . _ , . . . . _ _ . . ._ _ - . _ _ __
| |
| 132
| |
| .1 The test is conducted with constant liquid ^
| |
| i '$ 2- ' injected at the top of the tie plate at the constant rate d: '
| |
| :3 and they keep on there, and after a steady drain flow rate .
| |
| '{
| |
| 4 .is. established, saturated steam is injected from the bottom j l
| |
| 5' and the injection rate was gradually increased. !
| |
| 6J At the very beginning the drain rate is-not 7: affected, and that. period I refer to as the drain period, g 8 and sooner or later the drain rate is going to be reduced 9 and that period I refer to as the flow limiting and 10 eventually no water is coming down -- that is the flooding 1
| |
| 11- period.
| |
| 12 When I look at the flow regime computed in this 13 situation is that bundle, all the way bundle up to just 14 above the upper core plate the flow regime here has remained h'& 15 l-( ): film drop, flow regime --
| |
| g -16 DR. ZUBER: Okay. The droplets and film are '
| |
| 17 . coming down.
| |
| 18 MR. TAKEUCHI: Yes.
| |
| 19 DR. ZUBER: Go, ahead.
| |
| L' 20 MR. TAKEUCHI: That is che counter current --
| |
| 21 water is coming down through the tie plate.
| |
| : 22. DR. SCHROCK: So what is the ratio of the area of 23 the' vertical channel and the orifice?
| |
| l' l 24- MR. TAKEUCHI: Oh, ratio -- I don't recall but I
| |
| .25 have a number.
| |
| 4 ANN RILEY & ASSOCIATES, LTD.
| |
| !i: ;--D(s,/ Court Reporters l 1025 Connecticut Avenue, NW,' Suite 1014 Washington, D.C.. 20036 (202) 842-0034 t
| |
| . .. ..n- - - , - - -. , _ . - ,, , .,, .. , .
| |
| | |
| 133 1 ..DR.'ZUBER: .Well, let.me ask you, that' tie plate,
| |
| -h u
| |
| '2' Lit's not the.same tie plate you have-in your plant?
| |
| ,, 3 MR.:TAKEUCHI: But as far as CCFL in concerned the
| |
| '4' values'are similar,.quite similar.
| |
| 1
| |
| .5- I will show in the second presentation. I.have 1
| |
| '6. data in there.
| |
| 7I DR'. ZUBER: Okay.
| |
| 8 .DR. WALLIS: These periods are what? The steam l l9- injection rate is changing or something? l 10 MR. TAKEUCHI: -Yes. The steam injection rate is 1:L increasing with time like this. <
| |
| i 12- Then we've got-the flow regime in this cell. That 13- -is the leakage injection level. It goes'through a film l 1
| |
| 14 drop, churn-turbulent, and large bubble regime.
| |
| () ?15 DR. SCHROCK: At your lowest steam flow is that
| |
| : 16. ' orifice running full or is it like a weir?
| |
| 17 MR. TAKEUCHI: . Lowest steam flow rate -- probably 18 after this point ^it is the drain period. That corresponds.
| |
| '19 to this one.
| |
| 20 DR. SCHROCK: But I am asking how does it drain?
| |
| 21- Is it a weir type of draining pattern or does it fill the
| |
| -22' hole?
| |
| 23 MR. TAKEUCHI: Weir type? No, I don't think.
| |
| :24 This one is annular flow. That is the film drop flow regime
| |
| -25 is film liquid is coming down through the bundle -- no, fuel l
| |
| ; ' ANN RILEY & ASSOCIATES, LTD.
| |
| i Court Reporters
| |
| ! 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C 20036
| |
| '(202) 842-0034 i
| |
| > , u , _ - _ . _ - _
| |
| | |
| l 134 1 rods.
| |
| l 1 i
| |
| [~'
| |
| U) 2 DR. WALLIS: Not the orifice. But the question i
| |
| J 3 about what happens at the orifice, l
| |
| 4 MR. TAKEUCHI: At the orifice the same situation I i
| |
| 5 is taking place as far as flow regime is concerned. )
| |
| I 6 We cannot model so microscopically at the spot of '
| |
| 7 the tie plate. We have to look at overall behavior.
| |
| 8 DR. SCHROCK: It makes a difference whether any l 9 gas phase can get up. In this kind of a test you are 10 ramping the steam up. If you start with the hole plugged, 11 then you have to first unplug it.
| |
| 12 MR. TAKEUCHI: Plug it and unplug it. Well, 13 unless the steam goes up, there is no chance to have the 14 churn-turbulent flow regime, for large bubble flow regime
| |
| ,-m t ) 15 unless steam --
| |
| V 16 DR. SCHROCK: I never mentioned the term 17 churn-turbulent regime. I think to talk about such a regime 18 in the vicinity of that orifice is absolutely meaningless, 19 but it would make some sense to talk about whether the 20 orifice is running full or whether it is not running full at 21 the beginning of the test when you start ramping the steam 22 up. What follows will be a bit different.
| |
| 23 DR. WALLIS: Anyway, you are going to show us that 24 you can model that data.
| |
| 25 MR. TAKEUCHI: Yes.
| |
| /~'T ANN RILEY & ASSOCIATES, LTD.
| |
| (_) Court Reporters 1025 Connecticut Ave.1ue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . _ _ _ _ . . m . _ . ._ _ _. _ . _ _._ ..m _
| |
| 135
| |
| :1 DR. WALLIS: Whether orinot you know'what is
| |
| /N '2 : .
| |
| . happening at the orifice plate.
| |
| N~.]
| |
| : 3. DR.-KRESS: _ When you start this test; there's no 4 ._ liquid in the bottom. channel at all?
| |
| L i 5 MR. TAKEUCHI: Bottom channel? Here?
| |
| l 6 DR. KRESS: Yes, i l
| |
| 7 MR. TAKEUCHI: This one is empty. ;
| |
| i 8- DR. KRESS: The whole-thing is empty when you 9 start the test?
| |
| 10 -MR. TAKEUCHI: That's right, yes.
| |
| 11 DR. KRESS: You establish steam flow.
| |
| 12 MR. TAKEUCHI: Yes. i l- 13 DR. KRESS: Then you star the liquid injection?
| |
| 14 MR. TAKEUCHI: No --
| |
| (n) . 15 DR. KRESS: You' start the liquid first?
| |
| l 4
| |
| : 16. MR. TAKEUCHI: Yes. ,
| |
| 17 DR. KRESS: And then you establish the steam 18 flow --
| |
| 19 MR. TAKEUCHI: Yes.
| |
| 20 DR. KRESS: So there is water down there in that 21 bottom?
| |
| i 22 MR. TAKEUCHI: At this time, yes.
| |
| 23 MR. NISSLEY: But, Kenji, there is a constant l
| |
| l 24 drain out of the bottom.
| |
| I
| |
| :25 DR. KRESS: There is a constant drain at the i
| |
| t /"' ANN RILEY & ASSOCIATES, LTD.
| |
| [ \%,)s Court Reporters i
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| 136 1- bottom so more than likely that whole channel that is L /~] 2 restricted for it is probably empty of anything but steam
| |
| 'V 3 when you -- I mean it is probably empty when you first start 4 the steam flow. There may be some air in there. I don't j 5 know what the initial conditions are.
| |
| 6 MR. TAKEUCHI: The initial condition is that --
| |
| 7 the initial condition it's filled by steam.
| |
| 8 DR. KRESS: So that answers your question. When 9 it starts out it is not plugged. It's just an open hole.
| |
| 10 DR. ZUBER: What you just'said, you started --
| |
| 11 DR. SCHROCK: I didn't hear him say that.
| |
| 12 DR. KRESS: Okay.
| |
| ! 13 DR. ZUBER: You start first with water coming down 14 and you increase the flow of the steam or you start with l
| |
| 15 steam and then you go to the water in?
| |
| (("b) 16 MR. TAKEUCHI: To begin with, it's all filled by 17 steam, not moving at all,-and then the water comes in at a 18 constant rate, no change, a constant rate is coming, pouring 19 into it.
| |
| 20 After it's started the steady state is reached.
| |
| l 21 Steam is coming into it.
| |
| 22 DR. ZUBER: You start increasing the steam?
| |
| 23 MR. TAKEUCHI: That's right, l
| |
| 24 DR. WALLIS: Can you give us you next hour 25 presentation in half an hour? How are we going to do this?
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| , O)
| |
| ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 l (202) 842-0034 l
| |
| | |
| 137 l' MR. TAKEUCHI: Half an hour -- ah -- I can do 2~ but I think you have more questions.
| |
| f~')y M.
| |
| 3 LDR. WALLIS: We will break at between half past 4 12:00 or a quarter-to 1:00 -- some time in there we will 5' break. Have you:got a convenient time you can break in 6~ about 20 minutes or something like that? Or do you want to 4 l 7 break now and then come back? l l
| |
| 8 MR. NISSLEY: Break now?
| |
| 9' DR. WALLIS: Break now? ,
| |
| 10 MR. NISSLEY: Why don't we break now. We'll l
| |
| 11 filter that down and then we will draw some global
| |
| ~
| |
| 12 conclusions.
| |
| 13 DR. WALLIS: And then you can figure out how to I 14 present your second hour in half an hour.
| |
| ! i/~'i 15 MR. TAKEUCHI: In half an hour.
| |
| l- ( _ /
| |
| 16 DR. WALLIS: So we can then move on to 17 conclusions.
| |
| : i. 18 I think there are so many interesting phenomena r
| |
| 19 going on and so many questions we could go on for weeks with 20 this, and we have got to focus on what really matters.
| |
| i
| |
| : 21. MR. NISSLEY: How much time do we have for the 22' staff? Could I ask how much they expect to take?
| |
| 23 DR. WALLIS: The truth.
| |
| 24 MR. ORR: My real presentation is, I expect, I
| |
| '25 only have six slides -- 15 minutes. Can I expect to get ,
| |
| ['T ANN RILEY & ASSOCIATES, LTD.
| |
| \s ,)- Court Reporters 1025 Connecticut Avenue, NW, St'ite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l 138 l' tomatons and stuff thrown at me'for another 15 or so? I J
| |
| - f~' 2 .would say about a half-hour of that we could give up. We
| |
| , k_)r-3 . don't need all that time.
| |
| 4 DR. WALLIS: We can ask you all the questions we j 5 ' asked this morning.
| |
| 6 MR. ORR: And1you will probably get a lot of l
| |
| 7' answers like it/wasn't really relevant to our conclusions.
| |
| l 8 (Laughter.) i 9 DR. WALLIS: Okay. That's useful, that's very-
| |
| ! 10 useful to know because we need to focus on what matters.
| |
| : 11. MR. SINGH: Are we on the record?
| |
| 12 DR. WALLIS: Yes, we are still on the record, but 13 we'are not going to be for long if you will agree to make a 14 break now -- so we will recess for lunch and we will return i
| |
| -f* 15 at -- can we. return at 1:00?
| |
| 1
| |
| ( I think we need to return i ~
| |
| 16 early. Can we return at 1:00, please? Okay.
| |
| ( 17 [Whereupon, at 12:17 p.m., the meeting was ,
| |
| i l
| |
| 18 . recessed, to reconvene at'1:00 p.m., this same day.] I l
| |
| -19 l
| |
| 20 ;
| |
| I L
| |
| 21 2h2 l I
| |
| 23 24 25 l
| |
| s
| |
| ! ANN RILEY & ASSOCIATES, LTD.
| |
| k's T) ' Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 l
| |
| Washington, D.C. 20036 (202) 842-0034 1 l
| |
| l
| |
| | |
| _>,.. __ ._. . _ _ _ . _ - .- _ . _ - _ . _ . . . ~ . - _ . . _ . . . - _ . . _ . _ . _
| |
| 139 1- AFTERNOON SESSION
| |
| ('')
| |
| '2 (1:00 p.m.)
| |
| \g.
| |
| 3 DR. WALLIS: All right. We will. reconvene then.
| |
| : 4. .Are you ready? Let's go.
| |
| '5 MR. NISSLEY: .Okay. I B
| |
| 6 What we've done with the remainder of our 7 1 presentation is to condense it quite a bit. We have made ,
| |
| 8 some preliminary conclusions so far that the only basis 9 we've shown is some PWR calculations.
| |
| 10 We're going to show some comparisons with
| |
| ~11 experimental data, with a focus on substantiating these 12 claims that we've made, and later on, we'll follow that up with some overall conclusions.
| |
| i 13
| |
| ~
| |
| 14 The package that we've handed out has the full set 15'
| |
| :( of slides, but we have re-ordered this and removed much of 16 the material, so you might be better off just following 17 what's on the screen up here.
| |
| 18 MR. TAKEUCHI: This is the model for GE, the CCFL 19 test, I showed you before, and they have five tests, run 60, 12 0 - 61, 62. Those three tests are almost identical. Run 73 is 21 .almost the same as run 60. However, injection point is 22 somewhat different.
| |
| 23 The transient starts with injecting the sub-cooled 24 at the constant rate after steady state of draining is l
| |
| 25 established such that the steam is injected with increasing l~
| |
| i h ANN RILEY & ASSOCIATES, LTD.
| |
| i- . Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| .. Washington, D.C. 20036 (202) 842-0034
| |
| . .-. .~_ - . - ,-
| |
| | |
| 140 1 rate from the bottom. This is steam flow rate as a function
| |
| /~'h 2 of time.
| |
| () 3 This is steam injection rate, and after floodings 4 take place, the steam injection rate is reduced, and this is 5 the sequence of the transient.
| |
| 6 When the steam injection rate is not so much, we 7 get drain period here. Drain rate is almost constant. And 8 then drain rate is going to be reduced and then reach the 9 flooding.
| |
| 10 During the decreasing rate, the process is j l
| |
| 11 reversed. Then it reach steady, constant drain rate.
| |
| 12 DR. SCHROCK: These are experimental data?
| |
| 13 MR. TAKEUCHI: That's the boundary condition. And 14 then the experimental data drain rate --
| |
| ,m 15
| |
| ,'(U) DR. SCHROCK: Flow rate is what? Steam?
| |
| 16 MR. TAKEUCHI: This the steam flow rate. This one 17 is the drain rate.
| |
| 18 DR. SCHROCK: Liquid drain rate, meaning the 19 counter-current liquid flow?
| |
| 20 MR. TAKEUCHI: Counter-current liquid flow.
| |
| 21 DR. SCHROCK: So, that's not a boundary condition;.
| |
| 22 that's the result.
| |
| 23 MR. TAKEUCHI: The drain rate is a result.
| |
| l l 24 DR. SCHROCK: But my question was these are
| |
| ! 25 experimental measurements that you're showing.
| |
| (~')
| |
| (_,/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
| |
| ; (202) 842-0034
| |
| | |
| . . . . - - - - - - - . . ~ - - . . . . .- - - - - . . - . . - - . . - .
| |
| I 141 1- MR. TAKEUCHI: Those are the boundary conditions
| |
| -["T. 2 and the drain rate illustrated by computations, and the test i
| |
| \_)- . .
| |
| 3 . data comes in this figure.
| |
| 4 DR. SCHROCK: This is a computed result. ,
| |
| 5 MR. TAKEUCHI: Yes. ,
| |
| 6 DR. SCHROCK: By what? i 7: MR. TAKEUCHI: By COBRA / TRAC. However, this is 8 the test data.
| |
| 9 Triangle is test data, and bold phase triangle is 10- the steam > increasing phase and regular triangle is the steam 11 decreasing phase, and the prediction is shown by the
| |
| ~12 squares, bold squares for the steam increasing rate, phase, 13 and regular square for the steam decreasing phase, and this 14 is one test.and comparison of the test, but the sub-cooled j 15 CCFL prediction and data are almost the same.
| |
| 16 DR. FONTANA: We've got two answers for the same 17 ~-- 'the .07 looks like two answers, two predicted answers.for
| |
| )
| |
| 18 .the same -- okay. Excuse me. I read-it wrong.
| |
| 19- MR. TAKEUCHI: This one is sub-cooled. This one l
| |
| 20 is saturated.
| |
| 21 DR. WALLIS: The triangles and the squares.
| |
| : 22. MR. TAKEUCHI: Triangle is data, squares are 23 predictions.
| |
| I
| |
| ~24 DR. WALLIS: Okay.
| |
| 25 MR. TAKEUCHI: This one is regarded as i
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| . . . Court Reporters 4 1025 Connecticut Avenue, NW, Suite 1014
| |
| [ Washington, D.C. 20036 l (202) 842-0034
| |
| | |
| ; 142 1- conservative because the computed' drain rate is stopped much L , 2 earlier, with much less steam coming up. So, it's l' 3 conservative.
| |
| ! 4 DR. SCHROCK: I don't understand why you present 5 it this-way. You're making' calculations-to compare with 6 ~ experimental data, but'you do it at different points in the 7 steam flow curve.
| |
| 8 MR. TAKEUCHI: Different points. However --
| |
| 9 DR. SCHROCK: Different steam rates, if that's 10 your independent variable.
| |
| 11 MR. NISSLEY: In the calculation and the tests, 12 the test basically progresses like this and then comes back 13 like this.
| |
| l 14 DR. SCHROCK: My point is that you've done 15-
| |
| ) calculations for steam injection rates for which there are 16 'no data shown here.
| |
| 17 DR. WALLIS: If you go back to the previous E 18 figure, figure 5-3, that's a continuous code.
| |
| 19 MR. TAKEUCHI: The points are here.
| |
| 20 DR. SCHROCK: Why are there not two companion 21_ points.for each steam injection rate, one from experiment, 22 one from-calculation?
| |
| 23 MR. TAKEUCHI: In computation, this one is a 24 continuous calculation, and the time points are time space.
| |
| 25 DR. ZUBER: Why pick up something ad hoc?
| |
| l l - -
| |
| ; ANN RILEY & ASSOCIATES, LTD.
| |
| ,'' ~ Court Reporters
| |
| , 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l-
| |
| ]
| |
| i
| |
| | |
| l 143 i 1 MR. TAKEUCHI: Well, this is the computation. I
| |
| : 2 picked this point, this point, and this point and made the O(~'s 3 squares.
| |
| 4 DR. WALLIS: If you look at the figure below that, 5 those wiggles are because you've actually incremented by 6 five-second intervals.
| |
| 7 MR. TAKEUCHI: So, that one is sort of --
| |
| 8 DR. WALLIS: Why does it have those wiggles in it?
| |
| 9 MR. TAKEUCHI: This wiggle is --
| |
| 10 DR. WALLIS: You've incremented in steps?
| |
| 11 MR. TAKEUCHI: This is because maybe the 12 condensation taking place. It's not smooth.
| |
| 13 DR. WALLIS: No , but it's over a long period of 14 time. It's an average or something, isn't it? It's very
| |
| !n)
| |
| 'w) 15 strange.
| |
| 16 DR. ZUBER: You have 40 seconds on the right. You 17 have again some wiggles.
| |
| 18 MR. TAKEUCHI: This is the same situation as this 19 one.
| |
| 20 DR. ZUBER: What is it?
| |
| 21 MR. TAKEUCHI: Steam decreasing phase. It's 22 around here.
| |
| 23 DR. ZUBER: But what causes it?
| |
| 24 MR. TAKEUCHI: It must be the condensation.
| |
| 25 DR. ZUBER: Let me say something, what frustrates l
| |
| ! /''s ANN RILEY & ASSOCIATES, LTD.
| |
| , !s,) Court Reporters
| |
| ; 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 144 '
| |
| l 1 me always, especially with Westinghouse. You want to obtain l
| |
| /''h 2 some approval that you are doing a correct job. The first i
| |
| \j 3 _ thing, you have to leave impression that you understand what 4 you did and then try to explain it, and we ask you what it 5 is and you say this must be.
| |
| 6 You didn't even have the curiosity to try to find 7 yourself out before coming to tell us what we want to --
| |
| 8 what you want to have it approved, that what you did is 9 correct, and you cannot even explain what you have in your 10 calculations and experiments.
| |
| 11 I think this is something which the management at 12 Westinghouse should really realize, how poorly your 13 presentations are and how frustrating it is for us to give 14 approval for something which you cannot even explain, and
| |
| , - ~
| |
| 15 then you want to explain your arm-waving with nodalization.
| |
| (\_s) 16 You cannot even explain physically what you have there.
| |
| 17 I'm sorry to -- but I mean this is building up 18 over a period of time.
| |
| 19 MR. NISSLEY: We accept your comment.
| |
| 20 DR. ZUBER: Let me just say I would like to be 21 able to say you did excellent job, because I like this 22 industry, I've spent my lifetime here. It frustrates me 23 when I can't get really good technical results to say I'm 24 proud of what you did. I would like to say that.
| |
| 25 MR. TAKEUCHI: This drain rate and the steam l
| |
| p)
| |
| (,
| |
| AIM RILEY & ASSOCIATES, LTD.
| |
| Court Reporters l
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . .~ -. . . - . ..--..-..-.-.- - - - -.- - -....-.-.-..~..-. - - - ..
| |
| l
| |
| : l. 145 1 l
| |
| l
| |
| -1 injection rate are rearranged to.show and' compare with --
| |
| j 2 DR. SCHROCK: Wait, wait, wait. Are you going to 3- ; explain this historesis? I i.
| |
| 4 MR. TAKEUCHI: .This one? !
| |
| ;- SL DR. SCHROCK: .Yes.
| |
| I' l
| |
| 6 MR. TAKEUCHI: No. We're trying to skip l l-7 saturation situation, because no time.
| |
| 8 DR. WALLIS: I understand historesis in the data, 9 but how does the theory have a historesis?
| |
| 10 MR. TAKEUCHI: That I cannot explain, because-I
| |
| ! 11 tried to understand, but I couldn't find it.
| |
| 12 DR. WALLIS: You're just running the code here. q I
| |
| l
| |
| '13- MR. TAKEUCHI: Yes.
| |
| : 14. DR. WALLIS: And the code is predicting a l 15 historesis.
| |
| 16 MR. TAKEUCHI: Yes.
| |
| 17 DR. WALLIS: That's very interesting, because if 18 you just use the correlation, it presumably wouldn't.
| |
| i
| |
| -19 DR. SCHROCK: Your code functions in a 20 quasi-steady-state world.
| |
| l
| |
| ; 21 MR. TAKEUCHI: Yes. I 22: DR. SCHROCK: Is this forcing function too fast 23 for that?
| |
| i-l '24 MR. TAKEUCHI: No, I don't think that's the 25 reason.
| |
| 4 i
| |
| ! 'T k'%,)
| |
| ANN RILEY &-ASSOCIATES, LTD. 1 Court Reporters
| |
| : 1025 Connecticut Avenue, NW, Suite 1014 !
| |
| Washington, D.C. 20036 (202)'842-0034
| |
| | |
| - -._ __ _ _ _ . . _ . . _ _ _ . _ . _ . - . _ . _ . _ - . _ . ~ . . _ _ -_ _ _ _ . . _ . _ . . _ . _ _ - -
| |
| 146 1 DR. SCHROCK: What do you think is the reason? l i .
| |
| 2 MR. TAKEUCHI:
| |
| I couldn't figure out what's the 3 reason'for it. l I
| |
| 4 MR. NISSLEY: I know this won't be a very ;
| |
| [ 5. satisfying answer, but it's possible that the build-up of a
| |
| ! 6 pool in the upper plenum, in the upper region during the 7 flooding period -fus introducing or is leading to the 8 historesis given the nodalization.in the code.
| |
| I 9 DR. SCHROCKi Which means that this is too fast 10 for quasi-steady interpretation. Isn't that right?
| |
| i 11 MR. NISSLEY: Yes.
| |
| 12 DR. WALLIS: Well, experimentally, when you have a 13 pool, you get less flow down than when you have a weir-type 14 situation. I'm not sure the code is sophisticated enough to fs
| |
| ' f i 15- pick that.up.
| |
| V 16 MR. TAKEUCHI: The decreased drain rate and the 17 steam injection rate are rearranged in such a way that the 18 flooding correlation for sub-cooled and saturated CCFL is 19- compared. The flooding correlation is expressed by the 20 number like this,-and then we get sub-cooling effects.
| |
| 21 DR. ZUBER: Hold on. Does the effect of geometry 22 or the number of holes, the geometry of the holes -- does it 23 have any effect in your corollary here?
| |
| l 24. MR. TAKEUCHI: Within the geometry, all those 25 values --
| |
| i t
| |
| i I
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| a("T i
| |
| ,) Court Reporters 1025-Connecticut' Avenue, NW, Suite 1014 t
| |
| Washington, D.C. 20036 (202) 842-0034
| |
| | |
| )
| |
| 147 1 DR. ZUBER: Because the Germans did some
| |
| /~N 2 ~ experiments.
| |
| .V" MR. TAKEUCHI: Yes. That's the collection of the
| |
| -3 4 . parameters relevant to the CCFL.
| |
| 5 DR.-ZUBER': What is the thickness of the plates?
| |
| 6 The other one is'the number and the size of the holes and 7- the shape'of the holes. ;
| |
| 8 MR. TAKEUCHI: Yes.
| |
| .9- DR. ZUBER: All of them have an effect on the 10 . pressure drop.
| |
| 11 MR. TAKEUCHI: It has an effect, and thickness, 12 diameter -- that's the shape of the holes and so forth, and 13 those are the only parameter involved in the Bankoff 14 correlations, r\ 15 g)
| |
| / DR .- ZUBER: I don't know the geometry he tested.
| |
| 16 Did he test your geometry?
| |
| 17 MR. TAKEUCHI: No, he did not test the geometry.
| |
| 18 DR. ZUBER: Did you test your geometry? If you 19 ' don't know what he tested, fine. I don't know either. Did 20 you test your geometry?
| |
| 21' MR. TAKEUCHI: I did not test it.
| |
| 22- DR. ZUBER: Okay. Good. So, you don't have test 23 data to support this.
| |
| 24 MR. TAKEUCHI: I think I have enough data to 25 support this, because the correlation -- the parameters
| |
| :f ANN RILEY & ASSOCIATES, LTD.
| |
| 's Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| ._ -__ ___ .-._...._._..m _ -._. _ _ _._ ___._-.. _._ _ . _ _ _ _ ..
| |
| 148 1- included in the correlation is only those three parameters, l /~' 2 and~as long'as'--
| |
| ks ,)\
| |
| i c3 DR. ZUBER: .Let me say there were tests in Germany l -4 -
| |
| I viewed these tests -- in 1980 or '81, and it has an L
| |
| '- 5 effect. You don't know the configuration of Bankoff's,Jyou
| |
| '6 -didn't test.it, and you want to co..vince me that-this is l1 7l correct.
| |
| 8 DR. SCHROCK: You say.in your report that'Bankoff
| |
| '9 'did sub-cooled liquid injection on a perforated plate.
| |
| L 10= MR. TAKEUCHI: Yes.
| |
| 11 DR. SCHROCK: So, you do know what he did.
| |
| [ 12 MR. TAKEUCHI: I know what'he did, but Dr..Zuber
| |
| ! 13 Lis telling me,.if I did the test with my configuration, with 14 this one, I did not do it. However, I know that Bankoff's.
| |
| 15-( perforation plate,The has-perfect circles, 15 circles on the 16- plate about this size -- no, about this size. I 17- DR. ZUBER: See, the thing is not only the size.
| |
| '18 This has an effect, and the Germans did these. tests, and )
| |
| i 19 they reported it. They did it for us 15 years ago, and they
| |
| .20 used -- they did two kinds of tests. They used one for l
| |
| INEL,. and then they used Siemens configuration.
| |
| l 21-22 DR. WALLIS: Maybe when you write your report, you 23 -should suggest that they should also make a comparison with 24 --
| |
| 25 DR. ZUBER: See, they found out that the number of L
| |
| ; ("1 ANN RILEY & ASSOCIATES, LTD.
| |
| 3 Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 b . Washington, D.C. 20036 L (202) 842-0034 l~
| |
| I L3-
| |
| | |
| 149 l
| |
| l 1 holes and the shape of the holes, the size, I mean whether
| |
| (~~') 2 the holes are close to the walls, the configuration, all of
| |
| ~
| |
| 3 them had an effect.
| |
| 4 DR. WALLIS: It's just surprising that the 1
| |
| l 5 correlation here has nothing to do with geometry at all. l 6 The Bankoff equation has,no geometry in it.
| |
| 7 DR. ZUBER: That's the reason I asked.
| |
| 1 8 MR. TAKEUCHI: Bankoff has geometry in it, and if l 9 I apply those parameters in it, the Bankoff scale becomes 10 that number. l 11 DR. WALLIS: Which has no geometry in it.
| |
| 12 MR. TAKEUCHI: Which has no geometry, because it's 13 a limiting condition.
| |
| 14 DR. WALLIS: Yes.
| |
| [/)
| |
| \_
| |
| 15 MR. TAKEUCHI: When the thickness becomes fairly 16 small, it approaches that Reynolds number.
| |
| 17 DR. WALLIS: Shall we move on and look at your 18 comparisons with data?
| |
| 19 MR. TAKEUCHI: Yes.
| |
| 20 This is the Bankoff saturated CCFL, and titat's the 21 ---those two are Bankoff's sub-cooled CCFL conditions with 22 condensation efficiency of 0.24 and 0.6, and in the case of 23 Bankoff test data, 0.24, your use of 0.24 is recommended, 24 but with this GE test data comes up to 0.60, and the test 25 data are indicated by a triangle again in this case, and the
| |
| <N ANN RILEY & ASSOCIATES, LTD.
| |
| (( ~'
| |
| ) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| L l 150 prediction is.shown by the squares, and in the case of l 1
| |
| (-
| |
| : 2 prediction, looks like the condensation efficiency is 0.24.
| |
| 3 So, this is the situation.
| |
| 4 Dk. SCHROCK: As I recall, in Bankoff's 5 experiments with the perforated plate, the conditions would 6 establish where'some of the holes were passing all liquid
| |
| '7 down full of' liquid and that the steam up-flow was then f
| |
| l 8 restricted to other holes in that plate. This has been ,
| |
| 9 observed in other experiments, as well.
| |
| 10 MR. TAKEUCHI: Bankoff's report?
| |
| 11 DR.'SCHROCK: I think so, but I haven't look at it 12 for a long time, 17 years ago.
| |
| 13 DR. ZUBER: I think it's absolutely correct, and 14 this is the reason, the size of the box, whether it's three 15 by four or four by four or six by six, has an impact, and
| |
| {)
| |
| 16 this, again, supports what Virgil is.saying.
| |
| 17 So, the geometry has an effect, and I'm not sure i
| |
| i l 18 that the geometry that Bankoff tested is relevant to what
| |
| ]
| |
| l
| |
| .19 you want to do.
| |
| l
| |
| !~ 20 MR. TAKEUCHI: This is the-scaling test analysis.
| |
| R21: The scaling test data indicated by the open circle and the 22 Lcross and so forth. This break-through area becomes smaller 23 as scaling becomes larger, and those closed circles, closed 24 squares, closed triangles are those we have obtained by 25 prediction, and since we get scaling trends right, just as !
| |
| i'
| |
| /~s ANN RILEY & ASSOCIATES, LTD.
| |
| \
| |
| , 1025 Connecticut CourtAvenue, Reporters NW,~ Suite 1014
| |
| ( Washington, D.C. 20036 l (202) 842-0034 l
| |
| [ , .- , -- _
| |
| | |
| -- . .. ~ -,- - . - , . . - - . . - . ~ . - - -
| |
| I l'
| |
| 151 1 :you mentioned.
| |
| 2 So,.we are predicting those things, and I'm going'
| |
| : i. .3- on to the analysis of the core test, and the test is a model 4 just like PWR modeling. I 5 DR. WALLIS: You're claiming that your CCFL is 1
| |
| i 6 predicted from a.two fluid model or something with interface 1 7 terms? It's not a correlation of CCFL itself.
| |
| 8 MR. TAKEUCHI: This one is predicted for the GE 9 CCFL.
| |
| 10 DR. WALLIS: It predicts each one of GE's data 11 points here.
| |
| 12 MR. TAKEUCHI: GE data point is not included in 13 here, because scaling is so small.
| |
| 14 DR. WALLIS: I thought you claimed that it worked
| |
| ([ 15 for GE, your code worked on the GE data. That's your claim.
| |
| 16 MR. TAKEUCHI: For GE test, it works well. ;
| |
| I 17 DR. WALLIS: The secondary claim is that your code .)
| |
| 18 predicts Bankoff correlation.
| |
| 19 MR. TAKEUCHI: Yes.
| |
| 20 ~ DR. ZUBER: I don't understand. You didn't use 21 Bankoff's correlations in your code. You used your two 22 fluid interfacial in your code.
| |
| 23 MR. TAKEUCHI: In the code that we have, the basic 24 . physical correlations.
| |
| 25 DR. ZUBER: Hold on. First hear the question, l f~ ANN RILEY & ASSOCIATES, LTD.
| |
| i (,, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| _ . . _ _ . , _ _._m__ .- __ . . . _ . . ,, _ _ _ . _ . _ _ . _ . _ . . . _ _. _ .
| |
| i l
| |
| 152
| |
| 'l .then provi'de.the answer if you can. How thick is the plate?
| |
| f /"' 2.
| |
| g) . Half an-inch? And you.have.the holes. I don't know how big !
| |
| 3 they are.
| |
| 4' What' kind of flow regime map are you going tc use ;
| |
| h ~5 there? fBecause all your friction factor, all your !
| |
| 6: interfacial drag coefficients are based on. flow regime.
| |
| ; 7 MR.'TAKEUCHI: Yes. '
| |
| i L
| |
| 8 DR. ZUBER: How can you really use a flow regime
| |
| '9' which is developed'for horizontal pipe and vertical pipes L 10 two or three inches in diameter and. apply them to a small L 11' plate'with~ numerous small holes? What is the rationale?.
| |
| 1' 12 MR. TAKEUCHI: In'this case, it's a film drop flow l 11 3 regime, GE-test analysis. .)
| |
| 14 DR. ZUBER: In your plate, I mean how can you have j 15 ~ a -- you establish a flow regime in a half-an-inch-thick
| |
| '4
| |
| '16 plate, when data show some holes are' dry, some holes are 17 -full?
| |
| 18 MR. TAKEUCHI: In the computation, we do not model 19- .the microscopic points like a plate, but we just model in a
| |
| : 20. .more macroscopic' fashion.
| |
| '2 15 DR. ZUBER: Can you answer the question? How can 22: you have:these flow regimes in such a small plate, use two 23 : fluid models and interfacial coefficients to apply them to f'
| |
| 24' this case here?
| |
| 25 MR. TAKEUCHI: On this combination of the flow L
| |
| l ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| , 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036' (202) 842-0034
| |
| | |
| _. - .. . . .. . . ~ , . .. - . .- - . ..
| |
| 153 i 1 .areacand they hydric diameter.
| |
| : i. .
| |
| L
| |
| [~h; 2! DR..WALLIS: Well, your prediction that the flow V
| |
| 3 --
| |
| I..can't really:see what the nodes are, but across that 4_ ' orifice at the top, you have to predict.the flow rate )
| |
| ; 5: somehow.. Does the. code' predict' flooding there, at that i
| |
| -([ point? )
| |
| l 7 So, it's whatever models that orifice in terms of i
| |
| 8= flow rates through that whole, is predicting the flooding?
| |
| ~
| |
| 9 Is that the' case? ,
| |
| -10 MR. TAKEUCHI: I think so. The momentum is placed
| |
| .11 'at this point.
| |
| 12 DR. WALLIS: I think we only get the answer to 13 Novak's question by reading a lot of detail from somewhere l l
| |
| I 14, else.
| |
| ll 15 DR. ZUBER: Their approach cannot be applied to 16- 'this situation, and number two, the effect of force -- the 1
| |
| 17 shape of the. holes doesn't even come into the two fluid ;
| |
| 18 'model. There is no way you can make a case that this is a i 19 sound approach.
| |
| i
| |
| .20- DR. WALLIS: Shall we move on?
| |
| L 21 MR. TAKEUCHI: This is the comparison of predicted i.
| |
| 22 peak clad temperature and test data. Here is the test data.
| |
| 23 DR. WALLIS: Whose test data is this?
| |
| l .
| |
| -24 MR. TAKEUCHI: That's CCPF peak clad temperature 25- of run 72, and this one is the peak clad temperature for run L.
| |
| i ANN RILEY & ASSOCIATES, LTD.
| |
| p/
| |
| g,j Court Reporters
| |
| , , 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 j.
| |
| I
| |
| - ~.
| |
| | |
| . ..-. . . _ _ _ . . .. _ .. - . . . . _ . . ._._m. .
| |
| I i
| |
| 154 '
| |
| 1 76, single failure UPI. case, and test data are here,.and on
| |
| ~
| |
| L 2 .the right-hand side, this is the comparison of quench front 3 data against predictions. Predictions are' indicated by i 54 solid curve here.
| |
| 5 This is the void fraction on~ test 72 in the upper 6 plenum and lower plenum. ' Predictions are solid curves, and 7 test data are squares. This is the upper plenum, those two l
| |
| 8 are upper plenum, and'those two are in the bundle.
| |
| 9 DR. WALLIS: So, the prediction in figure 30.8 l 10 does not look good. The data are along the bottom axis.
| |
| I 11 MR. TAKEUCHI: This one doesn't look good, but I'm 12 wondering why the upper plenum is flooded so much, as data 1
| |
| 13 indicate.
| |
| 14 DR. ZUBER: Bacause you don't calculate it
| |
| / 15 correctly. That's very simple. I k.lY
| |
| '16 MR. TAKEUCHI: However, if you look at'this case 17 and this case, this case has twice as much UPI water 18- injected. So, by comparing those two curves, this one has 19 more water in the upper plenum than this case. So, they are i
| |
| 20 consistent.
| |
| [ 21 And-the comparison of a prediction and the test 22- data in the bundle region -- we always compute less water in 23 the bundle for this case and that case.
| |
| 24 Now, this is the flooding prediction against l 25 Bankoff sub-cooled CCFL applied to the low power region of E
| |
| j,
| |
| !./~s . ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 L (202) 842-0034
| |
| | |
| _ . _ . , . , , . . _ _ . . _ . . _ . _ . ....m- _ - . _ . . - _ _ _ . _ _ . _ _ _ . . _ . . - - . _ . _ .
| |
| 155 1: the CCPL model. .The prediction is~well behaved.
| |
| /'~S 12 DR. WALLISi Why are'there so many predictiona and 1
| |
| ~ '
| |
| 3 why do'they scatter so much? Why is'not there a smooth 4- curve? '
| |
| 5 .DR.:SCHROCK: I can't make out what it is you're 6 plotting there. It looks like the square root of kilograms,
| |
| .7- MR. TAKEUCHI: That's the number for steam.
| |
| .8 DR. SCHROCK- Okay.
| |
| 9 DR. WALLIS: It's' peculiar that your prediction 10 has such a tremendous scatter.
| |
| 11 MR. TAKEUCHI: I think some of them -- this is a 12 transient situation, and some of them don't reach the 13 flooding limit.
| |
| 14 MR. NISSLEY: This is a transient that starts out
| |
| ( 15. with full steam, the rods are. heating up. We're simulating 16' portions of that. Then the ECC injection comes on, there is 17 decreasing decay heat being simulated in here, so it is a
| |
| -18 very transient-test.
| |
| 19 You are moving from a situation of steam only, 20 then you start getting the injection and you go into 21' flooding, and you do have some breakdown and some down-flow, 22 as well.
| |
| 23 So, we're comparing -- each one of those dots 24 represents a point in the transient. The points that are on 25 the X and Y axes are corresponding to either no injection or ANN RILEY & ASSOCIATES, LTD.
| |
| ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . _ . ._~m . _ . . - . . _ . _ _ . _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ . _ _
| |
| 4 156 1 .a fully flooded case.
| |
| -, 2 DR. ZUBER: If I look at'the graph on the left, I
| |
| ,A-33, said,-well, most of them are just around point two, with the 4 exception of three or four points on top of it.
| |
| 5 ,
| |
| M R .~ .T A K E U C H I : In this kind of a transient 6f situation, as long as the' predictions don't go above the 7: CCFL limit, we are satisfied with.it. It is not designed
| |
| .s 8 for~the CCFL test.itself. It's a simulation'of the real
| |
| , 9 transient.
| |
| 10 'The prediction'did not go over the CCF line limit.
| |
| 11 If it go over, we have to start all over again.
| |
| : 12. DR. SCHROCK: Are we talking about GE tests here?
| |
| 13 MR. TAKEUCHI: No, this one is a CCFL test.
| |
| : 14. DR.'SCHROCK: What was their control variable?
| |
| L =15 MR. TAKEUCHI: Control variable here is 16 double-ended guillotine break.
| |
| e L17 DR. SCHROCK: Isn't it a separate effects test?
| |
| 18 MR. TAKEUCHI: No, this one is integrated effect L 19 test. Therefore, as long--- the points may scatter all 20 around, but as long as it doesn't step over beyond the CCFL 21- --
| |
| 22' DR. SCHROCK: But a point on this curve implies 23 that you have a determination of each of those variables..
| |
| -24 MR. TAKEUCHI: Yes.
| |
| I l- 25 DR. SCHROCK: The flux-of liquid, the flux of l
| |
| 4 ..
| |
| 1,q ANN-RILEY & ASSOCIATES, LTD.
| |
| l g ,j. Court Reporters
| |
| .1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
| |
| .;-- (202) 842-0034 p
| |
| li - . _ _ -
| |
| | |
| J - - o + = * -
| |
| i l
| |
| 157 I l
| |
| 1 vapor. I fv 2 MR. TAKEUCHI: Yes, 3 DR. SCHROCK: And you get those both from 4 calculations.
| |
| 5 MR. TAKEUCHI: Both from the calculations.
| |
| 6 DR. SCHROCK: If they don't lie in the right !
| |
| l 7 domain, what does it mean to you? If you get a point above, 8 what does that say to you about your calculation?
| |
| 9 MR. TAKEUCHI: This region is forbidden.
| |
| 10 DR. SCHROCK: Well, you've got a lot of points up 11 there.
| |
| 12 MR. TAKEUCHI: There's no points.
| |
| 13 DR. SCHROCK: Oh, I'm looking at the dashed line.
| |
| 14 MR. NISSLEY: That's the saturated curve.
| |
| ,m
| |
| (
| |
| ) 15 MR. TAKEUCHI: It should not step over this line.
| |
| 16 Some cool water was injected on top of upper core plate so 17 that this is the limiting line. It should not go beyond 18 this one. If the saturated water is the UPI, then it should l 19 not go beyond this line.
| |
| 20 DR. SCHROCK: Well, there are no longer any 21- steady-state possibilities once you get out here to the end 22 of this line. That's why you terminate it, sub-cooled 23 . flooding line?
| |
| 24 MR. TAKEUCHI: The reason is if I use this formula 25 towards the other side, then it should go something like I''\ ANN RILEY & ASSOCIATES, LTD.
| |
| (_,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 i
| |
| (202) 842-0034
| |
| | |
| 1 1
| |
| 158 1 this. This is the formula here.
| |
| (''h 2 MR. NISSLEY: That's why we terminate it.
| |
| %~sl 3 MR. TAKEUCHI: That's why it's terminated. The 4 formula becomes K(g) minus some number. 4 l
| |
| 5 DR. WALLIS: You have to take the negative route, 6 though.
| |
| 7 MR. TAKEUCHI: The flow direction was 1 8 counter-current, but it doesn't really mean too much. If I 9 use the notation like X and Y, this one becomes Y square 10 minus some number, one square, plus X, square root of it, is 11 equal to 2.0.
| |
| '12 DR. WALLIS: I think we should move on. I'm not 13 sure we can learn more from this figure.
| |
| 14 MR. TAKEUCHI: Yes. l
| |
| \
| |
| (~'% '
| |
| 15 Then we are going to the analysis the UPTF tests.
| |
| %.)
| |
| 16 In the case of the UPTF tests, there is no heated 17 core. The steam and water is sprayed from the bottom of t-he 18 CCFL regions, and we are interested in the flow dynamiis in 19 this upper plenum.
| |
| 20 DR. SCHROCK: I guess one final comment on those 21 last two figures -- doesn't it appear that the bulk of the 22 experimental data --
| |
| 23 MR. TAKEUCHI: I use a correlation as experimental 24 data, Bankoff correlation as an experimental data.
| |
| 25 DR. SCHROCK: I thought the solid line is l
| |
| l
| |
| /"S ANN RILEY & ASSOCIATES, LTD.
| |
| (_) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034
| |
| | |
| 159 1 Bankoff's correlation.
| |
| ! (~') 2 MR. TAKEUCHI: Solid line, yes.
| |
| V 3 DR. SCHROCK: And the predictions are COBRA / TRAC.
| |
| 4 MR. TAKEUCHI: Prediction expressed by square are 5 the COBRA / TRAC.
| |
| l 6 MR. NISSLEY: The dashed line is the saturated 7 limit. The hand-sketched line is allowing for sub-cool.
| |
| 8 DR. WALLIS: What haven't looked at the detailed 9 COBRA / TRAC interfacial shear and all that sort of thing, 10 have we? We don't have time to do that. Certainly, in the 11 AP600, the interfacial shear is made so that it gives the 12 right flooding curve. There's a feedback there. You don't 13 have that in COBRA / TRAC, do you?
| |
| 14 MR. NISSLEY: Can you repeat that?
| |
| 15 DR. WALLIS:
| |
| [v} That's different from what we saw in 16 the Westinghouse AP600 code, where the interfacial shear 17 correlations are adjusted so that they give the right 18 plotting curve.
| |
| 19 MR. NISSLEY: What curve is that? ;
| |
| 20 DR. WALLIS: Whatever is used in AP600.
| |
| 21 MR. TAKEUCHI: COBRA / TRAC is used for large break.
| |
| 22 DR. WALLIS: It's NOTRUMP, 23 MR. TAKEUCHI: Oh, that's a different code.
| |
| l l
| |
| 24 DR. WALLIS: You don't have anything like that, j 25 MR. TAKEUCHI: In the case of NOTRUMP, the I i i 1
| |
| r^N ANN RILEY & ASSOCIATES, LTD. !
| |
| ('~) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| )
| |
| l l Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| L 160 1 correlation is the same as the RELAP code, but we have more
| |
| - '('}
| |
| i (j
| |
| 2 basic correlation, 3 DR. ZUBER: You cannot model with a two fluid plot 1 4 and your interfacial package to model this process on the 5 tie plates. I think that's absolutely nonsense.
| |
| 6 What you could do and what is done by other codes 7 and other people, use a correlation, whatever it is, which 8 has a good database, and use this as the boundary 9 conditions, and embed these correlations in the code.
| |
| 10 When you come to CCFL, then you use the 11 correlations, and then you have correlations based on 12 experimental data you can trust.
| |
| 13 There is no physical ground to use this model and 14 interfacial package to this geometry, period.
| |
| r^N 15 MR. TAKEUCHI:
| |
| V) You might not believe it. However, 16 we have a unified way to model the flooding, and flooding 17 results -- the COBRA / TRAC always satisfying flooding 18 conditions for all geometries. One-dimensional pipe is 19 modeled with three-dimensional pipe. We get flooding 20 correlation.
| |
| 21 DR. ZUBER: There's a difference modeling in a 22 pipe and modeling on a tie plate. In a pipe, you have flow 23 regimes.
| |
| 24 MR. TAKEUCHI: And the correlations are also 25 different. Still, we get satisfactory results. That's N ANN RILEY & ASSOCIATES, LTD.
| |
| ! k'')
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D . C '. 20036 (202) 842-0034
| |
| | |
| 161 I 1 something --
| |
| L/~'N 2 DR '. SCHROCK: One more clarification.
| |
| \Q
| |
| '3 MR. TAKEUCHI: Yes.
| |
| 4 DR. SCHROCK: What are you calculating this for?
| |
| '5 What geometry?
| |
| 6' MR. TAKEUCHI: That's a perforated plate.
| |
| 7 DR. SCHROCK: You're calculating this for a
| |
| .8 perforated plate.
| |
| 9 MR. TAKEUCHI: Yes.
| |
| '10 DR. SCHROCK: Your code treats a perforated plate 11 as a single' hole?
| |
| 12 MR. TAKEUCHI: Yes.
| |
| l 13 DR. SCHROCK: So, it prohibits the process of i 14 counter-current flow through the mechanism of some holes
| |
| () 15 16 flowing full of liquid and other holes flowing no liquid?
| |
| If you look at the sub-cooled flooding line and 17 then you look at all of the points that you show there, that 1
| |
| 18 you say you calculate, kind of first impression is that the 19 calculations are done for flow conditions that are not 20 flooded, okay?
| |
| 21 But then you have to ask how could that possibly
| |
| '22 be, because you would expect, if the liquid flow is the 23' independent variable and you look at what you get in the way I 24 of steam up-flow to produce the flooding limit, you'd have a l
| |
| 25 family of curves that would be asymptotic to this flooding l
| |
| l
| |
| ;f''/
| |
| (,
| |
| g ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l' (202) 842-0034 l
| |
| | |
| i 162 1' line. They can't arrange themselves in any such manner.
| |
| i 2- So, what could they mean?
| |
| ' {']t s-3 MR. TAKEUCHI: If the same model of CCPF is 74 . subjected to the same condition as GE, CCFL, idealized
| |
| : l. ,
| |
| ! 5 : conditions --
| |
| l 6 DR. SCHROCK: You're relating this to GE?
| |
| MR. TAKEUCHI: Yes.
| |
| l 7
| |
| 8 DR. SCHROCK: How are you relating this to GE?
| |
| 9 You just told me perforated plate.
| |
| 10 MR. TAKEUCHI: GE is a perforated plata. yes, 11 CCFL, sub-cooled CCFL on the perforated plate.
| |
| 12 DR. SCHROCK: I thought you showed us a single 13 orifice for GE experiments.
| |
| 14 MR. TAKEUCHI: But that's a model of perforated
| |
| ) 15 plate.
| |
| 16 DR. SCHROCK: Your model of a GE perforated plate
| |
| -17 experiment.
| |
| 18 MR. TAKEUCHI: Yes.
| |
| 19 DR. SCHROCK: Oh , that's deceptive.
| |
| 1
| |
| :20 MR. TAKEUCHI: No, it is not deceptive.
| |
| 21 DR. SCHROCK: Well, it is, because --
| |
| ~22 MR. NISSLEY: It's a nodalization.
| |
| 23 DR. SCHROCK: Label it as such, okay? Don't say 24 here's the geometry of the system we're analyzing.
| |
| 25- DR WALLIS: Can we get back to COBRA / TRAC and the b
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| l .(,,
| |
| /''') Court Reporters 4 1025 Connecticut Avenue, NW, Suite 1014 I
| |
| . Washington, D.C. 20036 (202) 842-0034
| |
| | |
| I 163 1
| |
| 1 orifice? COBRA / TRAC has some equations in it and so on for
| |
| <~~
| |
| ; i 2 pipes and interfacial friction and all that. I don't 1
| |
| \s l l 1
| |
| i 3 understand how it models an orifice. I
| |
| ( l 4 An orifice has no thickness. I don't know what I
| |
| 5 you use for your friction. COBRA / TRAC doesn't model nodes I i
| |
| 6 of no thickness. I 7 MR. TAKEUCHI: Can I show you how we model those l
| |
| l l
| |
| 8 perforated plates? '
| |
| i 9 DR. WALLIS: Yes, may you can. I 10 MR. TAKEUCHI: Here there is a physical perforated l
| |
| 11 plate.
| |
| i 12 DR. WALLIS: Yes. I 13 MR. TAKEUCHI: And let me say this is the bottom 14 boundary of the CCFL region.
| |
| i /~~% i
| |
| ' i
| |
| %d
| |
| ) 15 DR. ZUBER: What is the dotted line below? '
| |
| 16 MR. TAKEUCHI: This is sort of the bottom section.
| |
| 17 DR. WALLIS: What do you do about the length? )
| |
| 18 DR. ZUBER: How do you determine the flow regime?
| |
| j 19 Because your interfacial drag depends on the flow regime, j 20 MR. TAKEUCHI: Okay. Just a moment.
| |
| 21 Then we make an imaginary jet channel.
| |
| 22 DR. WALLIS: Like a pipe.
| |
| 23 MR. TAKEUCHI: Like a pipe, with the same 24 configuration.
| |
| 25 DR. WALLIS: How long is it?
| |
| l
| |
| [~ \ ANN RILEY & ASSOCIATES, LTD.
| |
| (. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 j Washington, D.C. 20036 (202) 842-0034 i
| |
| | |
| i l L i 164' l l' MR. TAKEUCHI: We don't have a real reason for how 7
| |
| 2 much', but we have this one, and then this elevation is going i 3 to.be sliced into cells.
| |
| l 4 DR. ZUBER: Okay. !
| |
| [ 5 MR. TAKEUCHI: That's the model.
| |
| L 6 'The' void fraction is determined from this side of l 7 :the cell for the liquid. .For the void, it's determined from 8 this side. The vapor void fraction is going to determine
| |
| : 9. the flow regime.
| |
| 10 DR. ZUBER: From the flow regime, you determine 11 the drag. l i 12 .4R . TAKEUCHI: That's right.
| |
| l 13, DR..WALLIS: The drag has to be evaluated over 1.
| |
| L 14 some length of pipe, I think.
| |
| ;fe m .
| |
| lt 15 DR. ZUBER: This is how they start this artificial 16 model. It has no physics in it. It's wrong, t
| |
| 17 MR. TAKEUC'IA 'n the real world, there is a jet
| |
| : .18 'of steam going.up.
| |
| l 19 DR. ZUBER: Other codes have the same problem, and 20 they can use their correlations as boundary conditions. At L 21 least there is a database -- experimental database to L 22 support it. There is no really data you can really vary for L 23 lany of.these details.
| |
| l 24 DR. WALLIS: It appears that, by some method,
| |
| -25 COBRA / TRAC is adapted to be able to model orifices in some d
| |
| ANN-RILEY & ASSOCIATES, LTD.
| |
| :;xji
| |
| {~'V . Court Reporters
| |
| ;- 1025 Connecticut Avenue, NW, Suite 1014
| |
| : Washington, D.C. 20036 (202) 842-0034 l
| |
| l l~ , - .
| |
| | |
| 165 31' way,.as an effective pipe or something.
| |
| ["'h 2' MR. NISSLEY: By modeling'this.with what we call a '
| |
| ,\s ,/' _
| |
| l' l3, jet channel,.which'is an. extension of the geometry of the I
| |
| 4- plate,'our comparisons --
| |
| 5' .DR. WALLIS: And we-can' find that in your '
| |
| : 6. 'documentati'no somewhere?_
| |
| 7 MR. NISSLEY: Yes.
| |
| J 8 MR.LTAKEUCHI: You can think about.the steam jet E '9 going:through as if there is a pipe existing, when the flow l '
| |
| 10 ratefis large.
| |
| ; - 11' DR. SCHROCK: Does your model include condensation 12 on that interface?
| |
| t 13 MR. TAKEUCHI: Yes. Interfacial condensation is i 14 in there.
| |
| (m- ) 15 DR. ZUBER: On what areafdo you calculate your
| |
| -16 condensation?
| |
| 17 MR. TAKEUCHI: What area?'
| |
| 18 DR. ZUBER: On what area do you calculate your '
| |
| 19 condensation?
| |
| 20 MR; TAKEUCHI: Annular flow area.
| |
| 21 DR. WALLIS: This is something that you developed 2 2 '. .for the upper plenum injection problem?
| |
| 23 MR. TAKEUCHI: For all perforated plate, we use 24 this model. !
| |
| .25 DR.'WALLIS: Is this something that you're asking
| |
| '70m RILEY & ASSOCIATES, LTD.
| |
| (, Court Reporters j- 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . .-. - - . . . . - - . - . . -- - .----. . _ .. . . . ~ _ .
| |
| 166 1 the staff to approve?
| |
| (~') 2 MS. DEDEREV: This isn't really in -- even for v
| |
| .3 three- and four-loop -- we used this modeling in the UPI L4 . approach 10. years ago.
| |
| 5 DR. WALLIS: So, even if the experts don't like
| |
| -6 it, this is an approved model?
| |
| 7 b -
| |
| MS. DEDEREV: Well, we've been using this model.
| |
| 8 DR. ZUBER: This problem was not that important 9 for three- and four-loops, and if it was overlooked, it was
| |
| ]
| |
| '10 because it was not critical, there were other critical 11 ' things.
| |
| 12 Now, this problem is really the crux for this l
| |
| 13 plant, and if you did it without much arm-waving from the I 14 staff or from here, there'was a good reason.
| |
| () 15 Now you're trying to have this as a best estimate 16 methodology to approve. This is an important phenomenon, 17 and that's the reason we are asking these questions.
| |
| 18 DR. WALLIS: I think we have to move on, but the 19 only way we're going to resolve this is if we and our 20 consultants can actually look at the documentation.
| |
| 21 MR. NISSLEY: If I could make one last comment, 22 and then we'll move on.
| |
| 23 Our.first application of COBRA / TRAC was for upper L 24 plenum injection plants using the SECY 83-472 methodology, 25 which was in the late '80s, and this same modeling technique l t
| |
| I o
| |
| i
| |
| ./~ ANN RILEY & ASSOCIATES, LTD.
| |
| l k,,N) Court Reporters j 1025 Connecticut Avenue, NW, Suite 1014 y Washington, D.C. 20036 (202)~842-0034
| |
| | |
| 167 1 was used there; it was reviewed by'the staff and by ACRS at
| |
| (~'p 2- that time.
| |
| \/
| |
| 3 DR. WALLIS: But if we wish to dig into it again,
| |
| : 4. you can supply us with --
| |
| 1 5 MR. NISSLEY: Yes.
| |
| l 6 DR. WALLIS: Would you please point us to the '
| |
| 7 document page?
| |
| '8 MR. NISSLEY: Sure.
| |
| .9 MR. TAKEUCHI: Next is analysis of the NPTF test, 10- .and this square is the upper plenum we are interested in, 11 and the test. boundary condition is given by UPI water
| |
| - 12. ' injection rate from the top, and steam is coming through in- !
| |
| 13' one of the hot legs, and the core water rate is injected 14 from the bottom, and the same thing, steam is injected from
| |
| [V\ 15 16 the bottom, and those boundary conditions are show in this side of the table.
| |
| 17 The quantities we're looking at core leak
| |
| -18 down-flow rate, this one, and the hot leg leak flow rate and 19 the hot leg steam flow rate and the upper plenum 20 condensation rate, and test data for the phase one is shown 21 in this line, COBRA / TRAC computer results are shown in this 22 line. For another test, similar way, test data, COBRA / TRAC 23 calculation, test data, COBRA / TRAC calculation.
| |
| 24 DR. SCHROCK: What is the experiment?
| |
| '25 MR. TAKEUCHI: UPTF test, upper plenum test i
| |
| s ANN RILEY & ASSOCIATES, LTD.
| |
| 'L y,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 168 1 -facility.
| |
| (~'\' 2 DR. SCHROCK: It's not hard to write it on here.
| |
| ,O 3 MR. TAKEUCHI: A unique aspect of this data is to 4 apply the boundary condition to try to get the steady state 5 and see what's happening.
| |
| -6 DR. WALLIS: Let's see. What is actually measured 7 in the experiment? The condensation rate isn't measured.
| |
| 8- It has to be deduced.
| |
| 9 MR. TAKEUCHI: Condensation rate in the upper i 10 plenum is obtained by taking the difference of steam 11 injected and steam coming out.
| |
| 12 DR. WALLIS: All the other flow rates are 13- measured.
| |
| 14 MR. TAKEUCHI: All the other flow rates are
| |
| ,y ,)
| |
| ; 15 measured. You can see the core leak down-flow rate. The j 1
| |
| 16 predictions are always less than the test data.
| |
| 17 DR. SCHROCK: It's an integral measurement for the i
| |
| 18 whole core. It doesn't tell you where the down-flow occurs.
| |
| 19 MR. TAKEUCHI: It does not tell you 'here it will i I
| |
| 20 occur, no. It's a total amount is coming out. And if you 21 look at the hot leg leak flow rate and the hot leg vapor 22 flow rate, those are over-predicted. And if you compare the 23 upper plenum condensation rate, the predictions are just 24 right on the target.
| |
| 25 DR. WALLIS: Now, how should we read this? What's ANN RILEY & ASSOCIATES, LTD.
| |
| (''sq,,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| l 169 1 good and what's bad?.
| |
| I - [~ j . 2 MR. TAKEUCHI: Core' leak rate under-estimate is
| |
| %./ '
| |
| 3 the conservative side.
| |
| 4 .DR. WALLIS: This is a best estimate.
| |
| 5 MR. TAKEUCHI: I understand. Therefore, we are :
| |
| 6 . going to introduce uncertainty parameter and see what's 7 . going to happen.
| |
| 8 DR. WALLIS: The,most obvious thing is that the 9 hot leg liquid flow rate is not too well predicted. That's l
| |
| 10 presumably an entrained flow rate?
| |
| y 'll MR..TAKEUCHI: Yes.
| |
| 12 DR. WALLIS: And one might expect that to be not l 13 very well predicted, and I'm not sure how significant that 14 is. If that's significant, we might worry about it.
| |
| O
| |
| ( j 15 MR. NISSLEY: That will increase steam binding.
| |
| l 16 DR. WALLIS: Should we worry about that?
| |
| 17 MR. NISSLEY: Which would tend to be a, you know, 18 conservative direction, as well.
| |
| 19 DR. WALLIS: So, your strategy is to work to 20 improve the prediction or that or to say this is good 21 enough?
| |
| 22 MR. TAKEUCHI: We are trying to introduce 23- uncertainty multipliers, XY DRAG and XCONDU, and try to 24 cover -- improve the computed results.
| |
| l 25 The uncertainty parameter, we have a multiplier i
| |
| L 1 I
| |
| l
| |
| ~N ANN RILEY & ASSOCIATES, LTD. l
| |
| ! [A s) s Court Reporters
| |
| ( 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202} 842-0034 l
| |
| l L
| |
| | |
| 170 1 for the drag coefficient, XY DRAG and'XCONDU, and we have --
| |
| [''}
| |
| .2 in order to determine the range of the uncertainties, we 3 have done a study of GE CCFL test analysis, and the 4 combination of the values are shown by open circles in here, 5 and after all those tests with the study, what we found the 6 best situation is to go from XCONDU and XY DRAG equal 1.0 7 and make it XCONDU and XY DRAG simultaneously go down to 8 0.2, so from here down to here. That's the best 9 combination.
| |
| 10 DR. WALLIS: This means you're changing what you 11 thought was the drag coefficient by a factor of five?
| |
| 12 MR. TAKEUCHI: Yes.
| |
| 13 DR. WALLIS: So, the drag if 20 percent of what 14 you originally thought it was?
| |
| (q x-)
| |
| 15 MR. TAKEUCHI: Yes, something like that. And this 16 is leak drain rate against the steam injection rate for GE 17 CCFL test 20, with sub-cooled water injected, and test data 18 is the triangle.
| |
| 19 Now, if I apply --
| |
| 20 DR. WALLIS: I notice, in this case, some cases 21 here, the historesis is not predicted. For instance, the 22 one -- the second figure there -- there's no historesis in 23 the --
| |
| 24 MR. TAKEUCHI: This one?
| |
| 25 DR. WALLIS: The prediction -- the data doesn't.
| |
| i
| |
| ,[] ANN RILEY & ASSOCIATES, LTD.
| |
| (/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034 i
| |
| i
| |
| | |
| l 171 I
| |
| )
| |
| 1 MR. TAKEUCHI: Yes, prediction doesn't, but we i i(""T 2 have iustoresis, and the important thing we are looking for
| |
| '%-)
| |
| 3 -- this is the test data.
| |
| 4 Now, by going from XCONDU, XY DRAG equal to 1, the 5 prediction went over.
| |
| 6 DR. WALLIS: I think there's something wrong with 7 figure 24b-1. I 8 The historesis should be for steam increasing and !
| |
| 9 decreasing, and you've actually got -- I think something is j 10 wrong about your data and prediction symbols there to show 11 that there's a historesis. You should compare triangles )
| |
| 12 with triangles and squares with squares. Something is !
| |
| 13 wrong.
| |
| 14 MR. NISSLEY: There is no historesis in the data l I,r~s) 15 for --
| |
| v 16 DR. WALLIS: There's no historesis in the data? i 17 MR. NISSLEY: Correct. )
| |
| 18 DR. WALLIS: There was before, wasn't there?
| |
| 19 MR. NISSLEY: No, that was in the prediction.
| |
| 20 DR. WALLIS: Oh, okay.
| |
| 21 MR. TAKEUCHI: By varying the XCONDU and the XY 22 DRAG, we cover test data, and that's what we are looking 23 for.
| |
| 24 DR. WALLIS: Do you?
| |
| 25 MR. TAKEUCHI: It went over the triangle.
| |
| l l
| |
| !\(''i ANN RILEY & ASSOCIATES, LTD.
| |
| s ,/ Court Reporters i
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . . _ . ~. _ ._ _-_ _ . __ _, _ .m. _ . - . _ - _ _ . - . . _ .
| |
| 172 I 1 DR. WALLIS: The heavy squares?
| |
| T 2 MR. TAKEUCHI: Yes, heavy squares.
| |
| .-)
| |
| 3 DR. WALLIS: But the light squares don't.
| |
| 4 MR. TAKEUCHI: Light square did not.
| |
| 5- DR. SCHROCK: What in your model depends upon a J 6 change in the steam flow rate?
| |
| 7 MR. TAKEUCHI: Change in the steam flow rate. l i
| |
| 8 That's flow boundary conditions. We can change the flow l I
| |
| 9 rate.
| |
| 10 DR. SCHROCK: Each one of these is kind of a 11 ' snapshot, quasi-steady --
| |
| 12 MR. TAKEUCHI: Yes.
| |
| 13 DR. SCHROCK: -- is the spirit of your 14 presentation.
| |
| () 15 16 MR. TAKEUCHI: Yes.
| |
| DR. SCHROCK: But your model that you have 17 computerized is predicting when the steam flow rate is 18 increasing, it does one thing with your model, and when it's '
| |
| 19 decreasing, it does something else.
| |
| 20 MR. TAKEUCHI: Yes.
| |
| 21 DR. SCHROCK: My question is, in what way, then, 22 is the fact of steam increasing or decreasing enter into 23 what your model does?
| |
| 24 MR. NISSLEY: In the PWR calculation?
| |
| DR. SCHROCK: In any calculation?
| |
| l~
| |
| l~
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| l -(_
| |
| ,Q) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
| |
| , (202) 842-0034
| |
| | |
| 173 1 MR. NISSLEY: In the PWR calculation, when the UPI lC / '} 2 water comes in and you start refilling the core, you will go 3 from a fairly stagnant condition into increasing steam as 4 you're quenching more and more of the rods, so that the 5 steam generation rate would have a general trend of 6 increasing-until you've quenched a large portion of the 7 core, at which point it would start to decrease.
| |
| 8 DR. SCHROCK: I don't know how that relates to the 9 question that I asked.
| |
| 10 You have, then, this figure 24b-23, experimental 1
| |
| 11 data that show no historesis. You have predictions that 12 show a historesis. And what I asked is, in the physical 13 modeling that your code is doing, how does it depend upon 14 whether there is a ramp in the steam flow one way or the 7 ~.
| |
| r1
| |
| ) 15 other?
| |
| s_/
| |
| 16 MR. NISSLEY: Most likely -- and that's the best 17 answer I can give you -- the formation of a pool above the 18 plate during the period of flooding, when you're going up, 19 following the data, up in here, you're creating a pool of 20 sub-cooled water, and then, when you come back down, that is 21 having an effect.
| |
| 22 DR. ZUBER: Don't you think this is influenced by 23 how many nodes in that artificial stack you are making?
| |
| 24 MR. NISSLEY: I would think it would be.
| |
| 25 DR. ZUBER: And therefore, really, it may not have f
| |
| l
| |
| /~}
| |
| (s,/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenuc, NW, Suite 1014 l Washington, D.C. 20036
| |
| ! (202) 842-0034 l'
| |
| l l
| |
| | |
| 174
| |
| -1. any' physical meaning, because all your condensation, if this
| |
| )
| |
| '2 is used for condensation, depends on how many cells are you 3 going to take in that artificial pipe.
| |
| 4 DR. WALLIS: Maybe, when we see this again, when 5 -we see it next time around, you will have already looked at 6 whether or not.there's a pool and you'll have an answer to 7 why there's a historesis. I don't think we'll get the 8 -answer now.
| |
| 9 DR. SCHROCK: No, I don't think we will either, 10' -but I'd point out that these experimental data are taken in a system in which there will inevitably be variations in
| |
| .12 that' pool. But these experimental data show no historesis.
| |
| 13 So, there's something about the model which is not adequate 14 to represent the data in this particular geometry.
| |
| () 15' MR. TAKEUCHI: This one -- this is still UPTF 16 tests ---CCTF, flooding prediction or a flooding situation 17 on the CCTF at the low-power core plate.
| |
| 18 This is the base case, and this is the case of 19 XCONDU and XY is reduced to .2, and you can see that, in 20 this prediction, the prediction is all confined within the 21- . flooding limit, but changing the -- applying the multiplier
| |
| : 22. and reducing it to the .2, it can go over the flooding 23 curve.
| |
| 24 That's the purpose of varying the uncertainty 25 parameter.
| |
| A)
| |
| (- ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 175 1 And this is the effect of the ranging of those XY
| |
| ! [' b 2 DRAG and XCONDU'on the UPTF test predictions. Test data are U
| |
| : 3. shown by squares for the leak drain rate and vapor up-flow 4 rate and hot leg leak flow rate and hot leg vapor flow rate, 5 and test phases A, B, C range from right to left for each 6 1 quantity.
| |
| '7 And approximately 50 percent of the data are 8 . bracketed by varying XCONDU and XY DRAG.
| |
| 9 That's the situation.
| |
| * 10 And if we go to the' scaling effeJts -- I think I 11 have shown you this one before. In the 2-D, 3-D programs, 12 the NPR associates arranged the test data to produce certain 13 scaling trends of UPI parameters like brehk-through flow 14 area.
| |
| ( ,/ 15 DR. WALLIS: What is this break-through area ;
| |
| 16 again? ;
| |
| l 17 MR. TAKEUCHI: Break-through area is upper core i
| |
| 18 plate of this size, and only this portion -- in only this 19 portion water can get into. This is the definition of 20 break-through area.
| |
| 21 DR. WALLIS: Is the fudge factor on the flow area?
| |
| 22 MR. TAKEUCHI: Fudge factor --
| |
| 23 DR. WALLIS: It's a discharge coefficient or 24 something?
| |
| 25> MR. TAKEUCHI: No , this is not. The UPI was L
| |
| l ANN RILEY & ASSOCIATES, LTD.
| |
| t/
| |
| 's Court Reporters 1025 Connecticut Avenue, IM, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 176 l l
| |
| 1L injected to the upper plenum, and within the upper core 1 1
| |
| ['}
| |
| x _-
| |
| 2 plate, some portion of UPI water can penetrate, some part of 3 the core it does not.
| |
| 4' DR. WALLIS: This depends upon scale?
| |
| 5 MR. TAKEUCHI: That depends on scale. That's just 6 what Dr. Zuber has been talking about. If you make a test 7 scale of this size, then water can penetrate everywhere, but 8 when it becomes a larger test facility, it can go only 9 smaller portion of the area where the water can penetrate 10 through into the core.
| |
| 11 DR. WALLIS: Sounds very peculiar.
| |
| 12 MR. TAKEUCHI: Maybe because the steam can get in 13 from the other side.
| |
| 14 DR. ZUBER: It's really like in a down-comer, you t%
| |
| () 15 know. You may have a completely solid liquid coming in one 16 place.
| |
| 17 MR. TAKEUCHI: And the other side is just steam 18 going up.
| |
| 19 And the test data are shown by open circles, open 20 squares, and crosses in the square, and so forth.
| |
| 21 DR. WALLIS: Where does this break-through aren 22 come from? It comes from calculating a flow rate?
| |
| 23 MR. TAKEUCHI: If we look at -- we have a model.
| |
| 24 DR. WALLIS: Break-through area is --
| |
| 25 MR. TAKEUCHI: In some areas, it goes through. In
| |
| [3 ANN RILEY & ASSOCIATES, LTD.
| |
| (_) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| ., - . . . . ~ .- . . , . - - , - . .
| |
| 177 l I some areas, this is not -- it's zero or a positive, l
| |
| /'''Y 2 DR. WALLIS: You have to specify break-through
| |
| .V.
| |
| 3 area in your model? l 4 MR. TAKEUCHI: No. It's come out just as a result of computation.
| |
| 6 DR. ZUBER: Wait, wait, wait. 'Let me help you, 7 because you're not-citing your things correctly.
| |
| ~
| |
| l
| |
| -8 MR. TAKEUCHI: Oh.
| |
| 9 DR. ZUBER: The break-through area comes from 10 experiments. This was done at UPTF, and they measured,-and 11- this is what -- and this is how the fraction of area came.
| |
| 12 Now, how you calculate is a different case, but 13 that was the experimental basis for this break-through area.
| |
| 14 DR.-WALLIS: I see there's a Dartmouth point {
| |
| () 15 there. I don't know that break-through area was a concept 16 that was ever heard of at Dartmouth.
| |
| 17 DR. ZUBER: This is the UPTF experiments really 18 showed. You have area where just the liquid goes down and 19- the vapor completely bypasses and goes to other holes, and 20 by measuring where the water was collected, NPR came with 21 this percentage area, and those are experimental data.
| |
| 22 MR. TAKEUCHI: That's right, yes.
| |
| 23 DR. ZUBER: So, you don't have to argue. Those 24 are experiments. I'm trying to be helpful sometimes.
| |
| 25 MR. TAKEUCHI: Thank you very much. I appreciate l
| |
| l l
| |
| l
| |
| !/ ANN RILEY & ASSOCIATES, LTD.
| |
| l( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034 l-
| |
| | |
| i l
| |
| 178 !
| |
| 1 that very much.
| |
| (~N 2 DR. SCHROCK: Presumably, there's consistency in
| |
| -3 the definition'among all these experiments? I mean you show l 4 .UPTF, CCTF, SCTF, Oak Ridge, Dartmouth, Sami-Scale, all 5 .those, the definition of break-through flow area is the l 6 same.
| |
| 7 MR. NISSLEY: It's was NPR's application of their 8 definition to all those tests.
| |
| 9 MR. TAKEUCHI: And the predictions are indicated l
| |
| 10 by a' closed circle and closed square, closed triangles, and
| |
| .11 we can reproduce this kind of trend.
| |
| 12 And this one is down-flow rate as a percentage of 13 available water in the upper plenum. In this case, l 114 prediction increases with increasing scale. I'm sorry, data
| |
| ( ) 15 increases, and'also prediction increases with scale, 16 increasing scale, i
| |
| 17 The next one --
| |
| 18 DR. FONTANA: How do the predictions?
| |
| l 19 MR. TAKEUCHI: In the case of UPTF, it's easy to
| |
| .20 say, show to you as an example, because it's a steady-state 21 calculation. You have a down-flow rate.
| |
| [ 22 DR. FONTANA: Okay. Go ahead.
| |
| 23 DR. WALLIS: It must depend on all kinds of 24 conditions. It's not just a fraction of certain percentage 25 of water always gets down. It must depend on all sorts of i
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| i O) -
| |
| 1,_
| |
| Court Reporters 1025-Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i , . .. _
| |
| | |
| _.....__..._.____y._.____._._ - - _ . . . - . . _ _ . _ - _ _ _ _ . . _ > -... .
| |
| I 179
| |
| : 1. other things.
| |
| ~(~Nc 2 MR. TAKEUCHI: In the case of steady-state'UPTF,-
| |
| k !.
| |
| 3; !
| |
| it is fairly straightforward, j 4 DR. WALLIS: What's being kept constant? Steam 5 velocity or the pressure or what? There's all kinds of 6- things changing.
| |
| 7: MR .' TAKEUCHI: All those kinds of things change 8 during that transient, but -- how can I explain that? The 9 UPTF test is:a steady-state. Those numbers -- how do I
| |
| -10 explain this one?
| |
| l
| |
| -11 DR. ZUBER: Those are the best tests you can use' i 12 for verifying. It's a large-scale, it was steady-state.
| |
| 13 -- MR. NISSLEY: Right. For the PWR and the CCTF, 14 - you had to pick a time average. )
| |
| I 15 MR. TAKEUCHI:
| |
| ) A certain-timeframe. he don't 16 ' include broader period, for instance.
| |
| 17 This one is the hot leg flow rate carry-over as a
| |
| .18 . percentage of available liquid, and we have predicted fairly.
| |
| 19 well the tests there, and the final one is a collapsed leak 20 in the upper plenum. Test data decreases with scaling, and 21 the prediction is made the same way.
| |
| 22 With all those results, I have come up with --
| |
| L 23 DR. SCHROCK: Where is this collapsed liquid L 24 ~ metal? In what node?
| |
| !~ 25 MR. TAKEUCHI: This one is a whole upper plenum l l:
| |
| IO i."
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 I
| |
| l Washington, D.C. 20036 l (202) 842-0034 l J
| |
| | |
| , - . . - - . . . - _ .~ ,. - -- . _- . -
| |
| l1 L t. 180 l
| |
| ~
| |
| .1 average.
| |
| ['\ '2 With those verifications in hand, we believe we've V
| |
| '3 got those conclusions, predicted previously with the L 4' analysis of LOCA transient of UPI PWR. ;
| |
| L 5 MR. NISSLEY: We've just distributed a two-page 6: hand-out.
| |
| 7- I'll give the first slide, which is our overall 8 summary for the meeting, as we came here today, and the 9 second slide will be given by Jeff Koss of Wisconsin 10 Electric regarding some information about their background h 11 for why they need this and what their schedule is.
| |
| 12 Our conclusions is that we started with the 13 previously reviewed and approved best estimate methodology 14 for three- and four-loop plants. We've extended that to
| |
| <( p) 15 . upper plenum injection plants using the CSAU format.
| |
| 16 We used the PIRT to identify highly-ranked 17 phenomena that needed additional assessment, and we have 18 used those assessments to identify physical models which 19 need to be ranged to account for the highly-ranked upper 20 plenum phenomena.
| |
| 21 That's our assessment of the code and the models 22 ard what needs to be ranged in the PWR calculations.
| |
| 23 For the PWh calculations, as we've shown earlier, 24 .the split break is limiting. We do use a deterministic 25 break flow area for that.
| |
| l I
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| l[~'Y i
| |
| s._) Court Reporters l- 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1
| |
| | |
| =
| |
| t i E l 181
| |
| ]
| |
| : 1. "We showed or we had'in'our package a demonstration that' ranging of condensation in the down-comer and lower 2
| |
| .3 plenum as we did for~the three- and four-loop plants is not 4 .necessary for UPI plants.
| |
| 5 CCFL in the PWR calculations prevents draining at l 6 all assembly locations except those on the core periphery, 7- where they have low-power assemblies, and the cooling of the 8 hot assembly is by bottom-up re-flood, as it was in the 9- three- and four-loop plants such that the heat transfer I 10 ranging approach used for three- and four-loop plants 11 remains applicable. ;
| |
| 12 Our overall uncertainty methodology for upper.
| |
| 13 plenum injection plants does remain intact, with the 114 exception being there are adjustments to which global ~models ;
| |
| A):
| |
| ( -15 are ranged in the. plant calculations.
| |
| 16 A side comment we had here which was not directly 17 related to this, but our experience to date indicates that 18 the use of the CSAU concept has facilitated training of 19: analysts within Westinghouse, although the learning curve 20- for these methodologies is quite extensive.
| |
| 21 By using the same systematic approach, one 22 . learning curve for any one of the methodologies does allow 23 you to understand the other methodologies, as well.
| |
| 24 If you've been trained on three- and four-loop 25 plants, understand the CSAU concept and its application, l.
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| O- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| - . . s -_ . . _ - . . . - - . ... . - ._ . _ - . _ _ _ - - . - - . . . . . . - . . .
| |
| 182 l-i 1 -when you'go over and work on an AP600 or a UPI plant, you ifA/ } 21 don't'have to start over on_that learning curve.
| |
| 3' At this point, I'd like to turn it over to Jeff L 4 Koss.
| |
| L 5 MR. KOSS: My name is Jeff Koss. I work for ;
| |
| 6 Wisconsin Electric, and I'm the fuel upgrade project lead, j 7 and currently, we have a .40 outside diameter fuel. pin with l-l 8 zircalloy-4 cladding and grids, and we're going to a.new l .. 9 ' fuel product here with a .'422 outside diameter, ziralo.
| |
| !~
| |
| '10 cladding and grids, and annular axial blankets.
| |
| l 11 The primary reason for this change is better fuel 12 design and the economics associated with the new fuel 13 design. . We also expect.more uranium load into each 14 assembly, therefore allowing us a reduction in'the-number of
| |
| ;-( ) 15' assemblies over the plant life. The approximate number 16 right now is estimated at about 110 reduction in feed "17 assemblies.
| |
| 18 So, we're trying to meet this schedule right here, 19 hopefully. This is the proposed schedule right now. If we 20 can get the generic topical SER approved by NRR in February 21 of 1999, we can begin preparation of the plant-specific
| |
| : 22. submittal in February.
| |
| 23 If Westinghouse can issue the proprietary approved
| |
| . 24 version of the generic topical in May, then we would submit
| |
| \.
| |
| L 25 in June of 1999 the tech spec change associated with the I
| |
| i lf''}'
| |
| ~\ _/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters L 1025 Connecticut Avenue, NW, Suite 1014
| |
| )
| |
| Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 183 1 fuel' upgrade'as well as the plant-specific BELOCA. ;
| |
| . 2 DR. ZUBER: I have a question. When did you start 3, working on this program or show interest in this 4: methodology?. Recently or six months ago or a year ago?
| |
| 5 MR. KOSS: I'd have to say, initially, it was 6' maybe four -- three, four years ago, is when we initially 7 started, and then we've since backed off.
| |
| 8 We had some problems with our units, and we were 9 working on those problems, so we kind of put this project on 10 hold for a while, while we addressed the problems with the 4
| |
| 11 plants and tried to get the units back on-line, and then we 12- started this project again within the past seven, eight 13 months.
| |
| I 14 DR. ZUBER: Together with Westinghouse. When was
| |
| ) 15 Westinghouse brought in?
| |
| 16 MR. KOSS: -Initially. I 17 DR. ZUBER: Initially. Okay.
| |
| 18 MR. KOSS: Initially, yes.
| |
| 19 DR. WALLIS: The link between upgrade and UPI 20 analysis is the best estimate analysis method is expected to 21 allow you to show that, with some sort of upgrade, you will 22 be able to still be okay or something which enables you to 23- get some margin you didn't have before?
| |
| 24 MR. KOSS: Well, we were getting very close to our 25 PCT unit for the large-break LOCA. So, we were planning to O ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l 184 l' -get some margin back with this analysis associated with the 2 fuel. upgrade.
| |
| 3 DR. ZUBER: How much money would it -- financial
| |
| '4- benefit'you will get iftyou_ apply this methodology?
| |
| l 5 MR. KOSS: In the margin? '
| |
| 6 DR. ZUBER: All.together. I mean if you implement l 7- it in your management and fuel and everything, what is the 8- financial benefit?
| |
| 9- MR. KOSS: I don't have numbers. Dollars?
| |
| 10 DR. ZUBER: Yes, bucks.
| |
| I 11 MR. KOSS: Do you have any idea?
| |
| 12 AUDIENCE: Tens of millions.
| |
| ~ 13- MR. KOSS: So, again, we would like to submit the l
| |
| 14 fuel upgrade tech spec change request in June of 1999 and I
| |
| ()
| |
| o 15 then start to manufacture-tubing only for the 422V+ fuel 16 ' rods in December of 1999, recognizing that, prior to
| |
| : 17 : expected NRC approval of the fuel upgrade tech spec change 18- request, but we're willing to take that risk.
| |
| '19 Not very many dollars associated with the tubing.
| |
| 20 The real cost comes in for the conversion to the 422V+
| |
| 21 assemblies, which we expect to begin in May of 2000 for a 22 re-load of the new assemblies in the fall of 2000, October 23 of 2000 for unit two refueling outage number 24.
| |
| 24 DR. WALLIS: Where does the ACRS fit into this 25 schedule?
| |
| l' l
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 w - ,* y -
| |
| - -+---g 9-
| |
| | |
| . . . . . . . - . - . - . _ ...~-..., . _ _ - . _ . - . - - - . - . - . . . . . . -
| |
| .l 185 l
| |
| -l' MR. KOSS: I'm sorry? l l
| |
| M
| |
| , 2 ~. 'DR. WALLIS: What. actions do you expect-from the 3: ACRS or does_somebody expect from the ACRS in connection l 4' with this schedule?
| |
| l- !
| |
| 5 MR. KOSS: .Well, I assume that we would need your l
| |
| 6 approval of-this' prior to issuing the generic topical SER in 7 February of 1999.
| |
| : 8. DR. WALLIS: It seems to me.you want something 9 before February,'because you expect -- or the staff needs 10 .something.before February, because you expect the staff to 11' approve-your --
| |
| ~
| |
| . 12 MR. KOSS: Yes.
| |
| 13 DR. ZUBER: When did you submit something to the
| |
| ?
| |
| =14 staff? When was the first submittal transmitted to the ~
| |
| 15 staff?
| |
| 16 MR. NISSLEY: August of 1995. We received RAIs in 17 . fourth quarter of '96.
| |
| 18 DR. ZUBER: Let me say just -- I commend the 19 industry for taking this road. It is rational and l
| |
| 20 beneficial, and you should be commended for doing it. ;
| |
| 21 DR. WALLIS: Do we have anymore questions for I 22 Westinghouse? i
| |
| ( 23 MR. KOSS: Thank you.
| |
| I 24 DR. WALLIS: I suggest we move on to.the staff l 25 presentation, which is not expected to take as long. I I
| |
| i
| |
| , - ANN RILEY & ASSOCIATES, LTD. I
| |
| : Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
| |
| .(202) 842-0034
| |
| | |
| L i
| |
| l 186 l' 1 think that the' natural place to break is after the staff 2 presentation. Then we go to the research.
| |
| MR. ORR: I'll try to make it.as short as 4 possible. ~It's nice to have it short anyway.
| |
| 5 Okay.
| |
| 1 6 I'm going to skip the first thing and whip that 7 'up. It's mostly background. It shows up here that this l 8 came in in August of '95. hit's been on and off reviewed for 9 quite a while, at high priority, when Haddam Neck became a 10- problem, lost a. lot of priority when Haddam Neck went down i
| |
| 11 the tobes.
| |
| .12 The three- and four-loop methodology was approved ,
| |
| 13 in '96, and we got the final version of it in March of '98.
| |
| 14 We also reported to ACRS on our findings regarding the 15 documentation.
| |
| (f It turns out the documentation we found 16 acceptable, and you have a copy over there.
| |
| 17- We finally got a report frota our contractor. This 18 was not a hands-on review, reviewed under contract, but we 19 did keep in touch and involved in the review process on 20 this.
| |
| 21 The final conclusions of our contractor was that 22 we accepted what was put forth by Westinghouse, that this 23 was, indeed, not a change to the existing approved 24- methodology but, rather, just a couple of changes within the 25 methodology.
| |
| l l
| |
| ' \CT ANN RILEY & ASSOCIATES, LTD.
| |
| f s) Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| 187 1 Most of the stuff that I've heard that was
| |
| [~' 2 controversial here today is stuff that's already been Q))
| |
| 3 considered in other reviews.
| |
| 4 Westinghouse use of this in either best estimate 5 of COBRA / TRAC and best estimate or quasi-best estimate 6 methodologies goes back 10 to 15 years, and most of the 7 things we've been talked about have been hashed over within ;
| |
| 8 that.
| |
| 9 In fact, the previous SECY methodology was 1 l
| |
| 10 originally intended to be a best estimate methodology, and 11 it eventually was approved under the SECY approach.
| |
| 12 It did incorporate a lot of best estimate 13 features, even though it was not a best estimate I 14 methodology, but a lot of this -- and that methodology was
| |
| [v ') 15 approved initially for two-loop UPI plants.
| |
| 16 So, a lot of the technology between not only the 17 three- and four-loop methodology but the earlier SECY 18 methodology has already been looked and found acceptable.
| |
| 19 Now, what we did in our review was a limited scope 20 review. It's limited only to the identified changes from 21 the three- and four-loop methodology, which I already 22 mentioned we didn't find to be substantial and didn't really 23 change the methodology at all.
| |
| 24 We also verified that the CSAU process, 25 uncertainty methodology, statistical methodologies were all l
| |
| (~T ANN RILEY & ASSOCIATES, LTD.
| |
| ( ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| ; Washington, D.C. 20036 f
| |
| (202) 842-0034 l
| |
| l
| |
| | |
| _ ._ m> ~_. _m ._.._.m..-. _ . _ . _ _ - . . . , . _ . . - , . . _ _ _ . _ . . . _ _
| |
| ._ _ _ y 188 4
| |
| 1 left pretty'much'.the same L
| |
| L 2 They.did exercise.the CSAU process in the same way 13' they did in the three- or four-loop methodology which we i 4' approved, and.then we broke it down to the few technical l'
| |
| 5 inches-that we've been controversial on here to see if there 6 was anything in there that we really did not think, in the l
| |
| ~7: regulatory arena, for licensing purposes, was acceptable or. !
| |
| 8 unacceptable.
| |
| '5L We found that their. approach in developing the 1 110 . methodology was consistent with CSAU and very much the same
| |
| ; 11 process that they used in the three- and four-loop p ~ 12 ' methodology.
| |
| 13 We found that the overall calculation process was l'4 -- -- remained intact, with only.the exchange of a couple of 15 -variables in.the global response surface, same Monte Carlo l
| |
| 16 structure, the'same. statistical basis, same correlation 17~ ' factor process. That was all there.
| |
| l 18 So, really, we felt.that, if the three- and 19 four-loop methodology was acceptable for licensing, so is 20- this.
| |
| . 21 Two items were removed from the global response 1 22 surface, and they're listed there, and there are reasons l
| |
| 23 why, and then we've been hashing over the things that '
| |
| 24t replaced them in that response surface, but again, the
| |
| '25 Lresponse surface itself had the same number of variables in E
| |
| l' L
| |
| l
| |
| ! /~% ANN RILEY & ASSOCIATES, LTD.
| |
| <(s,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 L
| |
| Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 189 1- it, the same statistical approach'was used, the same
| |
| .' 2 correction factors were used. It's the same methodology 3 that'you saw in the three- and four-loop methodology.
| |
| -4 DR.-WALLIS: So,'what I gather from this is that 5 the only thing that matters to you is whether or not the XY I
| |
| 6 DRAG ~and XCONDU were~ treated in a satisfactory way.
| |
| ]
| |
| 7 MR. ORR: Well, there were other considerations, 1
| |
| 8 but they're l'ess_than -- really, when you look at it -- I, I 9 actually, having been reviewing this for UPI plants for a 10 very long time now,-have my own bias, and you know, 11 sometimes, having been here a long_ time, I like to put 12 premium on longevity and the wisdom you get from it and 1 13 everything, that I have a bias to thinking that things like l
| |
| 14 entrainment and condensation in the upper plenum are very j
| |
| ) 15 important parameters for a UPI plant that t.ay or not not be 16, as important for a three- and four-loop plant, and it seems 17 like the focus of what was done in this model was done 18 appropriately.
| |
| 19 Now, granted, there are questions about whether 20 this is an actual -- the type of -- part of our review --
| |
| 21 our responsibility in the review is to look at the 22 identification of what is ranged and if it has been given 23 appropriate ranges, and we reviewed, and those concluded 24- that,.yes, they identified the right things to range, and we 25 , felt that, for licensing purposes, those ranges were
| |
| /
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| g Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 190 1 appropriate.
| |
| 2 Now, as far as best estimate is about having your p 3 m<ean right in the middle of the range or something like 4 that, that isn't necessarily best for licensing purposes.
| |
| L
| |
| ! 5 We like to assure that we have as close to i
| |
| L 6 realistic as possible but also'with assured conservatism to l
| |
| l 7 cover for the uncertainty that we've got.
| |
| 8 This is our schedule for completion.
| |
| l 9 DR. WALLIS: I want to go back to that.
| |
| 10 MR. ORR: Sure.
| |
| 11 DR. WALLIS: The only that, for your purposes, 12' matters is the fact that they have. removed some items and 13 these new items, interfacial drag and condensation, are new 14- items --
| |
| 'V
| |
| [l '15 MR.-ORR: Well, new-multipliers added to old
| |
| :16 correlations, yes.
| |
| 17 DR. WALLIS: -- in the licensing business?
| |
| 18 MR. ORR: Pardon? l 1
| |
| 19 DR. WALLIS: They're new items in terms of the 20 legalistic or licensing worldi 21 MR. ORR: Well,-they're new -- they're multipliers 22- on existing correlations.
| |
| 23 .DR. WALLIS: Which have not been approved before?
| |
| MR. ORR: Which had not been -- well, they had 25 been -- the multipliers themselves are where the ranging l'
| |
| . ANN RILEY & ASSOCIATES, LTD.
| |
| ! . (, Court Reporters
| |
| )
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 1 191 ]
| |
| 1 1 takes place, and we had not reviewed that ranging of those 2 multipliers, and so,~now, we've got that -- w'e reviewed the 3 appropriateness of those ranges and the ranging of those 4 multipliers, but the correlations themselves already existed 5 in the three- and four-loop model.
| |
| 6 DR. WALLIS: So, the question is what ranging are 7 you going to require?
| |
| 8 MR ORR: Well, the ranging that they proposed was
| |
| ~
| |
| 1 9 this ranging of 1.0 times the correlation down to.0.2 times '
| |
| 10 the correlation.
| |
| 11 Now, our review found that we agreed with their 12 conclusions that it was appropriate to range down to 0.2 l 13 mainly based upon their conclusion that, even at 0.2 times
| |
| -14 those multipliers, they still remain conservative. .They 15 were closer to realistic, but they were still on the
| |
| (
| |
| 16 conservative side.
| |
| 17 DR. SCHROCK: Wasn't the exclusion of UPI plants 18 from the improved methodology, 1996 --
| |
| 19 MR. ORR: Right.
| |
| 20 DR. SCHROCK: -- a consequence of the view in NRR 21 that there were physics that were not adequately covered by 22 the code?
| |
| 23 MR. ORR: Well, it basically came down to this.
| |
| 24 In the three- and four-loop plant, it is non-conservative to
| |
| -25 have a lot of the -- a lot of entrainment in the core, l
| |
| I ANN RILEY & ASSOCIATES, LTD.
| |
| -p(,j Court Reporters
| |
| '1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 192 1 because you're pushing cooler water up the core and getting
| |
| (~'Nl
| |
| +
| |
| 2 additional cooling from it.
| |
| yj 3 But if you have a lot of -- that would be 4 non-conservative.
| |
| 5 On the other hand, if you put a lot of entrainment 6 up in that -- up in the upper plenum --
| |
| 7 DR. SCHROCK: You're confusing me. I'm puzzled by 8 conservative arguments when we're talking about best 9 estimate.
| |
| 10 MR. ORR: Yes, I realize that.
| |
| 11 DR. SCHROCK: Okay.
| |
| 12 MR. ORR: But in the licensing arena, I like to 13 protect -- I like to be -- you know, I'm still older and I 14 look at some of that stuff, and I sort of like to harken
| |
| , - -n
| |
| (
| |
| (- )
| |
| 15 back to there's comfort in conservatism. In 16 non-conservatism, there's not comfort, and when you don't 17 know the difference, it's better to have some -- you don't 18 know if you're -- you don't know exactly where the real 19 number is, you'd rather be on the conservative sida of that 20 real number than the other.
| |
| 21 You can do some calculations that come out to a 22 95th percentile anyway.
| |
| 23 DR. SCHROCK: Does this address the question that 24 I asked?
| |
| 25 MR. ORR: What is your question?
| |
| ' 7''T ANN RILEY & ASSOCIATES, LTD.
| |
| '',/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 193 1 Now, the question is things like the upper plenum
| |
| '(''}
| |
| ~
| |
| 2 ' parameters that we were talking about did not have -- in the 3 three- and four-loop plant -- did not have the same emphasis 4 it does in this version of the same model. That is why we 5 didn't have that included or two-loop plants included in the 6 approval for the use of that three- and four-loou 7 methodology as is.
| |
| 8 This adaptation here I would not approve for a 9 three- and four-loop plant, because things like the threa- ,
| |
| 10 and four-loop plants tend to be -- not always but tend to be 11 guillotine-limited, and you would want ranging of the break 12 flow there.
| |
| 13 So, that's why the two-loop plants were not 14 included in that approval then.
| |
| im 15 DR. SCHROCK:
| |
| (v) I guess you've touched on another 16 topic that we've had discussion about, which is split breaks 17 versus guillotine breaks, and I don't know how you core by 18 your knowledge for the split break.
| |
| 19 MR. ORR: We do not come by any knowledge. That's 20 why we allow them to do sensitivity studies to determine 21 which is the worst one. It goes back to a conservatism 22 again.
| |
| 23 We're back within the licensing arena of assuring 24 a certain amount of conservatism, even though we're getting 25 closer to realistic.
| |
| l l
| |
| /~'s ANN RILEY & ASSOCIATES, LTD.
| |
| '\ ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| ! Washington, D.C. 20036
| |
| ! (202) 842-0034
| |
| | |
| - _ - - . - . . - . . . - -. --. . _ . . ~ . -- -.- ~ . . - . _ _ -
| |
| 194 I l
| |
| 1 DR. SCHROCK: I don't know that you are getting
| |
| ~
| |
| : 2. closer'to realistic when you do calculations for a, quote, 3 " split break" which you have absolutely no experimental l
| |
| \
| |
| L 4 evidence. '
| |
| 1 5 MR. ORR: Well, I don't want to get into the trap 6 of telling'you I know what the flow is.
| |
| 7 DR. SCHROCK: Do you have any basis for contending 8 that you know anything about what is the limiting geometry 9- of the break?
| |
| l 10 MR. ORR: That's why we just determined -- let i 11 them do a calculation that -- where they take the most 12 conservative, mainly in terms of peak clad teraperature, and
| |
| -13 use that as their basis, because we're already taking what 14 we -- we have a high level of -- I'm not going to use the
| |
| () 15 word " confidence" but assurance that we have the worst case 16 flow, and then we do our ranging around that, not ranging in 1
| |
| 17 flow but ranging in the other parameters above that. '
| |
| 18 DR. KRESS: Is there some rationale, though, tnat i
| |
| 19 you can say that, in a split break, for example, that you 20 could get blow-downs from both directions that are 21' determined by the conditions upstream of both of those-22 directions and rationalize that some type of possibilities 23 or the limiting cases, and I think that's what Virgil is 24 looking for.
| |
| l 25 MR. ORR: I see what you're saying, and I agree l
| |
| l
| |
| ''}
| |
| i s_f ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| . ~ _ _ __
| |
| | |
| 195 1- with'you That's why I'was careful about saying that we ,
| |
| -?
| |
| '2- have a pretty good-assurance that that isn't the case, say, t''"p
| |
| =Y '
| |
| , 3 if'they've found something in'there, that certainly within ,
| |
| 4; the. ranging,t they'll be.the 1.0. case.
| |
| 5 DR. KRESS: You certainly found a. limiting case 6' .that their.model would give you.
| |
| 7 MR. ORR: Right.
| |
| 8 DR. KRESS: The question is does their model Si . '.really capture the limiting case, in reality?
| |
| 10 MR. ORR: I think that within the -- I think, 11 again, this model, their break-flow model, again, this has 12 pre-existed, this isn't a new break-flow model. The split 13 y .model has pre-existed anyway.
| |
| 14 Now, in that, we do have the 1.0 multipliers and
| |
| ~ stuff which assure that, within that spectrum that you're
| |
| ~
| |
| 15
| |
| : 16. calculating, that flow, whatever it is, is still 17 . represented, even though you are not -- or that case is 18 still represented, even though you are not varying the break
| |
| =19 itself anymore. You've biased that break flow.
| |
| 20 We have a high assurance that there's not some 21 singularity point or something out there at a different size
| |
| '22' .that affects the statistice, but I can't say that there 23- isn't.
| |
| 24 But I think that our assurance is a high enough 25: -level thht I'm willing to say that using the worst split ANN RILEY & ASSOCIATES, LTD.
| |
| ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 106 1 break is quite conservative, and right now, I know we're I
| |
| (~T 2 trying to get close to realistic, but again, I'll get back l G) 3 to Dr. Schrock's comment that I defy anyone to try to tell 4 you exactly what the break flow is or predict it. l 5 So, that's why we allow either bounding it or else 6 -- you know, in terms of peak clad temperature, or else 7 ranging it, like we do with guillotines.
| |
| 8 DR. ZUBER: Let me ask you -- this goes back 20 9 years.
| |
| 10 MR. ORR: Uh-huh.
| |
| 11 DR. ZUBER: We were never really able to calculate 12 well the critical flow, and it comes again now and at the 13 last meeting. When I was part of NRC, there were users 1 14 request from your part to our ears. Did you ever generate a
| |
| ,-~\
| |
| 15 user request for critical flow to put this problem to rest?
| |
| (d 16 MR. ORR: I didn't.
| |
| 17 DR. ZUBER: This is a question, because this comes 18 time and time again over the last two decades, and now we're 19 going to the best estimate, which is worth millions of 20 dollars to the industry, and there is the rationale to do 21 the best we can.
| |
| 22 If you were the user, ask Research to do this job 23 for you.
| |
| 24 MR. ORR: They don't do things for free.
| |
| 25 DR. ZUBER: The point is, though, this is a l
| |
| ! / 'N ANN RILEY & ASSOCIATES, LTD.
| |
| l ( ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| ~ _ . - - . - . . . ~ . . . . - . . . - . - - . - . - . - . . . . . . . - . . . - - . -
| |
| I 197 1 problem which was with.us, and now there is a good financial
| |
| .2 motivation, because the industry needs it to benefit our
| |
| -3 economy.
| |
| ! 4 MR. ORR: Okay.
| |
| 5 D;R . ZUBER: Why-don't we address this problem, and L 6- 'over,the next few years, obtain -- .
| |
| ?
| |
| 7 MR. ORR: I don't know. -That's a policy matter 8- over mv head. That's not really within the scope of this l 9 review.
| |
| : 10. DR. WALLIS: I was going to say this is not.really 11- within the scope of what we've been asked to review. We've 12 been asked to review the upper plenum --
| |
| 13 DR. ZUBER: Well, this'is something which always i
| |
| 14 come up,;and the question is -- we are listening to what R'esearch is doing --
| |
| 15 16 DR. WALLIS: You could ask Research.
| |
| 17 DR. ZUBER: But they are the guys who generate 18 that request.
| |
| 19- MR '. ORR: Those are our conclusions. We still 20 consider this methodology to be acceptable for licensing 21 use, and I have come into this meeting expecting -- I still
| |
| , 22 . expect to now -- issue an SER in about two months unless L
| |
| 23' something you have to say really puts a glitch into that.
| |
| 24 That factored in, we received a report from our 2
| |
| .5 contractor in 1998. W' re planning to issue an SER by the j i
| |
| t ANN RILEY & ASSOCIATES, LTD.
| |
| O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036 (202) 842-0034 1
| |
| | |
| . . .. - - - - . . . . - . - . . - - . . .. . - . - - - . . - . -.. -~ - ._.- ~.- .-.
| |
| 198 1- end'of February. We expect to get the approved version 2 back, proper documentation.
| |
| -3 I don't have the statement. I'm supposed to say ;
| |
| l 4 all' sorts of nice motherhood about all the nice things they ;
| |
| 5 did.
| |
| ~6 No known errors in codes, tests, and data or i
| |
| 7 documentation. Documentation meets regulatory requirements.
| |
| 8 Staff able to review all documentation, determined that the
| |
| , i J
| |
| 9 data, codes, and plant design meet regulatory requirements.
| |
| 10 DR. FONTANA: You must have a different document ,
| |
| i 11 than what I looked at. I've had recorts here with no 12 definition of symbols anywhere, 7 figure that's referred to i 13 in the text and the figures aren'c there. Figures are not 4
| |
| ~ 14 self-standing.
| |
| j 15 MR. ORR: These things here that you've got --
| |
| 16 that goes back to my other thing, and that goes back to the i
| |
| 17 chronology of this review.
| |
| 18 We are not interpreting this to be a stand-alone 19 document in itself.
| |
| 20 In our review, if you remember earlier, they said 21 that the original report became rather obsolete because so l 22 many things were changed as a result of the three- and 23- four-loop review and some of the better understanding of 24 best' estimate that Westinghouse gained in the process.
| |
| 25 So, a lot of the things you've got here are in the L
| |
| 1 L ANN RILEY & ASSOCIATES, LTD.
| |
| t'
| |
| . g,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034 l _
| |
| | |
| m . . _ _ _ _ _ - . . . . _ _ . . . . . . = - _ . - . _ _ _ _ _ . _ . . . _ . _ . _ - _ . . _ _ _ . _ ...
| |
| l L 199 )
| |
| i 1- fnature of question and answer. The question-and-answer !
| |
| l-
| |
| /''N 2' stuff is not. stand-alone. This document will have some b 3 definition in it when they are done and submit the approved 4 version.
| |
| i i
| |
| 5 What we're approving right now is the technology, i o .
| |
| i 6- This approved version, when we get it, we look at it to see j 7 if it has that type of: stuff in it,.just like we did'the 8 three- and four-loop methodology.
| |
| 9 DR. SCHROCK: How do you know'that it's' acceptable
| |
| )
| |
| 10 if you don't have it in suitable documented form?
| |
| 11 MR. ORR: Well, that's -- this is when we verify 11 2 that it's in an appropriate form. Westinghouse knows what 11 3 the appropriate form is. They went through this on the 14 three- and four-loop methodology.
| |
| () 15 We are not going to -- the one difference is that 16 we are not going to make this -- we agree with 17 Westinghouse's interpretation that thic is just an extension 18 of the three- and four-loop methodology, specially adapted 19 to two-loop plants.
| |
| 20- DR. WALLIS: Would it be true to say that's what 21- happened here is that INEL has essentially approved ;
| |
| i 22- something, and you look at that and say INEL approved it, 23 therefore it's okay?
| |
| -24 MR. ORR: No. I have looked at what INEL wrote.
| |
| :25 I do not disagree with what they wrote. I had Alan Levin, L .[ ANN RILEY & ASSOCIATES, LTD. I
| |
| ; ~' Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . . . . - . - , ~ - . - . . = . . - . - . - - . - . . - - . - . . - .
| |
| 200 t 1 who is in.the area of empirical formulations and things, 2 more experienced than I am -- or at least more authoritive, l 0("N 3~ and'we looked at the report, and others have looked at the 4 . report that INEL sent us. We looked at what we were given-5 from Westinghouse, and we concur with the basic conclusions 6 that INEL reached.
| |
| 7 DR. WALLIS: Okay. So, you -- so, NRR also 8 thoroughly critiqued the Westinghouse work.
| |
| 9 MR. ORR: Well' we didn't -- we pay a contractor 10 to do something besides just parrot back what we want to 11 hear, but we did look at it to-see if anything seemed 12 flagrantly wrong with it, and we didn't find any such 13- things.
| |
| 14 Based on the data they quoted, the correlations
| |
| () 15 16 and the way they were using them seemed to be supported by the data, regardless of how they expressed it here.
| |
| l 17 DR. WALLIS: Did you ask if anything was l
| |
| .18 flagrantly right with it?
| |
| 19 MR. ORR: Well, I.can't say about that.
| |
| 20- DR. SCHROCK: Did you get a solid consultant, 21 independent review of the Fineman report?
| |
| 22' MR. ORR: Well, it was more than Fineman that 23 generated it, and we did not hire another consultant to 24 review a first consultant.
| |
| l 25 First of all, INEL --
| |
| i i
| |
| ' N ANN RILEY & ASSOCIATES, LTD.
| |
| V Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 '
| |
| (202) 842-0034 l
| |
| l-i
| |
| . - . . . . , , , - , . . - - . .L
| |
| | |
| 201 J L
| |
| : l. 1 DR. SCHROCK: The kind of consultant I'm talking V
| |
| l
| |
| ('T.
| |
| l 2 .-about is'somebody:who has had experience in:the 1 i; \ss/
| |
| ~
| |
| -3 experimentation of counter-current flow with condensation '
| |
| 4 effects.
| |
| 5 MR. ORR: Oh, condensation effects. Well -- -l J
| |
| 6 Alan,.do you.have much experience with- I 7' condensation effects?
| |
| 8 .MR. LEVIN: A. litt7.e here and there. j
| |
| : 9. MR. ORR: See, some.
| |
| 10 DR. WALLIS: I'm wondering, have we reached the 11 end ot your presentation?
| |
| 12 MR. .ORR: I think we've pretty much reached it. I i i
| |
| 13 wanted to put up that schedule, because we're working to 14 Jim's schedule to accommodate the needs of licensees, and I
| |
| () 15 16 guess February seems like about it. If we turned around our SER in : February, gives Westinghouse some time to' generate a
| |
| '17 . report.
| |
| I 18 DR. WALLIS: So, what do you expect the'ACRS to
| |
| : 19. do?
| |
| -:20 MR. ORR: I don't know.
| |
| 21 DR. WALLIS: We listen to this, but what do you
| |
| ' 22' .want us to do.
| |
| 23 'MR. ORR: You listen to this. If you have some 24 feedback, if you have some criticism, that's fine.
| |
| : 25. DR. WALLIS: Are we obligated to do anything?
| |
| i
| |
| [. ANN RILEY & ASSOCIATES, LTD.
| |
| t' Court Reporters 1025 Connecticut. Avenue, NW, Suite 1014 Washington, D.C. 20036
| |
| -(202) 842-0034
| |
| : z. _ , . - -- - . - . - - - - -
| |
| | |
| ~ ~ . _ - . _ _ . . _ _ . . . _ _ . . . . _ _ . _ . _ . _ _ _ . _ ....__.. . _ _ . _
| |
| I 202 L 1 MR. ORR: I-don't know that you are. I came here l .
| |
| 2 mostly because --
| |
| ~
| |
| .3 DR. WALLIS: So, this is a learning experience for 4 us, maybe.
| |
| 5 MR. ORR: Well, I'll tell~you, there is a lot to
| |
| ~
| |
| 6 .be learned from this.
| |
| 7 MR. SINGH: Did we request a.-letter from you guys
| |
| ~8 to. review this?
| |
| l' 9 DR. WALLIS: No.
| |
| lx
| |
| . 10 MR. SINGH: You didn't request this review?
| |
| 11 MR. ORR: Dr. Kress had talked to Paul Boehnert, l
| |
| 1
| |
| .2 I had told Paul Boehnert that we had finished the review of ;
| |
| l 13 this, and'I said is ACRS interested in this?
| |
| )
| |
| 14 Part of the reason why I wanted this is because
| |
| () '15 .
| |
| 16 this is.rt.11, regardless of all the stones you throw at it and everything, itLis the most advanced LOCA technology we i
| |
| 17 have. l 18 When'you're doing something, it's the sharpest 19 pencil we've got-right'out there in industry right now.
| |
| 20 hit's much sharper than all the dull things other people are 21' using.
| |
| t 22 DR. KRESS: Yes. This is precisely the way it 23 .went,.and we said we wanted to review it mainly because we 24 expect to see lots of plants come in, or at least these 25 two-loop plants come in with requests for changes to the i~ :
| |
| L i
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| l Court Deporters
| |
| . 1025 Connecticut Avenue, NW, Suite 1014
| |
| ! Washington, D.C. 20036 (202) 842-0034
| |
| . _ . . ~ . ..- _. .. _
| |
| | |
| 3 I
| |
| 203' I
| |
| 'l licensing basis, upgrade their fuels, maybe change the power r's 2 rating, and this sort of stuff, based on this code, and in i p-3 order.to make judgements on those things, which we will have ;
| |
| to review, we needed to understand what this code had in it.
| |
| 4 '
| |
| 5 MR. ORR: There's other things, too, here. If you 6 notice here, we've taken-the. code, not -- and, in my 7 opinion, not changed the methodology at all. We've changed 8 a couple of variables.
| |
| .9 We're coming linto an era of risk-informed l i
| |
| 10 regulation. '
| |
| 11- This code, in addition to being the finest-tuned 12 pencil we've got, also coordinates very well with PRAs and j l
| |
| 13 other risk-informed techniques, because you're dealing with I 14 a probabilistic best peak clad temperature, your other i
| |
| () l'5 16 factors are probabilistically determined.
| |
| You've-got, in the plant conditions area and other 17 things, things like allowed outage times and things 18 conceivably should be fit in there without altering the code 19 itself, without altering the process. It fits very well in.
| |
| 20 We're seeing a lot of stuff in the licensing area, 21 in tech spec regimes and stuff, where they're calling stuff 22 and screaming PRA and yelling best estimate, and quite 23 honestly, it's not of the quality that we would get out of 24 using a code like this or a process like this for evaluating 25 them.
| |
| t-ANN RILEY & ASSOCIATES, LTD.
| |
| D)-
| |
| (, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036 l
| |
| (202) 842-0034
| |
| | |
| 204 1 DR. KRESS: Within PRA space, this is almost an 2 on/off switch, because it mostly relates to success criteria 3, of their ECCS, and they're either successful or they're not.
| |
| l 4 MR. ORR: If you look at it here, we're saying
| |
| ;5 that, 95 percent of the time, you're successful.
| |
| 6 DR. KRESS: That's right, and that's success 7 criteria that enters into whether you go further into the'--
| |
| 8 MR. ORR: You enter that 95 percent into the PRA 9 and that has something.
| |
| l- 10 DR. KRESS: You use the calculations of a code L 11 like that to make that judgement, along with other things.
| |
| 12 MR. ORR: So, conceptually, there are ways of 13 using this code that are clumsier with an Appendix K l l
| |
| 14 methodology, and I think that what is seen here is that this 15 ' code has a lot of -- this process has a lot of possibilities
| |
| ~16 .that Appendix K -- to handle things in a little better 17 fashion than Appendix K methodologies do, and so, that's why
| |
| '18 .I think it's of interest to you.
| |
| 19 DR. WALLIS: We have people waiting.
| |
| 20 MR. ORR: Certainly. I didn't mean to take up 21 your time.
| |
| 22- DR. WALLIS: So, you don't want us to -- we don't i 23 have to write a letter or anything. You just want some 24 informal feedback?
| |
| \
| |
| 25 MR. ORR: If you feel really upset about i l
| |
| l;
| |
| /%
| |
| V
| |
| , ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 j Washington, D.C 20036 !
| |
| (202) 842-0034 l l
| |
| | |
| l 205 1 something, I would encourage you to write it. I don't know.
| |
| /'')
| |
| %.J 2 Really, it depends on the process.
| |
| 3 DR. WALLIS: There's other people waiting. I 4 vonder if we could take up till three o' clock to go around 5 the table and get some one- or two-minute reactions to what 6 you've heard, and just try to be summary and not get into 7 details.
| |
| 8 DR. ZUBER: I would really again commend the 9 industry for taking this part and also the staff for pushing 10 this. I think this is our best technology and we should use 11 it.
| |
| 12 I am concerned, a little bit distressed about the '
| |
| 13 possible misuse, and I have no problem with the methodology.
| |
| 14 The problem here is the condensation in the CCFL, and this gs
| |
| ( ) 15 is the really crucial difference between this plant and 16 previous, and this is what should have been focused.
| |
| 17 The way you have done it is, I think, not 18 convincing. I think the problems we have seen, oscillations 19 after 80 seconds, the problem of historesis -- all this is 20 tied to the non-physical way you calculate the CCFL and the 21 condensation.
| |
| 22 My advice, in order to have really a robust method 23 which you can defend in front of an intervenor, is to modify 24 this approach -- I think you can do it by February -- use a 25 correlation based on experimental data and use this as the ANN RILEY & ASSOCIATES, LTD.
| |
| (/~N) s, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 206
| |
| ; 1 ' boundary conditions. ;
| |
| l l /''
| |
| ! km 2 I think other codes do the same thing. Don't use !
| |
| ! 3 the flow regime map and artificial tubes, artificial 1
| |
| 4 condensation to model this.
| |
| l 5 The only difference -- important difference l 6 between your plants and previous plants -- my advice would 7 be this could be done in two months and do it -- and you 8 have then a robust way, you can defend it and have approved
| |
| : 9. by anybody, and you could then be proud of your work you 10 have done.
| |
| 11 Otherwise, you have something which is very 12 wishy-washy.
| |
| 13 DR. WALLIS: Virgil? !
| |
| 14 DR. SCHROCK: Well, I've long been a proponent of f~h 15 best estimate calculations and deplore the idea that the
| |
| ( )
| |
| 16 Appendix K is grandfathered, evidently, forever. I think 17 it's sort of a disgrace to the agency that that situation 18 exists.
| |
| 19 But in terms of what we've discussed today, I l
| |
| 20 would have to say I'd be extremely hard-pressed to make any '
| |
| 21 recommendation that would be in any way an endorsement of 22 NRR's decision to conclude that there is no new modeling 23 needed in order to make COBRA / TRAC successfully perform 24 predictions on a UPI plant.
| |
| 25 I think that's just categorically incorrect as I l
| |
| (^'g ANN RILEY & ASSOCIATES, LTD.
| |
| (,,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034 i
| |
| {
| |
| | |
| l 207 !
| |
| I see'it today. I_ don't see any way that I could be convinced '
| |
| 2 otherwise on the basis of the kind of information which has-L o
| |
| 3 been provided in this meeting, ,
| |
| ! l
| |
| )
| |
| l 4~ So, I would leave it there. '
| |
| 5 DR. KRESS: I certainly agree with the i i
| |
| 6 consultants, with the exception-that I understand the-
| |
| -7 grandfathering and recognize that it's necessary,-that the to 8: staff probably has no option except to do that.
| |
| 9 The other point I wanted to make is I wasn't 10 convinced that the split break was very well shown, 11 demonstrated to be a' limiting one, and I think more can be 12 done there in a very easy fashion, and what I would do is 13 vary the temperature at the break between two ranges.
| |
| -14 One of them is what you get coming from one o
| |
| { 15 direction compared to what you get coming from the other
| |
| '16 ' direction, but when you vary that temperature, you also have 17 to make the flows coming from those two directions 18 consistent with it in a mixed -- in your model, you mix the 19 two and you could get a temperature going out.
| |
| 20 I would vary that temperature and make those two 21 flows consistent over the range of temperatures possible, 22 and then, at the same time I vary those -- I would do it for 23 each of the break sizes and see if I come up with some other 24 break size that's limiting.
| |
| 25 But other than that, I think I also agree that 1
| |
| ! s ANN RILEY & ASSOCIATES, LTD.
| |
| I \ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 y:
| |
| | |
| . _ _ . . - - . _ _- - .. ..~
| |
| 1 208 l
| |
| L 1 best estimate is the way to go, and I certainly support you
| |
| /~~ 2 going this way.
| |
| N.)T 3 DR. WALLIS: Mario? I 4 DR. FONTANA: Well, I agree with most of what's 5 been said before.
| |
| 6 I think the impression I got is that the upper 7 plenum flooding probably would work, but proving it and 8 making it convincing is more of a case of organization and 9 presentation, I think, and proper documentation.
| |
| l 10 That's all I've got to say.
| |
| 11 DR. WALLIS: Well, I feel the same way. I feel I 12 that a more organized, convincing way should be found to --
| |
| 13 if you want some endorsement of what you're proposing, 14 because I found that I had far too many questions which I
| |
| >, ~\ .
| |
| L) 15 need to get answers to, I need to dig into or something.
| |
| 16 So, we remain with some doubts, but I do thank you l 17 very much for making the presentations, and I think it's l 18 very useful to do so. I think we've learned a lot, and 19 you've probably learned something, too, and I hope NRR has 20 learned some things.
| |
| 21 Now, we should take a break for -- let's meet at 22 10 after three -- 3:15. I'm sorry, but Research will be 23 here later than expected.
| |
| 24 (Recess.)
| |
| 25 DR. WALLIS: We will go back into session, and if f
| |
| I /'N ANN RILEY & ASSOCIATES, LTD.
| |
| ( ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
| |
| | |
| 209 L 1. you're lucky, Farouk, you can get yours over before the L .
| |
| 2 consultants get back, We are really looking forward, 3- Farouk, to hearing all the' good things that are going on -i 4 here.
| |
| 5 MR. ELTAWILA: Well, I hope that I will not be i
| |
| 6 disappointed, but I heardLthat the branch have done a very l good job, and hopefully, we'll convince you of that.
| |
| '7.
| |
| 8 My name.is Farouk Eltawila from the Office of
| |
| .9 Research, and in our presentation, we're going to cover all 10 the areas that you discussed in your -- in the agenda. So, 1 11 I'm going to go over the budget and give you a relationship 12- between our program and the NRC strategic plan just to show 13 you that we are really focusing our program.
| |
| 14 I'm going to address some measures of success for
| |
| () 15 the present program and talk about the -- particularly as i 16 .related to the code consolidation, and then I'll give you 17 the balance of the thermal hydraulic five-year research plan 18 and summarize the achievement that we had last year, and i then I will be followed by four presentations, one on user 20 interface for RELAPS code, and that would be given by Ben 21 Gitnick from Scientech, thermal hydraulic.and codes that 22 will be given by Dave Ebert from the staff, and then Joe 23 Kelly will have two presentations about the consolidation 24 updates
| |
| =25 The fiscal year 1999 resources, as you can see I
| |
| ; ~% ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
| |
| ( (202) 842-0034 l
| |
| I
| |
| | |
| . . -- . . - . _ - - _ _ _ _ ~ . - . _ - . . - - - . -
| |
| I l
| |
| i <
| |
| p 210 l-L 1 here, is about - .slightly over two million less than we had 2 in the research plan, one of them due to a need to fund 3 other high-priority work at NRC, and the other one is due to 4- my own mistake, which I agreed to give up $1 million in 5 exchange for four people,
| |
| '6 They took the million, and.I never seen the four l
| |
| .7
| |
| ~
| |
| -people.
| |
| I 8 DR. SCHROCK: They took the money and ran?
| |
| 9 MR. ELTAWILA: So, that's the situation, and what 10 make things worse, too, that I lost two of the key people 11 from the branch,. Simon Smith and Vince Missou, and I think 12 it's a reflection about what happened in the agency last 13 year with the rumors about cutting 700 staff tomber, and 14 some managers say that we don't need thermal hydraulic l 15 research, we don't need research.
| |
| ( )_
| |
| 16 All that has a very bad impact and demoralizing to 17 the staff.
| |
| L 18 So, the good people left. I still have few left, !
| |
| 19 but hopefully, I'd like to see the committee help us.
| |
| 20 If the agency.wants to kill thermal hydraulic 21 research, do it at once, but don't do it slow-motion, you
| |
| ! 22 know.
| |
| 23 I would like to see the committee write something 24 about that, because I think, if we~ keep on that pace, I L 25- think that will be very bad for the agency and for the L
| |
| ; ff"N ANN RILEY & ASSOCIATES, LTD.
| |
| J ( ,/ Court Reporters 1025 Connecticut Avenue,-NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
| |
| l
| |
| - . . ~ - -- , - - -
| |
| | |
| ,- . .. - .- .. . - . . . . . - . - ~ . . _ . - . - . . . . . - . . . - . _.
| |
| r 211 I-l 11 ' people that are working here.
| |
| [ }
| |
| 2, DR. KRESS: That two million, Farouk, what percent 3: of the total is that?
| |
| f 4 MR. ELTAWILA: -That's about 30 percent of the
| |
| ! 5 . total, but the impact is much greater, beca'se u in '97, I had 6 extra money'. So, you can see it's almost like 50 percent-7- from-'97 money.
| |
| 8' Again, on one hand, we are saying that the agency, 9 in its strategic plan, say we will maintain a research j
| |
| 10- capability to provide timely and independent technical basis
| |
| 'll for NRC. regulatory decisions.
| |
| 12 That's taking out of the strategic plan, so that 13 NRC can carry out its mission efficiently and effectively, 14 and we will provide the staff with the best available tools 15 'so they can do an excellent job.
| |
| 16 And that's what we are trying to do, and I'm not 17 really complaining about that the' agency have limited j 18 resources and they have priority, but there is nothing wrong 19 in that, define the priority of the agency and go after them 20 and resolve them, but what's important that we have to look 21 at--the value of the research and whether we want to keep 22 that capability, like thermal hydraulic, fuel, any of that 23 stuff, going or not. i i
| |
| 24, The other things that we believe that the thermal 25 hydraulic research will have an impact on is the I i
| |
| l L ANN RILEY & ASSOCIATES, LTD. ;
| |
| Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034 i l-I
| |
| | |
| . . _ . . ___ ._._m-- . . . _ _ . _ . _ . _ _ _ . _ . _ . . . _ _ . _ . _ _ . . _ _ _ . . . _ . _ . _ . . . _ _ . _ _ . . .
| |
| 1 212 1 '
| |
| L . risk-informed regulation.
| |
| l 2 By providing tools that will have better best
| |
| )
| |
| L -3 estimate models, we will be able to assess the safety 4 margin, which the-industry has been asking for relief from l 5 all these conservatives. So, we are working on achieving l
| |
| 6- this core goal.
| |
| l.
| |
| 7 DR WALLIS': The best estimate -- from what we 8 just' heard this morning,;we~were seeing correlations that .
| |
| I' 9 are.over 30 years old for phenomena, and surely there's been 10 research done since then. Best estimate surely should be-l 11 based on the best, newesc methods, not on something that's
| |
| '12 old.
| |
| 13 So, there's great scope for improving. Best 14 estimate should be best, not just something.
| |
| ( 15- MR. ELTAWILA: Well, I agree with you, and I might 16 be speaking out of turn here, but I think you have to look 17 at the industry, and maybe because the licensing. process of 18 codes and review and things is a lengthy process, so they 19 prefer to keep things like the way it were 20 years ago, so 20 they don't go through the review process.
| |
| 21 DR. WALLIS: But it seems to me, if you're always 22; just following the industry and industry just wants to 23 submit 20- or 30-year-old work, there will be no progress, 24 but if NRC could be best and up to date, that's what we 25 should be looking for.
| |
| i O ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW,-Suite 1014
| |
| [ Washington, D.C. 20036 (202) 842-0034
| |
| | |
| .. . . - . . . . , . . . . _ . . . ~ . . . - . . . ~ . . - . . - . . . . _ . . - . . . . . . . . . . - . . .-- ... - .- - .~. .
| |
| 213
| |
| .1 MR. ELTAWILA: We're trying_to move in that.
| |
| t 2. direction in actual best estimate, but I don't blame anybody
| |
| :t -for saying that we don't have a best estimate tool, because !
| |
| 4 our tool is not best estimate tool, and it's very 5 cumbersome, and we'll try to do.something about making it 6; easier for the user to be able to use, so we can move in the l 7 direction of best estimate method, and then, maybe, when we l i
| |
| 8 develop our.own tool, the industry will;take noti'e c of that I 9 and see the benefit that you will gain out of using best i
| |
| 10 estimate methodology, they might adopt it.themselves,.and if I 11 the codes is simple enough, the staff might find that it can 12 do this review in less time than what it's taking right now.
| |
| 13 MR. KELLY: This is Joe Kelly from Research, and I 14 agree with you completely, Professor Wallis, and we're
| |
| () 15 16 approaching that on two tacks.
| |
| One is, as.part of the code consolidation effort, 17 we're formulating more comprehensive assessment matrices for 18 the' code-specific applications.
| |
| 19 We will then test the code and determine where the 20_ major deficiencies are and target those for model 21 improvement and, if you will, a follow-on to the code 22 . consolidation effort, but at the same time, we're beginning 231 small-scale separate effects tests that are targeted 24 directly at model development.
| |
| 25 There's a phase separation at Oregon State, there
| |
| ' O.
| |
| (%,/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 1 l
| |
| i 214 1
| |
| ! )
| |
| 1 is the'small-scale re-flood test at Penn State and
| |
| !- /"
| |
| d}
| |
| 2 3
| |
| sub-cooled boiling at UCLA.
| |
| So, that's why we're doing those. They're I
| |
| 4 small-scale, highly-instrumented, and that's why we're doing 5' these, to come up with better models for the future.
| |
| 6 DR. KRESS: Let me ask you a somewhat off-the-wall 7 question.
| |
| 8 If I'm a PRA man -- George is not here -- and I'm 9 calculating CDF, which is one of our PRA risk methods,-which 10 would I find more useful to me in calculating CDF, your good 11 consolidated thermal hydraulics best estimate code or a 12 severe accident code like MELCOR?
| |
| 13 MR. ELTAWILA: Without a doubt, it's going to be 14 thermal hydraulic codes to calculate the success criteria,
| |
| () 15 16 and then, if you want to calculate the risk itself, you have to go to the severe accident code.
| |
| 17 DR. KRESS: Why isn't that one of your needs? i 18 MR. ELTAWILA: Well, I put that as risk-informed 19 principle.
| |
| .20 If we are going to move in the risk-informed --
| |
| 21 there will be design basis accidents. It's not going to 22 look anything like what we have right now, and we're going 23 to look at a lot of scenarios and, from this scenario, 24 . determine the most limiting one, and that's when the codes 25 will play an important role here, f" ANN RILEY & ASSOCIATES, LTD.
| |
| Tg ,}f Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| c 215 1 -DR. KRESS: Thank you.
| |
| /'') 2' MR. ELTAWILA: Our desired goal, as we stated them l \_/ . .
| |
| L 3 before, is to have a single code that's capable of 4 performing accident analysis for both light-water reactor of 5 current generation and advanced light-water reactor like
| |
| ! 6 AP600, and based on our interaction with NRR and other user 7 groups like, for example, the Indianapolis meeting, item 8 _five, some of the important feature that they want to see in 9 the code.
| |
| 10 They want to have 3-D reactor kinetic capability, 11 that the' code should be able to be coupled through a 12 containment analysis code, the severe accident code !
| |
| 1 l 13 sub-channel analysis, fuel code, and CFD code.
| |
| 14 DR. ZUBER: I have a question. These people who I
| |
| () 15 expressed so many desire -- do they put some money with it?
| |
| 16 I mean it's easy to ask for something if somebody is going 1
| |
| 17- to pay, somebody else.
| |
| '18 MR. ELTAWILA: Actually, they sent the user need 19 letter, but again, when it comes to budget crunches, you 20 know, the big bosses are the one who make the decision about 21 what we do.
| |
| 22 The user office from NRR -- it's the same money, l 23 that's NRC money, but if you're talking about the member --
| |
| 24- the member are paying a fee. We will receive a membership j 25 from different country, total about $1 million from all L
| |
| i ANN RILEY & ASSOCIATES, LTD.
| |
| (,_
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i
| |
| Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 1 216 1 these foreign countries.
| |
| ~'T 2 So, we, every year, will go over their priority (G
| |
| 3 list, and we'll try to address the top two priority using 4 the money that they are providing for us.
| |
| 1 5 DR. ZUBER: Okay. So, they at least pay for 6 something.
| |
| l I
| |
| 7 MR. ELTAWILA: Oh, absolutely everything is paid 8 for.
| |
| 9 DR. SCHROCK: Is that separate from this budget 10 that you described?
| |
| l 11 MR. ELTAWILA: That's correct, yes.
| |
| 12 Now somebody will come and find it.
| |
| l 13 DR. SCHROCK: I'm sorry. I shouldn't have brought l l
| |
| 14 that up.
| |
| .[m't 15 MR. ELTAWILA: One of our desired goals is to have w.)
| |
| 16 the staff here, ready to respond to the emerging needs of 17 the agency. We would like to be the one that do the review 18 of the topical report of the industry instead of going to a 19 laboratory.
| |
| 20 We want to be the arms of NRC in terms of getting 21 the information, provide the technical expertise so we can 22 help NRR carry its regulatory function, and we want to do 23 more'research with universities, because most of the
| |
| -24 creative ideas are really generated in universities, and at 25 the same time, they are cheap.
| |
| l
| |
| /''N ANN RILEY & ASSOCIATES, LTD.
| |
| (_,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l
| |
| (202) 842-0034 l
| |
| | |
| 217 1 Nothing against national lab, but that's a true
| |
| '/ 'h 2 statement. The national lab have another role. And we want
| |
| 'v/
| |
| 3 to increase the staff technical involvement in the work that l 4 we are doing here, and as part of Joe Kelly's presentation, 5 his second presentation, he is going to talk about the code 6 modularization, which is completely done in-house here.
| |
| 7 DR. SCHROCK: Is there any possibility of 8 furthering your objectives by some kind of cooperation with 9 DOE on the kinds of programs that it funds?
| |
| 10 MR. ELTAWILA: Not at this time. The DOE right 11 now -- we don't know what they are going to fund in the area 12 of thermal hydraulics.
| |
| 13 Some of our contractors have been submitting 14 proposals to DOE for the experimental facility, but for code
| |
| () 15 development, I know there are proposals, but I don't know if 16 that's going to be funded or not.
| |
| 17 Jce just reminded me that we are cooperating with 18 another part of DOE, which is the naval reactor. Also 19 during modularization, modernization of the TRACP was done 20 by CAPO, and we took advantage of that. So, there are some 21 cooperation with that part of DOE.
| |
| 22 You asked in your agenda about what -- the measure 23 of success that we are using in the program, and I think the 24 most important thing is to have user satisfaction, and the 25 user will be satisfied if we have --
| |
| i l
| |
| l(\m- ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| , 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| r l
| |
| 218 1 1 DR. KRESS: Who is the user, Farouk?
| |
| \
| |
| i
| |
| [Y
| |
| \~
| |
| 2 MR. ELTAWILA: The user is NRR, Office of 3 Research, the international community that they are using 4 the code. ,
| |
| l 5 DR. KRESS: Now, are they -- now, do you know l 6 which things they're dissatisfied with?
| |
| 7 MR. ELTAWILA: In more ways than one, but they are 8 behind you. They can speak for themselves.
| |
| 9 Anybody want to say anything?
| |
| 10 [No response.]
| |
| 11 MR. ELTAWILA: The users have been telling us they 12 need improved documentation, and we are working very hard on 1
| |
| 13 improving the documentation for the code, improve the I 14 robustness, so the code can run to completion without rh f
| |
| s ) 15 interruption or without the user have to change the time 16 step and so on.
| |
| 17 The code should have a graphical user interface to l 18 help in preparing the input data and modifying existing 19 data, be able to run multiple cases so they can run 20 sensitivity and uncertainty analysis, and will reduce error, 21 improve the accuracy, and again, better physical model, and 22 most importantly, the compensating error in the code.
| |
| 23 I think Joe Kelly had a presentation last year 1 24 where he showed that the final answer is correct but getting i
| |
| 25 it for the wrong reasons, so we get compensating error in l
| |
| l-l l /~T ANN RILEY & ASSOCIATES, LTD. 1 (s-) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l (202) 842-0034 I
| |
| | |
| 219 1 the code, give.you a good answer, but the parts of that 2 formulation is wrong.
| |
| L '3 The user would like to see internal assessment of 4 uncertainty and try to. reduce the number'of options 5 available to the user'to choose from. There are a lot of
| |
| ,6 options, so if a' user could choose one model, another user 7 choose another model, and you can get the answer.that.you 8, want to do.
| |
| 9 So, you want to reduce that by improving the E.
| |
| 10 physical model's'in the codes.
| |
| 11 Continuing on with user satisfaction, again, 12 that's the measure of. success, improving.on time so that
| |
| ~13 they will be able to run many cases.
| |
| 14 DR. WALLIS: What are you looking at for run time j '15 improvement? Are you looking at half the time, a tenth of 16 the time, or 1 percent better or what?
| |
| .17 MR. ELTAWILA: Well, they are looking for actual 18- time. They are looking for the result of the CSNI meeting.
| |
| 19 that~took place in Annapolis, and Ralph Caruso was the 20 chairman of that session.
| |
| 2 11 He was pushing very hard for real time
| |
| ~22 performance, which we will not be able to accomplish it wita the current version of the code that we have. Maybe, in the 24 future, we can use parallel processing, and as the computers 25 get~ faster, we'll come close to that.
| |
| .]
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| (_f Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 o
| |
| | |
| t l 220 1 DR. WALLIS: This is an order of magnitude
| |
| ('N 2 improvement.
| |
| l 3 MR. ELTAWILA: Several.
| |
| 4 DR. WALLIS: What are you actually going to 5 achieve in run-time improvement?
| |
| 6 MR. ELTAWILA: I think most of the improvement 7 that we're going to get in run time is going to come out of 8 the computer's speed itself, not from any of the work that 9 we are doing right now, but in the long run -- but we are
| |
| : 10 putting all the foundation to improving the physical model.
| |
| 11 In order to improve the physical model, you need 12 to improve the numeric scheme, you need to do all this 13 improvement. That's not part of the initial phase of the l 14 program.
| |
| /
| |
| (m) 15 So, any improvement in run time in the next two to 16 three years would be coming out of the computer speeds.
| |
| 17 Hopefully, at the end of the program, we will be able to get 18 a better run time.
| |
| 19 I don't know if you want to add anything to that.
| |
| 20 MR. KELLY: Again, Joe Kelly from Research.
| |
| 21 There are basically three ways in which we would 22 intend to improve the run time, but again, as Farouk said, 23 this will really start after the consolidation or the first 24 part of this effort is over, just from limitation of
| |
| ~
| |
| 25 resources.
| |
| l ANN RILEY & ASSOCIATES, LTD.
| |
| {)\
| |
| q, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 i
| |
| (202) 842-0034 L
| |
| | |
| AA - A >J.,e -,6r.- a ;- e a,. w a b- .a5- . 9Rm- a-- =,.+4aws-L-at- A & J aS m.4-,ua ak A-m A. Lw =a 1---mua. m6 m at AD-,A 1
| |
| l 221 l
| |
| . l 1- .The three ways are parallel processing -- the i
| |
| , (~T' 2 other thing you can do -- and you can do it two different
| |
| ; %-) i 3 ways -- is make the time step bigger. If you can run with a 4 bigger time step, you can march through a transient much 1
| |
| 5 faster.
| |
| l 6 You can do that by improving the -- increasing the 7 level of implicitness in the code, but also, quite often the 8 code can't run at a maximum time step, because it simply 9 .doesn't converge very well because of things like 10 discontinuities in the physical models, numerical 11 -oscillations, . etcetera.
| |
| 1 12 1]R.'WALLIS: Or it's too stiff or something, it i
| |
| 13 bounces around. I l
| |
| 14 MR. KELLY: In effect, exactly. Yes, exactly.
| |
| () 15 16 DR. WALLIS: I got the impression from Farouk that you were waiting for computers to get better, when in fact, 17 there are things you can do to improve the run time.
| |
| la MR. KELLY: There are things we can do. We're 19 just not doing them yet.
| |
| 20 MR. ELTAWILA: In the consolidation effort, one of 21- the important tests that we are running that we should get 22 in all testing between the old version of the code and the 23 current version.
| |
| 24 So, we are trying not to make measured i ,provement 25 in the code in the phase one, at least, we will try to l
| |
| l l
| |
| i l ANN RILEY & ASSOCIATES, LTD.
| |
| <1 '
| |
| Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014
| |
| , Washington, D.C. 20036 l
| |
| (202) 842-0034 I
| |
| | |
| l J
| |
| l 222 I I recover the existing capability. That's one reason. l 2 The second reason is obvious, that we want to have l'']T L..
| |
| 3 the money and there are not enough people to start working 4 on all these things that we want to do right now.
| |
| 5 DR. WALLIS: You said three things. One is ;
| |
| 6 parallel processing. The other one is time step. What was I 7 the third one?
| |
| 8 MR. ELTAWILA: Physical modeling.
| |
| 9 MR. KELLY: It's still time step but two different 10 ways.
| |
| 11~ One is in increasing the level of implicitness, 12 stability requirements, etcetera, and the other is reducing )
| |
| i 13 what I call numerical events, things that decrease the time ;
| |
| 1 14 step because of, like Professor Wallis said, it being too (g) - 15 stiff and numerical oscillations, etcetera.
| |
| 16 So, really, upping the time step is the best way 17 you can go, if you can do it.
| |
| 18 MR. ELTAWILA: The measure of success that we are 19 using, too, is to conserve resources and investment and user 20 experience.
| |
| 21 Most of our user are RELAP5 users, and we will try 22 to make the transition for them very smooth, and by that, we 23 will be able to separate the input modeling from the 24 computation engine for TRAC, so a RELAPS code can be read by 25 the graphical user interface, restart file which will be l
| |
| !(~'y ANN RILEY & ASSOCIATES, LTD.
| |
| !(_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 223 1 connected to the TRAC-B code.
| |
| /^%i (J 2 So, that will make the experience in using the !
| |
| 3 RELAP5 less for those people, and so, we've provided the 4 separation of the computational engine from the input 5 process.
| |
| 6 That's ongoing right now and will be completed in i
| |
| 7 February of this next year, and that translation will be 8 started when we have some funds for that.
| |
| 9 Again, in conserving resources, we feel that, when i 10 we consolidate the code, we will minimize the nember of code 11 assessment that we have, we minimize the code maintenance, 12 and we will -- in preparation, input deck will be cut by 13 one-third. !
| |
| l 14 DR. WALLIS: Is there some sort of customer
| |
| [ \
| |
| ( ,) 15 support system for these codes? Commercial codes, you have
| |
| {
| |
| j 16 code developers and you have an enormous army of customer 17 consultants who help people actually use it.
| |
| 18 MR. ELTAWILA: We have our code developers at 19 Scientech and hcve what you call customer support or user 20 support, so when the user has a problem, they call and they i 1
| |
| 21 try to find out what's the cause of the problem, there is a 22 fix that can be made immediately, we will give it to them.
| |
| 23 DR. MALLIS: Who pays for all that?
| |
| 24 MR. ELTAWILA: That's part of our maintenance 25 cost, and for example --
| |
| l
| |
| [)
| |
| s' ANN RILEY & ASSOCIATES, LTD.
| |
| Court' Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| _.m. . . _ _ - _ . _ . _ . . . - - . _ . . _ . - . _ . - - _ . _ ~ . . _ . . . _ . . ._y_ _ . . . _ .
| |
| 224 1 DR WALLIS: So, you have to pay for that?
| |
| MR. ELTAWILA: I pay for that. Other U.S. Users, L
| |
| 3 if.the problem'is large problem and require a lot'of changes .
| |
| '4 to the code-and is not caused by.something wrong in the code
| |
| '5 itself but.is caused by their own application of the code, l 6 then we can ask them to pay for it.
| |
| 7 DR. WALLIS: I don't see why Research should pay 8 for customer support.
| |
| 9 MR. ELTAWILA: That's part of the charter that we
| |
| ^ 10 have in maintaining this code.
| |
| 11 DR. WALLIS: Okay.
| |
| 12 MR. KELLY: If I may, this is Joe Kelly again from 13 Research.
| |
| 14 We don't have a large team of people sitting at
| |
| ) 15 telephones. It's just there's a telephone number that can 16 come in and be routed.
| |
| 17- But even then, most of the contact isn't through 18 telephone anymore but it's using the internet, and there is
| |
| -19 a RELAP5 and there will soon be a TRAC web-page that the 1
| |
| 20 users can log in on to submit trouble reports, and the 21 trouble reports get put in -- there's actually a database 22 where you can check -- you know, search on different 23 problems and see what's being worked on and whether yours i 24 has been i im or not.
| |
| R25 DR. ZUBER: This will be then available to the j l.
| |
| F ANN RILEY & ASSOCIATES, LTD.
| |
| ' []/
| |
| }
| |
| \_ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036 ,
| |
| ; (202) 842-0034
| |
| | |
| 225 1 intervenors, too.
| |
| -[~ ) 2 MR. KELLY: Sure, yes.
| |
| . Gi 3 MR. ELTAWILA: So, that's what we've been doing in 4 the area of code integration or consolidation.
| |
| 5 We have, as you recall, in the thermal hydraulic 6 five-year plan, we identify some internal facility that we 7 are maintaining, mainly the Puma facility and the Apex 8 facility, and we continue to run tests in these facilities.
| |
| 9 We're trying to get as much data out of them as possible 10 while we still have funds.
| |
| 11 In the separate effect tests, we are -- initiated 12 a program on redundant heat transfer.
| |
| 13 DR. ZUBER: How big is that problem money-wise?
| |
| 14 MR. ELTAWILA: The redundant?
| |
| >,-~~
| |
| () 15 DR. ZUBER: Redundant.
| |
| 16 MR. ELTAWILA: That's at the end of the program, 17 will be costing $3.8 million.
| |
| 18 AUDIENCE: I think it's 2.7 over five years.
| |
| 19 MR. ELTAWILA: 2.7?
| |
| 20 AUDIENCE: Yes.
| |
| 21 MR. ELTAWILA: We received a report from Penn 22 State which lists the literature review, scaling study, the 23 result of the scaling study, the instrumentation 24 requirement. We have sent the report for peer review.
| |
| 25 We are getting the results from the peer review I
| |
| (7
| |
| \, -)
| |
| ANh RILEY & ASSOCIATES, LTD.
| |
| _ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
| |
| | |
| 226 1 right now, and Penn State is in the' process of addressing 2 these, and once these peer review or comments has been 3 incorporated in the report, we'll issue it in final' form.
| |
| 4 We-are doing - have very big effort in the branch
| |
| '5' here, one person that's doing computational fluid dynamics,
| |
| -6 but.we have two separate program, one at RPI and one at 7 . Florida State University, to start to develop two-phase flow ]
| |
| 8 modeling for CFD codes.
| |
| 9 DR :. ZUBER: Who is doing it at RPI?
| |
| 10 MR. ELTAWILA: Padowski, 11 DR ~. ZUBER: And the two-phase flow models?
| |
| 12- MR. ELTAWILA ': That's what Padowski is doing. The 13 interfacial tracking model -- that's been done at Florida 14 State University, and the name, Angi.
| |
| (v ) 15 AUDIENCE: University of Florida. l 16 MR. ELTAWILA: University of Florida. I 17 DR. ZUBER: How big is that program?
| |
| 18 MR. ET.TAWILA: They are a small program, about 19 $140,000 pei fear, 148 or something like that.
| |
| 20 DR. KRESS: Farouk, what happened to the interface J 21 transport model?
| |
| 22 MR. ELTAWILA: The next view-graph about some of I 23 the accomplishments -- that will be mentioned.
| |
| 24 DR. ZUBER: What is this interfacing tracing 25 model?
| |
| l.
| |
| l-l j''} ~ ANN RILEY & ASSOCIATES, LTD.
| |
| (- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| _..___.m.__ _ ______ _ .. -
| |
| i 227 1 MR. ELTAWILA: That's'a spelling error.
| |
| : 2. DR. ZUBER: What~ is the difference'between that
| |
| ( !
| |
| ~3 work and the work at Purdue?
| |
| , f' 4 MR.' KELLY: Joe Kelly again.
| |
| -5 One-thing is that,.at'Purdue, you're looking more 'I
| |
| ~
| |
| .6 like the traditional -- more like one-dimensional, two-phase 7- flow.
| |
| 8 The work at Florida is three-dimensional using a 9 CFD code, and what it's trying-to do - 'I don't remember 10 whether it's a density function or a level sets-type
| |
| '11 . formulation, but it's to;put. values with each node as to how 12 much water is there and then come up with a level between 13 that and then write relations for the movement of-that 14 level, f) 15 So, it's the propagation of a front through a mesh 16 ofLCFD cells. It's' entirely numerical.
| |
| 17 And the work at Purdue is mainly experimental, and )
| |
| i 18 that's to develop a database on interfacial area and the I 19 rate at which interfacial area changes as you evolve --
| |
| 20 spatial-involvement, not temporal -- and then the
| |
| : 21. development of the fundamental models for what's causing the 22 . changes in the interfacial area, typically bubble
| |
| : 23. coalescence due to -- like slug-type bubbles overtaking the h4 smaller bubbles.
| |
| 25 DR. ZUBER: Why are these two programs are going
| |
| [.
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| -Washington, D.C. 20036 (202) 842-0034
| |
| | |
| L l
| |
| I 228 ;
| |
| l 1 to be integrate. .a the code, because they use different
| |
| ' / 2 approaches?
| |
| l 3 MR. ELTAWILA: Lact year, the Commission asked us 4 to look for some innovative ideas in research and fund some i
| |
| L 5 of this research.
| |
| 6 Whether they, at the end, will be utilized or not, 7 sn3 are at the initial stage right now, so we really don't 8 know how we're going to use them or if they are going to L 9 produce'any valuable results.
| |
| 10 So, they are cutting edge research to look at 11 different ideas.
| |
| 12 I-am listing here some of the accomplishments of 13 last year,_so I'm going to go through them quickly. j l
| |
| 14 We focused most of our effort on this element
| |
| '(,.
| |
| O) 15 that's important for the consolidation. We completed the 16 modernization of the TRAC-B data structure, and we call it ,
| |
| l 17 now TRAC-M, and as I mentioned,' most of the funds came from 18 CAPO.
| |
| 19 DR. WALLIS: That's just a data structure, though, 20 isn't it?
| |
| 21 MR. ELTAWILA: That's the data structure and ;
| |
| 22 converting it to FORTRAN lines, yes.
| |
| 23 Actually, CAPO paid for the 1-D component, and the 24 lutC paid for the 3-D component, so we should get credit for 25 what we're doing, too.
| |
| I A
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| /''1 (s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| (
| |
| l 229 I i
| |
| 1- We. revised the solution'for -- !
| |
| E l l2 DR. ZUBER: How many equations have been in the b, / --
| |
| i 3 CAPO model? .I heard hearsay there were more than six.
| |
| .4 MR. KELLY: In the TRAC version that is used by us 5 -and CAPO, it's the standard six-equation model with the 3-D I
| |
| 6 vessel, but CAPO also supports more fundamental CFD-type l
| |
| 7 work, I believe also at RPI. Is that right? And that's l 8 where they have like'seven different fields or something.
| |
| 9 DR. ZUBER: Okay.
| |
| 10 MR. KELLY: They have different size bubbles, i
| |
| : 11 different size droplets. ,
| |
| 1
| |
| ' 12 MR. ELTAWILA: We revised the solution procedure I l l 13 to support the RELAPS capability, because Joe might mention t 14 some differences between RELAP and. TRAC in terms of-l O t
| |
| V -15' .nodalization and solution procedures. So, we have changed 16 that so that eventually we'll be able to read RELAPS that 17 can solve the problem.
| |
| 18 DR. WALLIS: What happens to RELAP5 down the road, ;
| |
| l 19 when everything is consolidated?
| |
| ! 20 MR. ELTAWILA: We still have a transition period,
| |
| [ 21 and we will have assessment to go through, and if everything 1
| |
| 22 -- if the consolidated code looks much better than the L 23- .RELAP5 code, we'll archive the RELAPS code. But that's yet I l'
| |
| L 24 to be seen.
| |
| 25 At the end of next year, we will have a ,
| |
| l 4
| |
| {k ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i u- . - .. -- - .. .- - ._ . .- . .. - - - . . .. - . .
| |
| | |
| . . ,_.m _. , _ . - . - _ _ _ . . . . . - _ _ _ . . - . _ . _ . ._
| |
| 230
| |
| 'l consolidated version of-the TRACP, TRAC-B, and RAMONA, and
| |
| () 2 we, unfortunately, will not have too much assessment, but'at that time, we will know how successful we are, but we feel
| |
| ~
| |
| 3 4 that we are in the right track, because as Joe will present,
| |
| : 5. one of his presentations, it looks like the process that we 6- are following is-a good process. !
| |
| 7 We enabled TRAC-M to be able to do stability )
| |
| 1 8J analysis. It was implemented for the 3-D component vessel '
| |
| 9 - but now it'u implemented for the 1-D component, too.
| |
| 10 We are improving the method of passing information 11 between the component --
| |
| 12 DR. WALLIS: .You've enabled the implicit method, 13 so you have actual evidence that it works, you have some l l
| |
| 14 sort of performance evaluation of this thing you've enabled? l
| |
| () 15 MR. ELTAWILA: I think there was certain test that 16 was run by Professor Mahaffey. That work was done by 17 Mahaffey, and some of the test problems that he looked at 18 proved to him that the semi-implicit is working.
| |
| 19 DR. WALLIS: You could have, for some purposes, 20 some sort of gee whiz before/after slides that said this was 21 TRAC before we did this and this is it afterwards, that was 22 the kind stuff'it produced before we did it, and it took 10 23 hours, and this what we have now, and it takes one hour.
| |
| 24 MR. ELTAWILA: We will have some of that now, but 25 I promise you, if we finish wd'.hin two hours what we i
| |
| I
| |
| 'D ANN RILEY & ASSOCIATES, LTD.
| |
| ,k/
| |
| s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| __ ,. , _ _ . _ . . _ ~. .,, ,,
| |
| | |
| 231 1 prepared for you,'I will come up with this information next l''I 2 time.
| |
| O 3 DR. WALLIS: It would be nice to.have. You know, 4 these words are nice, but if you actually have evidence, !
| |
| 15 before and after type, that.would be good.
| |
| i 1
| |
| 6 MR. ELTAWILA: Yes. This is just introductory 7 remarks. Hopefully, some of them will be supported later.
| |
| 8 DR. WALLIS: Maybe it's all coming.
| |
| 9 MR. ELTAWILA: We are improving the method of 10 passing information between components so we can modularize 11- the code further and we can couple'it to external components 12- like, for example, the CVT model or the PARC coupling and 13 things like that. l 14 We have completed'the deve'.opment of the code
| |
| () 15 assessment package which-will help the developer actually 16 run multiple cases and use the-numerical scheme that was 17 provided by Penn State to give you a measure of goodness of 18 the code between the old version and the new version, so you 19 run the assessment, it's all maintained in the data bank, 20 and you compare.all the results with the new results and see 21 if you have improvement or deterioration in the performance 22 of your code.
| |
| 23 This can be run less expensively than what's being
| |
| ! 24 done right now, and we coupled the PARCS code to both RELAPS 25 and TRAC-M, and there will be a presentation on that today.
| |
| t l
| |
| [~'
| |
| \-
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| . Court Reporters
| |
| . 1025 Connecticut Avenue, NW, Suite 1014 I
| |
| Washington, D.C. 20036 (202) 842-0034 1
| |
| l .. - . _ . _ , --
| |
| | |
| i 232 1~ DR. SCHROCK: Does this "M" designation mean t9
| |
| .\_/.
| |
| : 2. modern or what?
| |
| 3 MR. ELTAWILA: "M," modern.
| |
| 4 DR. WALLIS: Modular or modern?
| |
| '5 MR. ELTAWILA: Modern.
| |
| 6' DR. WALLIS: "P" is prehistoric or something?
| |
| 7 MR. ELTAWILA: We have completed the development 8 of processor graphical user. interface called SNAP, and there 9 will be a presentation on that.
| |
| 10 We've completed the development of an expert
| |
| : 11. system that would be linked to SNAP to help the user develop 12 an input deck from scratch or diagnose a problem that they 13 have with the input deck.
| |
| 14 We've completed the installation of the turbine g
| |
| t y ) .
| |
| 15 component which is taken from the TRAC-B and put in the 16' TRAC-M, and you will get a presentation on that.
| |
| 17 We've completed the experimental facility, built 18 the facility, 'and we'll' start collecting data.
| |
| 19 DR. WALLIS: Turbine component? What is that? Is 20 that a misprint?
| |
| 21 MR. KELLY: The code is for the boiling water 22 reactor version.
| |
| l 23 DR. WALLIS: But you're actually analyzing the i
| |
| L 24' turbine?
| |
| ! 25 MR. KELLY: It's one node.
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| 1 \(%- - Court Repor:ars 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 4
| |
| y v. - - - . . . .- , --
| |
| a ,
| |
| | |
| 233
| |
| : 1. DR. WALLIS:. I thoughtLyou were analyzing the flow I~h 2 in the turbine using TRAC. So, it's not a major item, b
| |
| 3 MR. ELTAWILA: We completed the test facilities at 4 Purdue University which' measured the interfacial area-5- transport in vertical pipe, and at Purdue, for horizontal
| |
| .6 pipe, and. started getting data on that.
| |
| 7 Some of the tests in sub-cooled boiling'at low 8 pressure at UCLA -- we're not having a presentation on that 9 .today because of-limited time.
| |
| 10 That is my_ introductory remark. I would like to 11 invite Scientech to give you a presentation.
| |
| 12 DR. ZUBER: I had a problem last time, and I 13 didn't see it in your hand-out.
| |
| 14 You.have.many components, many activities at I 15 'different places, and I didn't see.the picture, how this is 16 integrated, who is really calling the integration -- I mean 17 how this activity at Purdue or in Florida and Penn State, 18 everything, and especially with Mahaffey, how this is 19 integrated that you can meet the schedule by end of next
| |
| '20 year or two years from now, a program, kind of a flow-chart 21 with some dates, and I didn't see this last time, and I 22 don't see it now, and I wondered whether you have it.
| |
| 23 MR. ELTAWILA: We have a program. We have people.
| |
| 24 We know when deliverable is coming. We usually have one 25 coordination meeting with all che people involved.
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| "O- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1 .
| |
| | |
| l l
| |
| 234 1 1 DR. ZUBER: Everyone-is going to put something, !
| |
| 2 but there should be one honcho, I mean somebody -- program-3 manager responsible for integrating for delivery.
| |
| l 4 MR. ELTAWILA: You asked that question last time.
| |
| l I was not able to answer you, but I will try to be honest
| |
| '6 with you.
| |
| l 7 I'm getting help from everybody in the branch, but i i
| |
| 8 I am-under a lot of restriction here about assigning, having 9 somebody -- the best one to do this effort would be somebody 10' like Joe.
| |
| , 11 Unfortunately, because of the agency procedure, we 12 cannot assign responsibility to him that has advisory.
| |
| 13 function, and that's why I'll keep saying that I am doing 14 that work, but in reality, the help that I'm getting from l) 15 ' Joe, _from Jennifer, and other people inLthis area is helping 16 the program,.but I cannot nominate him to assign work to 17 people and things like that.
| |
| 18 This is a problem that I have at NRC. Our way of 19 doing business is this way. ,
| |
| L 20 DR. ZUBER: I have no problem in who's assigned.
| |
| l 21 I would like to know who is responsible and how these 22 different parts -- how do I know that the model which Ishii 23 or Cojal or Florida or the models which Penn State is going l 24 to produce is going.to be really utilizing that form in the L 25 code.
| |
| Tw O .
| |
| -ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut' Avenue, NW, Suite 1014 Washington, D.C. 20036 o
| |
| (202) 842-0034 u . .
| |
| | |
| j i
| |
| 235 l'
| |
| You know, you have these separate activities, and I''h 2 I really didn't see how this is going to give.us a pie, O
| |
| 3 pudding.
| |
| 4 MR. ELTAWILA: I think that's a good question, 5 first of"all that we don't have enough staff to review all i 6 this work.
| |
| 7 We thought that we had them,.but because of this 8 budget cut and the people leaving the agency caused us some '
| |
| 9 problem that we are really not overseeing averything that we 10 .have.right now, i
| |
| 11 Our control is through the -- what we call the 12 statement of work. We'll try to provide the statement of 13 work as detailed as possible.
| |
| 14 For example, the work that Professor Ishii is
| |
| /~N (g ) 15 doing -- we explicitly say that the model that's being 16 developed from that work has to fit in the two-fluid model' 17 that's employed in the TRAC-B code. So, that's about the 18 only guidance that we can give them right now.
| |
| 19 We don't want them to develop a model that will 20 require data that is not generated by the rest of the code. '
| |
| 21 So, that's as much guidance as we give them.
| |
| 22- DR. ZUBER: And echeduling?
| |
| 23 MR. ELTAWILA: Scheduling is controlled by the 24- statement of work, and in general, most of the contractors l
| |
| 25 meet the schedule.
| |
| b ANN R.' LEY &. ASSOCIATES, LTD.
| |
| , %- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
| |
| t I.
| |
| | |
| .=. -.-
| |
| 236 1 MR. KELLY: I think the situation is actually a l
| |
| [~) 2 little bit rosier than that.
| |
| V 3 Remember, we're talking mainly about code 4 consolidation and then a secondary phase being code 5 improvement, and things like the experimental programs that 6 you mentioned will feed into the code improvement. So, for 7 the moment, we're more tightly focuse.d on code 8 consolidation, and that is, indeed, planned, with schedules 9 and budgets, etcetera.
| |
| 10 So, one of the documents that I'm pretty sure we 11 gave to the ACRS about a year or so ago was the TRAC BWR 12 consolidation plan from Scientech, and if you look in there, 13 there is, you know, these are the components we have to i 14 consolidated,-these are the ones we're going to do first, A
| |
| () 15 this is the milestone date for them, and so, everything 16 that's on the critical path we're more tightly focusing and 17 figuring out.when's the best time to start it so it doesn't 18 interfere with other work and when do we have to have it 19 done so it doesn't cause something else to be postponed.
| |
| 20 So, it's better than it sounded.
| |
| 21 DR. ZUBER: Last question. Suppose that the funds 22 are going to be cut or may be cut. What will be then the 23 tool available -- the best tool available for NRC to perform 24 its future activities?
| |
| 25 MR. ELTAWILA: If it's completely dried out, I'd l
| |
| (N,s' ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| l I
| |
| 237 1 still strongly encourage NRC to consolidate the codes, but
| |
| -\f'',}' 2 while we are doing_the consolidation, we still have the 3 RELAPS code, l
| |
| 4 DR. ZUBER: Suppose somebody says I don't need i 5 this. You are still going to use some codes in the future.
| |
| 6 What is the tool NRC would have two, three, or five years 7 from now to use?
| |
| 8 MR. ELTAWILA: If we don't have any funds right 9 now, would be the RELAPS code.
| |
| 10 DR. ZUBER: RELAP5.
| |
| 11 MR. ELTAWILA: Uh-huh. ;
| |
| 12 DR. WALLIS: That's why RELAP5 is a kind of 13 insurance policy.
| |
| 14 MR. ELTAWILA: It's an insurance policy. We are f')
| |
| , 15 continuing to improve it, and we're doing assessment, and 16 -we're coupling it with other codes and so on, but it was the 17 wrong choice for the consolidation. That's the problem with 18 it.
| |
| 19 Ben?
| |
| 20 MR. GITNICK: Good afternoon. My name is Ben 21 Gitnick. I work for Scientech, Incorporated. As you know, 22 Scientech is the contractor for some of the methods 23 development and modernization work, and the piece that I am 24 responsible for is the development of the SNAP GUI, 25 graphical user interface.
| |
| ']'~}
| |
| \,_ /
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters l
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| - .. . ~ . . .-. ..- . . ~ . _ . - . . . _ . _ - . - . . - . ~ . .
| |
| t 238 i j
| |
| 1 DR.- ZUBER: Do you have a hand-out?
| |
| {
| |
| ,2 MR. GITNICK:
| |
| .Yes. There should-be some on the 3 table.right there in_ front of you.
| |
| i 4 -As you all know,1 the NRC is developing a graphical. '
| |
| :5
| |
| ~
| |
| user interface. The early portion was focused on providing i n.
| |
| i 61 a1 user interface for RELAP5, and now we're switching our l
| |
| ~7 ' focus over to TRAC-M.
| |
| 8 DR. KRESS: -Is-SNAP an acronym?.
| |
| 9 MR. GITNICK: Yes,. it.is. It stands for' Symbolic 10 Nuclear Analysis Package.
| |
| .11. You've all been told about its intentions, and 12 'certainly, one'of the main things is to' increase the ease of i
| |
| 13 use. I 14' -It's1al'so supposed to allow us to recover this
| |
| " O(j 15 .large inventory of decks that we have, this tremendous 16 knowledge-base that we have out there in RELAP decks and l '17 being able to convert them, as'much as possible, over to 18 TRAC, TRAC-M basis.
| |
| 19' Work on the RELAPS interface is mostly complete.
| |
| 20 . Essentially it's complete. We're just putting out for beta
| |
| '21: -tests now, and as we get comments back from our beta L . .
| |
| l 22- testers, we have things to correct and holes to patch, but 23 we're very' happy with this.
| |
| r 24 DR. WALLIS: What does beta testing mean?
| |
| 25 MR. GITNICK: Well, traditionally, alpha testing I
| |
| i i
| |
| . ANN RILEY & ASSOCIATES, LTD.
| |
| N)
| |
| (s- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| ! Washington, D.C. 20036 bc (202) 842-0034
| |
| . , . . ., . =- , - - - - . - . .. - -- . -
| |
| | |
| , . .- -. - _ - . . . - - . . . - . . - - - ..- - .- - ~~.
| |
| l 1
| |
| i 239 1- means that it's going to blow up in your face so often that f(~'t. -2 only a developer can stand to run it. After a while, it
| |
| : y/.
| |
| 3' doesn't crash that often, and you can put-it out on the poor 4 public, let them try it, and they can give you their 5 feedback.
| |
| 6 When you're developing a user interface, it's 7 extremely important to get beta testing, because a good.
| |
| 8 graphical interface makes things much easier, and a bad one l
| |
| 9 makes things'much harder.
| |
| )
| |
| 10 It can txa more awkward and confusing than just not 11 having a user interface.at all. So, it's very important to' 12' get that user feedback.
| |
| 1 13 DR. WALLIS: So, it's already passed your alpha 14' testing.
| |
| ) 15 MR. GITNICK: -Yes. We've had a number of 16 different versions. This is not the first time it's gone 17 out to beta testing, but this is the first time it's fairly 18 in a useable state that's gone out. l 19 Work on incorporating an interface for TRAC-M has 20 just begun.
| |
| -21 Now, TRAC-M, as you heard, is a bit of a moving 22 target, but we are going to be working tightly with the l 23 people from Los Alamos and the other parts with teams at 24 Scientech and the people at RES to define and develop the 25 TRAC-M interface.
| |
| J ;
| |
| I I
| |
| i
| |
| '[ ANN RILEY & ASSOCIATES, LTD.
| |
| \ Court Reporters !
| |
| , 1025 Connecticut Avenue, NW, Suite 1014 i l Washington, D.C. 20036 l l
| |
| t (202) 842-0034 l
| |
| | |
| l l
| |
| 240 1 1 DR. ZUBER: I'm sorry. I have a question. I i
| |
| (~'T 2 thought that TRAC-M was being developed at Penn State.
| |
| \)
| |
| 3 MR. GITNICK: Dr. Mahaffey and his team are 4 working on TRAC-M, that is correct, but there's work also 5 going on at Los Alamos.
| |
| 6 MR. ELTAWILA: The TRAC-M was developed initially ;
| |
| i 7 by Los Alamos, but Professor Mahaffey have done some work --
| |
| 8 for example, the systems server, the numerical scheme, some i
| |
| 9 of these things that I identify in the achievement were done 10 by Professor Mahaffey, and now, as part of the consolidation I 11 of the TRAC-B component, Scientech is also involved. So, we 12 have three contractor, would be Scientech, Penn State, and 13 LANL involved in developing model and improving the existing 14 capability of TRAC-M.
| |
| O 15 DR. ZUBER:
| |
| () If I may ask, what is the division of 16 funding between these three activities?
| |
| 17 MR. ELTAWILA: I think Scientech is about 18 $600,000, LANL was about $800,000, and Professor Mahaffey is 19 close to $400,000 per year.
| |
| 20 DR. ZUBER: Who is doing the work at LANL?
| |
| 21 MR. ELTAWILA: It's Skip Deering, but it's under 22 the direction of Brent Foyak.
| |
| 23 DR. ZUBER: Okay.
| |
| 24 MR. GITNICK: A lot of the work that's going on at 25 Los Alamos, as I'll discuss later, is separating TRAC into a i
| |
| ! [~S ANN RILEY & ASSOCIATES, LTD.
| |
| (_- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l (202) 842-0034 l
| |
| l
| |
| | |
| 241 <
| |
| 1 front-end and a compute engine. So,.it's more mechanistic.
| |
| l 2 So,.the development for use of TRAC-M has just
| |
| [~')E
| |
| \_ .
| |
| 1 l
| |
| 3 begun.
| |
| 4 Our goal -- and there was just a discussion of 5 goals and schedules -- is to have a version ready for beta 6' test at the end cf. calendar year 1999.
| |
| 7 This is just a preview. It's a-look at what'the 8 SNAP interface generally looks like,'and what I'm showing i
| |
| 9- here is one of the workshop problems called problem two.
| |
| 10 This is a very -- as workshop problems go, it's 11 one of about moderate difficulty, but it's fairly complete.
| |
| 12 It has a core, in this area, pumps, a primary lube, a 13 pressurizer. .These are time-dependent volumes.
| |
| 14 In the SNAP interface, you select components, and f
| |
| ( 15 there's a pallet of-components. You can click on them and 16 then drop them onto the drawing canvas.
| |
| 17 Then you can connect them using junction
| |
| .18 components, and each time you do that, it will ask you for 19 information about their physical properties, their size, 20 their area, and later on, it will give you a chance to 21 nodalize those pipes and volumes and provide nodal 22 information, if you will.
| |
| 23- Some of.those processes, of course, we hope to 24 ' provide a series of expert assistance to help them step 25 through the more complicated model-building parts.
| |
| I-l' i,
| |
| t ANN RILEY & ASSOCIATES, LTD.
| |
| ;* Court Reporters
| |
| ! 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034
| |
| | |
| I 242 1 DR. WALLIS: These numbers that overwrite the f~'). 2 details of the picture are in different colors or something, L,1 3 or is that an oversight? ,
| |
| 4 MR. GITNICK: Which ones?
| |
| 5 DR. WALLIS: Well, look at the fourth -- what is 6 that? I can't ever see what it is. I mean 210, is it, l 7
| |
| that's written over the crosses and the -- whatever it is. .
| |
| l 8 The numbers are over the details of the components, so you 9 can't see them. l 10 MR. GITNICK: Well, first of all, these aren't 11 details of components. These are -- it's an icon. ;
| |
| 12 Basically, it's a enrtoon.
| |
| 13 DR. WALLIS: But the numbers should be outside the 14 --
| |
| g- s :
| |
| ( ) -15 MR. GITNICK: Well, it gets very complex if you w/
| |
| 16 start -- which way do you put it, to the left or right?
| |
| 17 DR. WALLIS: That's up to you guys to figure out, 18 but I've got to read it.
| |
| 19 MR. GITNICK: It looks much easier on the computer 20 screen, and I do have a live demo. This, unfortunately, is 21 not the best projector here. I have a live copy of the code 22 on the little lap-top here.
| |
| 23 We're initially developing strictly in UNIX, and 24 we do have versions now for six different flavors of UNIX, 25 but from the initial inception of this code, we had designed l
| |
| l.
| |
| ''N, ANN RILEY & ASSOCIATES, LTD.
| |
| t ) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| l
| |
| | |
| 243 1 rit to_be cross-platform, and we're very proud of the fact '
| |
| L 'Q/ Il 2: that' itJnow comes up in Windows, both NT and 95, and it has 3 the same look and feel, g 4 DR. WALLIS: A smart sophomore _could simply drag 5- _these things- over, hitch them up,. and say nu1? !
| |
| -6' MR. GITNICK: Except-for.the fact.that you'd have i 7' to tell it what area and length, yes.
| |
| 8 DR. WALLIS: Well, it would probably:ask you. ;
| |
| 9 MR. GITNICK: It will ask you. That is exactly 10 correct. And you can tell it -- I won't fill it in-now, but 1 11 of course, before you start run, it should ask you again,
| |
| '12 but yes, it's essentially what it is. It's just a graphical 13' ' user-interface.- This is the pre-processor portion of the 14 code.
| |
| 15 DR. KRESS: What do you do about things like M[ U.
| |
| t16 -elbows and T's?
| |
| 17 MR. GITNICK: Okay. This is what we call the 18- RELAP interface, and of course, it uses traditional j j
| |
| 19 RELAP-type approaches to pipes and volumes. There is going j 20 to be a TRAC approach which uses traditional TRAC 21 components.
| |
| 22 The actual information stored in the code is not 23 stored by RELAP component or TRAC component, it's stored by i 24' physical properties, and we intentionally designed it that 25 way so that what_we're working towards in the long term is a l
| |
| I
| |
| 'Q
| |
| ;. \s ,/.
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters L 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 (202) 842-0034 L
| |
| | |
| l
| |
| )
| |
| 244 l 1 physical generic view where a user won't be dragging off a
| |
| /"' 2 RELAP pipe or a RELAP pump, he'll be dragging off a
| |
| 'v) /
| |
| 3 centrifugal pump or a Schedule 40 pipe or an elbow or a T of j i
| |
| 4 a certain type of fitting, and he'll actually be assembling !
| |
| 5 a system out of real components, and then the code will know ]
| |
| 6 enough to translate that either to a RELAP deck or a TRAC-M l
| |
| 7 deck, and that's where our long-term goal is.
| |
| 8 In fact, with that --
| |
| extremely in the 9 long-distant future, we ha,, other plans for it beyond that,
| |
| ~10 but you can understand that our goal is to get users away 11 from the constraints of the front-end that was built into :
| |
| 12 RELAP 30 years ago and looking more towards the physical ;
| |
| 13 system, he can concern himself with the physical aspects.
| |
| 14 DR. KRESS: And you're still doing it for RELAP
| |
| , m_
| |
| 15 just out of comparison?
| |
| )
| |
| 16 MR. GITNICK: Well, to get where we are today, it 17 took a lot of hard work, and we had to have something first.
| |
| 18 We couldn't go off and build the generic model before we had 19 a RELAP model, because we had to translate from one to the 20 other.
| |
| 21 So, first, you build the small wing to the left, l
| |
| 22 then you build the wing to the right, and then you can build 23 the big house in between.
| |
| 24 DR. WALLIS: Why aren't there arrows on the 25 junctions?
| |
| /~% ANN RILEY & ASSOCIATES, LTD.
| |
| (,,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . ._ _ ._- - ._. . _ . ~ . _ - - . . . - _ _ _ ... _ - _. _.._. . ~ _..
| |
| c 245 1 -'~ MR. GITNICK: They indicate the defined direction O) .2 of positive. flow.
| |
| 'N.J
| |
| .3 DR. WALLIS: But really, the flow could go both 4' ways.
| |
| 5- MR. GITNICK: Correct. 'But the flow would be 6 consideredInegative if it goes against the' arrow. ,
| |
| P
| |
| -7 DR. WALLIS: Oh. So, it's just a cign convention 8 thing.
| |
| 9 MR. GITNICK: Yes.
| |
| 10' DR. WALLIS: Okay.
| |
| 11 MR. GITNICK: It indicates a positive sign
| |
| .12 convention.
| |
| 13 I don't need to read this to you. Obviously, you 14 have all these thoughts in your mind about what GUI is good 15 for, but I would like to draw your, attention to this third 16 bullet, the user ~effect. I think this is really a major 17 reason'for going to a GUI, 18 If you ask two engineers to build a model, it's 19 well known that they'll come up with two radically different 20 models which will give different results, and some 21- engineers, based on their experience, will get much better 22' results than others.
| |
| 23 DR. WALLIS: Those are the ones they show you.
| |
| 24' MR. GITNICK: Perhaps. But with this, we're 25 hoping to codify and build in some of that knowledge base, 5
| |
| ; ANN RILEY & ASSOCIATES, LTD.
| |
| , q, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l 246
| |
| 'l- so.that the code will just naturally tend to select the 2 proper options. The user will have a chance to over-write 3 those choices, but it will record the fact that the user 4 made a selection against the guidelines, j 5 DR. WALLIS: It think it would be a great day-when l 6 the optimistic vendor can run a code and get results and the 7- . pessimistic, critical NRC can run the same code and get the !
| |
| 8 same results. 'That will be the great day.
| |
| SF MR. GITNICK: I think it would be, too, but 10 perhaps differ by a margin which they' understand.
| |
| 11- DR. KRESS: Will this allow me to -- or someone to 12 run a -- the node size variation to.see --
| |
| ; 13 MR. GITNICK: Yes. i 14 DR. KRESS: And this will make it a lot easier.
| |
| =( 15 MR. GITNICK: It's one of the things the code is 16 set up exactly to do, and there are other features, too.
| |
| 17 DR. WALLIS: Do,you have representation so the 18 naive user can do something and then the expert can go to 19 some deeper level or something?
| |
| 20 MR. GITNICK: We don't have hidden user options.
| |
| 21 It's like Windows 95 where if you pull down enough windows 22 you'll find a check somewhere.
| |
| 23 Generally, though, a good GUI is one where you
| |
| : l. 24 don't have to go hunting and pecking for the strangest 25 window buried under three other windows that have no A' ANN RILEY & ASSOCIATES, LTD.
| |
| , k~s ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036 (202) 842-0034 l
| |
| l c - _ - . ,
| |
| | |
| 247 L1 relationship.
| |
| / ) 2 It's supposed-to be one that everything naturally
| |
| -(s,/- -
| |
| 13 flows from the.other, and that's what partially! beta testing 4 will tell us. -l 5 We've gone through several different. variations, 6 we've improved it quite-a bit, and quite frankly, it still-7 needs.a great-deal of improvement.
| |
| '8 DR. ZUBER: When did you start on this work?
| |
| 9' MR.:GITNICK: About a year-and-a-half ago.
| |
| : 10. DR. FONTANA: Can you estimate how much time it-11_ would'take a person to.do an input deck this'way compared to 12 -the old_way?
| |
| 13 MR. GITNICK: To start and do a whole deck such as c
| |
| 1 E
| |
| 14 AP600?
| |
| ' f%'
| |
| +j -15 DR. FONTANA: Yes. l I
| |
| 16- MR. GITNICK: AP600, of course, took-two years to-
| |
| .17 build.
| |
| 18 DR. FONTANA: Yes. i 19 MR. GITNICK: I would say it would probably take 20 about a month to two months this way.
| |
| 21 DR. FONTANA: A month?
| |
| 22 MR. GITNICK: Well, AP600 is a big deck. The 23 listing of AP600 -- it takes 400 pages, and it's 23.,000 24 lines.
| |
| 25 DR. KRESS: It would take you a month just to get l
| |
| !O ANN RILEY & ASSOCIATES, LTD.
| |
| ' ' \s / Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 .
| |
| l Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| 1 248 j l' the description of the --
| |
| .2. MR. GITNICK: A simpler deck, like a typical PWR,
| |
| )
| |
| 3 which many of you are familiar with, you could probably do
| |
| ~ '4 that inside of a week.
| |
| 5 That is about 2,000 lines of input, and again, j i- 6 lines of input have no more meaning when you start going to i 7 a' graphical user interface. It's how many numbers do you 8 have'to know, how many numbers do you have to make 9 available, how many components do you have to describe.
| |
| 10 Here I find out what SNAP means. And it says 11 here, the preprocessor not only builds input model from 12 scratch, but it could also take in an existing deck, just 13 like in one of these -- any RELAP run. It will read it, it 14 will check it, it will display it on the screen and try and )
| |
| () 15 make some sense of it. That is quite an achievement because 16 RELAP does not know left from right. Things at the same 17 elevation, it doesn't know where they below, so have a lot 18 of heuristics in there to try and figure out how the network 19 lays out.
| |
| 20 It does a good job, it doesn't do a perfect job.
| |
| 21 I think we can improve it, but we are still pretty happy 22 with what we are doing so far. That is what we are working 23 'on right now, translate from RELAP5 to TRAC-M. Of course, 24 we are defining what TRAC-M is in the process.
| |
| 25 Besides the preprocessor, which we have all been l
| |
| l-l ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| ! 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 249 1 talking about so far, there are two other modules. There is ,
| |
| 2 a'run time.and post-processor. The run time is a
| |
| (~'/}
| |
| ~- j 3 convenience feature, it allows the user to select which !
| |
| 4 machine on the network is the best to run this case on, get j 5 status reports while it is running, have messages come back 1 l
| |
| 6 when it is run. Have it automatically archive the results j 7 for you, a number of other handy features.
| |
| 8 The post-processing is not intended to be a I
| |
| -9 monolithic piece of coding. The post-processing is supposed 10 to be a package of individual aplets that will allow the 11 user to do 2-D and 3-D visualization, different types of 12 data analysis. Basically, it is going to be written -- I 13 don't know many people here are really into computers, but a 14 large number of JAVA applications that wi'l interface with r% 1 15 the restarts, give a number of-different tools to allow the (s-)
| |
| I 16 user to do it, enhance visualization, so you can understand l 17 the results.
| |
| 18 DR. KRESS: You can watch it evolve it time then 19 like a movie.
| |
| 20 MR. GITNICK: That is correct. That is correct.
| |
| 21 One of the things that both the -- to a limited extent, the l 22 run time will do that also, but with the post-processor, you 23 can animate it. You can stop, you can replay. You could 24 put a mask up, if you have a drawn mask, and animate the 25 mask. So there are a lot of features that can be done.
| |
| /~' ANN RILEY & ASSOCIATES, LTD.
| |
| (, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 I
| |
| Washington, D.C. 20036 (202) 842-0034
| |
| | |
| J 250 1 Let's go on, just the next two slides, for a
| |
| (~N' 2 minute. I don't have to discuss this. The current input 3 deck method, for example, some of the larger models, the 4 newer models, getting very big. Harder for a human being to 5 even comprehend. The AP600 base deck, the one at least I 6 have is around 23,000 lines, that's about 300 to 500 pages, 7 depending on whether -- how you list it, what font, or 8 whatever you choose when you print it out. That's an awful D big deck for anybody to very knowledgeable about what is in 10 that deck.
| |
| 11 DR. WALLIS: Or even to check it.
| |
| 12 MR. GITNICK: Or to check it. You can see why it 13 takes years to build one, and then you are not quite sure.
| |
| 14 Even for the smaller decks, it is almost possible to check.
| |
| (A) 15 We have been running some small decks, a famous deck that 16 has been used for 20 years, throt t 3 NAP we have been 17 finding errors that have lain there 'ar years.
| |
| 18 The SNAP GUI has all the modern features that you 19 expect, which is drag and drop and tab dialogues, full 20 English descriptions of what these variables mean. Features 21 like undo. You are familiar with the tool set of a 22 graphical user interface. It makes modifications with 23 existing decks much easier and has extensive error checking, 24 and that is one of its strong points. And the models can be 25 stored orderly in the database. Let's go on to the next Q
| |
| (,,/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| \
| |
| l 251 1 one. i i-2- Okay. I just selected three examples. I won't 3 chave time to demonstrate all of them to you, but I will'try L 4 and get it to. quick demo. Example ~1 is to take a pipe L
| |
| -5 without heat structures, so it is a very simple pipe, and ,
| |
| 6 . change it from five to 15 unequal nodes. We had a couple of 7 ~ people try this', all of them pretty much expert users, and 8 itLtook anywhere'from an hour to an hour-and-a-half, roughly L 9 depending on who did it. The time using SNAP was about four l
| |
| 10 minutes.
| |
| 11- The next example, we are utilizing the same pipe
| |
| , 12 with heat structures. It took the expert user about. half-a l
| |
| 13 day. SNAP took the same amount of time, it didn't care
| |
| : 14. whether a heat structure is attached or not.
| |
| 15
| |
| ( And the last example was moving.the location of.a
| |
| : 16. pipe break, which in SNAP is just a disconnected junction 17 'here, plop it over there. SNAP took about three minutes.
| |
| 18 It took about 30 minutes in the deck. And, again, these are 19' for expert users. There are many people who aren't very 20 much up to date, it would take them considerably longer.
| |
| 21 .Okay. Can we bring up the demo? I am going to l 22 try and get a live demo. We started out, like I said 23 earlier, developing in UNIX and we still -- the UNIX version j
| |
| 24 is still our leading version, but we have recently gone and 25 imported it to NT. It runs Native on NT. Well, we can show ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025-Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034
| |
| | |
| 252 1 them by dropping things on the screen. I was going to ask 2 you for some geometry,.but I don't want to fill that in j 3 'right-now,. so.just close the box and drop. Okay. There's a l -4 junction. Drop a pipe. Let's see. A pipe grid to find it 5 in terms of segments, how.many of the same diameter or the 64 same roughness, or the scme axial, inclination angle are t
| |
| ?- considered segments. I thinx you told us there was only one
| |
| '8 . segment there.
| |
| 9 DR. WALLIS: What do you do with things like' tie 10- plates'or orifices?
| |
| 11 MR. GITNICK: This is basically the dictionary of 12 components. You have --
| |
| 13 DR. WALLIS: Orifice is a pipe of zero length with 14 isome --
| |
| ([ 15 MR. GITNICK: Okay. Why don't you complete the 16 loop there?
| |
| 17_ DR. ZUBER: What is the orifice?
| |
| ~18 MR. GITNICK: RELAP doesn't have an orifice, per 19 se, it is treated as K factor at a junction.
| |
| 20 DR. WALLIS: It.is a pipe.
| |
| 21 MR.-GITNICK: In a pipe you can put a restrictive
| |
| ~
| |
| -- 2 2 flow area and K factor, and internal junction, that would 23 model orifice. Go ahead, there's a loop. It is not a very 24 -- I would not run that model, since, if it is water solid, 25 that would probably give you all kinds of problems running f
| |
| I'
| |
| ,/~} ANN RILEY & ASSOCIATES, LTD.
| |
| \s,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| f-253
| |
| ; il it, but it.is a model. '
| |
| L. .
| |
| (( 2- -Now,.we would want to go back, determine the flow areas-and lengths. -But that is all it took to complete a.
| |
| 3
| |
| [
| |
| '4 loop. >
| |
| c j ;5 Let's.go ahead and import a model. It is under ;
| |
| 6 SNAP.
| |
| l-7 DR. SCHROCK: Presumably, that is a lot clear on l
| |
| l - 8 .. the screen.
| |
| t
| |
| ! 9 MR. GITNICK: You are welcome to look at that 10 screen. I'm sorry about the nature of the viewing machine.
| |
| .11 DR. WALLIS: Now, presumably, you can take some 12' element.in your circuit and blow it up, expand it, or 131 magnify and fill in details. This is fine if you.have six l 14 or seven subsystems. But if you have a thousand -- !
| |
| /
| |
| 4 15 MR. GITNICK: That is the problem with E16 ' demonstrations, then tend to happen a lot more at l 17 demonstrations than --
| |
| 18 DR. WALLIS: This should be perfect.
| |
| 19 MR. GITNICK: There you go. Okay. Now, this is 1
| |
| 20 actually two independent systems, obviously, a primary and
| |
| '21 secondary, with a simple model.
| |
| 22 Go ahead, back to full. I am going to show some 23 of the other features.
| |
| 24 It's organized by layer and with different layers 25 you can edit different things. The bottom layer is hydro 1.
| |
| l I ANN EILEY & ASSOCIATES, LTD.
| |
| . Court Reporters i . 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| _ _ . . . . __ . . . . _ _ ._. ~ _. _ _._ _ _ _ _ _. _.._.__. _ _ . _ _
| |
| i -
| |
| U L
| |
| l 254 1 and you can only edit the hydro when you are on the bottom
| |
| /"Ni 2 layer. Above that is heat structures; above that is trips U '
| |
| 3 .and controls; above that is kinetics.
| |
| 4 'We have not been asked to do.any work on kinetics 5 yet so that's-probably'the last one we will get to.
| |
| L 6 [ Discussion off the record.]
| |
| 7 MR. GITNICK: This is at 133 megahertz. For the 8 demo we were going to use a faster machine but it was l l- 9 . unavailable.
| |
| [ 10 'DR. WALLIS: I was asking earlier, could you take 11 one of these elements in there and go way and put more 12 details into it and then reinsert it? Is that relatively i
| |
| L 13 easy to do?
| |
| 14 MR. GITNICK: Okay. We're still changing from one
| |
| () 15 16 level to another.
| |
| 1 DR. SCHROCK: Could we hear the question again? I l
| |
| 17 DR. WALLIS: You seem to be going ahead. Can you 18' take an element in there and say I want to represent that as 19 'a more complicated' thing and then you can take it an element 20? .in there and say I want to represent that as a more 21 complicated thing?
| |
| l- 22 MR. GITNICK: Sure.
| |
| 23 DR. WALLIS: 'And then you can take it and work on l
| |
| ?
| |
| l 12 4 it and put it back in again?
| |
| 25 MR. GITNICK: Yes, there's copy and paste. You !
| |
| l t l L
| |
| *\
| |
| (~ ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| - 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 (202) 842-0034 l
| |
| t
| |
| , , ,r- .. , . , . , . . - - , , ,
| |
| | |
| 255 L 1 copy it to another area of the canvas.
| |
| ~2 DR. WALLIS: And you can expand it?
| |
| 3 MR.-GITNICK: You can expand it.
| |
| j '4
| |
| 'Actually, expansion is something we are working on 5 right now. You can take a pipe.anu you can break it up into 6- three smaller pipes. You can take a group of ringle' volumes -
| |
| 7 and combine them into one pipe. -
| |
| ' 1 8' DR. WALLIS: So it is easy to do that, go from one )
| |
| 9 level to another?-
| |
| ! 10 MR. .GITNICK: Correct, and that is something that
| |
| .11 we are working on currently. There is a certain function 12 there right now.
| |
| '13 Okay. Now why don't I click on one of those, l 14 Bill, just to show them. Okay -- if you want to raise that.
| |
| ,V!h .15 I'm sorry, this looks a little bit fuzzy, but what 16 .you see is when it parts it in the deck. It takes any cards
| |
| '17 associated with a component that it doesn't understand and 18 puts them in a comment field so these are the comments that 19 actually are scavenged from that deck along with the names 20 for the components and the numbers. RELAP of course uses a
| |
| -21 system of numbers to identify components just like you might 22 use " Fred"~or " Bill." They use "100" aud "10" or "200" and 23 "20."
| |
| 24 All the information is organized by tabs. This is 25 'the one that changes the angle of the different components.
| |
| I
| |
| ;: ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| ! 1025 Connecticut Avenue, NW, Suite 1014
| |
| ( Washington, D.C. 20036 0
| |
| (202) 842-0034 L
| |
| | |
| - . . . - . ~ . - . . . - . - . - . - - - _ . . - . - - ~_ - - . .-
| |
| 256 1 Take that big one and double click on the left 7'T 2 with it.
| |
| (b '
| |
| 3 Explode the first one with the geometric view.
| |
| This is a blowup -- no, go down.
| |
| 4 This is trying to show the geometry as it actually 5' exists in that piece of pipe. What you are seeing is that 6 it has.got two segments -- the pressurizer and it's got a l
| |
| 7 surge line underneath it.
| |
| 8 Now if these pieces angled around, you would see 9 that. The geometric view follows all the twists and turns.
| |
| L 10 Okay? Let's go in and renumber that one. Go up to 11 Renumber. How many is in there? Four? Make it 14.
| |
| l 12 DR. ZUBER: You can put tie plates in the vessel?
| |
| j 13- MR. GITNICK: Yes.
| |
| 14 DR. WALLIS: What's it doing now? Patching nodes?
| |
| ,( 15 MR. GITNICK: He just went'from 4 to 14 nodes. j
| |
| .16 That is fairly easy because these are evenly spaced nodes 1
| |
| 17 but go ahead and click on that one and move that node '
| |
| 18 around. Okay -- leave it there.
| |
| 19 Now he just changed where the node is.
| |
| 20 At the bottom -- you have to lift that more -- but 21 at the bottom is the dimensions. You can also enter an 22 exact number, so he just moved a node further down.
| |
| 23 You could also select a node boundary, a junction 24 boundary, and delete it, combining two nodes together and it 25 will do all the underlying math for you.
| |
| i (%,
| |
| ! ANN RILEY & ASSOCIATES, LTD,
| |
| (,f ' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l ' Washington, D.C. 20036 (202) 842-0034
| |
| | |
| -- - - _ . - . - . - . ._. - - . - . - . - . ~ . - - . . - - - . - .
| |
| I 257 '
| |
| i
| |
| .1- I guess that's pretty much enough of the 3
| |
| 2 demonstration -- we could show you bigger models'but it L R.
| |
| i ('')Y ' '
| |
| L
| |
| =3 .w ould just take longer on the small laptop, but it will run i
| |
| L 4- some of'the bigger models'we have and perhaps the-next time l i
| |
| 5 I get a chance I can'show you more. i 6 DR. SCHROCK: In actual systems you-might have 7L sharp-edge entries from a plenum into a plate, you might !
| |
| '8 have --
| |
| 9 MR. GITNICK: Yes.
| |
| 10 DR. SCHROCK: You will-have all of those in your !
| |
| 11 icons or how -- ,
| |
| 12 MR. GITNICK: Right now, on the up lefthand corner 13 of the screen there's some of the code you are using.
| |
| 14 If you are saying RELAP5,'it follows the RELAP5
| |
| () 15 16 conventions for modelling these things.
| |
| follows the TRAC conventions.
| |
| If you say TRAC, it 17 When you go generic, we are. hoping to go with a 18- physical base description. In that case the person would 1 19 describe I have an orifice, I have an elbow, I have a T, but 20 with RELAP what you say is I have a junction where there is 21 a sharp reduction in area and has a K-factor of. That is 22 the-way RELAP describes it.
| |
| I 23 TRAC Nas a slightly different way of saying the l 24' same thing.
| |
| l 25 DR. ZUBER: When are you going to have the last ,
| |
| 1 l
| |
| [ /'' ANN RILEY & ASSOCIATES, LTD. j l
| |
| (_j) . Court Reporters l 1025 Connecticut Avenue, NW, Sui'_e 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| - . - . . _ . . - _ . _ - - ~ . - _ . ~ . . = - . - . - . . - . . . - . - - - . - - - _ -
| |
| 258 1 one to say I want a pipe --
| |
| ~
| |
| 2 MR. GITNICK: Well, we are hoping to have TRAC at 3 the end of next year. What's that?
| |
| 4 DR. ZUBER: RELAP when?
| |
| 5- MR. GITNICK: Oh, RELAP?~ We think'that it is just "
| |
| 6' .about ready for RELAP now, just RELAP, not the port to TRAC.
| |
| 7 There are things that have to be added still but 8 it is basically ready for RELAP now.
| |
| 9 DR. ZUBER: And TRAC by the end of next year? 1 10 MR. GITNICK: Yes. The hardest part of course is 11 the translation between the two. There is a lot to it.
| |
| 12 DR. ZUBER: That would be TRAC P or TRAC M? !
| |
| 13 MR. GITNICK: TRAC M. l l
| |
| 14 DR. ZUBER: TRAC M, I see. 1
| |
| ()
| |
| 1 15 MR. GITNICK: At least that is what I believe --
| |
| 16 MR. KELLY: That's correct.
| |
| I 17 MR. GITNICK: So after that of course it's in a 18 lot of people's hands as we go on to other things, but I 19 would like to_think that this tool was very valuable and 20 will see a life of its own.
| |
| 121 DR. WALLIS: How many users are there for this?
| |
| '22 MR. GITNICK: Right now we have 15 or 16 beta 23 testers outside the NRC. In NRC there's people in NRR and j 24 there-are people in RES.
| |
| 25 MR. KELLY: ACRS consultants. j ANN RILEY & ASSOCIATES, LTD.
| |
| .4~ Q(,,/ Court Reporters 4
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| L 259 l
| |
| )
| |
| 1 DR. KRESS: Feedback from those kind of users?
| |
| \ 2 MR. GITNICK: There is a reading file that comes l '(% )
| |
| ! 3 with'the distribution saying that they send their comments L
| |
| ! to SNAP at SCIENTECH.COM.
| |
| [.
| |
| '4 ~
| |
| We have an E-mail address. NRC is welcome to monitor it.
| |
| ~
| |
| 51 We do not have a website however 6' for this.- It's just strictly done by E-mail.
| |
| 7 DR. SCHROCK: Are you getting a lot of comments?
| |
| ( 8 MR.-GITNICK: Sometimes we get quite a few because i
| |
| 9 we will release a version -- we haven't been releasing ;
| |
| i 10 source code because it's a complicated program and not 11 everybody has all these libraries and tools, so we release 12 it made for Sun, IBM, HP, DEC, SGI, NT 95/98 -- so one of 1.
| |
| l 13 these versions perhaps there's something that went wrong and 14 it crashes. We try to test them all but it is a big job, so O
| |
| (j 15. when something crashes we hear about it quite quickly.
| |
| 16 DR. ZUBER: Let me ask you what about TRAC-P, the 17 one you have now?.
| |
| j 18 MR. GITNICK: I would like to do TRAC-P but I l 19 think that-ils someone else's decision.
| |
| 20 Obviously there's limitations on budget. There's 21 also limitations on programmers. As you know, you have to 22 have programmers who are very knowledgeable. This is all l 23 done with programming using some of the more modern l;
| |
| L 24 techniques.
| |
| 25 -
| |
| There's a couple more pages of slides showing the i
| |
| r
| |
| ['' -
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| . 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l
| |
| ! (202) 842-0034
| |
| }.
| |
| l
| |
| | |
| 260 1 architecture of the system in your handouts. Anyway,
| |
| [' 2 finding programmers who are very good with object-oriented 31 programming, C++, Java, who know RELAP and who know TRAC, 4 that is a very~small population, so it's not like you can go ,
| |
| 5 out on the corner and find four more programmers to put on
| |
| '6 the job.
| |
| 7 DR. ZUBER: So there is no plan to implement 8' TRAC-P?
| |
| 9 MR. GITNICK: Not at the moment as far as I know.
| |
| t 10 DR. WALLIS: Who will this belong to?
| |
| 11 MR. GITNICK: It is the NRC's.
| |
| 12 DR. WALLIS: Does it belong to the public so it is 13 going to be out there and anybody can use it.
| |
| 14 MR. GITNICK: Anybody who wants it, we'll send it 15 to them.
| |
| 16 DR. FONTANA: Did your kid brother work on this?
| |
| 17 MR. GITNICK: Teri is my wife. She is a UNIX 18 programmer.
| |
| 19 (Laughter.]
| |
| 20 MR. GITNICK: But yes, we live this thing night 21 and day.
| |
| , 22 DR. WALLIS: Thanks.
| |
| L 23 MR. GITNICK: Thank you.
| |
| 24 DR. ZUBER: That's very nice.
| |
| 25 MR. GITNICK: Thank you.
| |
| I lN(^- ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| : 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
| |
| '(202) 842-0034
| |
| | |
| . - . - -- - . - -. - ..~.- . - -_ -. . - , - . - - - ~ - . ~ . - . - . - . - . . . . -
| |
| 261 1 MR. ELTAWILA: Thank you.
| |
| ?
| |
| ;( ,:. .2 DR. . ZUBER: -Very good.
| |
| 3 .MR. EBERT: I am David Ebert from the same branch' 4 as Farouk and like the presentation I am going'to have to 5 ' sort of do things myself and two machines at the same time, 6 so this ought to be a quite interesting exercise.
| |
| 7- First of all, we are. sort of hooking up a j 8 different computer to'the LAN' system and so I am going to be 9 using this monitor which -- this monitor and then display 10 some files on PowerPoint at the same time, but I can start 11 talking about my presentation. 1 12 This represents the work that has been undertaken 13 'in this branch that Farouk talked a little bit about, as far 14 as coupling thermal hydraulics codes and the neutronics O
| |
| g 15 codes.
| |
| 16' Most of the work has been undertaken by the 17 contractors but the branch and myself in particular have 18 done some'of the work also.
| |
| -19 The thermal hydraulic codes are -- at the 20 beginning is RELAP, but then also recently we have coupled 21- PARCS code, the 3D kinetics. code called PARCS, to TRAC-M, 22 .but I will be talking about mostly in this presentation 23 applying the coupled codes to RELAP.
| |
| 24, PARCS stands for the Purdue Advanced React'' Core
| |
| : 25. Simulator and as the question was asked, it will be
| |
| /~'y ANN RILEY & ASSOCIATES, LTD.
| |
| (_/. Court Reporters
| |
| : , .1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 i
| |
| i' (202) 842-0034 v
| |
| i
| |
| | |
| !t 262 1 available, it is available to the public now. The NRC has 2 lits own licensed version of the PARCS code.
| |
| 3_ My presentation will be -- the outline of my 4 presentation is as follows. I will be talking about the 5 coupling of the 3D kinetics with the thermal hydraulics 6 code.
| |
| ! 7l First of all, the design from the application --
| |
| 8 two applications that we have applied it to recently, the
| |
| ~9 main steam line break benchmark which !s being undertaken by H10 a number of organizations.
| |
| 'll VOICE: Please speak up.
| |
| 12 MR. EBERT: Okay -- a number of organizations 13 undertaking the main steam line break, and then also 14 application to a control rod ejection accident, hypothetical
| |
| () 15 control rod ejection accident, all of these at TMI-1 for 16 high burnup.
| |
| 17 Then the next item I will be talking about is 18 recently the Purdue University, who was the author-of the 19 PARCS code, has introduced an improvement into it called Pin 20 Power Reconstruction. Basically what that is is you are 21 able then to not only get out the nodal power, the 3D nodal 22 power,-but then also superimposed on it is a good estimate 23 of all the fuel pins, the power level of all the fuel pins 24- at all the time during the transient, so thic is actually 25 quite a marked improvement and it will be quite useful.
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 r
| |
| | |
| -. ~.- -~-__ - ~ -- - - - -
| |
| 263 1 It will be useful in several senses. It will be
| |
| (''j - 2- .useful -- you'can immediately think of it being useful in a
| |
| 'Q 3 subchannel analysis code and then it would also be useful e 4' for right.now as far as fuel failure type of information.
| |
| 5 We have applied that --
| |
| :6 DR. SCHROCK: What do you have to do to get a
| |
| '7. . sufficient 3D distribution out,of RELAP?
| |
| 8 MR. EBERT: Well, you can't out of RELAP but also L 9 the kinetics ~ code would be hooked to TRAC, the consolidated 10 TRAC code, so whatever 3D distribution --
| |
| 11 DR. SCHROCK: You have said that you are doing it l
| |
| 12 with RELAPS.
| |
| 13 MR. EBERT: Well, at the moment. The first 14 -coupling is with RELAP. That's further developed but we
| |
| () 15 also have a preliminary version of PARCS in TRAC-M.
| |
| 16 DR. WALLIS: PARCS is more three-dimensional.
| |
| 17 MR. EBERT: PARCS.is a three-dimensional nodal 18 code.
| |
| 19 DR. WALLIS: So you have to do some averaging or 20 something to'use it --
| |
| 21 MR. EBERT: Yes, there is a mapping that takes 22 -place between the neutronics and the thermal hydraulics.
| |
| 23 Generally you don't have as many channels in the 24 thermal hydraulics as you do in the neutronics so there is a 25 mapping --
| |
| ,O ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 L
| |
| Washington, D.C. 20036 (202) 842-0034
| |
| | |
| a A - a~ - - , --, ---a L--1-.~ 4 4 - - - -
| |
| ! l l
| |
| l 264 1 DR. SCHROCK: I believe there's a representation
| |
| ~
| |
| ; d( T 2 in the core -- it doesn't give you very much detail on the l 3 spatial distribution of the void so you need to do something 4 other than the typical running 3 by 5.
| |
| 5 MR. EBERT: Yes. Well in this particular case we 6 have actually for the main steam line break we have 18 pipes 7 hooked together to a plenum, so the nodalization in RELAP is 8 18 pipes, and then you stick in the kinetics code in those 9 18 pipes -- 200 and so assemblies you stick into 18 pipes.
| |
| 10 But as far as the neutrons go, they don't know 11 about pipes. What they know about is just water and fuel, so 12 they can go across pipe boundaries easily.
| |
| 13 The neutronics code is much more finely nodalized 14 than the hydraulics, I'll have to agree with that.
| |
| (~ 15 g
| |
| The next two slides here -- it's mainly to 16 demonstrate these are abstracts of papers that will be 17 presented at a mathematics and computation meeting, 18 conference in September of next year.
| |
| 19 DR. FONTANA: They have identical titles.
| |
| 20 MR. EBERT: No, they are slightly different. One 21 if development and the other is assessment so that is the 22 difference.
| |
| 23 I wanted to put these up mainly to show these are 24 the developers in this part here, and then there is 25 additional people, mainly Brookhaven National Lab, are some l
| |
| l l
| |
| ''T
| |
| ((,s/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 265 r 1 of the users so the developers -- there are developers and 1
| |
| ) 2 . users but there's also some users at the moment, and this 3l code is being made, has been made actually available to NRR
| |
| : 4. so they are a possible user too, eventual user.
| |
| l 5 Next, going on to the design of the coupled codec,
| |
| ~6 the general design is to have the thermal hydraulics code 7
| |
| l 7 sitting here and the neutronics code sitting there, and then 8 communicating with each other through a general interface, l
| |
| 9 so for instance the thermal hydraulics code would calculate 10 like temperatures and pressures and so forth and relay those 11 to the general interface to the neutronics code. The 12 neutronics code then in turn takes those, uses those to i 13 generate the cross-sections, neutron cross-sections, and l 14 then from that the neutronics code then calculates powers,
| |
| () 15~ power densities, and feeds that back into the thermal L 16 hydraulics code, in this case RELAP.
| |
| L 17 Then it just keeps going on time step after time 18 step.
| |
| 19 Okay, now I get into the presentation here, the 20 PowerPoint. The rest of them are in PowerPoint format, and 21 we are logging into the NRC internal network --
| |
| 22- DR. KRESS: You have to have a valid password.
| |
| o
| |
| '23 Actually you've been logging in and I can go on i
| |
| ! 24 - .here. Actually I can show some of these. Modern technology 25 here.
| |
| I J -
| |
| l
| |
| ;,N,; -
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014
| |
| [ Washington, D.C. 20036 i (202) 842-0034 i I I
| |
| | |
| 266 ,
| |
| 1 The first application to this couple code was the 2 main steam line break, and the main steam line break
| |
| )
| |
| 3 scenarios being under the standard problem is being "
| |
| 4 undertaken at TMI,'high burnup core at TMI. So the main l
| |
| 5 steam line break transient starts with a break in the steam line, and'-- I was afraid this was going to happen'--
| |
| ~
| |
| 6 7 fortunately I have a backup.
| |
| l 8 The main steam line break takes place, and then 9 there's enhanced heat removal, so therefore cold water comes 1CF in the inlet of the core. From the cold water you get a i 11 . positive reactive feedback. The core power level increases-12 until the reactor trip point is set at 118-percent power in 13 this case. The reactor scrams, but the most reactive rod is ;
| |
| 14 stuck out. The core power drops as a prompt drop. But D) t 15 you're still getting continuously more cooler water coming 16 in the core.
| |
| 17 And therefore the subcritical multiplication, the L.
| |
| 18 reactivity, even though it's subcritical, increases, and you 19 get a phenomenon called subcritical multiplication, the core 20 . power increases, and the vital question actually is does it 21 return to power, does it return to criticality, or both. So 22 the whole reason that this main steam line break benchmark 23 was started is to have a nice applicable problem, a 24- realistic problem to be able to test the coupling of the 25 kinetics code, 3D kinetics code with the thermohydraulics ANN RILEY & ASSOCIATES, LTD.
| |
| la- s-[~')' Court Reporters
| |
| ; 1025 Connecticut Avenue, NW, Suite 1014 o Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l 267
| |
| : 1. code.
| |
| ['T 1
| |
| 2 All right, here's, as I mentioned, the l
| |
| .Q) 3 characteristics of the main steam line break. There's about 4 20 institutions around the country that are -- around the 1 5 world that are participating in this. Here's the vital l
| |
| j 6 parameters, as a transient goes on. Temperature in the I
| |
| 7 broken loop here and in the unbroken loop, and then you get 8 corresponding pressure decrease.
| |
| 9 The power goes first of all down because you're 10 getting more voids, and then up as your temperature 11 decreases, you get a scram here. Then you're subcritical, 12 but you're increasing in subcritical multiplication. You go 13 up here and you're almost critical at this point,.at about 14 30-percent power. Remember, all the rods are in except the n
| |
| highest stuck rod, so you almost have gone critical again,
| |
| () 15 j
| |
| 16 but not quite. And then you decrease as negative feedbacks 17 come in more.
| |
| 18 The whole worth of this presentation is in the i 19 graphics, but -- l l
| |
| 20 MR. ELTAWILA: Why don't you --
| |
| 21 MR. EBERT: Okay. Later. Okay.
| |
| 22 Again, we have the graphics, but if you can 23 visualize this happening in time, I have very nice color 24 graphics of this, and the graphics makes the whole day. But 25 if you can visualize, here is core power at the beginning, l
| |
| ! /''N ANN RILEY & ASSOCIATES, LTD.
| |
| (-) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| : m. ..__ - -. - . - - - -- _- .... _ . . . - _ . - . - . - . . . . . - - . . . . ~ . . - . . --
| |
| 268 1 full power, this main steam line break transient, full
| |
| ;2
| |
| -d{''N power, and it's at -- there's a line here that's '
| |
| 3 corresponding to 100-percent power. Then as time goes on it 4 gets very skewed, radial skewing to the side where the 5 colder water is, the broken-loop side. And then also where 6 the stuck rod is.
| |
| L 7 And if you saw the animation, you would see this 8 is the location of the stuck rod, the side that's cooling,
| |
| '9 it actually goes up -- at about 60 seconds it goes up to --
| |
| 10 even though the reactor's subcritical, you've got all the 11 rods in, some channels are up to full power already. So 12 here you've got a full-power situation, your flow is closing 13 down, so this is a good candidate for DNB.
| |
| l 14 - Now if you' had the subchannel analysis code, and
| |
| ~
| |
| ) 15 also had the Pin Power representation, which I'll show you
| |
| .16 in a minute, you could do probably a very good best-estimate 17 analysis of this transient situation.
| |
| 18 The other transient that we applied the analysis 19 -to was a rod ejection accident at TMI, PWR, where the rod --
| |
| E 20 ejected rod -- is about a $1.18 worth. So you're getting 21- superprompt critical, you get your classical pumps coming up l 22: here about 80 second full width at half max, and then 23 there's an indication of the temperature rise and the 247 average temperature in the fuel.
| |
| 25 Now I guess I don't want to really talk too much I
| |
| L[]. ANN RILEY & ASSOCIATES, LTD.
| |
| , "(_/ '
| |
| Court Reporters
| |
| , 1025 Connecticut Avenue, NW, Suite 1014
| |
| [ Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l l 269 i'
| |
| 1 L
| |
| I.
| |
| about-how the Pin Power methods.are come about, but recently
| |
| /''\ 2 Pin Power reconstruction method has been put into the PARCS U
| |
| 3 code, and tested against three different situations. One t
| |
| 4 situation is'several years ago there was an international 5
| |
| _ standard benchmark for codes, mostly fine mesh codes, to try 6 to predict the power distribution in very simple lattices, 7
| |
| : and the lattice of concern for our case is the core called 8 C5, which is a mixture of MOX fuel and also your regular 9 uranium.
| |
| 10- Uranium is in the black sort of ones, and the M is
| |
| ( 11 MOX fuel. And this is surrounded by water. This is a core 12 representation.
| |
| 13 So the Pin Power reconstruction method was run on 14 this. It was a steady-state critical assembly. But the
| |
| 'q t
| |
| j 15 point is, and here is the results of the calculation, you 16 can see that there's the individual pin. Each one of these 17 little humps here is the individual pin. So not only do you 18 get gradients in the products of the assembly, one of these 19 big blocks is an assembly, but you also get little pins 20 sticking up here. So you can calculate the power level in 21 every pin.
| |
| 22 One of the important things is that there's a 23 great big drop between the MOX fuel and the UO2, and 24 diffusion theory quotes are notoriously -- it's hard to
| |
| ; 25 handle that for diffusion theory. And so theoretically or l
| |
| l
| |
| ! ("T ANN RILEY & ASSOCIATES, LTD.
| |
| l (,) Court Reporters i
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 270 1 you would expect that PARCS would not be good at predicting ;
| |
| '[ah 2 that. But fortunately, as it turns out for this particular 3 _ problem, PARCS does quite well. Here's sort of the measure !
| |
| [ '4 of the error between the PARCS code and experiment, and then 5 between PANTHER, which is a fine-mesh multi -- diffusion i 6 theory. code in the experiment. ;
| |
| 7- And you see here in the-last'line here that the 1
| |
| 8 . error is really quite small. I
| |
| ( 9 'Here is the main' steam line break again. This is i
| |
| ~
| |
| 10- the Pin Power distribution now at full power, and you can l 11 see quite a fine variation. And then at the maximum power t
| |
| f
| |
| :12 level after you scram the reactor, this is the Pin Power !
| |
| 13 distribution there.
| |
| l 14 DR. SCHROCK: Have you got all of the nitty-gritty
| |
| () 15 .little effects, gamma spreading of the radiant energy and --
| |
| 16 MR. EBERT: Yes, there's a gamma --
| |
| , .17 DR. SCHROCK: That includes all of the decay power 18 variations?
| |
| 19 MR. EBERT: Direct gamma deposition in the water.
| |
| 201 ' DR .. SCHROCK: What do you do -- well, more than 21 that, it's spread also in the fuel.
| |
| 22 MR. EBERT: Yes. There's -- the fission products 23 are there, and the gammas and the neutrons. They're all 24 ~ accounted.for. And again, in a nodal sense, but then the l '25 ~ Pin Power reconstruction method is sort of imposed on top L.
| |
| i l
| |
| 'a ANN RILEY & ASSOCIATES, LTD.
| |
| \ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l
| |
| Washington, D.C. 20036 (202). 842-0034 l
| |
| | |
| 271 1- of --
| |
| c .
| |
| j j 2 DR.' SCHROCK: What do you do'about the fuel L composition in these calculations?
| |
| 3'~ You have complex things 4 that arise and depend on fuel-cycle practices.
| |
| u 'S MR. EBERT: Yes, this is actually end of cycle at 6 TMI-1.
| |
| ~
| |
| 7 DR. SCHROCK: This is TMI-1.
| |
| 8 MR. EBERT: Yes, this is about 50 gigawatt-days 9- per ton on the average. So you actually get quite a 10 difference. Here's low-burnup fuel, here's high-burnup 11 fuel,-and so forth.
| |
| 12 DR. SCHROCK: What sort of detail is there in 13 that? These are pretty large blocks you're looking at 14 there.
| |
| 15 MR. EBERT: Well, they're assemblies. Each one --
| |
| 16 DR. SCHROCK: It is little assemblies.
| |
| 17 MR. EBERT: Is one individual assembly. Right.
| |
| 18 DR. SCHROCK: Pardon me?
| |
| 19 MR. EBERT: It's --
| |
| 20 DR. SCHROCK: Look like enough.
| |
| 21 MR. EBERT: I forgot where I put the marker.
| |
| 22 So this is not the full core. Maybe you're
| |
| '23 This is not the full-core representation. This
| |
| : l. confused.
| |
| 24 is just-around where the stuck rod is. So maybe that was 25 the confusion. But each one of these is an assembly, and ANN RILEY & ASSOCIATES, LTD.
| |
| l [\ -) Court Reporters 1
| |
| ) 1025 Connecticut Avenue, NW, Suite 1014 2 84 I b34 i
| |
| 1
| |
| | |
| 272 l l
| |
| 1 there's -- it's a 17-by-17 assembly. So there's 17 times
| |
| / 2 17.
| |
| \~
| |
| 3 DR. SCHROCK: You show some pretty good gradients 4 across some of those bundles, so presumably the detail of !
| |
| 5 the composition, is that the level of 110 or --
| |
| l 6 MR. EBERT: No, the composition is on the assembly I 7 basis. But the Pin Power reconstruction sort of takes in 8 those kind of global effects or gradients across assemblies, 9 amongst other things. It takes into account gradients 10 across assembly during the transient itself, and then 11 superimposed on that are individual pin, what's called I l
| |
| 12 heterogeneous function. So it takes into account the l 13 full -- pretty much it's fully representative of reality as 14 you get at the moment.
| |
| (A) 15 The gradients are even stronger in the rod 16 ejection case. This happens to be -- had about half a 17 second. The rod has been ejected and you are the maximum 18 power level, and the peak -- the peak rods here are about 19 five times full power. And so there is quite a bit of 20 energy deposited. And, in fact, we have calculated, briefly 21 calculated the energy deposition in the maximum pin and it 22 is about a hundred calories per gram. And this is just on 23 the borderline of fuel failure.
| |
| 24 DR. KRESS: What does the shaded bar represent?
| |
| 25 MR. EBERT: The shaded bars -- oh, colors.
| |
| ' ['
| |
| , i ANN RILEY & ASSOCIATES, LTD.
| |
| \.~sl Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| l t
| |
| | |
| . _ . ... _ _ _ . . _ 4 . . _ - . . . . ._._ ___.._.____ _. _ _ ____ __ _ _._ .-
| |
| l 273 l
| |
| 1 DR. KRESS: Yeah, they would be colors, j \ 2' 'MR. EBERT: If you saw that in color --
| |
| G 3 DR. KRESS: Yeah, if I saw it color, I would know.
| |
| 4- Do they. represent temperatures or what? .
| |
| 5 MR. EBERT: I'm sorry that things didn't work out :
| |
| L 6 too well but --
| |
| 7 DR. WALLIS: Factors at full power.
| |
| 8 DR. KRESS: .I thought that was what was on the 9 graph.
| |
| 10 DR. WALLIS: Five times, doesn't that say 11 anything? i 12 DR. KRESS: I mean that is the same thing that.is
| |
| {
| |
| j 13 on the graph?
| |
| l 14 DR. WALLIS: I think so, it is just in color. It
| |
| ; O
| |
| ( j 15- is a color code.
| |
| 16 MR. KELLY: It is a-scale.
| |
| L 17 DR. KRESS: It is a color code scale.
| |
| l 18 MR. EBERT: I did test this out about three times L 19 over the course of the week and it is not like I didn't do 20 my' homework.
| |
| l 21 DR. KRESS: It just helps us read the graph 22 easier, the color code.
| |
| 23 DR. WALLIS: This is a thermal-hydraulics and 24 physics subcommittee, isn't it?
| |
| 25 MR. EBERT: It is all there.
| |
| (~'
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| ' Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| L 274 j I
| |
| i 1 DR. WALLIS: So neutronics is fine. .
| |
| 1 2 SPEAKER: A question for Dave. We thought maybe
| |
| (}
| |
| 3 he would keep trying to lock in one of the --
| |
| l 4 RMR. KELLY: Well, Joe, is using the same system.
| |
| 1 5 DR. KRESS: It is same problem we always have.
| |
| i 6 DR. WALLIS: -You have-the same problem we have.
| |
| I 7 DR. KRESS: You have the same problem we-have.
| |
| 8 MR. . KELLY: If there are no questions for Dave,
| |
| -9 since we are running a little late, I will go ahead and --
| |
| 10 DR. WALLIS: Well, did you meet your objectives?
| |
| 11 MR. EBERT: Pardon?
| |
| 12 DR WALLIS: Did you meet your objectives?
| |
| 13 MR. EBERT: Yeah, we met our objectives on ?.ime.
| |
| 14- DR. ZUBER: When did you start on this?
| |
| () 15 MR. EBERT: We started about a year ago.
| |
| 16 -Originally,.the plan was to couple PARCS with TRAC-M, but 17 then the Kent members thought that they wanted -- Kent 18 members wanted PARCS hooked to RELAP first, so we decided to j '19- do that.
| |
| 20 Actually, once the interface has been built, it is l 21 actually quite easy then to apply it to other codes.
| |
| 22- DR. WALLIS: Who is the user?
| |
| 23 MR. EBERT: Who are the users?
| |
| l 24 DR. WALLIS: Yes.
| |
| 25 MR. EBERT: At the moment, I am a user, the 1
| |
| , ANN RILEY & ASSOCIATES, LTD.
| |
| ; Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
| |
| I' o
| |
| | |
| l 275 developers are a user, and David Diamond at Brookhaven 1
| |
| [')
| |
| Lj 2 National Lab is using this to calculate the energy 3 _ depositions in high burnup fuel. And we have got a 4 collaborative program with the Russians and also the French.
| |
| 5 They are doing similar pin power reconstruction type of 6 calculations. And then at the end of this, in about a year 7 we will three-way comparison of energy deposition.
| |
| 8 DR. WALLIS: The user, eventually, is someone in 9 charge of licensing.
| |
| 10 MR. EBERT: Yes. Right. Eventually, it will be 11 people at NRR. This will be -- can be used, will be used, l 12 hopefully, as --
| |
| 13 DR. WALLIS: So you reaching some objective that 14 they have set?
| |
| /N
| |
| () 15 MR. EBERT: Yes.
| |
| 16 DR. WALLIS: Trying to make the coupling.
| |
| 17 MR. ELTAWILA: If you recall, at the beginning of 18 my presentation, we had a list of the user conveniences that 19 the user requested, and part of this user community were 20 NRR, and the first one of the list was the 3-D neutronics 21 capability, so that NRR can include other users of the code.
| |
| 22 DR. WALLIS: Is there some statement from them 23 that is written down about what they want?
| |
| 24 MR. ELTAWILA: The statement I think that we have 25 from NRR, we had a user need letter to improve the code,
| |
| ,/'''; ANN RILEY & ASSOCIATES, LTD.
| |
| (_,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 276 1 make them more user friendly and improve the modeling
| |
| ( 2 capability of the codes. But explicitly, we don't have a 3 user need request for the 3-D neutronics.
| |
| 4- But this, what we see nationally and 5 internationally, and the trend to. increase power, for 6 example, the main steam line break that Dave is talking 7 about, it is critical for. power increase because the current 8 can go critical if you use point kinetic, but if you use 3-D 9 kinetic, it will not go critical, so you can eliminate that 10 from the scenario that you are concerned about.
| |
| 111 So although the Office of Research does not do 12- everything because it is user requests, we have a 13 responsibility to anticipate problem and try to see where 14 the industry is heading and to try to develop the tools to 15 be able to meet these needs when it comes.
| |
| 16 DR. WALLIS: Good for you.
| |
| 17 DR. ZUBER: Question. Do you have a list of 18 requests users need from-NRR?
| |
| 19 MR. ELTAWILA: We have a list of the user 20 requests, but most of them are related to AP600. Right now 21' the.only user -- don't forget that the office associate of 22 NRR said we do not need thermal-hydraulic research. So, by 23 saying~that, he will not see us a user request. You just 24 -can imagine --
| |
| 25 DR. ZUBER: How does that branch of NRC perform --
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L. . . . . . .. . . .
| |
| | |
| L 277 l' make' decisions without a quota? How do'they proceed?
| |
| ) 2 MR. ELTAWILA: I think that you should ask that 3 ' question of NRR.
| |
| l 4 DR. ZUBER: Okay.
| |
| 5- DR. WALLIS: . Well, I.think the concern we have is
| |
| '6- that you are doing some very useful work, but if someone who 7 controls some budget doesn't appreciate it or know about it, 8 then --
| |
| 9 MR. ELTAWILA: Yeah.- I want-to make it clear that 10 the Office of Research management is extremely supportive of l
| |
| 1 11 this program, but when it goes to the users and the user
| |
| .12- look at their priority, they say we don't need 13 thermal-hydraulic, we don't need severe accident and.so on, 14 and they identify, they list their own priority. But, h 15 again, as I mentioned earlier, we have our own charter, too, 16 and we want to be prepared when the need exists'for these 17 tools.
| |
| 18 DR. KRESS: That is where some words from us might
| |
| '19 help,'right?
| |
| :20 MR. ELTAWILA: Yeah, definitely.
| |
| L21 DR. KRESS: If they_are the right words.
| |
| 22 DR. SCHROCK: I think~you have gotten a lot in a 23 short time.
| |
| 24 MR. KELLY: Hopefully, what I will say won't
| |
| .25 ' change that opinion. I have two presentations, and I am l
| |
| I' l-
| |
| !. -ANN RILEY & ASSOCIATES, LTD.
| |
| \ Court Reporters
| |
| : 1025 Connecticut Avenue, NW, Suite 1014 i ! Washington, D.C. 20036 I ~(202) 842-0034 r , , - - - . - . - . , .
| |
| | |
| i j 278 )
| |
| 1 going to try to skip some of the details because we are
| |
| ()
| |
| y~
| |
| 2 already getting kind of late. Oh, and for the record, this 3 is_ Joe Kelly from Research and the Reactor Plant and System 4 Branch.
| |
| 5 The first presentation is entitled Code
| |
| )
| |
| 1 6 Consolidation Example: The Boiling Water Rreactor Jet Pump j 1
| |
| 7 Model. And I have a co-author listed; that's Biro Atkas of I 1
| |
| 8 Scientech. He is hiding in the back. Biro is the engineer 9 who actually incorporated the iet pump model into the 10 consolidated version of the code. So by putting his name 11 here as a co-author, I'm able to call on him in case I get a i l
| |
| 12 question I can't answer.
| |
| 13 Now, both of my presentations to some extent l 14 delineate the process that we're trying to use as we go to l
| |
| (('")\
| |
| 15 consolidate and improve the code, but also both of them do 16 show some results.
| |
| 17 I'm going to start by giving a little bit of 18 background about code development practices -- what have we 19 done in the past and what are we doing now? Then an 20 overview of how you go about consolidating a component, a 21 brief introduction into what is the TRAC-P jet pump model.
| |
| 22 I'm going to talk about functional requirements, and this is 23 coftware development language, what is it that the component 24 has to do? You have to explicitly state it. Then you put 25 together a verification testing plan to make sure that you t' 1 ANN RILEY & ASSOCIATES, LTD.
| |
| N/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 279 1 actually meet those requirements-when you deliver the L
| |
| l .y l''T 2 software and then the results of that.
| |
| L3 What did we do in the past? And'this.is 4 historical, so I'm sure most of you remember this. We had 5 four different codes. For each code, one contractor was ;
| |
| 6' primarily responsible for doing everything associaced with 7 'that code': all of the development, the maintenance and the j 8 ' assessment activities. RELAP-5, INEL, TRAC-P, et cetera.
| |
| 9 The NRC project manager was not expected to either
| |
| : 10. know the details of the code -- I'm talking about knowing i i
| |
| 11 either the numerics'or the physical models -- nor were.they '
| |
| 12 expected to be able to use the code. They just managed the 13 . details, getting the contracts out, getting the money, et 14 cetera. All responsibility for code configuration control
| |
| '(O/ 15 .was also left to the contractor.
| |
| l 16 Sometimes this resulted in no rigorous testing 17 between versions. A whole of changes would go into the code 18 in the developmental version. It would be run against the i
| |
| 19 installation problems. !
| |
| 20 First, the code had to compile -- if it could 21 successfully complete the installation problems, that was v 22 fine, That was all the testing that was done. No one went 23 and looked at the answers to see how much the answers 24 changed and asked the question if it made any sense.
| |
| 25 Also, software quality assurance was left to the
| |
| - (~'\ ANN RILEY & ASSOCIATES, LTD.
| |
| ! - --( ) Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| __. -. _ .___m . __ _m . _. _ _ _ .
| |
| I l'
| |
| 280 l 1
| |
| 1 ' contractor with little or no NRC oversight. So there are a
| |
| ' i
| |
| {} 2 lot cf undocumented code modifications. Things could change l-
| |
| '3 l'l:the u code on a daily basis with no paper-trail.
| |
| 4 Code assessment libraries. Those of you who were 5- associated with the 2-D, 3-D program and the CSAU study, you 6' .know a. ton of assessment was done. Well, those input decks
| |
| {
| |
| 7 and the. data were:not kept in a centralized storage 8 location.
| |
| 9- MR. ZUBER: Where are they?. l
| |
| '10 MR. KELLY: Well, most of the decks are still at 11 LANL, but they're for an older code version. And so we'll I'
| |
| 12 have to update those input decks-for the current code 13 version, and we'll have to, you know, make sure we can find 14 all the data.
| |
| 15 MR. ZUBER: What happened to PFl? Do we still 16 have access -- do you still have access to it?
| |
| 17 MR. KELLY: No, because it was -- TRAC PF1 ran on 18 a certain --
| |
| 19 MR. ZUBER: This is the only code for which we 20 determined the uncertainty.
| |
| .21 MR. KELLY: That's true. And it ran on the CRAY 22- operating system under the particular system that they had 23 at LANL and not on any other platforms Those computers i
| |
| 24 don't even exist anymore.
| |
| 25 So with that, we would have to do, in effect, a
| |
| /~' ANN RILEY & ASSOCIATES, LTD.
| |
| 7 Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 t
| |
| Washington, D.C. 20036 (202) 842-0034 l
| |
| r.
| |
| | |
| L 281 I i- l L
| |
| 1 lot of work to recover -- to get TRAC PF1 to work again.
| |
| 1 2- And I agree, that's a problem. That's part of this past
| |
| )
| |
| f 3 ' code --
| |
| 4 MR. ZUBER: This is what I really thought six 5 months ago. We have less capability today to.get I 6 uncertainties than we hadLten years ago.
| |
| 7 MR. KELLY: Unfortunately, I~ agree with you, and 8~ it's part of my job to make sure that we recover that 9 capability and-then get better.
| |
| 10 MR. ZUBER: How much will it cost to recover that i
| |
| 11 from -- the reason I'm asking, suppose -- I mean, I hope l 12 not, but suppose'that something happens, that your funds may <
| |
| 13 .be cut. We don't-have any code now for which we-know the 14 ~ urcertainty except that one we evaluated 30 years ago, ten
| |
| ,[) %,
| |
| 15 years ago. How much would it cost to recover that i 16 capability?
| |
| 17 MR. KELLY: I don't know that it would be in -- it 11 8 would be in staff years, I'm sure.
| |
| 19 DR. KRESS: It would be more than it would have to 20 have gone ahead with this.
| |
| 21 MR. KELLY: Well, not when you consider doing the 22 uncertainty assessment. But, you know, it would be on the 23 . order of staff years. It's not a five-minute fix. And you 24 couldn't be -- and because of the way the architecture of 25 TRAC was.done, it's very complicated, the way the container
| |
| ; s.(''\ ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters I
| |
| . 1025 Connecticut Avenue, NW, Suite 1014 I-Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 282 1 arrays and so on are done in order to bypass dynamic memory
| |
| (''')
| |
| U 2 management, you would -- to make sure you were getting the 3 same answer that you got before after you updated it to a 4 different operating system, it's -- you kncw, it would be a 5 lot of trouble and I'm not sure it's worth it unless we were 6 in a situation where we had no funding to go forward with.
| |
| 7 DR. FONTANA: If you wanted to run TRAC now, you 8 couldn't?
| |
| 9 MR. KELLY: You can run TRAC Pb2, but PF1 was the 10 version that the uncertainty study was done on.
| |
| 11 DR. FONTANA: Oh, I see.
| |
| 12 MR. KELLY: A lot of improvements were made for 13 PF2 where " improvements" is in quotes.
| |
| 14 MR. ZUBER: That's the point. I think the O 15 i j performance we saw six months ago on PF Mod 2 was very poor, )
| |
| 16 actually was not 1.
| |
| ]
| |
| l 17 MR. KELLY: That's correct. A lot of well 18 intentioned model development was done, but the type of 19 models that were put in were really ansuitable for working 20 with the numerics. Consequently, there's lots or numerical 21 oscillations that completely degrade the solution. l 22 Well, I ahowed you what the code development 23 practices in the past were.
| |
| 24 MR. ZUBER: Just last question that topic.
| |
| 25 MR. KELLY: Okay. l i
| |
| i I (~h ANN RILEY & ASSGCIATES, LTD.
| |
| \_) Court Reporters
| |
| ! 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| i 283 1 MR. ZUBER: When was the development of Mod 2
| |
| ['')
| |
| gj 2 stopped or when was it completed?
| |
| 3 MR. KELLY: I wasn't at the NRC then, but the 4 manuals for TRAC PF2 were released in '93.
| |
| 5 MR. ELTAWILA: And that's when the funds stopped 6 for TRAC. They were really running on less than $100,000 l 7 per year or something like that.
| |
| 8 MR. KELLY: Which is why PF2 was never assessed to 9 the degree that PF1 was. So I made some disparaging 10 comments about how we did business in the past. Well, what 11 are we going to do now to make it better?
| |
| 12 The first thing -- you can debate this, but the 13 code development activities are now dispersed between five 14 different organizations and seven different locations. The (G
| |
| w) 19 five organizations are LANL, Penn State, Scient'.i e o --
| |
| 16 Scientech both in Idaho Falls and here -- and Purdue 17 University, and the NRC staff. Two NRC staff members are 18 involved in the TRAC development effort.
| |
| 19 MR. ZUBER: Who is that?
| |
| 20 MR. KELLY: Jennifer Ewell and myself.
| |
| 21 DR. WALLIS: From the first slide, you would think 22 that would make things worse.
| |
| 23 MR. KELLY: It has pluses and minuses. The one 24 thing it does is it does allow you to take advantage of 25 qualified individuals. For example, Professor Mahaffey at
| |
| ('^)
| |
| \_/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| I 284 1
| |
| Penn State -- he did all the numerics in TRAC. He's the
| |
| /'') 2 world expert on that. But he by himself and with his
| |
| , '%,)
| |
| 3 students can't do everything, so this way, we were able to 4 use him for what he's good at, and hopefully we're targeting 5 that in the other locations as well.
| |
| 6 It does put more requirements on making 7 communications, and we're working that. We have fairly 8 regular meetings and we have a TRAC developers Web page.
| |
| 9 MR. ZUBER: How often?
| |
| 10 MR. KELLY: Six to eight weeks.
| |
| 11 MR. ZUBER: Every six to eight weeks?
| |
| 12 MR. KELLY: Uh-huh. That is right, isn't it?
| |
| 13 Okay.
| |
| 14 The other thing, and this was a change at the r~
| |
| (j 15 agency, the NRC project managers are now expected to become 16 knowledgeable code users. So if you're the TRAC project 17 manager, you have to be able to run that code and do a good 18 job of interpreting these results.
| |
| 19 DR. WALLIS: You said this is a revolutionary 20 change at NRC. Is this throughout NRC or just in a very 21 tiny part of NRC?
| |
| 22 MR. KELLY: I'm talking about my part of NRC.
| |
| 23 MR. ZUBER: Just a comment to this. It was a l
| |
| 24 dictum that the only thing the manager should do is write a
| |
| ! 25 good -- what was it? -- 189, 189 statement. Absolutely no l
| |
| /''N ANN RILEY & ASSOCIATES, LTD.
| |
| k ,)
| |
| m Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
| |
| | |
| . . .-. . . . . . _ ~ . . - . - - .- .-_ -. . - - - . - .--.. . _ . -. - ..
| |
| , 1 285
| |
| -technicsl. input-or direction from the headquarters.
| |
| 1
| |
| ['N 2 MR. KELLY: Yes. I have heard that.
| |
| 3 -The other thing along the came lines is technical 4' monitors were assigned to each of our programs, and that was 5 a person who was expected to be cognizant of the technical
| |
| : 6. details, to understand the models that.were being developed i
| |
| 7 or changed in the codes.
| |
| 8 This took a hit when Vince and Simon left, because 9 they were two of the staff who did have code development 10- experience.and good programming experience, so that'was 11 unfortunate. Code configuration control is also now 12 performed at the NRC and testing is done between every 13 developmental version.
| |
| 14 So we have a standard problem set that we run with
| |
| () 15 every developmental version, and there are a couple of 16 developmental versions per week. And this is about '6 -- 38 17 problems new and growing. Every time we consolidate a 18 capability, we add one or more test problems to show that 19 capability was put into the code right, so this test suite 20 is growing as we go through time.
| |
| 21 DR. ZUBER: But this is being done here?
| |
| 22 MR. KELLY: Yes. Now, Biro Aktas is the person i L 23 that does this for us, but he does it here on NRC computers.
| |
| 24 And as part of the testing, it is not just --
| |
| 25 DR. SCHROCK: It's doing things a little more l-('') ANN RILEY & AC90CIATES, LTD.
| |
| l'\ss/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036
| |
| ! (202) 842-0034 L
| |
| _ __ - ~ - - - _ - . ,_
| |
| | |
| _ _ . .= _ _._.. _ ._ _... . . _ _ _ . . . . . . _ _ . _ . _ _ _ _ _ _
| |
| 286 1- -complex than the nominal' oscillations.
| |
| /^\ 2 MR. KELLY: Yes.
| |
| .kJ L ~ 3 DR. SCHROCK: Thank you.
| |
| l 4 MR. KELLY: .There are two UPTF, for example. But 5 it is obviously;-- the testing of'the code is an area that I l 6 want to push and-where we have to have a really good 7 comprehensive assessment matrix, and that is something I 8 will talk a little bit more about later.
| |
| 9 Part of this testing is you compare the results h 10 against the previous version, and that is basically done by 11 diffing, if you will, the output files with the tolerance 12- factor and then anything that -- any change that is larger 13 than a'certain tolerance, which is usually to the sixth 14 decimal place, pops up a flag, and then you go look and see
| |
| ; () 15 where the answer changed.
| |
| 1 16 Some code changes, you expect answers to change l 17 and you check to make sure they changed how you' expected-18 them. Other changes to the code are supposed to be j 4
| |
| 19 answer-neutral, and so then you do a null and that is l l
| |
| 20 actually what I am going to talk about in.my second '
| |
| 21- . presentation.
| |
| 22 Software quality control, we are formalizing those j 23 . procedures and all documentation is being maintained online 24 at the NRC under our configuration control system. We are 25 also developing PIRT-nased assessment matrices, one for each t
| |
| ! ' ANN RILEY & ASSOCIATES, LTD,
| |
| : Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| n . . - - - . -. - -.- ~ . - _ _ . - . ~..- - . ~. - . . . - . - -
| |
| i i
| |
| 287 l' ' specific application of the code, so one for boiling water 2- reactors, one for pressurized water reactor large break 3 LOCA, one for pressurized water reactor small break LOCA, 4- and also the AP600.
| |
| '5 DR. ZUBER: Who is doing that?
| |
| 6 MR. KELLY: Basically, it is part of the contracts
| |
| : 7. -- like, for example, for the boiling water reactors at 8 Scientech, because they have the consolidation effort for 9 the boiling water reactor.
| |
| 10 Now, you have heard from Farouk that we suffered
| |
| ; 11 funding cuts. One of the things that the funding cuts are i
| |
| 12 affecting is the development of those matrices and getting 13- together the data and input decks and so on. So that effort 14 is stretching out in time some. We are still doing it, but r( 15- at a lower level of effort than we would like to do.
| |
| 16 MR. ELTAWILA: But Brent is in charge of the large 17 break LOCA assessment matrix.
| |
| 18 DR. ZUBER: Boiler.
| |
| 19 MR. ELTAWILA: Boiler.
| |
| 20 MR. KELLY: Yes. And they are working close 21 together. And Brent is in overall charge of getting the 22 documentation together on it.
| |
| 23- DR. ZUBER: On what, on that part?
| |
| 24 MR. KELLY: ON the assessment matrices. Yes.
| |
| 25 I am not going co spend much time on this slide.
| |
| i --
| |
| l' O ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D C. 20036 1 (202) 842-0034 L
| |
| b - - _ . ., __ __ ._ . - _ -
| |
| | |
| l I
| |
| 288 1 You know pretty much what the objective is, it is to recover
| |
| (')
| |
| V 2
| |
| current capabilities of four codes, and these are largely 3 redundant. Their capabilities greater overlap. They are, l
| |
| 4 in effect, one-dimensional, two fluid type codes. These 5 were the individual missions.
| |
| 6 The approach -- first you have to identify what it 7 is you have to simulate, what type of transients, et cetera.
| |
| 8 I will go into a little more detail in the next slide. We 9 are using the modernizing TRAC-P code as the base, that is 10 what is called TRAC-M, it is in Fortran-90. It gets rid of 11 some of the homespun memory management that was in TRAC-P 12 that was such a nightmare.
| |
| 13 Into this modernized version, you then implement 14 the required modeling capabilities. Those can be component r
| |
| (ms-)15 models like a jet pump, new capabilities like the 3-D 16 kinetics that you just saw, a different numerical solution, 17 and that was mentioned in the semi-implicit scheme.
| |
| 18 DR. ZUBER: I had a problem, I mentioned it last 19 time and I didn't get a reply, and I would like to raise it 20 again. Years ago, 20 years ago, Stan Fabik ran beauty 21 contest between RELAP and TRAC and TRAC was selected as the 22 main code. The problem was that the users, which was NRR, 23 didn't participate or was not very enthused, and we ended up 24 essentially developing two codes, and NRR was constantly 25 using RELAP.
| |
| O A ,/
| |
| m ANN RILEY & ASSOCIATES, L1D.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
| |
| ! (202) 842-0034 l
| |
| | |
| r> 1 i
| |
| p 289 1- When you decided to use TRAC-P, that was, again, a If'')
| |
| \j 2 decision. Do you have any agreement or tacit agreement from 3 NRR they will be using this code or train the personnel when 4 this code becomes available? The problem is you may again i
| |
| i 5 run into the same problem we ran 20 years ago. It is a i 6 political thing. You know, people in NRR use -- I 7 MR. ELTAWILA: No, no. No, I think that is a very l 8 good question, but when we developed the thermal-hydraulic {
| |
| 9 research plan in 1996, it was endorsed by NRR management.
| |
| 10 And in that research plan, we explicitly stated that we are 11 . going to use the TRAC-P code as the model for the 12 consolidation of the code. So there is an endorsement for l
| |
| 13 the process that_we went through. And, again, you can't 14 force anybody to use the code unless you make the code very
| |
| () 15 easy to use and modern and all this stuff.
| |
| 16 So that is what we are trying to do, or will try 17 to do all the things right in that code so it will be very 18 easy for people to switch. We will provide them the 19 training, we will convert the decks, and all that, so it l 20 will be easier for them to switch.
| |
| 21 DR. ZUBER: Okay.
| |
| 22- MR. ELTAWILA: So, in a nutshell, yeah, we have an l 23 endorsement for the TRAC.
| |
| 24 MR. KELLY: This slides shows the consolidation 25 process, but it doesn't show all the steps, because that
| |
| ! l ANN RILEY & ASSOCIATES, LTD.
| |
| [. Q(m/ Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 ,
| |
| Washington, D.C. 20036 '
| |
| (202) 842-0034
| |
| | |
| l 290 1 would be too messy and I couldn't describe them anyway. But
| |
| ('')
| |
| V
| |
| : 2. the approach, what we want is a scrutable, well documented I 3 process. And at the end of this, we need a comprehensive 4 PIRT-based' assessment with both separate effects and 5 integral effects tests, 6 And earlier this morning, Dr. Kress had a question 7 about, well, if in your PIRT you miss a phenomena that 8 actually is important, but you have it well ranking, and 9 then you don't carry it through in your assessment, how do 10 you ever know?
| |
| 11 Well, if all you ever did was separate effects 12 tests, you wouldn't. But you have to have properly scaled 13 integral test facilities and do the simulations on those.
| |
| 14 And if you misranked a model and you have a lousy code model
| |
| / \
| |
| 15 (v) for it, but it is really important, it is going to show up 16 here, and then you have to figure out which model it is and 17 go back and correct it.
| |
| 18 DR. KRESS: Only if that proper scaling captured 19 it, though.
| |
| 20 MR. KELLY: Right. That's correct.
| |
| 21 DR. WALLIS: It will show up if you are aware 22 enough to notice it.
| |
| 23 MR. KELLY: Well, if it is really bad, the results 24 will --
| |
| 25 DR. WALLIS: If it is really bad.
| |
| i
| |
| /"' ANN RILEY & ASSOCIATES, LTD.
| |
| t (,,%/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
| |
| | |
| l 1
| |
| ! . 291 l 4
| |
| ! 1- MR. KELLY: And that has historically happened.
| |
| I
| |
| #' \
| |
| -/ 2 DR. SCHROCK:
| |
| l b) 3 One of the problems I always felt the codes suffered from in the past is the. idea that when you compare against incegral test results that you can
| |
| ~
| |
| l 4 1
| |
| 5- attribute deviations between experiment.and predictions to 6 one specific model and go into code and start diddling with 7 it, and that practice was used over and over and over again. )
| |
| 8 Are you doing this in such a way that you will ensure that !
| |
| 9 that bad history isn't repeated?
| |
| l 10 MR. KELLY: Well, we haven't gotten to that point l
| |
| 11 yet, but I_can address it historically from the time when I 12 was at INEL and was working on the AP600. And one i 13 particular was the critical flow model, and people did try 14 to -- initially, when we had a problem with the ADS-4 flow,
| |
| () 15 16 people tried to make the type rationalizations and tweaking that you are suggesting. But you had to do two things, you have to make sure that you are, first, computing the right 17 18 upstream properties. If you don't have the right pressure
| |
| ~9 and quality or pressure and subcooling, who cares what your ]
| |
| 20 critical flow model is?
| |
| 21 So, we would try to isolate it down. I even used 22 measured conditions and input those to the model and 23 assessed the model that way, so that I knew that it was --
| |
| 24 you know, I wasn't -- but, I agree, I don't like that 25 approach at all.
| |
| l l
| |
| (''T iJ ANN RILEY & ASSOCIATES, LTD.
| |
| '' Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014
| |
| ! Washington, D.C. 20036 (202) 842-0034 I ..
| |
| t
| |
| | |
| 292 1' DR. SCHROCK: Well, what I am really looking for, 2 is there going to be a set of procedures that will enable
| |
| ~
| |
| 3 you to test individual models from integral test data in a 4 more. detailed way than-has been done in the past, not rely-
| |
| .5' just on a few global parameters as measures of success? In 6- ifact, 'some of the stuff written in the past espoused a 7 philosophy on the part of NRC that it didn't matter that the 8 individual models weren't performing terribly accurately on 9 a localized basis because it is-the global result which is 10 the key. ,
| |
| i 11 DR. ZUBER: Is the key. ]
| |
| 1 12 DR. SCHROCK: Right. Yeah.
| |
| 13 DR. ZUBER: 'We had it today on the PCT, everything 14 was justified on the PCT.
| |
| 15 DR. SCHROCK: Yeah. So, I guess as Ivan Catton 16- would have put it, that implies compensating errors.
| |
| 17 MR. KELLY: Yeah, I. feel pretty strongly about I-18 that. And I guess Farouk mentioned earlier a presentation I l 19 made at Santa Barbara, which was on the subcooled boiling 20 model in RELAP5. And if you just looked at, say, the bulk l
| |
| 21 condensation part of it, and I went and found a separate 22 effects test where they measured things like interfacial 23 area and inferred condensation rate from the decrease in 24: interfacial area -- or decrease in vapor flow.
| |
| l L25 : And if you looked at the RELAP5 models, you know, "Q -
| |
| (,,/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| ; 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 293 1 the' interfacial area, say,.in the code'was like this, and it
| |
| -_ ("'s
| |
| .g 2 'was like this in the data, and they crossed at one point, 3 -and that one point was where the model had been tuned to 4 match another set of data. And that -- then you don't have 5 any confidence in being able to extrapolate beyond that 6- .very,.very narrow base.
| |
| 7, So, I agree, but what you are suggesting is very
| |
| -8 -hard to do. To look inside an integral effects test, it 9 takes a very good analyst, and an analyst that is j 10 knowledgeable what is in the code, and those are very few 11 and far between, but it is something I am committed to l 12 doing.
| |
| 13 The.way I want to try to do that is, if you de 14 have a good, large assessment matrix, you know, we can do
| |
| {} '15 .that now. Computers are fast, we have a lot of them. We 16 also have something called the auto-validation system, which 17 makes it very easy to run repeat tests, so it is not, you l 1
| |
| 18 know, staff-time-intensive.
| |
| 19 What I would do for all the high priority ones, 20 identify a separate effects test, and make sure it works 21 over the range of parameters for which you need that model I 22- to work. And then -- but hay time you go to integral 23 effects tests, it is difficult and you just have to do the 24- best you can, and then you have to rely on how smart your j 25 analyst is. ;
| |
| i a
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| p)g
| |
| ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 294 1 DR. ZUBER: To do this - I'm sorry I am taking t''T 2- time, but I mean this is important. We d'on't have any more O 3- integral facilities. How well is our documentation, what is l
| |
| 4 available from the tests we have run? For example, we run 3 5 . tests.at INEL, we run tests at Oak Ridge. Do we have still 6 access to the data?
| |
| 7 MR. KELLY: Well, from my perspective, the best 8 integral facilities are ROSA and Bessie and, of course, 9 LOFT. When you go to the very small scale, but full height i 10' experiments, you bring in a lot -- like semi-scale, you 11 bring in a lot of other phenomena which make -- like heat- .
| |
| l 12 losses, which are the problem with the metal, heat storage 13 'in the metal, that make it much, much harder to get some i l
| |
| 14- good information out of that. And ROSA is still around and
| |
| : 15. still being used, and we have good documentation on that
| |
| ])
| |
| 16 facility in the various --
| |
| 17 DR. ZUBER: And those documents are still 18 available, all of them?
| |
| 19 MR. KELLY: I think so.
| |
| 20 DR. ZUBER: Okay.
| |
| 21 MR. KELLY: And we are working on getting that 22 ' data, you know, making sure the data is on the data bank, L 23' making sure it is usable and making sure we have the r
| |
| 24 documentation. And we are scanning the documentation into
| |
| ~ 25 .the data bank, so we will have it in electronic form as f~ ANN RILEY & ASSOCIATES, LTD.
| |
| . .I Court Reporters
| |
| }, 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 f
| |
| | |
| _. > _ . . _ _ . . . - - . . .s . _. . . . _ - . . . - . . _ _ _ . . . _ . _ _ . . . _ . _ _ _ _
| |
| 295 1 'well.
| |
| l'')
| |
| '%./
| |
| 2 .DR. ZUBER: If I may just give advice, I think you 3 should do'this.as soon as.possible, because people who run 4 -
| |
| who are familiar with LOFT are retired,. dying or leaving.
| |
| S' PeopleLhere are changing. I think this should be.really
| |
| '6 documented if you wantLto make use for this activity here.
| |
| 7 Otherwise, --
| |
| L 8 MR. KELLY: I agree completely.
| |
| 9.- DR. KRESS: Is that box you have up there labeled 10 accuracy?
| |
| 11 MR. KELLY: Yes. What is the judge?
| |
| 12 DR. .KRESS: Well, you know, in the past we have 13 sort of"used a pragmatic approach of how well does the 14 prediction fit within the boundaries of the uncertainty of l- 15 -the data itself, which is a pragmatic. approach, but it 16 always left me a little cold. Do you have more thoughts on 17 how you were going to judge the accuracy?
| |
| L l l 18 MR. KELLY: Well, Farouk mentioned something 19 called the automatic -- automated code assessment program, j; 20 and that is a tool we developed, or had developed at Penn 21 State University, and it provides a suite of numerical l 22 techniques for judging the goodness of fit, if you will.
| |
| ; 23. So, and they recommend two methods. For 24 stationary, good old steady state axial profile stuff, root
| |
| -25 mean squared, you know, that's about as good as you can get.
| |
| e
| |
| ;.,e.s ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 t-Washington, D.C. 20036
| |
| =
| |
| (202) 842-0034
| |
| | |
| 296 1 But how do you compare transients? That has always been a l
| |
| /'^) 2 problem.
| |
| V 3 DR. KRESS: That's right.
| |
| 4 MR. KELLY: And Professor Daria from the 5 University of Pisa came up with a technique using fast i 6 Fourier transforms.
| |
| 7 , DR. KRESS: There is still a goodness of fit, 8 though?
| |
| 9 MR. KELLY: In effect.
| |
| 10 DR. KRESS: The question is whether is that is 11 okay, though. It requires some acceptance level on the 12 goodness of fit and I don't know how you get that.
| |
| 13 MR. KELLY: And we are going to have to develop 14 those. And the idea is that for each -- when we get to the
| |
| (
| |
| n\ 15 point where we are actually developing the matrix, you know,
| |
| %)
| |
| 16 we have identified the tests, an analyst runs them, then an 17 analyst has to, using this ACAP tool, develop the numerical 18 procedure for quantifying it And what is currently 19 recommended is using something called continuous wavelet 20 transforms, which I just know the acronym and not much else.
| |
| 21 DR. KRESS: I know what those are, but don't ask 22 me to explain them.
| |
| 23 MR. KELLY: Yeah, I know they are used in digital 24 signal processing and that is in.
| |
| 25 DR. KRESS: Yeah, they are sort of another version
| |
| /''N ANN RILEY & ASSOCIATES, LTD.
| |
| () Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 297 1 of Fourier transfer, but not. They are another way to do
| |
| ['] 2 that sort of thing.
| |
| G1 3 MR. KELLY: And so doing that, you can get a 4 numerical quantitative value. And then at some point you 5 are going to have a person say, well, a value of this is 6 acceptable and this isn't, and then use that.
| |
| 7 DR. KRESS: That is the part I am interested in.
| |
| 8 How do you get that value?
| |
| 9 MR. KELLY: Probably we'll have peer review 10 process.
| |
| 11 DR. KRESS: It'll probably depend on figure of 12 merit to a particular code.
| |
| 13 MR. KELLY: You know, for a certain type of 14 ,
| |
| transient when you're trying to look for certain things, you x
| |
| ( ) 15 define those key parameters, you know, whatever they might 16 be , like core collapsed level.
| |
| 17 DR. KRESS: Right.
| |
| 18 MR. KELLY: And maybe half a dozen, 20 of those 19 for an integral test. For those 20 parameters you would 20 basically throw up the data, throw up a couple different 21 code calculations, go tnrough the suite of numerical tools, 22 and say okay, which of these code calculations is the best, 23 and how do these numerical rankings go with that, and select 24 it that way. And that's, you know, it's better than what 25 we've done in the past, but nothing's going to be perfect.
| |
| ! (~'s ANN RILEY & ASSOCIATES, LTD.
| |
| ,() Court Reporters i
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| _ .. . . .. ...~ . . .~ _ - _
| |
| 298 11 But we are trying.
| |
| ./~} 2 .DR.~ WALLIS: Sometimes what1you find is that the l G
| |
| F code which'has.the worst physics fits the data best.
| |
| 4 DR. KRESS: Yes.
| |
| 5, MR. KELLY: Compensating errors; yes.
| |
| 6 The focus of this presentation is really these two 1
| |
| 7 boxes, but since you ask about accuracy, okay, in the sense 1
| |
| : 8. of talking about the-code consolidation, an'd that's what I'm ,
| |
| 9' talking about now, not code improvement, and in particular 10 I'm talking about BWR components, our metric here is it has 11 to be at least as accurate as the current version of TRAC-B. ;
| |
| I
| |
| .12 It has to be at least as robust and at least as fast. And i L 13' so far for the test. problems we've been looking at, the 14' TRAC-M code run for BWR-type problems is five times faster
| |
| {) 15 than TRAC-B. So efficiency and robustness, we're doing 16 great on those so far.
| |
| 17 But accuracy might be a question for certain 18 things, like say void fraction and rod bundles. We may have 19 to improve that.
| |
| 20 Now this is kind of the overall process made 21 simple, but basically you figure out what you're
| |
| ! 22 simulating -- j 23 DR. ZUBER: What package are you using to 1 24 determine the voids? !
| |
| '25 MR. KELLY: Okay. I wish you wouldn't say "you !
| |
| I i
| |
| 4-l / ANN RILEY & ASSOCIATES, LTD. l
| |
| }; ( Court Reporters '
| |
| .. 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 l (202) 842-0034 L
| |
| i
| |
| | |
| 299 1 .are using, '' but what 's in the TRAC-P. code --
| |
| 1(~ 2. DR. ZUBER: I mean, this code here, or what are V.
| |
| 3 you going to use?
| |
| 4 MR. KELLY: CH1, I don't know yet Okay? Because 5 it's going to come out through the assessment. When we get 6 .to this point, and we're not there yet, one of the 7 components that's being installed now is the BWR Chan 8 component, basically the fuel element assembly. Part of the 9 assessment matrix for that will be things like the FRIGIF 10 test, which is subcooled and saturated void. We will do 11 quantitative comparisons between TRAC-M, with its current 12 interface drag package, and the TRAC-B code, and see which 13 is more accurate.
| |
| : 14. EDR . ZUBER: Well, that package was extracted from 15 a drift model or how --
| |
| -s 16 MR. KELLY: That's my understanding; yes.
| |
| 17 DR. ZUBER: It would be then the same package in 18 this code.
| |
| 19 MR. KELLY: Maybe. If TRAC-M is more accurate 20 now, we won't change the physical model. If it's necessary 21 to put in a specialized model for a boiling water fuel
| |
| ! 22 element, for interface drag, it'll be put in, but as a j 23 component-specific package. So it will only affect BWR fuel
| |
| , 24 bundles, and not drag in pipes, which is what's happened in 1
| |
| 25 the past.
| |
| 1
| |
| .(N ANN RILEY & ASSOCIATES, LTD.
| |
| i
| |
| (~) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ;
| |
| Washington, D.C. 20036 1 (202) 842-0034 )
| |
| c l l l l l
| |
| | |
| 300
| |
| >1L DR. ZUBER: 'In what way will the work at Purdue-
| |
| :2
| |
| ;with the interface tracking affect this modeling?
| |
| 3' MR; KELLY: Okay. TheJprogram'at Purdue is a 4 'long ' term exploratory research effort, with two goals' One .
| |
| 5 'is to develop a data base that we.can use to check 6 fundamental' quantities-like bubble diameter, interfacial :
| |
| '7- area. The other side of it is developing a way of modeling 8 the processes that affect the evolution of interfacial area 9 in time and' space. !
| |
| 10 I-have in the past written codes that included an 11 interfacial area transport equation that you then use 1
| |
| 12 ' instead of'the flow regime package. But that's years down f 13 the road. .That's not part of code consolidation at all.
| |
| 14 DR. ZUBER: But you would have also to take into f 15 account the' drag, I mean, the interfacial drag. That would
| |
| .16 be the whole package. It would be not just tracking the 17 interface, it would be also having a better or same model 18 incorporated for'the drag, interfacial drag.
| |
| 19 MR. KELLY: Right. From the interfacial area and 20' void fraction you can get something called the geometric 21 intensity, which can give you an idea of what the shape 22 looks like, the shape and size of the particles. Then from 23 there you can go to drag relations form and then do that for 24 a mixture of particles.
| |
| 12 5 DR. ZUBER: But would.this.then change the i
| |
| l.
| |
| ' ~
| |
| N i' IdRI RILEY & ASSOCIATES, LTD.
| |
| ss Court Reporters 1025 Connecticut' Avenue, NW, Suite 1014 Washington, D.C. 20036 L (202) 842-0034 t
| |
| l-
| |
| | |
| 301 1 approach you are doing now in this first part when the l p) e 2 structure of the code and how do you --
| |
| . %.J 3 MR. KELLY: It would make changes, yes, but not 4 radical. In the COBRA / TRAC code, which you heard about this {
| |
| 5 morning, and which I worked on when I was at Battelle 6 Northwest, has an interfacial area transport equation in it 7 for the drop size, and that's why he could talk about 8 bringing -- or, excuse me, they could talk about bringing in 9 a flow from a boundary condition as a continuous liquid or 10 drops, and could talk about bringing in the drops with a 11 specific diameter. You can't do that in TRAC or RELAP. i 12 And that was put in very straightforward. It was 13 very easy to do. And I think we can do that here. The l 14 catch is developing the physical models for bubble
| |
| ,o) i 15 coalescence and breakup, especially in things like a reactor !
| |
| 16 core. You know, it's hard enough in a long, straight pipe, 17 where you can approach fully developed conditions.
| |
| 18 DR. ZUBER: My own guess, it would be worse in the 19 upper plenum, because there you have a kind of -- drops 20 everything. I mean, that would be -- and this is really 21 important for these new applications. It's not only in the 22 core, it's the upper plenum.
| |
| 23 MR. KELLY: Yes, and one of the things we're
| |
| ! 24 moving towards is things that are more and more j- 25 component-specific, where if you determine that you have L
| |
| h, ANN RILEY & ASSOCIATES, LTD.
| |
| \ss/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ;
| |
| j Washington, D.C. 20036 '
| |
| l (202) 842-0034 t
| |
| i i
| |
| | |
| t 302 1- problems-in the plenum, you'go ahead and start developing a 2 constituent model package for that.
| |
| .3 DR. ZUBER: 'I hate to monopolize the discussion, !
| |
| l 4 but -this iis where the modulization comes.
| |
| l 5 MR. KELLY: Right.
| |
| j 6 .DR. ZUBER: .Are you going to discuss it today?
| |
| 7 MR. ELTAWILA: If we continue in this way, we may 8 be on Friday.
| |
| 9 [ Laughter.]
| |
| 10- MR. KELLY: That's in the second presentation.
| |
| 11 It's:actually an example of a modularization.
| |
| 12 DR. ZUBER: Fine, fine, fine, fine, fine. Go 13' ahead.
| |
| 11 4 DR WALLIS: Are you happy, Joe, to keep going j f-\. !
| |
| 15- until you finish?
| |
| i }
| |
| 16 MR. KELLY: I don't mind, because I spent the time 17 preparing these presentations, but you may mind sitting 18 here.
| |
| 19 DR. WALLIS: It's taking time, yes. I' don't want 20 to miss anything important.
| |
| 21 MR. KELLY: So for installing the BWR components, 22 it's really these two boxes here. You have to define what 23 the modeling requirements are and implement the model and 24 'make sure.it meets those requirements.
| |
| 25- Now as I show in the next slide, this is actually i
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| : l. O( ,/
| |
| , Court Reporters l
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l 303 1 seven' steps. To go to the. perform code assessment, that's i
| |
| ((~ 2 .when weihave.to have all of the BWR models installed. And
| |
| %)
| |
| 3 'so_'all'of this is downstream. This is going to start fiscal 10 year 2000. Well, okay, so this summer or this fall is when L5 we'll-be starting that. ;
| |
| 6 Enough on that slide.
| |
| ~7- These are -- to go from specifying the ;
| |
| 8 requirements to the implementation of the model, these are
| |
| '9. the seven steps. Define the functional requirements. I'm 10 going to show you those. From those develop a verification 11 test plan to make sure you've satisfied those requirements. j l
| |
| 12 Then you have to implement the model, and that's basically '
| |
| 13' .three steps.
| |
| 14 The first is you have to identify what TRAC-B 15 coding comprises that model, because if-you look inside the 16- code, there's no jet pump model, and everything that has to 17 do with the jet pump is in one place. Instead, their "if 18 test" in the momentum equations, there's "if test" in the 19 input, there's "if test" in the interface drag package,
| |
| :20 which says if this is a T and this T is modeling a jet pump, 21 then add this momentum source term or do something else. So 22 you have to cull out all these various exceptions from the 23 TRAC-B code, because the TRAC-B code is not modular.
| |
| 24 So once you know what comprises the model, then
| |
| : 25. you' create a component type in the data-base structure in l
| |
| i r ANN RILEY & ASSOCIATES, LTD.
| |
| s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036
| |
| !: (202) 842-0034
| |
| | |
| ~ . . . _ . ~ ... . . _ _ . . . - - - - . . . - - . - ~..
| |
| 304 1 ^ the TRAC-M code.. And now in this particular case it doesn't
| |
| [O L '2-3
| |
| .do.anything yet, it's just there.
| |
| the code for where.you're going to put it.
| |
| It gives you the space in Then'you recode 4 the model-into TRAC-M in component-specific routines.
| |
| 15' DR. KRESS: .It's the same model that's already in ,
| |
| 6 there.' i
| |
| '7 MR.' KELLY: Right.
| |
| 8 DR. KRESS: Just move it over.
| |
| g 9 'MR. KELLY: 'Yes. And the idea is to try to get
| |
| -10 . exactly the same result.
| |
| 'll DR..KRESS: You want to get exactly the same 12 result. Then if you need-to improve the model, you can do
| |
| '13 that.
| |
| 14 EMR . KELLY: Exactly.
| |
| p.
| |
| .( ) 15 Let's consolidate these four codes -- we can only
| |
| ~
| |
| 16 support one and so we can all learn how to use one.
| |
| 17 DR. ZUBER: Howiis it different in numerics 18 between TRAC-B and TRAC-M?
| |
| 19 MR. KELLY: TRAC-M now has two types of numerics 20 in it, something called SETS, which is Stability Enhancing 21 Two Step, and it is the method Professor Mahaffey came up 22- with to allow violation of the corant stability limit but we 23- reimplemented the semi-implicit method, which was the 24 previous version and what is good about this semi-implicit, 25 your time steps have to be lower than the corant limit, but l-i
| |
| ! gr o ANN RILEY & ASSOCIATES, LTD.
| |
| l\ -
| |
| Court Reporters L 1025 Connecticut Avenue, NW, Suite 1014
| |
| [ Washington, D.C. 20036 l .(202) 842-0034 I
| |
| | |
| l 305 l i
| |
| 1 mass and energy transport are explicit in nature, which I 2 means there's less numerical damping associated with it and
| |
| ( b(~%
| |
| 3 so the reason we did that was for BWR stability analysis in l l
| |
| 4 the future. l
| |
| )
| |
| 5 The TRAC-B code has semi-implicit but also l i
| |
| 6 something called nearly-implicit, which is an INEL attempt '
| |
| 7 to go beyond SETS, but as I told you, TRAC-M for these 8 problems we have looked at so far is five times faster, so 1
| |
| 9- we are not going to change that. I 1
| |
| 10 Once you get the model in -- l l
| |
| 11 DR. ZUBER: And you are going to use the same 12 numerics for PWR and BWR?
| |
| 13 MR. KELLY: With the caveat that you switch to 14 semi-implicit if you are doing something like a stability I
| |
| Q/
| |
| ) 15 analysis where you have to worry about numerical 16 dissipation.
| |
| 17 So once the model is in, you write a software 18 design description document, then you perform the 19 verification test plan and you compare the analytical 20 solutions and TRAC-B results. I am going to show you some 21 of this.
| |
| 22 You prepare a verification test report, then you 23 transmit the model and the documentation to NRC l 24 Configuration Control. The person here will review the l 25 documentation to make sure it is completc, install the model l-i (3 ANN RILEY & ASSOCIATES, LTD.
| |
| s ,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l (202) 842-0034
| |
| | |
| 306 1- in~the current development version of the code, and repeat 2- the verification test on a number of platforms at the NRC.
| |
| (
| |
| 3 These are things that have never been done before.
| |
| 4 DR.-ZUBER: Who is that person?
| |
| 5 MR. KELLY: Biro Atkas at Scientech is the one who 6T is doing this for us. '
| |
| '7 -DR. ZUBER: Okay. l 8 'MR. KELLY: And to give you an idea about what is 9 involved here, this is a completion report for the jet pump. l 11 0 It comprises four-documents in one -- the software and '
| |
| 'l 11 verification test results.
| |
| 12 The second one is for -- this is ridiculous -- and 13 the third one is for.the turbine.
| |
| 14 i. DR. WALLIS: Would you~ speak to the record and you i
| |
| 15 -can speak to us.
| |
| 16 MR. KELLY: It would certai...'y be easier with a 17 throat mike but -- I like to work the audience when I'have a 18 chance.
| |
| 19 So before we talk about installing the model,.what L20 is the model? So this is a caricature of a jet pump, the 21 ~ drive line, the mixing region and the diffuser, and I am 22 ftalking about the TRAC-B jet' pump model and as I said it is 23 not.a stand-alone component. It doesn't have.its own 24 numerics and.it doesn't have its own physical models with a 25- few exceptions.
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| 3O. Court Reporters 1025-Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l 307
| |
| .1 .Instead, it uses a. TRAC-B Tee component ~and that' '
| |
| 2 gives it the;one-dimensional, two-fluid numerics and 3- . constitutive'models, things'like wall drag.
| |
| 4- The Tee model in TRAC-B-is not perfect, so it adds 5 momentum source sink terms to account-for inaccuracies in 6 the: momentum flux differcing at area changes. That would be !
| |
| 7 here,'~here, and here.
| |
| -8
| |
| ~
| |
| l It also adds special form loss models for
| |
| :9 incomplete mixing and the irreversible loss at the nozzle 10: and-those'were backed-out at the INEL scale test.
| |
| 11 So if you look at the TRAC-B Jet Pump Model, it is 12 a tee. It changes-to the code database in I/O to allow for 13 a component called at jet pump. These momentum flux '
| |
| 14 corrections I talked about and the special form loss I '15 coefficients, so if we are now going to put that model into 16 TRAC-M it -gives you three very d'.stinct development
| |
| '17 stages - . create the place for it in the database, check and
| |
| [
| |
| 18 .see if you need the momentum = flux corrections, and implement 19 the form losses.
| |
| l 20 'DR. WALLIS: This is a two phar atodel?
| |
| !, '21 'MR. KELLY: It's a two phase model, yes, but --
| |
| 22 DR. SCHROCK: Is it liquid driven --
| |
| f- 23 MR. KELLY: I may have to have Biro answer this 24 'but all'of the testing that was done in TRAC-B was only with 25 single phase, single phase liquid, the 1/6th scale test.
| |
| l h
| |
| !O V 3 ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
| |
| I , . , . ,. , . . . . . - , . ._ .m. .
| |
| | |
| l 308 1 There is no verification existing for this model in two l
| |
| () 2 3
| |
| phase. That is a deficiency.
| |
| DR. WALLIS:
| |
| l Are you raying is it a two phase 4 prediction? l
| |
| . 5 MR. KELLY: It is my understanding that most of t 6 he times when it is important it is single phase but that is
| |
| ,7 what will come out of the assessment matrix.
| |
| i ,
| |
| 8 If it is a high priority item when it is two '
| |
| 9 phase, then we can look at maybe doing a small-scale 10 separate effects test.
| |
| 11 DR. ZUBER: But there is data on that, quite a few 12 data -- I think given the Russian data, actually even in 13 this country it was air and water mixing in jet pumps or 14 something like that.
| |
| /N
| |
| () 15 MR. KELLY: .I am not the jet pump expert but we 16 will certainly have to find data for it. I am glad to hear 17 that because most of what I have heard is that it is 18 proprietary, you know, GE stuff that we can't get.
| |
| I 19 DR. ZUBER: There are other data. I will look for 20 it.
| |
| 21 MR. KELLY: Did you hear that, Biro? Okay. So 22 one of our objectives is to have a scrutable process. The l
| |
| 23 first step of that is what does the component have to do?
| |
| 24 That is ralled functional requirements.
| |
| 25 Now there are going to be actually 12 of these but l
| |
| i
| |
| /]
| |
| \/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
| |
| | |
| 309 1 only 7 that really matter.
| |
| ; 2
| |
| ['L J The first one is hydraulics. I talked about the 3 first step being create a place for it in the database, so 4 you have created a jet pump component that is nothing more 5 than a tee, so that jet pump component should give you 6 exactly the same results as a tee. At this point you 7 haven't changed anything except calling it a jet pump and 8 storing the information in a different place so they should 9 produce identical solutions.
| |
| 10 Reversible pressure gains and losses, and I am 11 talking about Bernoulli type changes due to area changes, 12 and they should be accurately calculated by the 1D two-phase 13 flow model.
| |
| 14 Pressurized due to mixing of the suction and drive
| |
| /^x v
| |
| ) 15 flows, and again there was a deficiency in the TRAC-B model 16 for where the side leg of the tee, and so that has to be 17 handled correctly so that we don't need to put a correction 18 factor in for it.
| |
| 19 Irreversible pressure losses -- that's what I just 20 talked about.
| |
| 21 Being scaleable, able to handle multiple -- you 22 know, you use one jet pump to model, several --
| |
| 23 DR. WALLIS: How do you check scaleability?
| |
| 24 MR. KELLY: That is really difficult and in one 25 journal article there was some full-scale data and there
| |
| (''h ANN RILEY & ASSOCIATES, LTD.
| |
| 'w l Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 1 310 1 were TRAC-B and TRAC-M comparisons to that, but there are a 2 lot of uncertainties -- some of the geometry that had to be 3 guessed at and one of the things that we think we need to s improve is the loss coefficients. They are not physically 5 based.
| |
| 6 The loss coefficients from the 1/6th INEL scale 7 data, they are not in terms of thinga like Reynolds numbers 8 like you would expect. Instead, it is in terms of the M 9 factor -- that is the ratio of the suction flow to the 10 driver flow and velocity at the nozzle.
| |
| 11 That is not exactly correlating parameters, so 12 that is a place where we know we can improve the model.
| |
| 13 The last ones are IO requirements.
| |
| 14 As I said, for each functional requirement you
| |
| ) 15 have to have a test, one or more tests, to make sura that 16 requirement is met, so for the first seven of those, the 17 other ones were IO stuff, we have a verification test case 18 for each of these. I am only going to discuss three of 19 these, numbers two, three and four.
| |
| 20 Number two is reversible pressure gains and 21 losses. Again, this is looking at the pressure change as 22 you go through an area change because there were 23 deficiencies in the TRAC-B momentum flux model.
| |
| 24 They had to put in source / sink terms. Well, are 25 those needed in TRAC-M? It turns out they are not.
| |
| C, ANN RILEY & ASSOCIATES, LTD.
| |
| mj Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l 311 1- The momentum flux differencing is done more
| |
| [ [%.)~)
| |
| 2 accurately in TRAC-M. It is much more like a central 3' difference formula, but to test it we used the input model 4 for the 1/6th scale, single phase, frictionless, no gravity, 5 and we compared Bernoulli number for all the cells in the I 1
| |
| 1 6 diffuser and tailpipe sections and the cells in the nozzle '
| |
| 7 and riser.
| |
| l 8 When you do that, and here you can compare the 9 code result to a analytical value because the Bernoulli 10 number should stay constant, we don't have reversible 11 losses. We turned off the friction. I mean irreversible 12 losses.
| |
| 13 So the difference between analytical value and the 14 code was 0.1 percent and we deemed that acceptable, and so
| |
| '3
| |
| /
| |
| j t
| |
| 15 in this case there were no correction factors for momentum 16 flux terms needed in TRAC-M.
| |
| 17 DR. SCHROCK: You have a little embellishment on 18 there that's not on our -- I was just about to ask what is 19 the Bernoulli value. I am on the wrong page.
| |
| 20 MR. KELLY: Page 12. It's just 1 over Density 21 times the Pressure plus one-half V-squared for single phase 22 liquid.
| |
| 23 DR. SCHROCK: But you are talking about a two 24 phase application, so what do you --
| |
| 25 MR. KELLY: Remember, I said the model has only,
| |
| /'''% ANN RILEY & ASSOCIATES, LTD.
| |
| As _) Court Reporters l
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| .. . . _ ...-.__m _. . . _ _ _ . _ _ _ .___-_. _ __.
| |
| 312 1 in the entire history of TRAC-B, it has only been assessed
| |
| ) 2 for single phase liquid and so that'is what we are doing as
| |
| ( 3 part of this, because our metric is to recreate what is in l -4 TRAC-B and be as good as that.
| |
| 5 Now when we start doing the real assessment, then 6 'we will identify model deficiencies and target where we need 1
| |
| I 7 to make improvements. l 8 DR. WALLIS: This is not surprising if you put in l 9 the right physics and the math, ano then it should at least 10 come up.with a Bernoulli equation.
| |
| 11 MR. KELLY: That's true. TRAC-B did not and they 12 had to go back and put in a correction factor and if you ;
| |
| '13 went to the early -- j 14 DR. WALLIS: They should go back and put in the
| |
| ("%
| |
| .g -15 right physics.
| |
| 16 MR. KELLY: Well, eventually they did but if you 17 went and looked at the initial TRAC-P-1-A and some of the 18 input decks, and this is because of differencing, you would 11 9 see negative loss coefficients stuck in to make.up for the 20' extra pressure loss you got from the numerics. It is one of 21 the things that doesn't come out until you look ht a lot of 22' thought problems and that is one of the positive things 23 about doing something as simple as this.
| |
| 24 DR. ZUBER: Let me ask you, how long did it take 25 you to develop this thing from scratch? You started this f ANN RILEY & ASSOCIATES, LTD.
| |
| !lss -
| |
| Court Reporters
| |
| . 1025 Connecticut Avenue,.NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 313 l 1- and then you ran these tests?
| |
| l' e s i
| |
| ( 2 MR. KELLY: You-mean to do the whole process for ,
| |
| 3- 'the jet pump -
| |
| how.many staff months was it, Biro?
| |
| 4- MR. ATKAS: Would you repeat the question, please?
| |
| 5 DR. ZUBER: .How long did it take you to develop 6' this model from scratch to put, to state the requirements, 1
| |
| '7 then to put them down, to run it and to document it? I
| |
| ! 8 MR. ATKAS: Five months.
| |
| i.
| |
| L 9- DR. ZUBER: Five months?
| |
| 10 MR. KELLY: Yes.
| |
| l 11 DR. ZUBER: One person?
| |
| l 12 MR. KELLY: Yes.
| |
| 13 DR. WALLIS: Now this .1 percent, this error in P L 14 'over O plus half V-squared?
| |
| () 15- MR. ATKAS: To make the test simpler, I kept the l
| |
| 16 : jet pump horizontal so I avoided the effect of gravity.
| |
| i 17 DR. WALLIS: It's a little funny because you could l
| |
| 18 just put 1000 psi on everything and it would make you look
| |
| ! '19 better because percentage-wise but it really is not a good t
| |
| 20 test -- percent of what is an interesting question.
| |
| 21 MR. KELLY: Well, this one will give you a little l
| |
| 22' bit better feeling.
| |
| 23 MR. ATKAS: Let me be a little clearer about the l
| |
| .24 ' TRAC-B not being able to predict the Bernoulli gains or 25 losses accurately. It goes back to the use of Upton
| |
| $ ANN RILEY & ASSOCIATES, LTD.
| |
| ; Court Reporters
| |
| : 1025 Connecticut Avenue, NW, Suite 1014 I Washington, D.C. 20036 l
| |
| i-
| |
| -(202) 842-0034
| |
| | |
| l 314
| |
| : 1. differencing.for the momentum flux terms and the code was
| |
| ()
| |
| i 2- only using'Upton -- there was not a special treatment for l
| |
| 3- : area changes and. TRAC-M had this special treatment.
| |
| l E4- DR. WALLIS: A better test is the change in the
| |
| .5 Bernoulli number. compared with a half V-squared because that i
| |
| 6.. is what'is really driving it and that is -- the base is sort 7 of irrelevant. It is the wiggles.
| |
| l.
| |
| 8 MR.. KELLY: I.see what-you are saying and you will 9 see that to some extent here.
| |
| l 10- This was again reversible pressure changes but in 11 this case due to the mixing of the suction and dry flux and 12 .again this was a deficiency in TRAC-B that they had to put a 13 correction in for, so now we are testing TRAC-M to see if we ,
| |
| l- 14 need that correction, so you mix the two streams and lock at
| |
| ? >
| |
| ( '15- the delta-P across this. You can do the analytical and you 16 get 51.3 kilopa.ccals as the pressure change.
| |
| 17 The TRAC-M code gave 51.27. Again it is a simple
| |
| '18 problem but it is something that TRAC-B could not do without i.
| |
| E19 putting an extra correction there and TRAC-M could, so it 20- lactually made the jet pump model a little simpler.
| |
| l 21- DR. ZUBER: This is due to numerics?
| |
| 22 MR. KELLY: Yes -- due to the way the momentum 23 flux term in the momentum equation is differenced.
| |
| '24 And now some data, since I have showed a R25 comparison analytical. This is the irreversible pressure i
| |
| ,v [] ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters I. 1025 Connecticut Avenue, NW, Suite 1014 i l Washington, D.C. 20036 (202) 842-0034 c-. . $- , -
| |
| | |
| . . . . _ . - . . -_- - . -.- . . . . . - - - . . ~ . . . . . - . - . . - . . .
| |
| 315 1 losses, and we are going to compare it to the INEL 1/6th
| |
| /' scale test data,- and remember the form losses that were put V) 2 3- in were backed out of this test.
| |
| 4' Basically you ran TRAC-B, looked at the difference 5 bet' ween TRAC-B and the data and got a form loss that made it 6 comeLcloser to the data, so TRAC-B is tuned:to this data.
| |
| 7 What you see is the pressure ratio or N ratio-8 versus the flow ratio and we have positive drive flow here, 9 negative drive flow here, and normal operating conditions 10 are about in this range -- you know, normal, steady state, 11 . full power would be somewhere in here.
| |
| 12' You see the datapoints with an error range. Quite 13 often the error bars are smaller than the symbols I used.
| |
| . 14 It depends on how large the pressure drop was. If the
| |
| () 15 pressure drop is very small and you are taking the ratio of 16 two small pressure drops, then you see huge error bars like 17 this one.
| |
| 18 If your pressure drops are actually very high, and 19 you can measure them accurately then you are not even going 20 to see the error bar here so the blue curves are TRAC-M, and 21 F the red curves are TRAC-B.
| |
| 22 Over most of this region the results pretty much 23 overlap but there are some discrepancies. The biggest 24 discrepancy is here and it is in TRAC-B and it goes back to 25 something Professor Wallis said. He said well, if you have
| |
| ' -NJ[] ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| I I '
| |
| l 316 l i
| |
| 1 - got the wrong physics in -- in this case the physics was a
| |
| () 2 3
| |
| differencing of the momentum equation -- why don't you go back and change it? Well, apparently the history was they 4 came up with source / sink terms based upon what the
| |
| : 5. differencing was to make up for the inadequacies of it, 6 tuned it to the data. A year or more later someone went in i
| |
| 7 and change the way-the momentum flux terms were done and did -
| |
| 8 not go back and change the momentum source terms, so that is 9 why TRAC-B is so bad.right here, i
| |
| '10 Most of the smaller differences are traced to )
| |
| 11 differences in'the wall friction factor, and those are 12- places where the prescure drop is very small, so small 13 changes in what you use for the wall friction, and they are 14 different models. Those would approximately COBRA, for J 15 example.
| |
| 16 So in summary, the TRAC-B jet pump modeling 17 capability has been recovered in a. consolidated TRAC-M code.
| |
| 18 We're using a scrutable, well-documented process, including I 19 these four different documents that have to be delivered.
| |
| 20 MR. ZUBER: How many documents do you think you 21- will need? How many activities like this one will be needed 22 to have a BWR?
| |
| 23 MR. KELLY: Ten.
| |
| 24 MR. ZUBER: Multiplied by four, you're talking 25 about 40 documents. No. By three.
| |
| i 4
| |
| [~' -
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . . . _ . _ . _ - . - . _ _ _ - - _ ._. _ _ - . _ _ . _ _ . _ _ _ _m.. . _. . . . _ . _. _ _ . _ .
| |
| L
| |
| ! 4 317 l g 1 MR. KELLY: Yes. Well, this is four documents t- . .
| |
| I~'j 2 -combined, j
| |
| \_/ \
| |
| 3 MR. ZUBER: I see.
| |
| ~
| |
| L 4 MR. KELLY: So the idea is they come out as 5- chapters. One chapter is the requirement specification. 1
| |
| : 6. But the reason I list them this way is they are delivered to 7- 'the NRC'this way. They sent us the requirement l
| |
| l ,
| |
| 8 specification for a particular component. We review it. I I
| |
| l 9 And for a couple of the first ones, we went through two to j 10 three iterations on these first two. But since then, you l 11 know, we have de-bugged the process, if you will, and now ;
| |
| 12 it's not quite -- it's not quite so cumbersome.
| |
| 13 MR. ZUBER: But how many documents are you going i
| |
| 14 to need?
| |
| l
| |
| () 15 MR. KELLY: There will be ten like this.
| |
| 16 MR. ZUBER: Ten like this.
| |
| 17 MR. KELLY: And we have them in hard copy, but as 18 part of the delivery of.the component, it comes on a CD-ROM 19 with a PDF and a. frame-maker file for the document. So all
| |
| -20 the documentation is in electronic format, and you can go 21 actually on the Web page and look at the' documentation at 22 the same. time as-looking at the test problems and the 23 results of the test problems.
| |
| 24 MR. ZUBER: When do you think you will have
| |
| ; 25 everything finished, all these ten documents?
| |
| t.
| |
| !O F( /
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| [. 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 318
| |
| : 1. MR. ELTAWILA: This summer.
| |
| () 2 MR. KELLY: The end of this summer. And that's 3 when the real assessment will begin, including the integral 4 tests.
| |
| l 5- MR. ZUBER: That's good.
| |
| i ,
| |
| ! 6 DR. WALLIS: Joe, this is a very good process.
| |
| 7 This is a'very nice example. But it strikes me that it's an 8 extraordinary simple one. I mean, this is single-phase flow 9 with Bernoulli-type.or mixing-type behavior, and you already 10 found there were quite a few glitches to fix, if I can call 11 them that, or there were various historical changes which 12- needed to be found out and fixed up.
| |
| 13 MR. KELLY: Well, most of those were in the TRAC-B 14 code, and they were things that we did not need to bring t
| |
| () 15 because TRAC-M was better.
| |
| 16 DR. WALLIS: This is for a simple, almost sort of 17 . undergraduate type single-phase flow --
| |
| 18- MR. KELLY: That's true.
| |
| 19 DR. WALLIS: When you go on to two-phase, you're 120 going to find all kinds of skeletons.
| |
| 21 MR. KELLY: And that will show up probably in the
| |
| '22 next component to be delivered, which is mid-January and is 23 the Chan component, the BWR flow channel, because that has 24- thermal radiation, it has two-phase flow, it has CHF or --
| |
| : 25. excuse me --
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| : s )-~ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 g Washington, D.C. 20036 (202) 842-0034
| |
| | |
| _ ._ . . _ . _ . . . . _ - . _ . . .m._ _ . . _ . . . - _ _ , . . _ . _ _ _ _ _ . . . _ _ _ . _ - _ _ _ . . . . -
| |
| I.
| |
| 1 l
| |
| 319 !
| |
| l 1 DR. WALLIS: All'right.
| |
| 'l -
| |
| .2 MR.' KELLY: -- DNB.
| |
| lV
| |
| ! 3L And the last thing I wanted to say was that the
| |
| : 4. improved treatment of the momentum-flux turns in the TRAC-M 1
| |
| 5 code actually made this process easier than it was j
| |
| : 6. : originally in the TRAC-B code. And with that, I'll let Dave 7 give.hisidemonstration, rad then I'll come back --
| |
| l 8 DR.'WALLIS: How long --
| |
| 9 MR. SINGH: It will'only take five minutes, and we
| |
| ^ 10 think it's really worth going through it for you.
| |
| l l 11' DR. WALLIS: How is the committee's endurance? l 12- MR. ZUBER: It's okay.
| |
| 13- DR. WALLIS: Okay. Good.
| |
| 14 MR. EBERT: This is the beginning of the -- I'm
| |
| , f~
| |
| t 15 sorry -- this is running backwards, but this is a main steam s '
| |
| 16 line break transient, and this is, of course, faster than 17 real time. It's not being calculated; this is just a power 18 point file that was generated. But things go by pretty fast 19 here, so just sort of watch it. It's going full power, 20 oscillating around and the steam line breaks, then the rods 21 scram, goes down to ten percent' power, starts building up.
| |
| 22 Notice the changes in the radio power 23 distribution. It goes up to almost full power here. And 24 even though it's sub-critical, it almost goes up to full 25 power in certain channels.
| |
| l ANN RILEY & ASSOCIATES, LTD.
| |
| L'k Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 L
| |
| Washington, D.C. 20036 (202) 842-0034 l
| |
| l
| |
| | |
| i
| |
| ' 320 I
| |
| \
| |
| ; 1 You can see vividly why a point reactor code has
| |
| .l
| |
| ' /~'T - 2 no chance at all of tracking this transient.
| |
| V That's the r
| |
| 3- nodal calculation for the main steam line break. I H
| |
| 4 DR. WALLIS: I have a very simple question. If 5 it's sub-critical,. how does the power go up?
| |
| 6 MR. EBERT: It's what's called sub-critical
| |
| (
| |
| l 7 multiplication, and if you think about it, you have to.have
| |
| ;- '8 sub-critical multiplication in order to go'from low power, l 9 zero power, up to full power. And'what happens is as you L 10 increase-reactivity, the neutron -- the total chain reaction i 11 length increases until you're critical and -- 1 12 DR. WALLIS: You have to be awfully close to 13 critical, don't you?
| |
| 14 MR. EBERT: You have to in percent terms. When
| |
| () 15 you scram this reactor, it was about five percent 16 .sub-critical, so that means a K effective of 9. -- or .95.
| |
| 17- So.you're not really that far sub-critical even when you've
| |
| -18 'got all the rods in.
| |
| , 19 DR. KRESS: That almost just says you could have i 20 taken that region over there and, with proper design, could
| |
| , 21. have made it critical by itself.
| |
| p I
| |
| l .22 MR. EBERT: It is almost critical by itself.
| |
| 23- DR. KRESS: Yes. I mean, it just means that's 24 almost a reactor -- on its own --
| |
| 25 MR. EBERT: Here's the same situation with the pin i
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| s -Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| : l. Washington, D.C. 20036
| |
| [ (202) 842-0034
| |
| | |
| . _ _ _ . . _ _ _ . _ __.m _ , _ _ . . . ._ .. _ _ _ _ _ , _ _ . _ . _ _ - - _ _ _ . _
| |
| o 321 l' 1 -power,.so you can see the rods scrammed and you're getting
| |
| ' '('T - 21 ; increases in -- and you can see more vividly that.it's
| |
| . l\ ) '
| |
| 3 getting up to 100 percent power even though it's ---the 4 average core power is way down to about 30 percent.
| |
| 5' DR. FONTANA: All rods scrammed here? Did one get i.,
| |
| 6 ' stuck?
| |
| 7- MR. EBERT: Before -- where the big increase in 8 power is where the -- this is only part of the core.
| |
| : 9. DR. FONTANA: .Right'.
| |
| 10' MR. EBERT: And the stuck rod is right there.
| |
| - 11 DR. FONTANA: Oh, okay.
| |
| 12 MR. EBERT: Okay. One last one here. This is l
| |
| 13 even more interesting from my-point of view. This is the l 14- rod ejection case of pin power, and here the rod ejects in a
| |
| () '15- tenth of-a second. The rod is ejected, you're getting a 16 prompt jump here Notice you're at ten to the minus fifth,
| |
| - 17 and then you all of a sudden go to about a few percent
| |
| - 18 power". But it's building up and it goes up to about five l'
| |
| L - 19 percent of full power in a half a second. And so we can 20 . calculate easily'now the total energy deposition in the
| |
| [
| |
| ;. 21 maximum fill pin or any fill pin in this transient. So we 22 basically have -- we do see uncertainties in this rod g
| |
| j 23: ' ejection case. And again, this is for high burn-up, so this
| |
| : 24 is exactly what we're looking at in the fuels program, i.
| |
| 25 As I said before, this -- our work will be 4
| |
| . ', ANN RILEY & ASSOCIATES, LTD.
| |
| \ Court Reporters g 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034 4 - ,- , . . . , . , , . , , , _ , , . _,m. ,.x.- . , _ _ _ _ _ _ , - _ , , . , , ,-
| |
| | |
| 322 1 compared with the Russian work, which has a different code
| |
| ''N 2 that analyzes this, and also the French are running through (O
| |
| 3 exactly the same problem. So this will be a good 4 benchmarking of the pin power construction because it will 5 be comparing three different methods for the same reactor.
| |
| 6 MR. SINGH: Thanks.
| |
| 7 MR. EBERT: Okay. Thank you for sticking around.
| |
| 8 DR. WALLIS: Do we have other questions? I don't 9 think so. Do we get Joe again?
| |
| 10 MR. KELLY: Yes.
| |
| 11 DR. WALLIS: Are you going to finish up? You've 12 got ten minutes to finish up.
| |
| 13 MR. KELLY: Ten minutes? Yes. Let's do it in ten 14 minutes, f~'s
| |
| ( ; 15 DR. WALLIS: Let's not drag it out. I think we're 16 going to fade if you drag it out.
| |
| 17 MR. KELLY: Okay.
| |
| 18 So thic is talking about code modularization, and 19 I'm going to give an example of the interfacial drag 20 package, which is what I'm working on right now.
| |
| 21 Because of time, I'm going to go very light on 22 modular design strategy. I'm probably only going to show 23 the first two slides.
| |
| 24 A little bit of background on what's on the TRAC-P 25 package now. Then I'm going to show again the process, and ANN RILEY & ASSOCIATES, LTD.
| |
| (''/
| |
| (_,
| |
| )
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034
| |
| | |
| , - - . . . - . ... - . .~..- - . - - - . - . - -. ~ . . -- . . - - . - . . - . . . -
| |
| I ;
| |
| 323 i-1 then some results. So I'll again show software 4
| |
| (''y -2 requirements, verification testing, modular structure that
| |
| - }~;, )
| |
| 3' I'm going'to, and a surface-plot utility to show results.
| |
| 4 In the last couple of yea _a, as we've come in 5 front'of you and talked about the code consolidation 6 process,-the word " modularity" has been bandied about a lot, ,
| |
| 7 and it means a lot.of different things to different people.
| |
| 8 So I want to say something about what it is and why we're 9 doing it. So why does the modularity of TRAC need to be I
| |
| 10 improved? What types of modularity are we considering? And 11 what are the benefits? And I'm only going to talk about the 12 benefits with respect to the interface drag package.
| |
| 13 When you go out in the software development 14 industry, there's a lot of people out there that try to look
| |
| #G 15
| |
| ( j at what makes some software work and what makes others fail.
| |
| 16' Not surprisingly, the most significant factor is how complex 17 the software product is that has to be delivered. So this i
| |
| 18 little plot is the percentage of projects that either f ail 19 or succeed versus complexity measured in something called 20 function points, which I'll mention in a second. i 21 So these are projects that are delivered either
| |
| '22 early or on time, and these are projects that are delayed or 23 canceled. And there was a wide range, from computer 24 operating systems to homework problems. So a large number 25 of projects and,'of course, the smaller the project, the L
| |
| l-(
| |
| ? %
| |
| (~ ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters l1 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l t
| |
| 324 1 more easy it is to do it; the more complicated the project,
| |
| /'' 2 the more difficult. It's no big deal.
| |
| I V) 3 l Well, how does this translate to us? Function i
| |
| 4 points are a complicated software metric. Roughly one 5 function point is 100 lines of FORTRAN source code, roughly.
| |
| 6 And if you only use that as a measure, the TRAC code comes 7 out in this box, with roughly a 60-percent chance of success 8 if we were to do it from scratch.
| |
| 9 But, because the data management inside of TRAC is 10 so complicated, we would actually be over here, and have an 11 even lower probability of success. So how can we increase 12 our chances of success? Well, of course what this plot does 13 is it just confirms your commor. sense. Keep it simple.
| |
| 14 Well, how do you do that? You break the software
| |
| () 15 product down into a bunch of smaller products which each 16 basically do one thing, and they ahould consist of less than 17 about 100 lines of source code. So that keeps you over 18 here.
| |
| 19 That being said, where are we with TRAC today?
| |
| 20 This is a size distribution of subroutines. This is the 21 number of subroutines of a given size range. The good news 22 is an awful lot of the subroutines are very small, and 23 basically do one thing. So about 75 percent of the routines 24 meet the size guideline. But there's also bad news. There 25 are eight subroutines of up to 1,200 lines.
| |
| I l-(''}
| |
| (_s ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l 325 1 DR. ZUBER: Which are those?
| |
| /"'T 2 MR. KELLY: All the ones that do the work, 1D
| |
| ()
| |
| 3 interphase drag, 3D interphase drag, 1D momentum equations, 4 3D momentum equations. You know. So -- and that's the bad i
| |
| 5 news. Almost all of the calculational effort is spent in l 1
| |
| 6 these large routines, which means they are very hard to 7 understand and there's probably a lot of errors in them.
| |
| 8 There are four types of modularity that we talk 9 about putting in the code. What's called high-level 10 functional modularity. These are my definitions. This is 31 what has been extant in both TRAC and to some degree in 12 RELAP forever, and this is why people refer to the codes as 13 modular. And what that means is they are based on major 14 tasks, input processing.
| |
| (,,) 15 There are subroutines that read input. There are 16 other subroutines that perform the transient thermal 17 hydraulics, and other ones that do the output. But in codes 18 that are hundreds of thousands of lines long, that isn't 19 very modular.
| |
| 20 But we have two tasks to help this process along.
| |
| 21 One of them was mentioned earlier, and that's the separation 22 of the input processing from the thermal hydraulics kernel.
| |
| 23 They'll actually be kept as separate codes, and the input 24 processing will go with the user's interface. So it'll make 25 the thermal hydrarlics kernel much smaller and much easier l
| |
| []
| |
| \s /
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
| |
| | |
| I l 326 1 to find your way around in. That is ongoing now. And a l
| |
| [''}
| |
| %/
| |
| 2 separation of the matrix solver from the equation setup.
| |
| 3 Actually when they were setting up the equations, some of 4 the matrix reduction was done in the same routine. And that l 5 way we could not just say plug in a new matrix solver, 6 because you never had an identified matrix that you could 7 . ship to a matrix solver. Instead, ycu started formulating I 8 the matrix, you did some, you know, computations with it, 9 and then you go formulate it some more. And so we have 10 already done this.
| |
| 11 Low-level functional modularity. This is more 12 down at the physical model level. So it's at the lowest 13 level for a section of coding that does one thing, computes 14 say something like the bubble diameter, you separate it out t 15
| |
| [O as a function subprogram, and everything that affects that 16 goes in that one place, instead of having it modify it all 17 over in different parts of the code. And I'm going to give 18 an example of this by doing the modularization of the 19 interface drag package.
| |
| 20 The last two types are component-based, and I 21 define two types of that, internal and external. An 22 internal component is something that uses the underlying 23 numerical scheme of the code like the jet pump or the BWR 24 CHAN component, but comes in with its own constitutive 25 package. An external component is something that would l
| |
| t i
| |
| /O l
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| () Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| l i
| |
| 327 l 1 'actually run in a standalone mode. It would have its own
| |
| ; ('')
| |
| v 2
| |
| numerical scheme'and its own constitutive models. It could
| |
| ~
| |
| l 3 be independent of the code. And you.would then link it via
| |
| '4 .the network solver.. And this is something --
| |
| l 5 DR..ZUBER: Okay. That's-important. Suppose for I f
| |
| 6 one case you needed. inertial effects. It's a.large-break 7: LOCA. You would have the capability. Now'you run another 8 test which lasts for days where the inertial is zilch.
| |
| '9' .Would you be able to separate these two modes of 10 calculations? )
| |
| 11- You are carrying luggage which really -- the code-
| |
| ~12 was using -- doing all these calculations probably on the 13 right-hand side'of your complexity without really much 14 benefit.
| |
| () 15 MR. KELLY: I know what you're saying, but that's 16 an entirely different two-phase numerical solution scheme, 17 and you.could do.that -- we don't have any immediate plans
| |
| '18 for.that. That's not a simple -- that's not like the 19 'RELAP-5 accumulator, which I define as an external 20 component. What you're talking about is something much more
| |
| .21 fundamental than that.
| |
| '22 DR. ZUBER: This came in, AP600, you had this long 23L transient and it took us ages to do one calculation.
| |
| 24 MR. KELLY: If you're using something like the 25 SETS method, and it is able to run at its true maximum time l
| |
| j.
| |
| l' -
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| ! Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 328
| |
| ;' 1 step, which can be say a couple hundred times a carant
| |
| !l[~}
| |
| M /.
| |
| 2 limit, you'can start taking very'large time steps. And then 3 the only difference comes in is if there's something it
| |
| ! '4 would do to pressure wave propagation that is causing L 5 feedback to your solution that degrades it. That's l
| |
| i 6 something we could think about in the future, but we're not I
| |
| ! 7- looking at right now.
| |
| l
| |
| .8 DR. ZUBER: For example, in some of these they are
| |
| ]
| |
| ! 9 looking at certification, and you always carry these 10 inertial terms with you when this is really not important.
| |
| 11 MR. KELLY: Yes. Well --
| |
| 12 DR. ZUBER: This is something that we were 13 . discussing, modularization, something which really, discount i
| |
| 14 part of the modeling for cases you're not interested and you 15 do some rapid calculations and focus on the process you want 16 to look at.
| |
| 17 MR. KELLY: And we're going in a direction that 18 will make that easier, and using the network solver, which
| |
| '19 -is in TRAC now, gives you a place to clip in. So what I'm
| |
| -20 planning for the future is we will be able to develop 21 components that have different. numerical schemes that could 12 2 be used for parts or maybe even all of the system, and this 23 .gives you an environment in which you can plug them in. But
| |
| -24 that's not easy.
| |
| 25: DR. ZUBER: No , no , but this was the quantum jump L
| |
| l I
| |
| i
| |
| ! ANN RILEY & ASSOCIATES, LTD.
| |
| "s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| ' Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 329 l
| |
| 1 in the capability which this Agency would need.
| |
| (~'T 2 MR. KELLY: Well, we're trying to head in that V 3 direction, but no firm commitment to do that. I mean, we've 4 got enough to do at the moment just trying to get the 5 consolidation done. l 6 The other type is simply other computer codes. l l
| |
| 7 One computer code can't do everything, and, obviously, we l 8 shouldn't have developed a neutronics codes inside of TRAC.
| |
| l 9 That would be silly when there are good neutronics codes !
| |
| 10 already out there. So what we did was we identified one and 11 we built an interface between that code and the TRAC code.
| |
| 12 Actually, we built what is called a generic interface and we 13 used that to couple it to both RELAP and TRAC. That way, )
| |
| 14 the PARCS code doesn't have to know anything about the
| |
| [~}
| |
| s-15 thermal-hydraulics code and vice versa. Their database l
| |
| 16 structures are completely separate. '
| |
| 17 Because we are running so late, I am going to skip 18 over the next two slides, which were the benefits of low .
| |
| l 19 level of modularity, and go to what I call background on the 20 interface drag package to current status. All of the 21 interface drag calculations in TRAC are done in two 22 routines. There's one for the 1-D components and one for l
| |
| 23 the 3-D vessel. They are both over a thousand lines of 24 source code in length. They are prime candidates for being 25 broken up into small pieces.
| |
| (~T ANN RILEY & ASSOCIATES, LTD.
| |
| ! (,)
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| i 330 l
| |
| 1 But it is worse than that. For the 1-D I l
| |
| 2 components, all of the 1-D components are done in this one 3 routine, whether it is a vertical pipe, horizontal pipe,
| |
| '4- accumulator, pressurizer, pump, it doesn't' matter, it is all 5 stuck into this one routine.
| |
| l L 6 Also, until very recently, and this is something 7 Jennifer Euell did at the NRC, the interface drag l 8 calculation was buried in a much larger subroutine that
| |
| [ '9 actually set up all the terms in the momentum equations. I
| |
| ?
| |
| 10 mean so there was no modularity there at all.
| |
| 1 L 11 So, at least now, all of the 1-D interface drag is 1 12 in one. For the 3-D vessel, this is the routine, and this l
| |
| l 13 routine calculates interface drag three times. It does it E 14 for the three. coordinate directions. It goes through, in' 15 effect, an entire flow regime map for each coordinate 16 direction. It is not a very good approach. But one thing 17 that is good, they do make what I will call exceptions for 18 the different regions in the vessel. So you will see "if" 19 -tests. You know, if this is a downcomer, you do this. If 20 this is the core, you do this. So at least they do that 21 instead of treating it all like pipes, which is pretty much 22 what is done in RELAP.
| |
| l-23 I am going to make some disparaging comments about 24 'the TRAC coding style on the next slide. Before I do that, l 25 I want to say that it i r, typical of what was employed in the ANN RILEY & ASSOCIATES, LTD.
| |
| : (\ Court Reporters
| |
| . 1025 Connecticut Avenue, NW, Suite 1014
| |
| !- Washington, D.C. 20036 l (202) 842-0034 l
| |
| b
| |
| ~ ._ _. . ,. _ _ .. . _ _ _ -. ._ . --
| |
| | |
| t , ,
| |
| 331 l 1 '70s,-and it is very similar to what you will find in RELAP.
| |
| 2- Actually, it is not quite as bad as what is in RELAP.
| |
| l 3
| |
| i
| |
| .When you.look at the TRAC code interface drag, 4 what are the. problems?. We are talking just at a coding 5 ilevel now. The 1-D and 3-D interface drag packages are not
| |
| > 6- ' consistent ~. If you~,-in effect, model a vertical pipe as a l
| |
| l 71 single stack of axial computational cells with the 3-D l l
| |
| 8 component, you will get a different answer than for a 1-D ,
| |
| 1 L
| |
| 9' pipe for the same, and that's -- you shouldn't. They should l i
| |
| 10 be consistent when you reduce it down to the same noding.
| |
| 11 They also allow a selection of user options lL2 through what is called nameless options, you turn things on 13 and off and then special models pop on and off. So if you '
| |
| 14 are inside the 3-D vessel routine, there will be "if" tests
| |
| () 15 'that says if this node is in the vessel and it is in the 16 downcomer region, and the Blausius model is turned on, then
| |
| , ,17 you go do something different than normal. And this just i.
| |
| 18 gives. multiple user options, and there is no point. This is
| |
| ! 19 a development tool. When you don't know what you are doing l- .20 and you want to test out a couple of models, okay, put them
| |
| '21 in. But once you have tested and make a decision, then take 22 this out of the hands of the user.
| |
| ~23 The coding is enormously complicated. It is
| |
| -24 poorly commented and local variables are not defined. That 25 makes it unreadable. As a developer of the COBRA / TRAC code, i
| |
| I ANN RILEY.& ASSOCIATES, LTD.
| |
| I' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| ! Washington, D.C. 20036 (202) 842-0034
| |
| , . < - , . , , - , - ..-._ __.,,m._ . _ , . = . _ . - ~ , . - . - - ., . .
| |
| | |
| I 332 1 we were working under a tight deadline and we developed the l (~'T' 2 code for ourselves. We never thought about someone 15 years L/
| |
| 3 later who is going to have to maintain that code, and you 4 probably have huge grudges against me because of that. So I 5 will throw some dirt on myself first.
| |
| 6 But, typically, that is what happens, whether it 7 is in our industry or any other industry, the developers 8 develop it for themselves. They understand each individual 9 line of coding that they have written and that is it. They 10 don't think about the future.
| |
| 11 DR. ZUBER: But there is another advantage, you 12 are indispensable. You have a job guaranty.
| |
| 13 MR. KELLY: Well, unless the code is so bad they 14 can the program. The other thing, inputs and outputs to the
| |
| <~N 15 (v) routines are not necessarily defined. It makes it very hard 16 to go through TRAC and see where a variable changes its 17 value, because it can be masked through an argument list, so 18 it is really a maintenance nightmare.
| |
| 19 The coding is repeated four times. I said that in 20 a 3-D vessel it uses -- it goes through the flow regime 21 stuff three times, and then also for 1-D. So if you look 22 for the coding that defines bubble diameter, you will find 23 it in four different places, and that is just in interface 24 drag. Then you can go to interfacial heat transfer and 25 probably see it four more times. I haven't gotten there l
| |
| ('N ANN RILEY & ASSOCIATES, LTD.
| |
| (_,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| _ __ __ _ __ _ , . > _ . . - - -m. _ _ . . --. - _ _ _._- -
| |
| 333 I; l' yet. That is obviously silly, 2 'DR. WALLIS: When NRC gets these vendor codes that L 3 it wants to get, it is going to have the same thing It .
| |
| E l
| |
| 4 wants to get the source codes from the vendor, it is going 5 to find that there is a lot of. history of somebody 30 years
| |
| -6 ago having some correlation put into'something when the code f
| |
| 7 was first developed or something.
| |
| L8 DR. ZUBER: Well, you heard something this 9 morning.
| |
| t 10 DR. WALLIS: Yeah, that's right.
| |
| 11 MR. KELLY: And I have an example coming later 12 about that.
| |
| 13 My other complaint is that component specifics are 14 treated as localized exceptions. If you look through, say,
| |
| ! /~N 15 the 1-D routine, you are not going to see an "if" test and
| |
| ()
| |
| 16 'say,.if this is an accumulator, and then find everything to 17 do with an accumulator there. Instead, it is going to try 18; to-go through the whole flow regime map, and in the middle 19 of it, there is going to be an "if" test, and it says,.if 20- this pipe is really an accumulator, then change this 21 weighting factor to_some value. And so by the time you go
| |
| -22 to the bottom, you don't know what is being used for an 12 3 accumr'.ator. It is not traceable at all.
| |
| 24 If I am going to modularize this, what am I going 25- to do? So these are my requirements, the null effect, all I i
| |
| ' 1,C ANN RILEY & ASSOCIATES, LTD.
| |
| .l\
| |
| Court Reporters
| |
| , 1025 Connecticut Avenue, NW, Suite 1014
| |
| ;. Washington, D.C. 20036 (202) 842-0034 p
| |
| | |
| 1 334 1 am doing is recoding it and reorganizing it in a better i
| |
| ~
| |
| / '% 2 fashion. I should not be changing the answer. They have to V
| |
| 3 be component-specific. I will have a specific routine for a 1
| |
| 4 vertical p$pe, a specific routine for an accumulator, et l 5 cetera. I 6 It has to be independent of the TRAC database, and 7 I want this so I can run it with a driver code, and that i
| |
| 8 allows me to do some testing, and I will show the results of 9 that. It has to follow that development guideline. Each 10 routine should do one thing and be less than a hundred lines 11 long.
| |
| 12 I want to use reusable functions. If I am going 13 to have a function, I want~ to take my bubble diameter and 14 put that as one function and then use that function, instead f's 15 (O
| |
| 1 of recoding it every time.
| |
| 16 For variable definitions, I am going to require 17 that in each functional unit, all of the IO has to be 18 identified explicitly -- what variables come in, what 19 variables go out. All local variables have to be defined 20 and, wherever practicable, a function should return one 21 value. So if you call a bubble diameter, it shifts back the 22 bubble diameter, it doesn't change half a dozen other 23 things.
| |
| 24 I said that the first requirement was that it not 25 chaage the answer. Null testing. So within the tolerance
| |
| /~N ANN RILEY & ASSOCIATES, LTD.
| |
| ( ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 335 11 that.you get by~doing~ numerical computations in a different 2-. order, the answer should not change for our standard test
| |
| .3 problem suite. .The' rest of the. requirements will be met' {
| |
| .4 through coding inspection, and that .is performing a line by 5 line codi'g' n inspection. You cor, are the source, the TRAC 6 source _ code to what was documented in the theory manual. 1 7 All undocumented features, and there are lots, will then be 8 .added to the theory manual. Differences between the code 9 and theory manual will be resolved wherever possible by 10 going back to the original reference.
| |
| 11 Debug testing. As I am recoding this and putting 12' in individual functions, I run that function through the ;
| |
| \
| |
| E l 13 debugger. I get an answer for every line of the code. I I l14- make sure I.go through every branch, every possible loop, j g
| |
| () 11 5 16-get an answer for every line of the code and compare it to
| |
| . hand calculation, and put that in a workbook.
| |
| L .
| |
| 17 Well, what is the modular structure that I am 18 using? Everything above this dashed line would be the TRAC t
| |
| 19 code, which, obviously, I am not laying out. I am just 20 showing the one routine, which is going to be FRICIF, which l 21 'now only sets the local fluid conditions. And what it will 1
| |
| 22 pass to the interfacial drag module is component type, void 1 23 ' fraction,. phased velocities and fluid properties. What it l 24L gets back is-the interfacial drag coefficient for that l'
| |
| 25' junction.
| |
| , ANN RILEY & ASSOCIATES, LTD.
| |
| i Court Reporters
| |
| '1025 Connecticut Avenue, NW, Suite 1014
| |
| : Washington, D.C. 20036 i (202) 842-0034 i
| |
| 5 8
| |
| 2 -,. . - . -.+..--...-,;,,
| |
| | |
| ...m___.. __ __ . . . _ . _ . . _ _ _ _ _ _ _ _ . . . . _ _ _ . _ _ . _ . . . _ . _ _ _ _ . _ . _ . _ . _
| |
| 336 1 This routine, INTFR, provides the interface --
| |
| f''o 2 well, my printer didn't do too well here. But the next
| |
| -Q 3 level'down are the components. There will be a separate 4 routine for each type of component. The level down is the 5 flow regime. TRAC uses four flow regimes, bubbly, ,
| |
| 6 . bubbly / slug, a transition between that and annual. The 7 transition one actually calls bubbly / slug, calls annual mist 8 and uses a weighting factor to combine the two. If you go 9 to bubbly, then it uses three things to calculate the bubbly 10 flow interface drag, a profile slip factor, a bubble 11 diameter, and the bubble drag coefficient.
| |
| 12 If you're.in the bubbly / slug, you need to know 13 what. fraction-of it's in slug flow, and then after that you 14 call a routine called SLUG, which then calculates these same
| |
| !(s_ )
| |
| \
| |
| 15 three parameters. It just calls the same routines. So I'm 16 reusing the low-level routines instead of recoding them, 17 which is what's done now. 1
| |
| .18 When I do it this way, and I make this division so l 19 that it's independent of the data base, I can replace TRAC L 20 here with the driver code, where I can vary parameters over l 1
| |
| '21 a range. And this was the demonstration I was going to 22 .give, but instead I'll just show you.
| |
| 23 This is a surface plot of the annular mist flow l
| |
| 24 regime. The void fraction varies from .75 to 1, vapor f-l 25 ~ velocity from 5 to 25, this is for a vertical pipe, a 2 bar, 3, s ANN RILEY & ASSOCIATES, LTD.
| |
| i . Court Reporters
| |
| : 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 j- (202) 842-0034 i-
| |
| | |
| 1 337 l 1- and the hydraulic diameter is 2-1/2 centimeters. And this !
| |
| i
| |
| (} 2 vertical access is interfacial friction coefficient. This 3 surface doesn't look very interesting here. And what it's 4 showing.you is there's the main effect is on void fraction.
| |
| 5 It has very little effect on vapor velocity.
| |
| 6 So to let you see it, I'm just reversing the axes.
| |
| -7 So I have void fraction here now in vapor velocity. So you 8 see the almost linear dependence with void fraction; almost !
| |
| 9 no dependence on vapor velocity. And out at 25 meters per I 10 second you've got most of the liquid entrained, or at least
| |
| .11 you certainly should, 12 DR. WALLIS: Does that make sense, though?
| |
| 13 MR. KELLY: I don't think so. For me it's an 14 ' unexpected result. And the reason is, TRAC is a two-fluid 1 15 code. When you go to the annular mist regime, you should i
| |
| 16 have two characteristic velocities for the liquid.
| |
| 17 Obviously the drops move a lot faster. What TRAC did in 18 order to overcome that is a superpose of force for the drops 19 and a force for the liquid film. So you're able to get a 20 drag coefficient for the drops in a film, but to get the 21 force for it, you need a relative velocity. They used a
| |
| : 22. terminal velocity for the drops, and a drift flux velocity l .23 for the film. So in effect it makes the drag force on the 24 drops near zero, and entrainment doesn't matter. And that's 25 wrong. So you never see the large increase in interfacial j h ANN RILEY ac ASSOCIATES, LTD.
| |
| ' (_/
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| ;. Wushington, D.C. 20036 (202) 842-0034 4
| |
| m ,, -
| |
| w
| |
| | |
| - ~ -. . . ~ - - - . . .
| |
| 338 l l
| |
| 1 shear as your entrainment fraction goes up.
| |
| j''} 2 DR. WALLIS: .But even with no entrainment at all, V
| |
| 3 this code makes no' sense. If you have five times the vapor 4 velocity, it's 25 times rho v squared, you've got to change. l l
| |
| 5 MR. KELLY: Okay. This is not the ferce. This is l 6 the coefficient.
| |
| 7 DR. WALLIS: Oh, this is the coefficient?
| |
| 8 MR. KELLY: Yes, it's the interfacial friction 9 coefficient.
| |
| 10 DR. WALLIS: Ahh.
| |
| 1:L MR. KELLY: So to get the force, you multiply it 12 by one-half rho v squared.
| |
| 13 DR. WALLIS: So where's the friction coefficient 14 'being plotted?
| |
| f%
| |
| (,,-) 15 MR. KELLY: It's the vertical axis here. j 16 DR. WALLIS: Okay. Okay.
| |
| 17 MR. KELLY: I'm sorry about that, but I'm glad 18 now, because do you know what this is? Do you recognize l '19 - that slope?
| |
| ! 20 DR. WALLIS: Yes, I've seen that before.
| |
| l I 21 ' MR . KELLY: 75 times the liquid fraction?
| |
| ; 22 DR. WALLIS: That's antediluvian.
| |
| ! 23 MR. KELLY:
| |
| (
| |
| The other thing you can do is not just L 24- do a surface part of the quantity, you can take a gradient
| |
| ! 25 of it. So this is the gradient of that plot. This is as s
| |
| 1 iQ
| |
| , k,,/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters
| |
| : 1025 Connecticut Avenue, NW, Suite 1014 '
| |
| . Washington, D.C. 20036 (202) 842-0034
| |
| ...) -
| |
| ,c .. , ., . -.
| |
| | |
| -1 339 1 1 you go to a void fraction of 1, and that's just'hecause
| |
| 'f7k
| |
| ~ '
| |
| 2 there were some' limiting factors then.
| |
| ~
| |
| 3 .You notice these strange bumps? They shouldn't be 4; there. They're discontinuities. And it turns out they used l'
| |
| 5 two different correlations for the terminal velocity of 6 drops, and they used an "if test." Those correlations are 7 not continuous across the "if test." So this is one of the 8 ~ advantages that you can get by going to this approach. You L 9. can fully test a model, and'you can.look to see for 10 unexpected behavior and discontinuities.
| |
| 11- And talk about unexpected behavior, that was a 12 two-bar. This is exactly the same thing at 70 bar up at BWR 11 3 conditions. Now we see the interfacial drag increases with 14 void fraction until you get to somewhere like .95, and then
| |
| () 15' it drops way down.
| |
| l 16- What on earth is going on? It turns out code L
| |
| 17 developers like to try to use what they call defensive L
| |
| l 18 programming, so they like to put minimum and maximum on i 19 things that they think are physically reasonable. So they
| |
| '20 put -- and this is an undocumented feature; you won't find ;
| |
| 21 it in the manual anywhere - *, hey put a maximum value on it. l 22 They said well, if all the liquid is drops and the drops are 23 Lvery small diameter, that should be the maximum drag you
| |
| ;- 24 should ever get. So let's use this as a maximum value. ;
| |
| 25 Well, they made one little error. They forgot the I
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| '\
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i "ashington, D.C. 20036
| |
| . (202) 842-0034 i
| |
| ~
| |
| | |
| 340 1~. vapor. density out in that. And there's a huge difference
| |
| /~'N 2 between the vapor density at 2 bar and at 70 bar. So almost d
| |
| >3_ all of this curve is in error. It's a limiting value that 4 was. calculated. incorrectly. You can catch that this way.
| |
| 'S It's hard to catch that by doing even separate effects.
| |
| 6 assessment, because you'only see how it did versus a few 7 data points.
| |
| 8 DR. SCHROCK: That was in TRAC --
| |
| 9 MR. KELLY: This is in TRAC-M today.
| |
| : 10. DR. SCHROCK: TRAC-M tcday.
| |
| 11 DR. WALLIS: Well, TRAC is probably riddled with
| |
| .12 this sort of thing.
| |
| L .13 MR. KELLY: When I go through and modularize it, I 14 can start finding them. ' Arm it gives you two advantages.
| |
| q j 15 .You can find the problems, and when you go to fix them, 16 they're localized and you can fix it in one place.
| |
| 17 DR. SCHROCK: On your two-bar graph, you've got 18 negative friction coefficients.
| |
| H19 MR. KELLY: That better be the gradient.
| |
| 20 DR. SCHROCK: Okay.
| |
| 21' 'MR. KELLY: And so in summary at 6:25 in the 22- evening,.the modularization of the 10 interface drag package 23 -is in progress, and I believe I've J;monstrated the value of
| |
| '24 the approach.
| |
| .25- Code documentation. I want to make some comments
| |
| /' ANN RILEY & ASSOCIATES, LTD.
| |
| 5
| |
| . (_]f Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 1
| |
| (202) 842-0034
| |
| | |
| . . . _ _ . .._ _._. _ _ _ . . . _ _ . _ _ _ _ _ . - _ _ _ . . _ . . _ . . _ _ . _ _ . ~ _ . _ _ . _ _
| |
| - 341 J 1 here. The models as coded often do not reflect what's in i 2 the manual,.and I use the word "often." I believe this is a 3- problem if --
| |
| 4: DR. WALLIS: -We review something, we see an 5' ' equation, and we sort of have to believe that's what's 6- actually in-the code.
| |
| 7 MR. KELLY: Um-hum. RIf I told you that the Wallis ,
| |
| 8 ' annular flow model was used,,you might say.oh, it's an old 9 model, and maybe it's not' state of the art, but you 10 shouldn't get this. Okay? But if you look in the TRAC 11 ' documentation, you use the Wallis model. So I believe it's 12 a problem, if'your only source for a code review is looking
| |
| : 13. :at the documentation. You would have to do a full 14 verification effort, and very few people ever want to do
| |
| () 15 16 that under strict QA standards, uecause it's very expensive.
| |
| DR. ZUBER: Let me just make a comment on this.
| |
| 17 One of the recent -- about 10 years ago NRC started 18 international program, and we wanted to have all these l
| |
| 19 people look and use the code. This was never really ever 20 addressed. So actually people have to be interested in
| |
| : 21 finding something, not just running the code and say look, l
| |
| E22 'everything is fine. And this is a good example, what you 23 did.
| |
| 24 DR. WALLIS: You've confirmed a lot of our 23' suspicions, I think, and you remind me of my children, that
| |
| ! ANN RILEY & ASSOCIATES, LTD.
| |
| I ,
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
| |
| : ~ Washington, D.C. 20036
| |
| , '. (202) 342-0034 i
| |
| ,e- ,. . v, - - , , , - + w-, . - , , , , - - yw,.-
| |
| | |
| 342 1 when they were teenagers, I used to think that what I
| |
| ('')
| |
| V 2 suspect they're doing, they probably are doing, but I'm not 3 sure that it's the right tinie to bring it up. So when they 4 become 35 or something, you hear them say, "Do you know what 5 we did when we were teenagers?" And I say, "Yes, everything 6 I thought you were doing, you were doing." That's what's 7 happening here is that at last someone's coming clean about 8 what's really happening in the code.
| |
| 9 MR. KELLY: Well, I'm trying.
| |
| 10 DR. WALLIS: Or someone's finding out.
| |
| 11 MR. KELLY: And what I find is that there are 12 undocumented features, there are a lot of limiting values, 13 like the one that there was an error in. You also will find 14 either errors or deliberate modifications. You will see (n) 15 coefficients changed, or at least they're different than 16 what's in the manual. And you don't know if it was an error 17 or whether it was changed to make some data comparison 18 better. And that's one you can't trace. But what I'm doing 19 is documenting all those. And then when we get to improving 20 the code, we're going to have to address every single one of l
| |
| l 21 those.
| |
| 22 DR. WALLIS: When we get these vendor codes that i
| |
| f 23 the NRC is so happily accepting, should we assume that f 24 .. hey're free of all these kinds of manipulations?
| |
| 25 MR. TAKEUCHI: We do line by line.
| |
| /'s ANN RILEY & ASSOCIATES, LTD.
| |
| (_) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . . . - . . - _ . . .-- ~. . _ . - . . . . . . .- - . - - - . - - . - . - . . . -
| |
| 343.
| |
| 1 DR..WALLIS: You do,'but how about the NRC? Does
| |
| / Y 2- -the NRC do line.by line review?
| |
| Q ;
| |
| 3 MR. TAKEUCHI: We.do line by line review and write i
| |
| 4 it down, .the equations, and document it. That was the first 5 requirement the NRC imposed on us.
| |
| 6- MR. KELLY: And that is strict. requirement.
| |
| 17 MR. TAKEUCHI: That is very strict and we worked 8 so hard for it and that one is reviewed by INEL.
| |
| 9- MR. NISSLEY: Let me add a word to that. This is 10 Mitch Nissley from Westinghouse.
| |
| 11 If you look in Volume 1 of our code qualification 12 document where we document the models, we do describe these l' 13 limits, mins and max's so I think you will find that that'is 14 in there.
| |
| 15 I will not swear that every one of those max's and 16 mins is free of nonphysical behavior.
| |
| 17 MR. KELLY: So the last item, when I look at the 18 annular or MIST regime and that is all that I have looked at
| |
| -19 in detail so far, I just coded up the bubbly slug.
| |
| 20 I found unexpected behavior, the lack of -- didn't l
| |
| 21 see much of an effect due to entrainment. It contains l 22 discontinuities and errors and that is just an annual MIST 1
| |
| : 23. for vertical pipe, l
| |
| 24 DR. FONTANA: This raises a question. Why don't f(
| |
| i 25 you just start over, or is that essentially what you are ANN RILEY & ASSOCIATES, LTD.
| |
| 3 ['s,/~')
| |
| Court Reporters j- ' 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 4 (202) 842-0034 Av +- ,e- -,----ne - , - - s 4. w - --,m- s - - - -
| |
| | |
| 344 l l
| |
| -1 'doing anyway?
| |
| [']
| |
| s_/
| |
| 2- MR. KELLY: Well, remember my first metric was 3 nullLtesting:so for ID models at least I am going to recode
| |
| )
| |
| .4 the'models to give the same answer but document where errors ,
| |
| 1 5- are,.then file - .as part of our procedures I have to file l p '6 an error report on each of those and then go through and 7 =give the correction and then that way it.is traceable so
| |
| ~8 that --
| |
| 9 DR. FONTANA: But if you. start i.t from scratch you 10' wouldn't have to do two or three extra steps.
| |
| 11 DR. KRESS: I guess the answer to that part of it
| |
| {
| |
| 12- is there's a' lot of good things in that code that you want 13 to preserve.
| |
| l 14 DR. FONTANA: Must be.
| |
| 'Oj j 15 IMt. KRESS: And starting from scratch means you ;
| |
| l- 16_ just sort of do.what he is doing anyway but you'd want to ',
| |
| i l
| |
| 17 dig out all those good things. l l
| |
| i 18' MR. KELLY: .Yes. The bad thing is if I started i 19 from scratch.I might-have used what was in the manuals and l
| |
| h i-20 not what is in the code, and what you will find, it's pretty L -21 simple to write a simple 1D, two fluid code and do a pretty i
| |
| l
| |
| ,22- decent job but when you start hooking up pipes to plena, 23 pipes to 3D vessel and so on and so forth, and going through
| |
| ;24 all of the different things that happen like when, say, L 25. nitrogen.comes in from the accumulator, then there are a lot i
| |
| k
| |
| : ANN RILEY & ASSOCIATES, LTD.
| |
| I'k'],/
| |
| Court Reporters
| |
| ; 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
| |
| +
| |
| s'
| |
| | |
| r
| |
| ('.
| |
| i
| |
| ! 345 t
| |
| 1~ ~o f'special. tricks that have been employed in the code.
| |
| '2 Especially.in the numerics those are good. They help the
| |
| . Q[/~'I
| |
| ! 3 code get over things, but you won't find them documented in 14 any great detail.and so from scratch in my opinion -- well, i
| |
| L 5- I proposed doing it from scratch but there are pros and cons 6 either way.
| |
| 7 DR. WALLIS: Shall I say thank you very much now?
| |
| 8 MR. KELLY: Please.
| |
| l ..
| |
| '9- (Laughter.]
| |
| l 10 DR. WALLIS: Well, I've let you go on far beyond L 11 the original time because there were all kinds of good i
| |
| 12 things'here.
| |
| 13 I do think what we originally intended was 14 something not quite so detailed but we have learned a lot n
| |
| ( 15 and thank you very much.
| |
| 16 Other comments from the rest of the committee 17 would like rm make about the research program?
| |
| 18 (No response.]
| |
| i
| |
| ! 19 DR. WALLIS: Did you hear the kind of things you 20 want to know?
| |
| l 21 DR. ZUBER: I was really, after the last meeting I l :22. was really very perturbed. I thought RES didn't have their l
| |
| l' t
| |
| 23 ~ arms around the problem. I feel much more comfortable now.
| |
| p 24 I think you have made good progress. I think what you are 25 doing here~is very good.
| |
| .[~ ANN RILEY.& ASSOCIATES, LTD.
| |
| i (_,. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| .sa
| |
| !=
| |
| l 346 L ,
| |
| l 1 .The only thing I feel a little bit saddened and
| |
| / 2 disappointed is.that you will not be able to address the i
| |
| 3 modular approach that will give flexibility and speed to 4 NRC.
| |
| l :5. I thought that you would also attacx the problem l
| |
| 6 of looking -- to neglect some phenomena which are not 7 important, for example for this long transient there is no 8 need for acoustic calculations -- and this is where the code 9: grinds,.and I think had you done this, you would have really j 10 an excellent package.
| |
| 11 You have made great progress but I don't think you 12 are really addressing the final needs and I think this will 13 be probably the final code development for this agency so I i 14 would really, as an ex-member I would really urge you to
| |
| () 15 16 think about how to bring this modular capebility in this new code because if you don't have it now, they will never have 17 it, but it is good work, I think. l 18 The only thing I am surprised at, that you are 19 .really doing the work on the reflood. I never heard the l
| |
| l 20 rationale. We have so much data on reflood, why do we need 21 one more experiment -- which is quite expensive -- but maybe I 22 you will tell us another time.
| |
| l 23 DR. WALLIS: Novak, we don't any time for the 24 . answer.
| |
| I 25 DR. ZUBER: Good -- I have more discussions -- but lL T
| |
| _[ ANN RILEY & ASSOCIATES, LTD.
| |
| j \_- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
| |
| --+en -+ - , - - , ,,-
| |
| | |
| . . -- - . . . . . = - . . - - . . . . . . . - .
| |
| L 347 1 it was a very 17od presentation.
| |
| 'f['}
| |
| V.
| |
| 2 DR '. WALLIS: Would you put some of that in 3 writing, please?
| |
| 4 DR. ZUBER: Definitely -- because if I am 5 critical, I am critical, and if I have something nice to 6 say, I say it.
| |
| 7 I would urge you really look how to make it 8 modular. This is the last chance this agency has to have a
| |
| : 9. good code and it would be really sad not to have it, and you 10- have the capability of doing it.
| |
| 11 DR. WALLIS: Perhaps some other --
| |
| 12 DR. SCHROCK: I think that it was a marvelous 13- presentation. There's some very encouraging things in all 14 of it, b 15 The one thing that I still wonder about fe the G.
| |
| 16 adequacy of the knowledge of the physics that go into some 17 of the models that are in it, so at this stage you are 18 trying to recover everything that is recoverable from TRAC
| |
| , 19 and even RELAP. There's a lot of stuff in there I think l
| |
| .20 from our last in-depth review of TRAC that doesn't deserve L 21 to be retained, and I haven't heard much about what those 22 things are and what.is going to replace them and it seems to l 23 me that there are some areas where it is inevitable if you 24 are ever going to get a good model you are going to have to
| |
| ; 25 do some more experiments.
| |
| I I,1/''s ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
| |
| | |
| 348 1
| |
| I think I heard that would come in the second 2 phase of your program, after you have finished the
| |
| ['')i
| |
| 'R-3 consolidation, is that right?
| |
| 4 MR. ELTAWILA: That is correct.
| |
| 5 DR. SCHROCK: But shouldn't you be identifying 6 what those things are as you go along in this consolidation?
| |
| 7 MR. ELTAWILA: I think hopefully part of his 8 activity to identify all these models. I am hoping too by 9 next year I would be stopping spending money on TRAC-B 10 maintenance for example, so I can take that $600,000 and run 11 experiments with them, so that is how the program is built, 12 you know, so when we free sort money from the maintenance we 13 can use it for running experiments.
| |
| 14 DR. ZUBER: See, you really have a very good C 15 justification to get more --
| |
| (
| |
| 16 MR. ELTAWILA: I agree with you and I hope you put 17 that in your recommendation.
| |
| 18 DR. ZUBER: I hope you do the modulization also.
| |
| l 19 MR. ELTAWILA: No, I think we are listening to you 20 and I agree with you, but I think there are a lot of l 21 things -- you said it right. I think this is the last I
| |
| ! 22 chance that NRC is going to develop the code and we are 23 going to do it right, you know. We are not going to for the 24 sake of budget cuts or something like that to skim here and 25 there, so we are going to do it right and hopefully with l
| |
| (~)
| |
| ( ,!
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 349 1
| |
| your support and my management support -- I am not denying
| |
| ! [~
| |
| x_)\
| |
| 2 that at all -- we will finish that effort.
| |
| 3 DR. KRESS: I want to compliment the staff, I 4 think this is one of the well laid out programs I have seen 5 in a long time. They have extremely good objectives, a good 6 plan, a good approach, and it looks like very competent 7 people working on it. I think -- I hope you stick to your 8 guns on the part where you said we are doing some of this 9 not because of user need letters, because we think it is the 10 right thing to do. I agree with you completely.
| |
| 11 The problem I have is I presume we are preparing 12 information for use in our research letter. Now, we got me 13 and you and Mario here that is going to have a fight when 14 people talk about priorities to the full committee. And I l (f- s) 15 am wondering, how do we translate what I see as an extremely 16 good program that is well justified, should have a high 17 priority, should have more money than it has got, how do we 18 translate that to the full committee?
| |
| 19 DR. ZUBER: May I suggest, you heard this time and 20 last time how much money this is worth to the industry.
| |
| j 21 DR. KRESS: Yeah, that is a good point. That is a 22 good point.
| |
| 23 DR. ZUBER: And the crux of the problem, the crux 24 is if we have a good code, with good physics, you can 25 justify and the money, the benefits, the industry has it.
| |
| [~ \ ANN RILEY & ASSOCIATES, LTD.
| |
| i(/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| r 350 1 Because we have a methodology, and for the best estimate, 2 with a-good code, the industry can use it.
| |
| } )' And I think this l l 3 is what'-- !
| |
| 4 DR. KRESS: That is a good selling point.
| |
| 5 DR. ZUBER: So I think excellent. And you heard 6 this morning, tens of nillions of dollars. Last -- a month I 7 ago, they also said millions and millions of don ars, so I 8 think the justification is right.there. And they cannot do 9 a probabilistic analysis'without a g7od code. How much --
| |
| 10 what is the percentage of -- 95 percent assurance, or 60 11 percent? I think this is the --
| |
| 12 DR. WALLIS: I think one problem is that, for 13 licensing purposes, it has often appeared adequate to use a 14 code which wasn't so good.
| |
| () 15- DR. ZUBER: But then you are not going for the 16 best estimate and you don't want to then~ claim to make ten l
| |
| 17 millions of dollars with the best estimate.
| |
| 18 DR. WALLIS: As long as people are accepting the 19 code as it was, where --
| |
| 20 DR. SCHROCK: It was more comfortable for the 21 staff, the codes, too. It was easier, too.
| |
| 22 DR. KRESS: To go conservative.
| |
| l 23 DR. SCHROCK: Right.
| |
| l l
| |
| l 24- DR. WALLIS: Maybe ACRS should insist on higher l 25 standards for codes.
| |
| L i
| |
| i .l
| |
| ' k/O
| |
| ,/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| ~ , , , .-
| |
| | |
| 351 1 DR. KRESS: Well, I think we should,.and I think
| |
| /~'s 2 this looks.like a higher standard, b
| |
| -3 DR. ZUBER: . Higher standards and really worthwhile
| |
| : 4. as far as the economy is conccrned. You can justify the 5 changes.
| |
| 6 DR. KRESS: It's good for NRC and good for the 7 country.
| |
| 8 DR. WALLIS: Everybody benefits.
| |
| 9 DR. KRESS: I think somebody must not. I think 10 who doesn't benefit are the other parts of NRC you have to 11 take money away from to put into this.
| |
| - 12' DR. WALLIS: Mario?
| |
| 13 DR. FONTANA: No , I have no additional comments.
| |
| 14' DR. WALLIS: Well, I chink it has been said, we O
| |
| (j 15 are impressed with what you are doing. We'have to write a 16 report where this is weighed in comparison with a lot of 17' other-things going on, and we have got some information 18 which will help us there.
| |
| 19 DR. KRESS: You have given us some good 20 ammunition. We appreciate it.
| |
| 21 MR. ELTAWILA: Thank you very much.
| |
| 22 DR. WALLIS: Maybe le should go back to the 23 Westinghouse briefly, because we haven't decided what we do 24 now. You made a presentation today and NRR is intending to 2'5 . -- would like to approve your code in February and I am not
| |
| []
| |
| (.m/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters !
| |
| 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l (202) 842-0034 I l
| |
| __ l
| |
| | |
| . - , . . - . ..~ - -.- - . . .- - ..
| |
| i' 352 sure how we'come into that. From the reaction you heard
| |
| ( ) 2 today, I think it is fair to say that, a better job could be 3- done in selling your code. It is probably in the 4 . presentation area rather than, I hope, in the documentation
| |
| .5. area.
| |
| 6 Do we need another presentation? If we were to 7 advise NRR today, based on what we heard today, would we say 8 this is good enough?-
| |
| l 9- DR. ZUBER: Well, I have to agree, I wouldn't put
| |
| ~10 my name approving this, what I heard today. Let me say, 11 they'are_ setting a precedent. A best estimate calculation 12 is worth millions of dollars. It is good for the economy, 13 it is good for the company. And this is the cutting edge
| |
| : 14. for this' agency, and what they do, they come up with a f .15 ' nonsensical model, nonsensical calculations.
| |
| 16- Really, the critical point in this whole exercise 17 'today was the condensation and the CCFL on the tie plate, t
| |
| L 18 .That was the feature of this plant. And the problems which 19 we heard today, oscillations, and the hush and the -- this 20 all stems from that non-physical model. And then they add
| |
| ( 21 to that model some coefficient and rate it from 1 to 2 and 22 say this is good enough. That kind of approach will ruin 23- the best estimate licensing in this country once somebody 24 else hears it. And I think this is a disservice to the 25 technology of the future. It may be expedient to NRR today, l
| |
| ' /'')
| |
| \s /
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 4
| |
| 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (2021'842-0034 i
| |
| | |
| i 353 I I
| |
| 1 because they will meet the February date, but it will harm
| |
| (''; 2 the technology in the long run.
| |
| L) 3 If you want to do it to a best estimate approach, 4 which this country needs, and you have the tools, then do a 5 good job.
| |
| 6 DR. WALLIS: Can this be fixed up? l 7 MR. NISSLEY: Would it be appropriate for me to 8 make some general comments? If not, that's fine.
| |
| 9 DR. WALLIS: If they help guide us in what we 10 should be doing.
| |
| 11 MR. NISSLEY: We acknowledge that our 12 presentations were not as focused and did not address some 13 of the concerns that you had this n orning.
| |
| I believe some 14 key points that I would like to repeat, that we did make
| |
| , A (d' 15 earlier, were that the -- while you may take issue with the 16 method used to calculate flooding, what the calculations are 17 showing is that you have flooding at all of the interior 18 assemblies and the only place you get downflow is in the 19 periphery of the core.
| |
| 20 One result of that is that you get bottom-up 21 reflood and that the nume -i cal hash that results later in l
| |
| 22 the transient is after the transient is turned round. With l 23 that as a very cursory overview comment, I think it would be 24 possible to do better.
| |
| 25 Several observations. One is if you revise the
| |
| [3
| |
| (/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 1 354 1 model, most of these issues, I don't think would -- most of f~'')
| |
| v 2 those global statements would not be affected. I think you 3 would still get draining down the low power channels only.
| |
| 4 The hot assembly would be turned over by a bottom-up reflood 5 and if you eliminated the hash, it would be able the peak 6 cladding temperature was calculated to occur, so you would 7 be more comfortable with the numerical results, but would 8 not materially affect the results and how it affects safety.
| |
| 9 One more comment is that if we were to make those 10 changes and come in with a better model, as distinct 11 example, to directly implement a CCFL correlation in the 12 model and do that by the end of February, I think we would 13 be in a position where we have a utility that already has a 14 fully completed analysis. We have a methodology that (av) 15 already has a fully completed TER, that we would be back at 16 the beginning in terms of redoing code assessment work, 17 revisiting many of those RAIs and completely redoing that 18 plant-specific analysis which would, I am sure, cost that 19 utility at least one full cycle for each of their two 20 plants, resulting in millions of dollars of impact to them.
| |
| l
| |
| : 21 Now, that does not -- that should by no means 22 totally reconcile that there were shortcomings, not only in i 23 the modeling or in the presentation, but I think that is a i 24 bit of perspective that I felt an obligation to pass on.
| |
| 25 DR. WALLIS: I ask myself do you mean by best
| |
| ' [~')
| |
| (_ '
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 355 1 estimate? If best estimate is best estimate at peak clad (n) w.
| |
| 2 temperature, that is one thing. If it is best estimate of 3 what is happening, without compensating errors and things, l 4 then I think that is a much more stringent standard. If you 5 are happy to live with some of the inadequacies of modeling 6 because they don't matter as far as peak clad temperature 7 goes, is that what we mean by best estimate?
| |
| f 8 MR. NISSLEY: I believe in the Regulatory Guide, i
| |
| 9 Reg Guide 1.157, which was endorsed by the ACRS back in l
| |
| 1 10 1989, that there is a statement in there that there will be 11 -- it's acknowledged that some models will not be completely 12 physical, and may be not completely robust or may be i
| |
| 13 simplified, and if they do not invalidate the code for use 1 14 as a realistic analysis, that that would be considered
| |
| (~h
| |
| ( ,) 15 acceptable. Obviously I'm paraphrasing this.
| |
| 9 16 MR. ZUBER: I wrote that report, and this was for 17 the important phenomena and the range, but we had models and 18 we had experimental data. What you have here, the code, the 19 way you use it, you cannot calculate the important phenomena 20 and then you range it over a range which really applied to 21 something which is not applicable.
| |
| 22 Now, you want -- what you are asking -- now we i
| |
| L 23 spent the money, now approve it because if you don't approve l 24 it, it will cost more. This is a short-range and I
| |
| 25 shortsighted decision. You are harming the entire approach,
| |
| [)
| |
| \/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 j
| |
| | |
| [
| |
| 356 1- you1are compromising it, because something like this sooner 2 or later.will come up and it will bite you. It will bite
| |
| : 13. this agency. This agency,:if'they want to go to the.best 1 4 estimate, then that'should-be consistent. You have-a 5- -non-physical situation which you arrange and then say the 6 -calculations.are'okay, they will not-change. I hope this is i
| |
| 7 -- the' water will always come down, but.you didn't prove the
| |
| .- 8 . case.
| |
| 9 I think you.are going to harm it in the long run.
| |
| 10 You'are discrediting the entire approach. If-you know that, 11- 11t's much better to stay with the conservative approach. At 12 'least you know whe're you~are sleeping. I believe.that this l 13 'could be done in.a month or two months. l 14 As I'said, I put this in writing. I would not put
| |
| '15 :my name approving this --
| |
| 16 DR. WALLIS: We also have no independent 117 assessment. We have to go totally on the Westinghouse 18 presentation. There's no NRC calculation or anything, 19- there's nobody-else's calculation, we just have to believe 20 or not believe.
| |
| ; 21; Well, do we want to have any more co do with this 22 and just let NRR go ahead? Have'another meeting?
| |
| 23 MR. ZUBER: See, what I would really like is to E24 see'the industry does a good job, and we cannot prove it I
| |
| 25- because this methodology is needed and it's work to discount i.
| |
| i.
| |
| O' ANN RILEY & ASSOCIATES, LTD.
| |
| (_ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
| |
| a
| |
| | |
| 1 357 1 as far as dollars are concerned.
| |
| x i s,r~ ).
| |
| QJ 2 DR. KRESS: I think it would be wise to have \
| |
| 3 another meeting. l 4 DR. WALLIS: It would not?
| |
| 5 DR. KRESS: It would be.
| |
| 6 DR WALLIS: It would be.
| |
| 7 DR. KRESS: Yes. Otherwise, I think we're left i.
| |
| 8 with --
| |
| 9 DR. WALLIS: I think we are. I 10 DR. KRESS: Yes.
| |
| 11 MR. SINGH: When you say one more meeting, are you 12 talking about full committee meeting or subcommittee 13 meeting?
| |
| 14 DR. KRESS: It could be either. I suspect full --
| |
| ,, x .
| |
| (j 15 subcommittee meeting would be more efficient. If you 16 satisfy the subcommittee, pretty much we can convince tha 17 full committee to go along with it, I think.
| |
| 18 MR. SINGH: If you're going to have a subcommittee 19 meeting, how much time are you asking for? Half a day?
| |
| 20 Then we have to bring everybody in or we have to piggyback 21 somebody. If we have a full committee meeting, it will be 22 in two hours.
| |
| , 23 DR. KRESS: Yes. I think a full subcommittee 24 meeting day.
| |
| 25 DR. WALLIS: Well, we have a meeting in January
| |
| ' f) ANN RILEY & ASSOCIATES, LTD.
| |
| \~./ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 358 1 already.
| |
| I~'\ .2 MR. SINGH: Yes.
| |
| \._/
| |
| -3 DR. WALLIS: And we could stay for another day and 4 we could say,. Westinghouse, come back and do it right. '
| |
| 5 DR. SCHROCK: If they're prepared. If they're 6 prepared. And the idea of the subcommittee meeting would be 7 to' address these questions we have and convince us that it's 1 l
| |
| 8 still okay.
| |
| 9 DR. FONTANA: I was going to remind you that that !
| |
| I 10 two-day meeting is no longer a two-day meeting. One of l 11 those days is not going to exist. So you've got a day you I 12 could put it in. The 22nd.
| |
| 13 MR. SINGH: Okay.
| |
| 14 DR. SCHROCK: The 22nd was scheduled, but now it's 1
| |
| (~N 15
| |
| () no longer ---
| |
| 16 MR. SINGH: Okay. Before we do that, we've got to I 17 ask Westinghouse if they're willing to do that, okay? If 1
| |
| .8 you want to come back to me next week, that's fine.
| |
| 19 DR. SCHROCK: Yes, you can think about it.
| |
| 20 MR. SINGH: Think about it with your management.
| |
| l - 21 MR. NISSLEY: I do have a -- I feel a legitimate
| |
| : 22. concern of how much we can get done in that time.
| |
| 23 DR. WALLIS: Would it be different if you came 24 back in a month?
| |
| 25 MR. NISSLEY: I think realistically, we could f)
| |
| ' Am ,/
| |
| ANN RILEY & ASSOCIATES, LTD.
| |
| Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
| |
| , (202) 842-0034 4
| |
| | |
| ~- . . . . ..,_ - - _ . . - . . . - - . _ - - - - - - . . - - . - . - . -
| |
| 359 1 address a few of the issues. We could certainly do a better '
| |
| () 2 3
| |
| packaging and organization, presentations.
| |
| concerned about you having high expectations and us not But I would be 4 being able to deliver in one month.
| |
| 5 How about.if we get back in a week?
| |
| 6 DR. WALLIS: Yes. .I think we need to think about '
| |
| 7 it.
| |
| 8 DR. SCHROCK: Think about it.
| |
| 9 MR. ZUBER: Just a question. How long would it i
| |
| 10 take you to put this as a boundary condition and run the 11 calculation? -
| |
| 12 MR. NISSLEY: Put what as a boundary condition?
| |
| 13 MR. ZUBER: 'Having to put a correlation, a 6 CFR 14 correlation in POSIT and see how it affects the 15- calculations.
| |
| 16 MR. NISSLEY: We can't answer today.
| |
| 17 MR. ZUBER: As far as I'm concerned, it's not the 18 package; it's the way you approach the problem.
| |
| 19 DR. WALLIS: I think the concern I have is that
| |
| :20 here is NRR saying everything is fine, there's no problem, 21 piece of cake, and then there's this committee listening to '
| |
| '22 a presentation and saying, wait a minute, we're not really 23 convinced that everything is fine. That's what bothers me, 24 is the complete difference here and the impression.
| |
| 25 I. wondered if you could resolve that by doing a ANN RILEY & ASSOCIATES, LTD.
| |
| A Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| 360 1 cracker-jack job next time so that we think everything is 2 fine?
| |
| 3 DR. SCHROCK: That was my reasoning.
| |
| 4 MR. ZUBER: This is your future. This approach is 5 really the approach where we are going. When you start with 6 Ethe correct step, you have everybody with you.
| |
| -7 MR. NISSLEY: I understand your comments. I do 8 need to go back and talk with my management. We will have 9 to make some major resource reallocations.
| |
| 10 DR. WALLIS: Okay. So have a Merry Christmas.
| |
| 11 MR. NISSLEY: I don't think so.
| |
| 12 [ Laughter.)
| |
| 13 MR. ZUBER: Happy New Year, then.
| |
| 14 DR. WALLIS: So we will negotiate with you next 15 week or two.
| |
| 16 MR. SINGH: Before we quit, are we going to write 17 any letter next month?
| |
| 18 DR. WALLIS: I don't think we know yet. We don't 19 know yet.
| |
| 20 MR. SINGH: Okay.
| |
| 21- DR. WALLIS: Okay.
| |
| 22 MR. SINGH: Thank you.
| |
| 23 DR. WALLIS: We will recess until 8:30 tomorrow 24 morning.
| |
| 25 [Whereupon, at 6:53 p.m., the meeting was ANN RILEY & ASSOCIATES, LTD.
| |
| O' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
| |
| | |
| . . . . _ _ . _ . . . - - = _ . . . _ . . . _ . - _ . . .m. _ _ . . . . . _ _ . . - . . _ _ _ _ _ . , _ _ . - _ -._. _ .~..__.. .__._.__._ __
| |
| 361 l' recessed, to. reconvene at 8:30 a.m., Thursday, December 17, l
| |
| L
| |
| <p-
| |
| ; -. g 2 1998.]
| |
| 1, 3 .
| |
| i l l' 4 l 5
| |
| i I l 6 i 7 ,
| |
| l 8 l l l
| |
| ! 9 l 10 11 .
| |
| l i
| |
| 12 1 13' l-14 I s
| |
| 15-1 16 l 17: I 18.
| |
| 19 l l
| |
| 20 )
| |
| l 21 '
| |
| l- 22 l- 23 L
| |
| i '24 25
| |
| , ANN RILEY & ASSOCIATES, LTD.
| |
| ;. Court Reporters
| |
| ! 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 j
| |
| * \
| |
| | |
| REPORTER'S CERTIFICATE i This is to certify that the attached proceedings before-the United States' Nuclear Regulatory Commission in the matter of:
| |
| NAME OF PROCEEDING: MEETING ON THERMAL-HYDRAULIC PHENOMENA-- '
| |
| 4 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.
| |
| I
| |
| : m. \ 0 p n Hundley Official Reporter Ann Riley & Associates, Ltd.
| |
| I i-
| |
| [O
| |
| | |
| 4 INTRODUCTORY STATEMENT BY THE CHAIRMAN OF THE THERMAL-HYDRAULIC PHENOMENA SUBCOMMITTEE 11545 ROCKVILLE PIKE, ROOM T-2B1 ROCKVILLE, MARYLAND DECEMBER 16-17,1998 The meeting will now come to order. This is a meeting of the ACRS Subcommittee on Thermal Hydraulic Phenomena.
| |
| I am Graham Wallis, Chairman of the Subcommittee.
| |
| The ACRS Members in attendance are Mario Fontana and Thomas Kress. The ACRS Consultants in attendance are Virgil Schrock and Novak Zuber.
| |
| The purpose of this meeting is for the Subcommittee to discuss the applicatior of the Westinghouse Electric Company's WCOBRA/ TRAC best-estimate large-break LOCA code to nuclear power plants with upper head plenum injection; the NRC Thermal-Hydraulic Code Review Action Plan; and the status of the NRC thermal-hydraulic research program.
| |
| The Subcommittee will gather information, analyze relevant issues and facts, and formulate O proposed positions and actions as appropriate, for deliberation by the full Committee.
| |
| Amarjit Singh is the Cognizant ACRS Staff Engineer for this meeting.
| |
| The rules for participation in today's meeting have been announced as part of the notice of this meeting previously published in the Federal Register on November 30,199C.
| |
| A transcript of the meeting is being kept. It is requested that the speakers first identify themselves and speak with sufficient clarity and volume so that they can be readily heard.
| |
| We have received no written comments or requests for time to make oral statements from members of the public.
| |
| (Chairman's Comments-if any)
| |
| We will proceed with the meeting and I call upon Mitch Nissley of Westinghouse Electric Company to begin.
| |
| [ DOCUMENT NAME: G2oEHNERTVNTR1216.PB) 1
| |
| | |
| O O O f s, i j United States Nuclear Regulatory Commission
| |
| . _.~ ,e _ _a _ ,n _ _..:. .
| |
| l i
| |
| f Code Consolidation Example:
| |
| BWR Jet Pump Model i t
| |
| Presented to ACRS Thermal-Hydraulic Subcommittee J.M. Kelly' B. Aktas2 1 - NRC Office of Regulatory Research 2 - Scientech Thermal-Hydraulics Group December 16,1998 1 i
| |
| | |
| Code Consolidation Example:
| |
| BWR Jet Pump Model 1 r.a+ w mmxxsemwntm:2raxe: wwm na sc. . < . . ~.: . -
| |
| a CONTENTS:
| |
| +
| |
| | |
| ==Background:==
| |
| Code Development Practices
| |
| + Overview of Component Consolidation Process
| |
| + introduction: TRAC-B Jet Pump Model l
| |
| + Jet Pump Model Functional Requirements t
| |
| + Jet Pump Verification Testing Plan & Results
| |
| + Summary 1
| |
| | |
| O O O ~l
| |
| | |
| ==Background:==
| |
| | |
| Code Development Practices m Past NRC Code Development Practice: !
| |
| t
| |
| + For each code, one contractor was responsible for all development, maintenance, and assessment activities. i
| |
| + The NRC project manager was not expected to either be cognizant of the code's technical details (models etc.) or to perform analyses with the code.
| |
| + All responsibility for code configuration control was left to t the contractor => no rigorous testing between versions.
| |
| + Software Quality Assurance was left to contractor with no NRC oversight => undocumented code modifications.
| |
| + Code assessment libraries (input models & data) were not consistently maintained nor transferred to NRC.
| |
| I December 16,1998 3
| |
| | |
| ==Background:==
| |
| | |
| Code Development Practices m NRC Code Consolidation Project:
| |
| + Code development activities are dispersed between five organizations at seven locations (including NRC staff):
| |
| + take advantage of uniquely qualified individuals
| |
| + NRC project managers are expected to become knowledgeable code users; technical monitors to follow details of developreent.
| |
| + Code configuration control is performed at the NRC and testing is repeated between every developmental version.
| |
| + Software Quality Control procedures are being formalized and all documentation is being maintained on-line under the NRC configuration control system.
| |
| + PIRT based assessment matrices are being developed and will be incorporated in auto-validation system => easy to repeat.
| |
| D mber 16,1998 4
| |
| | |
| O O O NRC Code Consolidation Effort m Objective:
| |
| + Recover current capabilities of four largely redundant system T/H codes in one code.
| |
| e TRAC-P: PWR Large Break LOCA e TRAC-B: BWR LOCA & Transients e RELAP5: PWR & ALWR SBLOCA and Transients e RAMONA: Coupled T/H & Kinetics (BWR Stability) m Approach:
| |
| + Identify Simulation Requirements (plant & transient types etc.). '
| |
| + Use modernized (Fortran 90) TRAC-P code as base. .
| |
| + Implement Required Modeling Capabilities.
| |
| Component models (jet pump), new capabilities (3D kinetics),
| |
| different numerical solution, and improved physical models.
| |
| + Comprehensive testing to ensure simulation requirements are met with sufficient accuracy and computational efficiency.
| |
| December 16,1998 5
| |
| | |
| I BWR Component ;
| |
| Consolidation Process i
| |
| ~ -
| |
| ~ . . . . . - = .:e unm e n n .., .. . .
| |
| l I
| |
| l u Approach: scrutable, well-documented process with comprehensive PIRT based assessment: SET & IET, Presentation Focus Modeling Capability -
| |
| Model i
| |
| / Requirements implementation Transient Simulation y Requirements PIRT Based m Perform Code e Assessment Matrix Assessment 1r r ,
| |
| Accuracy _
| |
| ~
| |
| Improve OK? = Models yes it y Improve
| |
| [t Robust ? ;] no Numencs p Y'S it Eff ncy FINISHED c yo, n.
| |
| 6 4mber 16,1998 e e -
| |
| l
| |
| | |
| o o O
| |
| BWR Component Consolidation Process STEPS: Requirements => Implementation
| |
| : 1) Define Functional Requirements.
| |
| : 2) Develop Verification i est Plan.
| |
| : 3) Model implementation:
| |
| + Identify TRAC-B coding that comprises component model.
| |
| + Create component type & data base structure in TRAC-M. ;
| |
| + Re-code mode! into TRAC-M in component specific routines.
| |
| : 4) Software Design Description Document.
| |
| : 5) Perform Verification Test Plan.
| |
| + Comparisons to analytical solutions and TRAC-B results.
| |
| : 6) Prepare Verification Test Report.
| |
| : 7) Model & Documentation to NRC Configuration Control.
| |
| + Review documentation, install model, repeat verification tests.
| |
| December 16,1998 7
| |
| | |
| h
| |
| | |
| == Introduction:==
| |
| TRAC-B Jet Pump Model
| |
| : n n - - ,. . , - . ..
| |
| Drive Line m TRAC-B Jet Pump Model-TEE Side o,iyo uomi, , 'e9
| |
| + Not a " stand alone" component
| |
| '' model with numerics & physics.
| |
| wxino i + Uses TRAC-B " TEE" component
| |
| , for two-fluid numerics and for l
| |
| l -
| |
| constitutive models.
| |
| TEE + Adds momentum source / sink "Uf terms to account for inaccuracies in momentum flux differencing at l area changes.
| |
| + Adds special form loss models Tail Pipe .
| |
| 3, derived from INEL 1/6th scale test data.
| |
| D mber 16,1998 8
| |
| | |
| O O O I Code Consolidation Example:
| |
| BWR Jet Pump Model m TRAC-B Jet Pump Model:
| |
| + Convert " TEE" into "JETP" + 3 development stages i
| |
| i t
| |
| e :::: TEE lY & Y lux t input / Output Corrections P
| |
| u
| |
| * Special P
| |
| 4 Form Loss Coefficients December 16,1998 9
| |
| | |
| Code Consolidation Example:
| |
| BWR Jet Pump Model
| |
| ~ _ _ _ _ _ _ _ _ , _
| |
| m Functional Requirements:
| |
| : 1) Hydraulics: after creation of JETP component, but before modifications for jet pump specific models, the JETP and TEE components should produce identical solutions.
| |
| l l
| |
| : 2) Reversible Pressure Gains & Losses: reversible "Bernoulli" pressure changes due to area changes should be accurately calculated by the 1-D two-fluid momentum equations.
| |
| : 3) Pressure Rise due to Mixing of Suction & Drive Flows: the directional momentum flux component for the side-leg of the TEE (drive flow) must be handled correctly.
| |
| : 4) Irreversible Pressure Losses: the irreversible losses due to incomplete mixing and flow geometry changes should be calculated using the same loss coefficient formulas as TRAC-B (based on INEL 1,6th scale jet pump data).
| |
| ber 16,1998 1
| |
| | |
| o O o 'L Code Consolidation Example:
| |
| BWR Jet Pump Model i
| |
| a Functional Requirements:
| |
| : 5) Scalability: though developed based on 1/6th scale data, the Jt-I P component should be sufficiently general so that it can be applied to full scale jet pumps.
| |
| : 6) Multiple Jet Pumps: must be capable of modeling a single jet pump assembly or multiple jet pumps in parallel lumped together.
| |
| : 7) Flow & Conduction Coupling with Vessel: must model hydraulic coupling to the downcomer region of the 3-D vessel component and conduction heat transfer through walls.
| |
| 8-12) Input & Output Requirements: specifications for jet pump specific quantities on input & output files.
| |
| December 16,1998 11
| |
| | |
| Code Consolidation Example:
| |
| BWR Jet Pump Model
| |
| ~.--_ _ - ---- -
| |
| m Verification Test Plan & Results:
| |
| + For functional requirements #1-7, there was a verification test case for each of the requirements. Only the tests associated with requirements #2, #3, and #4 are discussed.
| |
| : 2) Reversible Pressure Gains & Losses:
| |
| + For TRAC-M, a "Bernoulli" correction factor is built into the differencing of the momentum equations, no special momentum source / sink terms were needed as in TRAC-B.
| |
| use the input model for the INEL 1/6th scale jet pump, test #96 single-phase liquid, frictionless, no gravity 2
| |
| compare Bernoulli number, () P + - V ), for cells in the diffuser and tail-pipe sections, and for cells in the nozzle and riser.
| |
| + Results: compared to analytical value, the largest deviation was less than 0.1% and was deemed acceptable.
| |
| l mber 16,1998
| |
| | |
| o o
| |
| ~
| |
| O Code Consolidation Example:
| |
| BWR Jet Pump Model
| |
| ..;- - _ ,.aaaccamamaw n . ,
| |
| i m Verification Test Plan & Results: ,
| |
| : 3) Pressure Rise due to Mixing of Suction & Drive Flows:
| |
| + TRAC-B does not handle the momentum transfer from a TEE side-arm correctly, so a correction factor is added as part of the jet pump model.
| |
| + In TRAC-M, this correction is not necessary as long as at least two cells are used to model the mixing region.
| |
| use the input model for the INEL 1/6th scale jet pump, test #96 single-phase liquid, frictionless, no gravity
| |
| " compare pressure rise in mixing region to Bernoulli value EE EE
| |
| ==
| |
| =
| |
| ; EE
| |
| = - ==
| |
| E EE
| |
| - AP--+
| |
| : + Results: the analytical value was 51.30 kPa, and the TRAC-M calculated value was 51.27 kPa.
| |
| December 16,1998 13
| |
| | |
| Code Consolidation Example: .
| |
| BWR Jet Pump Model L a Verification Test Plan & Results:
| |
| l :
| |
| l 4) Irreversible Pressure Losses:
| |
| + Compa:e to INEL 1/6th Scale Test Data & TRAC-B results.
| |
| ! l 3 .
| |
| 2
| |
| - jy h -
| |
| _ , _; u O -- fB 7- -E -
| |
| 4 -
| |
| -1 Negative Drive -TR AC-B - -- --
| |
| 2 --t r e a i i i
| |
| -1 0 1 2 3 M - Ratio
| |
| + Differences traced to wall friction model, an errorin TRAC-B, and momentum fluxes at a pressure boundary condition.
| |
| De er16,1998 14
| |
| | |
| O O O Code Consolidation Example:
| |
| ~
| |
| BWR Jet Pump Model a
| |
| ,...w_wmmmat. .- g m a m , .:x ,- - .. = w : r,_. . .= .; _
| |
| I a Summary:
| |
| t
| |
| + The TRAC-B jet pump modeling capability has been recovered in the consolidated TRAC-M code.
| |
| + A scrutable, well-documented process was followed:
| |
| + Functional Requirements Specification.
| |
| + Verification Test Plan. '
| |
| + Software Design and Implementation Document.
| |
| + Verification Test Report. l
| |
| + The improved treatment of momentum fluxes in the TRAC-M Tee component made the incorporation of source / sink correction factors (as in TRAC-B) unnecessary. l December 16,1998 15
| |
| | |
| o EXTENSION OF BEST-ESTIMATE LOCA METHODOLOGY TO PWRs WITH UPPER PLENUM INJECTION 1
| |
| O. PiaT aESutTS von vei PiiENOmENA M. E. NISSLEY WESTINGHOUSE ELECTRIC CORP.
| |
| (412)374-4303 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS THERMAL-HYDRAULIC PHENOMENA SUBCOMMITTEE MEETING DECEMBER 16,1998 l
| |
| : O L
| |
| | |
| j.
| |
| USE OF PIRT FOR UPI BELOCA O
| |
| IDENTIFICATION OF HIGHLY RANKED PHENOMENA USED TO GUIDE SUBSEQUENT STEPS IN CSAU (e.g.,
| |
| ASSESSMENT MATRIX, NPP SENSITIVITY CALCULATIONS)
| |
| SUBSEQUENT STEPS CLARIFY ACTUAL IMPORTANCE TO ANALYSIS RESULTS
| |
| . VARIATIONS OF HIGH'LY RANKED PHENOMENA OVER EXPECTED RANGE MAY NOT SIGNIFICANTL'Y AFFECT RESULTS g . PHENOMENA MAY BE SHOWN TO BE MORE IMPORTANT THAN PIRT RANKING SUGGESTS UPI EXPECTED TO PRIMARILY AFFECT UPPER PLENUM AND CORE REGIONS O
| |
| | |
| ^ '
| |
| O O O Table 2-2 (Cont'd) ,
| |
| PIRT for Large Break LOCA Refill Reflood Blowdown .
| |
| UPI ' W3 & 4 Loop CSAU UPI W3 & 4 Loop CSAU UPI W3 & 4 Loop Process CSAU Upper Picnum 7 7 Hot Assembly Location 7 7 9 9 9 6 Entrain 1De-entrain.
| |
| 7 6 Phase Separation 8 9 CCF Drain / Fallback i 8 8 ,
| |
| Condensation ,
| |
| Two-Phase Convection !
| |
| Hot Leg i
| |
| 9 Entrain 1De-entrain. t Flow Reversal 5 5 Void Distribution Two-Phase Convection ,
| |
| i t
| |
| t i
| |
| _- l
| |
| | |
| %_) LI s .)
| |
| Table 2-2 PIRT for Large Break LOCA Blowdown Refill Reflood ,
| |
| Process CSAU UPI W3 & 4 Loop CSAU UPI W3 & 4 Loop CSAU UPI W3 & 4 Loop Fuel Rod Stored Energy 9 9 9 5 5 8 7 7 Oxidation 5 5 8 8 8 Decay Heat 8
| |
| Gas Conductance Core DNB 8 8 7 8 8 8 8 8 Post-CHF 8 6 6 7 5 5 Rewet 9 9 9 Reflood Heat Transfer Nucleate Boiling Single-Phase Vapor Natural Circulation 9 5 5 3-D Flow 7 7 7 7 9 Void Generation /Dist.
| |
| 8 8 Entrain 1De-entrain.
| |
| Flow Reversal / 8 8 Stagnation Radiation Heat Transfer
| |
| | |
| l HIGHLY RANKED PHENOMENA UPI PLANTS 3-/4-LOOP PLANTS BREAK FLOW BREAK FLOW ,
| |
| i
| |
| ' BROKEN LOOP RESISTANCE BROKEN LOOP RESISTANCE !
| |
| l STORED ENERGY / FUEL ROD STORED ENERGY / FUEL ROD i l
| |
| CORE HEAT TRANSFER CORE HEAT TRANSFER
| |
| ; ECC BYPASS ECC BYPASS l
| |
| CONDENSATION (DC, LP) CONDENSATION (DC, LP)
| |
| ' ENTRAINMENT/ ENTRAINMENT/
| |
| j STEAM BINDING STEAM BINDING O ACCuMutATOR NITROGEN ACCuMutiTOR NITROGEN i
| |
| ; UP DRAIN DISTRIBUTION a
| |
| O
| |
| | |
| i 4
| |
| 1 UPPER PLENUM MODELING STUDIES O)
| |
| INJECTION MODELING
| |
| : 1) CONTINUOUS LIQUID VS. DROPS (VARYING SIZE)
| |
| : 2) OUTER REGION OF UP VS. SPATIALLY DISTRIBUTED i
| |
| : 3) COMBINATIONS OF 1 & 2
| |
| : LIMITING HOT ASSEMBLY LOCATION NODALIZATION ABOVE AVERAGE ASSEMBLIES e
| |
| NODALIZATION ABOVE OUTER (LOW POWER) ASSEMBLIES n c NOMINAL MODELING SUPPORTED I 4
| |
| V 4
| |
| J 1
| |
| 1 0
| |
| | |
| l INJECTION MODELING STUDY LHSI WATER ENTERS AS HORIZONTAL JET MOMENTUM LARGELY DISSIPATED UPON IMPACT ON SUPPORT COLUMNS AND STRUCTURES JET IMPACT CAUSES FORMATION OF DROPLETS, FALLING FILMS, ETC.
| |
| SOME INJECTION MAY PENETRATE TO INNER REGION OF UPPER PLENUM c HOW TO MODEL THE INJECTION?
| |
| NOMINAL MODELING: CONTINUOUS LIQUID MASS SOURCE IN AN OUTER GLOBAL CHANNEL, WITH NO MOMENTUM t
| |
| BASIS FOR RANGING STUDIES:
| |
| : 1) POROSITY OF UPPER PLENUM STRUCTURES IS <20%
| |
| FOR FIRST IMPACT OF HORIZONTAL JET, IMPLYING LIMITED PENETRATION INTO INNER REGION AT INJECTION ELEVATION
| |
| : 2) WESTINGHOUSE TESTS IN STAGNANT AIR SHOWED DROP SIZES OF % TO % IN. (0.02 TO 0.04 FT)
| |
| : 3) DARTMOUTH STUDY WITH AIR INJECTION REPORTED DROP SIZES ON THE ORDER OF 0.007 FT
| |
| , (BARATHAN ET AL.,1980)
| |
| : 4) DARTMOUTH DATA CLOSEST TO PWR TARGET FROUDE NUMBER SHOWED ALL DRAINAGE WITlHN O THE UP QUADRANGLE NEXT TO INJECTION
| |
| | |
| 9 b
| |
| U
| |
| \
| |
| t UPI 1 i
| |
| G ukle h be
| |
| ^ :: #
| |
| I /
| |
| o --
| |
| l l l /
| |
| n
| |
| 'd i l W!I iC1 '
| |
| i i j_,
| |
| OuwtRegca { I l
| |
| .-Jg L-a-.J 3
| |
| 's ' ' I I l
| |
| i l I
| |
| l I! O
| |
| % ,__ .__ _,q__;/ _... __,
| |
| 's ! I i
| |
| 'A _D.N .. 1_ L'J_q 1 8 1% I l l l /l ~I I I
| |
| _'_- 1 Igl / I I y __r_ -.
| |
| m -- H - ra --,
| |
| l l i 1 I% i i l l I l l g
| |
| J l I. _ _ .1 _ _ I .! _sL '
| |
| L . .I m =,: J
| |
| . _ _ ,._.i .-- - - - -- - --
| |
| r-- --
| |
| - - 4 *s- - '
| |
| s _ a__ __ Sk a__ __u__ a l I l 1 .
| |
| l Figure 491. Upper Plenum Configuration at UPI Injection Level i
| |
| | |
| . . _ . _ . . _ . _ _ _ _ . - _ _ _ . _ _ - - _ - - - ~ _ _ . . _ _ . _ _ . . _ _ _ -
| |
| O 4
| |
| 1 l
| |
| s s
| |
| ' \ \ i
| |
| - 4----g _L_ wa------------+-_-_____
| |
| i i ,
| |
| I l g \ \l I I g \ \1 I I \ 1 g l l g \ li I I
| |
| I (g
| |
| \
| |
| 4 l\
| |
| g i
| |
| l i g \ \
| |
| ' l O l 1
| |
| I
| |
| \
| |
| g g
| |
| 'I'
| |
| \l
| |
| \g 1
| |
| \
| |
| i l
| |
| 1 J g_ _ _ _ _ _ _ _ _ ,l -----___ .
| |
| - - - l- - - - - - - - - -\ _ _ _)1
| |
| , g g
| |
| ,\
| |
| l
| |
| \
| |
| l l
| |
| I l
| |
| \
| |
| I \ l\ l
| |
| \
| |
| I \ l
| |
| \
| |
| l
| |
| \ \
| |
| I i i i \ l I i l
| |
| \ \
| |
| I \ l l
| |
| \ \
| |
| i \g \ l
| |
| \
| |
| \; \ l
| |
| \
| |
| l- ) \ \ l .
| |
| I i\ \ \ l l l\ \ \ l
| |
| - - - i_ _ - _ _ _ _ _ _ _ _ _ _g _\ g_ _r- r--__q__-____.
| |
| , ' ' i i
| |
| Figure 49-3. First Impingement of UPI Water into UP Structure
| |
| - . . , , m.q-- . - _
| |
| | |
| c .
| |
| i O
| |
| [
| |
| l l.
| |
| 270 DEG.
| |
| t l
| |
| 50 S
| |
| d- .,
| |
| ... g. . _
| |
| %, ,,, - , - --. :\, ,
| |
| E O
| |
| 7 l l 57 SC } .l l 'h 180 DEG.
| |
| 053 --- ---*
| |
| .. GT .*
| |
| e--- ---
| |
| h ODEG.
| |
| 'Q g
| |
| ' h -
| |
| ' ~~~~---'
| |
| U U 1, .
| |
| 5.
| |
| 51 55 '4- -
| |
| . .. .... *- 52 -
| |
| M;.58v
| |
| "~~"
| |
| 59
| |
| . 48 90 DEG.
| |
| O Figure A-7. Point Beach ECOBRAffRAC Vessel Section 6
| |
| | |
| o Dj@
| |
| =
| |
| O@c @N94
| |
| @ l G +G if~ el L O D b Dl6'8 $@
| |
| / O g@Ol g88 . e g. .. - )
| |
| o j .
| |
| dj = 9.5 nus -
| |
| .. . .... .a ,u ,,., -
| |
| Angle of Incidence o' velocity of the Jet V) (m/s) 4 .g Freude Number 7, y Total Drainage from All Rods 86 7 Total Drainage from Rode 2*
| |
| under Measurement J I''' O 7 YI. /
| |
| Average of Measurements ; g*g , g, withinscatterEnvelope]
| |
| standard Deviation 70
| |
| ' *?otal Drainage in scatter /No. of Rods]2 Envelope 7. 2. :s &g.7//2
| |
| * Max. Later al spread j Man. Longitudinal spread (f Area Coverage f. (
| |
| Ms. of Rode in Envelope / g,,
| |
| i
| |
| !O Figure 49-6. Drainage Data Closest to PWR Target Froude Number
| |
| | |
| INJECTION MODELING SENSITIVITY STUDIES 1
| |
| SOURCE LOCATION DESCRIPTION, PCT (F)
| |
| OUTER GLOBAL CONTINUOUS LIQUID 1758(NOM)
| |
| OUTER GLOBAL 0.04 FT DROPS 1702 OUTER GLOBAL 0.01 FT DROPS 1736 l l
| |
| OUTER GLOBAL 0.007 FT DROPS 1759 60% OUTER /40% INNER CONTINUOUS LIQUID 1767 60% OUTER /40% INNER 0.007 FT DROPS 1778 40% OUTER /60% INNER CONTINUOUS LIQUID 1744 40% OUTER /60% INNER 0.007 FT DROPS 1768 CONCLUSIONS
| |
| : 1) NOMINAL MODELING IS ADEQUATE
| |
| : 2) EXPLICIT UNCERTAINTY TREATMENT UNNECESSARY l
| |
| l l
| |
| O l
| |
| | |
| l i
| |
| O i
| |
| l Point Beoch Split Break Cd =
| |
| UPI Injection L o c(o t i o n 0.6) Transient Comporison Peak Clod Temp / Orop Site Study eratures Case C8: Outer injection. Continuous Liquid Film Case OR1: Outer injection. Drop Size = 0.01
| |
| ---Case D R 2 i- Outer injection. Drop Size = 0.04 7800
| |
| : % v 4 ,'
| |
| 1600 l
| |
| . %.I
| |
| ^ 1400 l w \
| |
| b*
| |
| 1200 u b
| |
| " h 1000
| |
| - rm C3 u
| |
| * b ,h
| |
| /
| |
| e 800 ,'
| |
| a .
| |
| E -
| |
| v -
| |
| i 600 -
| |
| \ ,
| |
| 1
| |
| \
| |
| 200 0 50 100 150 200 Iime (S)
| |
| O Figure 45b 2a. PCT Time History of the Continuous Liquid Injection vs. UPI Drop Size 0.04 feet and 0.01 feet
| |
| | |
| ...._ __.._._.___._ _ _. _ _ __ _ . _ . . _ __ _ _ _ _ _~ . __ . ._. _._
| |
| .7 __ _ _
| |
| HOT ASSEMBLY LOCATION STUDY l
| |
| DIFFERENT HARDWARE CONFIGURATIONS IN UPPER PLENUM CAN AFFECT DOWNFLOW DURING BLOWDOWN AND CCFL DURING REFILL /REFLOOD 1
| |
| HIGHEST POWER ASSEMBLY COULD BE LOCATED AT ANY LOCATION EXCEPT CORE PERIPHERY (NEUTRON LEAKAGE) !
| |
| I l
| |
| i c NEED TO DETERMINE BOUNDING LOCATION i
| |
| l BASIS FOR NOMINAL MODELING: REVIEWED AND RANKED ALL INTERNAL POSITIONS, CONSIDERING
| |
| ; a) FLOW AREA AT UPPER CORE PLATE b) CROSS FLOW RESTRICTIONS IN UPPER PLENUM c) FLOWPATHS TO UPPER HEAD 1
| |
| d) NEIGIIBORING ASSEMBLIES i
| |
| l I
| |
| 4
| |
| -- s ,
| |
| | |
| NOMINAL MODELING: HOT ASSEMBLY UNDER FREE- 1 O STANDING MIXER, NEXT TO AVERAGE CHANNEL UNDER SC/FSM/OH I
| |
| SENSITIVITY STUDIES:
| |
| : 1) HOT ASSEMBLY UNDER SUPPORT COLUMN, SURROUNDED BY AVERAGE CHANNEL UNDER GUIDE TUBES f 2) HOT ASSEMBLY UNDER GUIDE TUBE, SURROUNDED BY I l AVERAGE CHANNEL UNDER SC/FSM/OH j CONCLUSION
| |
| ,
| |
| * NOMINAL MODELING IS ADEQUATE i
| |
| O i
| |
| i O
| |
| | |
| ~
| |
| O Point Beach Split Break Transient Results Comparisan Hot Assembly Location Study Peak Clod Temperature HA = FSW. connected to OH/SC channel
| |
| ----HA=SC. connected to GT channel
| |
| - - -- HA= GT. connected to OH/SC chonnel 1800 I
| |
| 1600 I ''
| |
| - \
| |
| ~
| |
| \
| |
| r.
| |
| ^ 1400 u_
| |
| '?,' ,s i v '7,-
| |
| s 1200 , ',
| |
| e '
| |
| 's., Nr,g
| |
| ' ' s' s 3 1000 ,
| |
| k,,'s ,
| |
| o ~_
| |
| ; u _ ,' '% , '
| |
| * 800 ,
| |
| : o. - , b
| |
| - t E -
| |
| ' . t l e -
| |
| ', t r 600 i
| |
| - i
| |
| ,. I
| |
| : - i a t
| |
| ! 400 i
| |
| - ''. t
| |
| . . . . . O. . .
| |
| 200 200 O 50 100 150 i Time (s)
| |
| O Figure 16-2. Peak Clad Temperature f
| |
| | |
| 1 ELETION OF CONDENSATION RANGING O IN THE DOWNCOMER AND LOWER PLENUM I
| |
| l 3-/4-LOOP ANALYSES CONDENSATION IN DC/LP IS EXPLICITLY RANGED, BASED ON UPTF DATA UNCERTAINTY REPORTED BY MPR (50 TO 100% OF NOMINAL)
| |
| EFFECT IS SEEN AFTER BEGINNING OF REFLOOD UPI PLANT i
| |
| l ACCUMULATOR FILLS DOWNCOMER AND
| |
| !Q LOWER PLENUM TO INITIATE REFLOOD LHSI INTO UPPER PLENUM DOMINATFS i
| |
| REFLOOD PWR SENSITIVITY STUDIES SHOWED NEGLIGIBLE EFFECT OF CONDENSATION RANGING IN DC/LP O
| |
| : t. .
| |
| l
| |
| !O l
| |
| l l
| |
| KN = 0.77. XCOND = 10
| |
| ----KN=077. XCOND = 0.5 1800 f\
| |
| 9
| |
| ~
| |
| >~,
| |
| 16CO r vvv
| |
| - s s
| |
| 1400 m .
| |
| ~
| |
| : u. .
| |
| v .
| |
| g 1200 \
| |
| e \
| |
| 6 \
| |
| =0 h ss
| |
| .sT E -
| |
| Ns N 9 e
| |
| ca. 800 5
| |
| E _~
| |
| * i w _
| |
| i 600 I
| |
| _ t{ i 400 i
| |
| ',*~,-
| |
| . Vy
| |
| , , , , , , i e i ' '' ' ' ' '
| |
| ' o 50 100 160 200 Time . (s) l Figure 23b-1. Downcomer and Lower Plenurn Condensation Study at Ku = 0.77 L s. Peak Clad Temperature i
| |
| .v,
| |
| | |
| I O
| |
| KN = 1.58. XCOND = 1.0
| |
| ----KN= 1 58. XCOND = 0 5 -
| |
| 1800
| |
| ~
| |
| lj %
| |
| 1600
| |
| - %f )
| |
| . -f s
| |
| . \
| |
| 1400 \
| |
| ^ . )
| |
| L .
| |
| : v ,
| |
| 1200
| |
| : O
| |
| $ 1000 .
| |
| y.
| |
| u %
| |
| _g = V
| |
| ; a 800 b E
| |
| T i
| |
| w -
| |
| 600 \I i
| |
| p 400 '
| |
| l
| |
| ~
| |
| es 200 ' ' '
| |
| O 50 100 100 200 Time (s)
| |
| O Figure 23b.2. Downcomer and Lower Plenum Condensation Study at Ku = 1.58 Peak Clad Temperature
| |
| | |
| . . - . . . . - . . . _ - . . - - . . . . - - . . - . . . - . . . - - . = . - . - - . . - - . . . - . _ . . . - . - - _ _ _ - ,
| |
| I l
| |
| : f. . .
| |
| O l
| |
| l r
| |
| l KN = 2 40, XCOND = 1.0
| |
| ----KN l = 2.40. XCOND = 0.5 1800 ..
| |
| si 1600 .
| |
| . s 3
| |
| - s 1400
| |
| : t. m . s L t.a . \g v . N 1200 h l . \s
| |
| :3 1000 '',k
| |
| 'nv o -
| |
| u
| |
| - \
| |
| . t'',
| |
| a 800 E
| |
| )i $
| |
| w -
| |
| I 600
| |
| . 's 1
| |
| I 400 ' l
| |
| - \
| |
| . \'. ~ . .k .
| |
| 200 O 50 100 100 200
| |
| ; Time (s) l I
| |
| l O Figure 23b 3. Downcomer and Lower Plenum Condensation. Study at Ku = 2.40 f ''
| |
| ' Peak Clad Temperature f-l l
| |
| | |
| i CONCLUSIONS
| |
| (
| |
| UPPER PLENUM PHENOMENA RANKED HIGHER FOR UPI
| |
| ; PLANTS' i
| |
| e HOT ASSEMBLY LOCATION (7)
| |
| ENTRAINMENT!DE -ENTRAINMENT (9) e
| |
| , PHASE SEPARATION (7)
| |
| ; e CCF DRAIN / FALLBACK (9) e
| |
| , CONDENSATION (8) f PWR SENSITIVITY STUDIES PERFORMED TO CONFIRM NOMINAL MODELING
| |
| !
| |
| * INJECTION MODELING 4 e HOT ASSEMBLY LOCATION j e NODALIZATION
| |
| * CONDENSATION IN DC/LP O
| |
| ASSESSMENT OF SET & IET NEEDED TO IDENTIFY PHYSICAL MODELS TO BE RANGED
| |
| )
| |
| v l 1
| |
| 4 n.
| |
| | |
| l
| |
| /
| |
| b .
| |
| II.5 Code Validation, Sensitivity of Consequences to Model TABLE OF CONTENTS:
| |
| 5.1 Analyses of GE Subcooled CCFL Tests ......... [RAI 36]
| |
| 5.1.1 Tests and Model 5.1.2 Drain Rate 5.1.3 Subcooled CCFL at A Perforated Plate 5.2 Analyses of CCTF for UPI, Run 72 and Run 76....... ... [RAI-30]
| |
| 5.2.1 Test and Model 5.2.2 Prediction vs. Data (PCT, Quench Front, Bundle & UP Void Fraction, and HL Flow) 5.2.3 Subcaoled CCFL ...... . [RAI-29]
| |
| Prediction vs. Bankoff Correlations 5.2.4 Quench Front ............ [RAI-27]
| |
| Prediction vs. Data 5.3 Analyses of UPTFTest 20 5.3.1 Test and Model 5.3.2 Prediction vs. Data ... [RAI-32), [RAI-46]
| |
| hf-n. (Flow Rates, UP Condensation, UP Liquid Level) 5.3.3 CCFL Prediction (Phase C) vs. Bankoff Correlation ....... [RAI-46]
| |
| 5.4 Ranging of XYDRAG and XCONDU and Effexcts on Validation Tests .... [RAI-24]
| |
| 5.4.1 Ranges Determined by GE CCFL Analyses 5.4.2 Effects on CCTF 5.4.3 Effects on UPTF 5.4.4 Application to PWR 5.5 Prediction of Scaling Trends ......... [RAI-37]
| |
| 5.6 Range of Conditions in The Assessment Test Matrix vs. UPI Plant Kenji Takeuchi We's tinghouse Electric Co.
| |
| (412)374-4263 Advisory Commitee on Reactor Safeguards Thermal-Hydraulic Phenomena Subcommittee Meeting Dec.16,1998
| |
| >{}
| |
| < frame /UPl/ACRS/ Pres 2. FIG >l
| |
| | |
| 2-l 1
| |
| o G l 1
| |
| l Code Validation Items, Tests, and RAIs I
| |
| Code Validation Item GE Test CCTF UPTF PWR 1
| |
| UP Liquid Distribution RAI-30 RAI-46 <RAI-18>
| |
| UP Pooling RAI-30, RAI-46, RAI-37, RAI-37 RAI-37 <RAI-18>
| |
| UP Liquid Downflow RAI-36 RAI-29, RAI-35, RAI-28, RAI-37 RAI-46, RAI-37
| |
| | |
| ===RAI-37===
| |
| Hot Leg Liquid Flow RAI-30, RAI-33, RAI-37 RAI-37 RAI-37 l_O Ue Ceedensetien <R^1-37) RAI-33 <RAI-37)
| |
| | |
| ===RAI-46===
| |
| Quench Front RAI-27 NA <RAI-9>
| |
| PCT RAI-30 NA <RAI-9>
| |
| Scaling Trends RAI-37 RAI-37 RAI-37 UPI Model RAI-45b (RAI-) Indirect data.
| |
| <RAI- > Subject Discussed w/o data.
| |
| 1 4
| |
| 0
| |
| < frame /UPl/ACRS/ Pres 2. FIG >2
| |
| | |
| 3 l
| |
| l
| |
| * II.5.1 Analyses of GE Subcooled CCFL Tests 5.1.1 Tests and Model TABLE 36-1 The GE Test specification l
| |
| l Wup; T f,in Subcooling , .
| |
| L.iquid Injection (kg/sec) ( C) AT ( C) l Kun 60 U.602 5/.6 64. Bypass Slots Run 61 0.753 37.8 64. Bypass Slots Run 62 0.934 37.8 64. Bypass Slots Run 69 0.608 93.3 8.2 Bypass Slots i Run 73 0.625 37.8 64. Spillover 1
| |
| l l
| |
| l p pressure B.C.
| |
| l l 1 12.0" 6.0 l 6.0 h 1
| |
| 2.875 h f'n. ject on I
| |
| 6.0 @ 5 g I
| |
| 9.063 g ;
| |
| 9.063 10.0 4 10.0 13.625
| |
| ~----~
| |
| Steam 13.625 g Injection
| |
| __{}___{p_.
| |
| l __ (3] _
| |
| q Figure 2-1 WCOBRA/ FRAC Model of The GE Test Facility V
| |
| < frame /UPUACRS/ Pres 2. FIG >3 l
| |
| l
| |
| | |
| +
| |
| l l
| |
| O(% 1 Table 3-1 Comparison of GE Tie Plate and Bankoff Plate l l
| |
| l 15 Hole Plate PWR UCP
| |
| _ (Bankoff et al., GE Tie Plate )
| |
| uter O H 1981) j Perforation Rutio* 0.420 0.436 0.308 Hydraulic Diam. 0.413 in. 0.455 in. 0.407 in. i Thickness 0.79 in. ~ 0.3 in. 1.5 in.
| |
| * The ratio of total flow area to the total flow and metal areas.
| |
| 1 Subcooled Flooding: i p-BC Oubcooled Liq. The superficial velocities ofliquid and
| |
| [' Injection steam are computed with the steam injec-tion rate (W,) and the drain rate (Wj) at the bottom of the test section.
| |
| Upper Plenum y O 0 8 " p,FA Pi P.
| |
| ll ll ll ll ll I'A W' JI = p;FA where FA is the flow area of the perforated plate and pi and p, are the densities on the Bundle Region top of the plate.
| |
| This m.thod of flooding analyses follows the test data analyses (Sections 5.2 and 7.0 N of Jones (1977)).
| |
| Steam Injection (W,;
| |
| I U Drain Water Rate (W i) l l
| |
| l t.g\
| |
| 7
| |
| <frameNPl/ACRS/ Pres 2.FIO>4
| |
| | |
| . - . - . _ - . . _ . - =_ - . .- .- _. - , .-. .. . . - - - - - -
| |
| N l
| |
| l l %J m. ;
| |
| ^
| |
| .oe A T i
| |
| { c:n
| |
| \ l x g _
| |
| \
| |
| l V _
| |
| \
| |
| g.
| |
| ~
| |
| 3B: .@
| |
| 3 : [
| |
| w
| |
| *b
| |
| : j v1 _
| |
| m -
| |
| \
| |
| ~
| |
| .O 0
| |
| O 20 40 60 80 100 120 140 100 Time (s)
| |
| Figure 5-2 Sream Injection Rate, Test Run 60 0 E 1
| |
| \
| |
| .5 x -1 v -
| |
| w 5
| |
| < - 1. 5 x -
| |
| 3B: -
| |
| O -
| |
| .J -2 w -
| |
| in 5 v1 -
| |
| < -2. 6 y .
| |
| i,, ,,, ,,, ,,, ,, ,,, ,,, ,,,
| |
| 3 U 20 40 60 80 100 120 100 140 Time (s)
| |
| Figure 5 3 Liquid Drain Rate at The Bottom of The Test Sec-tion, Test Run 60 0
| |
| < frame /UPl/ACRS/ Pres 2. FIG >5
| |
| | |
| J l
| |
| l 5.1.2 Drain Rate vs. Steam Injection Rate O
| |
| s 1.5
| |
| ~
| |
| a Data for stec m increasmg phase A Data for steam decreasing phase
| |
| ^
| |
| 0 0 Prediction for steam in< reasing phase O I rediction for steam de<:reasing phase o 1
| |
| * O 3 D EBB @Qg 46 ^
| |
| o ao eA 1 w .5
| |
| % 0 j
| |
| = -
| |
| z
| |
| ~
| |
| 1 a
| |
| < 0 : OCD SC ^^ 3.
| |
| m -
| |
| O pe
| |
| .5 ' ' ' ' ' ' ' ' '
| |
| O '.02 '.04 .06 .38 '.0 ' .12 STEAM INJECTION RATE (kg/sec)
| |
| O Figure 36-2 Liquid Drain Rate vs. Steam Injection Rate, for
| |
| \-
| |
| GECCFL Test 61 1.5
| |
| ~
| |
| A Data for steam inc easing ph2se
| |
| ~
| |
| A Data fcr steam decreasing phase 1
| |
| ^
| |
| ~
| |
| D Predict lon for steam increasing phase
| |
| . 3 O Predict on for steam decreasi ig phase m -
| |
| N m -
| |
| x .
| |
| w .5 bhUU% a
| |
| * O N
| |
| < ~ 0 A
| |
| % A A
| |
| ~
| |
| . ,M e i
| |
| < 0 . - - -
| |
| i i
| |
| e -
| |
| O .-
| |
| .5 ' ' ' '
| |
| 0 ' .02 ' ' ' .(4 ' .06 ' ' ' . 38' ' '..0 ' .12 i
| |
| t j STEAM INJECTION RATE (kg/sec)
| |
| Figure 36-4 Liquid Drain Rate vs. Steam Injection Rate, for GECCFL f
| |
| p Test 69
| |
| <frameNPUACRS/ Pres 2. FIG >6
| |
| .]
| |
| | |
| 7
| |
| ) 5.1.3 Subcooled CCFL correlation on a perforated plate The Bankoff flooding correlation (Bankoff et al.,1981) for subcooled liquid injection on a perfo-rated plate is expressed in terms of Kutateladze number for the tie plate of GE CCFL test facility.
| |
| (kg,,)1/2 + (kf)1/2 = 2.00 (29-1) where Kutateladze number is defined by r 2 3 1/4 pf kf e jf (29-2)
| |
| (80(Pf- Pg)j and kg,,* is the effective dimensionless steam flow rate,
| |
| "
| |
| * I' '"
| |
| k<-k-s s h,f N k p, *I.in (29-3) where Tfin is the injected water temperature kfin
| |
| * is the Kutateladze number ofinlet flow rate with the perforated plate area kg* and kf* are the steam injection rate and the liquid drain rate with the plate area Cpis the liquid specific heat fis the condensation efficiency (Bankoff,f = 0.24)
| |
| I f^j The system pressure is ~1 bar (15.0 psia)
| |
| U k = 1.9453jf . (29-2A) k ,, = k -f(AT)(0.04117)k ,,, (29-3A) 2 Ah= 0.00781 m ............ The flow area at the top of CCFL channels jf, f, = IVypj/(pf 3) A = 0.0477 m/sec --->
| |
| k* fin = 0.3044.
| |
| For All Test Cases,
| |
| \Vupi TJin subcooling . . fCpAT pf .
| |
| (kg/sec) AT l/> '_"
| |
| (oC) fi" h,f { I. i" Test 60 0.602 37.8 64. 0.1565 0.3044 0.3399 (f= 0.24)
| |
| Test 61 0.753 37.8 64. 0.1887 0.3671 0.4099 Test 62 0.934 37.8 64. 0.2357 0.4585 0.5119 Test 69 0.608 93.3 8.2 0.1517 0.2951 0.0379 Test 73 0.625 37.8 64. 0.1565 0.3044 0.3399 p
| |
| l _.Y l
| |
| < frame /UPI/ACRS/ Pres 2. FIG >9
| |
| | |
| U l
| |
| l l
| |
| 3 A Data in the flooding relationship O Prefiction of the looding relationship 2.5 -
| |
| Bankoff saturated flooding correlation Bankoff subcooled flooding correlation 2
| |
| s\ i
| |
| \\ l N N
| |
| ~
| |
| A N 15 M bA N O \ N i k, -
| |
| ~
| |
| t N
| |
| -s.
| |
| O I
| |
| N J= o 60
| |
| 'O \ '
| |
| l
| |
| " N 4= 0. 14 s
| |
| . a
| |
| : a a 0
| |
| O ''''.5 1 ''''.5 1 2 2.5 3
| |
| -k f
| |
| Figure 36-7 Bankoff Flooding Correlations vs. Test Data and Prediction, for GECCFL Test 61 l < frame /UPl/ACRS/ Pres 2. FIG >l2 j
| |
| | |
| h
| |
| ^
| |
| / )\
| |
| ( '
| |
| )
| |
| I 3
| |
| ~
| |
| A Daia in the floodiis relationshig ,
| |
| O Prediction of the iooding relati >nship 25
| |
| _ Bankoff saturited flooding <orrelation 1
| |
| l 4
| |
| 2 l l
| |
| \
| |
| $4 I
| |
| _p 1 1.5 *O t.
| |
| 06
| |
| * ,% f
| |
| % l
| |
| ~
| |
| O -
| |
| o 00 5
| |
| 0 E
| |
| tit 0
| |
| 0 .5 ''''.5 1 2 2.5 3 i
| |
| I
| |
| * g-k, Figure 36-9 Bankoff Flooding Correlations vs. Test Data and Prediction, for GECCFL Test 69 l
| |
| O l
| |
| < frame /UPl/ACRS/ Pres 2. FIG >14
| |
| | |
| l lD l i
| |
| l n II.5.2 Analyses of CCTF Tests
| |
| : v 5.2.1 Test and Model l Cylindrical Core Test Facility full height "$74-l%.. ?
| |
| ,, , tid 1/ll-th scaled (FA) UPI PWR g. .q !s s!*i.v--.s .I&'. g' ' %ad 32 of 8x8 assemblies a a='= a*: i eg 3 m inPm7 m i a* ~~#
| |
| *i N IN"!al *f~ C for Refill and Reflood e +. < % 1, ; s#' e- a-i Run 72 and run 76 for UPI .J e'ut...m.mme a
| |
| + "U !s*t *m.i:m a
| |
| +1 14.amme C m1 ti'6
| |
| ' +
| |
| * i
| |
| [.. .,,, l l iQi #r .%.i 8 g. ... g LOOP 2
| |
| * 8' LOOP 1 cIm ab him HN asba g .. p m. ... . . . ._..... _ .
| |
| g ,,
| |
| 7 .w. .
| |
| ," u 7,,*, s * ....'
| |
| f m.7 j -..~. .~~. . . .f . .
| |
| 4 i
| |
| s
| |
| ...i._... ...
| |
| -! . ..I.... .i. m . i %. 7..g.~
| |
| Y s 1 i MI l # g'; +
| |
| "a ""
| |
| 4.....y . ,D.. . .l.... 4 4-.
| |
| , ,, . O. . ...O. .
| |
| 9-4.. . . 4. . .~. e......e . .. . i . ), .
| |
| . . . ... . } *.m......,..-~.... . . . , . . . . . .
| |
| a m I l i 0.. } E....
| |
| s H3.D. E . .r ..E.._7 D. . { .D .G ~"" ..! .D. 7 7.G.
| |
| p ;
| |
| W %,
| |
| f j.T ,, ,
| |
| G ,,
| |
| N% e
| |
| == ' ."..'j..~ .,it.s j'.;.;..~i .e
| |
| ..q. .
| |
| ( / "
| |
| ., \ ..
| |
| ......-f...4.......
| |
| ' ..r i-a * " ' j .
| |
| .I . . . .. 3 .
| |
| ....h.I..... m 4-Q
| |
| ,," .W.j 2 j. 2.. W.iRi 3 mjs! -e. s s r s s 4-!s!# i ! '
| |
| ri g" C o--
| |
| 8 - o-
| |
| **
| |
| * Y "
| |
| a u
| |
| . Figure 2A CCTF VesselInternals Modeling LOOP 3 O- ' o' '
| |
| O-oc Figure I CCTF Prhnary taop Medeu.g Q p ouest V
| |
| ~
| |
| s/ e I-BMs,8
| |
| ( )
| |
| E\
| |
| # w
| |
| '89Ye SECTION 4: UPPEA PLENUM f
| |
| ! r--
| |
| 1 t
| |
| Figure 2B CCTF Ves.el Cross Sectional View of The Upper Plenum l
| |
| < frame /UPI/ACRS/ Pres 3. FIG >l e
| |
| | |
| //
| |
| /% 5.2.2 PCT & Quench Front V (Prediction vs. Data)
| |
| Predi': tion W N
| |
| , h .. l y u. . .
| |
| ,,,, PreJiction om l
| |
| E. 4 E /Wb '
| |
| 5 ll,'l /
| |
| ; :/
| |
| % .- l 1 - _- l i .. , .
| |
| n im. im. 3r. 's4 .o. . ,.
| |
| TlWE AFTER TEST INibrATION (S)
| |
| * Figure 30-21 Run 72, Cladding Temperature at 1.85 m outmen'iist .:l.*r ii. i.U t...) "
| |
| (6.08 ft) Elevation, Prediction vs. Data Figure 271 Run 72, Quench Front of IIP Rods, Prediction vs. Data 1
| |
| (
| |
| s .
| |
| 'redictio>n N #
| |
| N
| |
| .._ h -
| |
| g o..
| |
| ;t... Pre <liction s
| |
| ^
| |
| m
| |
| !,... 7 h i' -
| |
| /
| |
| !..,, /
| |
| , V *
| |
| . /
| |
| /
| |
| a,
| |
| ! , /
| |
| 11 lu. 2h. SU. 4W. Su. .' .
| |
| TlWE AFTER TEST IN TION (S) l Figure 30-24 Run76, Cladding Temperature at 1.85 : ,
| |
| (6.08 ft) Elevation, Prediction vs. Data '
| |
| l outdd nut r t r i*ri../le..)
| |
| * Figure 27 2 Run 76, Quench Front of IIP Rods, Predletion vs. Data
| |
| ( ;
| |
| '% /
| |
| < frame /UPI/ACRS/ Pres 3. FIG >2
| |
| | |
| /2 5.2.4 Void Fraction in Bundle & UP
| |
| .U
| |
| ] ) !) IredictiOf1 ,
| |
| f .1)ata PTCdiction l
| |
| .. . . eata .
| |
| z a . f !
| |
| l
| |
| =
| |
| 6 I
| |
| <e:
| |
| . - . i, 1 o ..
| |
| o .4 - .
| |
| O ,
| |
| .1 _
| |
| .2 ,
| |
| .. ..,,:. . e.:. . .
| |
| 0 til. .id. .i.
| |
| Tim (s)
| |
| Figure 30-6 Run 72, Hundle Void Fraction in Average Cha nel at 3 2 3.93m Elevation, Figure 30-16 Run 76, Bundle Vold Fraction in Averagi Channel at 3.32 - 3.93 m Elevation, Prediction vs. Data O
| |
| Pr: diction i
| |
| ) PreJiction
| |
| ., h Ij
| |
| = " ""
| |
| sij g{ I = = ca a
| |
| ;l i
| |
| j ,
| |
| ,g ,
| |
| e ; li <
| |
| k a L, t, y S
| |
| 2
| |
| { l d 0' i i .
| |
| < (ig a '' " "
| |
| y o ..
| |
| E .
| |
| o
| |
| ' . . . . . . . . . A .. .
| |
| , ,r. .". ..
| |
| ,c. , , , ju. , p . .; ,, , , , , , , ,u, Figure 30-8 Run 72, Upper Plenum Vold Fraction at The Bottom ofIlot Leg Elevation, Figure 30-18 Run 76, Upper Plenum Void Fraction at
| |
| , Prediction vs. Data The Hottom of Ilot Leg Elevation, Prediction vs. Data O-
| |
| < frame /UPl/ACRS/ Pres 3. FIG >4
| |
| | |
| i i
| |
| //
| |
| l l f 5.2.5 IIL Flow Rates l
| |
| Preliction u u ' ' **
| |
| Predic' ion s e a Data g
| |
| : a. 4 a. 4 0
| |
| =
| |
| =
| |
| g :. ., g :.
| |
| m ,
| |
| Q.x--n / ',
| |
| m ..'ly (I w)Ll .p tf a'' '1i a , ya = ' "
| |
| r j i,. ir. so. .o. .o. . ,. o io. im. .~. .o. .o. .u. 7u.
| |
| Time (s) Time (s)
| |
| Figure 30-9 Run 72, Total Liquid Flow Rate into The Figure 30-19 Run 76, Total Liquid Flow Rate into The 1101 Leg, Prediction vs. Data llot Legs, Prediction vs. Data O
| |
| y, Predi : tion
| |
| } Prediction C m si Data 4 mm Data
| |
| ,_ l ll ,
| |
| IL 2__.
| |
| f g, RI p
| |
| , 1 #
| |
| RE.
| |
| I la j_
| |
| : g. S .
| |
| d d C- '
| |
| y-5 5
| |
| -' u -*
| |
| io. im. 3 .. .,. , ia. , ,. .. 4,. . ;. .o. su.
| |
| l Time . ( s ) 4r. Time (s) l Figure 30-10 Run 72, Total Vapor Flow Rate int a
| |
| [ p The Ilot Leg, Prediction vs. Data Figure 30-20 Run 76, Total Vapor Flow Rate into The i
| |
| C) Ilot Legs, Prediction vs. Data f
| |
| < frame /UPI/ACRS/ Pres 3. FIG >5
| |
| | |
| 5.2.6 CCFL (k, - 1.578)1/2 + (kf )1/2 =2 (k,,,)1/2 + (kf )1/2 = 2.00 for Run 72 r 2 3 1/4 Pf
| |
| * 1/2
| |
| * kf s/ f (ks -0.861) + (kf )1/2 =2
| |
| <88ca(pf - p,),
| |
| for Run 76 ks , e ~-sk ' fCP (T,o,- Tf , f,,) pf *f i''
| |
| h fg Pg f= 0.24 System Pressure = 40 psia 2
| |
| AT = 170 F Wup1 = 47.9 lbm/sec for Run 72 Wupi = 26.1 lbm/sec for Run 76 AUCp = 1.186 ft2 , excluding GT area.
| |
| A U
| |
| 3 3 l
| |
| 38 O O % ,,,
| |
| DD m 5 N \ -
| |
| N N o
| |
| n oN, A I,, m N
| |
| ' yoN%% '
| |
| , 7 oN E, \. "
| |
| s a:
| |
| o
| |
| " n i
| |
| a o a: e
| |
| ' 3 n a 's \ ~
| |
| giao o .,
| |
| a q r t ('- k r . ) I %c 80 m o
| |
| 'o s , i.s 2 r.s a Figure 28-2 Run72 Flooding Behavior at The Top of LP Core Channel, Prediction vs. Correlation Figure 29-1 Run 76, Flooding Behavior at The Top of LP Core Channel, Prediction vs. Correlation I
| |
| < frame /UPI/ACRS/ Pres 3. FIG >6
| |
| | |
| /f O II.5.3 Analyses of UPTF Test 20 5.3.1 Test and Model D
| |
| [Y 2 D U@
| |
| @ l
| |
| - I tag m
| |
| .e 1 y (F.
| |
| ~
| |
| f law N
| |
| @ ' i 1-4
| |
| * a
| |
| > sb p4 .f
| |
| {; 0 'I @4 '
| |
| : 3) e p 5.
| |
| 1 Vessel @ Surgeline-Nozzle 2 ~ Steam Generator Simulator (Intact Loop) ECC-Injection Nozzles (Cold leg) 3a Steam Generator Simulator / Water Sepa- ECC-Injection Nozzles (Hot leg) rator (broken Loop Hot Leg) Core Simulator Injection Nozzle 3b Water Separator (broken Loop Cold leg) TV-Drainage Nozzle 3c Drainage Vessel for Hot Leg Steam Injection Nozzle 3d Drainage Vessel for Cold leg Drainage Nozzle 4 Pump Simulator 5a Break Valve (Hot leg) 5b Break valve (Cold Leg) 6 Containment Simulator Figure 1 Schematics of UPTF Test Facility O
| |
| < frame /UPI/ACRS/ Pres 4. FIG >2
| |
| | |
| m .. _ _ . _ . . __. .. . .._ __..._____ ,m m. __ _ _ _ > ._ . _ . . _ - _ _ _ _ ._ . _ _ _ _
| |
| a d
| |
| j
| |
| ; O
| |
| ~~
| |
| Break 1
| |
| 1
| |
| , Co m o Pump O''*" 80-
| |
| ' steam i
| |
| FM i ,
| |
| * SG i Loop is ,,
| |
| - - - - ,, ., L=,8
| |
| . re si a e' s ESSEL L 8 CL en m wei , si was.
| |
| FE Fu Fe Fu i gggg Rn es .=
| |
| Inner Channels Intact CL Core Simulator infectors '*
| |
| m=
| |
| L''' O Juncton Number O ComponentNurnber O
| |
| v . . . . .
| |
| =L =L =L sw w si w.=r := won sw.m won FW FG FW FN Fm FN Fe Fu RRRRRRRR outer Cm Core Simulator triectors Figure 2 3YCOBRA/ TRAC Loop Component Model for UPTF Test 20 0
| |
| < frame /UPl/ACRS/ Pres 4. FIG >3
| |
| | |
| IJ l
| |
| l l
| |
| 1
| |
| ,\
| |
| V =- . .. ; ., ;
| |
| 'rul W , E I Tn p~ -
| |
| uhe sec.
| |
| ean 7 1
| |
| l w; . +,x .ob iMj-A,. i, EIE) saa r. . ; r .. . .u. .. . . .c. .. . . . .
| |
| ;: - ' " : e . c.. s.~ s l
| |
| .i.-.e..+...- .
| |
| I"
| |
| ! af c. W
| |
| ;i.r. ~ , *.a a .. i. c, r ,. ...
| |
| ~~5 ~.S..E ? set tm.."%"
| |
| e . . u -a -
| |
| .\ *.
| |
| r: a : .-w ; _ : 3=. f .. e- t e... .g
| |
| -1.:.;c q. . .
| |
| 7.. ...g... ,
| |
| i..... :..L... ,q!;,q ....)......-...i_.......t.... . . ...:... .... . .i .p. S "
| |
| ..7 .J .
| |
| -. ,q ;
| |
| . n. _ . ...e ... .. .
| |
| m ,;'*.!.. .. .?............i...........9..............l .... A' ''6 . . . . . . .4.
| |
| . .. . se
| |
| ..E..,! .E. .h.- E... ... .i 8.......
| |
| . .. . 8... 8 . ... . ,
| |
| . ,. . M.. . . i. .. E .. i .. Eor.. .i,...
| |
| =5.n
| |
| ; p y
| |
| = = - y
| |
| : n. _ .r.:_qr.g.
| |
| p ,
| |
| ;m cxuse;
| |
| . . . . - . .q+., .... .
| |
| . . 3... ,
| |
| 5],h%'.N n,. _ . . _ . - . :n zummmmmes 5 i M i 8 3 !EM.8!5
| |
| . _..mmmmmmmmm . w >
| |
| 8*'*"'
| |
| I s& e e e $~5 ea _ . . . . . ...... . . . . . . . . . . - . . . . .
| |
| i I
| |
| --.+
| |
| . 310 - . . . . - .
| |
| m s.c on:
| |
| rr m m m 1
| |
| Figure 3 ECOllRA/ TRAC Vessel Component Axial Noding for UPTF Test 20 Os v
| |
| .. ) r~.
| |
| . . , ;g . , ,
| |
| 6':, ,, . . -
| |
| ~ ~ c: .
| |
| ./ ,
| |
| - GT**'e; $ i
| |
| : :o c: : : . ..
| |
| I., ' . h, , ''
| |
| '- I ;m l,
| |
| , ~ ,
| |
| ,~. , ,'
| |
| ~q ~ . ~ ~ . . . . . .
| |
| SECTION 4: CCFL END BOX REGION (1 LEVEL) chann i O oa Figure 4 ECOllRAfrRAC Vessel Component Cross Sectional Noding for UPTF Test 20 G
| |
| V
| |
| < frame /UITACRS/ Pres 4. FIG >4
| |
| | |
| t /f ,
| |
| l i
| |
| 1 es 5.3.2 Prediction vs. Data '
| |
| O 5.3.2.1 Flow rates and UP Condensation Rate Ccre Core Core Hot Leg Hot Leg UP Water Steam Liquid Liquid Vapor condensation Test 20 Injection Injection Downflow Flow Flow Rate (kg/sec) (kg/sec) (kg/sec) (kg/sec) (kg/sec) (kg/sec)
| |
| Phase A Data 22 81 298 25 44 37 WC/r 22 81 282 44 52 35 Phase B Data 19 91 298 22 51 40 WC/r 19 92 255 65 63 35 l
| |
| ^
| |
| Phase C Data 20 75 307 15 40 35 WC/r 20 77 278 27.5 49 34 UPI BE 78 (sf=2.1)
| |
| O UPI Water Injection Rate = 262 kg/sec [UPI BE 215 kg/sec (sf=2.1)]
| |
| SG Steam Injection Rate = 6.0 kg/sec UPI Wate y'
| |
| G Steam HL Liquid 1 -
| |
| 4
| |
| ~
| |
| HL Steam Upper Plenum 4-- Top of UCP 3
| |
| i i
| |
| Core Water ' Core Liq. Downflow V
| |
| Test Conditio Core Steam LJ
| |
| < frame /UPI/ACRS/ Pres 4. FIG >5
| |
| | |
| l
| |
| /9 I')
| |
| v ' 5.3.2.2 UP Collapsed Liquid level l
| |
| Range of Levels (summary of many data, see RAI-46)
| |
| Data Prediction Outer Channels 0.01 - 0.08 0.035 - 0.4 Inner Channels 0.02 - 0.2 0.03 - 0.2 I
| |
| - The predicted levels are generally higher than data.
| |
| - In the data, levels in the inner channels are higher than the outer channels. On the other hand, I predicted levels in the inner channels are higher than the outer channels, which may be the effect of the UPI injection modeled in the outer channel.
| |
| l
| |
| - The over prediction of water level is consistent with the underprediction of the liquid down-flow. <
| |
| l
| |
| ! l l
| |
| < frame /UPUACRS/ Pres 4. FIG >6
| |
| | |
| 20 1
| |
| g r II.5.4 Ranging of XYDRAG and XCONDU and Effects on Validation Tests ... [RAI-24] I 5.4.1 Ranges Determined by GE CCFL Analyses 5.4.2 Effects on CCTF 5.4.3 Effects on UPTF 5.4.4 Application to PWR 5.4.1 Ranges Determined by GE CCFL Analyses TABLE 24b-1 The GE Test Specification Wupi T f,in Subcooling Liquid Injection (Ibn/sec) (op) AT Kun 60 1.38 100. I14./ 13ypass blots Run 61 1.66 100. I14.7 Bypass Slots Run 62 2.06 100. I14.7 Bypass Slots Run 69 1.34 200. 14.7 Bypass Slots Run 73 1.38 100. I14.7 Spillover XCONDU l
| |
| A i i , i 1.0
| |
| -h_h__
| |
| i I I
| |
| ** Case l l 1 I l
| |
| 0.5 - _I ___ __ __ Figure 24b-6 Variation of the Global l
| |
| Parameters, XCONDU and XYDRAG l
| |
| l 0.2 _,
| |
| 1 1
| |
| __i___d--
| |
| I I l 0.1 0.2 0.5 1.0 3YDRAG XCONDU = XYDRAG = 1.0 to XCONDU = XYDRAG = 0.2 v
| |
| < frame /UPl/RAl/ Pres 5. FIG >l
| |
| | |
| . - . - . . . .. . . - - - .= .
| |
| 1 1/
| |
| 1 l
| |
| 1 i
| |
| V 1.5 l
| |
| A Data for steam increasing prase l A Data far steam de:reasing plase u -
| |
| O Predict on for steam increas.ng phase ,
| |
| % 1 C i1cdiuion for actm decreasing phase I s -
| |
| a to g w
| |
| [M pa%% eSE AA 4 c
| |
| 000 a 5 g A m -
| |
| 3 2 O a A A A 2 0 !" DD * .
| |
| m -
| |
| c
| |
| _,3 , , ,
| |
| 0 .02 .04 .06 .08 .lo '2 STEAM INJECTION RATE (kg/sec)
| |
| Figure 24b-1 Liquid Drain Rate vs. Steam Injection Rate, for GECCFL Test 60
| |
| ()3
| |
| 's 1.5 A Data fo steam in< reasing pt ase A Data fcr steam decreasing pbase 7 O Predict on for steam increasing phase l 1 O P. ud mi.un Tv. A o.u Jw.sming phase
| |
| ~
| |
| cn -
| |
| (a gar %D Mb a w 5 E
| |
| < a O A Otl
| |
| .c 0 O nn rde a-i m -
| |
| O
| |
| .5 '
| |
| 0 ' ' '.02' '.04 '.06 .08 . l0 .' 2 STEAM INJECTION RATE (kg/sec)
| |
| Figure 24b-23 XCONDU = XYDRAG = 0.2 for Prediction of Liquid Drain Rate vs. Steam Injee:lon Rate, for GECCFL (3
| |
| %)
| |
| Test 60
| |
| < frame /UPI/RAl/ Pres 5. FIG >2 I
| |
| 1
| |
| | |
| l Z 2- ,
| |
| l 1
| |
| 1.5
| |
| : A Data fir steam ir creasing ;hase 1 A Data far steam decreasing 3hase I T* O Prediction for steam increosing phase 1
| |
| O Predic:.on for .;a.aia dec.c. sing phase N
| |
| cn -
| |
| .M -
| |
| ~
| |
| w .5
| |
| $BOhhu'th a I
| |
| E
| |
| ~
| |
| o % a
| |
| = A A
| |
| o [ '= 8' M A M -
| |
| C hw
| |
| .5 ' ' ' ' '
| |
| ' .02 '''.M .06' ' ' .08 ' .:2 0 .0 STEAM INJECTION RATE (kg/sec) )
| |
| Figure 24b-4 Liquid Drain Rate vs. Steam Injection Rate, for GECCFL Test 69 V)
| |
| V 1.5
| |
| [ A Data for iteam inen asing phane A Data for uteam decreasing ph:se o O Predictioi for stearn increasing phase I O
| |
| $ iTedictio 1 for stean, uccreasirg phase N
| |
| ~
| |
| 9 o l w
| |
| ag aaamst a6a3a a
| |
| .5 a
| |
| +
| |
| a om a
| |
| < A N
| |
| 3 c. A AA l z
| |
| o g no ok e a O _
| |
| _,3 ,
| |
| 0 .02 .N .06 .08 .10 L2 STEAM INJECTION RATE (kg/sec) l Figure 4-5 XCONDU = XYDRAG = 0.2 for Prediction of Liquid Drain Rate vs. Steam Injection Rate, for GECCFL Test 69 '
| |
| sJ t
| |
| l
| |
| < frame /UPI/RAl/ Pres 5. FIG >4
| |
| | |
| L .
| |
| Z3 l L
| |
| L. ,
| |
| l t~
| |
| ()
| |
| (>
| |
| \
| |
| l ~ 5.4.2 Effects on CCTF !
| |
| 1 l
| |
| : 1 Table 241 Run Cases for Sensitivity Study l to The Interfacial Condensation and Drag Multipliers l i
| |
| RUN72, RUN76 XYDRAG XCONDU
| |
| ; Case 1 UP & CCFL UP & CCFL, Whole Core Case 2 UP & CCFL UP & CCFL, LP Core Channel ,
| |
| 1 Figure 24-1 PCT Predictions and Data Figure 24-2a Predicted Quench Fronts vs. Data, RUN 72 Figure 24-2b Predicted Quench Fronts vs. Data, RUN 76 Table 24-3 Comparison Results Sununary for Vessel Pressure, Void Fractions, and Hot Leg Flow Rates.
| |
| The predictions by cases 1 and 2 are similar so that case 2 is preferred, because XCONDU is !
| |
| l applied to the smaller core regions than case 1 such that the reflood heat transfer predictions in the O
| |
| hot regions are not directly affected.
| |
| (_)
| |
| Figure 24-4 to 8 RUN72 Flooding Prediction at The Top of CH-8 (LP), l Base Case, Case 1, and Case 2 1
| |
| I I
| |
| l t !
| |
| 1 (D
| |
| . L/
| |
| < frame /UPI/RAl/ Pres 5. FIG >5
| |
| | |
| , .mathcadCCTF/GLPCT2.mcad>
| |
| 8 RUN 72
| |
| .. f-
| |
| . 1600 , ; , ,
| |
| ) ,
| |
| 1500 -
| |
| ~:.:.
| |
| * Qg -
| |
| i, o i'.
| |
| I ~ '. l DAT2 ,'9. ~
| |
| o B2 g M4k Si a .,.,
| |
| DC22; 1300 -
| |
| Y' -
| |
| v.
| |
| DC2A2'. %
| |
| .g 1200 - -
| |
| 'h 1100 - -
| |
| o 0 2 4 6 8 to 12 Ellg RUN 76 O -
| |
| I# i
| |
| -.fgQ,;'%.2 I I I l' 'o o
| |
| 's, ia -
| |
| / o ',, '
| |
| j 's
| |
| . wi
| |
| 's DAT6; 1200
| |
| )
| |
| * 5 6 A \'
| |
| j 1000 - '-
| |
| DC2A6;
| |
| . ,(
| |
| 800 -
| |
| 600 - -
| |
| , I i f i 1 0 2 4 6 8 10 12 E14; O
| |
| Figure 24-1 PCT Predictions and Data
| |
| | |
| e 3V
| |
| , a)
| |
| '%J NHOUSE PROMUETARY % 3 CCTF-Quench Front RUN 72 M007A.1 (fCH-8 ^)
| |
| 25oD ~
| |
| CCTF Quench Front RUN 76 MO D 7 A [CH-8] .1( t 2 3 c o D C . 2 A ) 4 e I
| |
| \
| |
| l N % '
| |
| ''., l j
| |
| 3- 3 ao X 3 o r t. ."-
| |
| \
| |
| , , l l
| |
| ys, l
| |
| \\ i S r%
| |
| o .) #
| |
| tb , ;.!. l J, s
| |
| ..y :
| |
| ,' a s f.'
| |
| 0a Z.
| |
| > l .i y / '' .f ,p
| |
| .,, ~
| |
| g ,./ ,
| |
| ~~~ W p' \
| |
| / .de L t a=$
| |
| ,'[ t .
| |
| r r o'
| |
| y'
| |
| .a',~ -
| |
| ll -
| |
| l ?
| |
| 0
| |
| - e b
| |
| e U 100 200 300 0
| |
| -400 51 0 ti,0
| |
| " 'W8 26 0 sil0 .n0 3,,,
| |
| Figure 24-2a Predicted Quench Fronts vs. Data, RUN 72 # "
| |
| I i
| |
| t,v)
| |
| | |
| 26 warfoemouss encmaster m CCTF TEST-72 SIMULATION (t25aGLO Flooding Relations GL ( B A S E C o s e ))
| |
| O g D ot The 'OTOP of CH-8 vfg000il (LPiSC)
| |
| , 2 0 Lt0Us0 Ott$st?
| |
| %Y 3 1
| |
| 2.5 2
| |
| 7
| |
| & NN s O l.5 - ~ \ ; 3 i-o
| |
| \
| |
| N 0U
| |
| \ w N Q O O
| |
| 'g r, w
| |
| . \
| |
| w G O f 4 -- ,
| |
| a G 5 1.5 25 sqrt(-kle) 1 J 5}
| |
| 25 Figure 244 RUN72 Sees Caen. Fieseos Promesses et ne hp of CH4 (1.P)
| |
| \
| |
| 2 sx
| |
| - i.s \ c
| |
| <~ s U
| |
| ' } * \
| |
| 1 n O 0 %
| |
| 0 %
| |
| 0 C
| |
| 0 0 0 0 O O
| |
| 'I n ,
| |
| Q CN OCP "
| |
| ; o O o 0
| |
| 3 0 5 15 2' 2.$ J sqrt(-kfe) 25 Figure 244 RUN72 Case 1, Fleeshig Preectles at he Top of CH4 (1.P)
| |
| \
| |
| NN u
| |
| r ?
| |
| O I w x
| |
| g w %,
| |
| a O O
| |
| .s N y
| |
| ! 3 c 8 l
| |
| ., m
| |
| .c - ,.,
| |
| sqrt(-ki*)
| |
| ' \
| |
| e rigen 244 rum 2 Cae: 2. Fioness Preertsee se no Top of CH4 (14
| |
| | |
| <mithc d/UPTF/GLsens2> ,
| |
| l
| |
| - 5.4.3 Effects on UPTF Q 350 , , , , , , q C B A O
| |
| 300 -
| |
| o o -
| |
| I I :
| |
| 250 -
| |
| I -
| |
| Liq. Drain 200 - -
| |
| 5+'l Z
| |
| i i
| |
| b 1
| |
| a 0
| |
| FLOW RATSO - -
| |
| G (kg/sec)
| |
| M J CoreVap.
| |
| 100 - -
| |
| i 1
| |
| HL. Liq.
| |
| ]
| |
| {7 HL.Vap.
| |
| l I
| |
| 50 - [ 0 -
| |
| o II
| |
| ( tj U
| |
| 0 O 2 4 6 8 10 12
| |
| ! i l
| |
| _g Figure 24-10 UPTF Flow Rate Sensitivity to The Global Ranging
| |
| ,V END l
| |
| | |
| 37 l
| |
| i 1
| |
| () 5.4.4 Application to PWR l
| |
| According to the CCTF results, XCONDU will be applied to the LP core region and both multipliers will be applied to the CCFL and UP (see Figure 24-11). The range of eg. (24-1) will be applied in these vessel spaces. The resulting global model run ,
| |
| matrix for UPI plants with split breaks as the limiting break type is shown in Table 24-5.
| |
| I TABLE 24-5 Global Model Run Matrix for UPI Plants (Split Break)
| |
| Split Break XCONDU Case
| |
| * CD '
| |
| XYDRAG l GM1 Bounded
| |
| * 1.58 1.0 ;
| |
| 1 GM2 Bounded 2.40 1.0 GM3 Bounded 0.77 1.0 (9
| |
| u GM4 Bounded 1.58 0.2 GM5 Bounded 2.40 0.2 GM6 Bounded 0.77 0.2 Determined on a plant specific basis by ranging CD in intervals of 0.1 with KN = 1.58, and XCONDU = XYDRAG = 1.0 I
| |
| '~h (G
| |
| , < frame /UPI/RAI/ Pres 5. FIG >9 l
| |
| | |
| ~
| |
| M/
| |
| Appendix A - Point Beach Best Estimate LOCA Model
| |
| -1 Point Beach }.VCOBRA/ TRAC Vessel Noding Diagram
| |
| .4, a i] Secson ,9 as see-e i , .00 = ** *.* .
| |
| * A
| |
| ........g.,g... . ... '
| |
| A f .'. 8R .
| |
| 3 3 8 8 3 5*3=a, 37 b m n_ m es. .
| |
| m ;
| |
| .,Q ' 6[ ; R .qr.
| |
| E:'s .a ,. 3. . E .8..\ '-. s:.e...:.. .. .". , E' 6.;5...:59 ' s==. a..r su . .._ 8. . :.5 E..?_ . . . , ...
| |
| I!d g: ,
| |
| did pA u .- ~
| |
| o g._ .
| |
| a, z .., . . ..;.
| |
| / .
| |
| gg. &g=
| |
| , . . G. . . ...... ... .
| |
| . . . . .g>, ... g. g...
| |
| J , , , ,
| |
| g; .,
| |
| g, .
| |
| _}. .
| |
| ...0. . . . . . . . . . . . . . . . . . . . . ..
| |
| . . . . .#./.. . #. ..,..: . . . ..,E. .. . . o. . . . . . . '
| |
| . . g..: .
| |
| , ,, m _
| |
| s.s .... .
| |
| 2 'ig.
| |
| =as s=.si n. i
| |
| : n. m _ . . . . . . . . ;.. + . . ... .. . .s . . . .... .:.... ....:.....
| |
| 7 n 'un .
| |
| . fs) 43
| |
| -..@......su wi O n. m _
| |
| E n o hm
| |
| . .M. . .. .
| |
| ..(t
| |
| . .o. @ ~31 T, .;
| |
| S Q4 A. .x@ 'E:
| |
| a.2
| |
| --s c a. . .; . . .
| |
| . . . . . . y ..
| |
| ,27 . 7 == .... . ........ ........i............ ........ .. ....
| |
| ,2,0,0 = ****. . ..... .. .......I..'..........
| |
| ....i. ........... . ... .... ........1...... ... ...... . . .. ..
| |
| m,n . . .. .G. . . .
| |
| 4., ,n .....,..... ........
| |
| +
| |
| . r. m . .... .. ...,.....- p. : -
| |
| s..o.c, noion.s,
| |
| ...u_ ,..,. .... ... . . . . . . . . . . .... . .. . .. . .
| |
| ...8....... E;
| |
| .o _ .O. . ..: E. . .E.. . 3. . . .
| |
| : 3. . .. . .8........ ...5....... 8. . ..E. . .: E. ..
| |
| . . . LP onc
| |
| . . . . . . . . . .n.i . g . . . .o.f. .
| |
| * .0 3 6.= . . . . . . . .. . . . . . . .. . .LP..... .
| |
| . ,, m .d.:.d... .. . ... . . ...I........I..... . ..... . . .. . !.I. .. .M..
| |
| , , , , , , . _ ..... ........ .. ..:.4 . ...... ......... . . . . . .
| |
| . .u_ ...... , .... ........... . . . . . . .
| |
| ,,,=
| |
| II. . .. . . n . . ca . . . n , n . .. n . . .[-] .... n . . n - G...... .GII seenn2.
| |
| E 'I; 3;G 1
| |
| qf M. i I G I d. !Gf J-O Se.c, e..n..,1 4'
| |
| ' 0@
| |
| g *"' .
| |
| wtp e -
| |
| O Figure 24-11 The Regions of XCONDU and XYDRAG Application
| |
| | |
| yo WESTINGHOUSE PROPRIETARY CLASS 2 1 II.5.5 Prediction of Scaling Trend
| |
| -(J
| |
| , According to the cited reference, the scale effects of the UPI test analy:;es will be performed on the quantities of:
| |
| (1) Breakthrough Flow Area as Percentage of Core Flow Area (2) Downtlow Rate as Percentage of Available Water (UPI Injection, Core Water Inj. Plus Steam Condensation)
| |
| (3) Net Hot Leg Water Carryover Rate as Percentage of Available Water (UPI Injection, Core Water Inj. Plus Steam Condensation)
| |
| (4)_ Collapsed Liquid Level / Height to Hot Leg Centerline and the scale is made on the basis of the core flow area from Table 37-1.
| |
| TABLE 37-1 Scaling of Tests Based on Core Flow Areas Facility Core FA Scale Point Beach 3924 in 2 1.0 UPTF 8527 in 2 2.2 CCTF- 343.0 in2 0.087 (1/11.4-th)
| |
| O V GE 15.53 in 2 4.0 E-3 (1/253-th)
| |
| Scaling effects appeared in the predictions of UI'TF, CCTF, and Point Beach analyses will be obtained in this order. GE tests are excluded from the scaling study, because it is so specialized to the single effect.
| |
| i v-
| |
| < frame /UPI/ACRS/ Pres 6. FIG >l
| |
| | |
| l F/ l l
| |
| l WESTINGHOUSE PROPRIETARY CLASS 2 I.
| |
| 7 ,.,
| |
| A i O uere e oi rFprediction g 100 3 PWR prediction
| |
| !! CCFF y CCTFprediction g
| |
| e SCTF $ ORNL 3 Module 80 ORNL 1 Module f O Q X Dartmouth O Semiscale y o
| |
| @ 60
| |
| < X f k co y J 40 fo 3O
| |
| :: 8 E 20 e
| |
| ca5 0
| |
| .001 0.01 0.1 1.0 10.0 Scale Factor Figure 37-10a Scaling Predictions and test Data for Breakthrough Area r%
| |
| ( ,)
| |
| 100 1 1 h 80
| |
| % J
| |
| $ V 1 1 2 60 o cc O uirrF e uirrFprediction MN 40 g PWR prediction d .,0 y CCTF prediction V CCTF
| |
| & SCTF $ ORNL 3 Module c: .g y4 20 o ORNL 1 Module X Dartmouth D Semiscale g% @
| |
| X 6 6 0
| |
| 001 0.01 0.1 1.0 10.0 Scale factor Figure 37-10b Scaling Predictions and Test Data for Downflow into Core f%
| |
| .b
| |
| < frame /UPl/ACRS/ Pres 6. FIG >6
| |
| | |
| 35 WESTINGHOUSE PROPRIETARY CLASS 2 l
| |
| ^
| |
| 100 + 0 UI'rF e uf'rFprediction y g PWR prediction y y CCTF y CCTFprediction i u g 80 LX SCrF Q ORNL 3 Module )
| |
| By Q ORNL 1 Module O% A X DJ.rtmouth O Semiscale j 60 b
| |
| E E 40 g - 3
| |
| .t 'ir 3 it S ir e E 5 20 '
| |
| x8 tU
| |
| } U I
| |
| ZA 0 l
| |
| .001 0.01 0.I 1.0 10.0 Scale Factor l Figure 37-10e l Scaling predictior.s and TcSt Data for Hot Leg Water Carryover V
| |
| 100 O uvrF e unrrFprediction M PWR prediction y CCTF y CCTFprediction o A ScrF $ ORNL 3 Module g 80 Q ORNL 1 Module
| |
| -o X Dartmouth O Semiscale Mc 3 0 60 O 2@
| |
| EC ''
| |
| D y 40 ::
| |
| 'I
| |
| ?s E.
| |
| s.3
| |
| * 20
| |
| [
| |
| Oo u% 0 n 0
| |
| .001 0.01 0.1 1.0 10.0 Scale Factor Figure 37-10d Scaling Predictions and Test Data for Liquid Level in Upper Plenum o
| |
| b
| |
| < frame /UPI/ACRS/ Pres 6. FIG >7
| |
| | |
| 33 5.6 Range of Conditions in The Assessment Matrix vs. UPI Plant O
| |
| Table 111 (Revised Table 3 4) .
| |
| Range of Test Conditions Parameters Range CCTF 72,76" UITF GE CCFL UPI Plant (Table 3-4, Test 20" Test" (Reflood)
| |
| Reflood)
| |
| Pressure (psia) 20-70 35-65 38-41 14.7-22.7 20-50 Initial PCT ('F) 1300 1600 950 1450 NA NA 750-1700 Maximum PCT ('F) 972 2184 1100-1600 NA NA 1400-1800 Initial PLHR (kw/f) 0.67-0.95 1.40 NA NA 0.1-0.5 Rod Bundle Array 15x15,17x17 Similar to NA Slightly Similar to Sizes / Types w/wo mixing 15x15 lattice larger than 15x15 lattice vane grids w/o mixing 15x15 vane grids lattice Axial Power Shapes Chopped Chopped NA NA various (peaking factors) Cosine Cosine 1.66-1.19 1.91 NA NA Top Skewed 1.35 Bottom Skewed 1.45 Spacer Grid Types Simple and Simptp Grids NA NA Mixing Vane
| |
| , Mixing Grids UPI Water Temp. (*F) 97 86 5S 180 32-120 Subcooling at UCP 10-40 40 0-126 60-140~
| |
| (*F)
| |
| Core Steam Flows NA- (165-201) (0.08-0.22) 50-80 (Ibm /s) +2.2 *253
| |
| =75-91 =20-56
| |
| : Entrained Flows out (0-44)* 11= (33-55) NA 10-40 Hot Legs (Ibm /s) 0-484 +2.2
| |
| =15-25 UP Liquid Levels 38-48 % 0-11 % NA 12-40 %
| |
| (% of UP Height)
| |
| Core Liquid NA (657-677) (0-1.4)*253 100-400 Downflow +2.2 =0-354 g (Ibm /s) ,299 308 O
| |
| i 11-4
| |
| | |
| . . ~ .- _ -.. - .. . . .- .. - - . . - - . .. . . - - --~ . - .
| |
| WESTINGHOUSE PROPRIETARY CLASS 2 O
| |
| v Summary
| |
| : 1. Thus, we can conclude that the test conditions of UPTF, CCTF, and GE CCFL are qualified to be applied to assessment of UPI PWR.
| |
| : 2. Prediction capability of scaling trend by WCOBRAfrRAC brings us a confi-dence that the good prediction of each test is a valid assessment for the UPI PWR.
| |
| : 3. Prediction of PCT tends to be slightly conservative.
| |
| : 4. Hydraulic status in the upper plenum is well predicted.
| |
| : 5. Saturated and subcooled CCFL at the UCP can be reasonably well predicted by ranging the interfacial drag and condensation multipliers, XYDRAG and XCONDU.
| |
| : 6. Finally, the global model parameters and run matrix were established for BE LOCA of UPI PWR.
| |
| O
| |
| < frame /UPl/ACRS/ Pres 6. FIG >8
| |
| | |
| L. !
| |
| !O II.4 Details of UPI Code Physics 4.1 Introduction to WCOBRA/ TRAC 4.2 PWR UPI Dynamics l Sequence of Events UPI Waterin Upper Plenum CCFL at UCP UPI Water in The Core 4.3 . Physical Models of Interfacial Drag, Entrainment, and De-Entrainment Flow Regimes and Interfacial Areas l
| |
| Interfacial Shear Model l Entrainment and De-Entrainment i 4.4 Physical Models ofInterfacial Condensation O Interfaciai vapor Condensation Example CCFL (GE CCFL)
| |
| Kenji Takeuchi Westinghouse Electric Co.
| |
| (412) 374-4263 Advisory Commitee on Reactor Safeguards Thermal-Hydraulic Phenomena Subcommittee Meeting l Dec.16,1998 l
| |
| < frame /UPI/ACRS/ Pres 1>l
| |
| | |
| 2 l
| |
| 1 O WCOBRA/ TRAC MOD 7A, Rev.1 i
| |
| TRAC-PD2: 1-D, Drift Flux Model,5 Consevation Eqs. for Loop Model COBRA-TF: 3-D,2 Fluid 3 Field Model,8 Conservation Eqs.
| |
| where 3 Fields (vapor, continuous liquid, and entrained liquid) for Vessel Model UPI Specific Assessment .... COBRA-TF CCFL interfacial condensation interfacial drag entrainment / de-entrainment l
| |
| O Suschannei Formuiation Channels Gaps Numerics: Semi-Implicit, Staggered, Donor-Cell Method Cell Center --- scalars [p, 0:, p, ....]
| |
| Cell Boundary --- vectors [ v, gav, ...]
| |
| .O
| |
| < frame /UPI/ACRS/ Pres 1>2
| |
| | |
| 3 l
| |
| J O Code Assessment for UPI l
| |
| CCTF Run 72 and 76 (full UPI and Single Failure)
| |
| UPTF Test 20, Phase A, B, and C (UP Liq. Drain, UP Condensation, and HL carryover)
| |
| GE CCFL Tests (4 Subcooled CCFL and 1 Saturated CCFL)
| |
| Interpretation of Dartrnouth Test for UPI Jet Assessment of Scaling Trends, MPR (2D/3D Program)
| |
| The results are good 12 conservative.
| |
| l 1
| |
| O Table or The Next Subjects:
| |
| 4.2.1 Sequence of Events 4.2.2 UPI Water in Upper Plenum 4.2.3 CCFL at UCP 4.2.4 UPI Water in The Core 4.3 Physical Models of Interfacial Drag, Entrainment, and De-Entrainment 4.4 Physical Models of Interfacial Condensation i
| |
| I
| |
| >vn l
| |
| < frame /UPl/ACRS/Presl>3
| |
| | |
| . 1 4- l g
| |
| -v 4.2.1 Sequence of Events (Typical LOCA for UPI Plants) 0 - 20 see Achieve steady state conditions 20 see LOCA Transient Started 30 sec. Hot Rod Blowdown Peak 33 sec. ACC & HHSIInjection Started 50 sec. End of Blowdown Cooling End of ECC Bypass UPIInjection Started 67 sec. End of Refill 70 sec. Hot Rod Reflood Peak PCT 75 sec. ACC Emptied l Bottom 1/3 of LP Core Region Filled !
| |
| O 110 sec. Core Average Rod Quenched
| |
| (.') 210 sec. Hot Rod Quenched l
| |
| i
| |
| - LP Rod remained Tsatduring entire transient.
| |
| Summary Report Table 2-1 i
| |
| j i
| |
| O
| |
| < frame /UPI/ACRS/Presl>4
| |
| | |
| - .- - . - . - ... . ~. . . . . . .
| |
| f b WEP-UPI Spilt Break BEST ESTIMATE LBLOCA CD=0.6. KN=1 58. XC=1.0 Limi ting HR PCT f o r N o[d i n g Sensitivity RAl-18) (ps10)
| |
| PCT- 1 0 0 PEAK CLADDING TEMP.
| |
| ---. PCT 3 0 0 PEAK CLA0 DING TEMP.
| |
| ---PCT 4 0 0 PEAK CLADDING TEMP.
| |
| ---PCT 5 0 0 PEAK CLADDING TEMP.
| |
| - - -- TSAT 12 13 0' SATURATION TEMP.
| |
| 1800 1600
| |
| ; IU -Hot Rod n
| |
| u_ 14 00
| |
| -v i / OH/SC Rod o : 11
| |
| ~
| |
| " MR"A
| |
| [ 800 h
| |
| : c. :
| |
| ' n il, /
| |
| N.
| |
| 'd h 600
| |
| ~
| |
| ' ~ \
| |
| s -
| |
| fd\ \,/ "i ec LPF od
| |
| : '" N \
| |
| 400 \ ' '
| |
| ^
| |
| 5% ,
| |
| 3'L h- ''''
| |
| 200 -
| |
| O 50 100. 100 200 250 300 Iime (S)
| |
| Figure 2-1 PCTs of The Hot Rod, The Average Rods, and The LP Rod I
| |
| O
| |
| < frame /UPl/AC).S/Presl>5
| |
| | |
| [
| |
| - j l
| |
| Y WEP-UPI BEST ESTIMATE LBLOCA Limiting Split Break [CD=0 6. KN=1.58. XC=1 0] (ps10)
| |
| Collapsed Liquid level in Core LP LQ-LEVEL 5 0 0 COLLAPSED LIQ. LEVEL
| |
| - - - - L O-L E V E L 4 0 0 COLLAPSED L10 LEVEL
| |
| - -- L Q-L E V E L 3 0 0 COLLAPSED LIO. LEVEL
| |
| - - - L Q-L E V E L 2 0 0 COLLAPSED LIO. LEVEL 12 C
| |
| v 10 gp j 8
| |
| L fl[ I' ' I j - 1 lll yn' '% ll'u yt lji l l
| |
| l'il h I g , ,ll.
| |
| i l E6
| |
| , N2 Effect g I l [ . f i l 1 '"GT & OH/FM'SC j
| |
| ~
| |
| I j i I (
| |
| 10), O
| |
| ~
| |
| a l N .n L I l ,'u Is.
| |
| y 4
| |
| l' ''
| |
| C, -
| |
| l , l "YW yTpf lq] gyN o_ -
| |
| l lg lil
| |
| # ~
| |
| _, 2
| |
| \ In i j -
| |
| Il l N HA C -
| |
| r f 1
| |
| ' "D, "''' '''' ' ' ' ' ' ''' ''''
| |
| 0 0 50 88 100 150 200 250 300 h Time (s)
| |
| ACC Emptied UPI End of Refill Figure 2 2 Collapsed Liquid Levels in Core Channels l
| |
| f n
| |
| i
| |
| < frame /UPl/ACRS/Presl>6
| |
| | |
| 7 O 4.2.2- uri water in upper Plenum HL I
| |
| ; I
| |
| ,. ' s
| |
| ' N j
| |
| / ("% \
| |
| I \ SC i._, 7 L_. _.
| |
| I IGT I
| |
| ]\ s N
| |
| s
| |
| '_J,/
| |
| % ' N i
| |
| I HL I
| |
| PI O .
| |
| i j
| |
| Netto UPI
| |
| .- '- Opposite to UH
| |
| ' \
| |
| 'o USc\ Sh z
| |
| __q s , j ,. _ , L ._ ____
| |
| \ '"
| |
| 4 f 5s i
| |
| I Below UP1 I
| |
| Inner Global
| |
| .A U
| |
| < frame /UPl/ACRS/ Pres 1>8
| |
| | |
| 4 7
| |
| Figures 18-3 & 18-4 YOlD FRACTIOR in Outer CL below UPI
| |
| ~
| |
| AL 35 2 0 VAPOR FRACTION
| |
| ----AL 35 3 0 VAPOR FRACT10N
| |
| - -- AL 35 4 0 VAPOR FRACTION '
| |
| -------AL 35 5 0 VAPOR FRACTION
| |
| - - - AL 55 2 0 VAPOR FRACTION I u- T i
| |
| i
| |
| ) /#i, il
| |
| _I I',l I5 i ',,'i' ;'#') '5 ,
| |
| U Z lll il fs ',)
| |
| I igi '
| |
| I' I '{elllflI ll
| |
| ^ ' '
| |
| 6
| |
| ! h,,i o-b! hljfi,!:, jlh
| |
| ' yo ri'.
| |
| 5
| |
| '? : '
| |
| '' g{
| |
| o 1 4 ' ' ''i '
| |
| l.,,
| |
| '[
| |
| 8
| |
| , 3([vtP {p p' ,' "f.,.,,,
| |
| l ,':, . .,',[.,
| |
| 1 , 1ll ,' fifls jN , i 20 40 60
| |
| , , eccemroe3 80 100 120 JANN( 160 100 fm Time (s)
| |
| Q V01D FRACTION in Outer GL next to UPI AL 3< 2 0 VAPOR FRACTION
| |
| ----AL 34 3 0 VAPOR FRACTION
| |
| - -- AL 34 4 0 VAPOR FRACTION
| |
| ------- AL 34 5 0 VAPOR FRACTION
| |
| ---At 54 2 0 VAPOR FRAcil0N I ' e Z
| |
| _l j , y '
| |
| si j il i. s l
| |
| * Q ' , .
| |
| i p
| |
| ','' {e lj '
| |
| lj l ! ei i n ' l ' 'i i qfi];,plf ll
| |
| ..I k if I. , .' , ,, l *M , !.
| |
| 1i,f L ll i 'iii ' , 1'' i, , i,i =
| |
| ' I '
| |
| e4 3 S p ;; ; <
| |
| / . ; ,p'.' ai" .iI 'i,' w o l' ' / i i I 'h ' ,' ,
| |
| ti ,
| |
| h k-l') 9% ! !
| |
| f5,'f ?$9 ,$
| |
| [V 1 t 3l't !
| |
| jl0l jl i
| |
| [
| |
| 8
| |
| ,9 91 # %
| |
| ' ' ' ' ' ' '' I 0 2 20 40 60 1U0 120 inD 100
| |
| ! ) I ,80 V
| |
| lme (S)
| |
| < frame /UPl/ACRS/Presl. FIG >l
| |
| | |
| \
| |
| I. f Figures 18-5 & 18-6 p)
| |
| ( VOID FRACTION in Outer OL opposite to UPI AL 33 2 0 VAPOR FRACTION
| |
| ----AL 33 3 0 VAPOR FRACT'ON f - -- AL 33 4 0 VAPOR FRACTION l -------AL 33 5 0 VAPOR FRACTION
| |
| ---At 53 2 0 VAPOR FRAC 110N 1
| |
| c' o.,
| |
| e zi n,. :g p
| |
| 'q
| |
| : ': a.. s%.i : s n.u cu j!j iy' g-l
| |
| % '63 l"k: , $ y' "
| |
| m C
| |
| ij (. ' ]h. '.'; . ;A.,0 '" [/
| |
| s ,.
| |
| k
| |
| ,,,, . y' -
| |
| pi I iii i t . , t. ;
| |
| m in i n ..i... A /. L ., 'e <
| |
| O '4 ,. 'i .fi, i A I 4! t' ii \gi,6 /,gI,-'('. ' i''i.[;
| |
| 4' i id n -
| |
| o
| |
| " ) l ,s , s , ,. , .
| |
| , il C n! <! L t ''i" ltk".p, !'
| |
| N M
| |
| f Ull;,k )$i'y\l ,
| |
| 0 20 40 60 80 100 120 id 0 100
| |
| (" . Time (s)
| |
| ( . VolD FRACll0N in Inner GL AL 44 2 0 VAPOR FRACTION
| |
| ----it -
| |
| ------- AL ti 44 i
| |
| 5 "i m '"ti?l8" 0 VAPOR FRACTION
| |
| ---AL 56 2 0 VAPOR FRACTION 1
| |
| ' I [h
| |
| _~ l , h ,
| |
| IiI
| |
| 'd .l''l U.b. hi),! $, ldd
| |
| , ~ / ['i l. ' ' .'' i l rW 6
| |
| WF MWF m
| |
| pi j/'i. i,' A(
| |
| s
| |
| @UCF '
| |
| c '
| |
| l qij lw''lt 11 e
| |
| 'y a
| |
| e c4,,,,
| |
| ,i '.
| |
| yl 7
| |
| l I
| |
| f j i B
| |
| f f i '
| |
| ]'
| |
| { UP; 6 l
| |
| 0 20 40 60 80 1U0 120 140 100 l (j I,ime (S) l
| |
| < frame /UPI/ACRS/ Prest. FIG >2
| |
| | |
| /0 Fi brures 18-7 & 18-8 O
| |
| Q L I QU I D F L 01lf R A T E A T T H E O u t e r OH WTH00022 39 1 0 LIQ AX1AL WASS FLOW
| |
| ~
| |
| --- WTH00023 38 1 0 LIQ AX1AL WASS FLOW
| |
| --- W T H 0 0 0 2 4 37 1 0 LIQ AXtAL MASS FLOW m 2000 j
| |
| j
| |
| ~
| |
| _ cour te{cunent dou nflow v '- "-"
| |
| 0 '
| |
| w _
| |
| 'N_. %. Q'' N l s
| |
| N ' g'N 1 l
| |
| -2000 \_ s Q site to UPI l o -
| |
| N y
| |
| w
| |
| -4000 g *w '
| |
| E -sNext to UPI w
| |
| 2 -
| |
| UP.: s
| |
| -6000 '
| |
| o o -
| |
| elo'v UPI
| |
| , -8000 g
| |
| \
| |
| c _
| |
| - -10000 ' ' ' ' '' ' '' ' ' ' ''' ' ' ' ' '
| |
| )
| |
| 20 40 60 80 100 120 140 100 I
| |
| [ , .
| |
| Time (S)
| |
| , VAPOR FLOW RATE AT Outer OH ,
| |
| WTH00028 39 1 0 VAP AXIAL WASS FLOW
| |
| --- WTH00029 38 1 0 VAP AXlAL WASS FLOW
| |
| . --- W T H 0 0 0 3 0 37 1 0 VAP AXtAL WASS FLOW m 160 ,
| |
| E
| |
| .o
| |
| - 140
| |
| : u aflo"' __ .- 1ppcsite to UPI w 120
| |
| ~
| |
| '~
| |
| '~'_ ,
| |
| + -
| |
| s 5 ,, -A r
| |
| / Below UPI
| |
| =
| |
| o 3
| |
| \%,w "
| |
| j 80 u'. 3 f
| |
| w -
| |
| m 60 ~'~_
| |
| =
| |
| : I\ .- Next to UPI l \ -
| |
| s 4,
| |
| o e
| |
| l l
| |
| N , -f
| |
| % 20 - j w Jl l-e 0 c -
| |
| -20 ' ' ' ''' '''
| |
| 20 60 80 100 O
| |
| 40 120 140 100 Time (S)
| |
| < frame /UPUACRS/ Prest. FIG >3
| |
| | |
| H .
| |
| Figures 18-9 & 18-10
| |
| ~
| |
| LIQUID FLOW RATE AT THE Outer FW/SC WTH00034 43 1 0 LIQ AXIAL WASS FLOW
| |
| --- WT H00035 42 1 0 LIQ AXIAL WASS FLOW
| |
| --- W T H 0 0 0 3 6 41 1 0 LIO AXtAL MASS Flow m 1600 E
| |
| .o _
| |
| C 1400 -
| |
| w _
| |
| * 1200 l
| |
| < - cc -current '
| |
| * ~
| |
| uaflow
| |
| * 1000
| |
| ~
| |
| 3 :
| |
| 800 Oppos,te to Ul S-h,d m coLlDLer' ~ ,-
| |
| m -
| |
| current p/..'
| |
| y goo -
| |
| downflow Next to UPI g', / g.
| |
| - , - ~ y y 400 I -
| |
| u --- -
| |
| ,_s /
| |
| - -- --7 '-~'TE: low UF I 7 200 i c
| |
| 0
| |
| \r ''' ''' ' '' ''' ' ''
| |
| 20 40 60 80 100 1;!D 140 100 i
| |
| ~T Time (s)
| |
| (d i
| |
| VAPOR FLOW RATE AT THE Outer FW/3C WTH00040 43 1 0 VAP AXIAL WASS FLOW
| |
| - - - W T H 0 0 0 41 42 1 0 VAP AXIAL WASS FLOW
| |
| --- W T H 0 0 0 4 2 41 1 0 VAP AX1AL MASS FLOW m 500 E
| |
| .o -
| |
| v
| |
| /
| |
| w 400 w -
| |
| /s s :
| |
| Opposi I /
| |
| w -
| |
| o 300 e d -
| |
| upilow [/
| |
| m w ' ,'
| |
| /a,'N Below UPI
| |
| < 200 N s #'' Next to UPI I
| |
| m
| |
| ~ l n_
| |
| .~-_/ '/ /
| |
| C 100 _I . _ _ .
| |
| . :I c I
| |
| - ri i i i , , .,, , , ,,, , , , ,,
| |
| o
| |
| ' 1;l0 20 40 60 80 100 140 100 lJ Time (s)
| |
| < frame /UPl/ACRS/Presl. FIG >4 i
| |
| l l
| |
| : l. ,
| |
| 12 l l
| |
| r %
| |
| O ,
| |
| i i
| |
| I L imi WEPt i n g S(p Broo l i TUP BEST I) ESTIMATE LBLOCA i k [CD=0.6. KN=1 58. XC=1 0) (ps10)
| |
| Subcooling in Outer GL Below UPl TL 35 2 0 L10010 TEMPERATURE
| |
| ----TL 34 2 0 LIQUID TEMPERATURE L - --
| |
| TL 33 2 0 L10VID TEMPERATURE
| |
| ------- T S A T 35 2 0 SATURAT10N TEMP.
| |
| 1 500
| |
| - l
| |
| ~
| |
| l i
| |
| ^ 400 ,
| |
| L- _ l v
| |
| W _
| |
| a:: ~
| |
| V l r H 300 l ~.-., ,,,.
| |
| .W',''* .
| |
| w -
| |
| g i
| |
| .,s----
| |
| , o_ _ ',, , s - - -- . . ,
| |
| JE W
| |
| _ Opposite to UF h,
| |
| -s n ,y !g
| |
| . ' ,, n /C '~~~ ' ' ' / ' '' '
| |
| , . ,,, e es t i-- 2 0 0 s
| |
| -. n a_\
| |
| c
| |
| .u
| |
| -v Rext to l'PI \, w .y ,
| |
| s -
| |
| M(Below UP I ' '' 2' ' ' ' '
| |
| 100 20 40 60 80 100 120 140 160 i
| |
| Time (s)
| |
| Figure 18-11 i
| |
| l
| |
| ! /~h
| |
| ! O i
| |
| l l
| |
| < frame /UPUACRS/ Prest.FlG>5
| |
| | |
| - . . . . . . . . - - = . _ . _ _ - - - . - . - . . - . . . - . - . . - _ . - . .
| |
| i.. ,
| |
| 13 i
| |
| O Points ofInterest
| |
| : 1. UPI injected at a corner of UP leads to relatively uniform pooling at UCP.
| |
| : 2. - Circumferential downflow distribution reflects the liquid temperature distribution.
| |
| : 3. Liquid downflow into the core is controlled by subcooled CCFL at Outer OH.
| |
| i l
| |
| O l
| |
| i i O
| |
| <frameNPI/ACRS/Presl>9 l
| |
| | |
| l' If l-The subcooling effect is estimated, assuming the system pressure of 37 psia. The UPI specifica-tion is:
| |
| Wupi = 226. Ib,n/sec, and T f,in = 76 F. I AT = T .t - Trin = 186.2 F The condensation efficiencyf= 0.24 (Bankoff et al.,1981).
| |
| 2 Aucp = 2.668 ft , excluding the areas of the inner channels, which do not contribute to the flooding.
| |
| (k[- 3.742)'' + (k)) ' = 2 (3-1) l.
| |
| 4
| |
| ~
| |
| 1
| |
| ]' l
| |
| ~
| |
| Subcooled CCFL
| |
| ) - - Saturated CCFL 3 O O O Prediction O 1 k }
| |
| 2-A h N I
| |
| \ E g D l
| |
| f Ng cb o O 00 3
| |
| n A , m_.
| |
| O cl O 1 CD
| |
| \f i o.
| |
| hD U
| |
| . 0 g
| |
| O
| |
| - - - - ~ ' 'N O
| |
| E g- - - ~ --to" -
| |
| ; O 1 2 3 4 b
| |
| g Figure 3 2 CCFL Prediction vs. BankofT Sat.and Suc. CCFL Correlations l Top of Channel 22 (OH in CCFL Region) l
| |
| < frame /UPI/ACRS/ Pres 1>10
| |
| | |
| IV l 1
| |
| l 1
| |
| v 4.2.4 UPI Water in The Core [RAI-9] l 50 - 75 sec. 75 - 100 sec.
| |
| 3500 lbm 4700. Ibm a6 10- A 6s.
| |
| __ ,450.
| |
| ;U < 117.
| |
| _4J60.
| |
| I
| |
| " 65.
| |
| LP Channel LP Channel _M70.
| |
| l -*@- l AM1570. Ibm >24. 1 80.
| |
| I z 970. t
| |
| : 3. - - - -
| |
| g59,
| |
| (. / /
| |
| V , #.L ,' ..
| |
| V 130. V 1000. 600.
| |
| - -c> Steam Flow to Av. Channel
| |
| > Liquid Flow to Av. Channel Figure 9-20 LP Channel and Time Integrated Mass Flows
| |
| 'i n.b.
| |
| 67 sec... End of Refill 75 sec.. ACC Emptied l
| |
| l O
| |
| l
| |
| ) < frame /UPI/ACRS/Presl>l1
| |
| | |
| . _ . - ~ . . . . - . - . - - . - - - . . . . . - - - . - . . . _ . - . . - . . . - . - - _ . - . - . . _ . . - . . ._.-
| |
| lb O 75 - 100 sec 50 - 75 sec
| |
| / /
| |
| -C>2 Wtually No Cross Flow /
| |
| -H>5 I - -c> Steam Flow p
| |
| O Liquid Flow "O 100 ;
| |
| / i
| |
| . HA
| |
| !- 1 l Figure 2-5 Cross Flow Pattern at HA Channel )
| |
| t i
| |
| See RAI-9 for more information
| |
| .O l
| |
| l 1
| |
| i iO l
| |
| < frame /UPI/ACRS/Presl>l2
| |
| | |
| I s l l /7 l
| |
| O A, 4 A Top of Core l ,
| |
| ' ' j
| |
| ; l l
| |
| 8 U
| |
| % ~ N A
| |
| P s A
| |
| s _ -
| |
| I Core Channels i Avg Avg l CH HA CH l
| |
| LP LP l I
| |
| w )
| |
| 8ottom er core O u u ,
| |
| O Liquid Flow
| |
| ----+ Steam Flow Figure 2-3 Chimney Effects Computed with _WCOBRAffRAC l
| |
| l lO l
| |
| < frame /UPI/ACRS/Presl>13
| |
| | |
| .i WEP UP I) ESTIMATE LBLOCA Limiting S lit Brea kBEST [CD=0 6. KN=1.58. XC=1.0] (ps10)
| |
| TOT L LIQ. FLOW RATE AT THE HA Core TOP MTH00003 15 13 0 ENT AX1AL MASS FLOW 30 O 20 m
| |
| {E -
| |
| UPI
| |
| # p1 10 v _
| |
| o A- " i N h I . W Y $$
| |
| x : A s= -
| |
| O a -10
| |
| !C en HA Quenc bed g _
| |
| < -20
| |
| == -
| |
| -30 ' ' ' ' ' ''' ' ' ' ' ' ' ''
| |
| 0 50 100 150 200 250 Time (s)
| |
| : I j'll ' '
| |
| : l
| |
| : i .-
| |
| (
| |
| A LN W 11 i
| |
| )N I
| |
| 0 50 100 150 200 250 Time (s)
| |
| , Figure 45f-1. Total Liquid Flowrate at The Top of HA V
| |
| < frame /UPVACRS/ Pres 1>14
| |
| | |
| _. .. . . . . __ . - ~ . . ._ - .-- -- . _ - . . . . . - . .
| |
| /7 \
| |
| O Conclusions So Far i
| |
| i
| |
| )
| |
| : 1. Blowdown through the End of ECC Bypass is not affected by UPI.
| |
| : 2. A Pool of UPI water is formed at the UCP.
| |
| : 3. Liquid Downflow is controlled by Subcooled CCFL at UCP.
| |
| : 4. Liquid Downflow takes place preferentially into low power core channel.
| |
| : 5. HA and HR Flow Conditions in the Top 2/3 Elevation are the same as 3-/4-loop plants.
| |
| (Many Issues ate covered by 3-/4- loop licensing) l i
| |
| 1
| |
| (
| |
| O 1 l
| |
| l
| |
| < frame /UPl/ACRS/ Pres 1>15 l L
| |
| | |
| Pf) l p Major Contributors to The Uncertainty of BE LOCA Analyses:
| |
| ! x)
| |
| : 1. GLOBAL Class 1.1- Plant Initial Fluid Conditions (Group t)
| |
| UPI 3/4 Loop TAVG TAVG RCS Fluid Average Temperature PRCS PRCS RCS Pressure TACC TACC Accumulator Fluid Temperature PACC PACC Accumulator Pressure VACC VACC Accumulator Liquid Volume TSI TSI Safety Injection Temperature
| |
| .KACC KACC Accumulator Line Resistance 1.2 Plant Initial Core Power Distribution (Group ot)
| |
| UPI 3/4 Loop FQ FQ Hot Rod Total Peaking Factor FAH FAH Hot Rod Enthalpy Rise Peaking Factor PBOT PBOT Average Axial Power in The Lower Third of The Core PMID PMID Average Axial Power in The Middle Third of The Core p
| |
| J 1.3 GLOBAL Models (Group )
| |
| UPI 3/4 Loop na CD Break Flow Model Bias (UPI split BK)
| |
| KN KN Broken Loop Nozzle Loss Coefficient na XC Condensation Model Bias [DC and LP] (UPIi25 F)
| |
| GBk SP Break Type (Split or Guillotine)
| |
| XCONDU Int. Condensation in Low Power Core, CCFL, & UP XYDRAG Interfacial Drag in CCFL and UP
| |
| : 2. Hot Spot Local Effects (Group y)
| |
| UPI 3/4 Loop FQH FQH Local Hot Spot Peaking Factor KF KF Fuel Conductivity KB KB After Burst Fuel Conductivity HTCG HTCG Gap HTC EH EH Local HTC TMIN TMIN Tmin (Film Boiling --> Nucleate Boiling)
| |
| [ RF RF Afetr Burst Fuel Density RX RX Cladding Reaction Rate l PG PG Rod Internal Pressure h
| |
| TB TB Burst Temperature i
| |
| < frame /UPl/ACRS/ Pres 1>l6
| |
| | |
| 2) l e- BS BS Burst Strain l U's 4.3 Physical Models of Interfacial Drag, Entrainment, and De-Entrainment 4.3.1 Flow Regime and Interfacial Areas: 4 i
| |
| Flow Regime Definition Bubbles Drops Small Bubble Regime a y< 0.2 Ag,Sg, We=10 A ,i drop Large Bubble Regime 0.2 < ay < 0.5 Ag,a Transport A t.sa, for 0.2
| |
| * space Equation 1 bubble for ay -0.2 Churn-Turbulent Flow Regime 0.5 < ay < acrit Linear Interpolation Film / Drop Flow Regime acrit < Gv Ajjif, uniform film The critical void fraction for the flow regime is lower bounded by 0.8; Gerii = max [0.8, a),,,)
| |
| The critical void fraction for stable film flow is related to the film Weber number by a;,;, = 1 - a, - 2.0/ Wep;,,,
| |
| where Wepi,,, = pgDn lU,gl /a TRANSPORT EQ. for Drop Interfacial Area dA' 3 ' d' P + Ve(A;, g,opU,) = (Ent/DeEnt) + S7(PhaseChange)
| |
| &p~
| |
| < frame /UPUACRS/Presl>17
| |
| | |
| 23 ry 4.3.2 Interfacial Shear Models U
| |
| T"'i, vt = K ,I vtU vt (I)
| |
| T"'s,y, = K ,t y, Uv , (2)
| |
| (2): DROP / Vapor Interfacial Shear Model K,, ,, = 0.125 Cogp,lU,,l 'y#
| |
| where CDd = 0.45 as suggested by Ishii (1977) for the Newton Regime.
| |
| (1): Small Bubble Regime (g < 0.2)
| |
| Ki ,vi,sa = 0.125C o spglU,gl A where CDb can be the viscous drag coefficient, the churn-turbulent drag coefficient, or the cap bubble drag coefficient, depending on the bubble Reynolds number.
| |
| g (1h Large Bubble Recime (0.2 < av < 0.5) l La Ki , vt, La = 0.125 CospglU glbfy y
| |
| where CDb for large bubble becomes the churn-turbulent or the cap bubble drag coefficient.
| |
| (Ih Film / Drop Flow Regime (acrit < Gv) h" K ,,g,yo = 0.5f;p,lUygl g
| |
| wheref;is the Wallis correlation (Wallis,1969) for a stable film.
| |
| (1h Churn-Turbulent Flow Regime _(0.5 < yn < acrit)
| |
| K i, vt, cr = FcT K g, ,,, ,o + ( 1 - FcT)K g,,,, tg where the weighting factor is a linear function of av which is zero at ny= 0.5 and becomes unity r at a.y = ac ,ff. Kg,yf,go is calculated with the friction factor due to Henstock and Hanratty (1976) for unstable film.
| |
| Ylh Intercell Drag If nj > 0.8 and aj < 0.6, then the drag force between vapor in channel i and liquid in channel j is l F f, x = f; p,lU,, g - Ug,,l( U,, g - U t.j) A 1, x where fg = 0.08 and Af ,y is the intercell area.
| |
| i
| |
| < frame /UPI/ACRS/Presl>l8
| |
| | |
| l 0 4.3.3 Entrainment and De-Entrainment for CCFL Entrainment in Counter-Current Flow 1
| |
| It is assumed that liquid in excess of that required for a stable film is removed from the film and
| |
| )
| |
| entered into drops (Lovell,1977), )
| |
| * 1 Sc = (acrit" U)PtlU ltAx Entrainment in Co-Current Flow The correlation proportional to the sand roughness, k s, and interfacial shear stress, t i, is developed by (Whalley, Hewitt, and Hutchinson,1973). (Wurtz,1978), and (Paleev and Filippovich,1966),
| |
| k Sg = 0.41 'Ti lU,lpgP,AX 2
| |
| o l'" ,
| |
| V De-entrainment in Film Flow The deposition of drops on the liquid film occurs as a result of turbulent motions, bringing drops into contact with the films. The rate of the entrainment has been correlated by Cousins et al.
| |
| (1965).
| |
| Sog = ka AC P,AX where kois the mass transfer coefficient and AC is the concentration gradient defined by
| |
| ; a,p; AC =
| |
| a, + ot y 12 nou ko = max [3.0492(10 )o ,12.4910U"]
| |
| ("]
| |
| U l
| |
| < frame /UPI/ACRS/Presl>l9 i
| |
| L
| |
| | |
| i 1
| |
| /~T L) 4.4. '' Interfacial Vapor Condensation to Subcooled Liquid The interfacial condensation rate per unit volume, SF ct,is given by (hgAg )SCL(H g
| |
| -Hf)
| |
| 'L Cp;Hf ,
| |
| Small Bubble and Large Bubble Regimes (ny < 0.5)
| |
| The interfacial heat transfer coefficient is based on the bubble collapse model due to Rowe et al.
| |
| i (1965):
| |
| h,t3,3ct j = (2.0 + 0.69(Reb) (Prf)I'3)
| |
| Film / Drop Flow Regime (acrir < Gv )
| |
| (h;A g)FD,SCL h i, film b i, film + h ,i drop b i, drop For the liquid film, the interfacial heat transfer coefficient is based on Colburn (1973) analogy:
| |
| l O hi , film = fuuPICpglU,il(Prg)~ '
| |
| where fuu is Hughmark (1973) friction factor, fnu = 3.850(Re,)-2/3 for Ref < 1000, or fnu = 0.5402(Ref)438 otherwise. For the liquid drops, the interfacial heat transfer coefficient is due to Andersen (19'73):
| |
| h,,g7,p = 2.7 where rd is the drop radius.
| |
| Churn-Turbulent Flow Regime (0.5 < ya < ac7;,)
| |
| (h;A;)SCL F CT(hg A;)FD,5CL * (I '' ECT)(h;A;)LB,5CL where the weighting factor, FcT,is a linear function of void fraction which is zero at ny= 0.5 and 1 at CL y = acrit-I O
| |
| 'V i
| |
| l l < frame /UPI/ACRS/Presl>23 i
| |
| | |
| ' i 3F l 1 o V 4.4.2 Example CCFL (GE CCFL)
| |
| The Predicted Flow Regimes in GE CCFL Analyses Location Drain Period Flow Limiting Flooding Period Bundle and Immediately Film / Drop Film / Drop Film / Drop l above the plate Liquid Injection Level Film / Drop Churn-Turbulent Bubble Above injection Cell Film / Drop Film / Drop Churn-Turbulent y pressure B.C.
| |
| I 12.0
| |
| _ _ _ ,_ y _ _ _ __.
| |
| 0 0'O
| |
| .g _& 'qh___ -_.
| |
| 6.0 I a 2.875 I I "
| |
| @ fb"ct on
| |
| .____ a_w _ _ _ _ . J 6.0 3 5 9.063 9.063 10.0 4 10.0 13.625 13.625 g @ fnfe*citon
| |
| ._{}___4)_.
| |
| .M __l__ f31 Figure 2-1 WCOBRA/ TRAC Model of The GE Test Facility t
| |
| O)
| |
| <frameMPI/ACRS/Presl>24
| |
| | |
| s O
| |
| i J
| |
| ACRS PRESENTATION ON COUPLED T/H-3D NEUTRONICS CODES WITH PIN-POWER RECONSTUCTION !
| |
| METHODIMPLEMENTED presented by David Ebert - RES/ DST /RPSB December 16,1998, Rm T283 Outline of Presentation
| |
| =
| |
| Q Discussion of T/H-3D Neutronics Code Coupling Method e Design e Application to MSLB Benchmark Exercise o Application to PWR Control Rod Ejection Accident ,
| |
| m Discussion of Pin-Power Reconstruction Method e Design e Application to OECD/NEA/NSC Pin Power Benchmark e Application to MSLB Benchmark Exercise e Application to PWR Control Rod Ejection Accident O
| |
| | |
| DRAFT e
| |
| Coupled 3D Reactor Kinetics and Thermal-Hydraulic Code Development Activities at the U.S. Nuclear Regulatory Commission (for presentation at M&C'99 in Madrid, Spain, Sept. 27-30,1999) by D. Barber *, R.M. Miller, H. Joo, T. Downar (Purdue University),
| |
| W. Wang (Scientech Inc.),
| |
| V. Mousseau+, D. Ebert (USNRC)
| |
| Abstract The USNRC version of the 3D neutron kinetics code, Purdue Advanced Reactor Core Simulator (PARCS), has been coupled to the USNRC thermal-hydraulic (T/H) codes RELAPS and the consolidated TRAC (merger of TRAC-BF1 and TRAC-PF1). The coupling scheme was designed and implemented with emphasis placed on maximizing flexibility while minimizing modifications to the respective codes. In this design, the T/H and neutronic codes function independently and utilize the Parallel Virtual Machine (PVM) software to communicate with each other through code specific Data Mapping Routines (DMRs), and a General Interface (GI). The DMRs are subroutines, extemal to the separate codes, which transfer both control logic and solution data to the Gl. The GI is then responsible for both the mapping of solution variables between therma:-
| |
| hydraulic and spatial kinetic problem domains, and the communication of control logic between the two codes. These coupled codes may be used to audit licensee safety g
| |
| analysis submittals where 3D spatial kinetics and thermal-hydraulic effects are important. Examples of these types of transients are hypothetical rod ejection / drop accidents, ATWS events, PWR MSLB transients, and BWR MSIV closure transients.
| |
| One important feature of the PARCS nodal code which has been added recently is a three-dimensional pin-power reconstruction method that enables the calculation of individual fuel pin power levels under both steady-state and transient conditions.
| |
| Through the use of the General Interface, this feature could be coupled to a subchannel analysis code to obtain local fuel rod enthalpies under transient conditions, as well as aiding in the prediction of local DNB conditions within fuel bundles.
| |
| * Present Address - Scientech, Inc.
| |
| + Present Address - LANL l l
| |
| 1 O
| |
| | |
| I I
| |
| DRAFT l O Coupled 3D Reactor Kinetics and Thermal-Hydraulic Code Assessment Activities l
| |
| at the U.S. Nuclear Regulatory Commission (for presentation at M&C'99 in Madrid, Spain, Sept. 27-30,1999) by D. Barber *, R.M. Miller, H. Joo, T. Downar (Purdue University), l W. Wang (Scientech Inc.), 1 D. Diamond, A. Aronson (Brookhaven National Lab)
| |
| V. Mousseau+, D. Ebert (USNRC)
| |
| Abstract ,
| |
| The USNRC version of the 3D neutron kinetics code, Purdue Advanced Reactor Core Simulator (PARCS), has been coupled to the USNRC thermal-hydraulic (T/H) codes RELAPS and the consolidated TRAC (merger of TRAC-BF1 and TRAC-PF1).
| |
| Assessment of the coupled codes has begun and one of the initial assessment cases is I that of the OECD/NEA PWR MSLB Benchrnark exercise. This benchmark exercise is specifically designed to test the spatial dynamic feedback coupling between the ;
| |
| neutronic and thermal-hydraulic field equations. Models for both RELAP5/PARCS and :
| |
| TRAC /PARCS will be developed and tested for this benchmark exercise. In addition to the MSLB exercise, calculations of a hypothetical control rod ejection are also being
| |
| , performeci to obtain the maximum energy deposition and enthalpy rise in fuel rods O during this transient using the recently implemented pin-power reconstruction method.
| |
| Calculational results for both coupled codes will be compared with each other as well as to those obtained by the other participants.
| |
| * Present Address - Scientech, Inc.
| |
| + Present Address - LANL a
| |
| O T
| |
| 53
| |
| | |
| Thermal-Hydraulic / Neutronic Interface Design Thermal Neutronics Hydraulics input input '
| |
| : T/H Side [
| |
| y 4 Interface l te e Y
| |
| //// -
| |
| Input ! Input
| |
| "-T, T,(ro 4 rn) Tj Tl(ro
| |
| * rn) 4 q a-Tm T/H j Neut.
| |
| I General -
| |
| Data i Data i Interface i Map
| |
| ^
| |
| N
| |
| ( ! ! ,
| |
| \
| |
| +
| |
| 0 g 0 9gO (A)o(AB) (AB)e(B) O', OjpOac Memory .
| |
| Structure Memory Memory (AB) Structure Structure (A) (B) 4
| |
| | |
| O O O -
| |
| Application to Steam Line Break Analysis
| |
| :: m Accident Scenario o Break of a Main Steam Line in a Secondary Loop -
| |
| e Sudden (Secondary Side) Pressure Decrease in Steam Generator e Enhanced Heat Removal from Primary to Secondary (Easy Evaporation) e Coolant Temperature Reduction in Core Inlet (One Side Only)
| |
| I e Positive Reactivity Feedback Insertion to the Core e Core Power Increase Until High . Power Trip Set Point is Reached e Reactor Scram but with One Rod Stuck Out e Core Power Drop e Continuous Coolant Overcooling and Positive Reactivity Insertion e Core PowerIncrease e Return to Power or Return to Criticality ?
| |
| 1
| |
| | |
| i Characteristics of SLB Analysis
| |
| .. m Distorted Radial Power Distribution Evolving with Time e Over Cooling in one Side of the Core e Mainly due to the Assumption of Stuck Rod e Requires Spatial Kinetics Calculation a Strong Primary and Secondary Coupling e Use of a System T/H Code Essential e Adequate Test Problem of a Coupled System / Spatial Kinetics Code o OECD MSLB Benchmark Problem Established For TMI-I and Currently Being Solved by ~20 Institutions Internationally 1
| |
| l a
| |
| 2 O O O -
| |
| | |
| u e a
| |
| Flow Boundary Conditions in Exercise II of the OECD MSLB Benchmark Problem ;
| |
| w' 300 - m- m =v m ,
| |
| ~- m = < =w w~ -
| |
| 18 - v w, . . _ .. r -_ - - _ _
| |
| . - - ;p <
| |
| .3 O. . . . s: p, ;c ;;w.;&g, sgysz j.
| |
| - ., .24 _
| |
| .~ p '& g.p , ;-yM; fy eyj.g:#M ,7; ';.-y: _
| |
| y ~ 8; i w n 35 7 a s me,ev v
| |
| .e .
| |
| 290 --
| |
| 16 .. s .
| |
| r Wev-
| |
| .! .. e -intact Loop *i V 4.M cG .
| |
| m ew4> m ,.nm*M,m, ap, 4
| |
| m' a":W..
| |
| ,yy ;7 . 'e * "n " ", r
| |
| " " - v+"- =a. .,
| |
| : -Broken Loop ja 280 ,, -
| |
| . . .. - - 2 A_- c; N -, ;
| |
| .7,3w .. eg; .,..ss.;
| |
| .. ~. 4. :n;;. ,.a,;m. - # m m,u%,. e- m..+ M. _
| |
| ,, r., . .
| |
| o 3 Q W ' W 'W! C@
| |
| [;g m-;q.a @... g 12 ; ;[ , , ,
| |
| 270 .. -
| |
| - .. ., q- c~ ~ llE i - a <WM. C
| |
| * y, 6*. ^
| |
| * f , ,.%,n;$. 5Q C's -
| |
| _. , ' , i Lt.p:' Y , .,
| |
| '
| |
| * n 6-'j-1.f } y: r-
| |
| . - a e ..
| |
| , .~ >
| |
| y:~. .,:.- ~ ,,
| |
| o to
| |
| .c . -- -
| |
| !" 260 e
| |
| O s -..g.s.
| |
| n~ .- . . .
| |
| : n. n.. f. ., .
| |
| n; ~A-g t - , ,
| |
| s ..
| |
| 7 ry_ . c u y, eg ." 7y y,.g,o g
| |
| u g , ,
| |
| m-'~
| |
| e :&- w 5 , .u ~ vx .e nz: w , ~ - 3:; ;-
| |
| n 250 ., , w. . . -. ~ . . , ,n-
| |
| - m>. o" , .c : %..
| |
| av.
| |
| e-. y. , , t .gy % e. , t r m . a c. &-c, : + > g .s.c s.# %* :
| |
| -xx , .
| |
| .+ m . a w.u. . n se u 6 W-w.; %;M56;.M &f:%. ^, .W@ .c~w < ;"Ma G>W pleG.. .n- :fr ; s + , u - -
| |
| 240 v .-
| |
| Gr@ CW -
| |
| M N!?F ,
| |
| * p*' 3:Vf r 4~,6~-8vd"FW: . . _
| |
| +.-
| |
| .. MW k' N" ?'
| |
| m
| |
| ..y j- . $^$ %ty *w%<:i; Q f
| |
| -~
| |
| ~
| |
| m s .a L w, ~''',~ -
| |
| ~ ;~ yM[.pl g .
| |
| --@; ;.g' @& w h +:w ge y(5 1m.uf,f i:. . '%.y ;; -
| |
| e*g,472Mrs e;# %,ay&ny%;.-yy,gp .v p g y:Q- ~
| |
| r+ -:a;;Q. r W.n .r: %7
| |
| , p 7;4 gy,y.g%.mnu " -
| |
| 230 4_ ,, a # # s % e , u ,a v 'o
| |
| .t . S ,. , 9_.ng33 g ,
| |
| ''~ " * "4; ' '~
| |
| 2
| |
| ~
| |
| *bE.- Y : s <: : '
| |
| ~
| |
| -. .., .{, s,i ","> '- r;.:~g k.N N:hf(il: kNElh* -y:yu$r?Wf?:i$w?
| |
| a ,
| |
| ,w
| |
| ; nmg< .cg nu eg%;
| |
| , . . . , ., n g ;-- _m.u .
| |
| * m:nr.g.
| |
| g r .g g , , i 220 0
| |
| O 20 40 60 80 100 0 20 40 60 so too Time, sec Time, sec m
| |
| 3
| |
| | |
| RELAP/PARCS Solution : Core Power 120 i i i ' ' ' ' ' ' -
| |
| i 100 80 - ~
| |
| l 8
| |
| ! ti h 60
| |
| - ~
| |
| l e
| |
| 0 40 - -
| |
| i m
| |
| y 20 -
| |
| y' _f i e ' ' ' ' ' ' '
| |
| 0 0 10 20 30 40 50 GO 70 80 90 100
| |
| . Time, sec 4
| |
| 1 0 0 0 -
| |
| | |
| t _,David E rt ._f 0_0.jpg_ _ . . - . . . . -
| |
| _ ___b_e - --
| |
| Page 1 i msta Assedly-h er O
| |
| RPOat 0s
| |
| ]
| |
| 4, I
| |
| 3-2s 0
| |
| l 0 20 i
| |
| 3 0 . .
| |
| i 18 1.2 l
| |
| 16 -
| |
| ; a m -
| |
| 12 . 0.8 to . .
| |
| E 0.6 E. E U.[b 6 O. 4 4
| |
| d a a I - 0.2 2
| |
| 0 0 20 15 10 5 0 t
| |
| 1 1
| |
| i O
| |
| 1 1
| |
| l l
| |
| l t_- _- _ _ _ - _ _ , - - - . _ - _ -
| |
| | |
| -" eLg M., Ana m %4 m.M Anmem-4, > m W - w'm='-mw-=6"wasv-vA-= *w srw~~--M-wwmwx- w- w-w ww w=M,.Amw"4 O
| |
| O d
| |
| .. ..,.. . e 3
| |
| * a D, h ~
| |
| a i h
| |
| <r i re .
| |
| l r S
| |
| / / / / / / / / A
| |
| *3 $
| |
| ~
| |
| : : a e
| |
| | |
| r
| |
| ; Divid Epert - ching:s.ps _
| |
| - Page 1___
| |
| (3 LJ Change in Core Characteristics Pararneters during a Rod Ejection (TM11, EOC) 1.25 r-1.00 -
| |
| ee
| |
| .N 0.75 -
| |
| t
| |
| ,, g 0.50 -
| |
| c-0.25 0.00 +-
| |
| 6# 150 i ti a
| |
| f 100 O
| |
| O 50 ,
| |
| 0 ,
| |
| / O i
| |
| (N y g 700 2 ^
| |
| 3 600 -
| |
| g500 -
| |
| $ 400 -
| |
| 23% -
| |
| 200 0.0 1.0 2.0 3.0 4.0 5.0 Time, sec O
| |
| | |
| l l Homogenization and Dehomogenization i
| |
| Homogeneous IIIII ,
| |
| !! Intranodal Flux
| |
| .fd V
| |
| .#fd, V .A eeeee M 8
| |
| C:.
| |
| -- . .s~. m M eeeee -
| |
| i M
| |
| f
| |
| (
| |
| Homogenization h .......
| |
| :::s::::
| |
| !565i E
| |
| = ....
| |
| sss:
| |
| y I
| |
| :s:::::: :::::::::
| |
| X g-:::sss-:-:-:-:-::-:-: i:!8i31 . _ _
| |
| !..- ! i." !!.
| |
| E
| |
| ~~~~~~~' '"
| |
| C!$$i ...!$!
| |
| '"I"' '"I" EHH L _1
| |
| $$$i$$$$$$$!$$:sssssssssssss:- -- -
| |
| i:58i8i8s:5 83 g F F Dehomogenization l y ;
| |
| ...,... . .. y.. .
| |
| L
| |
| ^
| |
| J leOeet OeeOe -
| |
| Heterogeneous
| |
| = , -
| |
| eegee
| |
| *o oe f #(r)s $g(r) -
| |
| o**ee Form Function 8 5
| |
| 8 9 e .
| |
| | |
| O O O -
| |
| i I i NEA-NSC L336 Pin Power Benchmark (Nuclear Energy Agency - Nuclear Science Committee, OECD Organization) :
| |
| a Problem Specification l e Two-Group Constant Sets given for 8 Pin Types
| |
| > UO,,, PuO, and Burnable Poison Pins + Water, Guide Tube .m<1 etc.
| |
| e 3 Fuel Assembly Types Defined
| |
| > UO2 (k_=0.99817), MOX (k_=1.02665) and UA (k_=0.66068)
| |
| ?
| |
| e 5 Core Configurations Specified C5: Small Core with Reflector !
| |
| C3: Reflective Infinite Checkerboard J J J
| |
| J=0 i 20
| |
| | |
| - - - - - - - - - , . , . -- - - - . - - - a m .
| |
| O
| |
| . e gec0** ys, 1 #e 9
| |
| %' LA
| |
| \ \ i C i x-ggt c 6 o- '
| |
| p, O
| |
| /
| |
| /' DE i 10 ,/ 30 g \
| |
| g
| |
| \
| |
| ,~ ,/ / \
| |
| so / so
| |
| /
| |
| 40 0
| |
| \/
| |
| 24 O O O .
| |
| | |
| O O O
| |
| ~
| |
| i I
| |
| I
| |
| ! L336 Result Summary 1
| |
| I i
| |
| ~
| |
| Code Cl C2 .C3 !
| |
| d5
| |
| ( l l C4V
| |
| ' ~
| |
| Peak Pin f PARCs l 2.5 -0.7 ! -0.8 0.3 l 0.3 ^
| |
| --+
| |
| Power F- - ---- t - .-
| |
| m - '- - -
| |
| I Error, q j PANTHER l 3.3 l 1.7 '- l 2.4 0'.9 L ~1.6 y . _. .._..g. _ -
| |
| 4 Max. " '
| |
| PARCS 3.7 .~ 2.2 d 2.2 4.1 l 3.2 1 Absolute 1 -
| |
| Error, q PANTHER l 5.0 1 2.4 l 3.4 j 4.2 3.9
| |
| , v--- .; -
| |
| l RMS 4 i .
| |
| ! PARCS- 1
| |
| ! 0.71 - 0.42 O.69 - 0.66 -
| |
| 3 Error, % - i ., .
| |
| ; Cl:UO,/UA Checkerboard 1. 6 i
| |
| C2:UO2/MOX Infinite Checkerboard C3:UO 2/MOX Reflected Infinite Checkerboard :
| |
| :I C4V: UO 2/MOX Semi Reflected Core with Vacuum BC
| |
| _ C5: UO,/MOX with Reflector ,
| |
| i 2h
| |
| ) i i
| |
| | |
| .-.-u~..-a-~.a.wa.-a-_.a_e..una--+-.------an-~~---- - .-.- - - -- ..+---- .
| |
| --.-n --..--n.- .----+ .-
| |
| ^
| |
| David Et;ec - LOO c; 6 -
| |
| , \
| |
| \
| |
| ,3 o 'S' ss 9 i.
| |
| u
| |
| \' '
| |
| \
| |
| y s S. h
| |
| \
| |
| \
| |
| s \
| |
| OD
| |
| ( \
| |
| 4 s
| |
| i 0, \
| |
| \ ,
| |
| =
| |
| \ ID.
| |
| e, g(D L
| |
| \
| |
| i 6
| |
| 50 4
| |
| . so ao
| |
| \
| |
| gh
| |
| \ n
| |
| \
| |
| t I
| |
| 1 I
| |
| I I
| |
| l 1
| |
| l l
| |
| )
| |
| O 1
| |
| 1 l
| |
| L _ . _ . _ . _ _ _ - _ __ _ l
| |
| | |
| i i
| |
| i i.
| |
| !O 1
| |
| i 1
| |
| : T. N O W T N
| |
| ; - - - o o o o
| |
| .i f
| |
| i a
| |
| o O
| |
| O CO t
| |
| i b
| |
| {
| |
| '$ S O
| |
| l 4 O m
| |
| l
| |
| ; O i
| |
| gi ,
| |
| o
| |
| . o N
| |
| e ,, e f
| |
| I O I a.
| |
| t a.
| |
| co:
| |
| h i
| |
| W o o
| |
| j -
| |
| o co i
| |
| 8 1
| |
| i l
| |
| O N
| |
| } e / - q 1 9 "
| |
| 9 o o a
| |
| ! - o i
| |
| s
| |
| }
| |
| }O 1
| |
| l.
| |
| | |
| 7iaei 9
| |
| O
| |
| : , = _ e ~ : - :
| |
| =
| |
| /
| |
| \8
| |
| /
| |
| t 2
| |
| \8 a
| |
| m
| |
| ,E
| |
| .2 9 r+ o 2
| |
| is '
| |
| 3 2
| |
| ~
| |
| 0 v
| |
| , q
| |
| ^
| |
| s5= _
| |
| e -
| |
| \ -
| |
| d "
| |
| ,. E s
| |
| 8
| |
| \
| |
| O
| |
| \
| |
| 8 A J - &' ,
| |
| O
| |
| | |
| b .h- mAb--e4ll-M-hA_mE2,a-A3Jw--, A4A--- O3 AANBnL,.e.JO4 + ah_ d amme a A.\%,e.A& >&eaa,as6 Ms 4 A enn.- AA e M,ma da Ambnasa 41.ms-.--Aasbw---
| |
| amm e =gm-_w.,u.a-wma-.v-w,-mw----. -
| |
| a O
| |
| . . Y v CO CO N N & F o d d d d d d o d d d o O
| |
| N
| |
| /
| |
| O O
| |
| r-i e k b o
| |
| , c2 O
| |
| h O f
| |
| ?
| |
| s -
| |
| e W
| |
| c e
| |
| * ~
| |
| p (1.
| |
| V (1.
| |
| U
| |
| #7 o Q ./
| |
| 2 ra 1 O
| |
| O
| |
| > r=
| |
| \
| |
| o CD
| |
| ' 5
| |
| ' o sD O
| |
| V
| |
| \
| |
| o N
| |
| - ,3 M N & O O i
| |
| i O
| |
| 1 d
| |
| I i
| |
| ,e i
| |
| | |
| O O O f s,
| |
| [ } United States Nuclear Regulatory Commission
| |
| ...;~ m g ;,. w =n- ~ __ = ~ m m m y;w n w n,,, u.., .,.r r
| |
| i Code Modularization Example:
| |
| Interfacial Drag Package Presented to ACRS Thermal-Hydraulic Subcommittee J.M. Kelly Office of Nuclear Regulatory Research Reactor and Plant Systems Branch December 16,1998 1
| |
| | |
| l 1
| |
| l Code Modularization Example:
| |
| Interfacial Drag Package m CONTENTS:
| |
| +
| |
| | |
| ==Background:==
| |
| Modular Design Strategy
| |
| +
| |
| | |
| ==Background:==
| |
| TRAC-P Interfacial Drag Package
| |
| + Interfacial Drag Package Modularization:
| |
| + Software Requirements
| |
| + Verification Testing Plan
| |
| + Example: New Modular Structure
| |
| + Example: Surface Plot Utility
| |
| + Summary December 16,1998 2 9 9 O 4
| |
| | |
| 0 0 0 .
| |
| | |
| ==Background:==
| |
| | |
| Modular Design Strategy ;
| |
| m Software Development Success Factors:
| |
| + The most significant factor affecting the likelihood of success for a software development effort is the complexity of the product. ;
| |
| [ lg !! r = = .
| |
| i
| |
| ; 80 .
| |
| g m, - - -
| |
| es :
| |
| 60 - -.
| |
| g ,,. !y - - -
| |
| n na : q x T
| |
| i :
| |
| En ij ! S j ht l
| |
| { j @i[
| |
| d '
| |
| i;-
| |
| ~
| |
| ~' ' '
| |
| M p l il tf !
| |
| h i l !
| |
| 4 3 3 i :
| |
| 20 -
| |
| 32 -
| |
| !! I -
| |
| f bI i e i n [g i p; sw %
| |
| o l ''= M 1 1 10' 1 10' 1 10' 1 10* 1 10'
| |
| "%"JaiFA"J'
| |
| + Confirms common sense: " keep it simple"
| |
| + Guideline: assemble software product from simple functional units that do one thing and consist of less than 100 lines.
| |
| December 16,1998 4 5
| |
| | |
| ==Background:==
| |
| | |
| Modular Design Strategy m Modularity ?
| |
| + Why does the " modularity" of TRAC need to be increased ?
| |
| + What types of modularity are being considered ?
| |
| + What are the benefits of increased modularity ?
| |
| + Example: Interfacial Drag Package December 16,1998 3
| |
| * t' e-
| |
| | |
| O O O ;
| |
| | |
| ==Background:==
| |
| | |
| Modular Design Strategy i
| |
| a TRAC-P: how modular is it'?
| |
| + Size distribution of subroutines (Version 1.10).
| |
| 200 i.,.,i
| |
| . . . i.
| |
| . ,i
| |
| ! 150 - -! -- - - - -
| |
| 5 100 -- ;- -
| |
| us t 50 '
| |
| - i- ---
| |
| 0 ' ' ' ' ' '"
| |
| = a ! ! 2 Lines of Code (max.) ,
| |
| + approximately 75% of routines meet the size guideline, but
| |
| + most of calculational effort is spent in the largest, most complex routines.
| |
| December 16,1998 5
| |
| - _ - a- - - ' - - - - - - - - - - - _ _ - - -
| |
| | |
| ==Background:==
| |
| Modular Design Strategy i
| |
| a Types of Modularity: }
| |
| + High Level Functional Modularity: )
| |
| l + Based on major tasks, such as input processing, transient thermal-hydraulics, and output processing. (:
| |
| e separation of input processing from thermal-hydraulic kernel. !
| |
| e separation of matrix solver from equation setup.
| |
| + Low Level Functional Modularity:
| |
| + At the lowest level, for a section of coding that computes one thing (e.g., a heat transfer coefficient), use a function ;
| |
| subprogram and group into a library package.
| |
| e modularization of interfacial drag package. ;
| |
| l l
| |
| t December 16,1998 6
| |
| [
| |
| O O O i
| |
| P
| |
| | |
| ^
| |
| O O O
| |
| | |
| ==Background:==
| |
| | |
| Modular Design Strategy
| |
| ,. - , - - - _ ~ _ _ , -
| |
| 4 m Types of Modularity: (cont.)
| |
| i
| |
| + Component Based:
| |
| i
| |
| + internal: use two-fluid numerics with own constitutive package.
| |
| e BWR fuel assembly: the CHAN component.
| |
| + External: may use a different numerical scheme and coupled via network solver (e.g., RELAP5 accumulator). ;
| |
| + Other Computer Codes:
| |
| + lt is impractical for one code to do everything, for applications where another code is more appropriate but feedback effects must be considered, develop an interface using a message passing interface.
| |
| e coupled TRAC /PARCS code.
| |
| December 16,1998 7
| |
| | |
| ==Background:==
| |
| | |
| Modular Design Strategy a Benefits of Low Level Modularity:
| |
| + Code Inspections:
| |
| l
| |
| + Manageable size and clear statement of function make line-by-line inspections of the code practicable.
| |
| + Testing:
| |
| + Functional units can be thoroughly " shaken down" using a driver program to vary input parameters systematically.
| |
| e Allows graphicalinspection of output surface for discontinuities 4
| |
| and unexpected behavior (see example later).
| |
| i December 16,1998 8 .
| |
| 9 9 9 .
| |
| -l,
| |
| | |
| O O O
| |
| | |
| ==Background:==
| |
| | |
| Modular Design Strategy ;
| |
| m Benefits of Low Level Modularity: (cont.)
| |
| + Maintenance:
| |
| + Upgrading a model becomes the simple replacement of a !
| |
| functional unit that is easily identifiable and has straightforward ,
| |
| interface with rest of code.
| |
| i i + Reusability- !
| |
| + Functional units that perform a single well-defined task (and are not hardwired into the data base structure) can be used by other routines as building blocks.
| |
| e (e.g.) a function calculating the bubble dierneter could be used in both the interfacial drag and interfacial heat transfer packages.
| |
| December 16,1998 9
| |
| ^
| |
| | |
| ==Background:==
| |
| | |
| TRAC Interfacial Drag Package a Current Status:
| |
| + Two Routines: each greater than 1000 lines of source code.
| |
| + One-D Components: FRICIF e all 1-D components computed in one routine.
| |
| e until recently, the interfacial drag was buried in a larger routine that also setup the 1-D momentum equations.
| |
| + Three-D Vessel: CIF3 e one routine that separately computes interfacial drag for the three i coordinate directions.
| |
| . exceptions made for different regions in the vessel:
| |
| downcomer, core, upper & iower plenum '
| |
| i
| |
| + Note: TRAC coding style (see comments on next slide) is typical of i that employed in the 70's and is similar to that found in RELAP5.
| |
| December 16,1998 10 e . . . .
| |
| | |
| 0 0 0''
| |
| | |
| ==Background:==
| |
| TRAC Interfacial Drag Package
| |
| . .me.u maranomxcoanmaram:2merm . ~ = - --..
| |
| m Current Status: Problems
| |
| + 1-D and 3-D interfacial drag models are not consistent.
| |
| + Allows user selection of modeling options:
| |
| (e.g.) "Biausius" model for downcomer, Wilson bubble rise for plena.
| |
| + Coding is enormously complicated, poorly commented, and local variables are not defined => unreadable.
| |
| + Inputs and outputs are not defined clearly for each routine:
| |
| " difficult to ascertain what variables are being changed or to find where a variable has been redefined => maintenance nightmare.
| |
| + Coding is repeated four times => multiple mainte iance points (e.g.) bubble diameter coding is present in the 1-D routine, and repeated in three different places in the 3-D routine.
| |
| + Component specifics are treated via " localized exceptions", all coding for one component is not grouped together.
| |
| nearly impossible to determine exactly what is being used for a December 16,1998 11
| |
| | |
| )
| |
| Code Modularization Example:
| |
| 1 -D Interfacial Drag Package m Software Requirements Specification:
| |
| l !
| |
| l 1) Null Effect: this is a "recoding" exercise only, the calculated results should remain unchanged.
| |
| : 2) Component Specific: a separate routine should be provided for each component type (e.g., vertical pipe, accunaulator) and all factors that affect the value of the interfacial drag will be grouped there.
| |
| l
| |
| : 3) Independent of TRAC Data Base: the interfacial drag package l should be isolated in its own module with a well-defined interface to the l
| |
| TRAC code; should be callable by a driver program external to TRAC.
| |
| : 4) Software Development Guideline: should conform to guideline that each functional unit should do one thing and consist of less than 100 lines of source code.
| |
| December 16,1998 12 j e e e .l
| |
| | |
| ~'
| |
| O O O Code Modularization Example:
| |
| 1 -D Interfacial Drag Package
| |
| - ;w m . .W agM*MWW Y - ?k* tupw.w. ' -
| |
| a Software Requirements Specification: (cont.)
| |
| : 5) Reusable Functions: at the lowest level (e.g., bubble diameter),
| |
| the models should be coded as functions that can be used by multiple components, similar effort for flow regimes (e.g., the model for the bubbly flow regime should be usable by pipes, pressurizer etc.).
| |
| : 6) Variable Definitions: for each functional unit, all input and output variables must be identified explicitly, all local variables must be defined, and wherever practicable, functions should return only one value.
| |
| December 16,1998 13
| |
| | |
| Code tviodularization Example:
| |
| 1 -D Interfacial Drag Package -
| |
| u Verification Test Plan:
| |
| : 1) Null Testing: this is a "recoding" exercise, the calculated results for the standard test problem suite should remain unchanged (i.e., within tolerance due to performing operations in a different order).
| |
| : 2) Coding inspection: software requirements #2 through #6 will be verified by performing aline-by-line coding inspection. In addition, the TRAC source coding will be compared to that documented in the theory manual; all undocumented features will be added to the documentation; differences between code and theory manual will be resolved by referring to the original references (when possible).
| |
| : 3) Debug Testing: during debug testing, everyline of the source code will be executed and the code calculated value compared with a hand calculation. The results of this effort shall be kept in a design workbook.
| |
| December 16,1998 14 9 9 9 ,
| |
| | |
| o o o
| |
| ~~
| |
| 1 Code Modu arization Example:
| |
| 1 -D Interfacial Drag Package ,
| |
| ; ; -a~~.m a*:nwwe+m Km m~ l Modular Structure: i t
| |
| A FriciF i (sets fluid ccnditions) {
| |
| TRAC teemponent type. JL
| |
| 'a
| |
| - - - - - - - JfE'.",'a?;s. '"Ta!"if't --------
| |
| Interfacial fluid properties) 3g l
| |
| Drag !
| |
| Module g,,,d",tFr (etc.)+
| |
| V V V V Hor ontal Ve cal Accumulator Pressurizer i I
| |
| v V V V Bubbly / <
| |
| Bubbly Slug Transition > Annular Mist
| |
| /
| |
| i Slug l Fraction ;
| |
| Profile Factor h V i ubble Diameter Slug l December 16,1998 Bubble Drag b 15 l
| |
| I
| |
| | |
| Code Modularization Example:
| |
| 1 -D Interfacial Drag Package m Annular / Mist Flow Regime: P= 2 I' par)
| |
| Interfacial Friction Coefflcient 8, ,
| |
| 7s .
| |
| i 6s 5s ,
| |
| 4s ,
| |
| 3s -
| |
| 2, 1s - -
| |
| 25 -
| |
| 20
| |
| * 1 O.95 15 0.9 0.85 f 10 0.8 j 5 0.75 Vapor Velocity (m/s)
| |
| December 16,19 16 e e e.-l
| |
| | |
| o o o ^~l Code Modularization Example:
| |
| 1 -D Interfacial Drag Package -
| |
| m Annular / Mist Flow Regime: P = 2 (bar) interfacial Friction Coefficient i
| |
| 8s -
| |
| 4 7- -
| |
| 6s -
| |
| 5- -
| |
| l 4- -
| |
| l 3s -
| |
| j 2s -
| |
| ! 1- -
| |
| . 25 0.75 20 0.8 O.85 15 0.9 1o
| |
| , g 0.95 5
| |
| 1 Vapor Velocity (m/s)
| |
| December 16,1998 17 l l ;
| |
| | |
| b Code Modularization Example:
| |
| , 1 -D Interfacial Drag Package .
| |
| ft '[c h'' ' WN , w - ON N 'D~ ./ $%d$rv%9", ' ' ' ~', '
| |
| m Annular / Mist Flow Regime: P = 2 (bar)
| |
| Gradient of Interfacial Friction Coefficient
| |
| -22 . _
| |
| j
| |
| -26s
| |
| -28s-r
| |
| -30s-32--
| |
| .-34> ,
| |
| 0.75 20 ,
| |
| 0.8 -
| |
| 15 0.85 l 0.9 10 0.95 1 5 Vapor Velocity (m/s)
| |
| December 16,1998 Void Fraction 18 9 9 9 . ;
| |
| . - _ _ _ _- - -__m _
| |
| | |
| O O O Code Modularization Example:
| |
| 1 -D Interfacial Drag Package u Annular / Mist Flow Regime: P = 70 (bar?
| |
| Interfacial Friction Coefficient 20 15 1 10 0.9 5 0.85 December 16,1998 Vapor Velocity (m/s) 0 0.75 Void Fraction 19 s _____
| |
| | |
| Code Modularization Example:
| |
| 1 -D Interfacial Drag Package m Annular / Mist Flow Regime: P = 70 (bar)
| |
| Gradient of Interfacial Friction Coefficient l
| |
| l 200--
| |
| l Os
| |
| -200 ~ -
| |
| -400- -
| |
| -600s -
| |
| -800- -
| |
| -1000 ~ -
| |
| -1200- -
| |
| -1400 7 0.75 15 0.8 '
| |
| 0.85 10 0.9 -
| |
| 5 0.95 1 0 Vapor Velocity (m/s)
| |
| December 16,1998 Void Fraction 20 0 0 0 . ,
| |
| | |
| 0 0 0 ;
| |
| r r
| |
| Code Modularization Example:
| |
| 1 -D Interfacial Drag Package '
| |
| ..--- _ _-_n .
| |
| i a Summary
| |
| + Modularization of the TRAC 1-D Interfacial Drag Package is -
| |
| in progress and has demonstrated the value of approach.
| |
| i
| |
| + Code Documentation: models as coded often do not match the documentation or the original reference.
| |
| + Problem if documentation is primary source for a code review.
| |
| . Undocumented features (e.g., limiting values).
| |
| e Errors and/or deliberate modifications ?
| |
| t
| |
| + TRAC Interfacial Drag Models: annular / mist regime e Unexpected behavior -lack of dependence on vapor velocity ;
| |
| (entrainment) due to use of drop terminal velocity. ;
| |
| e Contain discontinuities - due to improper 'if' tests.
| |
| e Contain errors - omission of vapor density in limiting value.
| |
| December 16,1998 21
| |
| | |
| /("
| |
| ,i NUREG/IA-0126 GRS-100 C MPR-1345 S International f Agreement Report 2D/3D Program Work Summary Report i
| |
| Edited by:
| |
| P. S. Damerell, J. W. Simons.
| |
| l
| |
| ! MPR Associates. Inc.
| |
| l i
| |
| l
| |
| )
| |
| repared Jointly by:
| |
| Japan Atomic Energy Research Institute Gesellschaft fuer Anlagen-und Reaktorsicherheit Siemens AG, UB KWU U.S. Nuclear Regulatoy Commission les Alamos National 12boratog MPR Associates, Inc.
| |
| Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, DC 20555 June 1993 Prepared as part of Arrangement on Research Participation and Technical Exchange between the Federal Minister for Research and Technology of the Federal Republic of Germany
| |
| ; (BMFF) and the Japan Atomic Energy Research Institute (J AERI) and the Umted States Nu-cicar Regulatory Commission (USNRC)in a Coordinated Analytical and Experimental Study of the Thermo-hydraulic Behavior of Emergency Core Coolant during the Refill and Reflood se of a less-of-Coolant Accident in a Pressurved Water Reactor (the 2D/3D Program).
| |
| Published by U.S. Nuclear Regulatory Commission
| |
| | |
| , _ _ _ _ _ _ _, . . _ _ _ - - - - - ~ - - . . - -- --
| |
| 4.4 TIE PLATE AND UPPER PLENUM BEHAVIOR TESTS 3 4.4.1 Rationale of Tests
| |
| : Countercurrent flow phenomena at the tie plate are different with different types of
| |
| ! ECC Systems. For PWRs with cold leg or downcomer injection, countercurrent flow at the tie plate occurs during the reflood phase of an LBLOCA as water entrained by
| |
| ; the core steam flow de-entrains in the upper plenum and falls back through the tie plate into the core. (De-entrainment phenomena in the upper plenum and loops are discussed in Section 4.5.) De water is not subcooled.
| |
| l For PWRs which inject EOC into the hot legs (i.e., combined injection) or upper plenum, countercurrent flow phenomena at the tie plate involve simuitaneous two-phase upflow from the core and local downflow of subcooled water to the core. In
| |
| , PWRs with UPI, only low pressure pumped ECC is injected into the upper plenum; ,
| |
| consequently, tie plate countercurrent flow occurs only during the reflood portion of an LBLOCA. In combined injection PWRs, accumulator and pumped ECC is injected through the hot leg injection nozzles and tie plate countercurrent flow occurs during all phases of an LBLOCA and some SBLOCA scenarios. Small break phenomena are 4
| |
| discussed in Section 4.8.3.
| |
| : Countercurrent flow phenomena at the tie plate were previously investigated at small-scale test facilities (References E-471, E-931 to E-933, and G-901). These facilities
| |
| ) simulated the tie plate using small perforated plates. The maximum size of the facilities was equivalent to one fuel assembiy. Using the subscale data, Naitoh, Chino and j Kawabe proposed an ana!ytical model based on the pressure balance at the tie plate
| |
| ; and the principle of maximum water downflow (Reference E-934).
| |
| 4 UPTF integral tests with combined ECC injection indicated that countercurrent flow behavior at the tie plate at full-scale was quite different from that observed at small-scale. Separate effects tests were performed at UPTF to investigate tie plate countercurrent flow.
| |
| 4.4.2 Scope of Testina At UPTF, several separate effects tests were performed to investigate countercurrent flow phenomena at the tie plate. Test 100 focused on the countercurrent flow
| |
| . limitation for two-phase upflow from the core. The objective of this test was to determine the core simulator water injection rate necessary to produce a given net flow into the upper plenum (Reference U-903). The other tests (Nos.10A,12,13,15, 16, 20, 26C) investigated countercurrent flow phenomena associated with ECC injection into the upper plenum or hot legs. The specific objectives of these tests are listed below. The test conditions are summarized in Table 4.1-1.
| |
| 4.4-1 a u ..
| |
| | |
| 7F Determine the portion of injected ECC which penetrates through the tie plate to O, the lower plenum.
| |
| . . Determine the location and magnitude of the water penetration area.
| |
| - Determine the collapsed water level and the water distribution in the upper plenum.
| |
| - Evaluate the effect of water subcooling at the tie plate on water penetration (Test 10A).
| |
| - Determine the delay between initiation of ECC injection and initiation of local water breakthrough.
| |
| - Quantify the tie ' plate countercurrent flow hysteresis effect (Test 15A).
| |
| - Evaluate the influence of the momentum of hot leg water plugs on water penetration (Tests 158,16).
| |
| . . Investigate water entrainment into the hot leg and SG simulator for UPI PWRs (Test 20).
| |
| The data and quick look reports for these tests are listed in the bibliography for this report (Section 8). Evaluation of these tests by FRG is provided in References G-411, G-906, G-915, and G-925. Separate ovaluations of Tests 100 and 20 by the US are documented in References U-453 and U-903, and Reference U-454, respectively.
| |
| 4.4.3 Summarv of Kev Results UPTF Test 100 was a basic test for quentifying the combined CCFL behavior of the core simulator (i.e., steam / water injection nozzles and dummy fuel rods), end box and tie plate. In this test, the steam / water upflow from the core and water fallback through the tie plate were uniform across the vessel. The steam upflow and water fallback rates for Test 100 are plotted in Figure 4.4.-1. For comparison, the corresponding data from the Karlstein Calibration Test Facility (one fuel assembly) are also presented in Figure 4.4-1. The figure shows that the flooding curves for the full-scale and small-scale facilities are nearly the same. Hence, the CCFL at the tie plate for uniform steam / water upflow is independent of scale. The upper plenum collapsed water level did not exceed 0.3 m during Test 10C. _ The results of Test 100 were used in both the specification of test conditions and evaluation of data from other UPTF tests.
| |
| O 4.4-2
| |
| -. ~o --- + $e
| |
| | |
| For tests with ECC injection into the hot legs, flow phenomena at the tie plate and in
| |
| ;Q the upper plenum were heterogeneous (i.e., multidimensional). Specificall in Figures 4.4-2 and 4.4-3, water downflow occurred in discrete regions. The water downflow regions were located in front of intact loop hot legs where injection was occurring. Collapsed water levels in the upper plenum were also multidimensional with i
| |
| increased water levels over the downflow regions.
| |
| Figure 4.4 2 includes a plot of saturated water downflow vs. steam upflow based on
| |
| ! the results of Test 10A. The figure shows that water downflow increased with L decreasing steam flow, as expected. Due to the heterogeneity of tie plate countercurrent flow described above, water downflow with hot leg ECC injection was significantly greater than that predicted by the CCFL correlation determined from Test 10C (Figure 4.4-2).
| |
| In Test 12, complete breakthrough of subcooled ECC occurred at core simulator j steam injection .1tes as high as 284 kg/s, the highest steam flow tested. For comparison, the average steam production in a 1300 MWe GPWR during reflood is 150 kg/s. The maximum subcooling measured in the water downflow just below the tie plate was 70 K.
| |
| ( Tests 13,16 cnd 26C investigated the effect of core simulator water injection on
| |
| [ countercurrent flow phenomena by variations in the ratio of core simulator water and steam injection rates. A Kutateladze plot of the steam upflow vs. water downflow for h these tests is providedin Figure 4.4-4 (left graph). The graph shows that, for a given steam flow, water downflow decreases with increasing core simulator water injection rate. Hence, tie plate CCFL and water downflow depend on the total upflow through the tie plate and not simply the steam flow,
| |
| ;~
| |
| Test 26C demonstrated that existence of a two-phase pool of saturated water in the upper plenum at initiation of hot leg ECC injection had only a minor effect on the water breakthrough at the tie plate. Water penetration to the core region followed the ECC 4
| |
| delivery to the upper plenum without significant delay. Formation of a local subcooled water pool in the upper plenum combined with a local increase in collapsed liquid level provided the necessary conditions for water breakthrough at the tie plate.
| |
| Increasing or decreasing core simulator injection rates had a significant effect on the water breakthrough at the tie plate. In Test 15A the core simulator injection rates were decreased and then increased to investigate this hysteresis effect. During the period of decreasing core simulator injection rates, the water downflow rates in Test 15A
| |
| : (2 ECC injection ports) were the same as Test 13 (3 ECC injection ports), which also
| |
| ; had decreasing core simulator upflow rates (Figure 4.4-4, right graph). However, during the period of increasing core simulator upflow rates in Test 15A, water l downflow at a given steam upflow rate was much higher than in Test 13. This 9
| |
| I 4.4-3 j ._ ^
| |
| | |
| BV
| |
| /^ hysteresis effect resulted from differences in steam condensation below the tie plate for increasing and decreasing core simulator injection rates. Specifically, condensation was higher during the period with increasing core simulator injection rates because the
| |
| " initial" downflow of subcooled water for a given injection rate was higher. The higher condensation rate reduced the steam upflow and increased water downflow.
| |
| In Test 158 ECC delivery to the upper plenum was intermittent due to plug flow in the hot leg. The time-averaged water breakthrough at the tie plate was not significantly affected by intermittent delivery.
| |
| To simulate flow conditions representative of different times in an EM transient (i.e., single LPCI failure) for UPI PWRs, the core simulator and ECC injection rates were lower for Test 20 than for the other test! wi*, hot leg ECC injection, in this test, about 40% of the core simulator steam injection was condensed by the ECC-water; the condensed steam was returned to the core region with the water downflow.
| |
| Due to steam condensation in the upper plenum, the maximum water subcooling at the tie plate was only 25 K. The collapsed water level in the upper plenum was small (upflow region: 0.05 m, downflow region: 0.2 m).
| |
| A Kutateladze-type CCFL correlation was developed to predict water breakthrough with hot leg ECC injection (Reference G-906). In addition, an analytical model for tie plate countercurrent flow was developed based on the pressure balance in the water
| |
| ] downflow and two-phase upflow regions; this model shows good agreement with UPTF tests results (Reference G-925).
| |
| 4.4.4 Conclusions The UPTF test results revealed tie plate countercurrent flow behavior with hot leg ECC injecdon is quite different from that without hot leg ECC injection, even with saturated ECC injection. Without hot leg ECC injection, flow phenomena at the tie plate were uniform; for this case, tie plate CCFL was independent of scale. With hot leg ECC injection, flow phenomena at the tie plate and in the upper plenum were multidimensional. Specifically, water downflow through the tie plate and upflow from the core occurred in separated regions; the water downflow regions were located in front of the hot legs. Also, water accumulation in the upper plenum was greater over the downflow regions than over the upflow region. Due to these multidimensional phenomena, water d}}
| |