. .

VNITIDSTAT!S

NUCLEAR REGULATORY COMMISSION WASHINGTON. D. C. :OSSS

APR 1 1 1979

ATTACHMEfiT 1

ME."!CRA:IOUH FOR : Roger J. Mattson, Director, Division of Systems Safety, :1RR"

F~C~I:

suaJEC':' :

R. 0. l·!eyer, Reactor Fuel Section Leader, Core Performance arJr.C:t, Of•li sion of Systems Safety, ::RR

CORE DAMAGE ASSESSMENT FOR TMI-2

Attached is our assessment of the core damage at TMI-2 for use in the SER for natural circ~lation . It represents our independent evaluation of the facts available and of the industrj/vendor/licensee analysis, which we have heard in several briefings.

An ~!rlier estimate of fuel damage was made ~Y Rubenstein et al, and a recent meeting was held at NRC with fndustrJ experts. Memoranda describing those evaluations are attached to this coc~~nt.

Attachment: As stated

cc: R. TedeSCQ K. Kniel P. Check C... Berlinger D. Crutchfield (FOIA Ffle) o. Houston H. Tokar V. Stello 3. Grimes G. ~fghton J. '/ogl ewece D. ?ewers POR

......... . ........... // •/ ~ 1-• , // • • , /~,r.:- .. \..c.: ' .'-~ / ~ •• Ralph 0. Meyer,· Section Leader Reactor Fue 1 s Core Performance Branch Division of Systems Saf~ty

·~--~-----------------------------------------------------------------------r 1 i .!

I I

'i ; I

I I

·.I I I I

!

.· i ;

CORE OmGE

A. Introduction

for the usual analysis of hypo~~etic~l ac:icents~ initial core

conditions are assumed and consequences are calculated. ihis would

i:webe c:::::1!;( !he:-:::al-h;.r~rluii'- cllculations and fuei bena•1ior anai;.rses.

At ihree Mile Island, hcwever, sc~e of t~e conse~uences are known (i.e.,

scme info~at1on on fission prccluct rele!se, hydrosen gener!tfon, and

instr~nt readings is available), so we will use Mreverse engineeringM

as our principal method of back1ng out an assass~nt of core da~~ge .

~e start with ~~e asstmction that !~e core was ~ncovered and allowed

to heat up for significJnt ~eriods of ti=e. Figure 1 shews the

system pressure histor; for ~arch 28, ~hich includes three periods of

significant uncoverJ. ihe periods of uncoverJ correspond approximately

wf~~ ~~e major ~eriods when sys~~ pressure ~as below the saturation

pressure. We will ass1.1111!! that the first core uncover,. began shortly af~:-

92 minutes into the ac;icent at whfc~ tfme excore fen ~~ambers shew a response

spike corresponding to the loss of water shielding. Al t~ough the ~ later

per1ods of uncover} may have produced additional core damage, we will

focus on the first period bec1use decay heat was larger then and because

that period produced ~ie lar;e radiation instrument reading (at i50 ~inu:es)

fn ~ie contafn~~nt indicating ~Jor fuel daoage.

Because the fuel damage to !:e d1scussed below ~s so e:t~~nsbe, we •~til l

- ·· - ·- --· _ .. ·------------- ---·----·:-- ·-~- ----- ·

.· :

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

J

..

,I

g .... N

g .., N

Q c -N

e c c;\

8 .... ~sd '::Hnss::cd

g .., Q , c

::\

------------·

·:----~--------------------------~--------------------------------------~-~~ I -i ; • I . i ., i ' I : i I j I

conclude without d~~nstrat1on that vir:ually all of t~e fuel rods in the

core failed in the sense of exper~encins defec!s 1ar;e enough to r!lease

gas. Fu~~e~re, the rods probably failed by a lOCA-like ~allcon~n~-anc-

of ~ailure is ?robably immaterial.

~s a point of reference, iable I 1 ists r..elting !err.cera:ures of t.~e •tar!ous

materials used in ~~e fuel syst~~.

B. F•Je~

:.:.;J~ ~r:cuc: ana hyorogen measu~T.ents a: iMI-2 give important clues

about the condition of the fuel rods. We will deal wi~~ fission ?rccuc:

releases first.

Air and water samples containing fission produc!s have ~en analyzed.

While we have analy:ed bot.~ fer indications of fuel condit~ons, we have

concentrated on the Xe-133 concentration in the air sample. ihis

isotope was sele~-ed for analys~s fer sever!l reasons: (a) it is a ncble

gas and will not react, plate out or ccneense, (b) it has a relativeiy

lcng half life (5.29 days) and a high production rlte (6.3 atcms per 100

ffssionst and therefore w\11 be abundant !hus reducing measure!r.ent errors.

and (c) fission produc! release corr!lat!cns are mu~; ~ette~ es:lbl~shea

for noble gases t.~an for ot.~er fissicn products.

aettis (3APL) has evaluated the Xe-133 activity and concluded it is

equivalent to 31: of ~~e total core 1nventorJ. We have indecenden:ly :hec~ed -----···- ---·--

~~is calculation (but, of course, not t.~e samcle activity) and agree (~l. S!).

. • .. . ---.. 'f I

I !' • ! I . . i ' • j

i TJlELE 1. ~TUJj TE•mATI!FS . I !

f"AlEIJlL lEifEFAlU~ °C ffit'ff?AilJREI CF

~ 20C6 5E

Zlft-4 1.51 33E2

mGz 2715 4919 .. um~ns 12f:D-12SO 2..~23l:6 .

3t4$ 1359-1421 ~:{}-2..c::g)

Pt..1Jr'&4C 2Im 3€86

Pli-IN.fll EOO 1472

L~~· ZiS1 4~82

• T·o«l fuel assemb)fes contstned gado11n1a test rods.

-------.__...--..-_________ __ .. -- - -----·~·---

I ... .

Fission products including gases are no~aliy retJined ~Y ~'te uo2 pellets_. A nor.nal pelle: release to ~'te ~~el red int!rnal voidage fs

only· l or 2: (even for a successfully :e~inated LO~} so ~'tat a 30:

release indicates additional release frem fuel pellets not just a

re 1 use of the gaj) ac-;i'/1 ty •

F~el pellet releases are st~ngly ~e~e~~en-; en ~~cera:~re. and

r!gure 2 shews a correlat!en of release versus tempe~at~re for Xe-133

(f~ a recent ~IS-5.~ draf: standard). ihe c~rrelat! cn, howeve~.

is for stesdy-state releases and~ are dealing with a transient.

Fu~,er errors are ~~ssible tecause of kinetics changes due to

oxidation to U.;03 or u3o3• Ne•1er-;.'teless, 1: !s a l"!ascnable

apprcxtma:tcn and is consistent with recent shcr<;-t!rne anneal ing ex­

peri~~nts (private cemmunicat!on J-l0-79, R. A. Loren:! O~IL) and

earlier annealing work (G.~. Parker et al., O~IL-3981 -See a:!ac~~nt A).

Parker heated irradiated Sam?les in a furnace fer :.: hours. ihe

samples had burnups ranging f~ tr!ce to ~000 MWd/t (atout the s~~ as

iMI-2). Parter ~~asured releases of about:: at 16C0°C, 15: at 1200°~

and 40: at 2C00°C wi~1 an uncer+.afnty of about a fac:~r of 2 in

release . ihese exper!:~1tal releases for conditions roughly similar Cll'e in

t3 ~I-2, ~ut for different isot:pes,Afafr agre~ent ~1~1 Figure 2.

Ustng Ffgul"! 2 ·...e coul~ c::nc1uc:e that (a) :O'te 'fuel ·~as ·~n~f:~iJ hea-;eti

(unifor.n fn ax!a1 and radial directions) to !bcut 17:v°C, or (~}

20: ~f ~~e fuel ~eltad ~niie ;o: r!~ined ~eic~ 1200°c, or (c)

any in~tr.nediate cond~:~cn existed. geclus~ of :he ~ere uncover1 secue~ce, 3·l6l·ht~

.. ..

w U1 a: w _J w c::: (Jj

a: Q)

z 0 -(f) U1 ..... u_

t"") t"J -I

CD X

~\----------~------------------------~--~~ • • • • . • • I I • I • • .. • . • .

' . ' . • • • .

' . . ·•. . . '• .

' c ··. '• •, ., '• .. ... .. ··. ., ·-.. ..

······ ... ·· ..

'· .. •, . . .. ·. \ .

' .

~-,..o

' •, ~o-S ·. .. • . • . . . . ~Lg : .-r:;

·.•

the fuel rods probably did not heat up unifor.nly fn the axial

direction. It is reasonable, however, to treat t~e fuel rods as.

1sot~e~al in the radial direction because of the low heat flux.

Figure 3 illustrates thfs point with a comparison of a full-power

radial ~~pera!ure profile and a decay-heat-power t~peratur! profile.

ihere are physical limits on how hot ~~e fuel can ;et during the

periods of core uncover/ because the fuel reds have a lar;e

heat capacity and a low heat generation rate. If one ass~s

zero heat re!:'.Q'ta1 (!his would produce the most rapid heatup rate

possible) during ~~~ first period of uncovery, the heatup rate is

still fairly slow. Ff;ure 4 shows the adiabatic t~perature increase

with time for the peak-power axial location, for the low-power ends of the

reds, and for the avera;e location. Since there ~us! have been some

heat removal thus further slowing dawn the t~perature rise, pellet

temperatures prcbably did not reach the ~~lting point. Figure 5

shows the tem~erature ~ian;es wi~i time for a surface heat transfer

coefficient of 0.5 BTU/hr-ft2-°F, which 1s a verJ small •talue.

----·--- -·- ·- ·- - ·- - --· ·- The results o~ te!:lpe~~:Ure ~~s:_:1~u~1~n _a:e.! th!~.~?'::• not con~lusive.

I~ is unlikely ~iat fuel temperatures were uniform and no lcwer ~ian

liS0°C, and it is also unlikely ~iat any fuel {uo2) me1!~ng t~ok ~lace.

temceratures.

Oxidation of Zircaloy ~Y St!!m and ~ie attendant dec:m~osft1on of ~a:ar

provided the :najor source of hydr:gen in the ~I-Z •tessel and c:ntai rn::e~t .

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

. . . .

a... l:: w f-

_J w ::) L...

_J cr: -Cl cc c:::

11'1 r-------------------------~---------------re

~61 ;[il rn' r o) ;::; U; ,.r1flJ I \ . .

i I 0

l

•

• •' [I_ I • t l I. @r;.' I" I '.1 I\; ,., 'I 1 ii J lu...., l.J ~\J ~.iu ;_;,

01

~I

-0 tw c::: t..: > 0 (.)

w c:: c (.)

..... 0 0 (.) (.) c: I ....

en .... c zc.. ~en -c:: tJ=:l c.JC a:: I

t.Jlf) c:: • ~-

• I c a w -\:: -..

a.. $ ~ ttt ...... ~~~~ a: -taJ t"'fCI<D

~ X OHQ~ I I I

CJ D04

-= ..... ...... -cr i! a:l -a: ... ..... c a:

..

ClaiE CIIZ at D1 a l:J ~J 3WUlAG il Gl "t:nd

~~r~fRl r'r0 f ~ • 1: ((0 r I ru~~ ~ f::\ n ~·18170 ,_llj JL L! l~ L 'u U\.! ~

... . ....... --- ..... - . .,

~· ..

«1.

~ cr LtJ ttt ~

0 ~~~== -C"f~CD en ..... ~ 1-' Q)CQ ...

a: I I I D 0 .. -cc -2: a: -~

~ c a: -... >-..J a:: cr t&J z

rfJ rm~m @~LL ,~ft~£1 ~4CJl70

--- --· - .-... - .. __ .._._ . ..

- -- -------------------------------------------------------­..

The Contlinment Systems 3ran~~ has estimated the amount of hydrogen

~resent in ~~e plant (Attlc~ment 3) after the periods of core uncoverJ

that caused fuel damage. ihey included ~ounts (a) ccns~ed by the hydrogen

explosion (225 lb mole). (b) ~aining in contlir.~ent after the exolosion

(80 lb ~ole}, and (c) in ~~e primarJ syst~~ bubble (76 lb mole), which

was correc!!~ for radiolysis.

Ccmparing ~~e above amounts wi~~ ~,e total ~~unt of hydrogen tha:.cculd

have been ~roduced if all of the Zfrclloy in the fueled re~fon reacted

~i~, water, we get 41~. ~s with the temperatures, an amb~gui ty exists.

ihis could ~ean that (a) about ~: of ~'le cladding wall thickness is

uniformly oxidized throughout the core. or (b) jQ~ of ~,e fueled region

of ~ie cere has fully oxidi:ed cladding. or (c) any inte~ediate - . ccndit1cn exists.

Figure 6 shows ~ie time re~uired for total wall thickness oxidation

as a function of tem~erature (Cathcar:-Pawel correlation) . It is

clear from Figure o ~,at cemplete oxidation is possible fn cladding

se;r.ents that reached t~peratures of around 2000°C during the period

of core uncover] . It is also clear from Figure 6 ~~at all of the cladding

did not ex~erience sustained temperatures of around 1750°C else

it ·--ould all have oxidf:ed. ihis is fur:.'ler e'l1dence tha: fuel :ar.~oerlt:Jr'!s

were not uni for.n throughout ~ie core, and that tem~er!:Ur'!s locJliy ·~re

1er1 i\1 gh .

Based on early esti~:~ates !)y the )nalysis Sran~, of core uncoverJ, ·..e ·..,il l

ass~e s1mplffie~ uncoverJ histor ies shown in Ffgures 7 and 3 for the

r

-----·------ ·-- ...... - - ·-.•. , ............ -

..

w E -E-

z 0 -E-cc 0 -X 0

...J a: E-Cl E-

: . . . . , . . • . . . . ocCo1

.. • , . • • • • • I • • • • • . :z , , c • -c.J . • t-c:: .

• (.)::) : a::t- , wa:: , • C::Ej .

I , C!lC.. • , Z%: • -w ,

• s~- • .

I • Cl:\~ , -C : e,:)z , .

I •c I . ~- ,

I .......... I

Oc.J I

: t-Z . ::) , . a"- I :

"-a: , . , I

~~ I , ,

I • .... , I , ,

•' • , .·• , ,

• • . · . •' . • , . . . . • . . . .

cCol COl at" C S:UnNI~J :;~Il

~

§

~

-~

0 r-§ •

- C!l t.J c

§~ - :s .... a: c:: s w .. ~

-%: u t-

r-~

-~

8 --§

·s-

. 1.::1 Ul -\5 I!:

. ... ~<...· ' .. p··

...i~.o~ 1•.1

..

.. ..

w U1 a: u

)-o

a::: w > a C,.) z :::l

w a::: a u

. -----·-----------------------------------------------··-····--·~

,.· • . • • . • . • . • .. • . • .

... ··· ,

Qlot

-··' .·· ,•' .·' •' ... ··

... ···' .. ···· . . ·· .... ···

.. ··· .. ·· . .··· ...• ··

otl cOt . (S:i~:Nll

. . ca og or 13A~ lN~iOOO ~~00

, .. •' ... ··

• cz

.·· .·· ,

0

-f-

.. ..

N

w (J')

cr.: (.)

>-0::: w > 0 c..J z :::::>

w 0::: 0 (.)

r-~,------------~----------------------------------------~M • . • • • • . • • • • • • ··-...... ·-· ......... ·-.... ·-·. ·----..... ··-. -· -· ....... ----...... .

.. .. .. -· .. .·· .. .. .. .. .. .. .. .. .. .. ..

' ' . CH czt cct

!S:i~:NI l

.. .. .. .. .. .. .. -· ..

' CS OS.

-· -·

lN\:fiGOO ::~00

-· .. -· -·

cz

-· .. -·

' 0

0 .---..------------------------------ · ·· · --· -

. i 0 .

••

following calc~lation. Fuel that is covered will be considered to ~e

cold (i.e •• no cladding oxidation). F~el that is uncovered will ~e

allowed to heat up; fuel that hea:s up will be given a he!t t~3nsfer

coefficient that is adjusted such that the total integrated oxidation 1s

40:!. These c!lc:.a1ations gi •te the oxida~~on dfstri!:ut~cns shc~~n in

Figures 9 !nd 10. and these dfstribut~cns are insensi:i•t'!! t::l many of the

assurnpticns that were ~~de. Figures 9 and .lO thus are more probable

distribut~nns than lao: oxidation over 40: of the c::~re or JQ~ oxidation

over lao: of ~~e core.

Figure 11 is a recent best-esti~te embrfttlement correlation (~ssner

et al., ~IL) ~~at shows high-t~~perat:.are frsgmentation of quenched

tubes at about Ja: oxidation. Using this correlation, Figures 9 and 10

indicate that a fra~ented region of about 5 ft. in height exists near

the top of the core. It may well ~e right at the top of the core as a

result of simplifications in our analysis. rn any e•tent, at least

4 to 6 ft. of intact (but partially oxfdi:ed) fuel rods remain standing

at the bot:Cm of the core.

Figure 12 :ho•NS fra~~nted Zircaloy cladding after oxidation in a

simulated-LC~ test. Kassner (AtiL-78-ZS and ANL-78-49 reports that at

high t~'llperatures (> 12S0°C) many frac;ments are pMduced ;o,ne:-e!s at

lcwer temperat:.are the red r.4y simply break into ~o pieces. inas~uch

as nH-2 te.rncerat:.ares ·~re higher than 12!:0°C and oxication ',ot!S Si!'tel"!,

~all fra~nts of ~~e si:e shewn in Ff~~r~ 12 sr.c~i~ ~e !xpec:~d along ~fth

larger tuce-like ~ieces.

.; ,. ·----- ...

w (f)

CI: c.J

z 0 .,_ E­cc a ~

X 0

w 0::: 0 C,)

~ .'t •• -

·.

~------------------------------------------.,-N : • : .. .. · ... -·· ~= . . . . ................... . ~ .................. --. ....------ .

....... _______ .

-··-···---­···-····--·······-. .. ·-----.. ..

. . . a·a g•o ~·a z·~

OJZIOIXO SNIGCbi~ JO NO!!~~~~

·-· ... ·. . . · . •

c

• 0

~ ~· c.J • . •

"'

~rt

.= a

C)

e : ~ t.:..

. , .

.. : : .• .

! • •

!:!481'1'9

..

N

w U1 a: u

z 0 ~

E­cr: Cl ....... X 0

w 0::: 0 (.J

r-------------------------------------------~N . I I • . • • • • ,

• • • / .·· . .

~=

i-2

-------------·-.-· ·················· .

... ... .. .... .... ... , ....

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

a~ 9~ ~~ z~ 03Z!OIXO SNIOGY1~ jQ NO!!:~~J

·-.. .. .. . • • • . • .

I • • • • I . I I

en

~a

~ t.J t.J

"'~ % 0 -i.CQ 1-a: > t.J ..J

rout t.J

t.J . ~ 0

~· c.J

"""' ~N

~"--

.= 0

c:i ....., w

~ tZ:

S4.8.!b0

(c:>J

? ----_;"' ~--!~~I ~-·-:-.. --- - ...... Lo J f' J :....=:::J .---, --~. "'- ·- c::::::> c;;::::--] '--- j C....::.:: ~--;;; t;-;:-.... __,__. c_::;..J

[j==J

c.o ;;. ~ ~· ... Ch :--·It

i! 0 ... ~ IIJ a: CJ ~ 0

j u

~ .... ;J > 5 0 w

FAILURE MAP FOR ZIRCALOY-4 CLADDING DY THERMAL SHOCK OR NORMAL HANDLING

ISOTIIERMAL OXIDATION TEJ,tpERATtiRE ret

000 1000 1100 1200 1300 1400 1500 100r----r----r----r----r----r----r----r----l-----r----r----r----,

0 CLADDING INTACT

• e FAILED ON QUENCIIING

/'\0

I) SURVIVED OtlENCIIING BUT FAILED DURINfi IIANDLING

'T 0

60 - 0 0

601-0 0 .. 0

, .. •

~ 40 t- 17% LIMIT 0

~ . B> og . -o

C) ., 301- 0 0> 0 ~ 0 0

o 0 oo o \Y o •••--a8 _8:..sif--o-~.o~:cs>~-o'%""o ~..no Jh

- UOO CD .m 0{) tO ()(Q Or'Q> 0 'V 0 (S> Q lJ!'I 0\.. 0 . O • t_ 1477 K LIMIT 101- u '0 • . 0 •

• •

oO 6>o

cf}o

•

1 I I I I 0 ···- · UiOO 1700

ISOTIIERMAL OXIDATION TEMrERATURE t•Kt

Fra~ 11.

..

..

..

..

n3.m.:: Zlzcalay-l Cb~dLag &iter 't'hen:ul-tllcclc :'all• 111• SllovLag tccac1oa of T!wmoc:oupla ~c Praduced tllc Te:nF1tatae-¥t-'tlm• Carica La n5• m.:o. ANt. ~•!· ~. 306-71-223.

F!G!J\E 12.

------------

F1J. m.u Zlzuloy-l C::ac!dLag aft£: ~en:ul-thoclc !1!1• we Sl:lnoliag t.oac1oa of numocoupla ~C i'rad111:cd tl:c ':cm~u~rurc-n-T!:::c C:~na 1.::1

:If. llL!l. ANL ~•i· .~:::s-2!:_~': -··

.. ·-

_ ....

·;.

·.-: . ~ .. ..

. .

Fuel pellets normally crack during operation and crack healing can occur

at power. Ffgure 13 is a typical ex~ple of a cra:ked peliet. Quenching

during core flooding may also prcmote fra~enting of ~~e pellets.

Se•t!~ly fr!~nted regions are c=conly seen in fuel pellets as a

~sult of,.ex::- . _ . :. - ~ : . .. =.:~~ :-:..: : ~ : ;;.; : ·-::~-:::rs ~ ?cwdered

regions fn fuel pellets have also been seen in some ?SF tests,-but

t~ese :!!':j • - · • : : · ,·' · .: ~ :' '' 'J.::-:' ~i;!i ~:·..t::-s (>20 kw/'ft) and

verJ St!!p temperature gradients unljke the low-power unffor.n (radial)

tenper!ture TT-U-2 fuel. Therefore we would expect the TI-II-2 fuel to !::e

in ~~ i ; ;~~:~:--si:e granules and lar;er pieces including whole pellets.

C. ~~fueled Ccmoonents (Cont~l rods. aufde tubes. etc. )

Figures 14 through 17- show ~~e cont~l rods, the burnable poison roes,

the power shaping rods, and ~~e central instrument tute. A;l of these

rods and the instrument> tube are inserted into Zircaloy guide tut:es in the

fuel ass~~ly. ihe materials of ~hich ~~ese components are mace are

indicated on ~~e figures.

An important clue about ~~e condition of unfueled ccmponents fs-~rovided

from instrument readings. The fact that all 52 thermocouples 'NOrXed .,.csf

;.";roug:.out t.ie accident anci"ccnt!nue to gf'le c:-edfble infor:r.ation

suggests that a central tubular struc:!.lra 1 ::ll!!llber surti'lfed. !t is

:~;::~; :: c:r.c i4c~ :~a: ! l i Z!r:Jloy ;uice tut:es aiso surtive~. :u:

this may not be the clse s!nce ~~e ~~ermocouple is ~11 ~rotec~d ~Y

mul:fple barriers •

.! .· .

.. J

·.·

r_~ • ( . )

~

.. .

CCIII'UHO

-{-~ D_.,_s.s-r) __

Nllr.WCH .&aSCMM WAnltiAL-

-(I~;:_ It: I' .. • S"_G.J)

e-'..lfTWO&.ItCO- • -

-r

-

I! I !

ii iiii

II 1111

-._... -

.------o. "~ ,,~. 7D4-~S

(o.4111;, .-.. . ~o)"'"

o . ., '-; ;.o. >t o. r;.;, o D

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

..

.. .

TCI' VllW

FIGURE ]5,

~· ·- . 94~lb6

.. .

TO# vtiW

IUIUUIU POISON- · IUTUIAI.

( 1.,. C. .:!..IA/z. 01) .

FtG.JRE 16.

-

3ur::able ?o!.se:: :\cd .Use:!: l7

. Fixed SPND assembly cross section

FtG.:RS V.

INCONEL OVERSHEATH

CALIBRATION TUBE

BACKGROU~{D

DEiECTOR

.-JJ-+- SELF· P 0 WERE D NEUTRON DETECTOR (7)

-. · --·· ·-· ·---

. ! •

. I:IEL has made calculations of gufc!e tube tem1=eratures by parm!trically

var1fng heat transfer conditions (see Attac~ent C). Their results show

~~at ;uice tube ~eratures lag ~~e fuel rod temperatures by only

about 20°F. 9abcock ~ ~11cox has ~erformed similar cal~Jlaticns ar.d

concluded that there is a muc.., larger spre!d in temperatures. ~e

believe the I:IEl. calculations are :r.ore nearly coM"!ct and t!lat

temperatures of unfueled cemponents ~re close to fuel rod :e~oeratures.

Since fuel rod temper1tures are !:eli eve~ to ha•1e exce~ced 1 i:0°C

in the .hot region of the core, then in ~~at region (a) Ag-In-Cd and its

stainless steel cladding ~uld have ~~lted, (b) Inconel spacer grids

would have melted, (c) Zircaloy guide tubes .,ould ha'le oxidi:ed,

and (d) Zircaloy cladding of ~'1e burnable poison rods ·~uld ha'le

oxidized.

In the cooler parts of the core below about the~ to 6 ft. elevation,

we would expect all unfueled ccmponents to be intact, alt~ough perhaps

damaged, Just as the fuel reds are expected to be intact. Control

rod se~ents could have cnly fallen about 3 inches if severed ~Y melt~ng

in the hot region, and the Ag-In-Cd absc~er should be in ~lac!

because it is an insoluble metal. A1~'1ough ~.,e burnabie poison rods

would also be expected to be in place, their poison is probably

lost; boron is ~,cwn :o le!c~ cut of 3~C-A1 2o3 pellets ~nen e~:osed

to water in 1 rldiat1on environ~ent.

jJ

0. StM!ARY

1'1any or all fuel rods may ha•te t:allocned an<! n.:pt:.n·ed, ~ut !his

~<!e of initial <!efec~~ng is prcbably irr!ievant in light of later

more extensive <!~~se.

In ~~e hot upper centr1l region of the c~re, fuel :~perat~res prcba~11

e:cceeded 17S0°C rele!s1ng large quantH:ies cf fiss~c·n ;:r:<!uc:l; acou~

30% of the total core inventc~/ of noble gases ~as rele!sed •

.\bout .;o: of the Ziraloy cla<!ding re!c~ed '<'lit."' ·..,ater. i'nfs region

of severe cxic!a:ion ·..,as lc.:!l i:ed a!:o·1~ t~e ~ to 6ft eie'lat1cn and

may not have included peripheral bun<!les. ihe severely oxidi:ed fuei

probably frasmented int~ pie~es ranging frcm milli~eter si:e t~ ~hole

sec:ions· of rods.

The temperature of unfueled components lag~ed :he temper1ture of fuel --·------ --·-- · ·- -- .. -----rods by only about zooF so ~~at they also e:c;erienced temperatures --- . ··--··--· . ·---.. . .. above about 17cooc. Conse.,uently, fn the hot region of the c~r~ ·

Zircaloy components should ha•1e oxidi:ed, and c:rnponents '4itll Inconei,

stainless st~l. and .\g-In-C:t should have :netted. aecause of many la:ters

of protection, ~~e ~"'e~~couple tubes have sur1ived even in the damage~

. ~re region, al~"'ough :he outer sheath of the instr~ent tube ~a:t !:e

badly damaged.

Nearly all of the broken and oxidi:ed fuel de~ris shcuid remain :r!c~ed

in :he ueper core region :ecause ~"'e ~cper enc! f~::ings have a gril lage

that ·..auld act as a sc:'!en. Fu~~e~cre, :he c~cac:icn of fuel c!e~ris

34.81!10

1s limited because it is fabricated wf~~ a packing frac~ion of about

46: and the theoretical maxi~ packing fraction (for a bed of

spherical particles) is only a~out 63: . It i s ~erJ likely .that fuel

debris are also ~rapped in scme mfxfng cups (See Figure 18) c:ntributfng

to non-unifo~ thermocouple readings .

An earlier estimate of fuel damage in ~I-2 was ma~e at NRC by

Rubens~ein, ~eyer, iokar. and Johnstcn. ihat estimate is in general

agreesr:ent with the · present estimate although our current e•tal ua~ion

is more refined . A memorandum sumnari:fng the earlier est~mate is

atta~~ed as Attachment 0.

E. Recommendations

Reactor fuel is rugged, and it is unlikely that limits for natural

circulation conditions will be related to fuel behavio~. ihe general

criterion with regard to the fuel should be that additional Zirclloy

oxidation and fission gas release should be a•10ided .

Sfgnificlnt oxidation rates do not occur until 900 or 1000°C (See Figure 5}.

Significant fission gas releases do not oe:ur until even higher ~eratures

(See Figure 2). rnese t~~eratures should be avoided in ~,e (~latively)

undamaged regions of the nu-2 core, !lut these t!mperatures !T'! 50

high ~~~ t ot~er Hrni ts will pr,bab 1 y prevail.

3y now the adiabatic heatup rate is l ew (See rfgure 19) and !mpie t~~~ ~i l l be

~rovided to detec: fission gas or hydr:gen releases. T1~~ ! : : ·: ,

on-1fne met.',ods of sue:-. detaction, if feasibie, should gi•1e ade..:;uate

rfJ~ r~r~ . -- ·-®t o:n~rn-+o"'+Tm~r 0........,. '"+fl _____ _

~ u"'[j"[j ~u wwLS . - ---

.. ..

---- ____ ,.,. ... • • DA • •

·-. -.-:.- ..... ,' ~'fl:. ....

w ••

.L:>3rsns

" I )

-u s

eQ - ·

----·-·· .. --- -

-...

U1 -..J -0::: CL a: c::: 0 tL..

CL ::l ~ a: w :I:

(.J -~ a: CD a: -Cl a:

• t • • •

CXISZ CIXlZ CCSI CCOt C~

t:J ·s~cu ~nr=~c:l-G.!. -r.:n.:

~moooo ®~t~~u~&~

en c:::: ::J g

111--t.J ::: -

, c: c

A discussion of instrument ~sponses relevant ~ fuel !:eha•tior 'ltU

held .,1th a group of fuel e.'<;:erts from across the industry.

stmnarJ of those discussions was prepared by ~ . '1. Johnston and

is att3ched as ~t!~~~~ent E. One consensus of :hat grcup was

that in-core ther.r.occuple readfrgs ~hould be recorded continuously .

~ re~endation fer su~, datl recording 'ltiS made and is at:ached as

At:ac.~ent F.

rP 00 (ID [~ @ j~ L ~~u ~Jtj~

··-------------- ·----------