ML20138P154

Rev 1 to TMI-1 Nuclear Generating Station Natural Circulation Cooldown Analysis W/O Reactor Vessel Upper Head Void Formation, Topical Rept

ML20138P154
Person / Time
Site:	Crystal River, 05000000, Crane
Issue date:	07/02/1985
From:	Bucheit D, Irani A, Lynches P ENERGY, INC., GENERAL PUBLIC UTILITIES CORP.
To:
Shared Package
ML20138P139	List: 3F1285-12, Confirms 850125 Notification to Adopt Gpu Approach to Analyzing Natural Circulation Cooldown (Generic Ltr 81-21).TMI Topical Rept on Subj Encl ML20138P154
References
017, 017-R01, 17, 17-R1, NUDOCS 8512240333
Download: ML20138P154 (45)

Text

{{#Wiki_filter:_ y L. + myD: tn _.g; g; < - -x. f k ', . 3gg;j k J 3~. Rev. I b._ "= . Page il e fi: .n. y ..[ (TMI-l-lL 1 NUCLEAR GENERATING STATION U . NATURAL CIRCULATION ~COOLDOWN ANALYSIS' i. W WITHOUT-REACTOR VESSEL UPPER ~ HEAD, VOID FORMATION 1 -it. TOPICAL REPORT 017 REVISION 1

s.

w i ig; - PROJECT;N0i l540 '41274' kp#- 'T ~ ~ D. Ch Bucheit (Energy,Inc.). ~' (/A. A.-'Irani-7(GPU Nuclear)( pF' PfW. Lynches. '(GPU1 Nuclear) (s t jf _

July 2,."1985i

.t p )APPRdVALS-N: { ; ji 47 9 y is" [7

5ECTION MANAGER.

'OATE Y g 7 g~ pg

DEPARTMENT MANAGER..

DATE s y r512240333:851220 - f$ iPDR-ADOCK'. 05000302. _ {C:- ' P, ..PDR l g4 m

u. =

~

f(l Y I , ? ~ u$a, .m %n : y. + s . ; 6;,

n

. ge ~ + c ,..f. '~ A m og.

TR.017' 2 '.

s.t.; - y" ' #..- - Rev.1 Pace la: e k r

Q s

1. 1 ll DISCLAIMER OF RESPONSIBILITY ~ ~ W s eThisfdocumentsas; preparc.d ;by or for?the SPUN Corporation. L Neither:

GPUN Corporation;nor any..of!the~ contributors' to this document:-

~ 7 ,~ E -. a'.1 Makes a'nyLwarranty or creprssentation,= express' or implied, with 4 ~ ' respect :to theiaccuracy, completeness,:or usefulness of-the - A

information. contained.in this' ~ document,' or that,the~ use' of any

~ < M, R- .information ' disclosed. in, this (document may not. infringe' privatelyj ~ " owned rights;' orb ' ^ $i... : ssumeslany responsibility foriliabilityTor damage of any. kind which i s may result fromjthe:us'e of' any. information disclosed in this ^ a

document.

a 1 r ; ]-. d ? s f 9 E .:4 s .e . -. =, - r, .ii; ( >..,c.; .6 4 + u e } 4 I'.- Eh b a

• '~ -t

~, 1 ? -..h.. qQ g e 2 l r t 4% [_ I L + .n N ga -t [ 4 f h .x 1 + (. L.j[9;7. -' ;' fi;' 'yz , 4-..

w+-9= c,.. - m, N$c J:* c '1r:. . i.c c 3.

n Le t '

• a.

~~ + 'TR 017 M- ,a Rev. 1. m 4 y-E. I, g; PageL.2. ~ -' ap -TASLE OF CONTENTS + 4 ~ PAGE: n jl;0. INTRODUCTION'- 4-7 '1.1['Backgroundj 4 .:s gl 41".2 MNRC Concerns and' R'equirements.- .4 o, i_'f 14, ?l.3E TM'I-l ?N'atural? CirculationiCooling Procedure OP-1102-16

5.

i il~.4'-THI-l Natural' Circulation'Ccold sn. Analysis 7

Hithout-Reactor Vessel Upper; Head :Vold Formation'

-7 m 7 f. ~ 8 ~- 'I ?i2.0 t i REACTOR 4 VESSEL; UPPER HEAD COOLDOWN-PROCESSES ~; ;.. NN. ' 3.0 LHEATING6 MODEL OF THI-I UPPER HEAD ~ '14-E .s-3.1 HEATING 6 Code' -14" ,s

H'E TING 6 Limi tations-

' 141 63.l.1) 1 ~ +. 4 4 - j3.2JUpper; Head Model" 15 h: 13.3:-hssumptions: 1 15 ~ ^ m a" ~ }3 M Steady /Sfate Heatt ransfer- ~ ~ 22 T

• s m

s , 3 3;5 - fInt tialiCon'di fiods, '23 s: y 3'6) 8.ounda'ryLConditions;

24

= M ~ 74.0? ANALYSIS RESULTS-29 24;l.VolOme Averaging;of the Top Foot of:the Vessel Head 29 A 4.2IT 'CS'Cooldown at 10F'/HR-30 - ~ R 3 '.3 ?RCS-Cooldown at 50F'/HR 33- ~ 4 sy,. 15.0 J0PERATIONAL. GUIDELINES 36' Cf. i

6.0 : CONCLUSIONS-41 g;

4-17.0;iREFERENCES- -42 2 ~ fi

ppb 4 DIX'A2 43 A

2 E .( n 4 N _ - ~ '

+p= i r f(? ' g< g,

m 1

[, ^ z.. .TR 017' Rev. 1 .Page'3 ~ ^

SUMMARY

s ~ LThis' report contains an analysis for THI-l which simulates a natural circulationlcooldown without r.eactor v'essel upper head void formation.' A two-dimensional)cylindricalimodel of the THI-1 RV upper head was. developed and ~ s fused asLthe basis for the. thermal analysis. Computer aided an'alyses_were-performed; by means-of HEATING 6.(a-multi-dimensional : generalized heat conduction' code)ito demonstrate the thermal response of the RV head to RCS - ~

cooldown rates of 110F'/hr'and 50F*/hr.. The results-indicate that for an RCS cooldowniof 10F'/hr to 204*F,11t:would require _22 hours to reduce the coolant

temperature in.the. reactor vessel head to:429'F*. On the other hand, for an

. ~.7 .RCS'cooldown of 50F'/hri.to 204*Fs-the time required to cool the reactor vessel- ~ head to'429'F* is':on the' order of 7: hours. The results of these analyses may be : correlated into pressure versus [ temperature curves which can. be incorporated into' existing cooldown procedures. L*: 429'F is the saturation temperature which corresponds to the decay heat removal system. cut-in-point - 325 psig RCS pressure. r' 2 x w /^

p, w, _. y y y ~y {\\ ^ i TR 017 ,6 Rev. 1 Page // p .-1;0 'I'NTRODUCTION~ ~ v. 1j;j =BACKGROUNDL OnlJuneLil,1980, the St. Lucie reactor was shutdown due to alloss of -p . component cooling water to the reactor coolant pump: seals which also ] required shutdown"of. the _ reactor coolant pumps.- -The cooldown ~was x ~ --4

accomplishef.by' natural' circul'ation. ! At< approximately four; hours'into (the-event,' charg1'ng flow, which was. initiall'y being divided between the.

s* . cold legs and the ' auxiliary pressurizer-spray, was. diverted entirely to - s ~ ~ f the auxiliary sprayL to. enhance the depressurization and reduce-the l system q 6} [ , pressure on the' pump! seals. AtLthat: time, abnormally rapid-increases in, pressurizer levelL were observed. Detailed evaluation and follow-up y s - ' y analyses?have.; indicated that' the increases in pressurizer. level i [ indication werei the results of steam void formation in' the upper head h' ~ Tregion..ofithe-reactor vessel; The' steam void-was produced at the' instant ~ theLsystem pressure fdropped belcw the saturation pressure corresponding. - to th'e' upper' head coolant temperature.. Under con'ditions of natural ' circulation, ccolant..in-the. upper head 1s-expected to be in-poor thermal ~ m.. .lcommuritcation with coolant in the 'plsnum. Consequently, the coolant-ctemperaturesLin thelRV~he'ad will' tend to remain elevated above ~ itempefatures1 indicated by hot and cold leg instrumentacion. L g-pE 1.2 'lNRC-Concerns and Reauirements. -Because'of-the-unexpected occurrence the void during the St. Lucie r Levent, f ajid Lthe failure of-the operators. to immediately recognize the void ~ L ~ 't .u w-

1. r-m wa g

g 33;. ~ =, ~ } s.%'c<s,' 7 d ' t:e m' ' ~ TR 017. -Rev. 1 .r. 2 Page.5~. ~ fformati6n'and take corre~ctive: action, th'e questlen of.whether such void ~ i s a sformationll's~ properly accounted for in safety ~ analyses.has been an area Lof;concernLtoithe'NRC1 These concerns relate.to a) procedures and ~ (trainingtoenableoperatbrs'to' avoid;voidformation'(ifpossible),or I

recognize and~ properly react.tocreactor vessel hea'd voiding during s

I I ! natural circulation cooldown'and b) the possibility that significant head

voiding increases.the susceptibility of ~ the-plant'~to more'sericus laccident's. i n particular,' these. Issues' are1 contained in Reference 1 I

which requestsLthe:following.information: A ~ .v. [ 1. ..ATdetailed descriptionLof the. natural circulation _cooldown procedure ~ jand its.basist(it'sficuld include-guidance on. possibility,. ~ , ^, prevention ~,)and -mitigation of. upper head ' voiding and natural ~ circulation-interruption). p l:22:4 Demonstration by' analysis.or otherwise. that: ,w w a): ' Use of this ' procedure.will, not result;in upper. head voiding ,;g b); 'If voiding; occurs,: the prcicedure.lwill prevent any voiding at

the hot

; leg elevation-

-i . l.-3' 1 THI-l Natural Circulation Coolinc Procedure OP 1102 ~ Operating procedure'~ 1102-16' addresses RCS natural circulation ccoling for I THI-1Lin-wh!'chlanRCScooldownrategreaterthan10F'/hrbutlessthanfor y

equalLtoi50F'/hr i_siprescribed as~a means of preventing upper head vol'ing. iSection A.2.6 and;C.4.3 include guidance on possibility and d

JmitigatienTof upper head voiding as quoted below: y ,.4, I' .c .e .,,.S_, e aw " m --

yg s :. y- .= a g3 lQ a TR 017 ~ .Rev. 1 Page (o . Volds may occur int the Reacter Vessel head while m 3 .7 sdeplressurizingthe'RCSduetoheadwatertemperature . bel'ng higher..than RCS temperature. This.' condition may Jbe evidenced. by an increase in pressurizer level while 5 1 ' reducing RCS pressure even,though an adequate saturation- I' ' margin is indicated between. T., (hot leg temperaturet and'RCS pressure. Reference IE Circular 80-15. .w.. n s JIf void formation should occur-in the RV head, the head

1..

1 ubble -s'hould. be <condensad before centinuing the cooldown. b ~ i .Iffan RCP cannot be bumped,--then RCS pressure should be held constant or: slightly. higher 'untiltthe.-head bubble ' 1has condensed as. indicated by the return of pressure / ~ ~ centrol~to the pressurizer. ~ s

In

addition, the. procedure-will prevent any voiding at the hot leg

~ elevaticn by Steps A.2.5 and 5.2.5: x ~ Steam. voids..at the top of the hot legs can interrupt n natural circulation. T5is is prevented by establishing 3,4 ,[N and maintaining at least a 25*F subccoled margin afte'r k, Reactor Ccolant Pump trip. ( 9 1 s --r-7 f ' F t

W,- s:-: w Mcn. y. p. _, 3-4 s. < c w a,., _., y r M.,, n *:._ y, .,, 2 ) L l.?.h.. _ A ' . e{ - .L . g. ) A4

m*,

m, '~ gn_ n;; Rev.' 11 tm 4 [

a, Page-^9 j,.I s,

l,4.1T_T.HI-l Natural'Ctrculdtibn' Cboldown' Analvsis 'without Reacter Vessel -l--* R F,. _ _ " i k. _ 5 f -f y s-K, i.6 <# 7 dorer-Head Vold Formation; 7 ' ;!: 51niorder'. to --formu_lats procbdural re'quirements to' prevent-coolant flashing-t ~ . n. s ' in. t.he:reactorfvessel head,J at analysis was' performed to.' simulate a + I, inaturalicircUlationicoo' downf at THI-l ktwoWimensional: cylindrical i . 3 t ^. 2 ~ 8 w. $model;of-the: RVj upperthead wjts used 'as 'the basis for the. thermal analysisi ,y n hich5 utilized H'EATIIG6j(aimulti dimensional, generalized heat conduction w g {* e jf^ code). +, y ' . ll, 7 m M+ .% g -zy - gyAnaijfes 'w$re(performed forJRCS cooldown, rates of.10F*/hr and:50F'/hr to k .. p : ~-.. r .L jdetermineithe~minjraumctime.. required to reach the decay heat remcyal o $Yystem cut-_in.poiritf(325.lpsig' and RCSL t'emperature of 300*F). based en :the ^ i ', ? system designi:: !!n" order lio empicy, theidecay. heat removal systad, the .~ . 1 Ltemperature. of fluid Tin.the: vessel head mustzbe less than the saturation s I

l

~, 'G" temperat!ure.(429'F) Nhichicorrespdn'ds to 325 psig.' In :additlen, the RCS : q j# [temperatureHwa's:al'. lowed Lto dedline toL 204*F-;in.both cases. The results-Y ~ ~jhv Tof these f analysesiare{p' resented;in 5ection 4M. ~ ~ E*' /~ 's n , j'a. l' ? g a:W x. ~ ~~ ge, s _. fg ,p c. m m. 1 ;.ve p ff*? }' f jh g3, g .. c f, k -Y ; }_.j - 33 n. J[k:6 g' w .a ; q ~ 1_. } } -_ '?. N' ~ -.} k ' 7 g em g[ = . 'en f y y._ ?b A M -N--- -- -_' _' I '9'

~ ~. - - -~. . - - - ~ ~ ma n, s ^ c es ? s L fs;l j?%3_ ~ q/c _ 1 TR 017: "+' ~ s Page f Gi ~ Rev. 13 %.fa ' 9 .L'- e; w + 1210luREACTOR. VESSEL? UPPER'.HEADCOOLDOWN-PROCESSES' m,- - sf _

Duringl natural ccirculation,.the fluid in the RV upper head will remain

~ ~ rElatively?stagnint.s1nceitherreactor cooldnt system: loop flow rates will e. 1 - !b'e'significantly' smallerf th' n during forced circulation'.' The plenum S a ~ Scoverland : structural 4 compon.ents4 (shown on Figure-2-1). tend to' isolate the. ' g3V head fluid;f[omL.coola'n'tlin the;p1'enum.. Coolant [willienter the head 4 i.regl'oniatllowvelo' city.Lthrough.thdCRD: guide tubes which extend g_ ,m ... = ~ + approximately 20 inchesiabove:the ' plenum cover. Consequently,11ttle ~ . mixin~g is expected between entering coolant:andL the majority.of fluid in m.m - .L... f the: RV2 head:: dome. 'The : head. metal and water will cool. slowly.by.means N n\\

t"he1 heat' transfer proce'sses ' delineated below and illustrated on '

j-.,. V. P 'fl~... ~ .t ~

Figure 2-2. : Patterns of, fluid; circulation (which will likely result from, w

, 5the JeffectsLof' na'tural; convection cooling)' are depicted on Figure 2-l' ..~ .m w;,

g:

s n ~

e -

~

Thechea'd' cooling processes are

+- W ' Heat 1trans'fer lfrcm the exteriorJ surface of the mirror-insulation to g. ' ~ , ;, k icontainment is. considered tofoccur by means of free.cenvection and radiation. ~, .. ~- ? Heat' tr'ansfer across. the; three (3) inch ~ thick mirror: insulation is u

,yRQ J 7

Y ^ const_dered to'. occur: by means. of conduction. a s.g - 3 g Heat: transfer from the' exterior surfaces of-the vessel head through .;^,

the{ air space t'o the' inside surfaces of the. mirror insulation is m - '

' expected'to occur. by means of natural convection and radiation. p. "r p d b4 v b. I:g: %_& : fN+. %-

6~ f TR Oli : .Rv CoMNNENTS. - Y' ner.1 s a: ,,, 9 -s .f_. 4 8 ef A&lAT/hy +< ~U C. L .g I mruvmn n fin 7 .,__.,- =% : .J yn e 3__s-heemy e< / . ) f h< _l l ~ / ( gf o u _= pm j___r g*as 4 . ) J Cv [ J ')j c J -( 5 Raug -z f+; f

'f 1 --.

u g covee a _ = y C M im D. p / J S'Tilveme9i A, ,i. bh h,__ g ~n e ar g

(

4 ~ 3 m l tr D 4 -P"* /hr g J h f J ( l 1 "ime I ~ L t l7-f( / r s f* I*~ 4 sw~m 6

TR 017' RV NEniMar Tan'spae j,';; 1,o -6". 'O j 2b 4 L 40 1 3 y y$l1lt 5OIg y l l A ) I f' gM a l + g g-4, p 4 = _ t M* <aL'4 / J l,l.N l / p ' d'p N / ( ,N T d [ / ./ I7 i ? M Y st 7 ? \\ iV TEP Cf C&66 '/ / DRAWING LEGEND / @#,@,@,e -umva.- co we:nce / @.Racerkw f @ awauxx f- @. vmae abusun ?%x n w uss 1: f/GUEE 2-2 ,vsM

z., mg m-

, %+. y

s.s Wy f[

m;9[# -TR 017 ~~ 4

g g-Rev.

1.

f gg, ' W ~ .Page?ll p$Eh .s ,. 1>- x J 7mp;., t.; 7HeatTwillfbe Econ' ducted alongi he metal walls toward the CRD-housings;. t ~ q-1- y2s 11ocated-in th'e' vessel head. gg ( g - g i +g e. _ i f0' ~ ~ wheat from.the upper head wall.Lwill.;be. conducted along the vessel-m s ~ 1stielitowardVtheLcoolantiinlet which is71ocated in the annulus: f , _ f, imbe' tween the-vesseltwalls tandithe core -thermali shield.' : Heat will'- .g L v

ial'so be transferred from the1v'essel' walls to the cold 'and hot: leg -

s y

.+ ' m;n - u ' nozzles. % ~ - ' n a/ ...s 4', ' T Heatiwill?be5 transferred from the-air space to the CR0 housings by! -1 .7m p fmeans-ofinatural convection-; ~ } y y 2 s .?

During? the iccoldown'. of-~RV head metal,(heat will :be -transferred frcm

.n /In?radditien,: heat stil ~ be removed;frcm the~u'pper head water by ; ,~ lconvective heat transfer between the' ccolant and the CRD. lead screws. I f m s ( jM g 4 g <,- LHeatLtransfer betweenilayerstof upper? head water at different-A > ;e -.. 1 Ltemperaturestwilllinclude convective-' effects. The conditions under: ~ -wh'ith' convective heats transfer is expected are elaborated upon in ~ 1Section~-3gL y );; '^ f. -. .Inithe.-ishortiperiod ofJ time folicwing the 4 RCP trip, heat will be y itransferred1from the plenum cover and its structural members.to the .g ~ plenum coolant by means'of. forced convection. Throughout the w - ', - e,

remainder cf the cooldown period, ccolant which enters the upper

5. - ;

1 4 7 kJ

f 7," '.s 4 s - :u C., J - ~ - ~ ^ y} _.j? 1 TR 017' x. s vm.': -

g. -

4: Rev. l' ' Page /.L. - t, A (hdakvia'CRDguidetubes[is' considered-to'mixwith'andd'isplacethe gg ~ ~ 9, Ycool' ant'in the vicinity.ofLthe CRDLguide tubes.. Mixing effects were e 1 ,restricte'd to; the'.lccolant vol'ume defined by the.20.'5 inch axial a. fa,- - ~ m dimension:(the? guide tube height) above.the plenum cover'as shown on A e.1 .. Figure 12-3. ^ l ^

l,

--.a 3 -s.. ' . *t IEnergy. transport will occur.. from the upper head by means of' effluent C' c J. coolant'which will..; pass through the cutlet annulus to the hot leg. ~ JCoolant which exits'the-upper-head will be replaced by, coolant flow - l- ~ ,afrom thd1CRD guide tubes. 1 .+- u 3; ~ - .m+ - W ' jhea[d jill { decline with temperature Consequently, the RCS will ~ > provide.over-5000 lbms of makeup ~ coolant-to the head untti-the decay 'heattsystem Is employed. e. X c ~ ( i U f T,-- +

te' y

V -s_ s Y -\\ ? r ~.'i,_=' s t.- 9 - U n 'T-h 7 ..-m ~ J ,(,__; 2 5 w ,,,e u e ..g. , -, ~ ~. -, -+g 7 7-,.. ..m+ q

a a ' TR 017 Page 13-AV ' ISOMETR/C B50Lwby s. 'e-s N f_m /zourm ~ ^ mfg % og 2,h I /! j %,l 9 ~ dlj /.g 'y '(y -X f:ihC / /" c, f / 1!\\(, A hi (,i'[ / [ / p' y-l i 7 / f-N.4 s o,- v j V e 1 NGUEE 23 ~

w 4 - =. r= ~^ ~o r m.g s x Q Q _jf 4 ( N,4 - K. ? ? l LTR.017: ,e- .gey, 3: y? 1-Page N W [3f04EATINGUMODEUbFTHIil'UPPERHEAOL

!3 flL. (HENTING6 ' Code /

,1,. sHEATING6 isfa# multi-dimensional,:: generalized heat' conduction' code which E d ".may?berapplied to) problems whi.ch require' steady ~ state and/or transient .r_ ~ ' ~ 'soliitiossi*C iThei thermophysicallproperties' of various mat'erials may be) T

1. considered.:.tof be'ani sotropic, temperature or. time, dependent.j Boundary ;

~^ a conditions may be applied:by means Tof the following':

a

^ " fc [Coeffi c'ients ; for'. convectivef heat l transfer J-natural-co.nvection' g

• j

^ .c..forcediconvection 1[~ - t-

Coefficients 1forcradia'tiv'e heat: transfer-

.z a. ~ -Temperaturel m - ~ * ' ? 4 Heat flux n p .p , : UsNsuphlied 9 subroutines allcwiflexibility in the manner' in which the r 4 k' f ? methods of! solution.may be' developed; Both explicit and impilcit finite uf, I.(', 1 difference: nume'rical; techniques' are available to the user.- ^ F:jl ~ ~+ ^ 1.1.l L HEATING 6 Limitations 3 k Since HEATING 6 is a.' nodal-heat; conduction computer code, it is . limited by:the1following: i "I The mesh applied over Jthe geometry of interest is fixed in-4 t size. iThe. mass which occupies each unit mesh is assumed to be gl b ~ Lconcentrated at-the; gecmetric center. ~ .'If,each unit mesh is. censidered an elemental centrol volume, p w + .x -mass transport across centrol surfaces is non-existent. ue ~, j 9-7, ,,.I

n.

1. ~.~,', L q;:",:

? a ~ m ~ M gigg. e: . h ~ -? '-mgQQ:; ( r VW s:+DM - Ni? n- - " ~ ~ . = ~ 7 TR N17 m@ ppm ' ' " ^ Revl' 'Pa9e'l5~ L e-ML... GL,X: ~~ ~~~w wu; y mm,

x

+ A ',n : * ~_ i win 1 HEATING 6,f the; variation'ofimaterial den'st t' ;as a.; function of -- 2 y 7 h 3 + ;t'emperatureima~y" inhibit or prolongithe convergence of, the : X, e e

. :.. x.-. _

LiterativelsoluttorFprocedure. Consequently -the density lof- ?!. x ~ ei f icoolant in' the#ilpperp head wascassumed to be alconstanti . +a., 'm ~.' , (50.0fibm/f t') throughout.thisldnalysis;,This valueltends(to : A y w 2 r.(increase ~ the-thermal'capacttance:Lof nodes 'at high temperatures. - w y a; ~ n s :d_. Oj y.

• c e

s. ':In? order to include convective effects between layers'of ~ W iH c Li:oola'nt;atidifferent temperatures,-an effective thermal; ~ 2? m% h-w

conduct 1vity for;the.

coolant may be Idefined by means lcf.the{

~ ' ~ b7 : y ~ -following: " 'an xL- , -Kb,~ ,h e . ~f< .M,. wherechc;isf.theheat.transfercoefficient.fornaturai_ '-f: e c ~ ,convectiontand L is ~a characteristic length of.the nedal-

gecmetry.

3.23UcoerfHeid'Model: ~ IThe geometry off the rhact'or, vessel' upper head was:a'pproximated in a two ~ 3.di me ns ionall cyl i ndr i c al.icoordi nate 'sys tem. The'90' vessel s p etry about M m ' the verticalicenterline was' taken into consideration. During the, V idevelopment of*the!.ttermal analysis, care wasi taken to account fcr the 's g Wp _ M ' heat' transfer pathst and processes. discussed in Section 2.0. In addition, w. ? ' { j _, (' 4the;following corrections.were l applied to varicus convective heat .y, [4 transfer ' coefficients ~ used ~ in'this analysis in order to account for any q y. < w N;y i '[differencesfikcomponent : surface area between the actual and mccel RV y t-F chead gecmetry; y$ ~ 11; -The interior surface area e' the RV head; w ~ y, e 4 ~ r l ~ ny z yf..,,: -{

,= ,~ y W;% 37 h c =gg 94 ' I ~~ ^ TR 017' Rev. I 2 hs

Page &

o. +. m y p' - - .3. CRD lead-screwsLand housings. 12. The exterior: surface _ area of tthe.-RV head;. d 4 I-(Thelprimaryfcomponents(of'the! therma 1modelare: the plenum cover, the ~ L..

mass' of upper head coolantlithe vessel! walls -and-mirror. insulation.

? Mode'l' ccmponentidimensions were' estimated from construction drawings or 4 established by.' calculations. '.Gecmetric regions and nodes-were further ~ ~ b used.to: describe ~.the head in detail. Figures 3-1 and 3-2 illustrate the. + 4 V n (regions and: materials of the' HEATING 6 RV! upper head model. They.are d', - $ disc $ssedLindetailbelow. ?The. carbbn steell. vessel. walls 'were 'modeled from the mating surface down ito:apointijustabovefthehotlegnozzle. Water from the cold leg which ~ f O enters-:the vessel' byf means of the annulus 'downtomer was also included. r p. lThe= plenum cover and CRD guide-tubes were also.modeled. A variable my boundary, temperature.(based upon an energy balance between ecolant ..enteringcand leaving the upper head) was-applied to;the region defined' ? 7 fromf the Ltop off the plenum. cover to the top of-the CR0 guide tubes.

Applicatien-cf the boundary condition restricted the effects of ccolant -

'_; x e = mixingfto-the first 20.5 inches above the plenum cover. Heat' transfer to x; '.this region from the_ coolant directly above (adjacent ncdes) is ~ - Lconsidered to. occur by:means;of natural convection. Above the region of f -adjacent nodes, convective heat transfer effects were also included. The'CRD--lead screws and. housing were also modeled. The 69 CRD lead I screws-and housings-were grouped into four separate concentric x f cylinders. iihe lead screw and housing cylinders were distributed radially and: aligned on center with the CR0 guide tubes. [ ei

x M%.#s'

t -

T,,~* D7 i_ 7 a. g I tfpW Eg E rCS t t / AD n Sfg V U y, / 4/y 0 'l / l s G9 7 % R [6 E E 4 .* ; ( l A* g, y. m o. t A ra 3 I 7,x E 7' G 0 D 5 n I_ 1 0H L 0 E 3 0 4 3 9 S e S C n o. l vi 0 ] k' ~ R. g 5 E ) 7 r l} R ? f-U e g 0 3 h g g t g 06R 7 i f. f~ yA k's fjS l9 . 0 R n e c 7 U. i 6 N n o T 'p l? 4 ? g 4 5 L l9 4 r H hi & o 0 4 41 LO O B 2 6 -,J 6 ,)s 0 E T &3. y {5 hi ] Q1 0 R $ T 3 lZ Q 3 H 7 2 6 L . )!. QB L I-B T 4 0 ,lL L k 2N J S t' 5 ,j? E _7 M 2 ]b 1 1 1 8 ,- } f O 5 6 S 0 0 qgQf nf.RQN"9B&y- =.8~ 0 4 ? p

d [. ESTItMTC Or.T MI-l tyttM _VCS$fL FEfm NY. T .,,s,' Mi>p or rue MsrEnieis ~. '9. k~ .fJSS ' JSS 5-SS TNSI. y ft [Nst y ss-S.?. S$. Sg CS-C9xBW MES -o 4 as - Sm/Nass M y-gg g g g g mo-w9me 33 G - GAP (AbN97WAC 9 cs cs cs. us-rusmeron g- \\ }l20 5) Ihal Nay' G' cs G N I ret. yxgyisyyon g.,_.. 5I G ,. I n Hzo ' ' //ro .w INS 2-INSUtM/oAl R N k' h J t 4 ' CS i -9 f sg, Eg, f Cs .g-a c g a i 65 S N j WS .h I ssf W cs d' ssl ss v. 9 G 2 n CS G .i o, 8 ~- = " 0.0 15.0 30.0 45.0 80.0 75.0 -90.0 105.0 120.0 R-1 9 ~

P';j-y.< o -.

F/GURE 3-2 ~ 9m. M NI e 4

,+ w;,gy ra, f:n:W y 4, _ p-le l f 's y'~ 4 TR 017; .Rev. l= 'Page:/1 + > - A ',

Thefreactor; vessel; head geometry;was approximated by-means of two.

g y ,q (cylindricaljshells. ;The-top 'of the' head was' modeled as a disk. g Radial- -dimensionsLof various-head componentsfwere fixed from construction [ drawings. -The-axialfdimension-ofit,he head.was established from a - Ecalcu1Attoniwhich: considered; tot'al coolant volume to' be on the order.of: m ne

'522: cubic (feet.

LTheivess'eliinsulation ~was modeled as: a cylindrical'shell with an outer: jadius:of10321.nches. The ' top.of the' insulation was modeled as -a disk. LSince: the distance;bEtween the head and insulation actually varies with + ~ thefhead radius,--a mathematical average-distance was used to fix the c .:positioniofJthe; insulation. 23.3 Assumotions-

q 4TheLfollowing modeling assumptions were used

in the analysisi l1L '-Theisphericalf head was approximated as a disk on tcp of two

_ cylindrical. shells. -2;

The : resistance :to heat transfer offered by. the 0.125" stainless

'steeli ladding on the interior =of the vessel was assumed to be c .~ . negligible. p 3.' The effects of mixing between stagnant fluid in the upper head and y," coolant whlen exits the'CRD gut.de tubes is assumed to occur in thet first:20.50 inches above the plenum cover. An energy 'Jalance over ~ that region-was use'd to establish a time dependent temperature 'which 1

= m. = m a w %. - wn.' K; L ' : W m ~ ~ n.=k Gc %wh f,

• L ;.

J'b 7:,'. v'm x,. u' } C:; h 7l m@ T Q;p. a-N[swmMk i N h,.bb [,, 1 . gygge gy:q c .-b, l ' cTR 017 = . Rev.-.l. mW-,jwm m -w Page.20 &fl$-&Q, ~ g&; e nj. l _;. b yf y;,iled Tasi boundary condition =at thel tcp of the -CRD guide:. ym ;g x:: '_ _ yasyapp %. v ~. v, + ~ -..

s..m ~

,a., m n + ~ Qv -. -itubesa The: effects of; coolant flow from;the CRD guide tubes were W t s e m. _s m,- .y .., e % W7 f_ z - Jcon s i defredil i mi t e d I to l tha t i r'e'g l on ; &w w+i4' r + rp. w, m,:, u. 2" R.?m -am y J 4.c ' - ' ~ -t.. = a:. yh, w ~ ks n.%y - + M~lw.OHeatdtransfer fromithe" plenum cover-to coolant in the plenum was 4 =v, o - 1. 7 gge ; gf;n f 2 %g ,i WIP M. based 'on a forcedi. convection theat: transfer coefficienttfor fluid! t 9 .um ~

f'

'< flow?over i flat".. apl'atei

m.,,1

= - - - c 4 -.e_ 1 m ' h f k' y;.y y d # . 'y + L g: ..+ v.

• f

.:. ~ ( ,f s s l51 JHeatitransfer:from; upper head. coolant.to the head metalf was.assuined a

,w.n' C 3o occuriby'meansCofinatsral. convection.

- Ojj h ' w m. yn - y; ~. ,r.*e ? V.L-$. 4 jf 1 M ?6 s,1 JHeatstransfer from1the CRD housingsito the service structure regica 1.. &i

arid :containmentiwere"notrincluded. -A bousdary temperature ofil20Y is e,

4., .1 ~ - .r, qD >i ~ ~ 5 f wasipplieditodhe CR0 housing 31/2 feet above the REhead. lThe. g g'..-~, e. temperatu're:was: b'ased on ai si?pil' le'd heat ' transfer analysis - 4 __,L l _ e f J n ,tv, ' js m Qg y ' ", ~ l performed'o'n;a1CADilead' screw; g - :,; 7 ?P gw ~ mp< s- + ~' @ ~ 7r 9Heatitransferfto/or from the CR0 housings' toithe'alr located between 4% . w a'W ,f Lthefvessel head 'and 11nsulation-was neglected. 3 y .%s w ~ <m ,_m c8i !!~A'short subprogram was used to include the effects of convective J eatitransfir71n' the? axial and radial directions'between ncdes at h , w + 4 + ..different.: temperatures. Heat transfer between nodes in the radial o %,m ,i 1: /'. 4 j ~. 7 ' w: , -directionlis/ discussed in' Item 9. For the axial direction, the s ,y relativef position between a specified node and its neighbor were 1 f 1 (detiermined fifst. Convective effects were included if the folic *g c 2 '.' +1 m.

cr_iteria were s'atisfied

\\ h., f ._m.- } ,.[ J .._.'( .f., i ) e-.r g f Y

~ as g;g;gn, - c y 1

M M 9f(

< m c-.- g<3 _h-W' 't ,,' l ~ TR 017: g: m

g,, _ 3 Page' M e

Tq pgs .,7. 3 :f ? Ja)1 ' Neighboring)nodellocatedLabove.thslspecifiednode-_ temperature ~.., .-v s of~the~ neighbor _less?than that-of the specified node. , t Lb)) JSpecified nodellocated f above' the neighborf Etemperature of.the. ~ ~- r 2: - .[ specified node lessfthan,that of'its neighbor. e _ Otherwise.:theLeonvective: effects between nodes were considered to_. y, m- ,n 4. h ted to laminar natural convection. I [' ig,-- ~

A somewhatiless. conservative.'but-realistic assumption'.is that

somes 1~

.' convective Theat' transferfis: expected between adjacent' nodes (in the: q g, p 4 radial direction; 151nce cooler. (denser) fluid willitend to fall-in u ~ ' 1 warmer;surrou.ndings,- _it ;is' not unreasonable to expect fluid velocity ; ?. Lcomponents inithe ' radial (direction. iConsequently,ilt appears c ..+ realistic (tolcarryJover. thefinfluence 'of convective effects to the ,' radial direction. w y

10. Tithermophysical properties of. the._ stainless" steel mirror

-insulation were estimatid from its thermal conductivity. and ~ J-- : iknowledge of its construction. - -d f, ' I'l. Priorito the onset of-the. transient, the temperatures of upper. head - coolant:and-metal. were initialized to_604*F. This temperature includes measurement uncertainty"), and is the approximate temperature'of ccolant in the -hot leg under conditions of 1007. power. ~12. Subsequent; to the 4 RCP. trip and flow coastdown, natural circulaticn flow was assumed-to beccme stable and invariant at 37. cf rated RCS t 4 . flow at 1007. power. p 6 7 y b

W '- i ~' ~ l' ~ ~ ' ' ^

1. W & ' W : " =

y~ 7 ~,

~ fww'n wA = '.

• W.m s ;; q:w;.e"

^~ ~ g; e Ny&eff u ' ', p: - i m-c $p;,C Tf, g' . "YA sq = ^' TR 017' -gay,.i-a_ m P. - b, n;u ,~ ~ J I _Pagef A - ,l El3 - - ~ - + ,, ;m 7,a ~ .'132 ECdolantifl.ow Lintoethe upper hiad fromLthe-CR0 guide. tubes was; y - m 9.'., 3.Gassumed to7 e;S".;of'theiRCS.:flowjat any givenl time."t .._......n. _. R. m + b t ~ T ,.2 s ,Y, ,a.3 ,y-w_w 114u JAffourfpumpfcoastdoEniwas; assumed,to= occur!followi_g th'e reactor.

w

.n 4.t.: '"y i s 4,.;.,.., r ~ g ltripl? iThe.ieffectsfof theicoastdown were, included only'inL the energyi 4 ~ + ,W' jbalanceidescribedsby11 tem 3,LSection 3.3. ' Operation of:the RCPs I}~ , hf ter?alreadbr[ trip;will tendito substantially re' duce-the a, c ??,, J,

temperatures of ;uppersheadl. coolant;and, metal. Since

the temperatures y

-: off coolant and metal fin?the RV.h'ead' tend ;to follow.the hot leg -

m.,

. e ~ c 1 itemperatureiduringeforced flow conditio sn, the amount of temperature-E ' ;reductionftln;the upper' head 'will;be dependent upon'the duration of: m' ' ' w

R,CP[operationassumedllafterthe'reactortripand.thepost-trip.

_m

cooldownLrate..of, the. RCS.

m IS2-{Theia~mbientf(reactori building) ' temperature was. assumed constant at = m'N T120*F. L"' .d. f' s il6L -T'he=finsulatilen was censidered to be. completely sealed agains't air. .n ; ileakageT y w sq" m 'r ,.s.- ,r i .u ~ .. ~. ? Steadv State' Heat Transfer ' 1, L3.4 %m Y: J y The steady state;heatTtransfer_ rate from the.RV head was estimated from

the' manufacturer?sispecification'for the heat flux associated with the bo s

lnsulation. Convective coefficients for the films which act at the 4 exterior' surfaces iof the insul' tion and betneen the vessel head'and the

+ a 2 interiorisurfaces of the insulation were determined by means of an-Literative procedure which included the folicwing constraints: A 2 S i.. t,~ Y ' ~ " ' 'j

mn s me Wg 9 T "i kM jf l o k[. [ 3 9 9 ; '$f ~ > %y$:W @ W~ %..y6 c; J ~- " ~ '>- c%X ':^ .~

M f4 2.

~ " g" '. TR.017 m ' $$D 's, O r. :

f' f %

.Rev.11E q;m _- m s _ FPage >$. s 2:::g " i L w

c

g'

: 7 Q _ P' :il):?.tRadiative heatirans'fer was Lassumed ko. :take ' place: between the' N:

n--~n= ~ a:af ~. mJ ~ + - ?; s exteriori surf ace s fofj mi rrorfinsul ation s and ithe icontainment y+, 7,= ~ fsuFroundings. 4The ins'ulattenLwas? assumed to be a. gray body which: r, M~...,... '- ^ (transferred heat:-to blackisurroun' dings maintained at 120*F. (An:. ^ ~ x. remis' ivity[for #18-8l polished: stain 1'essisteel'(evaluated at 200*F) _, - s 4 Y J5 Wa'sfassumeddoib'eadeqdate. - g i d, m_ 1.2)? ;Theifilm' resistances between the RV head metal and coolan't were 4.. <.. -. -.. s. ~ x-Mi.

[ ~ M '

;cordideredWbe'zero;at steady state.

+

o,

r9-3 );yThe-:RV head metalisas ~ considered 'toLbe :at= a-temperature of:604*F. - _ 3 ~ " )4)[ !Thi{ temperature change :acrosslhe.. insulation (approximately 380F') ,w 4 fs was determined from:- p g.C ; +

~ m w y..

i a)? JHeat flux:specified by th'e insulation. manufacturer.. The heat ^ 1. ' T (- 3 flux lwasiassumed iapplicable for' steady state conditi ans. g _ J V

b);. iThermal'cenductivity cf mirror insulation - assumed ti be

- + Mo

independentlof temperature.

~ 9: J 5) ,:W iContainment' airitempe'rature' was assumed to be invariant at-120*F.

• b s;, e

- + "{The-heat transfericcefficient applicable 'between:the-vessel' and ne ~- ,m . 'ilnsulationisas initially assumed-andlitera'tedlupon until all constraints

were satisfied.

y- .O ~ 1,. m t . in _ V

3.54 Initial Cc'ndttions

? ji (Several.~of the initial conditicns" imposed on the-thermal mcdel at the g.; .g, ~ Monset of the reactor > trip have already been outlined in Sections 3.3'and' a. t L3.4b 7 Consequehtly, they will be summarized by the folicwing: p 4 3 3'

4..

f-

g mm - m-- r t w [_ z'- / ~ r e, >gmyy s hi l

Ij.

TR 01T kg a +< ~Rev.31. - ~ ^Page 2 ll ~ =,.

?.

-.....X (t m~' u w' eThe plant was assumed toibe operating at steady state ccnditions at %+ x 1001Epower. ~ J XIns'ulation: temperature was;taken-to:be 408*F. W joldfleg.temperaturewasconst.dered'tobe555'F) ~ e-4 K- +c p .a, 93 These:aredhownonFigure3.3. t w ? g a,

At time ((t -?O+), Ethe ' transient!was initiated with. the' following' events.

e. 7 ~- W Jimposed: l V y;. . Reactor? trip 'withiinsertion of control rods' achieved. py~

u. -

2 f E' 1 *' lFollowing th'e' reactor-trip, 4 RCPs.lare tripped. (Subsequent. to the-tfip'of the = 4' RCPs,'RCS flow was taken to decay 1with pump coastdown. < EP ~ .7 3 . y 3.6 -Boundarv Conditions-Various boundary conditions were applied to components of the upper 4 'Jvessal-head'model. The types ;tha.t were used are delineated belcw and ~ A.111us'trated on Figure 3-4. 1 4 ,j '3.~6.1:~Saecified Temaerature / A's : explained in Section'3.3, a constant temperature (120*F) boundary <conditionLwas maintained at the top of the CR0 lead screws and CRD - hous i ng:. = (BCS). A time ~ dependent temperature (which is a function of the; RCS cooldown' rate) was applied between ecolant in-plenum and hot leg-metal (BC2). - A. time dependent boundary temperature was defined at the tcp of the n. -CRO guide tubes. The methed used to establish boundary ccnditicn

(EC6) is delineated in Section 3.3, Ites 3.

.g .); _. s \\ b %.t.s ml

_.e_y. _;. j. c7..;; e A i ESTIMRTE OF THI-I UPPER VCS$tL Km SNIETRY Map of rits retrial awoir7sas N / I R i 9 5-1 2, 2 g I tegepo, 7 r 9 1 - lef *F f i i 1 2 . fog p y-f 9 1 1 1 I N~ 'Nors: g ~~ f f' NETAL, WATEQRNOJNSYXAT/0N ~ o i f g f 7EMPERaMPES SET 7D lof g-bof'F A@ 408'F RESMMfly. f S7F,RDysrArc 7Eypsygrygg g. DISTrisurroxis osr,o as g-t i i i I .I s i THEfxtrias m;77gy ~ ,2 W Tfss'Egyaj~7tygao, q ~ 3 j SEE 17M tor ro.g Mrsyz3, / 9 Z -g / n- ./ 1 i i/ s .8

- -- R g

B E 3 3 0.0 15.0 30.0 45.0 80.0 75.0 go.n 105.0 120.0 R 4 % f'o. AsuxE 3-3 + g,t e

.~ ESTimvrt & Int-1 urPER vtsstL Htna ornntyny g May of 1Ns Bwnwy coxwnavs

• B

/ cf .,.ge,g, L a i ,5 Bc Be Bc B ... 9... m,g o,_9........a.9......... - "7" "J' ""/"......J........., g ~~ y-Bcu Bcn Ben leggyo k &..k........Scy.L.......... ect-roxceoexmer,sv 9 M...,( ~

,.,.. tw....

c ecz-nns umunr ._m. g. g,,, i i scjo irnmenwxe t j rm Bcs-- NRmW CcNiccrim l0.' 'l y Bc8 l ;/ eay: 9 B c . ;-zawm act,ac7,Bcio Naruxat ~ a-M

/y coxvecnwu ncs - ra ro re n u,,sivs, n o. -a,-

pa ,e Bcy,, z.R f c7 l Bcs',Sc7,geg yyyypz(gyggy, B 6 1 BChy t M il-k mgoz g yygeyg ( l .VlScjis n(tw

p

• M> Bel 3-Nawssi a1wgc77og M

Anotanov rawxes su Bc1l 5-Bci l

1

• fnre BcL 1NellitEs note tyrencesr-l R

BC2 Ec3-k, I gwnphrif TEnftRantee. Sed \\ 6~ Ilf l eccnox 3.s. o R Bee 8

.m,,

an r, =n o.o 1s.0 so.o vs.o so.o 7s.o so.o -, eas.o saa.o gc R 1 Fisuns 31 1

yy 9 m.< e

) u-- P

p-}7&Wb gy;; =r 'n W~ ,TR 017J ~ R, x -Rev. II .Page21 c3.6.2 Convective Heit Transfer a AL-heat; transfericoefficient for. forced convection was applied 'as a ~ iboundary condi tion sto the.-lower surface of-the' plenum cover -(SC1.). Y_^ ~ Similarly, al natural. convection heat transfer coefficient was - fapplied in ~the ' annulus 'downcomer between thezvessel wall and the' core thermal'shleid. ~ '~

In-the upper head,'cnatural convection. heat transfer coefficients Twere ' applied at all metal'-water surface interfaces.

i Natural (convection ccefficients~ were also applied to the following '

o 7model compone'nts

To the outside' surfaces of-the insulatten - To' the airc spacc between the vessel head and the inside surfaces-of-the insulation s -As mentioned earlier. the convective coefficients were calculated by -means of-an iterative procedure performed on the conduction and- ~ convection heat transfer relationships. A known steady state heat flux:across the insulation was used in the procedure to predict 4 . temperatures on the' interior and exterior surfaces of the , insulation. Subsequently, the temperatures were used to predict -suitable convective ccefficients. It should be noted that the convective coefficient used between the vessel and the inside A

9, ..; +. ; -l r-y' : ;, s 2, ; 5 TR 017.' y J... Rev. 1' jjvf11(l ', e>, .~. Page J P - ~ c surface of =the -insulationfis-considered;to include l radiative heat = f i;jtransferieffects "Sincefit was difficult to determine an ~ c ppropri ate L rad i a t i ve iheat l transfer;coef fi ci ent for the ' ccmpl ex a ^, F - M (gsometrylof the air space, the 2 lumped convective. coefficient was 7 ladopted 'and the1 air space ~was modeled as an area devoid of material. J 4 '~ 4 b E 4 s S 4 t i g 4 h.,, I te. I O 8-A a h ? I t b 6..

gg;b;. t ~;,

• m::.

~ ~ n, s. s 2 ni -. r - u.h-gm e .s -c 4 r e M :X..v n '

x

~y y y.;r y ;p $ ,p: + m. ^:7"' ~~ TR 017. ?, . D,'

Rev.-1

~ km? p ~ Page 4 9; w gr,.. 4 w ? 4'. 0:TANALYSIS~RESULTS1 @::p w, ~ ~ ~ ' [m Q MTwo"analysesfwere. carried.oufftoidemonstbateithe'thermalresponseof;ther J

t. ' '

JRV hehdjas :al$unctidniofSthe:cooldown rate:in the RCS. 5The' ccoldown- ~

+ p.c

,w m a:t, m .,irates! !mposed(on the RCS were'10F'/hrT.and '50F'/hr. s:The;10F'/ hr~ rate was; c.." <y c ' imposed;for. aL40 Lhour; period; - The 50F'/hr: rate'.was. applied for an 8 hour w ~ - =. g < < 1_ l period.- 4Thereafter,LthelRCSitemperature was he'id at 204*F for' two. _s a ihhurs; TheVresultsfareshownonFigures!.4-1.throughi4-4'and~are ~ ,~~ 5.__ m

idtscussed?in

Sections 4.2 Land'4.3.

g, ~ m 2The:volumeaveraged[coola'nt=temperatureinthe.topfootoftheve'ssel 3_ s L head..was'~used. toLrepresent the overall/coolantL temperature as explained

a

. v; 'in.the following; p a

2. i 1 1 HVolume' Averacino of the' Too Foot of the ' Vessel. Head 4

y z ^ ' + N Neat transfer by means.~of natural : convection is a complex -process which? n', V invol_v.es; mass -and energy transportLat relatively '1cw fluid 'velccities. Fluid ~cticul' tion iriinatural' convection is attributable to Luoyantiferces: ( [- a ^

which arise from' temperature

variations in the fluid. - Consequentily - free-Jconvective flow is compressible flow. J n$ this process..the convective I

Lheat transferDcoefficient -.is characterized by the Rayleigh number which-1 11sf a productLof the. Prandtl"and.Grashof numbers, the Grashof nurber being- ' - proportional.to the ratio--of bUo~ ant to viscous forces. y 4 [_? 1 'eI lDuringla natural circulation ecoldown, convective cooling of u::er head ~ i metaliand'ccolant will: result in buoyancy driven fluid circulation. '41 th M i. n ^E.'. r V O 4 f 3 ^

w 7:n.- n 1-r ( m M qQg i;: ? ~ TR 017 Rev.:1 [M lm;- PL95 30 ?

continued cooling > ofifluid layers at different temperatures,. ccolant 1'

ctrculation lis expected to 'developlas ' depicted on ' Figure 2.1.. y f }pl; TDuring the. cooldo' n.lit is-expecte'd: that buoyancy driven fluid - w ~ 'l circulation:will; propagate; from the head walls 'and lead screws 'toward the-

centerline of~the head. L As~ mentioned previously,.wlth the decline of -

~ ~ m [+,,

coolant temperature in the RV head, it is expected that.over 5000 lbms of

. i coolant'will -be 'provided to the -head by means of.the RCS. Consequently,. !.it-appears.to' be' unrealisticito' treat ' he cooldown of the head. strictly t ias:a. conduction problem.

Since'ccavective effects are included in'both ccordinate directions,' the

. use'of a volume averaged ccolant temperature in the top fcot of the upper.- head appears to be a conser 1tive representation of the ccola.t conditions in.the RV head.' 4.2 RCS Cooldown at' 10*F/HR - ' Figure 4-1 demonstrates the rate of' change of coolant temperature in the ~ .RV head in' response to:the'10F'/hr RCS cooldown rate. The results

ind' icate. that the coolan' temperature decreases at the rate of 8.86 F*/hr t

4? lover the.40 hour-period.. Figure 4-2 demonstrates the rate of changejin ' the saturation pressure of head coolant throughout the ccoldcwn. At this

cooldown rate, it appears -that it would take approximately 22 hcurs tc

~ cool the head to the OHR cut-in point without coolant flashing. The

• average change in coolant temperature over that time is en the order cf

-8.23F'/hr; w

- e (.;'[ 'r ' ~ ' ~ TR 017 ,.4.; n_. g :. _.

Riv. 1

p s,

a

.pgg, 3, s. TEwMRRruRE vs nie - shdw ~ f es mn @) 0-40nes

0) Q;r2Lrnwh7E-Ar-/0/7m 4.

t O 6 4 h 1 g-P k o 6 4 h_ ,+ - g'*Ae 4 / p I//l5(NES) loc + - i. # 1 i i e i e 4 l 4 i i i t.. i O' 10 20 30 g ficuRE 4-1 9

E.,.; ~'" g - ye TR 017. -

• • W8 w

em ~+ lSR70RATIMWJff (M'AT Rav~. I lfl page fz Rymsq702977CN/kmuMWTruc-Wg/u At2 Ca1.awd: (, b* 1 (j)$Ntes-Ats.tona f m.enre v g w 9 k l ' ~ ]J00-x i fjp

1:

\\ .W <1 A 4 400-sw-wwwn). ' )g ,h h 'b5 hS h 6 'O . o FKd)RE 4 2 b

=

we , [gdC 4 s ~ +. TR 017-ip s Rev.'l givre;- n Page g3 4.3 :RC5 Ceoldown'at 50F'/hr-m g Figure 4-3: demonstrates the. thermal response of RV head coolant to the; J50F?/hrL. rate' of change: Imposed.upon the RCS. As you will note.'the RCS x

cool. ant temperature was held constant at 204*F after.8 hours. The.

results indicate' that. the; average rate of decline in coolant temperature:

over Jthe_ first 8 hours is approximately 31F'/hr. Thereafter-(in the

interim ~between hours 8-10), the rate of ' decline in coolant temperature shows the effects of-the hold at-204*F.

d Figure-4-4 demonstrates the rate of change of the saturation pressure of .s T head coolant throughout-the, ccoldown.. At this ccoldown rate,.the' results s ^ lndicate' that it would take approximately 7 hours to reach the DHR cut-in-spoint.. L -At.this peint',ilt?should be noted that the head coolant saturation .pressura versus time ' data cbtained for., this cooldown rate were ccmpared 2 - t'o'RCS pressure data cbtained during~ the natural-circulation ccoldown event at St. Lucie. The results of this analysis ccmpare quite favorably 'to the St. Lucie data in predicting the t'iste at which onset of ccolant 4 flashing wi.t. 1etected. The ccmparisen is elaborated upon in Apcendix A. 4 a .IU ) s: -

-w, ,J .TR 017 ~ . -;j. -Ray. 1 -{. (' ,..Q Pega 3'{ g ~ .( g I - y Lg j -4 c gy n u -k; x-Q.' w -4 gs g v A 4 d D' j 5. -4. n

y[,%

-t a..,y,,.s--- m n :

g. a y g:e s,m.,m z ~

x gs e ...SS g . _h y r;.: 4 4 + -> g -to k g k* -4 _y

g

-m ~ -~ "N 8 ,o o o g o N S G 4 m a '4 Ls -

s {SARMA710N fMlA5[ / .;gy ggy $gyggmy&$$JEm W~ MS f;.lWlf;f X 9 - *. t; a. M(WWN:-. me /!DO~ ,1.,. . ( yyy x !. 7 m p.,- p 9,p y>ip-7 J t/. Q~Rj_&;A A-H

• l TR 017'

/590 0~ "i? Rev. 1 /S32 1. PaSe LT 13C0~ m 2 ,jg4 3 y -. .I .j y ~ =-. Glo S ~ .,y 106 .6 .8 3~ 1M> ts5 7 ~ e: g.

1. 9 8

g _0 og z W xp:, 't > " ' N,$ g' 5/ ,) # f .W , (*'s'i, A - ~ 1, s -t, ,'5 ,t \\% u s. ~ ~. i e .y n -:fcc_ e. . 1,, S 4 e 500 - b -t 't n / .\\' c jg ?. v._. .[ '5 ,f g -S ll _ _W (. - y* r, q a r,1, 0 e 3 a.. ^, '100-(.. f 6 j.. 'h I ) h b.' 6 1 e p., F/ GUM 4-f ( s , g .'.p r .g %.jee

2

' s

,_ m .-a ..a. W WWj$n $5f Qg l A f W % ^^ M. ~ %. = $U$Y.}a dh ?=' %.;0; g;% "

• Aks

'W _m a.

y. nnlllllMq 2 - -

((3, s.. -. ;f, ~ f =- s' 3 ~ - <= - c W v nc n, p g % yg* '.,, ' wt ~ sTR 017. A 1 4 an,- m:m a u- 'Rev 1 fyQn.e g*ru; yy;~nmr 9, ? p: % g,wh , ye ;. mas page 3y Jt a J s 1 M, w. i W !s: M S',"OsOPERATIONAL" GUIDELINES ~ ~ y y;;7 /r SI j h" h 7 T$Nmeth6doldgy?il previousIseetionslof 'this, report 1will > provide" the x; ~ ,qwq jeF$,f baitsifor[o. perationalig'u'ide11nei to)be devel$ ped lfor a1 natural 2;:;\\M e M,lation[cooldown;, ~ ~ Q1; y;7 y i circu m_ u ~ =. w e, ;,yg",'.. y p. 4 G gy [1 des'irabikto avoidireactorivessel upper-head voiding in-1 [ [r( M [fjfponfomergency[cooldownesituations ?the followi

z

@3' su .x wlA,.,W4, 1. =k ?thislphenomenonlin perspectiv': ..=.. - ^ s e n w

m

, Inat0ral'ciicu1 Atton~'cooldownlis an unilkely occurrence' J s,s 3h : ^ A a r. j:a / ^ / % ' ~

• * ?The(two sostilikely; reasons Lin'which the RCPsLwill not he available-g 7-g@-

s - g lapiioss off powert tc(the,RCP motors and loss l of.servicesJcool.ing. a k / 1.

• f h..,

y lwateriand[seaMinjection)hto?thi RCPs. 1 xperience has shown-that E ..x c :

f. J Q^;

L__ w ,)b'oth of these <are unlik'elyLforlan. extended time. JM- .S a m pm .~ y;; b:t M.'", > / 2 L. "A steam 3bubbleiln the RV uooer head is not a safety orcblem v. 4 ,v ^ M 1st'eam' bubble;in' the RV upper head I,s;not in itself a safety-w

1. ' G '+ 'g,'s

_a l problem,Dbut: a;pla' nt Tcontrol-problem for. the operator. The' u n nexpansion!of the voidsinto theihot leg to interrupt natural-e / lgy p, NircMation ~1s(unlikelk s'ince :.the regions below the upper. head are M ysi - . J. c y M subcooled Tand(expansion of,the void into these regions.would result 9, <., ~ ' ; 1g ~ 'in: condensation,'thus restricting the vold to the upper h'ead region 7 e , 'm

~ ', e only.'

s E, 3 f_, -e i I - xy a, 5 J' Controlled -transt'tlon to natural circulation or interalttent 4 5m M t...- m ->r f f -[, ':N coperation1 of RCPs

m i

, F In:the' highly likely event of a controlled transitten to natural 1

==: n.. q icirculation, operation of the RCPs after a reacter trip will reduce t 4 f: Y qL b-t' f f:_.' t.' s [ ' M ,( p, p-

i c: c. TR 017 Rev. I .b Page W x. .the.RV upper head:t'emperature. RCP operation for several minutes- ~ beyond the reactor trip will'be an effective methed in reducing the '. ? . saturation pressure >at which coolant flashing will occur. In ? addition, either intermittent' operation or ' bumping' would be an 3~ : equally effective alternative in reducing the potential for coolant flashing. ~ n

F1gures:4-l and-4-3,of the' previous'section demonstrated the thermal W'

~ responsas of. thef RV. head.to 10F'/hr. and 50F'/hr cooldown rates in 'the RCf.'. L The RCS could conceivably be cooled down at any rate up to ~ ~ ~50F'/hr. Figures 5-1 and 5-2 represent the minimum (limiting) RCS y- , (presxres^ req' ired tol prevent coolant flashing in the RV head. }4 u ll [

Thesefcurves we'r,.e developed by correlating thelRCS temperature

~ against the app'ropriate: head coolant saturation pressde. For a' -~ A gijen;RCS Ltemperature, as.long as the.RCS pressure is ma!ntained F above the; saturation ' pressure line, coolant f' lashing in the RV head $~. " N.. ? n v ~ W, will?be prevented. ~ g,. .g .Ploitsd onithe saturation pressure.versus RCS temperature coordinate

• ~

-9 . system,' the: data obtained for theLthermal ~ response of. the RV-head to aarlous RCS:cooldownLrates no longer appear explicitly in time. p - - Consequently, ~ Figures 5-1 "and.5-2 are pr'ocess diagrams. If the . I 3 _ 1(P. vs1 T)' dataf for -the 10F'/hr cooldown (Figure'5-1) were to be' n y 7 a,'

super-1.mposed;onto Figure 5-2, it ;is apparent (for a given RCS 1 temperature).....

fN m r that the 50F'/hrl rate 'Is more pressure limited. As you / t; i s ?. jp iwill-nots.Lthe' impor. tant parameter on the process diagraft's the w M *: A 4 ,1 -."y a, ~ ^' _ %[:.- j;; h

• Vg 'As y,; x a

~ ~ - ~ - - - ~ -

g -- U-'lIWD p mjrpnen PoeMsuM (M/m "Sa7uRR770N Aesssum (AV/MAC) Ys TR 017 ACS TENf6?A70RE ' - 6///gg Rev. 1 F Page 31 gy, $$#$ mis )?5-jofst JSW-7 w ~ meenAtyxe v

g

12%-

~ 1}M- _gy ~ g_ E geo_. W -1600 u g

4m-

.;M-L

-200-

?100 (. EC7 7:vc )70M(sc) 00: ' CO ' .bo ho 500 00 ]bo (, ~ F/GURE S-L. c s-

1 f-y i f' j & f $ SRTUECT/0V PREMURE (8'G). g MK& AQ!T4D)W TR 017 gg,, 3 9 Page$1 hbi C2%%WA1 ' $5 0 ^ fY$ hhMwdbrE-fC5-8 RA PArl)RS \\ o., -, r jp Isu y IS!7.54 W ,3g y 1500- \\ gjg 4h M1 404 ' :ff00 E?! 3 54 p .4. 1 3c4 241 2S4 110 0-. 12 4 20$. . 73 204 a .a ed '900-800-j.. 1700-W e .p-D 4 ' goo. YCh6DAT2C#f JCG-RND 154SPS/G o FOR 2hbuRS O Ats 77MGR7theE(4) O r ,_ ~ _g j b0 600 KC g a 10 0 ; 200 .3 0

ico f/GURf S -

..,_..c_.

h2 ~ y,; 7.s >: ... ;.y e , yqq S: g TR-017 7 Rev. 1. 3m Page R0 v J 't

slope of theicurve,t.lje. <the change in Fvt head-saturation pressure n

wlth-respect tof theich$nge in RCS temperature. Nith 604*F as one

, _'.' 'j .~. [ - -.7 reference!. temperature the' slope'of the 50F*/hr curve is obviously o:. ' %f Mr tsmaller than' the [ slope of the;.10F'/hr. evaluated ov'er the.same' c .,.... K As-a result : system pressures at.the 50F.*/hr rate must ' conditions s sw ' f ]be ma'intained suhtahtiallyl above those. for the.-10F'/hr case. 2 t . :-) ![ N Y-

The data-obtained from:the 50F'/hr ;cooldown: analysis will be p5 includedin-thejressuFe/temperaturelimitsfora_ natural

.:n Should there be.a cenflict between the ecirculation.'planticooldown. ,.:50F'/hk datban~d hie" fuel-pin-in compression.llmits whereby-thel ~ ~ ifuel-pinicompression411mits-proveito be acre limiting over-a

• r-g 4

-particular-Lregion' of the process;diagra, a ccmposite diagram _of -the - o Etwo willibe developed as(the411miting curve ever that-region. + - s y p-r f .l M j s, + ,9 ~ =~ i + ~ i ,j j W, ]S ' = x. s ( ~ d_ 7 + ^[ ., YM_ 's ,~5 ~h? y] s ~ : .w v,e "g-R i [ h' - 4 K, 4 E6 - n ~,/ .). 4 I s. m ~- <

= s z. ., -1"4 TR 017 ~_ Rev. l Page 41 '6,0 lCONCL SIONS SeveralLimportant conclusions of this report are: "In orderito initiat'e the-DHR system,~ the conditions -in' the RCS must

.+

~ be no!greaterithan1325 psig and 300*F; To prevent coolant flashing 2 (in-the <RV head during!RCS cooldowns at 10F'/hr 'and 50F'/hr, the head' LwillWe required to be. cooled for. at least 22~ and.7 hours 'respectively. Provlifed jhe-system pressure is maintained above the Indicated - ' minimums.during :the;.cooldown (saturation pressure shcwn on process ~ ^ diagrans:5-I and 2), flashing of RV head ccolant will be prevented. 1

The results of 50F'/ hrs ccoldown analysis -(Figure 5-2) will be incorporated into the~ existing.ccoldown procedures. Shculd'there be a potential conflict between' this data and the fuel-pin-in-compression iloits over any P-T region,-the more limiting P-T data

~ will be.used over that region. ~ b D t 3 m i.h. 1 h

Y 1 . '* h ~' j. l ~ j

,, =c TR 017.

'Rev. 1 ,Page 4/2. 7.0.EREFERENCES, m, D$ .l. MRCJLetter,- J.;F. Stolz= to H. D. Hukill, Request for Additional-

Informatlon, Natural. Circuiation' Cooldown (GL 81-21), July 20,1953.

HEiTING6[ A Multi-Dimensional Heat Conduction Analysis with the 12' T r Finite-Difference : Formulation, RSIC #. RSR-199. w y w M L3[ t TsI-1}FSAR, Updated Version, Volume.2, Chapter 4. i

Aff,

• ~

C 'Ji Smotrel, RV: Internal Flow Velocity, B&W Document # 51-1146582-01, f Augu s t -- 31', 1983.- hJ 5. LHAat Transfer Elst Edition. Holman,.J. P. ~McGraw-Hill Book. ~" Company. 1981. 6. 1 Principles ~ of Heat-Transfer, 2nd Edition. Kneith,-F. Internaticnal-Text: Book 1 Company..May 1981. '7.- ~- Momentum', Heat andl Mass: Transfer. Bennett, C. 0.' and Myers, J. E. ?McGraw Hi11-Book. Company, Inc. 1962. r lonvectionHeatTransfer...Arpact, U. 5. and Larsen, P.'5. Prentice 8.

Hall, Inc.

1984.

9. - _ Handbcok of Heat Transfer. Rohsenow, W.~ M. and Hartnett, J. P.

'McGraw Hi11: Beck Ccmpany. 1973; Ijg _

10. l Heat Transfer. ~Sucec, J.~

Simon and Schuster. 1975. l I ', e ! Convective Heat Transfer. Burnmeister, L. C. John Wiley and

• ~

Sons,:Inc. 41984. .12. :HeatiTransfer: Pocket HarIdbcok.' Cheremisinoff, N.' P. Gul.f o

Publishing. Company; 1984.

L13., Process l Heat.' Transfer. Kern, De R. 'McGrawiHilf Beck Ccmpany. '1950. s { 14.c. Fundamental of Classical' Thermodynamics. Van Wylen, G. J. and , l Sonntag,C R. E. John Wileyf and-Sons, 'Inc. 1965. ' 4 15 4 Analysis and Evaluation of LSt. Lucie Unit 1 Natural Circulatten ~ f,, .Cooldown, NSAC-16/INPO-2. December 1980. 6 16; :- ASMEf Steam Tables, 5th Edition. American Society of Mechanical. - 4 Engineers. '1983, o ej. i A' y j- ) 1 .! - f - 7, y' ~ I- ' - ^ '

Q :;-;, f.

n u. _

s. g.

3 rr TR 017 Rev. 1 1 Page: Ll3 APPENDIX A' x. As' mentioned.in"Section14.3, lthe thermal; response of the analytical model to Jthe}50F'/h'r RCSLcooldown.was compared to RCS: pressure data obtained during the _ -naturalDcirculation-cooldown:at St..Lucie. The. saturation'pressare of.RV head ladfL(referto' Figure 4-4)iwas:overlayed on to the RCS coolant pressure J versus;timt:; history for the first;5l hours as shown on Figure A-1. lYou will-s ' M fnote therel appears'?to be reasonably close agreement in the time during which i theTonset.ofl flashing. cf upperJhead coolant is thought to' have occurred. L eferencel15 indicates"thatithelonset~ of RV head coolant flashing 3 appears to R Thave occurred at 6:15' AM. : The authors have^ also noted that the RV head at - 2 ' 'St. Lucle? appears lto' have.ccoled ali a rate of 14F'/hr over the first 3 hciurs ~ ofithelevent'.- LThe.'RV head temperature versus time data for the' first 3 hours _(referito Figure'4-3), predicts the RV head tc'be ccoled at the rate of L15F*/hr. zYou' will alsoinote that-the cross-hatched area (beycnd the intersection' of the two; curve;s)...on Figure A I_ ' demonstrates :th'e presence of- . ' team voldssin the_ upper head. In addition, temperature data (cited in s ~11e'ference' 15) 'onained from the core exit thermocouples throughout the-event-i iat St. l.ucie..' indicates 'the! var atico in RCS coolant temperature over time to ~ be substintially reduced beyond the 3 hour mark-into the-event. t Y 4 4 j s. ? 5 +. .-,-.s s-,,a

E , -.4 ;, ; ..R....... r ef = = o.- TR 017 Rev. 1 Page 9/ 9 g j ,3 ~ i 4 ' k 0 I A I iu', . 22 2 e-f] I ( 1 l I l '1 4 6 c. ST(UC/Es kWR6 W.M, . - ). l 0 L i lI THT-f:WAVZMat. CA' t249/7ar' *

g fange l

p presaws roccreer,,; gyrygg7foy } -- ggy s* 19CC 1-j. reference 82. l-(EV N69 bht.DNE- 'o l \\\\ l 'l l I-V I % Looww b ir = Sof % p l.- l i 3 _a a _a l l

1 l

l l l l l j j _l '7"' i i qj i-i 1 i-1. ; i Let - i j. ; -1800 ; -N l (l I I I i 8-1 e i l - O - ST.ludtDazn l 1 1500 l.cwor timet of presswo recorder-g 4 -NE3vus-7kNl ' l 4 l e 1 i i /m4ws I i (!\\ -i -l r 1 i l i L 1l I l i 13:0 1 - (. \\i 1-I l 1 I i i l-c "co -1 t' ,1-1 ' l1 I <i i t e ~ 8C N"Qk h I l ' l ' l. l l l ' l, l-l l-l a M/hteSu7 Ant 10C09 t.-- -. - r- -- dt I 10 I D 3 ses wies. range p'ress.wo n. ac,any I i, i L 'i 1 1 I 'I I .- I I 'I y:. g:7eo l_-l l; l I.l ll h7#hg ' pressure es*:Between 12 00 noon and 12:2.reacar_ Jmated from anutepun -ccoing syvem prosaurs recereer l-l. Il i 'l ', Y//b I I i l I.I I I I ~ i: i.l. i I i ilV'i IX I t i l-i ) i l 1

i-( Nli i

l I X -i i i l t I 1- ; i

ri P e% l 1

i-i I ;

f* l. i l ~ l.I l i!.I i 1 i-l i %QNI i T-@ 4). 1 o < 1: ! -! l- ! l ti i L! I i i l I ! 1000 l ) l f f l i l i f i i fi i 'O .1. 2 3 4-5. 8 7 8 '9 10. 11 12 Time #ter Tric'(hours) 2:33. 3:33 4:33 5.33 8:33 7:33 8:33, 9:33 1C:33 11:33 12:33 1:33 2 33 l r-w o-s FIGURE R-1 -.a -... ~

---

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