ML19308C796

Technical Safety Issues for Large Nuclear Power Plants

ML19308C796
Person / Time
Site:	Crane
Issue date:	09/16/1971
From:	Hanauer S, Morris P US ATOMIC ENERGY COMMISSION (AEC)
To:
Shared Package
ML19308C792	List: ML19308C791 ML19308C794 ML19308C796 ML19308C798
References
TASK-TF, TASK-TMR NUDOCS 8002070528
Download: ML19308C796 (16)

Text

M M,M kqyWW W - t:*f 8, McAnmrW - /

.e

. b-

,,_m_,

.. v. t

.M

^.

e,

,o t) $

4 TECll5IC Al, SAFETY ISSUES FOR

'\\

vn.

m

[

i.

]

M.Qs-

.LAltGE NUCLE Alt POWEll l'l. ANTS I

sg

\\

a 5.!!. !!A NAUER, P. A. MOPMS Ms

.vm U.S. Atomic Enegy Commission, Waeinpor.. D.G.,

g.

United States of America l

gs -

g Aktract-Resume - Armorauu-Resumen

.d.

..n s

hlUW

" 'd.

TECm.K AL 8 ArrTY IS5Ct> fW LAftCE NUCLEArt POWLR PLANih.

k Although the basic rt enrements for and means of achievse; reactor sahty rerum unchan,cd. mucti h2s been learncJ fro *n scscarch anJ Jctclopment projccts and from analysa and esptrience in tl.e Jctiga.

. action, and ops rarson cf pomer re. actors. 8afety scvie=s @nn ths pist fa years by apuca d* and p,

the t*5 Af C rcrulatory staff. tocether with the US AIC Advnory thimim: tic on fte tetor ufs. orar. have y

rewlted in a numbc r t.f changes m piar.1 design or opct.ition. I hn paf:: i.ncaocs some ofr en mi

.i

,Y ?? ?

symfacant of ths,c et*arp s and the reasons for thcm. f.namp;cs are elet n tiot ret.* to aire and s miron-t -[w g

# *J' msntal char,setcristics, rc actor design. coolar*t sy. tem des:pn and ps rfernia ice. Vcsign of t'ie t orit.inment ed ott:(r struciares, dcsign and pc formance of engmtered s.tisty features, smtrumentaten and powe r-ayitem q

e design. and qual.ty-asserance procedurcs. Althou,;ts much propcss h.u been madt so rc.ector s sfety, rnorc y f h I rematt's to be vanc. Lspuience with and analysis of actual syst< ms unco.u new s ifety im.c and os *.upcets

.s me of the n.ost important such tschmt al

. M* ;

of old sa ety assets that isqune for+her study anJ resolution.

r o

ufsly usues currently unks considuation arc ducusscJ One of the most cl.allengir g sucstiom as thc y

qpopriate balancity cf enk and buaf at. For the safety revin wer tha ms.ms estabus'ung acetptacic rms

]

[

for accaJen's aad anticip#ed opera'ional occurre *ces and t%ref~e reliaNiity crucna for p etectne systemi

J, '

and engineered safety featurcs. Available tcchnology and mformation arc rsvacwcJ. and pm.ciplcs arc stated for considenag ruk analpis in the abscnce of adecuate dal.s.

t he Frfermante of protective *ystems

• I~

6t#

aid engmccred safety features must be considercJ as well as their icita%ty. Protecin c sp! cms w hoic dI e

performance is stuessJ incluJe the reactor prolcction tristrumentatmn spfsm, ths crecrp acy-cors -coolmg

. [ [

• 8 nyttcm sysicms to control cornbustab!c gas mistures in the corn.immer.t. anJ sy sr ms to toutsc! tL< r. idio-y s

activity in Inqu1J avh! gaseous effluents, it is becommg ciiAnt that hinun cerors.md commo i-moJs f.ulares

!l h

I'

% if sif scJundant cicmevi may be the prmcipal causes of protcetoc-system f.ulure. tlc u<c of wartous types

'f of diversity as Jascuued.

.qJ N

DONNELS 1[CIWt8)UEb H L A

• A LA s[CLTJ IE Dt s ilE AM Ltm t UCLI Atil5 h DE GR ANDi. PUISS ANCI..

bacn que Ics m6thodes et let donnits fondamentaics pour atturce h secunte ics #Eacis ars n' aient p.n t

Change. une information consaJirable a ete acqune grJcc aun pro) cts dc rscherche et Uc civelopps ment et au cours dc P analyse de la construction et du fonctionnement des tsacteurs ac puissance. Au cours we ces y

Jce nJres aantes, les rapports sur la s(curiti d6s rtacteurs, etabtn par les equipet pro;ctant la conuruction fy3h de noaveaga riactcurs, par les responsabics ue la rignementation i rl'SALC et par le Ccmrt comultatif sar la QJg secunte des pales de r Us AEC, ont aboutt 3 un granJ nombre de changemcyts dans h comtrucuon et

' fr,. g r esploitation ucs ruacteurs. I e memoire docute les chancements les plus impo tanti et Ici ranons aat L'

ics ont aments. Des stemples sont donass qui ont trait aan caractientiquei du site et de r cnvironnement, T

'/SQ*

J la construction Ja reactcur. J 1.s construction at au fonctionnement du 9stime de refroi.uwmcat. Jh cor.struction de l'cnccmte f tanche et I autics structor", J la construction et sa fouettenwent des anpositifs oc sicuritt, i r instrumentation aun dnpoutifs de pununce, eI aus meta.edes pear rnanntenir g.?

• 2

,{%p.f la qualath Ju mathnet. Picn que la s(cunt 4 des riacteurs ait conudcratiement procressi, il reste encore id beaucoup J f airc.

I' un Jes probitmes les plus(pmeun est I(qulibrer les nsec s c e les acant ues. rour rciamen de la sicurus cria sigmfic qu tl f aut dif ante des nsuus accepimes ( accident a t autres

% p..

esinernents pouvint se poJutre au cours de r emploitation et par comduut nt As cr.it,o A sisunti de

hpf;

D fonctionas ment pour les dispourifs de protection et les systtmes de securite. Les v.farmatiens et tectinmucs anponsh!cs sont stammtes et les pnocrpes (une analyse des inauci, ca r bence R Joaacs safinntes, h

t i

sora itabhs. Le foncuonncmcnt ces systJmes de protsctwn et Jss Jnpont.fi W sceents 9 itre e umme, Q'

ainsa wuc 1 ur siircil Lcs systs'mcs oc protectm i dont le fonctionnement cu ttuati coirt a,cnt r instrumentation pour la protection du r(acicurs, ics systemes de refresansement Ju ci.s ur un cas d' ac nder.t.

-[

i e

les dnpontifs pour le controle des milanges gazcun st:flammables dans r enccmtc tranche et les systemcs cc Te-

. Jg u.

POOR OM

~

?_

S N

e-

., ~

P. L,

-,a 1

k h

g y

. m wp e ggg+w xc e. h.w.

.,e.s m,. w,. x. a. n,; b.,,',4g4 p v w

'f W

m,m..~ w n. w g p wpy

. ;w g

w

< arm

. e a w g y w ) M-am

.,.: a ;

p va.

n wwww.

, p s. : e.9 n.

%3[ww,,i 5

dDbKQ ' g( Y & } **':s**? Q n ; %, t Q W. Q %sll %q' [g'M J I'm.,L.

_.dE, t.,

- (%f,.

,t' _b y,.,. c

{.'.. ~.,, [ * % sh.., ;Ol.d. 7i.

.} ff y,,

.o; A,

r m,.. yn

_g w

,~

.S ii&

G Ct ffh W!g <*? Qg)Q%. [.y j,

y

, q f.

n

,e p

t y.. u m.. y,. m...., m..:. m, w.tt m %;%93p. m m/w...N.y.

~

rx

, n-M&hhk, f

,. w&m h

wegym.co,&f.g~

wy G%g* C%

p ; g s r w s &p & q &,, ~;g::n y%ge% QQQ ye s

s & nK %

t.4,W.,. Mm., ;;.

,f e

~m>

s

.s.

.o,y

. jp _jga,

,m

_ m aw

-- je

-.y..-

l:L k-

1. -r

.L m #t

,s

.~'^f 4',

l m

ne n,

m.

]

9..

q

1. t. l.. i. r e,< -. ci J.

• dria u

.m *. t ' i,

a

.sn 4

3

)

...pe at iu c.u 4 p.g.a.

t 1

i p

.g 4

y~

.t

.m mm. ss.

,r, Q,

4

.,y ir x17imt T A 111: A(JMI:b T Li 1:l%II ACf!' W rif f I'V:!! N '

\\?Il1iN" m-n e. x..

r ; is... n, aet u e s noi n ue ri e u.r..ne n a+ <.i *a "

e re e y-

y. f '

n;., r am e = eo n aucimuust crymec t h1em'e utt le arW ii <

• a man +<-*. Wily

• 9*wt-4

> remn, a r a..e vs. n.c sence a.o.i.u.. n nonow e a.m< uim r uo i, w nn.. wm y>'.'su+ti edo

.I i t ;.,n t e e, r t e.a n n e.n iya r-site n surp ei n we m s g ea k t i-;

r. ~ >

e l't..y s ia n u ++.*ymn.. p...irm, t e. Ge so.ut r m a t r. r g + o s. r., i n< 4 v4 tew eoe meseam.seeteu: -

. f," *,

[

).cn yit:arew-

+"',7 3

q, tyrw

v<sen n Jas on@it t tv uuut ers an.s>ta k N - ( ;

i <oss errn r

~,

i t.t t.u 7 f4-

f

t

., ip,,,,am 3 cio., g g. e g;t A r.o, a;. nt ne w Ce.ecnar nce t o j c.o t r,.,, e rm g ; tt tu sn n e 4 1.1 si 6

- 5 pour try whe si ri nape.ac se r nlyittame ye t.enones. D > > w ae g r st ai n rn. r e n seen t & ruc 1C ;*.

i s.

ni si 4oice ryter teenun x ushienensiis n 56prN6 nst six Hits;e s = iW hoJe a c e ei nave! 6.

• Aca' i '., 3. 5 g'

umneca x..pa* *epnc f un reaosa.wam ais cig.oute is.c t ea u op i y

...m.en e s asu. stac t; ya en ; m-desig i

3a u.n t unD < % im e r eas t s-heis; si is,g a, new rg ymune n p16i.,r ts crc teuw t enionor s f ein. nur t; ye a 4

k, r~<.

g n ar ys u wo-ictrys u.io, or.r/,eimor ten mouc tg ymimh n (c ame.r tu n x g#,ot u, sou tl y* mu ni;g;

,m,.,r o i c n>uen"e iu',imara'rru

> -r> oemi-a < >'oe e u. a ' a

• u e rca + 'w er -

J-w o n 6 t c t s,a. lier uus ps i+a to isto ye e moctornyt savsni e e n t n * ;t seo.iu e npvGle u si oGec tie = renu s 6esoinst.i'.'c ti$ se e% s ut..ec ne n t u c i tig u t. sang e h

• OCi '

senna e aton od iactn. motopue tperyrt caocro sati.nen.act o n '

n n t-emenne. li Jossame D'. L.

3arg ar.spactc.s g!as uan6oace ma e nut te runger au x ac uen ce

. ymne e..o.-

anna e.icmac on in, namo2w w v e e*

r"----

py par.reemce sn euw n t rasme p t ruotrenus. Oamne u 3 c auus e

.nu s penruon sn i*c te w mor-9eS E pete, t pym.ee r enor ianicune perma o sus cau. /11e 9c194.cma.,anewamer r..t a songoc am' /.cao-

.1<

am nae noe rn..sto o ma eact yetaur+nenne nppestientum rnre ati e e ty eae abat en n uosuo = nu t "*

c f "p*'

IT.1 a

a 1 *. *[

citlyataanoemu x c 1y saiinewies), a norouy n o pet erre> pase u m,< va samptnu t airies, o remon%cc -

>. ( <4Y p ux ar ner sva 6esun < nut tu. Par cuatg.unactes conrettennas tewo tos us u cut uu.ser g en. 59; w a-3 ; w..:3.

un9. ef 3 rat amtr e ripususessu an.einsa gw na a cayteae etr.yi r.t ime "evnoas.unrx sanssp x.

l'a4st s sk.i,, ~'

~. -

4' c,' !

y 3.es.nirn ux cetteu si t exnst**eczne acnent u Geaonat.noc to a,., r s o u pu.v.wa n g ni a r s w s.uce. re a

c me npor.aun o,ee ne.eeneen e : naacanocin. Itay.eacu pe 3 iantieu e e.nr teuu e m uo%sn un t euy r

.e "8P e

d CW * '

M.t NM" sontpo ono-usucpercionon armag,a ryru no samnte retant ora, c re i vre.sy sie.,*enest.,r o g.se xt 14 *i'-

. w,,

3 maines au tutiood s<mu teamtog a, o,cteu p Aim montruie ci g. ass e.. nea n ; armx e asceu s e ure est 7

.eg g teO. n noa 34..t noi, om,ie..r.n pe a. wp a t a n. e o.c t eu p n.,t.,m e., a 2.s... u n.oe. rn n..,2 -

e q

ant n rasonu x us uv.zas. C t.mobotce o seent. nun, aero own?mn 2.'t.yr moeuuc wrieteput, n Vi3 V

OGene uenoiasan y unor ownciennux 21euenton uur yr Cu t t ri naipun.q u..eiun eu w-sta.I

,gg g as na ctpe curieup aamttu. Pact uatpneanric s psirwue e au anapanu un r et yanar.

g Nh]b TIMi.15 M As 1i C.it.t P 1)l. M 61Tilut)l'J M A';t)b i:t *:i s Au s

• l's !! 3m -

g.e ; 3 P.

4 Auuvue lis toi>Jiaono t>n,J.unent..ivs y !vs nu dws pau lor..e 1. % cuna.J a L.. ec is ts 1. i pc...u ccn g r;i g a[ a. tjg 3p y 1T; analterablo, *c li.i proeruado mutl.o goce n a bs Acin 4JaU t we msut;gacgon y Jg t srgo.

.UQg}

4 (ir a mper 6ncu t u matensi os pro)t cto, constmcc />n y f.memnim.s ato re,euro Oc pot. m u.

1.n i +t.. a r rperJ*

s de se gundaa realcados s:urants los 6ttimos ancs, par los soleut-m'o A wto'azaciwe y il po+en d s t y !s ui

.53..

', g3-w de la ITAIC, pint ime:ste Con el r:omHC Conwitno merc Sais ecuatut.o es h-actors < A la i S A!i

. J is:o O -h :r*

/ T[

por resottaJo una terns ale moWficaciotics cri la concept ab tice tca & :..) pimt.it o e n m fucci.n.m a o rg$C1.

4{

la li rremer:3 se (Larnman algunas Jt its moufacaciones ml't Senkfic.givat y las ra; or en a uw oli mu.

C V.'.

• J.

1 ^7 7 Se cyone n tjcmples scieremes a las caractcri ticas ect cmpiazir.aente v Al n.cdm.mhets, a1o C C'" ?!

7 0.'

pirticnbriha s Al ts ictor, a in caracteristien y <ficacu as t snie m, e ri-fn cu inan,.i s.s p a. 1 a.a a, i y,^j N-A h e.ontcncsSt y de viras s structurn, al tipo y 6 fac ua a :n doppix.m wac..mco,se aes,rp,

la.

5 % -t i -

% a.T, caracts rbtien Al butNmental y de la ecJ cllctrica y a 145 umi n.tv c inti n Ii calada J.

k

• E s s w

tVCn*

11.6.narr ada mucl.o en h scpridaJ de los reactores, cs foe t=I4 m'e la a, i.,m o p.p h +c s 2

e La c ip6:a naa aJantflJa conlessutemas actu.nles y si anAla as Jc JacFos sets mas wtu' acuem pro"h mas Je as enrow y th0 JOE

.F" A

=

.npcctm aniditos i!c pictiema, uuc regincra n mfis ampho cist.ua y seme h.

.w p sa in nu.e..!wM A EVen -

S */"

{s aW los mSt smpertantes prottertias teenacce de scpridad actualmcee s a cc: uo, t na A las cuc.tsonu inas re a : *,

e ra si s sps eto i n rtic ularies a:

COnO!*

j Jaficiles t's concruir uri ci.1:aliFrio adecuado entic tic 5ges y e m. f ac 6 >

a is punJLt. < sto ilentfica fspr los res spot admnatics de acc an u s c e,q. gei o p.es s e;,, Js N acua.wny fm

}

4'_

y por t arn.. stth ect los crsts tws de faahlblad.'phcaPla 3 a t a p, u 3Jm y a ! i ep.gues

=

,pq,

IC c tac ci G( r.' c ar &Ja d.

Sc pau scvnti J b6 conectmsenrot c Oc e n.

p., ;. ) w. s pn m gm,r gi

. W, ~

pra ct asiten a twsps, cu mdo is carcts Js datos aJsc @

t. (.

usr.nsmi..;..

.si,,

Ly

r.-

w i n t t. :-

g %,7 uitem o prof cciorts y de los appostinos t2cnicos & scf.ind v ni ce

.a es euL. w is tw wu d,

( :j f estudi.i D cficac64 de los siguicntes sistsmas prottetores: t.cr%11 m pgcccio:, A 1 ts.4 tor. utm.i ec 05:,!

s

,$.,47 refriget.icab Jc tmcrgencIJ del nuciso, Hsicmas para cont:01.: ) e, em --l 4 c 6,out s edueh i n b

}.n ; 's 1 -

. A.

.A

,e

- ~--

,m.--

g

.. s. 1.

,.~e.

,,* s

-s 4. WM p' V. r%M,10 QW ! WA.T %iy% *,

-~

x...

~s.+.

v x. 6u. ~n,,n

% > x.

,.a,

.' i.r.v s d w,y'y vge [ d sidwf W

.d>~

. y. -7

• e M. '. $ h'.

4

,4.#

yg,, '.

t

,.aM 4.

,.5,.,,

I*

A I.K e',(.AM N e

'y, c -/

• .,,,.[*,

[

t '-

,,. [4 4

+

r.

a s 'rx%t M.*QY.v.n w m w - m :. v.

a, ~

u>..

..u.

w.

a t

• r llQ}ff.t 9

• g*,e k t ". ' "".

-"si

<,t

~

I}+

v,Q '.

L o

. M,,,x, % ). w, _a.-... e.,,, <.

J7., 4..

- e Ma.'

.M, e

I ( f,i *.r w

v

-.?*-

• Ne,

1. p4 M'

• • , 4.

p

,34.<,r t. g*." p- =+ b.f m e,ar'. =e.ps, w w a vu N g.,r 'w,.w v. eCe

.i,

a f

t Ns,.3*s,^.~

s c

.r a cy N.- t.

.. yj.

sf? d.,

4

- j.

.s.

. u e e..

r

. s..

W4w. m m+.

., ;.~ %,. -~'s..;g ncy Q 7..&. e s

\\.

q,

,p$ m. m..,... ',,

, *w-

.,.,o

.G.,

~.

~

ir

,f;'

.. f

'4'

.I m

w 4.c

. ~~ -... w, _

w-

.. ~.,%w o, 9.

% ~ qz,

.e' w *F"J %+ %e ex

=a^"^"'

" ' ~ ~ - - 'eA.h ' ""

~ ' ^ ^ ^ ^ ^ ^ ^-

-a.'

-s

. y.

.t O

s su G.

f' t "

s:

u ros im.-

an A,

3 m

w..

.- e o s a " i.i

- o.is y%..

,m,,,

,4 i...,

we 4+

i,.. i u.

mr x,.m, h'*

s

.u, - i n i tm e m. t a n.i unc

=

t p

. m '

.e n -

we s ee g e 1.

I '.TP. T D' J.* I 3 '

c,.

im

'E y "d-Feacter safety - the avcidance cf - ue riri. to the purlic

,? g hed:n and

.a f t ty f rom tha cperatien of nuclear m a:ters -

tp "m"**

cal an.t nontechnical issue.

Set *in-the

'f' A -

inye iv :s 3-r w +i s

vi tar m

,7eals, r.umerical er otherwise, for acceptible ri:> a mt safety g.f

,. n,

is primarily a :s:ill and political task.

T ra n s l T t i n *. these

..y.

T.p.N. *

u.. c a.

goals into criteria arainst which o e.valu ate.pt e resals f o r. the w

- sm-design and operation of reactors requires social anl technical m.

a r m fu*

h ci ;iens arrived at with creat difficult /. Men cf the tech-r, y

' Q p.

' 47"aa nical effort in rea:ter safety evaluttion is hvoted to a more

' f * " P mundane task: cc parisen of specific reactern ac tinct uch m C

. bF

' ' * " ' * ~

criteria. This technical effort is perfor ned principalD by

'* %.k

,j"[ * **j.

the plant owner and operator himself, or throuch his cor.tracters, a

since the owner is respcnsible for the health and safety of the

. gl h.u mm public as well as his employees.

q',ly 7

-u e s t.;<.

nacoor-

.v

.on cese In the Ur.ited St a t e s, a s i".n other countries with nuclear N M.

reactcre, additional responsibility for the health and safety f f-1

' m ss-J-

't*5 4]

cf the public is Pcid by an atency of gevernment. Such recpen-

' m = 'c-

"'" 10 8 " *

• sibility necessarily involves safety evaluations of the same M **

type--if.not of the srme scope--as those conducted by t.he owner.

3 In addit 10n, the governmen.t agency plays a ma.jor role in the

- > n p-m se

.e s.

...ct a wy q.%e development of the criteria ref erred to earlier.

In this paper

.. e an

'the authors discuss some of the scre important and interestinc M M.?

swa.

technical issues that have come to the attenticr. of the Ad-

~M M w.n visory Com:rittee on Feactor f af e;;uards and the AEC regulatory h

1 MW so-staff in the last few years.

- y[f;r}.'N na.

Some of the issues discussed ar'e considb cd to have been i

resolved; that is, both the criteria and acceptatic desirn i]$

pomanecen approaches to meet them are available.

Ideally, experi % e and

%]k J.hus y measurements will have confirmed the dec;icn, construction, and waos operation of the devices er features involved in such a " resolved

'i. N it competente issue".

By no means all safety issues are in this happy state.

i /,.4 M@ *

. hui dado Others remain, to a Ereator or lesser extent, short of complete 4= mo.

resolution. Lack of resolution often occurs becnusa of the in-

$ M, Mi$

• Jec e n.

evitable time delay between identification of at issue and be completion of the work needed to resolve it.

./ se6. 6 Qg.

meatuwaoci 8

.a. a la for some kinds of issues, complete resolution is nst fere-

.$q%f we ie seeable as a pra:tical matter.

Such prcblems cually involve t

R-N opmencia events, or combinatiens of events, of such low probability that c' J.

% T,$,

waaa y thousands of reacter lifetimes would not be expected to encompass mnos o even one observed event.

Some postulated occurrences used for nmh reactor saf ety evaluation f all in this class.

~hus, design, M7

• es ee construction and operating experiences are not ex p.> c t e d to cen-

7. tW

iew.

m o firm the resolution of this kind of issue.

N$f. 4 epomnes

.y M

. pngmi Such rare postulated events--for example, upts.rc ;f a fy acu ee in lari;e primary system pipe, er failure of the centr:1 red: to iN-Q-g insert when needed, or occurrence of a severe earthqnl <>--revire w

'Q $y uncms

• car eful censideration in reactor safety evaluation even though s en ta knowledge about them is an.1 must remain incomplete.

'.?. '4 w%

34 rn i.*, ;. -

POOR @ MAL F,

]N

.,(f

-f 9,.

.it\\Q':'. t

<f

.. [*

?

b

, ' s j 't.f',.-e * **otq~%.s %yg*M%y

1. 8.pN'% 4 b, f'

.~b%%.

M 7

,g

..y 1

<' Jev y ta h.r#1 7 W'-

-C--

5

'JW*

ys 3

4.

N

• i' 9

• \\

U y*p V.. &% &'y.'o).x p... ~

.-a eW en :,,e.A w'.,

% #w'f.,[.h. "e

~

V.-

-. 5,/' a

-. /.'

k l.

~

k -%7-e v.-

4.

' Sg. ~ ;

QQ Q*[s '? .pg 1 3Q~[ g*;Ce'f,WlQ?,4 Mk w,

e4>

~'(

t

. 9.. ~,..s,m 2..a.,Ne. r*%a..,.n.~.

m m.

a.

. s,N c.., ~,

. m

.. s,e @,. 9.s.-.*R - (sN w g. 4 +'[Ja+

%Q;.g d,,. a ist s 2"

M4 - J...,.,.. -, y '-

.n i, gy

• g;

,.j e-1, 4 -

s*.,

s e

f a

.. e W.3. w...,, !oe, y. p. 7,*

J q, -~ *V A Kap prQ $.,im e.* ='y *M-[v.

"l,p,

..s w +

y..-,-# p 4&rfi1 y,y 6N f%. - %.'. - *-

..., p ',%

4*s.d_ u,..,,..<

e-

.9 9

.w.

n ur.s 7:-1 w-

..y.

n w.m,g.o.y. ppf,g.;, ga,

.,.4 W~, a y ;. 9 e

.y.

-n.

w,.

~,. -

9,l

.w,

• m' - - N.

n

, e*, 6-T-; *.

.n.,.

e:

p3--

4 W

t'

.. -.- A -.'M.>L.i..y_ N..R~, R.'fm.,A - _,.. _._.._

a s

g g,.,4 c--w-m - --

L }t?.W p

Am-mv WdI4

%W % W-V

'M yg -

504 it mm,m.: umts

~

w s.m

.g4 y_

D, $j It is noch issues is th"so that conn t.t um t h.

- e s t W!*1de,l challene,ing and difficult portions cf safc:/ evalsatico.

Ceciciens pb+WW on such issues necescarily involve j ud.

• he ust M nh os-e

.iecut ccmplete information - jutgr+ntr,

M*" fore, ntj": to Sj

[-

. M-chance as additional knowledge is gaine*

This paper discusses the general cubice of reactor cafety

% ;"7 goals and criteria, core examples of technical safety ico ms thn 52' S#

have been resolved and finally, scme examples cf technical caf ety P

', k issues that are unresolved.

,h.."

+a-g

.c_

g 2.

FEACTOR SATETY GOALC A*:D CRITEP.IA ex.

-t s:.

2.1 Safety Goals g;Q The ultimate reactor caf ety goil ic eliminatien of risk T*

to the health and safety of the public. The principal potential sir hazard associated with operation of nuclear reactors lies in the "o

possible release of the radioactivity of the heavy elements and evt fission products produced in the fuel.

The safety goal, there-ri fore, is prevention of the exposure of people to this rsdio-activity. This goal can be achieved with a high degree of assur-ance, but not perfectly, since the possibility of fuel failure ci bl.

'[.

and consequent release of some radicactivity to the environment I"

cannot be precluded absolutely. The ultimate reacter cafety c a.,

goal is thus theoretically unattainable, and a certain residual WOf risk is unavoidable.

anc.

IUY The minimication of risk is a realictic, practical goal, but "O

it introduces a dilem:na: What is an acceptably low level of risk?

the What is the basis for choosing it? !!inimization is not enough.

asc.

By choosing operating parameters with greater conservatism, o.

adding more safety f eatures,. usin ; more remote sites, one can obtain at least the illusion of greater cafety. This precess I.

has no end except diminishing returns, or at worst a decrease in overall safety as a result of overcemplication. What is P8' needed is some delineation of what is enough. The necessarily imperfect goal must be translated into a guide to the designer and i.i the person charged with safety.

th:-

se" The decisions as to what risk is acceptable - as defined bcb l

piecemeal in criteria, standards, code,s, guide: -- have, in the er k

past, been guided largely by experience and judgment. for e c 7.

some aspects of reactor safety, formulation of criteria in this w

way is a relatively tractable tack.

for the rare postulated events discussed previously, however, thic method of establishing let 8D -

criteria is difficult for two reasons.

First, the probability -

Gaf met expected occurrence rate - of these events is far too small to

[

permit the accumulation of experienc< necessary to proceed in a as I

meaningful way.

Second, the possible consequences of some postu.

ap; lated failures are so severe that the whole procedure is not lat pg;g -

bac applicable. Accumulation of experience - even if it were y

possible - is unthinkable for highly unlikely events that might ti.

.W.

4 kill or injure a very large number of people. Since the accept-2.

ability of nuclear reactors, or anythine, else, dependo en the fact that the unthinkable events are very rare indeed, an alter-native must be found for dealing with such events in the estab-lishment of criteria, saf p

4.a2'ap2-P00RDE M

.n 4

w. w,, -

f*

e U

5= T * '

, }tV F

{? W

_ w(gn.~hh.m.M.,g

~

~

f'} '

^ hN Y f-w g & i; W n g[.j.,, & &.I D h;N,.w w.%

yhk$ 1 h.1wy'N' OfS'q,,. " !O$

m,Q' g': eqgwe

~

s,.

A*

Y p

y %m.

a [ Y[ h h,y # h. m~

. m.hM

'sd M

~

L g

as n. n.v.

n, n

(h fh h

'.[,.

bh *5kb NbN k-M% M%c.g.hvr y M. &g,_..

m 4 r w w w g g., w ? $,_ G h

n u n..

~

~

y HM.

X lf m m. im 10s

, g ;g g

2.2 rin

a l vr i c Mg Ecisions Mf

, semincl/ mp:t attrictive. nee for

,1 : e

,-..r G.

g n

12 fe y gcais ar.J cr iteria 10 b ine, c" 2

n ri ai

, _ it;m

,y

'~ '

cf ':

x.

In i t s s i:r p b o t f or:r, the thecry @fn. ' ri 9

, the

b product of the prcbability of occ.:rrenco cf an <.- ru :ir a tho B

gg,77 cer.sc w n e -f that event, also suitably quantifio l.

li-ilar,

- $,;[7 Y

' oro comp.icatcJ meth?ds alto have toen trop m d.

T h._

- t d ri-thht' l

1 sdety

30. the 3r. of the c11culated rick:

' everythiq that car h ppen.

.nis approach 2nd c thers of this nl.ure are n:*.

the id ir.ite i"N%

tacks they would at first zicht appear to be.

A fm ev m : are h:'(pW found te dcminate the analysis. For plant to Le accepuble, N

extreme cases of hirh probability or (at the other end of the u

scale) sericus consequences must be found to i::volve

-1 Irw rick.

,, k,v,s.s, e.

i N

he potent 12,. advantates of such an apprvich are e

.icu,.

4tm' ~

m rish The evaluaticn autcMically emphasizes event

th it c rc :- a. '

P'os atSntial significantly to the. risk, which is where the e phasic balones.

s in the No decisions need be made about credibility, since " incredible"

' b:((:' ~

nts cnd evente eliminate themselves from consideration by their low

. th;r.

ricks arising from their low probabilities.

'$1 P

dio-g@

l

.A s

of cscur.

Closely associated with risk evaluatico is t h.- techn logy y k1W ailura cf reliability enalysi=, used to predict the required proba-

.QM; conmsnt bilitic of cccurrence. It is in this area that prcblem: arice, t%E*7 c

in the authors' opinion, which decrease the utility of risk jf, 6-calculation: in reactor safety evaluations. Tcr all iti proved M.Mi.g g

worth in many areas of technology, application cf probability g].

analysis to the t y;. e of rare events considered in nuclear :.afety nt w

.M;. h f involves the follcuing uncertainties:

(a) There is little or goal, but 1 of risk?

no concrete evidence,that the events under consideration obey V WQ;;

"" ""h*

the laws of probability that underlie the thecry. There is no

-;J M

a:surance that accident events are random in time or independent 42 N.'

s f each other. The intellectually satisfying idea of reliability h Q

oc =an analy:Is is no more than a hypothesis f or these events; (b) Even ML ss ry if the framew rk f the theery were correct, the values cf the 2j E parameters are larcely unknown.

d J y.

g %.z

, 1s,ly v

stri The method now generally used to predict these probabil-i< #( MWup

sign:r and ities -- cascading of probabilities of the individual "f ailures" that make up the event -- 10 known to be inadequate. The

'd $4 serious or potentially serious events that have occurred have i/i ME:

f, d3 ined been characterized by concurrent failures, usuadly interdependent M,LML

,2.n the er caucally related. Thus the theory's assumption of indepen-Tor dence of failures has not been borne out by experience.

/ ky?u in this

?

}W -d' lated Despite these chortcomings, reliability analysis and, k' h e QL to a 8

tablishing les er extent, risk calculation techniques are uced in reacter sbility -

cafety evaluation in the United States.

In scne arcas, such as gaW

f. ff.h.9 f.all to i

meteorology, there ic a substantial body of kn:wledge to serve as a basis for the calculations; these results are thcught to

? WM nd in a ema postu.

approximate reality. Where the basic is les: f irr., the calcu-MI 4,U@~

0 not j

lation are nonethele s useful in making ccmpariscns en a relative ba. sis, as oppoced ic an absolute bas:s., thereby directinc atten-i,s 2t might.

tion to the c. ore impcrtant censiderations.

w-gy

.p,w<,1.

. 7 e accept-i 2.3 Desirn Basis Accidents fhf '

/.

n the In alter-df* $$

.e etab-1 The principal tool used in the United States for reacter y

1 saf ety evaluation is not risk calculation but analysis of a W

i M-

.AlW<,;e.

.up 86,5 Y;

s ^4 l

h. fhib e k

$h

. ' % %.pp&p %w :M W w/?p%.>9,wkw.m us x,

s sl vs

• ~ ~,

~D w

wa w:.ww

~~

? <,

. 44 f y r.c M.at** b ts;44: -

k., ii SY c....

..u p&fi'.24EQ'.&s,ff.,4,W *~7:

M Wu.m 1 W-9.w :p~su.s ? a w n -

tla m

~

u v-3,1 ~5* 4 q O s*&'x.

&fUl. * "-

? '% $ s

'Y~

.Y -

• ~

M-g n d MN. & w.u..N'O**-*45' h-w=' V'b' :h' ' k" G ? * ; 4h ' 2 t.J. -

G AVW.4Et<3 N W

' 4 " ": ' -

Q* ?

• • M.W N5ENk b5Nd$'

1. 1. ^

t

• ?

, - *^~

kc 1

2 0-

',s a. m..v, n: g'n y e g-

.ww

- a+. - n

- m~.

m-

+.-

w.

~.1w v. w -yr.-

..ua

- _ =

-~

n.- m.u :.-

Q.

Q, ' ~ ?M?pm:gampy

,.g g j

k [,hh hhlff MIM '

(

. sa. t

' h.n :.. yc..

smrw g cr~d 3

.m a

A 4 a,. i imm.% me UMJ 20G

,g,,-

4-pectrum c:

'e sirn tau s -1c:1d. nt ;.

'k r.. v r..-

of c c ; an:.

% :^

s.

?.

g.

f.

plant to nch or tho e pestalated etent; 1: y,dtrary decic.

.k. -.m g : y,r gf radiolacical councquences are ccmpu e uitr.

se i...

mintj by f[7 one for events op ;t-t

>mr M

Two such criteria are used :

evcnts, frequent 11 during the reactor li' c

• im, t h-

!?er for ry u::-

1 ts 30 severe hg lively events.

M-livision of cis.;sec c + p ;tulateq curely possible and has been prc;;ce1: the present state of ovents n]: te

' *h7 knowlege appears inadequate to the authces to justify panti-

. p,M fication cf intermediate event;, but qualitative compris:ns

~

n..

appear to be meaningful.

nl,.

str fc+yg ;,

'NN P,

It is evident even from this brici dicca ion th d despite Le decir the explicit re ject ion of quantitative r i.k m ily. i' 1"r ufety occurre:

[.[?

cvaluaticn, notions of probability de play in in pwr mt mic A

ef nic M

more careful investigation sugge ts not two but t our clu ait ic l-

.ticcu m ussuran-tion; of events:

Frctecti transic:-

Cla:s 1.

Expected occurrences whoce effects or 3

consequences must be negli:;ible.

to prew

$j Class 2.

Occurrences whose frequency 1:.

the enz-y Inter-f a g i.,;..

meliate between Class 1 and Cla:: 2

,,_.4,,

that are only beginning to be defined b'y y

' wa

[

and treated separately.

enginee:

d-thit thm

%qq Unlikely severe accidents used for design bases.

Rch l i -.

Class 3.

eifecti M

Class 4.

" Incredible" eventa acainnt whose

example, consequences no protection is requirod.

any - sin.-

L-The classification is based on probability, even tho mh If there les:1y, 1

d'

,the values of the probabilities'are not well known.

i:: any doubt, an event can be placed in a more protabic classi-acciden.

fication to assure that the evaluation is on the co.servative catatlit anato,z side of reality. The classification ccheme is seen to be a "incredi crude approximation to risk analysis.

a requir e v o n +I t. '.

I 2.4 Defense in Depth T

s1..<.c<

M of defense in depth i *another way of express-lack a o

The s.oncept

.~gj ing the foregoing ideas. Reactor designs are regt Med to difric.,

s4g include a number of barriers between people and th( radioactivit/

theref :

contained in the fuel.

In water reactors these barriers are pv pf$

~ DQ the f uel-matrix itself, the fuel cladding, the primary system press',Jre boundary, and the containment. fxpected oper.iting 3*

c translents and other Class 1 events are required not to jeopardi e 2.2._

. y 12Wa :

any of the barriers. Any failures in the fuel or its cladding

,g.,i,j woul. release radioactivity to the primary system, frem which t ha,s

,3 m

• My;.,

only small leakage would occur; a radicactivity cleanup sy? tem

'CIl 5

• 13 I {F f *W would prevent any significant release of radioactivity to the r ?? '-

e nv iror.me nt, for Class 2 events, and mo: Class : events, de ter Mf' engineered safety features protect the primary system, although

cases, Ng some of these events could lead to radioactivity releases.

dOIiCM Severe postulated breaches of the primary sy: tem (the most

.a su.

severe Class 3 events) could lead to rupture of the fuel cladding J

e,j and release of radioactivity from the fuel and through the breach into the containment; emergency ccre cooling systems and 8

iP6 h

.d other engineered safety features would limit the radioloC cal m

i ane beern consequences to persons off site.

2.gg b.n w/p 46s-y.n x...

NW 'T%

y, l

u m &.g.<'p h,.$,.%.f ~ j..' &.. 3 : &...? Q ;,

l f5*EM*Y%C(Q'hjQ*^g&., Wy&* '

Y

,bQgw.

.m.

.g w..s.,4 C'

ll a ~

-~

R f

%L th, p 3-g*

g

((g,.,

....m My av,_ggg<y7 ;g,g.[]

r g

,m' % %.-v,.,w Q p Gygf M gy

. ~

_J.,

~ M MWuha ese*h[ M%

{g i

• • 9

. s n n. n,w,,

u, ts ' " - % eu.

e.

s

_ m.e

n

~y

.b l

., u p..

Q..

i.ht! ^

us r.1. r.tr s u

.E.

J,

,@r-

. r:

~

Cle: 1<.en!3. ire expect 2f te hqr

'bP u -

n-

.F%.

, ^{ {,7.,',.

.w -

cequen:o: m te nocligitle or

rly to.

i l.

5

rite..a.

dc.,D, of fost res to cope wi.th C:.as 1 e e?'

. 11 t -

M t

7 t;ned by

.cn/ s cc:rideratien;.

s c r -. a n '-

c

, p 7 g.,,

/ e f. t O. *ho 12MptaLIP C c h C e T IC W. n : al' if

• +V

>7 g,

vents %.

1 ccvero comme.r.:.:ra t e wi

• h t heli ty tha tir Icwer prcLih. iti 0-

[Wh s i e O s, events havs uch low probabi their ccmceqm a nc e,.

9%a-

, quanti, not be continered.

.w

' # ", e n.

i. 6 4

4

~he principal defence againnt accim nt in prcr -*;cn.

.e<

is11 structure:, systems, and comp:nentn i~pc Mn? te C W tv ru '.

1. g W pit,-

be deci:ned, built, and operated se that the pr tabuity

[/st h

, 0f h

oc:urrence of an accident is very ::all Tha key

• .c chte i:en*

q A

of this objcetive is an effective quality assman-e

.A.(

prorras, lassifico-discu sed at longth elsewhere.'

liowever w "llen t e v a l i t :.

c.

U

!.} Y assurance prcrrar, it nu:t be acknowledred to te cc, r ' ^ r !.

Frotectiva sr te c are installed thereforc to dN '

'h w trancients and f ailures as rnay occur despito a ll tha* i dene 5y e.

ta prevent then.

A third echelon of the defence in L ;. :.h its the enrincered safety features designed to cope with ur.litely

~y N

failuras that g f.

go beyond the capabilities of the wcI&nt prc-vontien and protective sy tems, as wel' as hie.ly unlikel.;

g; y

failur<c of ?N other defenses themselves.

.t y

The lesigns of engineered r.afety feature:

,N Uh'T are evaluated to provid' c m '.;e thu they will function properly unJer accident conditi ns.

Ign bases.

Each line of defense must

.W.

be well designe I an.1 everutol fer ef fective implementation of the defence-in-lapt h concept.

p.1 F

enmpl e, system prrformance is evaluated acuming a fiilure of For Q $1.

i, any single active component in any engineered cafety feature.

TR

h%yc.s i though In principle,

f there Ics:1y, analo o.is defense-in-depth can Le prolife: ate ! en !-

Dimini. shine returnsto the possible prolife n ti n of d e i.n basic

i. C

te classi-accidenes.

+

M from wch preliferatien licta c s:rva tive establishment of a limit analogous to the distinction between "creditic"to the required de f ense-in-dept h, a;Mi rd f*f

) bs c i

<*?

" incredible" (class 4) eventc. Thic limit, exp: 9 003 L. cithe (Class 1) TnI hM a raquirn.ent for depth of defence

";. p or an array of crH ihl..

for which protection is required, event:

~.N is one of the mor.t L

K.

difficult technical cafety iscues to resolve.

f upres:-

lack of knowledge regarding probabilitics is reopensible for the f 0 As utual. the i to difficulty.

Judgment is rendered on an inalequate bu is, and

.i C'.

d $

ijicactivity therefore is subject to change as additional knowledge is gained.

y,$

.r3 are

> system rp g.

mting 3,

gn 73}g;ycAL 3Arg7y 3

• e to jeopardire 133t;g 7 g }g r 3,g gg3:37 g Q q., +.y,

1 adding a which Listed in this section are examples of individual ar-icna

p. w.n r

that have been taken on technical safety issues.

'p system Many of these MM actions were 'taken to achieve conformity or cc pliance with to the existing criteria an1 guides of the AEC; many were the re: ult MpD

nts, decisiens made on matters that arose in the centext of lDUl

. although of ind iv it n ]

[I The:e examples are provided to describe cases.

ocs.

h the un t-of dh actions that bases fer tau. ave been taken, not to coment on er anal /:e the

.w

~- **

el cladding ng such actions, Wrk_

. the t

e

.?:y.p.

Mt:ms and bp.

. logical

• ^(c. for ("+c "o T. r, k..

Role of tie rued ', tate Atom m rg n.mm-m in c.wem M

t

,ng gcc,,m g,M gr,, Nn,,c,c,g,, gugg,,y gg (ggy, pg ;),

• A tQ.e

!Q.3pMf9

• i w' b ~

hw36)Nbb i

f w

, G H,...S 3 Q, P 7 W p K-Q p; g w g.5g& r y!:, K W k [ -

pp~

n r

n yr

g..

y - ?.fg**

• Nb / 4. g

,.+. ~. W W 3,-C a,,

s 4* f

id....^-Te

,G.: G.c ps.

4 A..

a4. ; -

M N,. s[-

.s

w. W u,..:

-n

, ?e 1

n.

q %. i.. b.?*M d._./*

f a #.

,q YD-1 i.'

• • N 4

-* % J.s v

.ew f.

b i

.'y, m'b.). Jgse :.c* cQ>pM.n3*?'MW W*

~k).,

Ihg* ?, N.r. As A

~ p@:; % 2, q ".3-v r.

)'

- 1.'1, +

-.~

q 2

..r 3.

~

d

,e :.. e. g: & q.r. y. 4,, g< c,..p..;.a -

M

.. m n

s wg s.g%.,

a '

-: 3.'

~

+

4, y ~.v.9 o..

y-f M & M W.4@fWWM. _ @' -- m*-

~ "_.--.-mum 1

. [b,

~;'

~^ ^

a.ac~en cWN.Mm:$%W; t!&7mit?t:9'ryP!:tWfgi,W

,4..,3 - FADW3lQT/.~:J,7%?Mg{C?'g);iYegnff)g@f,fgz%qy;&x2M&W% d{QQLZyy.c'g. qdfE p2Z n

-u _

g; A;r.;; 1, '

,q ;

G. ~ c,. 4,&, a >

m.

gy s.

34,j t

- h. : :r m r#sur- +

pm 208 in w rn u 4 "mc

%?t;%

A-3.1 Site and Envirentental Matters we 4.,,.

kh.h Design specifications for facility s*r r ture: a r.d equipment to acccm.Mata hicher carthquake acceleriticn valuco than cri-1:Oc g$/f&

M7 ginally proposed have been required in many inc t ance:.

p q-w Design mefifications to provide hicher f2ced protection

%@1

gEQ, levels (including hurricane, stcrm surge, seiche and :suna 1 n.n*

pratectica), than originally proposed have teen required in many rev..

,@1 e,74, VT

.M M

n m nee,,.

MW s, l"~.

Frotection of vital structure: an$ equipment actin:t ter-

.M ;. ',9[

nadoes is required. The design bacon for spent fuel pec'.s have

'fYUS 4QF been mcdified in a number of ca:es to protect Main t Ice: Of

.bE* '

water, from pool damage, and releace of fi Oion';rc htis from g;N*"*J 8

fuel damage as a result of missiles prcrelled by torn 2 die winds.

- -Yt dur:

h.g l

P dnT Ter one plant site located near an airport, design modi-Mhw fication: were required to protect against impact and fire h4W}

effects of aircraf t crashes.

2-e.e d.,?.f 4

red m

u-'T 0 2

$Tgawp Vent stacks have been required to be located away frem b2 containment buildings to avoid damate frcm a fallint Stack N D M,L

.I where the stacks were not designed to withstand tornadic winds.

M.te.,n-N sta "q.s M%

wp~#@8 Ii stallation of strong motion ceismcgraph: is now a 92i!

N requirement for all plants.

j' Extensive soil and rock drillinc pre; rams have in so :e V055-k, cases shown the need for remedial action for colution cavities, f0'l u17*

for relocation or redesi n of structures, er f or replarnment and C

d compaction of soil to provid'e suitable foundatien conditicns.

lear k

Desi n changes, such as relocatien of eqJipent, have been requ C

required to accc=modate the effec *3 of the failure of upstream

dams, be-3 r.

5 Instrumentation on towers has been required onsite to Pro' 2 '4 obtain local meteorological data to support adequately the era-M.,

red. !

assumptions made f or atmospheric dispef cion of radioactivity.

w t j> A mdj Environmental monitoring programs have been required to be ELE M bre men augmented.

N P A u' fai1 h q;b4e yMg 3.2 Reactor Design NY$d$

GV-4 Part-length control rods and reviced instrumentation were f0T MfMy provided to assure the capability for observing and controlling quir-M"$1 the potential effects of xenon oscillations and other power 3.4

.%.. SW* %.?

distribution anomalies.

H,$ 'h" ~$h Tixed reactivity shims were provided in the reacter core to reduce the initial value of the moierator coefficient of Wa5

~"

vactivity from a positive value to =cro, so as to reduce the esctivity that would be inserted during a loss-of-coolant

$4 d dent with the positive coefficient.

wer<-

y 3

a cs The capability to include fixed, incere flux detect:rs alv has been required where this capability did not exist originally.

3Y37 v..-

h, '. tJ D.. %D,Q

~ - h Q

h4 W4le' i

6] m +i!'

u u'e d

h,, d (%,V w U V 23 ' - ~

%}ti n. c.w. i i

ei y

y.e Q[

][,%

9

,m

[:...s.~a n. nr 1 f

.h,bMgV r..

y } W 6-% WyWiy,*?&. _. -

.d lgjQ M QHy h *

,,pg;[i?.y ?p$Q?Q*pMy*Gi.h;4&t*fFrQ@7%j,a'.V6.l2 + Qg JQ W.

..x,7 >.h w

V 5;*J;^ W q,

. A..w

g. f.;. T ;..w@. _

G 7

,,% : ; 3 q

., w.aw, f.o..mg.:W'm.ipw y w. 5 5 W.., _m. < :..,:. %.g. g% tc a

M,,

~

y

..... sw4_

c e,sm.y4,.u 4 6 q

.. v.s s. y,

.x pL 6 m %u-ppr @p.y,P.arn.,gyygTM M M,g @+/,gMtt gxc L u g g h *g#wugf *q

'p

- q-_

W"" t'_. Myfqq*Wg

/

s w wy, sw.~m

.: a,

" - - g Q q m.. L'." 5. Q Q Q - g-;ff];5, 5'

,QASCh

~

W '-QWWM':&:l,lWlMWQ [i:D;% f$gik(cpq,,..:s.

T J:;g _

y

~ m 4

.V...c,::.

p s.,.

. pv,e s

y

.f' : _*jl vs sa

.n w

WI

• n.,

.oh.=

';ibrition tect prccr ams c rs on ' ly 1

a' het

• r. tin.7 at tb ce u

,'T-

, l.t s i r.i t i a ; c'._r equipment 10coptibl p-r ier nanc ' of rea r; t or i n '. e r '.1 1

.m

• +--

.an ort-loo:e part:.

.;s'Q

'W t

't5 3.1

.:m.:-

Ce o l.i n t System f ecire an :

• ection y,

unami M d i t ie.n a l pipe hangers *nre inct ille !.i: a reca.

' u

• f ed in many-reviced dynamic analycic cf system pipin:.

N,.

"hyl,x ccicn of main steam lines wu require ' to inclou

[Q.hg'..

i inst tor-simplified dynamic analysis and improved quality contrel

,n cols have' 11*1d WSAd3*

'F % J

',oSs of

t %

t ts from

, tress analysi: and inspection of primwy ecolant ;;ng

)p a=

WW dic winds.

flyyheels durinc !abrication and the capability fer in pc ticn

) %[?.,

r*

durini; the service life of the flywheels is new requir e ' f or al' e'n modi-plants.

j.,9 m,.y fira

.m Flow restrictions hava been required in miin nea* lir.es to O *ft reduce the reactivity transient followine, c pos t ula t ed :t elu lino Y

,1 l trom gr g' '

n.a seu

gin

.e

. Material samples for monitorint the int errity of senci+. i ed r..: /$

ic winds.

stainle:c etcel within the primary coolant ::y:t.-n have been re-MPM

W e quired.

• %fe'

'.qp$ &

ir The decir.n bases for reactor int ernala insid' t.'.e tea:ter k*p*'

d f n som, vessel were molified to include an allcwance fcr Licwlown ferus

'3 lf,f[@3 3

cevities, following a lo: -of-coolant accident.

cement and iG

. The censitivity and reliability of dctection of :rlint y

g

'itions.

leaking from -hc primary system and frcs emertency syste:

f.,f[,n?j; ware have been required to be improved.

7 upstream i

Y Additional pipe restraints and inservice in pecticn have D

been requir ed in the vicinity of steam gc.neraterc to reduct the (b

6 te to pt obabili ty of a primary sys tem pipe tre G: caucing a ste r. ren-(*

the crator failure.

s.

%p g.

tivity.

=j

,, Additional restrainto have been requir d on prir:.ary system S

t. *M'

.irzd to be pipin;; to prevent the possibility of pipe whip, followine a

^ y" $;,.

break, rupturing the containment boundary or cau,ing eth."

pipint failures.

Replacement, or other measure: to minimize the retent ial N.

'i' tion were f r cracking of furna:e-sensitized stainless steel have been re-

M ntrolling quired in several plants.

%, *&d power g9 g# gm 3.4 Containment and Structurns 7

tcr core Diagonal reinferecmont fer the primar: contain="nt buil!ing h$

. p? g8g nt of was required, to accommodate seismic shcar forces.

'uct the

' A.

!t ilint Aluminum components used in:ide the containment buildinr, W N2 were replaced to the extent possible with cteel to minini::e the A.

amount of hydrogen that mi ht be generated fro.m Octrocion of the T(

f 6

• g - f ecters aluminum by the chemical solution of the containment cpray

'y,q '&

originally.

system.

pg, A,

p A'g t,.,s-fit -

,a m POOR 0 & R,_,, P_

M!

x +.... - _

.Myf; 4/

Q N 5.yJ N -i g M P ! /~ '-

t g1",S h 9'IM 4 O Q 'Q pl$, M._

- y m, m. ::,ww%.4u m s w [ ( &g }w %s a4 y

u.m>* v +s t,m:= e.. r..~

~

~ ~

~

M=x

.w d.,.ip d.-. m.3u,I e

.-d"*

~l'~.

/c*,o p#.f:t% s,-- sFt.

~Q 's2'.V F 4 *

• M e

W

je%':h~r+G'%

%"Aa- 'i "W' h -

=

'.' ? ? 'R ' '

O' !' " r "'* :

dr.' '..

4 QL' Y. W,J * '^' M S i:. Y' A ~]~ Q:f' 4 "* & y?,". l;'li?'.*

y; Q Q. f;ft$". & G y'<~? T &. c }y. Q,2h"'; d. h "5:

• f wp 8

Q' *1'k

~

2

~

.._a A

Y, g

I 7_

i....., _

7...-. 4 3,.je G 3. s gg,,ypys 3{p'. i:%pfgf.

w

...x-

f. w,w t

. v. m.a g W t.*r.,

• e < /.3 2 oeJ..c@ 4,v. % c4 M.,...,vy-)("d.c M,5

..s..

m..

..n

."5p t'e. yY M,.4 %s.. y e r...

.m.unam.h,n, r,e < ;.M A.,

t

.>~

v 't t

.i. w.:m.~~m-t**4 w

+-

"~'r-.

v.-

-w

- hw u rh u.

se %;-s' f,w*' c.C 1

f. *$r m.1 4

,p ce-,J wm

.t-t

,,a 't 44. q~'

210

w. o n u m m 1, +,,

• /-d aprevej containment builling pe*ie: liisn deci'n 5 cils, p-in:;u Jing h ar testing ability, were :W : ^!

f er a ' witr cf of en.

i

1. 1We plants.

P#i Vf.*31. e s grpgr g AJit tional protection of fuel stcrm p ^, I s M L^ine PMuir5, have 8

a, M

t, protect ar.ainst the consequences of dr ;;pinr a snippiu cask of al!

1Lg into tr.c pccl.

y "q..

fr _ F Contrcl room shielding was required to be in:reased in simulat NM thickness to allow normal occupancy folicwing a postulated 1035-

$fp@{

"g of-coolant accident.

f

'(

ventila gt".fg A fresh water storage tank was relocated since failure at fccts ::

@e ;l It.s original location could have caused th" failure of two v

'pw"g diesel generators.

m s.3;ys-i autsmn-E W

~

The intake channel and intake structure fcr the condenser pump ar

<.coling water were modified to withstand the desir.n b-icis earth-a rod-m

. g@An~$

quake and to provide emergency cooling at minimum water level.

s

'wy.;r+

onsite E ' '

A containment building design was mo lif led sc thn c:n-

% K;G[7J tainment prestressing cable anchors would be easily accessible Iin"C#'

for inspection during opcration.

$4 t

'h'*-*FiW All pertions of the post-accident decay heat,removil

.',]c. t ^

ar t rb fM i

systems are teing required to be designed to seismic Class 1 M

specifications.

.,g '-

te the V

The gaseous waste storage tanks are being requir.'d tc be prcduct designed to seismic Class 1 specifications.

in the

+

'/~ 7 The response spectrum for' seismic design was required te be revised to provide a more conservative design.

automat 4

the design of fuel storage building was required to be 5

l revised to include provisions for controlling leaki. e of gaseous standby activity and for charcoal filtration of gaseous effluent.

r.

Trovisions are now required to deal, without the necessity ment sy ps of venting, with potential combustible gas mixtures generated,

?,, ;

following a loss-of-coolant accident, by metal-water reactions

. b'd.y radiolysis of emergency core coolant, and corrosien strain:

in the core,

(

in the containment.

ond ccr'

,, w u,

of plan-9g.99; 3.5 Encineered Safety Teatures

.d@M

. %5&M The designs of emergency core cooling systems and vital vity Ic WIE heat removal systems have been required to be revised to increase sg] k reliability and margin and to meet active and passive failure

,s criteria.

p k*.t?h t a.wjp -

Emergency Ccre Cooling System pump motors ware requir N, to

..(JM be protected against the spray of the centainment coolinr, system.

Chemical sprays or filters (charcoal and particulate) 3.6 are L

being required to reduce iodine i*ventory aftar a loss-of-coolart gg accident to meet reduction factors required by conservative c

assumptions on fission product releanas and meteorology.

instrur vh. m

%*ers fN k

900R W[

QC e

~n

-- ~.. -

~.

y, t.

((

c

.s.: '- nn s

wu. m.s:n Am. w rm:.ww

o. e.

pM g.+..a &w+.r n" ~,-~w.:. m .=. a w.=;.": u m..

U' W 'D

'"Ep'*et'. v,-

% W%%..

a?. 45pi 907*b1 *7).MP.%

IN 8

MNENM hh# * *YM

'**I*

MA**I

<*D,"

IQ192T

• [-

b T-f*

W

' ' ~

u Mw%wsM.&p.32mmwc er u -He v

wmu 1

?

m

Si!.V 'u'. :.% j W:r:A p,q',;3*:.ybque1M%%M% eW.%2.:.N w9 Sm :.}c :";W1.%Cmqq%

pg:."!'

-er y;%

a v.1. m f & ~

• T re v <-

~e' 7a.

6,;.%

,W.

ww

f. C II

, m..

A tt (i\\l b/T i

g,

,c; O ML,,

nb4-

• tail '

e.

la n t itecol, ener atern.w e<

-.n' y-

'I of cf ecc: te schic a Fquate reliarilit-cf e'

• ^ rm -

~r 2.

',f L

.cc : ~. : housinc "ritical emerrency e nto ~,: 1. * [ c p-i.d4[

i.

c

, 4 requir+1 have teen required to be scaled or ce;arated to ;revent i ic t.a nc nq cask of all epi; ant as a result of a sin;;1e f ailure

7.. ;~

Tnats of vit11 equi;-ment have teen r g in 1 un!cr een'itic.n

".ki -

1 in simulating those attending an accident.

I M-- e 9 od loss-

; ]y.

$pC.

Tilters were required to be added to tho luxiliary bui. din:;

ventilation systems for a number of plante to titira te the ef-

[',

,.ur2 At fcets c: 2 fuel hindling accident.

%o

- / y%

a e

ny t i! D.I' h'

T% 4 irr. van mo lified in a nuiber of enen tc in-lu le automatic ico15 tion of the main condenser mechanical vacqua Thnser Pump an'! Cland Geal exhaustcr: to reduce pctential docos f r em W,9

Carth.

a rod-drop accident.

lov31.

The amount of fuel for emergency diocel generator storm!

'4' h~

onsite is r equired to be cufficient to allow orcration er en-p Q.D',

con-asible fineered cafety feature: for one week.

,{f.

lMNb The centainment air cooler design wa: requirel to !.c

.eti-

.4 f *= H fic! to allow detection and isolation of failure of the service

.I

~- T

, gd@-

water coolant lines.

s 1 Fadiation detectors were required to be pl20o4 sdjacant

if x,

to the opent fuel pool to provide quicker detection cf ficcion 1,

{.9 % g, to,.):

pro %ct release and resultant isolation of the reactor buildin;-

M-in the event of leakane.

' [9.-

red to be The control room ventilation cyntem wa: re w i rmi to doce automatically on detection of radioactivity in inlet air.

l[:j g.j be Electrical heat tracing wa: repired in i. provided on the r

caseou:

v..a nd by liquid poison control injection linn b ron colution).

d.

!z Physical separation and redundancy of standb'/ ras treat-

~

/

^cessity

.o n t systems were required.

...G.. [

- /.;

mted,

' tion:

Poro conservative limits were required orI th.- stre ces,

'I

~

rrosion.

Otrains and deformations permitted in Class 1 cy tems, structures 3 N' and components under seismic and accident conditions in a number

%:.h of plants.

WA tv %. ~

g ssdditional equipment has been required to reduce tho acti-3ptd eital vity lovel in of f-gas releases.

pg increase 4

heme ilurt The design leakage rate for several containment buildings

M y*Mg.

"y was required to be reduced.

.. W AC-f.hN{N Installation of independent overspeed pr:te :icn for the

. ired to turbine cencrater has been required for several plants.

Ug system.

%dh

.g ge 3.6 In:trumentation and Power Svrtom

,f g.

. \\ *:h.W'

'-coolant iva Changes have been req (tired in a number of planta to achieve

'Nd Ocparation of cables servicing redundant control and protection j QK instrumentation.

M..

P O 0 7s @v. M % W-] U.3, A

g s.-

b

.c m

g%p.

s, h.,f.L-fff t,

w w.

l y.

.',I

,,a

...,, g....o,.w.',',w;.m.

. _ R * ***.% NW! g&y.4s*s p.$b. go. '

.c.l,m,nq.ag;- yS. d, %"M lW Ng m m:m,t., m, w

qw.. a w.. y

a..o.. x.

%jO

)

' ~

~

sw ng;v,A.. ;Q

,1 V.,

.. z.....

w.

~

.w-7 ;+,

. %p* 7 ~M.g q g,p,..v.... -,., -lm., h,;g, l.ck.*#g y 4,,4.g4,g g g4 g.j e e-W)gM

,,.e 9, g,i...

I

.h w

t

,,. gy.n y s

s 6

71. y 4 ', nw..*. 3 >

.a '.p %.f eyym 9.e.-

& pewt e

g,,.-

z.ww. :a,; m.

u t.Q,

Qy e,y p,

m.,, j

,q,,g.. ; y. y._4.c....

i

,M %4.;.r.> g..jyp.vs p n

.v y s.

m.. K

. g lpp,h W y y*%>

p+

@n n

~

- w 1

.S.

M

%~.,,-.%-.-

w

-*A*b. hi I WE.

f '. ' Y M ' l ' "

a

^

t

\\

-1

-,w-.

r u.,..% - - y.

41,./w.n Mt, MAN: sort #OM./@$pp&a.

,',h_.,# # &y$ly.g'. *s,6Nz.$$N%i$p:M(4 uaf%@'!*'!]$n.m.-

.'?.tWM...N.M. ~,PM.;f9jJp 2MB;pUl%%%,M,,.,ty..... 4rg ve P W A..&.% s.,.4s.y m f@ m WS

,7 g, _

TNh Q*hyE?N p; ;m Ji TC.&{.Q4. 4%.i'.U$. &Nj&.

c (s*.

__sQ$,cs; M h p rWiiliOy

. % pt,.kr f

J,., w }.

.u w 2.;.o ja <. m.

.p

9j 212 uWMF JW Wr*

t '

4..

i

_1 Charca have been required to imp: c ee.* " testability cf p,m s; vital instrumentation.

that fa:

a t; t si D.A7~.D prevent t

Y..

Cwitching of equipment loads betwocn e. era n:y die:al

> </ ' '.

generatcrs was changed from automatic to manual.

• 2""' -~;yd with re

$% M)

Incorporation,of the split bus arrangemant for emergencf fer car.

[Q.r.,g.j

't power supply has been required fer a numter of plants, resultinc j

4

,M,

in the addition of another diesel generator and redesy;n of distribution p.

p d,p, -l..

sy stem s, the con-

-I

/p^$

Circuitry for a flow-bia se d, flux scram was required to te cs

+

7 ir to :tec t protection system starlarJ. for a number for a r..

c.,.stallel and pg~g reactor plants.

the num.

I

,M y..

, The circuitry for the rod block monitor sy: tem and auto-u, c:

relief system were required to meet protection system standards p9 v 4cen for several reactor plants.

e i

. m.Ag

~ h.$.0 The control rod position readou+. system was required to

$3

.,t.w!M j be m.od.ified to give indication when reds are of f their demanded y.)

c...

.~

Po s i,s i o n.

i.

are use.

ig/l

• ~%

Instrumentation is now required to meet seismic design 4.1 Th I,

standards in all plants.

~

Ar f

A redundant station battery was required.

consider water i The ability to perform a hot shutdewn and the potential at 300CJ capability to perform a cold shutdown from outside the control 25cc. 7y

- si roem is currently required.

calcula.

will n;.

.g Diverse signals are required for actuation of the emergency two way:

,' yp core cooling systems.

after,m; l

mater;a.

v m

'. Y Means for direct measurement of primary system c oolant Therefcr bh.h g,

flow rate were required.

conserv:

ss-hW$

An alternate circuit to provide oYf site pcwer to the T}

R emergency buses was required to cope with postulated failure of ultra-c.

h@."M the startup +ransformer of several plants.

of thici

!+

kncwn.

Ip; r

Actuation instruments for enginetred safety features are lon;; ir:

fjpjg required to meet protection system standards.

severel.

! g y/r M analytic

! 4j Con'.rol systems for emergency power were required to be reduce -

i j. %..,

redesigned to meet protection system standards, elapse.

p

'yg'% -

t t

w j: /:n la V-s g.,rge.

% 2.W whet-

'fi.

G 3,l Misec11aneous WydQ informa-h,I and rer Improved development of Quality Assurance Precedures and in ala-u

{

organization chances to implement these quality assurance pro-indicab M

S.t ; 4..%

grams have been required in many cases.

en Design changes have been required in radw2ste systems t for s-gg reduce normal operational releases.

resclve 4:.L W hb WQ,L. M il POORD M f ',,bf

's,

t... L6, Q

[* ** ' #

?

7

.aw a -p -.,.. y.,. %,*., t,. --W?~

. A v.-

'M I

3 i 5'.

y. w w a :+. m..- s.w.

.,_r.3rc.5 h u.j f.'."-g w n 7,$%di f f.i

N %nwm'Jy..*-

gns.wsw 5Yf..

+ <; 5. Noiaf -IA-2%"N enwa ;s

.+. w..

_ al \\ U L's : _. l* *a.. Q * ". '.., : h M

.f M' 'M5.

A*iR (f{G_ jQ,,

y g g i' T:y C_t'h*NT'P;Y.,*?.-f.:f.h;).;..=mcg g g.

.Q*y-

.=

.,t m, y~.. m. 7..,.n :yg.,

..... q..,. qww e..

q,,0.y v._p,(

g.

.. s...,.,. m.. m. -....y., y,.qv -

m.

r

.. l.n 4 n.-

m _

s

,m,,

s e-_.

1

,c

, s_

fo p.m

W r

-].p" ;

A, /s ( P,), lW I NI?

23

,74 3

r e.

M ility et g,.:ip of fire protecticn :yaters ha: U

+

th at failure cf Sei:-ic Cla:: 5 p:rti:n Of tb v te : cl'

.st

't r

seismic 01ss: 1 equ g -.t

.w r e v.. r. t the functioninc of

le33 4,. m Interlockc ani procedaral restrictions hwe bran repirec Q'. n.p q,

y wi' h recpe ; to the crane ahich haicts to minimize th' pctential

' WC C U'"/

fcr cash drcp which could damage fuel or the fuel po:1.

a-rzeult:nc

'.dh.

1. M "UO" Fest-cperaticnal integrated leak rate and s t re n;;t h test: Of the containment building arc required.

,N.

dwt air d to be Chan:;ec have been required in the operatinc orrJnizatic'

'3%

, $g[(. ~

4 number f or a number of plant te include an additional supervisor and

, Qy the :.umMr of licenced operaters.

And cuto.

4 FCP.E TECH::ICAL Carl'"( ICCUES TliAT HAVT.f?T SUC: P E C L70

^ M*f ^

i,$ b$%..s s :

stand rd, h

Conc examples of technical issues that are not resolved

.;- p.

b h-are discucsed in this section. Adequate assurance of cafety M8hb drW to do<*s not require complete resolution of these r:at ter: (se: Section 1.) prcvided approximate conservative ancumpticn: and procedures OQM dImanded are used in areas where knowledge is li.nited.

' q.g.f..-

dIsign ie.1 Thertal Shock from Emercency Core CoolinE

. h 7

As an example of an " unresolved" technical cafety issue,

-Wi@

f.1 -

consider the consequences of injection of emercency core cooline

[ h'Q water into the reactor vessel. The reactor vessel, initially d

stantial at 3000C, is sprayed or flooded with cooling water initially at pgi control 250C. The resulting thermal stress in the vescel wall can te n,

p.

calculated; it is not difficult to chow that a perfect vessel J.

3 will not be damaged.

But real vessels are imperfect in at 1 cast 7

ae cmargency two wayc:

(a) flaws smaller than scme acceptance limit remain

.D

//

after manufacture and inspection; and (b) tha propertins of the L NY c

th material chant,e during life because of the neutron irradistien.

coolt.nt Therefore, the stress calculatien must be performed with seme K,* rig conservatively hypothesized flaw present in an entrittled steel.

w 1 the The only way such calculations can be made presently is

E P' M failure of ultra-conservatively, since the fracture toughness pecperties T

kh of thick (15-2 5 cm) sections of irradiated steel are not well i

known.

These over-conservative predictions chew that following i

tures are long irradiatien such an event might cause the vessel te crack y,gij ecverely. A large-scal.e program is underway te develop both the q g, ;

analytical tools and the experimental information needed to

, pp N@g n ed to be reduce the conservatism now required, but several years must C

elapse before this information will be available.

Uf 4p'+

Today's vessels are not highly irradiated, because today's large reactors are young.

So, one can put off the determination

'.Wg of whether something must be done about thermal shcek until the

-In information is available. If the present conjecture is disproved, Mr and remedial measures beccme needed, the vessels can be annealed

' !M.

Q in place as their neutron doses reach the point where this is

4 3 *'x*

inc2 Pro-indicated.

..c.,"

'P,f.m,'

'p l.g Thus this problem has -- and needs-- no definitive solution intims t for some time to come, and an appropriate prcr. ram is underway to

@N.'

rc clve the iscue when needed.

Wp y

t q.

nm. m&p:nm.%m. %y., s%.f.tcg.y*py;.-e..%.~ ;.Q ~ y Q:s'~yy 9.,

.g a

.9 g.m'..,% n.

e g ss U

.,g pe

.,. c,.

...a,w.

~ ' '

v;

~,.

n-v se,.$.rrm. s '

Mg., M.,g..~

4 9,

.s_

b n.

~

1.

s y

n

.~.A $4 *Ng s.-. m m. m mw.n. m g. m.f

.%h

. hh.reh nh

• he f5C

,.5=l.h,6

,5 a.l.) l{

.,,z } s.,L-

!N hY i.k

+,... '...,,

a

,V m~..

w.- e

.e,m gew :.y;s,, -n;.,m.w.-p

..y w

n.+.w.~,. ~

e m;

..~.p.;

> a +. +.

y

,;.~ e

.a 4

s.m y @ yy.--;y;g;2,.y w f g..). g,pcm p4,p y, c.% ym m m.< >m -9 m.g.mp.my w

1 a

3 w.

w.._, (,;.r w. ::t. :. ~.r,y. p.

y -..~-- s y.,x. a;m,.p 4A g,._y.r,, x. ;..,.p ;. q. p+.m.t

• v, m y.. -: n

.e- ~yt, m, wL y

+c v...

a mw y -

u p,,,, u,

yw y. c s.

.t 4

g)3

.m.

,y bPr a

.n" m I

79 tm: a nr.na n t :

h.

d

,y

[# -

in.r M radine La its t c.

p.

inc t ' -r t. ;hnic11 i:Lue, ti,ia t ill n

.H D: "finalP/"

r+-

fi{.}.

b.*

i eir.y ga;;

wiv d, rollte: to th" pon r ratinc c: a

• s.2 7 '

< ^ +

4.

Vip"N'.,

hi r'eac i t t : c:rt tated very cons wvatively.

M' en 'ract ft - i

'w

..:n r. tim _a at t he se levels, Mv" 3pN at pt rir ms w _ u n t 2ker W~.)

'o chew, theoretic.tlly and experimentally, tha hi.- ra t ite cig wer o Ptstifie1. In many ca:es, the increatei pet i at.J pcb?r c. ;,,,

M.M 3Fpr:vej r*itt:re, e

p.u'ed with there of previoS:l a f e, g[i,dC

1. r.' ' t eun j u:t if ied on the basic that t he sifet/ sa r.; uy a, p ro -

ef i7

~,

densit/, at f

p.,CQ;

/.'

vi uely evaluatc l have been fcur.d 10 bc 1.ir r e r tMn or trinlily,

decir-i s 7,;

M g.

12 ; iev. r'. This reevaluation is usually t ased,on measureuntc in

~hE*. C th? newer operating reactors, or mor-soph; ticats! calculation b, tien la N 'a ir 77 1.

t.:hniT;e r,, er the data from the devc ic p

n' p c'/

v. --

aces on all cf these thin *:. In rany instin:es, the r.W inf:r-of"pp, "V

2t ion ar.4 techniccies do indeed show that the wproved perfcr m co lelve: nor-ine not significantly different from those formerly

,.. c thaurht to' exist.

However, it in also undeniably true that th*

s > t a= -

)drMR a;tual caf ety margin will have been dininished by the urratin..

anti.

M are s.)

$111 If the puwer increase is juctified by leaving t h-p"at to t'

"~f.#

p a ner +:n :i ty unchanged, but runnint the rest of the core at, gin h

hh:hc r power density (improved f lat tenind, then t he i s ::'e 1:

g '.'

3 p m'-

t a ther dif ferent. The hotte:t part of the core is m enstraly whi -

y'..)

• d r,o hotter than before. The rest of the cerc, thuuch hotter than c n '.

Sh before, is still cooler -- has a,greiter safety margin -

,'".3n 31;..

W L

the hot spot.

But any accident involving overheating wil. in-u i Q*

l volve a latter portion of the new core than the old enc, tocause a greater f r action of the new core runs hotter, even theu.;h the hot epot decs not.

varic.

2 i

queno

) rl MJ What i:; an appropriate limit for this procedure?

L. en ~'y dictate: that the core heat generation should be c's nearly uni-

,44 term as practical, so long as the result i:: cafe.

Ce far, pcwer u,4 uistritutions are not so flat as to give sericu; cor.:ctn,. :t a

, F limit of flattening has not been determined. Further study i:

M needed on the flattening question so that development of hi.her is tp M [,

rated, more economical cores will not be inhibited by lick of hurri t

i knowledge of this safety i= sue.

Clea:

' gg diff;

,,M NM M '.

u.3 common-Mode Tailures in Frotective ';y s t em occa Q

verv M

Protective systems and engineered cafety feature: are com-this

1. M' posed of redunJant comp:nents, arrangel so that r.o single failure l'

rec ~

DM f will result in system failure.

Yet, when the syster failures theY QF.!

that de occur are studied, it is discovered that tho majority of seisi

>c4 If such events involvc multiple failures, for which the pro!>ibility seve:

' f *iB M

f in calculated to be very low.

The actual failure are found to cern.

N -

involve not the concurrent independent compenent failures eva,_

usually con:1dered in probability calculatiens, but causally re-revir f[,'

lated f ailures from design error, exposure to hcstile environment, knem er human error. These common-mode, or cyrtcmatic, failures are

?*F

]W h '

distinguished from the random failurcs usually censidered.

3 Pedundancy offers little defence arainct cem-on-mode fail-hNd The only technique presently Lnow5 to rejuce their prcba-ures.

(* *;t.)

bility of occurrence is diversity; i.e.,

the ability to perf orm late.

sa:

the function in different ways.

For the reactor protect 1:n syster.

ecol; Y

diversity can be applied by monitoring dif f erent precerc variables inst =

em

' Q,:&.n.-

w. +.+.

2 B_

POOR ORIGER 9

..fg y

" '.. -.. +

hk)< ((

3 f

.(A.% /),

h, J. d.:

I-w...,.,.,'.w.e.,

y I

d

'1r 7,

m : y :n n@ m & g,%p.,] b.~ r:, n u,,y,d,,tyg#g, g...w

, e n;.n,.e.

2.n.ww,

W4p :.,...,. n. 9.

w.

Qp G W /f g Myggg

. pty 4

jpn4-fgg.g. M.., gp".g4Q.:y:,,

.y r

v-3 m m. m.w:n gg.m y m..g y.g= _.5.;ysy.m, g..y n

m. - i

.~s e _.,p

,.<*n y gyf;{M g @ K G Q K fg,.gy]QLf]G, l(;QQ4Q a:-

.',','6 Q W W ~ R st. %. T. M_ ;j

• h P.t.. W h _4., W q l f Tr.?, M C 2.y M :. @.;

N

.f.

a

't 5;P T

215 prev syr i

.- ( -

% Mta.t a c re -

.e nd i ' ic n (ri nal divn ai* D.

1.

i r.-

f&rmt U

• yp : cf e p i;t e to perferc the s a m.' 'un::: = j c..;-

- 6 r-f a a l,1,y., t e -

.ity), M Ly e ployiv different mom of ccr retien n*ic:

d ience hw[*7 (a c t t.Jter dive r';i!y).

/,. '

[@-

undertahn a

r ratints m stMy of th se proble.s is in its a f m:y.

A few in-pr cicents hwa ben analyzed; a few prctection rynterc Mve been p-nd pe; ar tu jie i systematically to identif y potential cr.:.on-Me failure:;

k.

I NCIC'PS '

c few e d ples :: ci,:nal diversity have been desigr.ei The a.c u t

'#5 P*

.riginally cf improvement in rellibility attainable by the use of diversity i f{A is net kncwn, since no systems are yet known tc hwa teen urements in de ;irned in the 'Jnited States with the full deliberate applica-Ub;

alcu la t is na l tien if the three typen of diversity. Morcover, the areunt of

. K;b

--in,sp."

ir;revement rep ired, if any, is not kn:wn because tb reliability

- lf '( e:'k y

"*""*0" of present systems is nct known.

ed perfcrmance I P**PlV s

The controlling factor in requirements for protection-Ob is tha t, t he system reliability is the need for protecticn against certain Q.'

uprattnc.

anticipated crerational occurrences. Examples of such occurrences Pyg are turbine trip for the boiline-water reacter and loss of pcwcr

y' b

• he p;ak to the main circulatinr. pumps for the pressurised water reactor.

eore at.

Sinto such events are expected to occur with a frequency of

.m/

i S S" appecximately ence every few years, any events in this class for

. ML monstratly which failure of protective action would have unacceptable

T

. I

. hotter than consequences will require a very high protection-system relia-

n -- thin will in-bility in order to keep acceptably low the probability of an l

iO.cb;c ur.;rotected accident.

,ne, becauro W.

though the

.[5h Study is continuir.C on both the probabilities cf the various f ailures and occurrences and on their potential conse-

,', g,,n quences.

.e JU

.?

Econsmy

'm y warly uni-Ji M.:

i far, power t.

4 Probability of Environmental Events

.27h:

ern, but a

$.p study is One of the design criteria for safety of a power reactor M 6 of hir,her is that it be able to withstand the most severe carthquake, Wt.T; lack of hurricane, fi ad or tornado that reasonably can be postulated.

.3h.

Clearly, each site must be evaluated in these regards. The

O T

I@ k-difficulty in establishing the criteria for a specific site occurs because the recurrence interval of such severe events is

.0 very large -- thousands of years or longer. In the United States,

~!

@D ys are Oc

• this is longer than any historical record, and the nenhistorical

h

""-le f a i b record is often hard to read.

The criteria fer a given site

M.T'" '

failures therefore must be based on the best judgment of experts in najority of seismology, meteorology, and so forth, based en frequency-i$M.

probability severity theories necessarily unchecked in the regien of con-

"QM

e found to cern.

The result is that occasionally a site must be re-evaluated in the light of new information and the criteria bg)L

.ures

'ausally re-revised (--

always upward, it seems.

In the present stato of d'

' envirenment,

knowledge, this appears to be unavoidable).

"Y.w.

eilures are

'cred, v.r r

4.5 Terf er mcc of Emorronev Core Ceoline-Systems

'h>

rmoda fail

'b heir preba-The loss-of-coolant accident (LOCA) i:: a Class 3 postu-to parform lated improbatic accident (see Section 2.3); the emergency core g, '1 cetion system, cooling system (CCCS) is one of the engineered safety features N

ess variables installed to mitigate its consequences.

Qj-7 i.> :

M,.

POORGEG M i_

I

,, 'e'

' [*. Qf!&

f.

{f-t il'

• A4.M,y.f.-

g < -

[h.

• N*

. ' ".E h

?

• 4O %z b, NO. '*

f e.

~~.f'4.

ll,

--p

^'

.., U '&T[#' "

^ Y.e

, j. '.

W2 C.... M{ ton.' & a n;%.&e

~

.n.

M L.

m e,c w -l.

- - ~ e....

• +**^k

~

s Ws.

L.,w.d [*...m :

ip**d b

F-.'M.d,*, d<o.*.n.ap..,8W.*5 f#AFM'hk.Ds

.-m

~.. *~ n ~ '. m

• Y <s**m,*~"*N "+w.r'k= %' 7' 7 4 WM %K 4 4 *** * % Q V '

9A3.) %w%n. J

• gcq = W so* *..b** *.~. ~. v 'a..

i

-w.J

, p.,.. c m m..p. q.; m.Q c w w w.*7?

M Y'

n.

M

+

ww%

m

.7

. V T3

.~rA.we

.w r

u

- w s4 y

" '"en

.. V

• +

^

+ '

~

9-J O q $.*l ~.* :.

~

• W.

.s Y. $

~ ~ '

owe'en =~.* eau ** - -

-v u

= + 4.g.,ee.-M A. *,ev_1 m

- -+

v~v - -- --- -- - -

.7 m

,3 a

.n ur, -

~..

,,. 3

- a w + M.,.. 9.

~,

,,s.y

, j;,, ;, - g.

w.w

..,g g,5. m.y - q.-

a m

.-a..y d-a.r; r;,.~.t wuram.;.%.uy$,

? w. :

-.q :.

s

ww s..w 4'

t s -

y

__m

p. _

c A

2'6 IW A.

i -

Y Ch- ::C C in raquirm! tc.

er i,.

tc 1

'- E uc.

i' b

(a) te liri-th. per.J. f ue l cit.dd ::#

up'ut_

+

i.

-tion c rr. : : 1-j i M e n ' to 1:e 1?GC 0, (t) n 1*

'b a'

..d 0

T

)

I t-m : h c!:.'iinc anj w2ter- %

'n 1~

t h.n th

-14

~i tot n mat of claddinc (c) o ' '"

m-*C

.t'.

tran-

., r c r e p u. *r a v u:-

T l

cient of tr 1.idding se that the :cre 4...,

l

. ling ar.d Lefere the claddity; i: ro Frit*' : ss tu fa;l

.p m

-r.

Ling, and (d) to reduce the

~ra *c ;e rst ure and r emove py -

riccay heat for an extenicJ pericd Of tire

. ~...

im a re ult cf an exten::ve :<

v~ i n I N., program; w re tw.+.

4 initiat<.: t y tu Atomic Ener gy ccq:n.-

! t h< nud ; x in-Q u.

. i dq:try leadinf to 2=provem^nt in pr :.r j cyst em it.* cerity,

.% w j

' b"T '.

developnent of irnproved analytice.1 wth:J: fer prclicting U 00

.$s$fM performance, and acqui:ition of ex - ' t. n t a l inf er.at ion to j A; ca;p rt anl cenfir:n the analyces.

' y ;..

are Jc0crite in e

a J yZ'-

The "xperimental and analytical p:cCrum:

J.'o ".

a cctpaniet. p a pa r. '

The related prec e;:e : Of dcrelcinent Of

/

MN A 1 analyti.el tcol; and confirmator; n;* rimcntaticn are ' ::1 a t e J

. Q,[g'?...

to cont.inue, lisever, use of th.e new,

ra 20phirticate,

E e

N w.a pc : f ormance in, ?icter ry analycir, techniq mo for evalu:; tion o,.

. gp,ag..c c/ctems ha tetun.

In view of the large at ant of new in: creation

,4h 4

l h %i R available, the Atcuic Energy Cc=.icai:n aR ain Ocn!ucted, in 1971, "U

mr

.N '.

a review cf the current state of ECC: techn: lory,

'. F '%gu+% 9, prw u -

Ideally, cne would have availat,17 analysis netheds cap able f

w.,

g* p j

i of det2iled realistic prediction of all phenomena kn:wn cr sus-d h "*

Ng8, Mc,

pe:ted to occur <!urinr. a LOCA, cup;crie in every a:pect by N

tc t he LCCA.

I.n the

-M E

lefinitive_ experiments directly applicabic Mw w+ A.

a mv aLcence et wch perf ection, adepite c::.:ranc~ of

• fc/ is h*hflh o!!2ined frem an appropriately conur eative analysis ta xd en

,^,

availatile experimental infor:2:lon. In areas cf ine pleta gg g*q Pnculedr<, conservative ascutpticn: r precedure: cro gp'. led.

z.,.L.J.'

3 # ".

M further experir.cntal i:if er.atien c: it; t eved calculat igny.1

' %y^# "< 'E k k$

technipce t ec0r:c available, the een tati:r a previca:1y ir-

[{

J o.e 1 can he reevaluated and a mcre re al: tic ipproach th en, lph,E.

t O r

batified by the improvement in knew 1 4.co.

pc._

~. ~,

~

The result of the 1971 review wu the : :ve le pr.a n t of

%y.

61 a

'7,,

interim ev31o gien =cdels cen i n inc Of com uter cedes, auh

[,(

r,p t00cther with its set of sultatly Ocn %rvative 2:577 tien? ani e

3 4.s,..s precedurcs, to M applied to each plan.

~

_s.

b..

.Qt h,"

y

.s r.

g,n.

m.

d.

~

r%

h5 N

y 1.e. 4 i Y es

. f ~* ' n Cr7a Q~.+h$7.'n.

.i m,

v.~.,.< a m.. - i A 3T t ;

A

.M y,e t. ;

b,v,,

o v:

M a(,.

b, 6t D "

• 4,,

' l'%h & F 1 I. G. T. et al.. d4% i f *C.(m. f.g r.

CTU t

p.

g, H

11,

i POOR ORMNAL 1

'9%

.*+ +..

s. ~.iw' s.

U_

,.w

.e w

s., - ;f 4

O(

MDt *

• (t,,.

h g ( h Y.

G.!$$

%'4

""4 a

-.'.n

,.,.. s ;s ~..., g. cg,. ? p g s s

a P us y :f ^; r * ;

',u e.,I m uus

- _" '" - Q

M-

^ * *

. en

-w...-~.

s.

,m,,._

c w

.