ML20136A978

Draft Analysis of Thrust on Pressure Vessel After Core Melt-through at Elevated Pressure

ML20136A978
Person / Time
Site:	Three Mile Island
Issue date:	05/18/1979
From:	Dykhuizen R Battelle Memorial Institute, COLUMBUS LABORATORIES
To:
Shared Package
ML20136A983	List: IA-85-498, Responds to FOIA Request for Specified Rept to Be Placed in Pdr.Document Listed in App Placed in PDR ML20136A978 ML20136J284
References
FOIA-85-498 NUDOCS 7908070141
Download: ML20136A978 (9)

Text

, , , . . . . .

. . ~ ,we 7 - + , n m -

+~r'~~_c- - ~ . + - ~ + - - -

- - ~ '

e ~~. _~ .. ..

~

.. . .- ? l 1

4. , ,

. .3,

.' :.. . '. . i : -

-! i. . . !L l A};*' it P L * ' .  ;! ..;

A! L .E'.'.".T C . .E L 5 M ha by R. C. Dyihutzen, R. S. Oenning, and P. Cybulskis Introductien l In some of the inportant accident sequences investigated in g the Beactor Safety Study as well as the Reactor Safety Study P.ethodology

• Applications Program melting of the cere is predicted to take place with the primary system at elevated pressure. If the systes is not depressur-E 1:ed prior to the dropping of the core debris into the reactor vessel bottom head, the failure of the latter will result in the discharge of l high-pressure steam into the reactor cavity and impose an upward thrust en the reactor vessel. The purpose of the analyses presented here is to i

enemine the negaite'o d of the thrust sad to evaluate the capability of the restreista la abe system to prevent the vessel from impacting and possi- ,

bly teiltag the aestaianeet hamadary.

s r

s h &alal Eenditisme e

r ., . .

m tdellestag asempticas unto made to determine the initial t .. _

[ esadistene for this analysis.

Ble'engtse p% system is filled with 2500 psia steam. This ,

epper pressure Semit is metatsaamd by the pressure and/or safety relief salves es the preemeriser. It is assumed that all of the liquid water bee boom eseperated.

Figure 1 obswo the geometry used in the analysis. The vessel is restrained by six pipes. These are the two hot less (36-in:513 and 2.875-

anch "-L " t if ckness) and the four cold legs ( !-inch ID and 2.25-inch

} qq63d'1hb1l , r, . 1  ? ^ '! j

' , q f T' -

l .' [ - _ - - _ - - - - -

- - - -z 4 3 p_ _ _ _ -_ - -- p .-

m --- -'

i I

! i 4

e t

am O

, Cap 0.2'

/ .

. ~

g l

f .

A f , i L 1 Reactor Vessel i j

3 /  : i 'a taa=

+.a. v / Q M P

~

. I h ,16y

,; ' cm o F ' 4 Egt T k' g k_ $

- + .; > ..,

> $f

y;g .

s an =

9 d' b

a ,..~ a.* #

1 s g%k % . , t

• I l t

, ,, m T' -

l 4 &' , , , N 4't

~'

V . L.:

+-

=

/ I 1

e /

+ v .. .

- e la;g N ,, * ,

n, ., ,

t - .4. .

J -

so- e n . I l

' ~ 4n a

1. O* .. ,

• ) i

• * , ,h I '

g I

~ .

tk i'a# [.

>-?',9 t .

P002 CINAL

" .r ir P

..s ' ! !;..ci :.es s ) .

Thex- leu -

lLr ..: .. ~

• t .'. e react.r. The effect ok t!.e vessel st ;; pert sitra, ts e:t ; . . '. . t . i- t r.i s analysis.

The pipes are conne:ted to no::les en the reaet;r vesse: hich are about three feet in ler:st!.. The distance from the no::le ta : Le shield

• t wall is approximately one foot. and the concrete shie c wall ta apprcxi-mately five-feet thick at the penetrations. The gap betwen the ;:1 pes and the per.etratict.s through the shield well is taken as 0.2 feet. These values

,  : were obtained by scaling distances on available drawings. '

g t

Finally it was assumed that the pressure vessel bottom head veul' '

j

(

li-l instantaneously spea op to a hole the size of the reacter press re vesse: .

4 i g,g

,,,g a . _ - _ It '

f blowdows Calculation l

i A hE1AP run was set up to model the entire pri ary system. The i

system uma modeled eith a staple sevem-volume, eight-junction network. The break use yled as am lastantaneously opening leak junction with a 15

.! Po&a stok peames,. .

[ ,

0 . Slima ,

, of the p r em the 11 ewing equation can be o s.e~, , ,e.e.e n :r mm s.e . .he sessel,

-' , . , ,,s , 33 y ra

, , _ .y.;

r: ..~

wn

. .~

.w - - {gg

,. a '

h. ~ c4.

. . 3.g.- .g

.-w h I,

, . . . 3 c, Gjbg&.- . W seen. evalasted at the emit i.

= +F.. . ~

j;.

h F8ppe.2. De lattial forse la

e. . .n. .

1:

r te..es n pseessee w differential _2500

.~ .

i.[

Entatally th48 Serse drops off I V das shas less septity as the other

q.  ;, -

-L

' :1,. 1 -

t Lp Was >

4 d she seasser veces1, 8.0 s 105 ,,,,,,,

• ~

,s.

o s

3 % .. ,

I .E P ( 98Mae Jape Forees Caused by RIowdca and 7 F G n -Q W 7,7 taw }

l , 9 .

v.--< > . Stese. *ASE,'J. Flands regineering, september 1973

1. _y . , .sg t,

, , r.

s t

VJI )

3 {

i. '

t j i

e i i g 1 ,

. i s

' \

7 . .

1 x 10 N, '

\

l 7-

\ '

4

( l I

l l

e. '

s

, e I

D L,

l l i,

u 3-I he y o l

l i

l t  !

i. ,

0 j<;.. . 1 a 10 -

t . l l

1 7

n  !

m .

.- g A /

A

>,: l

, w .

s . . l

e. . , .

s;: , .

g~,-

3a \

4. , j

)

_,,. J w

. y .. . =- y .

.i Lyn 1 - - -

, 'O .2 .4 .6 .8

'~ . .. s4,w.-* c'

ic- . ..;, .,.

4 g ,.

h (eseemds) it,j 2 , .  ;. . ,

R -

W.n' ; ,

M 2. 153037 05 VESSEL *.*ERSUS TUI

e 4. ' . . .-

^

-L' f

-' ' l *D #)

p;':9 ,

1

, %.w" p. e A. q_,

I k- '% , .

pt; 4 ' .{; -

s.

---

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

g. -

) .

. *.* .. J c . s

.. *f. rat

. , ". ~ :. ' 4:. '2 l* s's .i : e .

.. t r.c *

• th.

• 6 ' . *f ? ..s'. . . = . <-

7 .a . . c : s eealaa:t : . ..is.r . .:.. ter- c ;_ re a 1.  :. : e..:.* : *5c

r .c

n vir. M rt :: s

. e .cas, b.: .:.:t r,r . .' -  :: c .::e'*

a u !. .c.  :.J -e +t. st c.

l l

k Strest 7.11 - 3 I .1 t i" P.

i i

l A vertical gap of 0.2 feet was assumed bet *.een the pr'r.ary a coolant piping and the shield walls. If the thrust force remains at its I ialtial value, the time to impact of the pipes against the shield va*. s is s p onds. As can be seen from Figure 2 at this time the force will esill be near its initial value.

-o Osing the shipping weight of the vessel, the velocity at 1:;act I

se g . and the kinetic energy of the vessel is 1.14 x 10 ft-lbs.

If we assume that the vessel is stepped in another 0.2 f t of travel (caushtag of the shield well, and some deformatica of the pipir.g) the maximum y, . .r .

6, Neue e.s. es 5.. deteentmed. Here we soeuse the force decreased linearly

v. . .. ~ - . t ._3

[ 4 k"

['

• v+ u . . , i.u' s :a -

s L. ,J Q ;

t,

s. .-: e g.,

p g

@n + j$ .dfg.%,f

$m' ,:) *. ( 2 <3

- j99 k' f5 1 a 1.14 a 10 /8.2 = 1.14 x 10 lb g

- w. .Not p . W.

u ' . ; q'

m. . , ,w*:5m . Y .a> ,, ,

~

l

$$+Di.[~ t'

.a _.s k. . n'ef telemleties the resistive force of the W

l jpaetsy eeslama pipes are equally leaded, -

l

, . \

,is;eged as the ultimeta sessile m

l m

- : +N,,- eseuse that the en-l $****8 *'E***=*2 '*8as *ha .!

b eaaastated as the asesmed shear I l l .s' n. u;- .

ames of the pipias us11st l 4 & 'g W : 1. , W.&fi % 7 39M;h U M ' , M .m k M = 1.99 a 10 8

1h g

%g ~47 w w

t;h 42

y. - ,

f- 3 rq$7hc thMj n

' SQ - -

[M Q[; ,

effesseets af the pipias with the shield well (0.013 g ft,D h 'is egual se the impact feces plus the thrust c'

Q*- /- ,.

hqs gOddl_ M ~'l 5N;rl ],:

'~ AMM&.?N?&fM?$$g?;.'

"y : ga y w 9,JJ&

a W.%.. . .

~~]

, p. , - w, .

a -

l}

• 1.1- 4 **'~ + ..% x ) .  ; . n .' x .

3 13, ,

<ance the a::ir.: ! cree is r.rcate tc.an tr. rests:;ce :,:::e .; h theu ass ; tiens. the pipe:, vou!d be predicted to shear. ~he "en el vodd t!.en continue tc s: cele; ate upwards and could threater th conta:

ment. .

n-

' The acceleratin:: ,

feree sculd be less than shes, in Fi;:ure 2. af ter the pipes shear, since the a - "1d now be blevieg cut the e-g a*N acaties and the vessel vould be separated f rom the rest of the

1. arf-k cociant system which was adding stems to the vessel as it depreuurized -

i -

Calculational Uncertainties f

(

0 d

i the is11 swiss uncertalaties should be noted when evaluating the .,

abyse eslaulettees. - May both dearease and increase the likalthood of the voisiel breaking . s ises of the primary emelaat piping. y Some of these uncer-tessiamo saa be'seestved eduply. ty ebestalag more detailed plekstesolved. or the data f .

4

p>

k,- ,[. ,,vh. m mW h,m,)an'estamatve amount of re r-g'-Q t2[N;v, , . xM , , , .M % y-.

N8' '

• hose failure was assumed.

pW' -

Anttial thrust actias

,3 g -

g,.y . -

, 'E * .

"E

,. / 47: . v . .o. and the akteld wall is fjf . ~

• g .

4880 14 leSS that the i,I

]g' MN[.g g.g

r + .

. e m .,,- .s.4 'c.

yn% #.. w

.a v...u.-+.

.. . e .

..m. m m . m. - . . o.

.>p g

e - -

,1 - ..

og gm , song,gg,,

f.,

g

...:... . 611a _ enerly

'b m, .s

.; h. ~ : w,... she sal-

+

4

.o

.aEeO

,_hN t

.< tho ' . . . m,e .~1st11 rip

?c' .7 .

- , , ,. v _p, . r.

. fail at & emett m,.

ass 11er .~

,-w asen - .tle strees la optimistia. ..~

Nf . , ... .; ..

J f s <

r f f. .M .h-m :w.+ g g y

, NM Walaam, ' .$

?ft) f f< 3 1

,s

~>

,k ,* *n,o,

. s i.r. A . ' .. y~m

.\ .. *: .49> 4- r 4

,} .

x <

5$: A *& . p p _ ;.xy s- ",

~.

.np f a g, K >m.#M,4';7,Wllh "fR, ' f; .*p -

.n y$ ^ V.:  :. e . \5y C' N y ' .

x ,

d!:

$g % q

.\%;. ,

l,'  ?>>.;*c,fip?l&'}y.;%., f&& __- _ - - L A,.-' y" g

. ~

t ce :. c . - *-

.. - . l, S:

i . . . .

;. .: t

.:  : .c .e ;.;:.:  : :e . N ; c l J ; r. g 3 '. . -

a is t : ' .. .

fCJ-

- .... . a re: n i e'. r..r effect.

Ta a;1 7 , :t va, ass : d taa: the:e ;s r.;ta.n, n.'.; r. -: -

e m e! exce.r: the prinary coolan; r arir,'. and the c pres s f *e r . re ,e .-

tive fcrce as they are accelerated up with the vessel over the in t ; ;t.1 0.2 feet. In addition, the crack opening time and mede of : rack prepagation were not considered in the analysis. The -all cf the vessel in contact with the fuel vould be expected to be heated fairly unifor 1y anc significant yielding vculd precede failure. The initial failure would occur at either a hot spot or iccal defect. The crack sculd prob-ably run axially up the vall and down around the lower head rather than running areund the vessel and effectively un:1pping the lower head.

j, Typical running velocities for cracka in ductile vessels are 800-900 ft/sec. He separation velocity for the edges of the crack would be an order cf magnitude slower. Thus the openlag time could be slow in f comparison to the predicted time of impact in the analysis and the force at impact could be significantly less than calculated. The maxi.sua thrust force ta Figure 2 would alas met be achieved.

?

o_

' .,:t ., :.

, s . ; m,.[u ~. s.m, esm ertene e stay 11 stas, bet realistic onergy

, , ; ;, _ m . - .

(. '

M .". s. .< r Q:.I,-

W

+, s .. .n ;; $~ . essmag to prinsey emelant pipes fait

- t t>, ,

s : s . m,.,

@% , ' 'I ^ .

e r

)./k'W[Okij[ @ Y Ge pee m os unseal. p ity and tk h ring 4

l

? '. 4 ,,

.Geest asse spense. It we aeo ne the veneel is W;r.. - m

..s - - .

~

w m g_,fe. 't. e'eems e,f to peteery speeum, it will blow does quickly, 7 SMesNe w. su wM M % eely e ass 31 pecess of the system steam, and more hetes *

. ,.i

\

5 ,. l ieGL.'ge lW  %

.a@> m.p a&&.sesses h e peesoure. The fellewtas is an esttente 4 . .. . ,

}

p't u st e

- s elsee GWc.ee a, ehet'usend result as a functies of time a v .f ;.

, M. y,y .r i

i k . 'i e' F's N7 a 10 ear (-11.267t) t = seconds. h

'[ l .

,.l** -

'y'RC n , c , . js u g- . t g . '- h

+

\

+. . ' * ;? ,

A

.O *

.8 r,  ;

4 4 j }

% , t. r

,, q- f p .y -1

. * .. # #yC7~ .. * *U. r l

l - ' ' '

' ' ~ ~ ~ ~ ~

.. . 6 .'a ; c.

• f ; ' . 2: 1. * . at.:1 .

.t. .. *:

f. i' -, f. I ; ." . :  : . .e  !! ...r.: : .e is - .'.' t;. is-- .

g ,

cal:alat:r- - s...: .: :..:. . , . cre : ..c ; :. '

.( u.:. -

te.3::c s t ren.; h use.! ; rev w : . P.:s 13 a re. 1: e:' t.c ...: .  : . : ., : . 2 J.

i s t !.c difference '.et.e,..  : tens:;c s:ress an: a s!.e r st rs s . and t.;c s e e . r. . .

i ,

1/5. crudely :akes :,n:: acco n: ti.e ; essibi;ity : hat ::.c m; es gra Nally rip apart instead of an instantaneous shearing of thcar entire cross section.

There are three periods of interest: the ci.e period before shearing *, daring I

shearing, and after shearing. The force during these times on the vessel are given as follows:

F(t - 0, .015) = 5.7 x 10 esp (-11.267t) - 8 x 10' i F(t = .015. 0.20) = 5.7 x 10Iexp(-11.267t) - 1.5 x 10 m9-F(t = 0.020, 0.38) = S.7 a 10 esp (-11.267t) - 8 x 10' The fiaal time. 6.30 eseemds, is the point where the acceler-

% e

.  : asses chasses as.a dauenced wies.

a. . . :.

• i N.O,V'.  ; M k M.,.s... ..

Iyu

- e v. -

,:. 5 c:mese betas ehet. et 8.g,"Somens and omleettag the offoct en the vessel

.r- ..

eso'nado, the pressare vessel has risen 55 feet

. . ..$: ~~ . > ,m n .,

1. ,.: .h ~ M- g %,vy . 0l W ,~

8 s ~ ..

. n'idges.

w.

op, h..!J}I,@, . ' 1[j. _.m .n. m- . . M4 m MseMier ptplag et tiedowns that may hold

' 5 Mn tgly, g sb8ppies seight of the vessel was used.  ;

[n

[$ g should add to the inertia senseland other structures would have ru .m.

D:A.

u .a AR98;,

Eh ~-

Ee *;. i *Y S k  ; ,

,*s

, maym.-

4, : .

t.' w -

.3 l:,,;*U q .; . + k .

Jry4 ,  !

y.- 4-($g  ; rg t e , :.sf' ,, ,

4 .

2 4%- ' {g

• li k t] ~.{: .. m ~ .

.31 \ ' ~

f * ," e j *- 't

l

.U.C  ; . &,;, ,

~ .

^

$c w . a ,. :fl :::;; .~ -

v1M.. #-. Jm. 'n g.. c %. .y .. , .:

~~

~g g~ . m

,. . . i .' sr ,.; s',

.. ~

. . . . ,j. 's. : g

'S .N. $ lh "V hy kb

• e' *. ' '\I $ 9.,'

• e e

4

~ *'

s esc %Q s r. . : : a s e .s s i . 1 * : s T. s

• t . s S n: .

fr :. r : r e W r..; e e: .r rtr.:  ::.. .u .

4

. .e- ,-  :  : .tt t ..a :

i j e a r l s :a il u r e o f t r e c c r.: :r.r.er.t i <. c ric . . .- ; e . AJJit u . 4 , ...r.: a .4 : 2 .i r. a r are detailed cale21ations ate required :. dcternine C. ether er r.. t this is i a likely containcent failure scde for a s;ecific plant. Resolutten of this t

question is important for all current plant designs since it cculd effect the quantitative results of risk analyses and, of : ore signifi ance could influence the proper course of action of an operator in the event that ccre salting beecses inevitable.

L useenclature a.

A L,/.

e

a. r9 - u i e., A - area 5

1.-

..~ '

~

- 'T = force F-a=.spp1&ed feces u y

kh.mYh.N,.

.s ~

. ~

due se asevity f

. =<

m w..

.,  ; . , ,, ' . . ~

+,.o. .}.ik ' M Q;<- y

'N '

g

,%, ,'_ em W. Np - - +

....c.r,j A , . .. ,g. .o ' ~ t j;

s u g( n ,

a eiu w - . :

~*  :

. ..e .M, ym.C., L J+ ,g,..-?,< . ,

o

.r; s ,c s,'tl.

^;nb s 48' 4

. ] J * $, 'E *g

. "v r . .*

. d g* A e D# I ,I

, ,1[ .

%*y 5 ),l' . *s 9Ja$. .

4 M. I,;.j. . 14+*'r.a . # '.

4g,,' n %pd-p.~

2, r ....-..E* ,. s.e ,

+,'A  %.

p * -

. A. s ,

'. v Ot" l - g y*

.Q

. r..

u u*ky o,; " . M 7* *

&" e

.I 4 . -- .,i

• '--. ..I+...  :*

- n a,-ft; .tg.g) - ;ils qs L'%e y , a. % ~~ ..

su .h J

,y k, g- / ~y3 m% s. . Cs..of.

Mb, e g*g. . ,, **:f.;qf,. v* .

w s u , .p..;; a+

,,g .[ D 'e bk.

e ? .. -

'g o.

t

~ I 4.f*.'i *. % k m

a

.xd1g,'

61. f ' , # 1 " i

4. ;; g,n[.kcm

,, s

-y v. 3 4..,  % 3 **Q' ,f42 qh y s*g y:l. ,

g a;

%- ,Q

m. Y ,., r *

$. n. . .# u.A . , V, 9. . %. .y **,L W, I,,:e"M,ri 3.s#ie.M:~rs..

4;

- fAJ,Pr* ,f.t.a **,

.#. s. 5i1 307 1,

4. e9 ,% . * *

• e # e: .+ - .. .

s

, s . A, a e . 5.k. , ,

e+

ta: Q,nN'v.,",y'*d.

. 7,.

.ser* O2s * **-

g' *,l ,

r, .t

.- "glf - , f 4 **r ; .

• 1- 1 ..L, g ?y.?; y, >g .

@t.' . . . . ,

M. "

( & t c-:. ' , ,

.,: u.

, ' ? :s: w +&Y2 m.y..'k .a y ,, 3,g g<,y.. e . a , g ,, s, , , ,, n ,Q, a : g y

91- .a. , , - : s ._ - .

r.

. . -