ML20147C806

Discusses Rereview Re Ability to Cool Core Assuming Cooling Water Absence of Full Seismic Design Capability.Because of Various Assumptions,Staff Must Have Reasonable Assurance of Items Withstanding Seismic Forces

ML20147C806
Person / Time
Site:	Vallecitos File:GEH Hitachi icon.png
Issue date:	10/04/1977
From:	Schwencer A Office of Nuclear Reactor Regulation
To:	Goller K Office of Nuclear Reactor Regulation
References
781004, NUDOCS 7812180411
Download: ML20147C806 (11)

Text

._______--_ _ _ -____ _

October 4, 1977 l

NOTE T0: Karl R. Goller, Assistant Director for Operating Reactors, DOR FROM: A. Schwencer, Chief, Operating Reactors Branch #1, D0R

SUBJECT:

GETR REREVIEW - ABILITY TO COOL CORE ASSUMING COOLING WATER ABSENCE OF FULL SEISMIC DESIGN CAPABILITY Since it appears likely that both the staff and the licensee will be required to expend substantial amounts of manpower in the review of the geological faults and seismic design of GETR, to assure the availability of containment or of water to cool the reactor fuel in the event of an earthquake, you asked that the ability of the core to be adequately coled to prevent melting following a seismic ,

event be investigated to determine whether the radiological conse-quences would be acceptable in the absence of full seismic design capability of the facility.

During a recent discussion in my office with Fred Burger, Brian Grimes and members of the Reactor Safety Branch, it was felt by both Brian and Fred that, although the GETR core is small by compar-ison to power generating reactors, it would be extremely unlikely that the core could depend on being air cooled only any time soon after loss of cooling water without risk of meltdown if one assumed the initiating event to happen while the reactor is at power (50 MWt).

A brief literature search by Fred Burger has confirmed this feeling.

T. J. Thompson's "The Technology of Nuclear Reactor Safety" on pages 692-693 (ccpy attached) describes the results of experiments on fuel temperatures reached on loss of coolant in the Low Intensity Training Reactor (LITR) using 15 aluminum alloy MTR-type fuel elements (of the type used in GETR and the results of later experiments using 21 elements in stead of 15). Based on these experiments as shown on. figure 5.1 and in the text of the reference, the potential for. fuel melt in air " exists for any reactor which is light water moderated and cooled and uses plate-type MTR elements and operates at power levels of over 1.5 to 2.0 MW. Thus a sudden loss of coolant from such cores during fuel power operation should be regarded as a potentially serious accident."

, TlllS DOCUMENT CD;:TAINS

[ POOR QUhUTY PAGES 7 stetson \\

9

. J

Karl R. Goller -2 October 4, 1977 For comparison GETR, which has 21 MTR plate type fuel elements, operates at 50 MWt at full power while the T. J. Thompson article ,

states that maximum power from 21 elements before melting would  !

occur is about 2250 kw (or 2.25Mw). This threshold value is about )

22 times lower than the GETR's 50 MW rating.

Since the GETR SAR, starting on page 9068, analyzes an accident which asumes fuel melt, it is useful to see what dependence that analysis makes on seismically susceptible structures remaining functional from the standpoint of limiting the radiological conse-quences of a fuel melt.

1. It assumes one of the six 3" diameter reactor pressure vessel i bottom head nozzles fail,

2. It assumes an immediate reactor scram on low pressure,

3. It assumes the reactor pool remains intact.

4. It assumes that core melt and subsequent release of its radio-activity would be delayed one hour by making up water lost due to the nozzle break. This water would be supplied from other sources including the emergency pool recirculation systems, the Vallecitos site storage tank and a source of demineralized water,

5. It assumes no breach of the containment building (other than a 1.4%/ day leak rate).

Using the . hove assumptions and applying TID-14844 assumptions, the licensee calculated 272 Rem as the limiting organ dose to a person at the site boundary (206 Rem to a person 2 miles away).

i l

l

4 2

October 4, 1977 <

Karl R. Goller ,

It will be noted that assumption 3, 4 and 5 depend on the structures and systems involved being able to withstand the effects of the initiating event. In order for those assumptions to be valid, and the resulting doses to be within 10 CFR 100 guidelines, sub-stantial credit must be given for the structural integrity of several

- structures, components and pipe runs. Because of this, it would appear that the staff must have reasonable assurance of the ability of key items to withstand seismic forces.

s~

,([. / htMYd6

'A. schwencer, Chief Operating Reactors Branch #1 Division of Operating cc: - V. Stello 1 B. Grimes l F. Burger i

1 11

!1

'l j

l l

. j i

V9NffL A tl, N f I

,,,,,, GE..N.A.U..$

COLU.das t POS C04C8tti

1. ~ \ .

= , .. - ./n ' 1_ '

, =,=: S -

2..

' .g. b.P CCUEa JOJOO=%'s

..l.C ex

. . . - 0 _. . ., ,..,.,. .,. , L_ , 1 x' s ,

aDDD ^. .

6,i +-l %

g_

1

.y D D D D D Q,q\

I' 'l 'kftts "

.~w 1w.'a'r es aim e

' _ 2

.i j .c'.a.w..

m 1 l

., p .

sas,msm, soU,*~ent I- h* }

i:

y[Q g 0' ?'

f soom g

- - ;- .,l i

\

Q .

[ l-* ,

= iC0000?./

i, "  !; 1 l e. . 'e. *: .

a loEEEE /

\.

g%. - ~g .

DDMCODi LuCEC of;./  :

l l

I UuEFp/ / ~

-h

, "}p*

$Ul6OING

,, y 5 HILL VAL 6,91 FIGURE 4-3. REACTOR BUILDING BASEMENT PLAN I,

4.2.3 Vacuum Relief System The containment building liquid leg vacuum relief system is an engineered, safety-related system designed to relieve air automatically into the containment building should the containment shell limit for vacuum be reached. This system is essentially a large-scale manometer and is part of the containment barrier.

The liquid, leg vacuum relief system is located outside the containment building at the

• ventilation inlet. The system (Figure 4-7) consists of a 4 ft by 4 ft by 0.5 ft tank, a 6-in. n diameter " vacuum" stack connected to the containment building, a 6-in atmospheric stack .

open directly to the atmosphere, a 5-gal reservoir which contains the manometer liquid

-(fully concentrated ethylene glycol antifreeze fluid), 'and associated liquid level Instrumentation and sight gages. The tank, two stacks, and supply reservoir are fabricated of 6061-T6 aluminum.

4-7 -

18.' 4 ed ow .T

j NEDO-12623

+

• i 1

[ ENCLOSURE

' ~~

/Nk t . ** ; rl.: a A.7

. ;Q,9'4f/

s.

n .

,pl:'. :l ' .

fizj' ,p., < , , .

,,,.;,m1 s7.,";

a ,

, ,;; , ', f q;.h..  ; . i n .,, . _ s s ,;;.7.',

.;. .. . ., . . y i c :q: ... 8 ., !:. y

..- / . c'.)[v.,.Li,yj j p i .,,o., ,a.;., ; .;. ,, .,p.:j . 5 ,

EXPCSEO STEEL

,. ,.3, ' " [ , ' ),hW' ALLIN5ULATED (g; : A g? /f.h' 33 M i ln. RAOlWS [.k

,' " "

• lg;.g

. (TO ENSIDE FACE OF . .

jE1 '*

N, . . *.'!:::":...,..',*-

5 TEEL ENCLOSURE) ',-

' . ' ChTj"$'M / ""

n

. 'g. g*'Q .,

)h; & G" ~

"' ~

IS TON CRANE ,

d

/ EQUIPMENT HANDLING ..

i - 64 ft 2 in. $P AN p_= . . .

._..=.-. .- - - - - . - - . .

'L FUEL H ANDLING 'g l ZXPERfAENTAL # \

PLAT FO RM g / FUEL \

\. !- /'HANOL

/'v ac^8NGm . . ....i  :/

l

/l N

/

g N jl/ ; 5Aem '

I

,j '

THIRO FLOCR wayggggygg g ,

l ,b ,

ELEVATION 611 ft 7 an, ,

CANAL * .

=i g *

, ] (FUEL H ANDLING) ,

Wel

... ..... j j .

i t Ieght w. ' .. '

EQUIPMENT

• LIFT REINFORCEMENT ' {'

amos .

7 e.., s..N} EACTOR

... . 4 h!!

gg 5' ace

• I I

FACILIT Y PIPE PIT SECONO FLOOR

, N 'd ' *k* I" . l

• hl i , ELEVATION 589 tt 3 in.

CONC R ET E , . .I eLOCKS- 'T . PROCESS +

4

. PIPING

'{

e i

l$

j

-.f* AREA .

1

, . L . -. _

% I ,

j ..-

.U h,$*,h* 'I N " EGUIPM EN T) . d ..m g. }I t IH

>  ; LIFT SPACE 6 ." .E O,V,..i p'.* E.*.e

.. . .r . :

• o .. ..

-\ " ~

• 2

.- I '

' Aree

. A 't

~

,.' "'.i '.l.Dt :3.f'.~ *

• g;.-.. a, FIRST FLOOR * ~ ' ~ '~~~~~' '~ ****

ELEV AT ION 't:"'

j . .. ~Q

• seF ft s in. '

l -

h

• l'"".

' D I s *.

i CONTROL f

j PROCESS PipeNG AREA ROO ACCE5$

M *

• Twgaggg I 1

Asee SWO

,e GALLERY f .

l -- 't'

• I "

f:

ei

[ *

..). ;

.l ' ' *. ', .,

ELEV ATION ,

546 ft 3 an. .

FIGURE 4-4. RE ACTOR EUllotNG CENTERLINE SECTICN 1-4 va6enes .'

f

./ .

}'

e 4-8 I.

1:

U oom* w4 hY

. NEDO-12622 I '

i

. +.. ...

i .'

t j

.

• CANAL *

]

• EQUIPM ENT ROOM goygpugNy * ' , ,

pooM g

e a, ,

• /.

t, EL.E.V,ATO R . . .

e:.: ..

e

v. ., EAc,. ....

g CE,

-j . . . . . ' . ' .: .. :: aC *,

T:

En.-==_~ JJ : . . loa. E. >./ . REE

==

rf .

-ELo. 7

.h

• ta _ EXPE RIM ENT .. SEAM PORT

] EXPERIMENT

• .L CDENING q ==;l_ A - R EA ,/

j==.l A R EA= p!-y :-c .CQWN .: _r

- = = FLANGE I *

+# .. l =.

a=:' '--- L ----

(l .)'

.4 . ,

.'...-i l C_ _ e.4 '

."== ::: l~ ' Iid- I.it A " ': '

w l 1.sia.ru.a p I

.L. !;. SS_ 2 1 l

.

• i - 55 TON j 8 '

CRANE ACCES$ 4 I wa6 some i

• FIGURE 4-5. REACTOR BUILDING sE0cNo FLOOR PLAN t

it the pressure becomes sufficiently subatmospheric in the containment building, the liquid level will drop to the bottom of the atmospheric stack and air will bubble into the system through the open atmospheric stack. The system is designed to relieve at a vacuum of approximately -0.1 psi to ensure the building design limit negative pressure differential of

.-0.20 psi is never reached. In the event of a pressure buildup in the containment, the pressure would force the liquid into the atmospheric stack. The system is designed to ,

contain' pressures to 8 psig.

Redundant, hermetically sealed magnetic switches are installed to rnonitor the liquid !evel continuously. These switches actuate an alarm in the control room if a high or low liquid level is reached. Class sight gages are connected to both stacks to provide a visual indication of the system liquid level. The complete system is checked every 12 montns :o ensuie its eperacility.

4-9

~~ .

ed

?'"*' *r

o *

• NEDO-12622 8

i a

'~

h

/ I \ .-

\ -

FLOW OISTRI8UTOR j

• 4

[

..l $

SNL,ET CI AM 12 in.

pf ,

=> .

l WATER IN i

j U l{-.* /

t-l .. a 9

HOLO 40WNBAR s j 3.s it , ,j p LOCK NLIT l 1F

}

f1** _n I

SUPPORT 8AR i "

fI 3%

- 5 L..l 1

! CONTROL ROO GulOE TUBE

(  !

4

• * *w

.' l J.

• !E iN

~

g "

l PRES 5URE VE$$EL

, )

• }ij  !

25.5 in. CI AM a 0.75 in. THICM WALL

s ll/ l 1

+

. .; , U

f. j SPACER PACS y i: i. rf l-it y*p m* 7,3 1

o p *

,/,, PERIPHERAL PIECE l

i n SERYLLIUM FILLER PIECE

• CORE s .-

CONTROLRCO j '

h!!'

FUEL ELEMENT IP i, b d' -

9.625 in.

f "a #

g.a 're War~, * /

dI f hl'l GRIO PLATE SUPPORT CYLINCER l l' MT  ! f4 ,

CONTROL RCO GUf 0E TUBE

.q in.

4 l';

9 ,

> 12. in.

• OI AM WATER CUTLET

) ll 1:

7 fl w * '

p g o'" '

i

• ca m/ 80TTOM HEAD i

-l  :

UUU '-

1 e

- FIGURE 4-11. REACTOR AND CORE ARRANGEMENT -

p. E a.

I F?

6 4-24 ,

• meiw+ vk il

NEDO-12622 A5c A ",

~Y

{5 C

i l

go

, o O O O o o h/ k, s1

~_~

f 't 8 'l Q __

'O ,, ;eenymoggggugny DC$lTlCN,

; C l

{i I,-

3

~~

I LLER EC FCA  : 8 6ASMET O R EXPE RIME NT A.,

ASS EM BLIES io f-o x of[' i o l oy/' N \_=SMALLIRRAof l

l _ _ . ATION; i . - PRES $URE VESSEL i k <

",~'[ g ' CORNER BERYLLIUM REFLECTo p l

,,,.. , enEcts l l

s e ol e l

r a w FIGURE 4-12. REACTOR CORE PLAN FoR TYPICAL 21 ELEMEN1' CORE 1

i Depending vpon the in-core and pool experiment loadings, the number and location of fuel elements may be modified as necessary to provide an adequate reactivity balance or f'ux distribution or both. The normal fuel loading for the past several years has consisted of 20 or 21 fuel elements and 6 control rod follower sections. Figure 4-12 shows the 21-fuel i

element, skewed-core loading used most frequently in recent years. A 20-elemen- Icading i gives a symmetrical core arrangement with the fuel elements in the F-7 and D-7 pesinens removed and replaced with beryllium filler pieces and a fuel element in the E-4 position replacing a single-hole beryllium filler piece. ,

Square beryllium or aluminum filler pieces and associated capsules or experiments occupy matrix positions not used by control rods or fuel elements. Specially shapec aluminum anc beryllium peripheral pieces surround the fuel-filler array and round the core into a cylinder. .

The eight standard beryllium corner peripheral pieces have 1.5-in. diameter holes to

. . accommodate experiment or isotope capsules. The four aluminum side peripheral pieces 4-26 .. . , ,

5 ou y.

NEDO-12622 ,

0

=

EEE- -

f. i.

'.==D

/ SECTION AM

,**aa# #

l

• PRESSURE '.

VESSEL CONTROL t

h , , ', ', ! Y . GUIDE TUBES

1. ll ll ,*

' I .A.'p WATER l A 1

Sa in. DIAM A 'g

. y; c. 4 OUTLET

• OUTLET g

• i

$l b g@. =i R~~ c% i.

[

gc@'7 i,  %

i . r A.

{ ,

RECEIVER i

' 7 SECTION

, SUS #1LE ROOM g.

e e; -

(

• I MOTOR MOUNTING i PLATE f~

,, ,. - / ' , ,i n . , f.

k ' i ' ** CONTROL ORIVE .

3 , 4 ._, J _. 4/ / ocToa d l! g ,

n= = == '

MC M W '

at  ! ,

CONTROL ROO ,

ma m m ACCESS GALLERY CONT R. OL.O R. IVES. ,,

tt

~

FIGURE 4-13. CONTROL Roo oRIVE ARR ANGEMENT . 1 i

tolerances for finished fuel and control red fo!!owers. Sampling and testing plans are described in the specifications to ensure the actual fabricated fuel meets the design and quality requirements. Chemical, radiographic, and visual examination methods are used during various stages of fuel manufacture to maintain rigid quality control.

. T,i, Fuel elements are replaced or changed primarily on the basis of reactivity worth. An .

i axial average burnup level of 50% is not exceeded for this type fuel assembly. New fuel is usually inserted around the periphery and, after partial burnup, it is moved to the central ,,

region of the core. Normally, each fuel element is used in the core for about 6 to S power n.

' runs where each power run is typically 10 to 13 days of full power operation. Fully enriched uranium-aluminum, plate-type fuel has a long trouble-free record of performance in the GETR as well as in other test reactors. I l .,

4-28 n a.e ,.. tag - p M

1

- ,y

e a ..,. _,._ _ ,

^ ' - * ~ g _ a,._ . - - -

T. J. THOMPSON 692 .,

imum temperthre observed in the 21-element had been setuated. The operation of the fidat case).

shut off the demineralizer pump, but the leakage The observed points extrapolated to infinite continued by back flow through the pump into running time for a 15-element core are plottea the deminerallrer room. Even if back flow had in Fig, 3-1 (92L lt is evident that the 13-element continued as long as poss!ble the level would core should melt on loss of water after cperatin; have been maintained several feet above the at power levels above about 1330 kx. Wh a M-sctive core, _ element eore ee mimum ~"e mer m mbebe __ m nu In November 1963 a supervisor at the Texas  ! on 4lane s_acced of %m wom ne t al '

Agricultural and Mechanics College Reactor en- ' pomts in ng. 5-1 anow cae cen r i tered the reactor building to find that the pool fuel plate temperatures reached in tests in wnich water level had dropped eight feet [50l. A gasket water was sprayed over the fuel at a rate of J to 6 in the demiseralizer tank access hole had fat'ed gym (0.13 to 0.36 liter / sec; until a 500 gal and water had flowed, ultimately, to the hot (1893 liter) capacity tank was empty. Spraying sutop. From there it was automatically pumped kept the fuel plate temperatures below 212*F to a holdup tank from whence it overflowed, during the 2.5 hr it continued. It was cemenstrated letting most of the water optil onto the ground. in these tests that the LITR could re;ect heat From there it went to a dry gully. The pool continuously without melting fuel at a rate of water level had an activity of 43 ascim1. No 8 kw. Beall [92) estimates that at least twice serious consequences resulted. this rate is possible before meittg occurs.

The equation obtained for the fission protract <

power in a sicgle MTR-type element with 140 g l B.3.2 Fuel Temperatures Reached on Loss of of U"' was (81b Coolant in the LITR (81. 32. S3) ,

4 14P (t-o.2 - (t + TFo.aj ,

An early set of experiments done at Oalc j Ridge has provided some idea of the power levels where q is in Btu /hr, P is the reactor power l

.! at which melting might be expected in a tank-I (kw) before shutdow11 t the time (sect after 3 ' type or a swimming pool. reactor due to loss of shutdown, and T the operating time pec) be-j coolant (61 82). The Low Intensity Training fore shutdown.

f Reactor (LITR) was employed for these tests Beall concludes that as much as 75?: of the i using 15 MTR-type elements and three control- fission product heat is lost by conduction to other j

safety rods la a 3 x 6 lattice reflected by beryl- parts of the reactor before the fuel reaches r,

11um on all but one face. The experimental pro- a maximum temperature. He niso points out that

• cedure consisted of running the reactor at a calculations given by Poppendfek and Claiborne constant specified power for the desired length [S3] estimate mciting at much lower power levels.

r of time, shutting the reactor down W draining He believes tnat the difference is ;,rter.rt:y due out the water (no drop of control rods). and then to improvement in thermal conductivity 3croA5 following the fuel plate temperature in the hottest

• heat conduction gaps due to water in the ga;s.

element until a maximum was passed. Thus, most of the heat seems to be carried away About five minutes before the run was to be by aluminum conduction to the metal base place terminated, the cooling water pumps were stopped (weight = 1000 lb or 454 kg).*

and the inlet and outlet valves closed. A port i was opened in the top of the tank as a vacuum 5.3.3 Control Rod Experience (78) j vest. Then a 6 In. (15.2 cm) remotely operated valve syns opened. It required 2.5 min to lower In the University of Michigan pool reactor the level from the tank top to a point 1 ft (39.5 cm) (78) and in other above the fuel plates. At that point the drop in observed that somereactors as well.flat of tne relatively it has blan-been water level began to atfect the reactivity and type control rods havebecomewaterloggedandnave rapidly shut the reactor down. It required only expanded to assume a more cylindrical ah:;6 12 sec more for the water level to drop below The expansion appears sometimes to be due :

the fuel plates and 30 see more for all water to gas formation in boron-containing compounas .ina drain from the tank. The valve ar.d the port sometimes to corrosion gas evoksien.

In.the top were then closed. Temperature read- This expansion has led to binding of contrn . 1 inps were continued for two hours.which in every rods within their evides in the fuel elecer.J.

case up to 300kwwas time enouchfor the maximum There toen exists the poss.bilit: that the f :5.

- temperature to be reached ana passed. Nine 2-hr element within whien the contrci rod moves wi.. .

power runs at specified levelsThe from effect0.5 kN to of longer be picked up 'as the rod is picked up. cnly to hea 200 kw were carried out. 1 ster dropped back into the core resu; ting ::

runs was measured by runs of 2. 6.5. and 24 nuclear translent. This same type et contr4 rn hours at 150 kw. expansion might also prevent a control rod frcm Later experiments extended these measure-mente to 1250 kw and 150 hr operation. These produced fuel element temperatures as high 6 'Recently R. Panter. AERE Harwell. has com-249'c (no' F) and end-box temperatures above pteted work for the " Dido"-cians reactors tume.

boiling. This later set of experaments used 21 use MTH-type tuvl platts m varmus cc .ngurnt.cM :

elements instead of 15 and the rouits had to be on fuel element temperature > carme now-uK. ..

nortnalized by compartr.g neutron flux measure- accidents and during fuel tranuur operawns.

ments in the central elements (1.5 times the max-

• W yms-

ACCIDENTS AND DESTRt/CTIVE TESTS 95 693 i' 7 r '<eq and material failures can and do occur. At least aua=== =cm four of these failures have had the potential for am--. ..- - ,-- " - '

causing a serious accioent, although none actually

did occur. Other failures of a more minor nsture l occur more frequent:y and are usuaUy not re-

. I , ported. Five well-documented examples of this i g more minor type are briefly desertbed as typical.

'l

• • • ' SW1 (APPR) Closure Bolt Failure f MI. After g %

10.5 Mw-yea rs of operatico of tnis lo .'. twit),

s r 1200 psig (S2 atm) pressurized water reactor, the I

/ e amena6 tar s eaneem ~

vessel head was removed for a core examination.

At that time it was discover 1 that rso ad;2 cent

. .[a o enmi e< coe n.a.=4:s me nat ws vessel head studs were br had failed, a serious ace additional studs

.!d have resulted.

' ' C **" * '" The breaks were det we been causec

.'. m ' a by stress corrosion he Type 410

• ** " " '' ** stainless steel studs heat-treated

• "* by heating to the ra: ' ' (9 3, ;o nc.r s.1 ps foeTempersese Winue rwauns nnume.

or cemer runt elemem uTa-twamenc 3 com mer 'a982*C) and held for 0.f core.vs.mus empered at 1000*F (53S*C) fon *-

Brinell hardnese aumi dror;Mg back into the core. This problem is 5(ardness f:ic) = 25.5).

Escussa 1 further in the Mechanical Design chapter. showed snat hardness r. - j BHN (18 to 41 Re). (Dirit. .nn C

yments. Conclusions, and Recommendations it was also found that the . . e boren absort>ers in four shim rods ely cracked (1) It is wise in any reactor where loss of in the high burnup region .s attributed water can occur relatively easily (mostly in re- to hellum gas formation.)

sestth and test reactors) to provide alarms to Vallecitos Boilice Water Reactor Main Steam alert personnel to the problem end to take pre- I.ine Valve rallure (85i. On Surch it. HD. tr.e ventae action, if the reactor is to be unattended l eve plug separateu irom the valve stem in tne - j for an extended period perhaps the alartn should main steam pressure-reducing valve. The inthre give a switchboard or telephone alert. simost instantaneously stopped the flow of steam

. (2) It is well to remember that in many re- while the reactor was operating at 30 Mwit). The actors-especially research and test reactors- reactor pressure rose, and the core void fract!:n the potential exists for ejection of plu:ts from decreased. This gave a positive reactivity effect reactor access holes. They may be elected either and the reactor power rose to 33 Mw(t), the set by water pressure uncer relatively normaloperat- point for an overpower scram and the reactor ing conditions or during the course of an accident, scrammed in 5-10 sec niter the valve inilure. The Such plugs should be secured in the shield by reactor pressure had risen by about 50 psig 12.7 means of some strong latching device. It is also atm) above the nominal 1000 psig (63 atm, oper-possible that beam tuces can collapse causing sting pressure. No serious consequences resulted.

serious reactivity increases. SPERT-III Pressurizer Failure 6. m On (3) At ! cast one radiation alarm should remain October 26. 1061. duttag a series at test runs on I in its sensitive state in each potential rautation the SPERT-!!! pressurized water reactor, the sys-  !

area around a reactor-even when the reactor tem was brought to 221*C (430* F) and 2 60 psi; is off. (167 atm). Abaut three hours later, smoke was (4) The potential to snelt fuel exists for any [ ebserved coming from the vicinity of the pressar-reactor which is 11ght-water-moderated and-cooled trer. A normal shutdown was star:ed and the fire and uses plate-type MTR elements and operates department was alerted. The plantwas cooleddown at oower tecels of over 1.5 to 2.0 Mw. Thus, a and depressurized without need for the firece; art-

. sudden loss of coolant trom auen cores during ment or further incident. j full power operation should be regarded as a inspection showed that the smokecamefromtbe potentially serious accident. The beta-gamma fabric covering of the blowdown-line riser, which

, after-heat can give core melting. It is !!kely that was in line with a maior steam leak in the pren-

• ~i single elements and wicely spaced elements will surizer. A 3 $ in. (u.95 cm) wide, 2 3 % in. 4.;-

not melt until somewhat hie.er power levels are cm) long hole was founc in the central 41:-h Je2=

reached since there is less interaction heating weldmetal. A 1 in. (2.m. cm) diameter oo;t wnten  ;

between elements. tightened a stabilizi.:q band around the vesse! was *

(5) Special care should be taken that control broken. Corrosion on the broken face of tne c01:

rods are u ell desi;ncd. fabricated, and used and subsequent investigation indicatea that the ooit ,

so sa to ensure that no expansion or distortion broke in an earlier and separate expansion os tne .

nceurs in such a manner as to reduce reactor vessel. I i anfety. The pressurizer is a 33 in. (93.52 cm) tre.er diameter.16-2/3 ft (5.08 m) hich all-welded vessei ,

5.4 histerial and Mechanical Failures of ASTM A-264, grade 3. 0.04 ) max. carbon stee! .

with

• 304 L stainless steel fittings and internal Evidence continues to show that mechanical cladding. The backing plate was ASTM A-212 t

D 8 ..;e ,

'ri' '3 ""

'P'm ' ' '" m 3 ~ = - -

W

.M s-.* A p'