ML20096F054

Forwards 920427 Memo to BNL Re LOCA Outside Containment in Advanced BWRs

ML20096F054
Person / Time
Site:	05200001
Issue date:	04/27/1992
From:	Knecht P GENERAL ELECTRIC CO.
To:	Hsa C, Poslusny C BROOKHAVEN NATIONAL LABORATORY, NRC
References
NUDOCS 9205200162
Download: ML20096F054 (7)

Text

"

. # APR k hah M 53P[ GQC h sI - ~ ^ - ~^

~

^-

~~- - -

< > -.. ~

...;70:.

~4

~

P 1-w,

,,,4.

-N-

%.O GE Nuclear Enemy l

1 SYSTEMS' INTEGRATION ENGINEERING 1

$AN JOSE,sCALIFORNIA--

RAPIFAX= '

(408)1925-1289.

OR. DIAL COMM 8-425-1289

)

TELECOPY TRANSMITTAL

.l Ma*7/FE II DATE:

u

-T0i c.

AhvA-Ad"t-)

~

. V c. /%>w spy /6. / css /y Gase)

FROM:-

A /#;4*M NA4 )

a THIS TRANSMITTAL INCLUDES COVER SHEET +I PAGES COMMENTS:

/dm e+x 3/ y n

?# M (p p ' &

&4

+

s rk

,f:in

• 6; v.

W

^

T --

+ -

m nn e-w r

~

=

4 f

t9205200162 920427

-PDR TOPRP EMVGErJE

)

.C:

p99 4 6-925-1289 04-27-92 07:00 PM PO:

~..

f(R 27 !92 C4863PM GCC R&AS P.2 7

c

~,

3 April 27; 1992 QC B.G To:

Charlie Hsu cet JD Duncan Chet Fostlosny trookhaven National Lab S Visveswaren Glen Kelly 1.F Ytedvick i

.From:

PD Knecht

' subject:

140Am Outsida Contaimmt in ABVR

References:

1.

" Followup on Open Items from the ABWR PRA prEn and the March i

heating in sen Jose". L9tter Kelly to Duncan, April 9,1992.

2.

" Physical Properties of Fluids and Flow Characteristics of valves Fittings and Pipe", Crane Engineering Divisien. 1969 Reference 1.$.ncluded several questions regarding 10CAs outside of containment e

and the bypasa study included in $ action 19E.2.3.3 of the ABWR ISAR.

The

-follouing responses are provuled to these questions.

Please note that due to AAWR SSAP revisions Tables "19E.2 12" and "19E.2 1)" should refer to Tables 19E.2.20 and 19E.k.21, respectively, and Tigure "19E.2 8" should refer to L Tigure 19?. 219.. The statement of the questions provided below have made

'these corrmatiens.

h ranea 1 Ouantiana "0 49 1;

Sons of the bypass probabilities listed in Table 195.2 21 oppear to have been undereettuated because coman cause failures do not appear to have been taken into etnaideration.

For example, when calculating the bypass

probabilities of feedwater line, SLC injection line or the vacuum breakerr. comaon.cause fatture of check valves appears to have been.

ignored.

RD PONSE

~ Comon causes were considered in the estimate of some failure probabilities listed in Table 19E.2 21.-

For example PS was assigned a value of 1.0 to reflect the comon cause potential for loss of all AC power 6 ring a station Blackout avanti P1 includes consideration of a l

commen ;;use. affecting both MSIVs in a sin 6 e line. A commor cause 1

failure emong i. hack Valves was not considered.

With regard to check valves, industry failure rate data associated with allowing complete reverse flow was und.. Only Teodwater and the SLC paths contain more than one check valve.

If common causes are conpAdered for these lines (with a Beta factor of.18),.the bypass probabilities for the. lines would be increased by a factor of about 21.

llovever, due to the low contribution of these lines to the total the y'

-total Bypset, fractiet would only be increased by about.08%.

Thatefore

. such conaon causa effecta can be considered' insignificant.

4 e

.PPOM 406-92',-12J9 04-27*92 t7:00 PM T02

p g_

y

/,

. As indicated by Aq. 6, CE's analysis is based on the presumption that a 2

' core damage event has occu red.

It is not clear, however, whether some of the dtta such es J13, p14, and Pl$ shown in Table 19E.2 20 represent the failure probabtisties before a core melt or the conditional failure probabilities, given a core zeit.

RESPO:fSE The values used are conditional probabilities, given a core nelt.

In seneral these probabilities are not affected by the core melt.

The break failura rates were detsreined from VASH 1400 which provit.ed a mean break failure rate for a line less than 3" of 8.fE.9/hr. segment.

The failne rate of a larger line was given to be a factor of ten lower.

For an individual bypssa line, the line was assumed to consist of four segments outside of containment. Because it was presumed that an undetected furoak in an unpressurized line could c: cur at any time, the conditional probability of a bypass path was then taken to be the same as the failure rato during a one year period (which was estinated to be 7000 hours0.081 days <br />1.944 hours <br />0.0116 weeks <br />0.00266 months <br />). This approach of estinating p! a failure probability is P

judged to be conservative and the consequences of an unisolated 1.0CA outside containment is considered negligible (see response to question 4, below).

3.-

It-appears tnat split tracticas (a crucial parameter in obtaining CE's results) were calculated using Eq. 12, which was derived from Eq. 10.

The detail of how Eq. 12 was actually used to obtain split fractions shown in Table 19E.2 21 is not explained in the SSAR.

For example, no information was given regarding the actual numerical values used for the geos.stry dependent expansion factors, Y, 0 4 Ao resistance R

coefficients, K-for the broken area, ais af we penetration lines. No mention was made of how the different!rt noture..dP which is time dependent, was evalus.ted for each of % y;stration lines including thos's leading to the suppression pool.

'RESPONTE In tho' evaluation of flow split fractions in Tcble 19E.2 21. Equation 9 was evaluated using a coreputer program developed to ease input and calculation. Ths most significant assumptions (including a discussion of the dP used) vere included in notes listed in section 19E.2.3.3.3

-(Page 19E.2 32) of the SSAR. Other valuaa used in the calculation are listed in Table

'ow:

e 6

sf "

lable 1 Additinnai samumariana in Flev fpiit _ calculaelen Paramater Amaumad Value_

Emain Resistance Coefficient (K-fL/D) friction factor (f")

.011 to.014 Reference 2 (Pg A 25)

(Stre dependent)

Line length (L) 83 ft/5ft Note 5 Line Diameter (D) various Line site (Table 19E.2 1)

Other resistances (K)

Referenca 2 (Pg A 30) gate valve 13 check valve 135 globe valve 340 Entrance offsets

.5 Exit effects 1.0 i

Expansion Factor (Y)

.6 to.9 Reference 2 (Pg A.22)

(dP.K dependent) 4 Since GE has already identif!ed the major bypass paths (see Table 19E.2 21), it should be straightforward to identify those piping systems outside of the pressure boundary-whose brosk can lead to loss of coolant that is not automatically isolable. A simple fault tree analysis can then be performed to estimate the frequency of LOCAs outside of containment.- Event trons similar.co those shown in rigures 19E.2 19A through 19E.2 19K cen also be constructed to estimate the frequency of LOCAs outsideLcontainment. Once the frequency of LOCAs outside containment is determined, a LOCA avsnt tree can be constructed to

analyze the associated core damage sequences.

RESPONSE

The evaluation of bypass paths in section 19E.2.3.3 is based on consideration of the relative contribution to offsite-risk-rether than coreLdamege frequency. The approach used was forased on the r(1=tive frequency of releases which would have a high-associated source term due to a lack of suppression pool scrubbing. The frequency of LOCAs outside containmer,t een be stimated from the information in Table 19E.2 21, but several cansiderations make this approach not as useful.

1).Not all bypass. pat 1s require a LOCA outside of-containment; An open Main Steam line. for instance, can result in condenser failure which is not traditionally considered a LOCA.

2) Bypass paths from the Drywell do. set cause a trancient and are only of eignificance following core damage.

The evaluation showed that there is more significance to thesa paths than LOCAs outsido containment from the standpoint of risk.

-3) Ignoring the effect of flow splitting over estimates the risk

~

i:

of LOCAs outside containment.

L I

.EOM 400-925-1209 04-27-92 07:00 FM P04

, ffR 27 '92 04855PM GCfC RtAS P.9 4

If the suggested approach were taken, the initiating event frequency for LOCAs outside containment could be based on the bypass probabilities for Intermediate and Large lines from the RPV incated on Table 19E.2 21 after adjust.i.ng for Comnon cause Failures ano factors previously introduced for SB0 events (see attached markup Table).

This approach results in a total initiating event frequency for 'inisolated LOCAs outside containment of about 2.6i'.7/yr. Applyin6 this frequency in an event tren similar to Figures 19D.4 13 and 19D 4 14 yields a core damage

. frequency for unisolated LOCAs ovtside containment of abvat 7E.12/ year' and orders of magnitude less than the total core damage frequency. This evaluation also ignores the benefit from the flow splitting effect which provides an additional basis for excluding these lines from further consideration.

O l

a PFOM 40P-90!-10!9 04-27-92 07: 00 TM 105 j

t k,,

AP 27 '92 Vs155PM GDC R:.AS P6 oennt mecine compeer ABM PROPRJETARY1NTORMATION HAstars Standard PInni c w in N

,e C

Table 19E.2 21 7

Suminary of Bypass Probabilities

+'~

IJans from the RPY av s

'1",

Bypus PlowSplit Probability Bypass Bypus Figure EAdim fIntion Janalbm.. pg,,A Er.otaMla ZIAchm lidL2.12 f

Malm Stoast 6.7E 1 4'Pl*

+ PS)

(63k5'/,t&7 1.1E 5 A

, Nrrr* + raj

,,l g, # " ' : n '_: i,_

_L.i 4.4a 2

O7

-A dg '

cr# - *.#7 Fcedwater 3.2E.1 2'P9'PPP15 (2@ *

• MF $.8L10 B

q p # =~r 'r !Ae.

.1.3

.,u + w s

.c? 5 isra a

1 t

HPCF Discharge 1.1E 1 2'PP'P10'P14 1.3E 7 1.5B-8 C

p y# %^T"'xm p =

1 ^? ?

r:0 ria r;.a

.r;; -

7,;;;-i; c-yM "c '$- t-I

% Ac.t

!.Mara:g-1,7g.,

(3.ct-Sb id RCIC Steam Sopply 6.9E 2 1')('P14 (Hr5/'8 M 1.1E-6 E

)

PJUL LPCI Discharge 1.7E 1 2'P9'P10'P15 6,7E 8 1.1E.8 C

, ~# FC '"..pMai.- -1.0E4

. 2:Fid=?Wrn

-E.74 6 7.GE u C

i

,j (s.6 DSM W)

RWCU $tution 1.2E.1 1'N'P14 SfBc3 M *i 10E 6 E

I

(

F *C' ' '-" 'h;.

3.ie$

v rd'3"y a.un-o

-2J510 3

W F

PoeWW4-hr,-"*?

1 ^E

• 4 rs'P13 9.6E 4 s.L7 J

l I

LDS ' "- r r:

?.?! !

0'.'" JAM t.00 ' -

s,, c.. i e

-D

(

reta :

5,6,(-7 l

,u Total 1.5E 5 o

l,l These lines may be excluded for station blackout events 1*

l."

Amandtnent 8 19E2.o?

,4.

m

, a m 3 m es w.

mi u