ML20154C211
| ML20154C211 | |
| Person / Time | |
|---|---|
| Site: | Peach Bottom  |
| Issue date: | 11/30/1986 |
| From: | Howard R, Mehta H, Ranganath S GENERAL ELECTRIC CO. |
| To: | |
| Shared Package | |
| ML20154C177 | List: |
| References | |
| NEDC-31339, NUDOCS 8805180079 | |
| Download: ML20154C211 (102) | |
Text
--
- - s '
- v :* ~* .-..-~sa 4. s.*
- r> .wv. gm ~'
. . . . ..:',x. *., e ~ . . . . 4s .** - . J. . ,- ,;. *s > ~ &; *.. .
^
- i- s.- .
es - .**4
-n f* 6;'.'.*ita .
y;&m. . &. . ;, y ' ;,1 ."i .b. .A,*A s,. [ f. O.[ .y f;$ ,g*g.= . g ' . L' * <G S '. ... p . S'S.
o 74h' W .
. , a .,, .
'? k.,,.' _,' . .., ...~ S. .?a'M+Y,. .; .' ., ., . . .
. p i,Q.,.. a)
'# "*i- G. ,.%.*.$. :. ,. ).*);ud ~'. '." y;*r'*. . . . s _ f w& *,; ;
' f . . ,%
-;, G. ,y .i i ., .< ;Q D q- 15
.y&.,w%;i.f.Q.Q3.hy';;.*%
- ,1.,....
.1 @;. c ,J" ? fM, j.;$.- N .Q.M.$f. QDCe.3.1,3&Q. j,*" t
- a ..;m p,.7 @.# w ~C ASS Q
.* ,s .
A.
pg" .mfy;%.,.p.,u % 4 . M z,.g.,j; p "% m . ,,,.g.;*:pW/.2% ,
+ av .
j ;b. .=ym 1w U Qcs.-1;' ,.y t.c. Qf..9'W<'J2 p.::.g.,/S.
N VEMBER L q E WIV Mpf,"k;7.M'm&ls IM 9 E pk;$ %u m;Q.w. .w2 m .1986 W.'tmK.pNup'M4-Wfghh.Ml@E;'U$. k&f**Q *QW mt dMwwwg W8. a;.m'4*t:.5. N n'},W.e'* r
- g, w.
s' I ~.:* n****Q^y*i.j. m a g ..aQpQt.*.~4.Q.f,R, h
w 3.n F M.';;,)c .$. w%.g& g.f a fC .. Q
.> v . .., m s .
'Qi. w,.;%%l}T
..'...n. s Q. ~e;Q%% h,,ss?k.,e 'n, .
- 6. .-f. f.a&i.:
.2*Q .p'O , .lyi . }'Q:' 'wl&,.~
, J .f.. e +3 ,r % aVr.e .N=+ . s..e ..
...&'4 $. f*;f.*g,M c . , . c
- ir- ***- 1 .,.?*l';
.f.E=w=*
- y
% .i
-4. W .M. . .1h.
. *' a.fe u-.r ? ; s ? f *at y *,;b
- 5;,
& ",,.*.i.
=a- - "
u+
..s-:i.&~. 2; '&,4.W r,* .
('Y, 5.".r. M.. 7=*bf *p. d.#a a .?nd.'& H t.M.m **. , --+ :."h. p,7 * =.W. . 9 3 .,. : Sld '.,.
cu - M : wow' ?~' l a ,R.".A'. - - s.* .~ ~> ., ~
.p .A *rs,.,=. 4J.'...-x s- . - -- -
.v..
s.. s . -***. ..? ;n.*'*
s ti i:;M ; N. ' *-J.n. .W?; ' e S e;.t -# :' s e M .: .
j.g .Cp qm R,.y @ nu N. m r x.M.,.x.%.c. - ;",.e Fr :". . q;, .. _rf.y y'w n nh(q;u.'ilW**rrla.';jd;,M.,.G.N i ,M@.7.U;.'E %.g.;W{*u.7,'.W.~.:
I
- y. w. %. . .iu. p. n. 3 ,:,.: . G *d %.. er ;swy.
M..,St 'rprs~3 u .
. w y q, p ;y s: ..u
- ,-+ .,p,.. 1. a s ,,
,.
- a . .., . .,v. o;. , ..*y, * .:.: g.g
. ~ : 7. ... .. . .m .3.,. ,, , e r..y.p. 1a.w a .,.. ,. . : uA. , y.r.e. . #. . . ,,J<s m
, g 2.-..4 w.;q..--w . . z4 . o
.,, '.,~*s.
.6
,Wa%**
.; .: y: .:y, L A' ~**p ,c4f C y,. y;g- :t 'i p W.* Jr ~* . .i wt. ( 3.,
M:". ' %.: ,.h ,:,.
- **'; : Mw-.x';ms.. .. 2 .'. N.,.;.,e.. u=,' ,sa;. , :.% - < ?* -* M.v O.f.%. ; i* a
.. "* 1 - . *a . . ' - . , ' ~8' sa. .
"E
. .as. .
.. r,,. J*. ,** ',.,V,*'
. . ,n . , . -
..v;, .r sa =. . . L. 4,.,;, ,...e,.. . ; ..s , JefT.
r ,h-, . m. .-
. % }. c .p . . C; . ,*. *.%x~W. ~
e s ,
,>.. ,.,g'
. ..r cm.
z * ." o.%. '. , ,.. , . ;. ; S g. 'p , ., 9. C. .*n*
A ,2 *, ~ -., .,p, 'm. 4 i;t . . b
.;..4,,.'s.i.,',4..
- > v C r, .%.
..._ . J, ,e.#,C.'. m5r >
, . ., ., .4. . 3.#.. u. . <*. . . -a vw n W .a.- .q i ,* . g.s.s .
s , j ,a n.
..s.
. - fra . - t* .- ..- ' >
.a.s c C *- - ,e, :
- 4. ; q4, f.' -
.' '.. . r**
.r.m..m. -r J. c.,r. .;' m.. p.*..~..ic,..
. .J ,:,.,. y 6.=.c- ,.yn : . .$.. l,M.+. Qf. .- ..' .T/w,,9. Ac : .;y.n.*. . ;s..j. ;.-l.**M.,. ;,
-... ,m,/
. . .g5, y./ Af. y
. ;, J .
. m..-
- g. .
w..,a
,.,.w * . + .. u .e,? * ,y.,.,
. ;- A .w , _s n n~ -e.e A..';*
s': .
r.r..
r ,,
,9- ' . rg:2. .e + q j. ; .- wy?*,. , . f.L.y,;.e n,, ;. ,
- , ,e *- Y, , \;, ,
4
. g e.$ .,,'..; 7.; :;y-
/
. ,
- x .w s,,s';V,.i ;% p .:.y &s , . -.y,. ;* ' g; Q, y **;,? ' -Q g*b .. 9..k . . , y ;; n *;* L')'<f'r &
, c. .
Q %,. .').+-f*
- ..t
. r .,'4~& W on ~~ *.. m.M - ..
- f. . .& yiss on.us'.p:;.) o* * *.
- y,\ yt .me.:ry-t.
,- a # y. . w.
o-
.. . t*:4 f.a%L'.C.D.0
,. . a
.s,,). f . .
-.~- ;. :. .v.i.,, ,'y ,, ,, /[ * / . ., Q,,* ',. ; . . f f* 3 '* f . a.*.*r' J..",
r,)
. L
. .q , , . w. . .r .sg. . . T . . . . . . J .< ; .. ..
, y yo..,.s.% . . .. , , e ar.,,4 g y ,V,,,.Q.',. , ,
- m. m,. w.n.. , . . _a.
8 J. %. e. .
, . _ , . . .- . . [
.p+. ,.,. n.'ww4,t ...y g'T9 m :w~. ., m. .~..w.*.m.x.
q r .ca~ , .n.n.: ,n,,, ca4 w.m. ~ .u, me..
~ . . - -
.w p w . . .
l .s,a
_s.u. m .w .
. .k -
- m. - - -
'.>4g .; & *o
, v L
' - s.% 'e v* . e ?"tTr '.e.'y '. . . : . '7't .
'.. . . t . ;* *j2. .
- Y
'Q' ,,. v
- y w. v.e. - n . . .; ,. .w, v.;v ' ..! .f 'a.h? e r % *. / .e . -
- e. *.-- f. '-% c .-c.., -lM : %. c3 \.,..a. -.ll,1 N< J
. d ;u.*s*e.+ ,. *,.s.r* k w*.a.,. - 3,
, s . . .u .
. n .,3.V. . A.;:a. - s*
- .v. 9 . ..u. , .;. . . - y.. .n., .m-%
- ;.;.yM:: 3,.. ,
n, qf: ~p a y ' w ,*- 2., -- .v y ~, *n:).o . T.. ,
1 OOP.SA W+
BWR' OWN;ERS+' G..R,D*V~FCD.SSES
~ -
i.0;PgiD iWC hi "j,kM.EP.E..GENg -
gpgyg, . .s.%.. *.g.g.3.
o- ~~...a. -
,s .
~ -
. ,g. , ....,
. gg. g.
.,.g4...
. O yN -
t e
. D O.. . or.
.y.
- e. ,q. * ,9 - rs . ., . . ,: a- ,
- - . .. .p-i . s.. .
g . p*V , 4.f e4.IN.BO.nlsl.ih, , ?WAT,
- u. . - -
. . . , .n ; ,. ..
E R. ,o,, .-. .R.EAC.T.O RSJ,. - M,.-~.
i..
.,3..
...m....,,. . . . m . m..e
. . . , , .. . _ . . . . .~
w .. 1
_r.c. -...- .
.-:m. . . . . . . o .... .
- y. . L
.e . .a , r . , , . ~,.; _.
. . . ~ ~ .. , :
m ..v ._ f.9, .. v
..c.;v ..~.
. .. . .. _. . r. -....., .
, , . ,.. yuf, w-z .na.
.... r -,.<...
-s
~
_ . .g. .. .a. . a. . ...s s n q.,
.c.s . f.):. .,.Y
.%d P,W
- s. . ..;.2 x ; .~p . . , yg .. . ...,a - - ;...? : :::.t ..:e , - . . ..
- .,
- , .a.+ ,
..;c ...H. S. M E H. T. A .
1 4 ,A,N.
L. .+ ': :# 2'c.i: .c. .. L..
- ; .y*.;, ,7;.m ,. . R. W. HOWARD : --
1;f
.m-.~ .
- * #. **. ^
- *, .,M;
. . - ,e. . - .
. a. ,.._t e*..
.,g ...- . ~ . ..
z .. .
.~
s... .
+..
. . ',.,.* l 'p'.
- v. ..e s . we. .r ' y .
- l'_. * . Q , ._ g - .. t ,
. & . .J
, l,., ,
- .r h - e . .e. . s .
. ..g_.
,s.,,,... -., , , r. ; . .
- 7. ,:. .
. #g ..9 3 7 ,
. . g .;. .. . . . .. . . q, . . . .. .
,1, :
-. , . . ~ . . . , .,.,.'*4 . (- 'd'*#
de' "%
"'...~*a,.'*
.w . . . . .
. . .-[w.. O*b
. :- ) . .
- / 4**w .#*. . ' . .t' --~
.n
.c.~,;.,..,. ..;.,.'..
' ** M # 9 '
e .s*=-+f
- * , . . .*g,,,,
.-#...,..~\ .m
.-~.y-.., ..,,.. , , . , ,,
..7. ,. . .-* -.~. - . r .~'a
.s .: + = . . ... - *. w s.*. e . . . . , . . , . , .
,r , ; . , , y ', ,, u..
.ag em. ;, .. . , . =.e %.g p*m. . .
j.g u . , w Q Ng e
- ,?
nu.,4 3 m. /. , .. G eseisimm&4. . m.n. . . .s2. .. - . .s.,.e . ,, wsm,. .mmeasug,,,,,g,,, L .,. , ,.m.._ ..o y
. + am e. 56 rb ( , e en e . ., g
+ * . . * 'ere % *.y*"' '
- . *
- p y .7 . e . e -- 4
~,'...'.m, ,- . g . . g. ,e p . . ;4* p ,.m g an;..s .. .g-,3,, , ' . .
a ,'a m' .e . ** en eag
.~ ..
........:..,.,,,w'.^s...,,,.8.7.. -
. - - .,,.e','
, , ;; ,. .,J 7.q@.;*m .'vg m*
.nd
. y. cs .$.i.,
,p .34. p'" s'.v 'J
.+v er ,., ..p
..3 . . . ?. . e r . . ,,,,7..s
- 6 .
_;- . , . . . .. .. w.y .,,m.. , ,
f'.. . . . , ..
- p. ., , , .....* e . p;w *
~ * .- . a .vv y.u - . - - - . . . .s.. .e ...
- f - J.'* y . , . . , . , ,
.y.:,, "?%. .. . ... ..:h e: z t ,. q!A
. 5 .;.j ;, - .g , .. n . 3 ., ; , , a u , \
..y. ..
. a~ : q.3.... g.g.. . -
mr. .. ". .. . :f * . , ..
. .p. . . -
.- . ,g. g , y , . .:.:s ,,. . . 4.y.3.v;y9 ;
.::; . : 2 Hw%.~2.y.m;,d,
'C. yll W.pg+...u.y;M.7 ;p%g,:.t%.g.3.m:y.m ar u:. . ~:+.::y . .. ..nc .u ,m.%; g ,w.
vw,*>y.
. , ,s ..,
.,- ;:s r s: -w,~ .. r . .. . - , l
. . et .
e .*
. - - . . .r
.v
^
, -,, . . a . ,4. .. a
.'s,. - ..', . - l;.g , :
1
- 'g .j+...-.=g- . . g=,. .
- e. .
'.,y_ ... ., , , _
,% ., Q, ,
.p.- * * "** .. , , , , , g% g - s_
m,g..
.....s~.
, , -- -..o. .
y , . , . . . , .
.w.*%* -.
.,f.
c.
..a. z :.; , . . . , ,- ..,; ..;. : y_ ,. _..: n .
. .,. .,,m. . .. ...;.., . . m .. ,
. . m. -
.,.s.
sd..:
s
.x r .
- .4
.- 7,.. . .
d.4 m
. . ;,y
~ . -
. .. n .. .. ~.
" = - - ' " 3,;.
- .4 s'
a
- , p. . ,
... .%. m.i iN gN,e (v.r fE3 n o
- Sg
- C L,
. . " t r *:cio.* *' ) .
i 8805180079 880511 '
DR ADOCK 05000277 DCD
NDC-31339 DPJ- 137-0010-86-57 CLASS 1 NOVEMBER 1986 4.
1 BVR OWNERS' GROUP ASSESSMF.NT OF EMERGENCY CORE COOLING SYSTD4 PRESSURIZATION IN BOILING VATER REACTORS Prepared by: L Hardayal S. Mehta Principal Engineer s
k / m ised -
Robert W. Howard Principal Engineer Approved by: --"L
S. Ranganath U Manager Structural Analysis Services NUC.E AR ENE AGv 9' JS: NESS OPERATIONS e GENER AL ELECT A C CCMPANY S AN JOSE. C ALIFORNIA 95125 G E N E R A L (3 ELECTRIC
NEDC-31339 i r.
' l IMPORTANT NOTICE REGARDING CONTENTS OT THIS REPORT PLEASE READ CAREFULLY 4
l This docu=ert was prepared by the General Electric Company. The information centained in this report is believed by General electric to be an accurate and true representation of the facts known, obtained or provided to General Electric at the time this report was prepared.
Neither the General Electric Company nor any of the cortributors to this dccuzent makes any representation or warranty (express or implied) as to the co:pleteness, accuracy or usefulne;s of the information contained in this decunent or that such use of such information may not infringe privately owned rights; nor do they assume any responsibility for liability or damage of any kind which may result from such use of such information.
r
NED2-31339 l
TABLE OF CONTENTS Fare ix EXEC 1'IIVE SLT.ARY
- 1. IN!kODUCTION 1-1 1.1 Eackground 1-1 1.2 AEOD Report Conclusions 1-1 1.3 EVE Owners' Group Scope of Evaluation 1-2 1.4 BW7. Owners' Group Committee Activities /
Objective./ Approach 1-2
- 2. CONCLUSIONS AND RECOMMENDATIONS 2-1 '
2.1 Conclusions 2-1 2.2 Reco=mendations 2-2
- 3. AEOD REPORT ASSESSMENT 3-1 3.1 ECCS Piping and Cocponents Affected by Overpressurization 3-1 3.2 Frobability of Pressure Boundary Rupture During Overpressurization 3-7 3.2.1 Piping Integrity Evaluation 3-8 3.2.1.1 Hoop Stress Burst Margin 3-11 ,
3.2.1.2 Liefting Axial Flav Lengths in Piping at a Circunferential Butt Veld 3-11 .
1 3.2.1.3 Evaluation of Probability of Pipe l Rupture at a Circumferential Butt Weld 3-14 3.2.1.4 Conclusions from Piping Integ-ity Evaluation 3-19 3.2.2 Evaluation of Valve InteErity 3-21 3.2.2.1 Valve Body Integrity Evaluation 3-21 3.2.2.2 Body to Bonnet / Cover Bolted Joint Evaluation 3-27 3.2.2.3 Conclusion froc Valve Integrity Evaluation 3-28 3.2.3 Evaluation of Heat Exchanger Integrity 3-29 3.2.3.1 Heat Exchanger Shell Evaluation 3-29 3.2.3.2 Shell to Tube Sheet Bolted Joint Evaluation 3-31 3.2.3.3 Tube Integrity Evaluation 3-32 3.2.3.4 conclusion fro: Heat Exchanger Integrity Evaluation 3-32 3.2.4 Overall Syste: Rupture Probability Evaluation 3-34
. 3.2.5 Probable Results of Overpressurization 3-35 3.3 Probability of Interfacing LOCA 3-37 4 REFERENCES 4-1 4
-111-
NEDC-31339 TABLE OF CONTENTS (Continued)
APPENDICES A-1 A- .EkT PLANT DATA SLTJJtY i
B-1 B- LLNL PIPING RELIABILITT MODEL C-1 C- BEHAVIOR OF FLANGED JOINTS DURING PRESSURIZATION D-1 D- SAMPLE COMPUTER OUTPUT OF PROBABILITY EVAI,UATION f
E-1 E- PARTICIPATING UTILITIES - Bk'R Ok'NERS' GROUP ECCS PRESSURIZATION COMMITTEE e
t f
1 I
L 4 ,
i i
I i
i I
.i
_ 19 i
NEDC- 31339 LIST OF ILLUSTRATI0SE Title Pace Tirule 3- Typical Configuration of RHK, CS and LPCI/CS 3-2 3-2 Typical HPCI Configuration 3-4 3-3 Typical RCIC configuration 3-5 3-4 Typical HPCS Configurat.on 3-6 3-5 Surst Test Data 3-12 3-6 Schematic cf 150-Lb Pressure Rated HPCI Valve 3-22 3~ Scheratic of 63-inch RHK Heat Exchanger 3-30 3-f Chart-Tube Wall Thickness versus External Pressure 3-33 1
l l
l l
-v11-/-viii-1 1
1 l
NEDC-31339 LIST OF TABLES Table Title g. .Page 3-1 Calculated Hoop Stresses During Overpressurization in Representative ECCS Piping Sizes 3-10 3-0 Limiting Axial T1. rough-Vall Plav Lengths in BWR ECCS Piping at Reactor Pressure 3-15 3-3 Axial Stresses used in Probability Evaluation 3-17 3-4 ECCS Calculated Rupture Probabilities per Circumferential Weld During Overpressurization 20 3-5 Valve Body Minimue Thickness 3-23 3-6 Hydrostatic Shell Test Pressures 3-26
-v-/-vi-
l
'NEDC-21339 i
EXECCIVE SL'MW.ARY The Office for Analysis and Evaluation of Operational Data (AE00) issued a
- sse study repert of operational events involving actual.or potential ovsr-pressurizations of emergency core cooling sys :ms (ECCS) in boiling water ,
reactors (Bk'Es). The probability of an interfacinE loss-of-coolant accident (LOCA) is the product of overpressurization event frequency and the probabil- [
ity of system rupture given the overpressurization. k'hile the overpressur-iration event frequency is supported by actual industry occurrences, the prcbability of systee boundary rupture stated in the AEOD report was "judg-rentally assigned",* and believed by the participants in this study to be !
unrealistically high. This report was developed to present the results of a Bk1 Owners' Group (Bk'ROG) assesscent of the probability of ECCS failure due to l overpressurization.
ECCS cc.nfigurations for 19 dorestic Bk1 plants were reviewed. f roc. which ECCS piping systec cocponents subject to overpressurization vere defined. Proba-Filistic cethedelc.gy, developed by Lavrence Liversore National Laboratory l (LLNL), was applied to the typical ECCS configuration to assers the prebabil- !
ity of pipe rupture during overpressurization from the presence of latent j circumferential veld defects. The rupture probabilities for other compenents, j such as valves and heat exchangers, vere approximated with reference to the i piping probability. Fro = these. evaluations, the expected probability of an interfacing LOCA vas essessed. Additionally, deterministic evaluations of safety earEi n during overpressurization vere perforced to show that these l targins are greater than those specified by the ASMI Code to provida assurance I
against gross rupture.
i j The Sk'F CVners' Group (Ek'ROG) Cc==ittee has concluded that the conditional
! probability of 3k'R FCCS pressure boundary rupture during an overpressurization event is no greater char 3.0E-5 per event, two (2) orders of eaEnitude less than the stated "judg= ental" AEOD probability. This report demonstrates that
- Section 4.2 of F.eference 1 has "juderentally assigned" values to rupture prcbability due to uncertainties in undetected flaws, component leakage and waterhancer potential.
i 9
-ix-i l
1
NEDC-31339 EXECUTIVE SUVE.ARY (Continued) l the resultant probability of an ECCS interfacing LOCA is 3.0E-7 per reactor yur co: pared to the range of 1.0E-4 to 1.0E-5 as judged in the AEOD case '
4 study. Therefore, the expected probability of an interfacing LOCA is not l
significantly different from earlier industry assessments (References 4, 5, 6 and 7 of Reference 1), even though the frequency of overpressurization events may be greater than previously assessed.
This study was funded by the Bk'ROG. A list of participating utilities in this a,tivity is provided in Appendix E.
1 I
i I
(
a L
'E 4
-X-t
't I
_ _ _ , _ . . . _ . . . _ . . . _ , _ . - _ _ . ,-.m. . _ _ _ , . , _ _ . , .,_ ._ . . . . , - . , , _ _ , _ . _ . , - , , - - . . _ . _ - _ _ . . , _ - _ _ . ~ _ .m,_m. . . . . _ _ ,
5.33
' ' ATTACNu!K7 2 5.35 October 1987 Appendix R Audit !
5.36 Open items Not Reauirina NRR Review 5.40 Fire Damper Operability (Unresolved Issue 84-40-01, 84-19-01 5.41 discussed on page 4 in Inspection Report 5.42 Nos. 50-277/87-30 and 50-278/87-30.)
5.45 A fire damper program is being formulated to evaluate existing test 5.47 data and damper closure with air flow data and to address fire brigade and 5.48 training procedures tc provide reasonable assurance that the fire dampers will 5.49 satisf actorily perfom their design function. We plan to meet with Region I to 1 5.50 discuss /fomulate such a program by April 1988.
(Page 11 in Inspection 5.53 !
Incorocration of NRC Comments on Procedures Report Not. 50-277/87-30 5.54 and 50-278/87-30 5.55 :
l i
5.58 c SE-10, Piart Shutdown from the Alternative Shutdown Panel, is fru hr/
5.61 currently being revised to reflect changes caused by the completion of Appendix 5.62 R modifications. During this revision, operator, training, and NRC coments In addressing NRC coments, 5.64 were reviewed and incorporated into the procedure.
5.65 sign-off spaces have been added where needed, and the monitoring of the reactor cooldown rate has been enhanced.
I
Attac9 eat e Page 2 5."
A::a:: S ' ' ' t:, :' 9 0" : .t:a-d St aa- :::' at ': '.'a' . a ta-i.
6.4 (Fage 11 in Inspection 7seport Nos. 50-277/57-30 and 50-275/57-30)
The HPCI Inboard Steam Isolation Valve panel was originally provided 6.7 6.9 with slotted screws which required tools for access. The slotted screws were 6.10 changed to thumbscrews to allow an operator to access the panel without the use of tools. To address the NRC concern of overtightening, flat washers were added 6.11 6.13 to compliment the thumbscrews. The washer addition will provide a smooth 6.14 contact surf ace and enhance the operator's ability to loosen a tight thumbscrew.
(Page 13 in Inspection Report Hos. 6.17 Fuse Reolacement Controls 6.18 50-277/87-30 and 50-278/87 m)
Administrative controls for fuse replacement are being actively 6.21 6.22 reviewed. A modification has been initiated to generate a controlling document 6.24 for fuse replacements. For the interim, a guideline document is being added to i 6.25 :
the operator's handbook to assist in the current practice of replacement in- ,
I kind.
1 I
I
- l. INTRODUCTION
.1 BACKGROUND :
The Office for Analysis and Evaluation of Operational Data (AIOD) of the Nuclect Regulatory Commission (NRC) issued a Case Study Report AEOD/C502, "Overpressurization of Emergency Core Cooling Systems in Boiling Water Reac-tors", dated September 1985 (Reference 1). .This report summarized the AIOD analysis of operational events involving actual or potential overpressuri-zation of an emergency core cooling system (ECCS) in boiling water reactors
();T.) since 1975. The operating BWRs reviewed were product lines BVK/2 through BWR/6. Reference I focused on overpressurization or potential over-pressurization events that occurred in BVRs due to testable isolation check i valve failurec in all ICCSs. It concluded that these overpressurization events indicate the likelihood of an interfacing loss-of-coolant accident (LOCA) to be higher, by two to several orders of magnitude, than had been previcusly assessed. The BWR Ovners' Group (BWROG) authorized a com=ittee L
~
activfty to evaluate ECCS overpressurization and tssess the capability of the ECCS cc:ponents to withstand overpressurization. Thic BWROG report su==arizes
(
that assessment effort.
L 1.2 AEOD REPORT CONCLUSIONS ;
The AEOD review identified eight (8) events involving the failure of a testable isolation check valve, and expressed concern that these operatienal events should be cor.sidered as a precursor to .an interfacing LOCA involving the reactor coolant system and an ECCS. The AEOD report states the event frequency to be 1.0E-2* per reactor year combined with a "judgmentally assigned" probability of ECCS boundary, rupture of 1.0E-2'to 1.0E-3 per over-pressurization event, resulting in an interfacing LOCA probability of 1.0E-4 1 l
l l
l
'The SWROG Cotrittee believes this event f requency to be overly conservative
, based on a continued operating data base without additional occurrences and j an increased industry awareness of overpressurization events and the recogni-tion of the need to reduce overpressurization event frequency.
1-1
- NEDC-31339 This AEOD "judgmental f requency is two (2) to te !.0E-3 per reactor-year.
three (3) orders of magnitude higher than the frequencies previously assessed by the industry. d 1.3 BVR OWNERS' GROUP SCOPE OF EVALUATION The objective of the BWR Owner's Group evaluation was to assess the failure potential of ECCS systems, piping and components when subjected to everpressurization. The objective did not include evaluation of the conse-quences of discharge of fluid from relief valves or leakage from pipe cracks, gtskets or flangad joints on the basis that these discharges, in most events, can be co:pensated for by increased feedvater system output due to the lov fluid volume discharge rate from the sources relative to feedvater capacity as Additionally, it is deronstrated by the events discussed in Reference 1.
leskage from these sources judFed that there is a very high probability that car,be isolated.
Stresses in the lov-pressure side of ECCS piping and ce:ponents were evaluated based on the syste= infor ation received from the participating utilities and GE in-house information. Safety margins were evaluated and compared with ASMf Code-specified values. Quantitative evaluation of rupture probability at a circu=ferential butt veld was evaluated and the rupture The system probabilities for other cocponents were qualitatively evaluated.
rupture probability was then assessed.
1.4 BVR OVNERS' GROUP C0KKITTEE ACTIVITIES /0BJECTIVES/ APPROACH The !WR Ovners' Group objective was to evaluate ECCS overpressurization and assess the BWR ECCS capability to withstand overpressurization without rupture.
The frequency of ECCS overpressurization events in the BVR is well The EWROG response to the AEOD case study has docurented in the AEOD report.
focused principally on assessing the realistic probability of low design-pressure system pressure boundary rupture given an overpressurization occurrence.
1-2
- .
- 1. INTRODUCTION
!.1 5ACKCF0rND The Office for Analysis and Evaluation of Operational Data (AEOD) of the Nuclear Regulatory Commission (KRC) issued a Case Study Report AE0D/C502, "Overpressurization of Emergency Core Cooling Systems in Boiling Vater Reac- '
tors", dated September 1985 (Reference 1). This report summarized the AEOD anslysis of operational events involving actual or potential overpressari-zation of an eeerSency core cooling system (ECCS) in boiling water reactors (IWE) since 1975. The operating BWRs reviewed were product lines BVR/2 through BWR/6. Reference 1 focused on overpressurization or potential over-pressurization events that occurred in BWRs due to testable irolation check valve failures in all ECCSs. It concluded that these overpressurization events indicate the likelihood of an interfacing loss-cf-coolant accident (LOCA) to be higher, by two to several orders of magnitude, than had been !
previously assessed. The BVR Ovners' Group (BWROG) authorized a committee activity to evaluate ECCS overpressurization and assess the capability of the ECCS co:ponents to withstand overpressurization. This BVROG report summarizes ;
that assessment effort.
1.2 AEOD REPORT CONCLUSIONS The AEOD review identified eight (8) events involving the failure of a <
testable isolatior : heck valve, and expressed concern thet these operational l events should be considered as a precursor to 'an interfacing LOCA involving
! the reactor coolant system and an ECCS. The AEOD report states the event frequency to be 1.0T=2* per reactor year combined with a "judgmentally !
assigned" probability of ECCS boundary rupture of 1.0E-2 to 1.0E-3 per over-pressurization event, resulting in an interfacing LOCA probability of 1.0E-4
- The SWK0G Coerittee believes this event frequency to be overly conservative based on a continued operat4.ng data base without additional occurrerces and i an increased industry awareness of overpressurization events and the recogni-tien of the need to reduce overpressurization event frequency.
1 1-1 4
NEDC-31339 to 1.0E-5 per reactor-year. This AEOD "judgmental" f requency is two (2) to three (3) orders of magnitude higher than the frequencies previously assessed by the industry.
1.3 BVR OV5ERS' GROUP SCOPE OF EVALUATION The obj etive of th" dVR Owner's Group evaluation was to assess the failure potential of ECCS systems, piping and components when subjected to everprescurization. The objective did not include evaluation of the conse-quences of discharge of fluid from relief valves or leakage from pipe cracks, stskets or f.i.anged joints on the basis that these discharges, in most events, can be compensated for by increased feedvater system output due to the lov f;uid volume discharge rate from the sources relative to feedvater capacity as Additionally, it is de enstrated by the events discussed in Reference 1.
judFed that there is a very high probability that leakage from these sources car. be isolated.
Stresses in the low-pressure side of ECCS piping and components were
' evaluated based on the system information received frum the participating utilities and GE in-house information. Safety margins vere evaluated and co: pared with ASME Code-specified values. Quantitative evaluation of rupture probability at a circumferential butt veld was evaluated and the rupture
- f The system
' probabilities for other components were qualitatively evaluated.
1 rupture probability was then assessed.
1 -
1.4 BVR OWNERS' GROUP COMMITTEE ACTIVITIES /0BJECTIVES/ APPROACH
" The BWR Deners' Group objective was to evaluate ECCS,overpressurization and assess the BWR ECCS capability to withstand overpressurization without -
rupture.
The frequency of ECCS overpressurization events in the BWR is well <
docueented in the AEOD report. The EVROG response to the AEOD case study has l focused principally on assessing the realistic probability of lov design-l pressu'e system pressure boundary rupture given an overpressurization l
j occurrence.
i 1-2 4
. . - - - - - . _ , . - , , , - ,--,e%r_._
NEDC-31339 The overall ECCS rupture probability during the overpressurization event is the su: of the rupture probabilities of piping and the associated compo-nents such as valves and heat exchangers. It was judged that the most significant mode of rupture for the ECCS piping is that due to the presence of later.t veld defects at the circumferential butt velds. Rupture probabilities for this mode were determined based on methods developed by the Lavrence Livermore National Laboratory (LLNL) and previously accepted by the NRC in the studies of pressurized water reactors (Pk'R). The rupture probabilities for ether components such as valves and heat exchangers were approxirated with reference to the preceding probability of rupture for the circumferential i
veld.
l t
i j
i l
l l
4 3
i l
i I
)
l-3/1-4 l
- o. ',
- 2. CONCLUSIONS AND RECOMMENDATIONS
2.1 CONCLUSION
S Industry-accepted methodology from Reference 10 was applied to evaluate the probability of Bk'R ECCS f ailure due to overpressurization. The evaluation has led to the following conclusions:
e Deterministic evaluations of safety margins during overpressurization vere perforced to show that these margins are greater than those speci-fled by the ASME Code to provide assurance against gross failure.
e The realistic conditional probability for Bk'R ECCS pressure boundary rupture during an overpressurization event has been estimated to be no greater than 3.0E-5 per event. This probability is two to three orders 2
of eagnitude less than the stated "judgmental" AEOD probability of 1.0I-2 to 1.0E-3.
e Assuming the AEOD event frequency, the realistic frequency of an ECCS interfacing LOCA caused by systee overpressurization is 3.0E-7 per reactor-year compared to 1.0E-4 to 1.0E-5 claimed in the AEOD case study, e The cost probable result of overpressurization as indicated by events reported in Reference 1 and the evaluations in this report would be the discharge of fluid from relief valves and possibly leakage from bolted joints, and smoke generated by oxidizing paint on piping and equipment.
The consequences of such discharges and leakage are expected to be einical and vill most likely result in early operator termination of an overpressurization condition due to activation of high area temperature alares and/or visual observation of leakage by plant personnel as reported in the Reference I report. Activation of high line pressure alares and s=oke alar:s and plant personnel observation of smoke are also likely to result in early operator termination of the overpressure condition.
2-1
i The frequency of overpressuritation events documented in the AEOD report However, the probability of an
' is higher than estimated in previous studies.
interf acing LOCA determined from the Bk' ROC evaluation does not justify these safety implica-Ek'R operating events being classified as having "significant tiens" a. stated in the AIOD case study.
1 The AEOD recognites that none of the documented overpressuritation events has led to significant damage of the lov design-pressure system piping, pumps or valves. However, the report cautions that future events may lead to fail-ures caused by pre-existing flaws. The analysis performed as part of the EVROG activity confirmed that low design-pressure piping and system component failure due to overpressurization by full reactor pressure should not occur Furthermore, because of the design margins that protect against such failure. '
i flav analysis indicates that the limiting flaw length required to promote i
pressure boundary rupture vould require a through-vall crack of nearly three (3) inches. Field experience has shown that the probability of incurring This a
l crack of this size and having the crack go undetected is negligible. 6 the lov design-pressure Bk'R ECCS and RCIC l
provides further assurance that 4 System integrity would be maintained should overpressurization events occur.
2.2 RECOMENDATIONS l
- a. The current frequency of overpressurization. events it unnecessarily
.l 1
high. Corrective action should be implemented to reduce their occurrence. Individual utilities should consider specific actions I
to reduce overpressurization event frequency, including making plant operators core aware of these potential events and their causes as 5
vell as evaluating reduc'.ng isolation valve testing frequencies.
4
- b. Due to the Bk'ROG Committe2's assessment of the low probability of low-pressure boundary rupture caused by overpressurization, the issue should not be classified as having "significant safety impli-cations" and should be addressed accordingly by Industry and the f
I NRC. The interfacing LOCA probabilities stated in this report are consistent with previous industry studies and should be considered as a more realistic assesstent of Bk'R ECCS capability during over-pressurization rather than those stated in Reference 1.
]
J 2-2 f
2
, - . -- , . - . -. - . . . - - , . _ . . - . .-,- .n. ,.-n--.,
.. = .- - . ~ = . . . -..
t
'I NEDC-31339
- o.
- i
- 3. AEOD REPORT ASSESSMENT I
3.1 ECCS PIPING AND COMPOSENTS AFPECTED BY OVERPRISSURIZATION f This section summarites the ECCS and Reactor Core Isolation Cooling 4
(RCIC) System piping and components subject to overpressurization by the types [
of events described in the Reference 1 AEOD report.
The systems potentially subject to overpressurization aret Core Spray j (C') (BVR/2 through BVR/4); Lov Pressure Core Spray (LPCS) (BVR/S and BVR/6); [
I P.esidual Heat Removal (RHR) (all BVR/4 through BVR/6 and most BVR/3);-Low Pressure Coolant Injection and Containment Cooling (LPC1/CC) (some BVR/3); l High Pressure Coolant Injection (HPCI) (BWR/3 and BVR/4); and RCIC (all BVR/4 through BVR/6 and cost BVR/3).
As illustrated in Pigure 3-1, the RHR, LPC1/CC, CS and LTCS piping ;
sections and co=ponents subject to overpressurization include the piping !
sections and coeponents located downstream of the check valve (s) on the (
systems' main purp(s) discharge. The piping and components upstream of the esin puep discharge check valve are not subject to overpressurization because, during normal power operation, the systems are required to be aligned with the suction valve and flow path from the suppression pool open. This alignment provides a flow path to the suppression pool that has a large cross sectional flow area epstream of the check valve. Therefore, the piping and components upstream of the check valve cannot be overpres'surized.
There is a high level of assurance that the check valve on the pump discharge is closed and has a lov leak rate due to the methods utilized to caintain the discharge line full of water. Some plants utilize the condensate transfer system to e,aintcin the discharge line full. On these plants, exces-sive leakage of the check valve cannot be tolerated for an extended time period because of the excessive processing demand placed on the Radvaste System. The de and results from pucping the inicakage out of the suppression 4
pool in order to maintain the suppression pool water level within acceptable licits. On other plants, a lov flow capacity "keep full" puep is utilized to j
l 3-1 w , -, -- . - - - . - y y e . - -w., . ,-m -.
,-,my- - - - - - . - - - , ,,,.-~~-w- y.--c
- DRYWELL TEST VALVE PROVIDEO ON
- p -
- MO$ '* **
AO MO MO REACTOR UI N C2T22:22;
~ h PS
- H VESSEL M g 41GH OESIGN ** LOW OESIGN PRISSURE PRESSURE d I:
=e-- KEEP FULL SYSTEM
- 74; ::::(:::
lt 1;
H H NUMEROUS H H BRANCH " 4
--_~',r CONNECTIONS U HEAT EXCHANGER PROVIDEO
" I MINIMUM FLOW '
BYPASS UNE l [ON RHR AND LPCl/CS u n r
\ r 1
u n -
t MO
". H n Sv'PRESSION H POOL ~
O I ::r 3
& I PUMP DISCHARGE CHECK VALVE
+-*+ PiPsNG SUBJECT TO POTENTIAL OVERPR[$SURiZATION
- P:PtNS NOT SUBJECT TO OVERPRESSURtZATION
- OvtRPRESSUR.Z ATION POTENTIAL 15 UMITEO TO PLPlNG' COMPONENTS OOWNS VALVE
- PtPING< COMPONENTS UPST REAM OF PJMP 0:SCHARGE CHECK VAlvt CANNOT BE OVERPRES$UR; ZED TME OPEN FLOA PATH TO THE SJPPRESSCN POOL he., SUCTION UNE FROM SUPPRESSION POO'. AND min' M FLOW BYPAS$ UNE)
- PWVP O!SCHARGE CHECK VALVE '.EAKAGE RATE sS ASSURED TO BE LOW BECAUSE; R
- ON PLANTS THAT UTIUZE THE CONDENSATE TRANSFER SYSTEM TO KEEP THE DISC EXCESS!vE CHECK VALvf LEAKAGE t$ NOT TOLEPABLE FROM PLANT OPERATION A OF THE RESULTING FILLING OF THE SUPPRESSION POOL THAT REQU:RES EXCESS!
THE RADWASTE SYSTEM ARGE UNE FULL THE
- ON PLANTS THAT UTIUZE LOW FLOW CAPACnTY KEEP FULL PVMPS TO MAINTAtN THE 0:SCt:
FILL PVVD CANNOT MAINTAIN PRESSURE ABOVE LOW PRES $URE ALARM SETPotNT LF TH LE AKAGE r$ ExCESSfvE Pigure 3-1. Typical Configuration of RHR, CS and LPCI/CS Piping Sections Subject to Potential Overpressurization 3-2
raft.tain the lines full. If the check valve leakage is excessive, the "keep full" puep cannot maintain the discharge line pressure above the low pressure '
& Jar setpeint. Initiation of the alarm would result in operator action to reduce the check valve leakage in order to obtain an acceptable discharge line pressure.
As illustrated in Tigures 3-2 and 3-3, the HPCI and RCIC Systems' piping and components subject to overpressurization include the piping sections and cocponents between the main pucp suction inlet and the normally closed valve in the suction line from the suppression pool and the suction line check valve fre: the condensate storage tank.
The High Pressure Core Spray (HPCS) System (BWR/5 and BVR/6) lov design pressure suction piping and components are effectively prevented from being everpressurized by the check valve on the HPCS pu=p discharge. There is high level of assurance that the check valve on the pump discharge is closed and has a lov leak rate. This is because, normally, the leakage rate of the HPCS injection valve is much less than that of (1) the check valve on the pump I discharge and (2) the valves in the HPCS suction lines. Thus, the line forward of the check valve vould be maintained full and pressurfred by a low flow capacity "keep full" pu=p as illustrated in Tigure 3-4. If the check valve leakage is excessive, the "keep full" pump cannot maintain the discharge line pressure above the lov discharge line pressure alarm setpoint. If the alarm is initiated, the operator would be required to take corrective action J to reduce the check valve leakage in order to clear the low pressure alarm.
If the pucp's discharge check valve f ailed open and the leakage rate of the valves in the HPCS suction was less than that of the system injection valves, j leakage back from the reactor vould prevent the lov discharge line pressure alare and would result in a high suction pressure alarm. Operator action, to tereinate the high suction pressure alarm by opening a vent to depressurize the line, vould result in initiation of the lov discharge line pressure alarm, thus alerting the operator to the check valve failed condition and the need for corrective action. The relief valve in the HPCS suction line vould prevent overpressurization of the system suction piping in this event.
Therefore, there is high level of assurance that the discharge line check 3-3
7h Uce t
n ' !
r '
I' ' !' 3
) l
[
v E -
= V
= LR A t .
V t R g
t O n r
t i 0
8 P t y g
4 OO t e
O : MtAg _- T S
M C.
I t
tO t
GS$
y ,- C Df EO t
OC : r T l e Ae l QV Ru t
t bie P
J r r 1 D !
= O OD S
% = M MN s
! M =
0 1
A n
- E T OM r ie Se MiS f 0M NY t
o 0u I f OS c 8P = t
- =
it S i t g <: n ;; '
CA 1 l
<r '
- Ut g T y: 1 = SS i S LD s
C ON l
i g
O T P
= : OA I N N PL '
T 'M G G : : I S ^ U 8 F t t :
MG 0G E P 5 f R sR 4
- s T t u SN o OU S OS SO t
i HS ns tl R
t _- GE WtS E R't t l
e e
f !I 8 R OR PC m A o , ,
O HP tP P UIN t
s: i W u g M rr D , v S f I O i e E
E M r n OV o T t e
gs s
I s A i - i t KV f r N Ay C K nr 0 A e
8 ns B C ie g T
A uS S8 4 Cr t c e t
i 2
f S DC l v R t Nl CO .
W ' U R t i -
Y R
O _- S P i S MC(
'f i i -
M . S R l D t t t e.
R V N K l i
_ POO R
E O ft T N
A e
- t rn _
I N V T A l
e 2 E i
. O I OT t G g t v se
,,_S O S L OC T r RA UR T l' J
~ RO P P P
C U E S ES R
J 2 U B T Pf -
S U O RfA 3 E
- S N V S G
N N G ON E e
r I I LD a P PI AN I
i g
P P TO NC l
'l L E E 2 T D S
S
- ON PA E :
- V y
- DRYWELL U
I 'EEDW^tER atyTOR ,
s I MO MO 4 g >d ;
ogt NL 1r COOLER
' ' r TEST TO CST $ N X syt: -Y i
-- -- s
-'9 u
n gg
~ n GLANO n HIGN DESIGN StAL a SYSTEM n PRESSURE h
tj f ::-
NIs b ~
U
_._._._._.;_._.__.;; MO
$UPPRE $$10N M
HHl D
~~
POOL ll
>4: r Ih4 CST
- PIPING $UBJECT TO OvtRPRESSumt2 ATION Piping NOT SUBJEC1 TO OvtRPRE$$URIZATION 01 VALVI l e POttNTIAL OvtRPRESSURt:ATiON EXTENOS BACK TO SUPPRES$10N POOL SUCTION MOTO; AND CONDENSATE STORAGE TANE iCSTs CHECK VAlvt INCLUDING GLAND SEAL SYSTE l
Tigure 3-3. Typical RCIC Configuration - Piping Sections Subject to Potential Overpressurization 3-5
(
gV 3 -
E G
T A t E S s H
C A a sE i
T T M L T S O S I E S E
C G R HV A O R F T A L P L
AE G D. V E T L HV N G U I P EK A
= F CL TC P P S e A I
P A K
, EM D K V a N EH S
E A .
o EU y1 O C E y b, , - KP PC ME I
T L
LE EG E
R i UH C O s PC '
U WRA S M l io DH M t o , S C NC O i.
w f
P AS I T r
, , H DD E a
ng E F E P S U
^ L O OLU M D I e N N f
n D O CP O o C _iT-
- I SEM I
M T H
T A
C
_- C A I ET R Z Z S T
R I
V LDLA I
R C O U ANA U
- N N g N S VA E S li G G A S S S t I
e E KY R SM E SR E U I
N E
R CTU E I S E
R P
l DU P e.
T 4t DS O S R H CS R e HS S I
T I
I R
B O MC E O S
g v
S P U S UGN T P I T T L S N NL N S E StPE E M V TMG V .
P E 8 UR M 4
R R P H E
PA P l E H O , V O
E V
HWC TOS GN .
L -e M ,v S A E L D I I P r C V CA N I
P s
A SW i
P K n I H C R ALO NO i
- l v
E H UH SP A I
T l
C C E
SMo A t U s L G FPV UO S O L E
R A OLA EH A ,v W N H L L EU O T Y C VF T N R S D
O.
S I
HEEO t
S O P L V M Gt EK E
RO L X T PP P U H HT P
E B
A VV E
AT S F L E S C S I E S U EA I
L V E UM P E K H RAE RC R E ECT EE OL H HE A HH T E T TRR TC CS * =
- AS EE RV ye i
j j1i
NEDC-31339 valve is closed and is well seated. The probability of the injection and suction valves leaking at a rate that would result in preventing the low discharge line pressure and high suction pressure alarms coincident with the cischarge line check valve failing epen is judged not to be significant.
A more detailed discussion of the system's configuration and design data is presented in Appendix A.
3.2 PROBABILITY OF PRESSURE BOUNDARY RUPTURE DURING OVERpRESSURIZATION 1
Based on the ECCS piping configurations defined in Section 3.1 the syster piping, valves and heat exchanger components were evaluated for poten-tial rupture during an overpressurization event.
In evaluating the piping integrity during the overpressurization event, the following failure codes were considered (1) burst due to high hoop stress; (2) rupture due to latent axial defects; and (3) rupture due to latent l
= veld defects at circueferential butt velds. Each of these modes was evaluated as follows: (1) the hoop stress from overpressurization was calculated to a
compare with a conservative value of pipe burst hoop stress which was based on General Electric test data for burst hoop stress and a review of available technical literature; (2) through-vall flaw lengths that the ECCS low pressure system pipes can tolerate during the overpressurization event were determined (the purpose of this evaluation was to demonstrate that these limiting flav lengths are large compared to the flaw lengths that are norcally detected by nercal in-service inspections); and (3) the probability of a duuble-ended pipe break (DEPB) during overpressurization resulting from latent defects at circumferential velds was calculated. The evaluations in (1) and (2) above are deterministic and, therefore, were not directly factored into the proba- l bilistic evaluation perforced for (3). A piping reliability model developed by Lawrence Livermore National Laboratory (LLNL), with appropriate rodifica- i 1
tions incorporated by General Electric for B'a'R applications, was used to 1 calculate the probability of a DEPB at a circueferential veld.
l Two types of deterministic evaluations vere perforced for the valves and 1
RHR heat exchangers, which represent screvhat cocplex structures coepared to l
l 3-7 i
t 1
, ._. _. - __ __. -_)
. '. NEDC-31339 the piping. The first evaluation consisted of examining the body or shell thickness and the hydrotest pressures to demonstrate that the rupture proba-bility of a valve body or a RRE heat exchanger shell vould be less than that of the connected piping. The second evaluation considered the bolted joints.
Potential dynamic loads such as those resulting from earthquakes and safety / relief valve discharges were not included in the analysis because it was concluded that the likelihood of their occurrence simultaneous with an overpressuritation event is extremely small. Other potential dynamic loads during the overpressurization event, such as waterhammer caused by reactor vater filling a partially voided ECCS line downstreas of check valve, were also not included in the scope of this evaluation. The probability of parti-Therefore, sily voided lines is extremely lov due to the "keep-full" systems.
it was concluded that there would be no significant dynamic loads being applied to the ECCS system during an overpressurization event.
3.2.1 Fiping Integrity Evaluation The pipe size, schedule and the caterial information defining the various F4R ECCS and RCIC piping systems surveyed were reviewed. From this informa-tion, the following general conclusions were dravn:
- a. The ECCS piping is of seamless construction and the material is typically SA 106 Cr. B carbon steel.
I
- b. The largest piping diameters in the surveyed core coolics systems ,
range as follows:
Core Spray: 16-inch !
RCIC: 6-inch r HPCI: 16-inch RHR: 24-inch
- c. The Code classification is generally Section III, Class 2 (Reference 2) or ANSI B31.1 (Reference 3). l 3-8 i
- o. ',
The first step in assessing the rupture potential is to calculate the axial and circumferential stresses in these piping systems. Table 3-1 shows the calculated circumferential stresses for typical ECCS pipes when subjected to nominal reactor pressure of 1050 psi. A corrosion allowance of 0.08 inch .
var used. A review of the circumferential stresses in Table 3-1 shows that
}
they range from a low value of 16.3 kai for the 6-inch pipe in the RCIC System to a high value of 34.5 ksi for the 20-inch standard schedule RHR pipe.
The allevable stresses for various service conditions in the ASME Code are expressed as a constant times the Code specified allowable stress, denoted by symbol S. For Class 2 and 3 piping, it is the lesser of five-eighths of the yield stress or one-quarter of the ultimate stress. The allevable stress values, S. given in the ASME Code are essentially identical to those given in the older piping codes such as ANSI B31.1 used in the design of earlier Bk'R 1
plants. The value of S for SA 106 Gr. B material is specified as 15 ksi for te peratures up to 600*T. This is 1/4 of the Code specified minimum ultimate stress of 60 ksi.
The Level D or faulted condition stress limits are relevant in this case, since these lieits, while permitting some gross general deformatten, still assure pressure-retaining espability of piping components. This is consistent with the requitecent for pressure integrity of the ECCS piping during the overpressurization event. Therefore, the calculated stresses were compared to the allevable stresses for the faulted condition.
I The C1sss 2 and 3 pipes are sized such that the hoop stress at design pressure is less than the ASME Code allowable stress, S. Since a peak pres-sure of two times the design pressure is permitted during Level D (f aulted) conditions, the allowable circumferential stress during Level D conditions is 2 S. For SA 106 Gr. B, this allowable stress level is 30 ksi. An examination a of the calculated hoop stresses in Table 3-1 shows that, except for the 20-inch pipe, all other hoop stresses are less than 30 ksi*. Therefore, it is
- Even though the calculated hoop stress of the 20-inch pipe exceeds the Level D allevable, further evaluation of limiting flaw size and burst data (Subsection 3.2.1.2) shows that a sizeable crack length would be required to cause pipe rupture.
i 3-9
. '. NEDC-31339 Table 3-1 CALCULATED HOOP STRESSES DURING OVtRPRESSURIZATION IN REPRESENTATIVE ECCS PIPING SIZES NOMINAL HOOP STRESS
- lok'ER BOUND AT 1050 psi BURST HOOP BURST **
THICKNESS, 7 PIPE SIZE STRESS (ksi) MARCIN (in.) SCHEDULE (in.) PRESSURE 16.3 54.0 3.31' 6 STD 0.28 23.9 54.0 2.26 14 STD 0.375 27.4 54.0 1.97 16 STD 0.375 34.5 54.0 1.56 20 STD 0.375 28.9 54.0 1.87 24 XS 0.500 I
I r
t I
I 1
- Thickness used f or hoop stress calculation is (T - 0.08), where 0.08 inch is the corrosion allovance.
- Ratio of Lower Bound Burst Hoop Stress to Hoop Stress at Reactor Pressure (1050 psi).
550'P code einieue values for SA 1063 5,= 60 ksi S,- 27 ksi 3-10
N EDC-31339 l . .
t concluded that, generally, the hoop stresses in the ECCS piping during the overpressurization event vill be less than the .*SKE Code specified limit for level D conditions.
3.2.1.1 Hoop Stress Burst Margin The safety margin relative to burst type of failure in ECCS piping systes ;
during an overpressurization event is the ratio of the expected hoop stress at burst and the calculated hoop stresses at reactor pressure (Table 3-1). The hoop stresses due to any thermal gradients are not included in this evaluation because such stresses are displacement controlled and, thus, do not directly contribute to burst failure or rupture.
Based on extensive sets of test data, Rodabaugh (Reference 8) noted that for seatless pipes the hoop stress at burst is essentially equal to the ulticate stress of the material. Burst test data reported by General Electric (Reference 9) specifically on seamless 106 Cr. B pipes were reviewed for this case. The pipe dia:eters in these tests ranged from 4 to 12 inches. The l yield and ultimate r.trengths were also determined for each pipe tested.
rigure 3-5 shows the hoop stress at burst as a fraction of the measured ,
ulti= ate stress for that pipe material plotted versus the. ultimate stresa. It is seen from Figure 3-5 that the average burst hoop stress, is equal to approxicately 90% of the ultimate stress. This conclusion is considered independent of the pipe size because the burst hoop stress, rather than the burst pressure, was used in the evaluation. The ASME Code-specified minimum value of ultieate stress for SA 106 Cr. B steel is 60 ksi to a tecperature of 1
600*T. Therefore, the expected value of burst hoop stress is (60 x 0.9) or !
1 54 ksi. Table 3-1 shews that the burst hoop stress margin ranges from 3.31 (6-inch) to 1.56 (20-inch). Even the minimum 1.56 margin is greater than the faulted or Level D safety margin of 1.4 of the ASKE Code.
1 1 3.2.1.2 Limiting Axial riav Lengths in Piping During ECCS Overpressurization 1
4 A qualitative ceasure of the assurance of pressure integrity of ECCS piping during an overpressurization event is the length of the axial flaw (latent defect) that can be tolerated without rupture. Since SA 106 Cr. B 3-!!
N DC-31339 I
)
l i
I
' 3 C
l l .
n .- ;
I e.
E
' i .9 i n e v. -
1 E e..
f . 2 0 v.
- I I
e :
I I _
=* 5 1 f_ l 1 l i I I i l i 1
- w. a.
~ -
o e e ~ o e , m n -
ote :
- - - o o o o ci o e o o s$ s
$t<
nsusers ecow 2 sung $eE5 3-12
NEDC-31339 carbon ateel is expected to behave in essentially a ductile manner in the range of tecperatures expected during overpressurization, either the elastie-plastic fracture mechanics (EPTM) analysis or the limit load approach is appropriate in such an evaluation. In comparing the two methods, the use of EFTM analysis requires information on the saterial stress-strain curve and ,
caterial toughness in the form of a J-Resistance turve, while the only mater- l fe' para:eter required in the limit load approach is the flow stress.
- Reference 4 indicates that both the EP m analysis and the limit load predic-tiets are in excellent agreement with the experimentally detettined instabil-ity pressures of SA 106 Gr. B carbon steel pipes. The temperatures in these tests ranged from 538'T to 675'r, and the average value of the flow stress used was approximately 47 ksi. This empirical value of flov stress is appli-cable to the evaluation of axial flavs. A different value of flow stress ;
^
(2 S or 36 ksi; page 3-19) is used for the evaluation of circumferential flavs. Since the flow stress values are empirical quantities backed out from correspending test data, the use of different values of flow stresses for axial and circurferential flaw evaluations is justified.
I The limit lead approach is also the basir for the recently proposed ASKE Code procedures for the evaluation of axial cracks in both the austenitic (Ft:ference 5) and ferritic piping (Reference 6). Therefore, a limit load i approach, with a conservative value of flow stress of 47 ksi, was used in the following evaluation to determine the largest tolerable axial crack length.
An empirical forcula developed by Eiber, et al (Reference 7) relates the i hoop stress, ho , at failure f r pipes with axial through-vall flaws, to *he l I
flaw paraceters as follows:
f e
=g (3-1) l vhere c = flov stress and, M is a curvature correction f actor given by f
M = [1 + 1.61 1:!4 rt]I (3-2) 1 3-13 9
. - _ . . . _ . . _ - . . _ - , .. . - . y _ . _ , , , - ___,_;.._.-.-__m,,y,_-r-._,- ,e _- _____
vhere: ;
e P
I = axial crack length r = radius of the pipe t = nominal thickness l
An algebraic manipulation of Equations 3-1 and 3-2 yields the following equation for allevable axial crack length:
, y 1/2 (3-3) i- 2.46 rt () -1 l h
Table 3-2 shows the maximum tolerable crack length during the overpres-st.rizatien event for various representative pipe sizes in ECCSs ranging from '
^
2..
inch to $.8 inches. This indicates that large through-vall axial cracks ,
Since a through-vall crack j vould have to be present to cause piping rupture.
the ef such length would likely be detected and repaired, it is concluded that pretability of rupture of ECCS piping from unstable growth of latent axial defects during overpressurization is negligible.
Further evaluation of ECCS. failure during overpressurization considering j latent defects at circumferential velds is presented in Subsection 3.2.1.3. !
1 j
3.2.1.3 Veld Evaluation of Frobability of Pipe Rupture at a Circumferential '
i Butt 8
e j
This section evaluates the probability of failure during overpressuriza- '
tion due to a latent defect in a circumferential veld in the ECCS piping.
Since the ECCS piping is seamless, the most likely locations where a latent defect esy exist would be the circumferential butt velds. T:Asbilistic i
f l
cethodology, developed by Lavrence Livermore National Laboratory (LLNL), was
) !
! used in this evaluation. i l l I
l l
3-14 :
1 I
i
NEDC-31339 Table 3-2 LIMITING AX1Al THROUGH-VALL F1AV LENGTHS IN BVR ECCS PIPING AT REACTOR PRESSURE HOOP STRESS LIMITING TIPE SIZE AT 1050 PSI CRACK LENGTH (in.) SCHEDULE (ksi) (in.)
6 STD 16.3 3.3 14 STD 23.9 3.7 16 STD 27.4 3.2 20 STD 34.5 2.4 24 XS 28.9 5.6 3-15
. o. '
As a part of the effort for the resolution of tJnresolved Safety Issue A-2, "Asyaretric Blowdown loads on Reactor Primary Coolant Systems",
LLZ developed a probabilistic fracture mechanics methodology for the assess- j l
tent of double-ended pipe break (DEPB) probability resulting from both direct {
l a
i and indirect causes (Reference 10). The DEPB probability ' assessment fros l direct causes considers the growth of as-fabricated surface flaws at velded ;
} (
3 jcints, taking into account loads on the piping due to normal operating cenditions snd seismic events. Other factors, such as the capability to i
j detect cracks by nondestructive exasination and the capability to detect pipe leaks, are also modeled. Flavs which become through-vall but do not result in l rupture may produce a detectable leak when the calculated leak rate is above the detection threshold. The ratio of the calculated DEPB probability to the i detectable leak probability is a sensure of the leak-before-break probability. ,
LLNL has developed a computer code (FRAISE) whith incorporates this a
rethodology.
4 Even though the LLNL investigations were limited to pressurized water t
reactor (Pb7) coolant piping, the techniques are sufficiently general for i
adaptation to all light water reactor piping systems. General Electric has codified the FRAISE code for EVE applications and has included a more general limit lead-based failure criterion. Brief descriptions of the LLNL piping
]
j reliability codel and the General Electric modifications are given in 1
1 Appendix B.
1 i
.he pipe rupture probabilities were calculated for a typical girth butt l' l veld in the low pressure ECCS piping segments that would be pressurized during an overpressurization event. It was conservatively assumed that the prob- ,
Further, no ability of existence of a fabrication defect at a veld is.1.0.
i credit was taken for any preservice or inservice inspection.
)
l The axial and bending stresses considered in the evaluation were those due to pressure, weight and thercal expansion. The axial cembrane stress, i
which is essentially due to the reactor pressure of 1050 psi, is given in Table 3-3 for various pipe stres. The bending stress, due to thert:a1 expansion, is icpesed on the subject piping as the system heats up during overpressurization. While an exact value of the weight conbined with the l'
3-16 i
NESC-31339 Table 3-3 AXI AL STRESSES USED IN PROBABILITY EVALUATION Pipe Diameter Stresses (ksi)
Membrane bending *
(in.)
6 7.9 7.0 14 11.7 7.0 16 13.5 7.0 20 17.0 7.0 24 14.5 7.0 "Assured (includes weight + thernal expansion) 3-17
'
- NEDC- 31339 ther-al expansion stress at a particular veld can be calculated from the inforcation provided in piping system stress reports, a representative bound-itt value of 7 ksi was used in this evaluation. The pP.AISE code evaluations need enly nosinal stresses (i.e..vithout any stress intensificatien factors nor: ally used in code coepliance evaluations). The calculated value of weight plus thereal expansion stresses at sete locations in the ECCS piping systems Neverthe-may exceed the assuced representative upper bound value of 7 ksi.
less, it is judEed that the following conservative assumptions still assure a bounding systee rupture probability at circumferential velds:
- a. The thermal expansion stresses are displacement controlled and, thus, are classified as secondary stresses in the ASMI Code. Since only the primary (i.e., load-controlled) stresses such as pressure and weight stresses can cause pipe rupture, the inclusion of therr.a1 expansion stresses in the failure criteria of the PRAISE code is conservative.
- b. In calculating the system rupture probability, seee bounding stress level is assuced at all of the velds in the piping systec. It is judged that the increase in the calculated probability at so=e velds due to higher than 7 ksi bending stress vill be core than offset by the lower calculated probability at a eajority of the systec ve3ds where the stress is less than 7 ksi. In other vords, the calculated systen rupture probability, assuming all of the velds to be stressed at 7 ksi, is expected to bound the calculated probability in which actual bending stress levels are used.
Table 3-3 su::arires the stress magnitudes used for va:ious pipe sizes in the calculation of circu=ferential veld rupture probabilities.
As described in Appendix B, the failure criterion used in the probability evaluation is based on the lirit load approach. The material flow stress is a key paraceter in this apprcach. Based on Reference 6 flov stress was censervatively assured as 2 S where gS is the caterial design stress inten-g sity specified in the ASME Code. Thus, the flev stress value of 36 ksi, based l
4 i
3-18 !
. o. l l
tr e . = , vu '
the en the 5, value for SA 106 Cr. B at reactor l avaluatien.
Table 3-4 shevs the calculated rupture probab. e.<s for each pips size !
j considered. The probabilities range fros 5.4E-9 for the 24-inch pipe to !
9.6!-8 or approximately 1.0E-7 for the 6-inch pipe. It is seen that the !
i rurture probability is highest for the 6-inch pipe, although it has the lovest i axial ee brane stress. This is related to the differet.ces in aspect ratios l 6
4 (half crack length / depth; f) of cricical crack sizes (crack sizes for which i failure ic predicted) as a function of pipe diameter. The aspect ratios of ,
critical cracks in larger diaceter pipes are larger compared to those of the ;
s: aller diameter pipes. Since the rupture probabilities in the LLNL model are rd ated to the inverse of the exponent of 8 (see Equation B-2), the larger :
1 1 diameter pipes have lever calculated rupture probabilities compared to the i 4
B s-aller disteter pipes at the same stress level. Eased on the results shown it. Table 3-4, it can be conservatively concluded that the conditional
- rupture ;
- rebability/ veld for pipes greater than 6 inches in diameter is bounded at ;
1.0!-7 i Due to the asser:ed log nor=al characterization of the probability distri-bution for the crack aspect ratio, the LLNL piping reliability.codel is not l appropriate for velds in pipes staller than 6 inches in diaeeter. On the -!
1 Fasis that the smaller pipes have lover pressure-induced stress for the same j f pressure level, it was judged that the rupture probability at circumferential butt velds in pipes smaller than 6 inches in dia=eter is no greater than that
! fer the 6-inch dia:eter pipe.
3.2.1.4 Conclusions from Piping Integrity Evaluation ;
I 1
I
{
Eased on the preceding evaluations, the following conclusions are dravn: (
i i i I
- i. :
i l 1
i.
- "Condittenal" pretability ceans the probability assuming that system over-i pretsurization has occurred.
3-19 l
. . NEDC-31339 Table 3-4 CALCULATED ECCS RUPTURE PROBABILITIES PER CIRCUP.TERENTIAL VELD DURING OVERPRESSURIZATION BASED ON PRAISE CODE Pipe Site Conditional Probability / Weld (in.)
-8 6 9.8 x 10
~9 14 9.3 x 10 16 6.3 x 10
~0 20 2.0 x 10
~9 24 5.4 x 10 Note: Axial membrane and bending d<resses listed in Table 3-3 vere used in the calculatten.
l l
l i
i 3-20 i
l
- a. The calculated hoop stress burst targin in the ICCS piping fer an overpressurization esent is greater than the ASME Code specified safety cargin f or level D conditions which assure pressure integrity.
- b. Through-vall axial cracks of significant lent:h would have to be present to cause piping rupture. Since such through-vall cracks are likely to be detected and repaired, the probability of rupture of ECCS piping from unstable growth of latent axial defects during the overpressurization event is negligible.
- c. Given that overpressurizatier has occurred, the probability of a rupture or a DEpB at any circumferential veld, in any ECCS systee, is conservatively esticated at 1 x 10' .
3.2.2 Evaluation of Valve Integrity A valve is an essemblage of several subcomponents including a body, stet, disc, bennet, gland, yoke and operater. Therefore, the pressure integrity evaluation of a valve is more complex than pipir.g. A quantitative probabil-istic analysis method for the valves similar to the !.1SL piping reliability codel for piping is not currently available. b'evertheless, several inherent design features of the valves were examined to drav qualitative conclusions that the probability of pressure boundary rupture at a valve during the everpressurization event is expected to be no higher than that for the con-nected piping.
Tigure 3-6 shows the cross-section view of an Anchor / Darling 150 pound-rated II.-inch motor-operated gate valve used on the low pressure side of a typical HPC1 line. The two cost likely f ailure codes by which a large break crea could result during an overpressurization are: (1) rupture of the valve body and (2) f ailure of the body-to-bonnet joint. Each of these modes is evaluated separately below.
3-21
. , NEDC-31339 l
6%
l m, .
Ol i
, V - -- ,
t p - - ,
IL
,,.x -
3 - f 0 l
.. }g
, Q4, . . - .. gg ..
-'I &
8 - I iin ,. H a ly -
,--pa
%e
@J l _
-~
f J -- =
i
+$'.*"
1~
1 8 -v -wa i
- e.s. ;
I i l
! I l
Tigure 3-6. Schenatic of 150-lb Tressure Rated HPCI Valve i
1 3-22 J
)
3.2.2.1 Valve Body Integrity Evaluation t
A review of the data in Appendix A indicates that the valves in the KHR l and the core spray low-pressure systess which would be subjected to over- :
pressurization are rated at 300 pounds. The HPC1/RCIC suction line valves are I rated at 150 pounds. Table 3-5 (obtained from Reference 2) shows the required cir.ieut body thicknesses for the valves of various sites and ratings. It is l
) seen that these minieum specified thicknesses are considerably higher than l these specified for the same nominal diameter standard schedule pipes used f typically on the low design pressure ECCS piping. For example, f or a 14-inch l valve, rated at 150 pounds, the specified minimus valve body thickness in .
l Table 3-5 is 0.42 inches. (The specified minimum wall thickness for the valve i
! ef the same rating shown in Figure 3-6 is $/8 inch or 0.625 inches for a
~
cargin of 1.5.) This also represents a greater than 10: sarsin above the i neninal for that of a 14-inch standard schedule pipe whose thickness is i '
O.375 inches. The larger thicknesses in the valve bodies are intended to i litit deforeation to make valve function properly (e.g. to assure leak tight-ness at the valve seat). Rodabaugh (Reference 8, p.11-1) has observed that because of this, the valve body is, in most cases, rigid to the extent that the pipe section attached to a valve vill yield prior to its being able to (
iepart sufficient forces to cause a pressure boundary failure of the valve. A I reviev of the reported field failure incidents en piping systees also confirts ,
j thfs observation. Thus, the likelihood of valve body rupture during the over- j
< pressurization event is less than the rupture of the connected piping. Addi-f tienally, the valves also benefit from higher allevable stresses. The typical j carben steel caterial specified for valve bodies is SA216 WCE which has a j l
i cinteue room teeperature ultimate strength of 70 kai compared to 60 kai for ;
1 1 the carbon steel SA106 B used in piping.
I Another feature in the valve design that provides additional assurance of pressure integrity during the everpressurization is high rating of hydrotest i pressures. Table 3-6 (obtained from Reference 2) shows the specified hydro-test pressures for typical valves. Table 3-6 data indicate that 150-pound I carten steel valves are subjected to hydrotest pressures of 425 psi. vhile j 300-peund valves are tested to withstand pressures of 1125 psi. Thus, 300-I i !
I
- 3-23 .
4 l l
l i
l !
l Table 3-5 s VA1VE BODY MINIMLH THICKNESS (Table NC-3511-1 of Ref. 2) t Inside Minimus Wall Thickness.t ,. in.
Diareter _
- d. Primary Frissure Rating pr
. Ib. f g 2500 150 300 400 6 00 900 1500 l in. I
.10 .10 .10 .10 .10 ;
0.1 .10 .10 .22
.10 .10 .10 .10 .10 i 0.2 .10 .18 i 0.3 .10 .10 'O .10 .!2 .22
.43 .13 .16 .16 .23 l 0.4 .10 .10 .27
.10 .16 .16 .19 .19 0.5 .10 .30
.12 .16 .16 .21 .21 0.6 .10 .22 .33 0.7 .11 .14 .16 .16 .22
.17 .17 .24 .24 .36 ;
0.! .13 16
.18 .26 .26 40 C* .15 .37 .18
.28 .28 44
.16 .19 .19 .19
.25 .25 46 46 .79 r
.?6 .25 1.14
.28 .31 .31 42 .66 3 1.47
.31 .38 .38 .51 .83 i 4
44 44 .63 1.02 1.81 l 5 .36
?P 44 .50 .74 1.21 2.15 !
6 ...
2.51
.50 .57 .83 1.41 7 .30 ..
2.03 5
.56 .63 .93 1.59 E .31 ...
3.17 l 47 .63 .70 1.03 1.76 9 .33 3.51
.50 .69 .77 1.13 1.54 10 .34 3.85 ;
.36 .53 .72 .85 1.24 2.12 11 4.19
.38 .36 .75 .92 1.35 2.31 :
12 4.52 !
40 .61 .81 .97 1.46 2.50 13 4.86 1 42 .65 .84 1.03 1.56 2.69 14
.68 .88 1.11 1.67 2.88 5.20 l 15 43 '
.71 .91 1.18 1.77 3.06 5.54 16 45 75 .94 1.25 1.86 3.24 5.88 17 46
.78 1.00 1.31 1.96, 3.42 6.22 18 48
.81 1.07 1.39 2,07 -
3.61 6.55 19 .50 6.89
.51 .84 1.10 1.46 2.17 3.79 20 7.23
.53 .88 1.13 1.53 2.28 3.97 21 7.57
.54 .91 1.17 1.59 2.38 4.15 22 7.91
.56 .94 1.20 1.66 2.48 4.33 23 8.25
.57 .97 1.24 1.72 2.59 4.51
'24 4.69 8.59
.59 1.00 1.28 1,79 2.69 25 8,92
.61 1.04 1.32 1.85 2.79 4.87 26 9.26
.62 1.07 1.36 1.91 2.89 5.05 27 9.60
.64 1.10 1.39 1.98 2.99 5.24 28 9.94
.66 1.14 1.43 2.04 3.09 5.42 29 10.28
.67 1.17 1.47 2.11 3.19 5.60 30 3424
, , NEDC-31339 Table 3-5 (Continued)
VALVE BODY MINIMLH THICKNESS (Table NC ;311-1 of Ref. 2)
Inside Diameter Minimue Wall Thickness,t,, in.
d, Primary Pressure Rating pr . Ib in. 150 300 400 600 900 1500 2500 31 .69 1.20 1.51 2.17 3.30 5.78 10.62 32 .71 1.23 1.54 2.23 3.40 5.96 10.95 33 .72 1.27 1.58 2.30 3.50 6.14 11.29 34 .74 1.30 1.62 2.36 3.60 6.32 11.63 15 .75 1.33 1.65 2.43 3.70 6.50 11.97 36 .77 1.37 1.69 2.49 3.80 6.68 12.31 37 .79 1.40 1.73 2.55 3.90 6.87 12.65 36 .80 1.43 1.77 2.62 4.01 7.05 12.98 39 .82 1.47 1.81 2.68 4.11 7.23 13.32 40 .84 1.50 1.84 2.75 4.21 7.41 13.66 41 .85 1.53 1.88 2.81 4.31 7.59 14.00 42 .87 1.56 1.92 2.88 4.41 7.77 14.34 43 .88 1.60 1.96 2.94 4.51 7.95 14.68 44 .90 1.63 1.99 3.00 4.61 8.13 15.01 45 .92 1.66 2.03 3.07 4.72 8.32 15.35 46 .93 1.70 2.07 3.13 4.82 8.50 15.69 47 .95 1.73 2.11 3.20 4.92 8.68 16.03 46 .97 1.76 2.14 3.26 5.02 8.86 16.37 49 .98 1.80 2.18 3.32 5.12 9.04 16.71 50 1.00 1.83 2.22 3.39 5.22 9.22 17.04 J
6 3-25
NEDC- 31339 Table 3-6 HYOROSTATIC SHELL TEST PRESSURES (Table NC-3512(c)-2 of Ref. 2)
(All Pressures are in Founds per S;'.are Inch Gage-psig)
MA1 TRIAL Territic Steel Austenitic Steel Carbon Types Steel Carbon (Lev Carbon 1 Cr- 1( Cr- 2% Cr-347 336 310 304L 316L Class Steel Terep) Holy b Mo \N 1 Me 321 304 425 425 425 425 425 425 425 150 425 425 425 425 425 425 1125 1125 1125 1125 1125 1000 1000 300 1125 975 1050 1125 1125 1125 1500 1500 1500 1500 1500 1325 1325 400 1500 1275 1400 1500 1500 1500 2250 2250 2250 2250 2250 2250 2000 2000 600 2250 1925 2075 2250 2250 3375 3375 3375 3375 3375 3375 3000 3000 900 3375 2900 3125 3375 3375 5625 5625 5625 5625 5625 5625 5025 5025 1503 5625 4825 5200 5625 5625 9375 9375 9375 9375 9375 9375 8350 8350 2500 9375 8025 8675 9375 9375 3-26
l i
pound valves have already been hydrotested at a pressure greater than that !
expected during the overpressurization event (i.e., 1050 psi).
Troir the preceding discussion, three conclusions are drawn. First, the likelihood of rupture of ISO-pound valve bodies during an overpressurization event appears no greater than that of the corresponding diameter standard schedule pipe. Second, for the same stresses, valves have a higher design cargin due to higher allowable stresses compared to piping. Third, because of high hydrotest pressures, the rupture of 300-pound (or larger) valve bodies during overpressurization also appears highly unlikely.
3.2.2.2 Body-to-Bonnet / Cover Bolted Joint Evaluation Another part of the valve whose failure could lead to a breach of the pressure iaundary is the bolted joint between the body and bonnet or cover (Ffgure 3-6).
As noted in the preceding subsection, the 300-pound pressure rated valves are required to be hydrotested at 1125 psi. Therefore, only the bolted joints in the ISO-pound pressure rated valves were evaluated. This included valves in the 6, 14 and 16-inch sizes. The body-to-bonnet joint in the gate valve shown in Figure 3-6 was also determined to be the bounding case and therefore was used in this evaluation.
The bonnet in Figure 3-6 is attached to the valve body by sixteen 3/4- ,
inch 10NC-2 studs on an 18-1/8 inch bolt circle diameter. Appendix C describes the theoretical relationship between the bolt stress, pre-load and i the p 'ssure loading. Based on Equation C-1, the pre-load stress in the studs '
was esticated as 45,000//0.75 or z52,000 psi. The average stress in the l studs, due to internal pressure of 1050 psi and no pre-load stress, was calcu-lated as :51,000 psi, k' hen the effect of pre-load is taken into account (using Equation C-2), the bolt stress with 1050 psi internal pressure is calculated as 57,700 psi. This represents a small increase (approximately 11ll) from the stress experienced by the studs under pre-load alone. This confirms that, in cost flanged joints, the major stress applied to the studs or bolts is that applied in tightening the nuts. It follows that, if a bolt 3-17
,,-- )
NEDC-31339 er stud did not fail during tighteninE, then it is not likely to fail during service.
If the stress in the studs due to pre-load is less than that estimated by Equation C-1, the calculated stress during the pressurization to 1050 psi vill be even lover. Thus, the average stress in the studs during an overpressuri-zation event is estimated to range from 51,000 psi to 57,700 psi. Since the ASME Code implied minimum yield strength at 550'T for the stud material (SA 193 B7) is a87,000 psi, the minimue cargin for the calculated stress in the studs during an overpressurization event censured against the ASHI Code is 1.5. (The ASME Code implied factor of safety is 1.4 for Level D conditions in which the pressure integrity of a component is the only concern.)
The inherent structural redundancy of the bolted joint provides addi-tienal assurance that, during an overpressurization event, loss of coolant veuld more likely result free valve leakage than rupture, which is likeJy to be detected long before valve integrity would be co=procised. Results of finite ele =ent analysis, on a steple bolted joint given in Reference !!, are discussed in Appendix C. Such analysis clearly illustrates the load shedding ar.d redistribution characteristics when complete degradation of one or core adjacent studs in the joint is assumed. Reference !! shows that the stress increase in the stud next to the failed studs.is small. It should, however, be noted that a review of General Electric service experience data base on BWR
}
pressure beundary caterials has indicated no reported incidents of degradation l
in the lov alloy steel (SA 193 B7) bolting used in ECCS piping system valves and heat exchangers. This was not surprising, since most of the factors identified in Keterence 11 (i.e., the presence of borated water, stress cor-resion cracking and fatigue) are not likely to be associated vith the typical operating conditions in the parts of EVR ECCSs being considered in this evaluation. l The preceding discussions lead to the qualitative conclusion that the yielding of the bolted joints in the ECCS valves on the icv design pressure side during an overpressurization event are more likely to result in a leak rather than a gross rupture.
3-28
'
- NEDC-31339 3.2.2.3 Conclusions froc Valve Integrity Evaluation Eased on the evaluations presented in the preceding subsections, it is concluded that:
- a. The likelihood of rupture of the body of a ISO-pound pressure-rated valve during an overpressurization event is less than that of the corresponding diameter standard schedule pipe. In the case of valves rated 300 pounds or greater, the prescribed hydrotest pres-sure of 1125 psi assures that its probability of rupture during an overpressurization event is negligible.
- b. The likelihood of gross rupture at bolted joints in the low pressure side ECCS valves is extremely small. Leakage of fluid through the bolted joints is the more likely consequence during an overpressuri-zation event. The potential effects of Icakage is discussed in Section 3.2.5.
- c. The overall rupture prooability of a lov pressure side ECCS valve was judged to be no greater than that at the circumferential butt veld between the valve and the connecting pipe.
l l
3.2.3 Evaluation of Heat Exchanger Integrity Several RHR heat exchanger designs were reviewed with the most limiting one selected for this evaluation. The shell inside diameters of the heat exchanEer designs reviewed ranged from 40 inches to 63 inches. The typical design pressure for the RHR heat exchangers was 450 psi. The 63-inch inside
~
diaceter design was found to be most limiting and Figure 3-7 shows the outline of this heat exchanger. During RHR system operation, reactor water enters the heat exchanger at opening A and exits at B. The shell tube sheet and the channel are connected together through a flanged joint. The process cooling water from the channel side circulates through the tube bundle situated inside. The three parts of the heat exchanger that are stressed during the RHE overpressurization event are: (1) the shell; (2) shell-to-tube-sheet 3-29
1
-.S...-.
.. :: ".~* :.' . +
r.*---
p- m.... .q a i,. D b f,-, a .\,sl , '
t x
u
. . .i
\
."r f
A_A Ml J')p.;fsp.
- i. (?;,T;.. 2 g,
-"N7 C ;"= --. '
'~ T C;_ 2 C2 t u.-..' ;= 4 g,MIM[
.1w.2, v-- s.., .'
/ \ 1O.
f~
i 7 _ ....
e
/ -
I
. o ] e } o_j -
. = = .q . .tt..
- . , f - .. .:.~.r.: u .2 }
I k' ..
} .e py N : o .i
%. H , , .
/ 1
.x -ww 4 i
_ .4
/
l \ so:/ k -s
.4 ' ' W ,
- )
' .' ~I/ . ../ ... t.
i I:sc :
e.
. . .J.b.
B '"2 -I
\
= . ..
i f;,,:g.:" % .,
.g N'"*I"i. / ..-/;' :i.:a m u .:
, y
\ * )
y *
', ' s/- n-4 0
- rI
>-4 i
/ 8 1 -
=s j 1
\
l o\ i t js '
J V .. z .
3 ..
- 's c
_. l
, i . _./
.x
% .'P"'1.7
-a.a ..- - @l
_--=.','*m..m.
1 $t:?.t t i
Lr. . . . .m - , . . -
-...J..-- =
(.y. r ;4' . q 5 r, . '. . .e w i. ..6 "* .J'"A$H feri rvue Tigure 3-7. Schematic of 63-Inch I.D. RER Heat Exchanger 3-30
-5 O Why or why not?
- 14. What Steps are Planned to Prevent Recurrence?
Prepared by: Date:
TES/cic/04188809 f
l 1
i h h
l NEDC-31339 flanged joint; and (3) the tubes. Each of these was evaluated for large break potential during overpressurization.
3.2.3.1 Heat Exchanger Shell Evaluation The heat exchanger shell is a cylindrical structure with a top elliptical head. The hoop stress in the cylindrical portion is expected to be governing.
The nominal shell thickness for the example case is 1.0 inch. For an internal pressure of 1050 psi, assuming a standard corrosion allevance of 0.08 in., the hoop stress was calculated as 36 ksi. The material specified for the heat exchanger shell is SA516 Gr. 70. The faulted condition or Level D allevable pressure hoop stress 2 S for this material is approximately the same value as the calculated hoop stress. Furthermore, it should be noted that the calcu-lated hoop strest cf 36 ksi is for the limiting case. For most cf the other FFT heat exchangers, the calculated hoop stress during overpressurization is less than 30 ksi. The primary cembrane stress in the other regions of the shell is expected to be less than that calculated for the cylindrical. region.
Overall, the calculated hoop stresses, relative to Level D a13ovable values, fall essentially in the same range as those for the piping (Table 3-1). Thus, it can be qualitatively concluded that-the burst failure.of the vorst case heat exchanger shell during an overpressurization event is no more likely than failure of the connecting piping (Section 3.2.1).
3.2.3.2 Shell-te-Tube-Sheet Bolted Joint Evaluation The heat exchanger shell, tube sheet and channel are held together by a flanged joint. A review of such joints in various RER heat exchanger designs indicated that the flanged joint in the 63-inch shell I.D. heat exchanger, censidered in Subsection 3.2.3.1, was liciting and, therefore, was selected for evaluation. The sealing of this flange joint is assured by 68 1-3/8-inch diameter bolts equally spaced in a circular pattern. The same procedure used for valve body-to-bonnet joint evaluation in Subsection 3.2.2.2 was used to calculate the bolt stresses in the heat exchanger flange joint. Based on Equation C-1, the minieu pre-load stress in the bolts was esti=ated as 45,000/>'l.375 or m35 ksi. The average stress in the bolts, due to internal pressure of 1050 psi and assuming no pre-load stress, was calculated as 3-31
)
NEDC-31339 5 psi. With the bolt stresses expected during overpressurization being this gres:er than the pre-load stresses, some leakage of reactor coolant at In fact, in the Vercont Yankee j eir.t is expected during overpressurization.
incident (Reference 1) the RHR Syste= overpressurization did result in leakage of steae and water mixture from the heat exchanger tube sheet-to-shell flanFe area. While some leakage may occur at this joint, the overall integrity of this joint is assured as demonstrated by comparing overpressurization stresses with the bolt yield stress. The ASMI Code minieum specified yield stress at 550*F for bolt eaterial SA 197 E7 is sS7 kai. Thus, the margin against flange bolt failure is 87/54 or 1.61.
It is concluded, therefore, that the tube-sheet-to-shell flange joint is ccre likely to leak rather than fail during an overpressurization event.
3.2.3.3 Tube Integrity Evaluation Heat exchanger tubes are subjected to reactor pressure on the outside surface during an overpressurization event. Therefore, the principal failure code would be tube buckling due to the external pressure. The tube materials range fr:: type 304L stainless eteel to copper-nickel alloy. Because of the lover yield strength, 304L stainless steel tubes were evaluated as the limit-ing material. The evaluated tubes have a 1-inch 0.D. with a thickness of 0.0'9 in. (corresponding to 18 SWG). The collapse pressure of these tubes can be esticated using the procedures given in Paragraph ND-3100 of the ASME Code (Reference 2).
Figure 3-8 (Figure ND-3133.8-1 of the ASMI Code) can be used to graphi-cally detercine the collapse pressure of heat exchanger tubing. Figure 3-8 illustrates design pressure as a function of design stress for various heat exchanger tube thickness-to-diameter (T/D ) ratios. The design pressure has a built-in factor of safety of 3.0. Thus, the expected collapse pressure vill be three times the design pressure determined from Figure 3-8. The design stress for 304L stainless steel at reactor temperature is 14 ksi. For a typical T/d of a0.05, the resulting design pressure from Figure 3-8 is 550 psi. Therefore, the expected collapse pressure (i.e., 3 x design 3-32
)
'O # i i i ; iii i.'/ 1 i e i i e i i e i i/i 1 5# i e s i /ii 'f i
I l l l l I/ I l I [' '
',o I[
~
i l l I I Il '}
l l Sb I lN l
\ 2 7 $ l [J f T P il 0 l' l
' t I /-
/ /!/\f\/WV
/
/
/
/
/ * /
/ /
l
- / / Vt lllo'l('L Q
- / t / OY s Jf i Y_Y'/
f l
- i /
1 / / / \
// Vi n+f / \ Y s : A i / / / / Y VI /' f : Y '1 / !
! ((..
!( _ // / / //VI /J$# M** Ifl I i 2 1
}
~::~ i/
/ / / /// /
L//)./,
/ /// J/l l / '/ V'>}
l
~
/ /~/ l_/ /// / /- /// '
... ! /
~T / /// /// /: $C////
._!s' j
'~'
/ ///
,s j
////
j -j ,7 7
/V////I 7 ,/ n o$ 9 0 7 ,' ,_
/ ,' / / ,' / / ,' 1' / / z' / l
/ /_ / / n0Y .sf s'
/~s' /// / s l t / //// '
s' s' s's's'r' 10 s'/ s' l if / / f__/ / // / 4 0 70 16
/
- 1
,e,- ! / / /^ ~~"/~" f. / // / / /
,/ // f / / // / / /_ c 0 ' **
_,, f
/ ! \
f// / / / / / / /_, / _
00 ;*/
j
// / /// /// / .
1 P/ / _ _ __ __
/ / I
.. 7 /_ _ _ . . . . _ .
' :1 I l !' i .
l
, [' / /, l . ! il: ; i i >
l 10 15 'O 20 40 5.0 10 0 00 0
- etige. stren ksi
- s. .e.:e: :n : * :n .5, , :n.,e stini ,3, se...ns vue...,
Figure 3-8. Chart-Tube k'all Thickness Versus External Pressure (Figure ND-3133.8-1 of Reference 2) 3-33
en-NEDC-31339 pressure) is 1650 psi. yielding a margin against collapse failure during overpressurization of 1.57.
3.2.3.4 Conclusions from Heat Exchanger Integrity Evaluation The preceding evaluations lead to the following qualitative conclusions:
- a. The~1ikelihood of rupture of a RHR heat exchanger shell during an overpressurization event is of the same order of maEnitude as the connected piping (discussed in Section 3.2.1).
- b. The tube sheet-to-shell bolted joint, which would be stressed during an overpressurization event, is likely to leak rather than experi- ,
ence a gross rupture.
- c. Heat exchanger tubes have an inherent safety margin of three against collapse during an overpressurization event.
- d. The overall rupture prob.)bility of an RER heat exchanger was judged to be no greater than that at the circumferential butt veld in the connecting RHR piping.
3.2.4 _Overall System Rupture Probability Evaluation The overall ECCS low pressure piping system rupture probability during an l l
overpressurization event is equal to the sum of the rupture probabilities of In the piping and the piping components such as valves and heat exchangers.
the case of piping, it was judged that the rupture probability at the circum-ferential butt velds was the dominant contributor. Thus, the rupture proba-bility in the piping was defined as the product of the per-circumferential butt veld rupture probability (conservatively estimated as 1.0E-7 from Sub-section 3.2.1.3) and the number of circumferential butt velds in the system.
The rupture probabilities for valves and heat exchangers was approximated as equal to that at a circumferential butt veld as discussed in Subsections 3.2.2.3 and 3.2.2.4 (e.g., each valve and heat exchanger counts as an addi-tional circumferential butt veld in the system).
3-34 l
NEDC-31339 The number of velds in the portions of ECCS piping affected by over-pressurization depends upon the syste: configuration. In a limited plant survey, the number of welds in an RhR systet were determined as 112 for pipe sizes three inches and larger. Similarly, the number of valves 3/4 inches and larger in a typical RHR system was determined to be 91. On this basis, the nu ber of "equivalent circu=ferential velds" was conservatively assumed as 300. Thus, the ECCS low pressure system piping rupture probability is esti-cated as (300)x(1.0E-7) or 3.0E-5. This expected probability is at least two orders of eagnitude lower than the range of 1.0E-2 to 1.0E-3 stated in the AE0D case study.
3.2.5 Probable Results of Overpressurization As indicated by the above analysis discussion, failure of syste: cocpo-nents due to overpressurization is not expected to occur.
The results of such an overpressurization event would most likely be lir.ited to:
(1) Discharge of fluid and two-phase fluid fro syste=s relief valves.
(2) Leakage of fluid and tuc-phase fluid froe bolted joints and possi-bly failure of some gasket (s).
(3)* DischarEe of fluid and two-phase fluid fro: small undetected cracks l in velds. l
- It is judged that only small cracks in velds cay fail to be undetected during nereal plant surveillance and that the' number of such small cracks would be few. The effective flow area of these cracks would be small. For example, the effective flow area of such cracks vould be cuch less than the two-square-inch effective flow area of a large 25-inch 1cng crack of critical size in a 24-inch diaceter pipe veld. (A crack of critical size is the length of crack that can be present without resulting in a guillotine failure of the pipe during an overpressurization event.)
3-35 i
~~ - -- . __ _ _ _ _
- NEDC-31339 (4) Activation of line high pressure' alarms.
(5) Activation of smoke alares due to oxidation of paint on piping and components.
(6) Activation of area high temperature alarms due to steam that results from the~ discharge of two-phase fluid.
The volume of fluid discharged from bolted joints would be largely lietted by the minimal flexure of the joint and bolts. Following termination of the overpressurization condition, the joint would likely reseat and stop leaking unless gasket rupture had occurred. If gasket failure occurs, additiona) fluid discharge could result and may continue after termination of the overpressurization until the system sufficien6Ay depressurires.
The volume discharged fro: relief valves, bolted joints, undetected cracks, and gaskets if they fail, is judged to be small compared to the volute ef associated equipeent rooms such that the leakage would not be expected to result in substantia) flooding consequences for overpressuritation events of l 1
durations similar to those reported in Reference 1.
The most significant consequences to be expected would be the additional some hazard to plant personnel from discharged fluid and the potential that equipment eay be rendered inoperable due to spray from leaking co=ponents.
The equipment that may be rendered inoperable due to spray effects is expected to be limited to equipeent in the same division as the overpressurized systec l because of equipment divisional separation.
Activation of line high pressure alares, smoke alares, high area tecper-ature alares and personnel observation of smoke, spray and steam discharge free relief valves and' leaking co:ponents would likely result in decreasing the duration of,the overpressurization event by alerting operators to the overpressurization condition. These pheno =ena resulted in alerting operators to the overpressuritation condition in the events reported in Reference 1.
3-36
NEDC-31339 There is a high ifkelihood that operators vould be able to isolate the discharge free relief valves and leaking components by closing the systen it.jectier. valves. The HPCI and RCIC Systees' injection valves are generally located outside the room in which the systems' low design pressure components are located. This further itproves the likelihood of being able to isolate discharges and leaks from HPCI and RCIC Systems because of the reduced potential environmental eff ects of the discharges and leaks on the system injection / isolation valves. The likelihood of operators being able to isolate such discharges and leaks is, to some degree, demonstrated-by the fact that the operators vere able to isolate the discharges and leaks reported in Reference 1.
3.2 PROSA31LITY OF AN INTERPACING LOCA The expected frequency of a Ek'E interfacing LOCA involving the reactor coolant syste= (RCS)'and the ECCS and RCIC system is detercined as follows:
LOCA Press
- Rupture where:
I P = r a y an n erfacing LOCA between the RCS LOCA and ECCS.
P = Probability of overpressurizing the low pressure ECCS Press and RCIC system piping.
P = Conditional probability of a rupture in the ECCS R pt re piping given an overpressurized conditior.- .
Substituting the values of the probability of overpressurization and the conditional probability of ECCS rupture, the probability of an interfacing LOCA is, therefore, determined to be:
a
- 3-37
NEDC-313?9 l
l P = (1.0E-2)* x (3.0E-5)
LOCA P = 3.0E-7 per reactor-year LOCA The value of 3.0E-7 per reactor-year for the expected frequency of an interf acing LOCA is acceptably lov, nearly three (3) orders of magnitude lover than the expected frequency of a large break LOCA described in the k'/.SH-1400 Reactor Safety Study. It is, therefore, concluded that an ECCS overpressur-ization event poses no significant threat to the saf ety of the BL'E.
I
]
l
- Tor purposes of calculation, the frequency However, of system overpressurization is it is believed that this assured to be that stated in 7.eference 1.
event frequency is overly conservative based on a continued operating data ,
= base witheur additional occurrences since the issuance of Eeference 1 and l the increased industry awareness of these occurrences and the need to reduce !
cvent frequency.
3.-38 T-s-r -+wg - .7 p.
NEDC-31339 4 REFERENCES
- 1. Lac, P., "Overpressurization of Emergency Core Cooling System in Boiling ,
Vater Reactors", Report AEOD/C502, September 1985.
- 2. ASME Boiler and Pressure Vessel Code,Section III, Division 1, Nuclear Power Plant components, American Society of Mechanical Engineers.
4 V11kowski, et al, "Su= mary of NRC Phase I Degraded Piping Program -
Instability Analyses and Review of Experimental Programs to Evaluate J/T Instability Predictions", ASME PVP - Vol. 95, 1984
- 5. S. Ranganath, H.S. Mehta and D.M. Norris, "Structural Evaluation of Flavs in Power Plant tiping", ASKE PVP - Vcl. 95, 1984 I
- 6. A 2ahoor. H.S. Mehta S. Yukawa, R.M. Gamble and S. Ranganath. "Plav Evaluation Procedures and Standards for Perritic Piping", EPRI Draft Final Report on Research Project RP 175'/-51, July 1985.
- 7. "Investigation of the Initiation and Extent of Ductile Pipe Rupture",
R.J. Eiber, et al, Battelle Columbus Laboratories, Report BMI-1908, June 1971.
- 8. E.C. Rodebaugh, et al, "Survey Report on Structural Tesign of Piping Systecs and Components", TID-25553, December 1970, National Technical
. Inf ormation Service.
- 9. M.B. Reynolds, "Failure Behavior in ASTM A106B Pipes Containing Axial Through-Wall Flaws", General Electric Report GEAP-5620 April 1968.
- 10. "Prebability of Pipe Fracture in the Primary Coolant Loop of a PWR Plant". NUREG/CR-2189, Vol. I through 9.
4-1
- 11. !;1ckel, R.T., R.C. Cipolla and E.A. Merrick, "The Use of Leak-Before-break Criteria and Assessment of Margins in Addressing Closure Integrity
!ssues", SMIRT-6, Paper D6/3, August 19, 1965.
1 l
4-2
S 8 APPEhTIX A B'a PLANT DATA SL?O'ARY l
l i
l l
l l
l
- NEDC-3132C APPEhTIX A PLANT DATA
SUMMARY
The data received f rom _ nineteen (19) Bk'R plants that were utilized in the study to evaluate the probability of an interfacing 1.0CA are succarized in this Appendix.
Table A-1 lists the utilities, plants, BVR type, and containment type for the plants considered in the study. I Table A-2 summarizes the system design pressure, hydro test pressure, largest pipe diameter and the largest diameter pipe radius (r) to pipe thick-ness (t) ratio for the piping sections identified in Paragraph 3.1 of this report as being subject to potential overpressurization. ,
The eaxicum pipe hoop stress due to pressurization is equal to the r/t ratic reltiplied by the pressure to which the piping is exposed.
Table A-2 and Pigures A-1 through A-31 sum =arize the reactor vessel isolation valve configt.tation for each system. The figures do not differenti-ate between gate and globe valves (i.e., the same symbol is used for both).
The piping and valve configuration upstream of the reactor pressure vessel (RPV) isolation valves are plant-specific and vary in detail. However, the piping sections subject to potential overpressurization are, in all cases, the same as illustrated in Figures 3-1, 3-2 and 3-3. RHR RPV isolation valves that are utilized for the low pressure coolant injection (LPCI) function are presented in these figures. Other RHR RPV isolation valves that are utilized for the reactor shutdown cooling function only (such as thos'e on the RER suction line from the RPV, RPV head spray and shutdown return to the recircu-lation system on or feedvater line) are not presented because these valves are interlocked to prevent their opening when reactor pressure is above the design pressure of the connected lov design pressure piping, they do not receive any auto:atic opening signal and these valves are not tested for operability during norcal power operation.
A-1
~
NEDC-31339 All piping caterial is sealess A 106 Grade B or A 333 Grade 1 or 6 carben steel.
All valve body caterial is A 216 VCB carbon steel.
All RHR, LPCI/CS, CS and LPCS valves downstream of the pump discharge check valves, with the exception of the RPV isolation valves, have a minimum pressure rating of 300 lbs. All HPCI and RCIC system valves upstream of the cain pump suction inlet have a minimum pressure rating of 150 lbs.
A-2
Table A-1 UTILITY, PLANT NAME, BWR AND CONTAINMEh"I TYPE CONTAINMENT TYPE BWR OWNER PLANT BVR (MARK I, 11 OR III)-
Besten Edison Company Pilgrim 3 I 4 Ccrclina Power & Light Co. Brunswick 162 4 I Co==onwealth Edison Co. Dresden 2&3 3 I Quad Cities 161 3 I LaSalle 162 5 II .
Hatch 162
~
Georgia Power Co. 4 I General Public Utilities Oyster Creek 2 I huclear Gulf States Utilities Kiser Bend 1 6 III Illinois Power Company Clinton 1 6 III 4
leva Electric Light 6 Power Co. Duane Arnold / I Nebraska Public Power District Coeper 4 I hiagara Mohawk Power Corp. Nine Mile Point 1 2 1 ,
Nine Mile Point 2 5 II ;
Northern States Power Co. Monticello 3 I l
Pennsylvania Power & Light Co. Susquehanna 162 4 II Ph11adelphia Electric Co. Peach Bottom 263 4 I Limerick 162 4.5 II Public Service Electric & Hope Creek 1 4.5 I Cas Co.
Tennessee Valley Authority Browns Perry 1,263 4 I Washington Public Power Hanford 2 5 11 Supply System J
A-3 6
s._ w -- _---.,4. ,. , ~ w,, . , . -
y . --., - - . . - - - . ,y._
I NEDC-31339 Table A-2 B*a0G PARTICIPATING l'TILITY PLANT DATA ISOLATION LARGEST VALVE DESIGN HYDRO PIPE LARGEST CONFIG. PRESS. PRESS. DIA. PIPE PLANT SYSTDi (Fig.)* (psig) (psig) (in.) (r/ )
Browns Perry 1,2&3 RHR A-1 450 24 27.6 CS - A-1 500 14 22.7 HPCI A-12 150 16 26.1 RCIC A-12 150 6 15.6 Erunswick 1 & 2 RHR A-1 460 690 24 23.9 CS A-1 460 690 12 24.5 HPCI A-13 150 16 26.1 RCIC A-14 150 6 15.6 Clinton RHR Later 500 750 18 LPCS Later 600 900 14 RCIC Later 75 113 6 Cooper RHR A-1 450 450 24 23.9 CS A-1 500 500 12 20.6 HPCI A-15 150 16 26.1 RCIC A-15 150 6 15.6 Dresden 263 LPCI A-1 350 18 24.1 CS A-1 350 12 20.6 HPCI A-16 150 16 26.1 Duane Arnold RHR A-1 375 20 32.9 CS A-1 355 10 17.9 HPCI A-12 125 14 22.7 RCIC A-12 125 6 15.6 hanford 2 RHR A-2 500 625 20 22.8 LPCS A-2 470 588 16 26.1 RCIC A-17 125 110 6 15.6 Hatch 1 6 2 RHR 375/400 469' 24 23.9 CS HPCI 125/140 156 16 26.1 RCIC 100/125 156 6 15.6
/140 Kope Creek 1 RHR A-2 410 615 20 22.E CS A-9/A-3 500 750 14 23.9 HPCI A-18 105 158 16 26.1 RCIC A-19 105 158 6 15.6 1 1
- !dentifies the appropriate figure nu::,ber conttined in this Appendix.
A-4
- ' NEDC-31339 Table A-2 (Continued)
B'n'ROG PARTICIPATING UTILITY PLANT DATA ISOLATION LARGEST VALVE DESIGN HYDRO PIPE LARGEST CONFIG. PRESS. PRESS. DIA. PIPE SYSTEM (Fig.)* (psig) (psig) (in.) (r/t)
PLANT LaSalle 1 6 2 RHR A-4 500 750 18 17.7 LPCS A-4 550 825 16 18.0 RCIC A-31 100 150 8 16.8 ,
Literick 1 6 2 RHR A-4 420 630 20 32.9 CS A-9/A-5 500 750 14 22.7 HPCI A-20 125 182 16 26.1 RCIC A-21 125 182 6 15.6 Monticello RHR ~~A-6 275 413 16 26.1 CS A-6 303 455 10 17.9 HPCI A-22 23/30 90 14 22.7 RCIC A-23 50 75 6 15.6 Sine . vile Point 1 CS A-7 470 12 20.6 Oyster Creek CS A-8 300 10 17.9 Feach Sotter 2 6 3 RHR A-9 450 675 24 27.6 CS A-9 450 675 14 22.7 HPCI A-24 150 225 16 26.1 RCIC A-24 150 225 6 15.6 Pilgrim RHR A-10 500 $35 18 29.5 CS A-Il 300 550 10 17.9 HPCI A-25 80/60 110 16 26.1 RCIC A-25 80 6 15.6 A-1 400 510 18 24.1 Quad Cities 1 6 2 RHR 20.6 CS A-1 500 594 12 ,
HPCI A-27 150 188 16 26.1 )
RCIC A-26 150 6 15.6 1 River Bend RHR A-2 500 750 18 29.5 LPCS A-2 600 900 14 22.7 RCIC A-28 90 135 6 15.6 I
18.7 i
Susquehanna 1 6 2 RHR A-9 450 563-670 24 CS A-9 500 635-660 14 22.7 HPCI A-29 100 165 16 26.1 i RCIC A-30 150 200 6 15.6 l l
A-5 l
o . NEDC-31339 AO MO MO RE ACTOR PRESSURE FIGL*RE A-1 VESSEltRPvi OR , RHg (Lpcip, RECIRCULATION M ,,,i r, C
HIGH DESIGN _ _ LOW DESIGN PRESSURE (HOP) ~ ' PRESSURE ILDP) 1 AO MO s A w a RHR (LPCll.
RPs. -X y i -,
CS OR LPCS FICl'RE A-2 ; LDP HDP O HPCI INJECTION 0
1 RPV M 7 Q- -->@ C$
mm >
V m HDP LDP AO FIGl'RE A-3 s
AO MO l I
RPV Mh 74 d RHR (LPCll, LPCS l m2
- AO j HDP e - + LDP FIGGE A-4 .
I I
A-6
- . NEDC-31339 HPCI INJECTION AO MO MO
_ . _ ch __ - . _
FIGURE A-5 Rn W4%d ;; pc cs HDP 4 - + LDP AO l
t AO MO MO FIGURE A-6 RN OR RS
HOP = = LOP !
MO ,
i MO
-> 4- --
FIGURE A-7 RPV -
--Q Q- M CS .
t
-5 4- , r MO J L MO N HOP = = LDP AO MO Mc M ?; --
FIGURE A-8 RN W -
-CC CS
->4 AO MO g),as- + LDP i
i
-l A .7
NEDC-31339 A0 MO MO ;
FIGURE A-9 RPV ORR$ *>Q- 4d ;4 >Q RHR tLPCI), CS AO HDP = = LDP t
MO MO F1GURE A-10 Rs >< '/ >c ;; RHR cLeci> _.
HDP c - LDP J
MO MO FIGURE A-11 Rev X /
' ;4 -Q C cs HOP c ; LDP TO APV il h
w, -
V m - "
A0 MO MO MO
~
F1GURE A-12 'd ; ; ><
HOP c = LOP 1
A-8 l
NDC-31339 l l
TO RW FIGURE A 13 mm _
1 MO MO MO
? 4 ' >< 'M
_ ,w \
HOP = = LDP TO RPV n
wa -
"g
~
FIGURE A-14 MO MO MO
'M g ? 4 >C N #
REACTOR WATER CLEANUP HDP-c p-LOP
?$ =
AO MO MO MO
~ ~~ ~~
FIGURE A-15
-M >1,>< 7g U.)
HOP c ; LDP MO li
' REACTOR WATER CLEANUP tRCIC ONLY) a TO RPV - yj = ;
i h
AO MO MO MO i i
I
\
nGexE x-ie - , . l 64DP *--> LOP l
A-9 i
- EEDC-31339 Mo MO
->( *
^0 FIGURE A-17
'/ '/ ;A HDP c r LDP g,y
- HEAD SPRAY M0 M
To RW m m r,
' ?$ &
^C M0 Mo Mo FIGURE A-16 - --
u, 4
T= >< A 3
= cs ]/
"0 " " * *"
To Rev uo l
5t
- i F]GURE A-19 uo uo uo
?O >< / ]
- ny j HDP LDP To RPV !
i
i Ao
~
- 1 l FIGURE A-20 uo uo uo Vr > V } l
- - es uo 7
MO - b M Hop top
& ->l: 'q Ao
=m A-10 l
- -; ='
I~' i F]GURE A-21
- IMD MC d'0
> /'
- rw TO RPV HDF c ;- LD*
- 2 =
F]GURE A-22 M0 AO MO Mo
> < . L 't--DC ]
G=
L rw HDP 4 LDP
'70 RP V h
II
- r]Gypt. A.23 AO MO MO MO Md ?4= ;4 ]
~~*
I
- . _. 7 HDP m-+ LDP l
TO RPV '
1 i
b l
AJ g FI G'JPc. A-24 7
4 AO MO MO MO
-+ i-&4- 4 < ]
- x_ 7
-M HDP W L LDP A-11
N DC-31339 il l
1 1
iI M h '
TO RPV , ,
d -
F]GURE A.25
'/ ;; >< ]
_ ,W
]
HOP C O LDP l
TO RPV =
d MO MO MO
~~
A0l ~~ ~
FIGURE A-26
--L d u ;; 4C '
=] 1 u
db d bu2 _ f r, -
= .
l d
'C t
) Ao Mo Mo Mo t
-- -- I FIGURE A.27 4d ?; X '
T
- FW HOP c O LDP P
i
=
m
>I<
A0 AO MO
-- LM0 _ _
l FIGURE A.28 1 / I / k.2 m
/ I / f V ' l es s .
b
- p. zgy HDP -e - > LDP l i l
t I
A-1l!
[
t i
to nrv Il ,
I
. bb ,
FIGURE A-29 ;; >< '/
- FW HDP = r LDP To nev -
a ,] ,
5
_ _. h__ Mo t FIGURE A-30 ;; >< /
O'
- FW HOP e - + LDP m2 A0 AO M0 Mo 1
FIGURE A-31 _44 44 ;;
]
i I
l l
! l l A-13/A-lI. l l
l
0 g AFFENDIX B DESCRIPTION OF LAVRENCE LIVERMORE NATIONAL LABORATORY (LLNL)
PIPIhG RELIABILITY MODEL AND GENERAL ELECTRIC NODIFICATIONS
. . NEDC-31339 L.1 GENERAL Tigure L-1 provides a representation of the Lavrence Livereote National
- Lateratcry (LLNL) piping reliability codel (5.13). An initial population of circucierential cracks is considered. The depths and lengths of these initial ,
cracks are described by appropriate probability distributions. The initial cracks are expected to be detected with certain probability during the pre-service and/or in-service inspections. Cracks-that escape detection and !
repair are modeled to grow suberitically due to fatigue and/or stress corro-sien cechaniscs.
The crack growth calculation is based on the stress history induced by the norcal and abnorn! operating transients, earthquakes, and other cyclic loadings. The paraceters defining the assumed fatigue and stress corrosion crack growth laws are characterized as normally distributed random variables.
3 The critical crack size is determined using a net section stress cri-terien in vhich the caterial flov stress is assuced to be a randoc variable. i
] The probability of failure cf a pipe is the probability of a crack growing to the corresponding critical size. The codel also includes the influence of a leak detection systee and r hydrostatic proof test.
E.2 PROEABILITY DISTRIBL'TIONS AND FAILL'RE CRITERION f f
i The probability distributions and the values of the parameters used in the LLNL codel, as they pertain to carbon steel piping, are described below. ;
i a
4 B.2.1 Initial Crack Size Distribution The specification of circueferential surface crack geometry requires two parameters (see Figure B-2): depth (a) and surface length (1). Therefore. I the initial crack size distribution is characterized by a bivariate i distribution.
l i !
B-1 l
. - _-- ._ . - - - - . - - . , - . - . . . - - . . -...._\
- "* / c.ometry i,s i d *d nosnt l mewit: Womies f -
Irwtel crect taas f
R6noomey detriDwtions h5h gt w.ct a .c. P t..bi l
's== -
4/h ;*
g
- Prmerv c. . 0 [ '#D t inspection (P$ll I 1 I ' m htees of g
' WMeetion Pq e Hv orostatic fi bia = 10-proof telt
- Stw hinery j - ;estuteted transaents
- *nmet events 1
I E
C'**"
growtn ath /k i Ciaca growe k (;E' O o ,,b 1 l
- ['
3 i
chw actorinia se j ,j - Inservice espections itsil l Dg ya;Iti. e JK/11. Ap
-- ' ~re P , . .s,,ri.
- Uttitel net Mon Stral
- tear at moewtus mitsoility l
9 .
Lasa de rection
.r Leam detection P,g capabihty
,,onoitionai w isen anc
- GE GE pressoditivi Ltan witmEo=
, * ** Lena .,o E os OE GB .ith Een 0 k ' ""o MGB a,o Eo=
" or obabihty i
,7 GaHNintt systeen e p,g 4hre prooloihty gp Figure E-1. FIcwchart of Frobabilistic Tracture Mechanics Model Implemented in FFAISE Computer Code E-2
NEDC-31339 houlN AL Stat 13 1%?et wwCmACKED SECTich ot Pitt CR ACK LENOTM f
- 2R4 ] ,,,,,
9 *-- 8 Low sTR Ess. e' ,
, l--
s x == ,
2
/. *-
1 4
-/g,I\ i ,/ ,
a 5 ,y .
~ ,i 7t\ ___i
-= j i
\ \ \\
+ 'l _-__ I ,g m
a i
i II 7;"
l + !*~'"
d
$? m t !1 Di!* a n etiCN l*s Tat Omatat:St:?*C% a?
Tag *0 INT 08 :CL ..a*11 P.* A**.it: utvts ast sta t!11% UNCR ACE t t SI TION
'n
- A**..t: s ts iNG sta t ts is v%:p aca t c st :* ion 1
l l
l Tigure B-2 Scheratic Showing Stress Distribution at the Cracked locatien in a Pipe at Limit Load E-3
NED0-31335 The crack depth is expressed by the exponential. distribution:
-a/u -
(a) = _(B-1)
~T* p().e -h/u) i where h is the vall thickness and u = 0.216 in., a constant. -An option to specify lognermal distribution is also provided in the PRAISE code.
The initial crack length is expressed in terms of an aspect ratio (6)-
equal to 1/2a. A n.odified legnormal probability density function is assumed for 6:
r 0 6<1 P (S) 2
) C
/ S>1 (B-2)
, 6
-(inh)2(2A) 16(2r)I ,
where constants 6, = 1.336. C5 = 1./.19 and A = 0.5382 E.2.2 Crack Existence Probability It is assueed that an initial crack is produced by a fabrication process _ ,
such as welding. Therefore, the crack existence probability is related to the volume of veld and heat-affected zone, V = 2tDh8 , where D is the inside diaceter and h is the pipe thickness. The probability of having an initial j crack in veld volume (V) is esticated as: l l
- (B-3)
P(1) = VPy , e where P y , is the rate of cracks per unit volume.
i 5.2.3 Crack Detection Probability The probability of not detecting a flav (Pg) is a function of the flav size, the caterial inspected, and the instrucentation characteristics. For E-4
'. . NEDC-31339-the case of vrought austenitic stainless steel and flaws that are long (i.e.,
il 5), the following formula is suggested:
l l
Pgp(a) =
0.5eric(vin {,) (B-4) where a* and v are constants and eric is the complementary error function.
h.0.4 Patigue Crack Growth Parameters The fatigue crack growth is characterized using the equations:
1 0
ff = C
.(1-R)1/2 .
(B-5) l l
vhere:
LK = range of applied stress intensity factor (ksi /in.)
R = load ratio defined as v. min! max in which K,g and Kmax represent the minimum and maximum stress intensity factors, respectively, a = crack length (in.)
N = number of cycles Cm = constants related to the material and environment.
The exponent, m, is treated as a constant, while the constant C can be specified as a constant or as a random variable with specified median and 90th Forcentile values.
B-5
' NEDC-31339 ,
L.O.5 Failure Criterion A net section criterion is used. This criterion is stated as'follows: ;
(B-6) c LCA p"#o (A -Acrack) where:
e " "*A*1 * *P "'"' ** * * "E# " #'
LC A = metal cross-sectional area of the pipe.
p A
crack
= cross-sectional area of the crack.
3 a
e, = eaterial flov stress, which is a function of the yield i and ultimate stress.
The load control stress, cLC. includes the longitedinal cembrane stress due to pressure, and the mecbrane and bending stresses' due to veight and 3
seismic inertia loadings.
k E.2.6 Leak Rete Calculations i
The leak rate calculations are based on an assu=ed single-phase flow I model of initially subec)1ed liquids through narrov passages and is described i in Voluce 5 of Reference B.1. The cathematical relationship between crack i length and leak rate built into the PRAISE code is based on typical PWR coolant loop pressure and temperature conditions. ,
! B.2.7 Nueerical Simulation ?
i l Eecause cany parameters are treated as randoc variables in the LLNL f study, the Monte Carlo Siculation technique is used to evaluate the leak and t double-ended guillotine break (DEGB) probabilities. A stratified saepling scheme is used to increase the accuracy and coeputational efficiency of the nueerical sieulation, j 1
l B-6 {
I
_ - , . _ --m_-- _ _ _ _ _ - - _ , _ , , , . . . _ _ _ _ . . . . . - , , ,_ , y. ,_, ,_,- ~.,
'. , NEDC-31339-E.3 MODITICATIONS ISCORPORATED BY GENERAL ELECTRIC
~he codifications made to the PRAISE code by General Electric are in the j arets of failure criteria and leak rate evaluation.
(
1 t
B.3.) Improved Failure Criteria ;
An examination of the net-stetion failure criterion, as represented by Equation E-6, indicates that it does not include induced bending at the crecked section due to the eccentricity produced by the presence of the crack. ,
The following limit load equations developed in Reference B.2 include this 4
effect:
(t-od/t) - (P */c g)v E= (B-7) 2 2c P = (2 sin 2 - d/t sin a) (B-6) ,
i 1
vbere
}
t, = pipe thickness
- a = balf crack angle as shown in Figure B-2 1 & = angle that defines location of neutral axis j e
g
= caterial flev stress ]
F, = applied membrane stress in the uncracked section ]
P g = applied bendies stress in the uncracked section i
The resulting failure criterion line is schematically shewn in 3
Tigure 3-3. The key input in Equations B-6 and B-7 is the caterial flow stress, e . The lover the assumed value of flow stress, the more conservative g
(i.e., failure predicted at sealler crack depth and shorter crack length for
]
the sane stress 1 vel) the resulting failure criterion. Therefore..a requi-i rite level of conssrvatism can be assueed in the failure criterion by re3ccting an ap;.ropriate value of flev stress.
B-7
NCC-31339 i
._ p n
\
. e ,% . .s
\lw ' 1,\.\
l
\;-t
\ \\\.: \ ,' ' ,
l A
1.0 , , , ,
k i
Tailure Line Based on l Net-Sectien Collapse l
( Schematic Only )
- 0. 3- -
l t
I I
$ I
=
2 _._ _ - - - - B
'7-y 0. s -
Tailure Line Assumed In Probability Evaluation E
E
_ 0.4 _ _
0 N
5 0.2 _ _
0 ' ' i i 0 0.2 0.4 0.6 0.8 1.0 FPACTION OF C!RC'JMFERENCE s/-
1 l
l 3
Figure B-3 Schematic Illustration of Tailure Criterion Used in Piping Rupture Probability Evaluation l
l B-8 l l
NIDC-31339 Since a large ntaber of repetitive evaluations of failure criteria under the Monte Carlo Simulation scheme are involved in the PRA!!E code, the failure criterion line schematically shown in Tigure B-3 was simplified as shown by the dotted line in the same figure. A key advantage of this simplification is that it can also be used with the methodologies that provide information on the failure flav sizes only at the ende of the failure lint (points A and 2 in Tigure 3-3). For example, the elastic-plastic fracture mechanics based methods can readily treat only the flaws that are either through-wall or 360*
part-through.
B.3.2 teak Rate Prediction The leak rate model currentif tuilt in the PRAISE code corresponds to PVR
=ain coolant loop conditions. Therefore, the code was modified to accept an externally supplied crack length versus leak rate relationship. This provides complete flexibility to the user in the selection of appropriate leak rate calculation model.
3.4 RITIRINCES 1.1 "Probability of Pipe Tracture in the Primary Coolant Loop of a Pb'R Plant". NURIG/CP-2189. Vol. I through 9.
5.2 Ranganath, S. and Mehta, B.S., "Engineering Methods for the Assessment of Duct 11s Tricture Margin in Nuclear Power Plant Piping". Elastic-Plastic Practure: Second Sy=posium, Volu=e II-Tracture Resistance Curve: and Engineering Applications, ASTM STP 802, C.T. Shih and J.P. Gudas. Eds.,
American Society for Testing and Materials, 1983,.pp. II-309-II-330.
B-9/2-10
APPENDII C
- BEHAVIOR OF FLANGED JOINTS DURING PRESSURIZATION l
hDC-31339 C.1 GESTRA1.
Bolted connections, whose failure during an ECCS overpressurization event could lead to a large leabge area, are a part of the valves and heat exchanger assemblies on the low-pressure side of the ECCS piping systems. The questions that need to be addressed in the evaluation of the integrity of such bolted connections during the overpressurization are: (1) what is the maxistz stress in the bolt during this event, and (2) vbat is the impact of the failure of one et two bolts in the overall integrity of the bolted connection?
This appendis briefly reviews the theoretical background and the necessary equations to address these questions.
C.: BOLT STRESS The bolt stress in a typical flanged joint is primarily a function of the preload and depends, to a lesser ex ent, on the applied load and the relative stiffness of the belt and the flange. In most flanged joints, the major stress applied to the bolts is the preload that is applied in tightening the nuts. Reference C.1 states that the folleving empirical formula provides a fair estimate of bolt stress, S g , d e to preload:
45000 S = (C-1)
E where d is nominal bolt dia:eter in inches.
~he total st"ess S , in the bolt is given by Equation C-2:
T 5, = S +
T
- b (C-2)
' i A ug Ed f
where:
F = applied load / bolt j A = net cross-sectional area of bolt net g = bolt stiffness K = flange stiffness g
C-1
'
- NIDC-31339 vpically, the flange stiffness is eight times the bolt stif fness (Equa-tion C-3).
Accordingly,
- (C-3)
ST"31*A not New consider an example where the bolt diameter is 1 inch. Equation C-1 vould indicate a bolt pre-stress of 45,000 psi for this case. Assume that the applied load / bolt is such that T/Anet is 36,000 psi. This means that the applied lead would produce a stress of 36,000 psi in a bolt with no preload.
On the other hand, the preloaded stress in the bolt based on Equation C-3 would only increase from 45,000 psi to 49,000 psi. This represents an in-crease in bolt stress of only s91. This confirms that, when a flanged joint is subjected to pressurization, the bolts experience only a small increase in the stress o.er and above that induced by the preload.
- ten the applied load or the flanged joint is such that the bolt preload is cenpletely overcone, the bolt stress is then simply given by I
S.' (C-4)
A net This occurs when (Kg g )
T = (Sg xAg g) (C-5) g, C.3 IMPACT OF BOLT DEGRADATION ON TLANGED JOINT INTEGRITY i
A review of General Electric service experience data base on !VR pressure boundary caterials indicated no reported incidents of degradation $n carbon l steel (SA!93 B7) bolting used in ECCS piping systen valves and heat exchan '
gers. This was not surprising, since cost of the factors (identified in Reference C-3), such as the presence of borated water, stress corresion cracking and fatigue, are not likely to be associated with the operating ,
l C-2 '
i
NEDC-31339 conditions in the parts of the 3'n'R ECCS systema being considered in this evaluation, nerefore, the only possible scenario for the failure of a bolt during the pressurization event is the folleving: an undetected defect or crack at the thread root exists such that the bolt failure does not result during the preloading but occurs during the small incremental loading during pressurization to 1050 poi. ne analytical results presented in Reference C.4 are helpful in assessing the impact of f ailure of one or more bolts on the overall integrity of a bolted joint.
As a part of an effort to ertablish the leak-before-break margins in a steam generator sanvay closure, Reference C.4 reported the results of finite elenent analysis on the load shedding and redistribution characteristics of this bolted joint vben the failure of a number.of studs was modeled. Tig-ure C-1 from Reference C.4 shows the load redistribution curve for three adjacent studs as a function of a number of failed contagious studs. It is seen that even if three contatious studs vere to fail in this joint. the stresses in the nearest stud would only increase by 222.
nis clearly illus-trates the redundant nature of a bolted connection. Turthermore, failure of a number of studs would lead to leakage which is likely to be detected.
Based on the preceding discussion it is concluded that:
- a. Bere are no inherent environmental or other nechanisms present j wh.ch could cause degradation of bolting in the ECCS system valves l l
and heat exchangers. I 1
- b. Highly redundant nature of bolted joints vill result in leakage rather than failure in the unlikely event that one or two bolts were to fail.
l C-3 I
NIDC-31339 i ,
i 2.6 -
II'd '"I* -
2'4 -
OOQO ** * '= n 2.2 - 9- original strent -
o* e (he sagra sa tion s 1808 Ma a==a r *'
2.0 -
Laal tors Pr e i ea d pg * !!!! p s i g ( 11. 4 Ma l ste a re n t ,
g'g _
l{e l ghter e 16 -
o j l.a - -
~. - .n oas me e rv e s
-n.2 - -
l i
i.,
oi -
(
there f.eertit c.6 -
l c.s -
c.2 -
i i t i I
- g 8 I
, c 1 2 3 noi.e r vi s.ntiq...s 14 leo .t as l
- Figure C-1 Load Redistribution in the Three Nearest j Studs due to Stud Degradation C-4 1
NDC-31339 C.4 RITERENCES C.! Rodabaugh, E.C., et al, "Survey Report on Structural Design of Piping Systems and Components", TID-25553, December 1970.
C.2 Shigley, J.E., "Mechanical Engineering Design", 2nd Edition, McGraw Hill Book Co., 1972.
C.3 NURIG-0933 "A Prioritization of Generic Safety Issues", Safety Program Evaluation Branch, Division of Safety Technology, U.S. NRC, New Generic Issue 29 "Bolting Degradation or Tailure in Nuclear Power Plants",
December 1983.
C.4 Nickel, R.E., R.C. Cipolla and E.A. b rrick "The Use of Leak-Before-Break Criteria and Assessment of h rgins in Addressing Closure Integrity Issues", SMIRT-8, Paper D 6/3, August 1985.
C-5/C-6
8 9 APPEND 12 D SAMPLE PRAl$E CODE RUN
NEDC-31339 3 . .
1 i
r
?
-$.e.w.A
- e. c:.
W e < WWWkw 9Pzrs
- E E
- - E.chhI.
e y . s.ea. s. s.
- esos hee-r a - w e- - h
- i ' s' ' [ -
E w -[fff
- c. . ..... .
- ..- - g !
- . e e -
I'Yv6w
- <
- : E
. . w t ,, zrrr a e ~
. 53rr oc??
v ce e
E-we-we-c .=- -c- - - - -
eaves i *
- we
- <e-wb*.3
. - h. .k . -
- :< f,
- r : nee eco -
k' .s a
- 8. e e. o.woo g <
E.ce-. E.e. e e.
{
veyt y w ee -
I -e-n r .
. -e
- n 8 eece.
E5 E mg 6 ge 5- --
re c-c e ~ ne vo e. - X-
= -c 8c.AA e. c.
w c E=c e e me- c. z -e - <
w wtweew -L .
. .z - < e -- 6 . W V, r, - .
E ,.
r e < :-
e z
e T
.S - e s-
.- g.-c es-es bp
- se w-et
{s..
,h-e y .
- ea s 5 e e <-
- E.<sa-<u vc-r e v --
-3 w c .es:- s - =
e .< er m . - - -- - - -
- > -c < cw w -r = -
-g cNA-e e v.
e e-
- p a w-:==R e-hB -c. e g. r l
-w c.
e k.
IE g<
a : ; : := er - v
<f6 -LEtw -rt.
r r. ws e r w -=
. e r
, = eg c
<w
- r* <- c.3
- r r r.w.ste w *
- ge- Ew %-
'g Sb.is. .. e<
-.- e a E
e- w w e
.ec
<s c-s e"
e=- -<<<<, 'ssss5 e 2- w w w" GW<* = Wkk ' v Etr e .-
s- rra g<g.- 1 4
v , . E : 8x - t- z = g . < -
. e y- -ww z -
y
<g:<! v <r evo w . . <-w = . .
r
- <e .e c e-
-s e o c<w<
e<--6:
- v. - ----< st
< e* u. -
c ,. .3 =
e e tw D< te.
-< 6-e <e r:g.
c< C s===* ' r '.f. IE5r . <e=ce E t'--- e >c
=
'.& l'e- t-we
- o. f. r 5- c
- < g-2 <=L r, - wav-wv cevuuc ecp66 -w-c
-s .-
e-nck c - e- - u e v <
v w oc < g w g y e a - U - .
y
'. c
- s.
c-
= <
e
=
e
- r. I*
C =t V % v v \
- W -4 w 4 - 6- C "
- L -< c < - c h* w
< = Ze e e c W - [ *:
v w L -u < v c s - -
i l
I 1
D-1 1
4 s.,. A e m. x us - L. u y 6 - as--. _: sw NEDC-31339 e .
r 1
F Y
I P
5 I
O T
i C * ,
8 (
. C a O. m e C -
e= = w3
! N m :
- r a
=.
I 3 -
e-
~
w ; k,'
[ g 3 -*
. e -
= -
z E ,
l 4 E
< o. m -=
a wg L
> w
. a w y c r -* !
4 g ,
C - - - 5 ;
=
C. =(
C g gn Wm s s.
- j. y u ve.4 = g i
. I W -
- i. < c w ,
e > w' i 2 E E C E j - w m
( E C se i U E F i <
- E- y .
en
. u. .
...g . . k .
a .
t w U = l
< E we w i
, ..- v 5
,1 . E 6.
. eE* t 1 . C ,, 6 h
- ( -* .=
- f > gbw g i E . > .
1 = m C 4 1 g - . g
!
- E e (
, .* == -
L E
E -
=
i 1
- k
_ L.
2 - 2 e -
I l '
I 1
N I
1 i
)
f 4
D-2 i
, . - - - ,, y - , - - - , - - - , -
,, ~, .. .--.,---- . - - -
- - -. , . - - ,w,-- . ..-, ,w-. -,
i 1
e .
N DC-3&339 1
nc s
i C-E-~
.h*
g
- ?*
M TE
~,a m *.
4p
- _ nc L @ <s e A n
= .E
- C =
= v g 6 %
+
y
.w.
c E E
< .E E E =
t, :
w <0 0 =gC
(. Y t.*
a W *a = l 3 e= ;
u j
=t,
~
w!r L
>=<
- V. C l
D-3
t i . N DC-31339 .
i e i w
(h.hbbhb 2-bbbbb bb bb b I
e i
TW
> s e m N N N s e n N N - e e n ~ e r. ~ g ~0 0. C.
C. C. C. O. O. C. 0. C. C. C. 0. C. O. O. O. O. C.$ 0 0. .
-wwwwwwwwwwwwwwwWwwwww8.w aeoe.enn-eE.s-eeee
- e e e ~ = - c e. ; ne -~~ e ~ ce o . s e. .
suceNeie e
<ee~~n gennesen e ~8newee.~.~-*I.anse e,susechunne,e.n, c-eo~ in E e. n c ~ ~ e e s - e m - eie e. - e - ,a n c e c . - . . c. c. ~. n. u. n. e. . ~. h. u. N. a. s. a. l . - ne. =. =.
i.
i
- s. f e f !. ? f f f f e s t ; ; 3g g g g g g g r y gy 7 ,
.fx xx ffr::
t
= .!: !. ! & E. G E. y g a: - A. ' U -= b. ; &. 7 ;. 3 ; ee:. . s. e. . E - K :
r . - - e 8
v -
~ ,
w e o !
, w t
< - ~
- e. 8 ~e .g g e ;
- - n .p 8 e 8 {* *m E p e g gwe eegs8 n. i.
e :~ c SEgE.E .-
2
. z - Bee.. ~esc
, e c, c.~,...e.ee..88.b.t5 1
e 4 i
- = !
- L
= v ~ g.- C C g C C g C g g g C o p "5 8gCCgIg u -
.cc e g ? eE. 4. .; e. e . . . : . e as~~s r
4 e
<. 8~ w ceseeewee e. e. s. e e. o. ee e e e f. c e 5 c 5 v w ,
=: ,
- f. f. f. f. 9. Ec o- c t.ce l 9 9.5 f '
esecce,e,ree f8 RE f 5c. e.
e< s.
r E e.e e.r o.e r v e,r.e.e r o e.s s. 5 cs. u ~ l w ..
g
? g
~
-AnveQ~ce:-Wn-ecAte==hn r . ----------NNgN w I - W j
- I i , .
I e
1 h
D-4
k NEDC-31339 e e r i
e e m e N l C C C C C
- g. O.
y 6
i.T T
[
e t
g C r = = ,
1 . e e. n.
e . . . . . . . . . . . N. . .
O C. O. O O. O O. O C O C O O O O O O. O. ,
e i
m e N . e e a N
. C C C C e C. C. O.
. .C k' I r
i e
& 5t = t y .
l N = c - =
a g 8 W i e h a fE d d d d d d d b " d d d d d d N d a
E e A N N N N m e e N w v C C O j,
E C. C. C. C. C. . C. C.
4 W I b
'9 W k W W k s - 5 1 :* aO Cg fE eg g
! A E e =
, - g -
- _v. e e e - - -
5 E ~ c. s ee -
n E.
3 . . . . . . . . .
1 O C C. C. C C. C. O. C O O O C. L
., $ l a = i
< s l C = e e- e A t i < C C C ,
t g g . C. C. ,
.e .- tg ky W 4
" g
- I 5 . i
). < . C m e c c
e A
- O -
. . . . . =. . N. . .
O. C. C C. O. O O C. C. C. 0 0 0 0. C C. C. C
< e N = m e m N
=. C. C. C. C. C. C. C.
. WWW W h y W I 1
- ! *
- E 5 s -
7 c
2 ; E d & M h - h E d d d d d b d d d d d d d d d d 1 e e. N N N A e e p. A -
x C I
< C. C. C. C. C. C. C. C. C.
- Y W t W
x W I %
Ip
- x x x f. : x e i.n 7
- ~
- - -snn c 5 5 L F 4 E
r, e e e, _
. . n. .
C C. C. C C. C. C. C C. O O. C. C.
i d
t C C C C
. - C. C.
C 0
. C. C.
C C.
C . C. C. C. C. C. C. C. O. C. C.
j - C C C C 9 C C C C C C C C C C C C C C C C C C j
1 . - N M T O C A C & = N C N C
- N M j . .O = = .e .t O. . . .S .P 6 P. N N W
4 l
j D-5 a
a e- - , . , , , -
--.,r,,.-,-----.,3__- __.__-,,-,--,---.,-.m. - ~ , - -. .-
. , N DC-31339 r
m M e e N
< O C C. C. C. .
g
~
!,. E, I. i
[ C w.
e c. . . . E. . . . . . =. . . . . .
o o o.
e o o C C e. o O. o C e o e o e !
L W l A ~
- a N - a e a r
- = C C C 6 C. C. C. C.
b$* Y Y-t , ,
el ,}
i.= er .
O C O
. . s. . .
O O O O O OO O O O O O C r
- l
< /
b i
= m h h h N N m e e N m' C C O C C O C.
$j.C
}
4 O.
ka O.
g' 3 N e C &h e g
- = I h . ,
c . e 2 e . - .h n h .
d d dd d dd NN dd N d P
$ P n P P k' w C C C C C C r
. g- -
V i
.e
- t.w E 9 w .
i I
e w
G.
e e C
. . n. . . .
- u. .
f . . . .
L C O C .. . C C.
- O C. C. C C C. C C C C C. C C
.I
. I m N - e n e w .
w o* C. C.
o e. e. c. f g I Y I h W W >
eE ! i,
- E ,
c .
,E .E C
= C e . . t g e e i n u . u b 6 6 N d d d N N NN d d d b d d I I
e N 6 N N N m e c. h- ,
t C. C. C. C. C. C.
C c c t
< C.
I I W I k Y f i 5 i
- E = E >
- r. 7n -~C. -*N E ev eP.
E.
e 5 4
.e h
. h I
t b $ $ $_d d b b d 6 $ d $ !
I b
y C e C o e o C O o o o e c o o e o O c c C c C
- C o i d C d d 6 6 6 d 6 C 6 6 6 d d i 6 6 6 6
. i I
- - N e e e o a e e .e
- n. .n
, e e e
.c
. w a !
- . - . - - - t. A 6 t, ,
6 i D-6
. - . . . _ . ..- .. ._. _ -. - _ . . .=. . . _ . _
i 1
N E -31339 l
\
l 1
i l W I a u .O .C .C .C .O .
C..... .
" $.ceN
$ N. N N,. ,
.hseee t
. S. S. S. S. S. ,
i hhhhh i k '
l !!.!E.S.T.
'W WWW e en- <
d *NNh '
1 j .
g eggs ii
$ . 4 . = N.
-No...
A V.
i e a
w gT9999 a
a m- N 4 0. C. C. O. O.
g
= 1 s '
= sE emer a t e E A N E. ev b et e 6 = =.
y) & P O P..
j 4 ; w Nh=TT i j -
E !
3 1 !
3 : -
2
- k. - < E. e o g gc c. . . t 3 4 -
1
(
= CWWwww ,
.ThG44 1 g e m a gae r I I=
- = .,
i
Is,e%%~ l e < N. N. N. A h.
i g ... . ;
, g i 1
i L e i
gS$53! !
l W
g+
ww r
=
C h t. N# %
-esI*ec
' . c .e ~* h. . *, F ee l C
> O. I. ts O = .
t e <=AF..
'! Y t i 2 l 4 P
~
'1 C MeNNN ,
- +
> x 0 C. C. C C j % wkwwww
. ece 9 . C*em ,
W CSA eoe J 4 >Gh=Nc ,
< P. t - P
- j
.s... t
. l
) W t O. C O O C..
l
~OCC*C '
].
- =hEt i
< i l )
2 I
l
]
4 l
I D-7/D-8 f
l 1
-_-------.---n.- - ~ , ~ - - - - , , , _ . _ ,,_,._...,-n_ , , , . . , _ , . _ , , , - . - . - , , _ . . _ . - , . - - - - - - , , . . .
,.a~..n,-r-- - - ,, -
NID0-38339 f
APPENDIX E PARTICIPATING UTILITIES - Bk'R Ob'NERS' GROUP ECCS PRESSURIZATION COMMITTEE l
1
)
l i
l 9
1
NEDC-31339 APPENDIX E i
PARTICIPATING UTILITIES - EWR OWNERS' GROUP 2CCS PRESSURIZATION COMMITTEE This report applies to the folleving plants, whose owners participated in the report's development: ,
BWR OWNER PLANT i
Boston Edison Company Pilgrim Carolina Pove.r & Light Company Brunswick 1 & 2 Commonwealth Edison Company Dresden 2 & 3 Quad Cities 1 & 2 ,
LaSalle 1 & 2 Detroit Edison Company Termi 2 ,
Georgia Power Company Hatch 1 & 2 General Public Utilities Nuclear Oyster Creek l Gulf States Utilities River Lend 1 Illinois Power. company Clinton 1
!cva Electric Light & Poser Company Duane Arnold Nebraska Public Power District Cooper ;
Niagara Mchawk Power Corporation Nine Kile Point 1&2 Northern States Power Company Monticello Pennsylvania Power & Light Company Susquehanna 1 & 2 l Philadelphia Electric Company Peach Bottom 2 & 3 Limerick 1 & 2 Public Service Electric & Gas Company Bope Creek 1 Tennessee Valley Authority Browns Perry 1, 2 & 3 Vashington Public Power Supply System Banford 2 l
l E-1/E-2
ATTACHMENT 3 October 1987 Appendix R Audit Open Items Not Requiring NRR Review
- 1. Fire Damper Operability (Unresolved Issue 84-40-01, 84-19-01 discussed on page 4 in Inspection Report Nos. 50 277/87-30 and 50-278/87-30.)
NRC Comment:
The licensee's Technical Specifications require that fire dampers be inspected visually. The NRC raised the concern that a visual damper inspection does not provide assurance that the fire dampers will be able to function properly during a fire. This concern was raised because: A) The licensee could not provide Q.C. records indicating that the fire dampers were drcp tested after installation, as called for in the engineering packages; and, B) a recently issued 10CFR21 letter highlighted the concern that the type of fire dampers used by the licensee may not close under air flow conditions.
The licensee addressed this concern by revising the fire fighting strategy procedures giving the fire brigade the option to de-energize the ventilation systems involved. With no air flow presumably the fire dampers will close. The licensee's actions did not satisfy the original NRC concern for the following reasons:
- 1. The inspector observed a fire brigade drill. Although dn attempt was made to verify whether the fire jumped to areas above the hypothetical fire scene, no attempt was made to find and isolate the ventilation equipment.
- 2. Assuming that the brigade does turn off the air handling units there is no assurance that the dampers will fully close after the air handling units are turned off. This is because the dampers may drop and bind in a partially open position before the air flow is cut-off. To assure that the dampers close, the licensee must provide assurance that the dampers will close under air flow or that the air handling units are de-energized prior to dropping of the dampers.
This item continues to be unresolved. Considering the above concerns the inspector questioned the operability of the dampers.
I
Response
i A fire damper program has been formulated to evaluate existing test data and damper closure with air flow data and to address fire brigade and training procedures to provide reasonable assurance that the fire dampers will l
Attachment 3 Page 2 satisfactorily perform their design function. The evaluation program will be completed by August 1989.
- 2. Incorporation of NRC Comments on Procedures (Page 11 in Inspection Report Nos. 50-277/87-30 and 50-278/87-30 NRC Comment:
Procedure SE-10 "Plant Shutdown from the Alternative Shutdown Panel" was reviewed and found to be adequate. However, the team commented that some steps in the procedure may need signature checks to assure control. For instance the steps monitoring the reactors' cooldown rate and other steps that operators perform in the attachments to the procedure do not have sign-off blocks that the operation was performed. The licensee in subsequent discussions committed to review the procedure and add sign-off spaces where needed.
Response
Procedure SE-10 is currently being revised to reflect changes caused by the completion of Appendix R modifications. During this revision, operator, training, and NRC comments were revies i and incorporated into the procedure.
In addressing NRC comments, sign-off spaces have been added where needed, and the monitoring of the reactor cooldown rate has been enhanced.
- 3. Accessibility of HPCI Inboard Steam Isolation Valve Panel (Page 11 in Inspection Report Nos. 60-277/87-30 and 50-278/87-30)
NRC Comment:
During the walkdown of procedure SE-10, it was observed that the breaker panel for the inboard steam isolation valve of the HPCI system has a cover fastened on with wing nuts. The team observed that if the wing nuts are too tight the operators may not bc able to open the panel. The licensee stated that either bigger wing nuts or a tool will be provided to assure panel access.
Response
The HPCI Inboard Steam Isolation Valve panel was originally provided with slotted screws which required tools for access. The slotted screws were changed to thumbscrews to allow an operator to access the panel without the use of tools. To address the NRC concern of overtightening, flat washers were added to compliment the thumbscrews. The washer addition will provide a smooth contact surface and enhance the operator's ability to loosen a tight thumbscrew.
l 1
i Attachment 3 Page 3
- 4. Fuse Replacement Controls (Page 13 in Inspection Report Hos. 50-277/
87-30 and 50-278/87-30)
NRC Comment:
Ouring the review of the licensees circuit coordination study it was identified that the licensee does not have administrative control procedures in place to control future fuse replacement activities. The licensee stated that a procedure to control fuse replacement is currently in the process of being written and implemented. The licensee further explained that fuse replacement is currently performed by either "replace in kind" or using the Control Room mark up drawings which call for the type of fuses to be used.
Response
The following. administrative controls for fuse replacement are being initiated. A modification has been started by the Nuclear Engineering Department to generate a controlling document for fuse replacements.
Additionally, a guideline document will be added to the watch standards guide to assist the operators in the practice of fuse replacement. The guide will be revised by December 1988 to reflect this practice.
l
. -,