ML20101K533

Forwards PRA Input to Itaac,Item PRA-2 Re Common Cause Failure & Item 19.X Re Tier 1 Treatment of Design Features Identified as Important by PRA

ML20101K533
Person / Time
Site:	05200001
Issue date:	05/30/1992
From:	Duncan J GENERAL ELECTRIC CO.
To:	Kelly G NRC
Shared Package
ML20101K532	List: ML20101K533
References
NUDOCS 9207060008
Download: ML20101K533 (11)

Text

-

. a tir g# y p e

[ GENideeEnergy ABWR

,$ Date 1* *AY U?

To dLtS/4 u/ .

pgx yo, AAc Thispageplus l0 page(s)

From dM'9* Mali Code 175 Curtner Avenue San Jose, CA 95125 Phone (408) 925- Yf FAX ( 88)925-1193 or (408) 925-1687

~

Subject f'M spd 4 n ue. I L P M .3 cm m o>J o>M Psuuc 7% c3lof Message - -

76 ko A=rpox r<o oG//a ,'

t % 5 ,p d d> ,z % + c : pi / 4 y fxl4, sb abs & L ~~r;  :

SSM*& dS' /6~ PAfW P4 .4 , /, 2, 7.

~

hh ADO O 20 001 A PDR l

o r en c1nuu 3 mm_2rL___m ---- _

i Preliminary Draft 4

PRA INPUT TO ITAAC 19.X Tier 1 Treatment of Design Features identified as important by the PRA.

As the PRA was being finallred during NRC staff development of the Final Safety Evaluation Report, the PRA was resiewed to idendfy the most important PRA related ABWR features. Thejudgement of several engineers was used to identify those features and capabilities which are most important in maintaining a low core damages frequency ,

and in mitigating the consequences of an accident should one occur.

The results of this review are summarized in Table 19.X 1 through 4, divided into  :

4 major categories: Prevention of Core Damage, Avoldance of Supprenton Pool Bypass, -

Maintenance of Containment Integrity, Minimize Threats from Floods and Fires. For each feature, reference is provided to the corresponding verifying ITAAC by indicating the system number followed by the entry number in the corresponding ITAAC table, In addition, key subsections of Chapter 19 are identified to allow a reviewer to appreciate the general significanc'e of the feature beyond that identified here.

a h

9 s l

i. 1 l

b 6#

• I e _

, - , . _ - - - _ _ . . _ , . _ . . _ _ . . . _ _ . . . . . - - - . - -_._.,,_.s.,... -

J *

? -

Table 19.X-I. ,

FRA INrtTrTO TrAAC: PREVENTION OF CORE DAMAGE Chapter 19 (Notes For Now)

General Capability Spctific Feature / Capability Subsection ITAAC Reference RedundantSystems

• Three separated divisions of 19.1.2 2.4.1 (RHR) - 1,2, 3,8,9.18 .

ECCS and decay heat removal. 19.65

, ECCS pumps able to pump 19DS.I1.3 .

saturated water 2.4.2(HPCF) - 1,2,3,4,11,10 ~ -

• RHR vessie injection valve which 19J.3 2.4.1 (RHR)- 7 admits fire water to the RPVand drywell spray valgt have handwheels forlocal manual operation without power.

• Automatic depressurization for 19.1.2 transients and LOCAs Divenity

• RCICcapabicofoperation for 19.1.2, Number of hours not critical from PRA several hourswithout AC power, 19F973 importance view (because of fire water) and ability to override switchover 2.4.4 (RCIC) - 6 says isolation fails as is to makeup water source from on loss of ac. Am=6 WJEufw m

"* M 2.12.12 (Direct current power supply) -

later

• Combustion Turbine Generanor, 2.12.11 (CTG) -1 connectable to atleast one of three safety divisions to provide M Power h 2C .

~

- 1

. i g d i d M **d k S4Ifase d Tab 4e 19.3C.1. -

posJa

) -

PRAINPtTrTO ITA 6C: PREVENHON OF COREDAMAGE (ch) 4 I

Chapacr 19 (Notes For Now)

General Capability SpecEc Feati we/ Capability Subsection ITAAC Reference i

i

• Sciwnicallyq lified AC a c. 19.1.2 2.4.1(RIIR) - 7 independent act addition -

system. inclu dedicated 2.15.6 (e ,m)-N g3

, devene diese manually e

operabidvalves. Calculation fW

= g 4 1 -

1

[- - for vesselinjection, k;-ca. {

and with RPV pressure at i r

- for . k a-- .

,l and with drywell y.-- at j i . t l

Minimize Poacneial for

• Reactor Protection System-RPR 193.13 2.2.7-laser by RPR Failure ao Shusdown please define this one ,

• Alaemate rod insertion system Laecr byRPR i (key featureslaecr, probably i already tm,d) m - Standbyliquidcontrolsystem 193.13 2.2.4(SIC-by RPR 7 (key feaeures later, probably a) t already f.

[

r

\

IE -

i, I

• nt 2

s

3 i

i A .

~

.i 1

I

' t

+ -

hM i "rable 19.X.t.

m

' FRAENPtTFTO TTAAC: AVOEDANCEOFSUFFIWSSION FOOLBTFASS Chapeer19 (Notesp' Now)

General Capability Specific Feature /CapabiEt, Lharction ITAAC p ence

!. Avoid Unisolatable RWCU

• Reactor water clean-up Isolation 19.3.2.6 2.6.I(RWCU)-3.% RWCU EQcassy

Break Vahes suust be ~p% ~

in Table 3.0 orI~eer I anaecrial.

cgualified (including scissnic)" for  ;

i: expecsed duty

,2- Control Uni-dar=Me 1

• Reactorwaserclean-upsuction 19.3.2.6 2.6.1 (RWCU) -4

$ RW C U Break, nozzle noust be at least 5 feet j* .

above the planned cievation of 6

- the top of she active fuel.

a l'l . Control Uni-J===Ne

• Reacearwasercleanupdrain Ene 19.3.2.6 2.6.1 (RWCU) -5  ;

RWCU Break tie in no the maction line sumust be  ;

4 at an elevasion capsal ao or above ,

1 she macsion mozzle.  !

t Avoid UM===Me RHR

• Seisnaically qualiGed RHR 19.6.3 See RHREQeneryin Table 3.8 of Break isolasion pool =h rJoe  ; Tser I naascrial j i i

$ ~

I

!?,:

i:

lE N-  !

iK l 1

3 3

I

---.-e

. . 4 . , _ -_ _ _ _ _ _ _ _ .

v -

^

. l .

Tab 5e 39.X.3.

FRA ENFtTTTO FTAAC: MAINIINAN(X OF CONTAWeMBfT INTEGRTIY .

Chapeer 19 (Noecs For Noor)

General C- ^:- SpeciGc Femeure/C ,- "':, -

Sehecceiose ETAAC Reference Avoid Hydrogen s telM

• Provisions se provideisecreed 19.6.6 2.146 (ACS)- I 1 Threses containsment 19.6.8 Avoid CM= cent

• Containementoverprmn 19.2.4.3 2.144 (ACS) -5J-6.-8 g SerM Failmsse proeection sysacne wish napsesre g ,

disk. set-point esaablished at L

90 psig and nosnimal ihnr raec of s: XXX when containwient pressure b is YW. (XXX. YW. laarr fuese ,_

CEB).

-- '- _ _ **Y %

f Missimire(' "_-gc so Containsnest Hooder syseene: g , No TrAACsecaissa yet.

- M (number) vahmes which open, kneerdrywels :- ; - _

w et 509"F

- YWoosnimal ihmr raec pervabe ,

- N.YWlaner frosse CEB Mainae==a of

• RHR heat-:--E jM 19J.3

- , , - - Pool tweegrity

~--

capacity E

1 4

_ . _ ~ . - . - - - . . . . - . . . . . . . . . . . - . . . - - - -

4 e

0 m.- - _ _ _ . . . _

Table 19.X4 .

FRA INPUTTO ITAAC: MINIMI7E THREATS I

OODS ANDFROMHRES IN I t KNAL H GeneralCapability  !

Chapter 19 '

Hooding Sgulic Feature / Capability (Notes For Now)

Su b 6on

-

• Normally closed watertight door ITAAC Refc.cuce betwcui turbine builtling and 19Q. tater i

Later by McSherry, Ehbert service building tunnel.

g

• Control building kr u floorIcM 19Q. later sensors which alarm at 0.15 meser Later by McSherry, Ehbert and trip RSW pumps and close g j

RSW isolation values in affected division at 0.8 meter.

• ECCS rooms have waser tight dem which open into corridor 19Q. tater Laser by McSherry, Ehbert

• Reactor building cc idor (Roor 19Q. laser XXX) volume sufficienflylarge btcr byMcSherry,Ehbert (YiYcubic rneters) to contain largcx nood source 1 Fire l bterbyMax -c!!,Ranc.if i

i

~

5

I .

Tabte 19.X-3. -

F5tAENFtJTTO FTAAC: MAINTD4ANCEOF00tGANGE!NTDrfEGELTIY -

l i

w Chapeer 19 (Noecs For Noor)

General @Mity Specifu- Feasesre/CapabiEty Seebeeceiosa TTAAC M.uxe Avoid Hydrogen PA= ness . Provisions to provideissereed 19.6.6 2.144 (ACS)- I

~

~Derezas connaisennent 19.6.8 Avoid Casseminneeset

• Consainsnentoverpresseure 19.2.4.3 2.144 (ACS) -5.'-6.-8 Serenceural Failusse proecction syssene wiele rupensre

/, . disk ses-point established at j 90 psig and noneisaal floor raec of XXX where containement pre ===e isYW. (XXX. WY. Immer frosse CEB). _D.Scepoint neay change.

a Minansre N" Casseminsment

, es - *

~

Hooder syssene: g p NoITAACsecsioso yet.

t (noenber) vahses winicle open, -

l' looserdrywell a ; _. - m exceeds 500"F

- YWesosnimal floor rase pervahe ,

- N.YWlaser frone CEB Maineenasoce of

• RHR Incat ---- _"- - j eri==mir 19J.3 Sempyression PoolIsseegrity capacity  ;

Jl i

~

D 4

g, - gg,, - ._ w wTf

Tame 89.E4.

FRA B8FUT*IO rTAAC: MDGdEZE*11ERFA13 FROM BfTEENALMDODS AIMD HRES l' '

I -

Chapeer 19 (Noecs For Noor) cencras capahisity specir.c Feae e/capaisiity - S.b.ectio. rrAAC Reference W -

= NonnaHycknedwaecraightdoor beineen turbine buisping ==8 19Q. Iaeer LaserbybecSheny Ebbert

, service building sunnet.

l ,

• Control 1-2 " y,looserfloorlevel 19Q.laecr Laser by McSteesty.Ehbert

' s,

. sensors which alarna at 0.15 inceer

and trip ItSW poseps and close -

. :.- RSWisolation valuesin afIcceed

- division at 0.8 nececr_

j ..-

• ECCS roosns havewaoer tight 19Q.laner I =arr by McSteeny. Ehbert doors which open into coner

- acacearboiadingcorridor(noor ise. inner n .by W.Ehbert xxx) vai-ne suaiciemey annue onnrc bic.c ers) no contain largest flood sonne Fore  ! Laser by ManseM. Rahery l

l .

.g.

l ..

.  ; 5

_--m_

. t. , "

n 'Si

3(

May 2 8, 19"'-

nm A

J.D. Duncan .

M/C 754 Enclosed is our response to the NRC comments C-3 and 0-5 on the treatment of comon-cause failures in the ABRR PRA.

The response consists of reevaluating core damage frequency with the addition of co mon-cause failures at the component level in the four system fault trees that use redundant divisions or trains. The CDF of the CCF run is 19.2% higher than the CDF of the (revised) base run. This increase is not very significant. The major contributors to the increase in CDT are inter-division CCTs of the reactor building cooling water system, particularly due to their affect on HPCF and RHR (core flooding

. mode).

. u pu ;

l'L.G. Frederick M/C 489 ,

cct R.P. Raftery ,

0 e

O

+

4 h

I j Rgggenaa to NRC Outstanding Itama c-3 and e-B outstanding Items C-3 and 0-8 requested further analysis of equipment common-cause failures at the component level. In discussions with the NRC, it was agreed that an acceptable approach at this time is to perform i

the updated PRA Level i requantification without addressing CCF (with no l

additional common-cause failures), then requantify with component-level i CCTs added to see the effect on core damage frequency.

Ccts that were already included in the hasa quantification aret EsF logic backup scram relays transmission network (MUX) pressure sensore sensor and transmitter miscalibration APRMs.

output logic unit diesel generators digital tria unit batteries -

trip logic unit offsite power source main scram load drivers- safety relief valves These CCFs were retained:in the base run. /

Component _CCrs for the following systems were identified, evaluated and included in the G2 run (in addntion to those above): ~

S' "

MPCF (2/3 trains) (14 componente) l RHR RHR SPCore Flooding Cooling Mode Mode(3/3(3/3 trains)trains)

(25(24 components)*

components)

• RBCW (internal to each division) (4 components)

RBcW (between Divisions A & 5) (6 components)

RBcw (between Divisions- A & C) (6' components)

RBCW (between Divisions 3 & C) (6 components).

RBCW (between Divisions A, 5 & C)- (6 components)

These are the systems where component-common-cause failures might have a i significant effect. RP8, CRD, Electrical, and Instrumentation and

, Control systems already include component CCFs in the base analysis.

The component CCFs that have' been added to the system analysis include' pumps, pump auxiliary equipment, manual valves, motor-operated valves, check valves, room air conditioners, spargers, strainers, circuit-

, breakers, flow transmitters,_ heat exchangers, and temperature elements.

For the RBCW CCFs internal within each division of RBCW the component ccrawereadded'it'appropriateplaceswithinthefaultkreestructures.

For all other cases (inter-divisional"or between trains),--the individual component CCFs were summed and added-in at the-top as a CCF. module. The

-RBCW interdivisional CCFs-were added-in;at1theitop of the fault. trees for all systems that use-RacW.  !

component CCFs were identified wherever redundancy occurs in the fault  !

trees (generally, for every "and" gate). The component CCFs were quantified using the " multiple Greek' letter" method, and using the.CCF facters given in the EPRI-ALWR Requirements Document. Where common-cause-factors were not given for specific component types, the recommended l.

,,m, ......,n n. . w. w n .

• w *an

"generien factors vere used. For those cases, the results should be 4

considered as abounding" and are probably conservative.

The numeri' cal results of the analysis in terms of cCFs are given below:

HPCF CCFs (2/2 loops) 2.462-3 RHR Core Flooding CCFs (3/3 loopr) 1.00E-3 RHR SP Cooling CCFs 13/3 loops) 9.78E-4 1

RBCW CCFs Within a dLvision (1/2 pumps) 1.53E-5 RBCW CCFs between Div. A and B (2/2) 6.48E-6 RBCW CCFs between Div. A and C (2/2) 6.48E-6 RBCW CCFs between Div. B and C (2/2) 6.48E-6 i

RBCW CCFa between Div. A,8 and C (3/3) 5.93E-6 ,

The numerical results of this analysis also can be viewed from two different perspectivest the effect on system unavailability and the l effect on core damage frequency. Effect on system unavailabilityl System Base 3I SIwith CCZa t increase

. HPCF 2.33E-3 4.79E-3 105 RHR (flood) 9.65E-5 1.10E-3 1040 RHR (cool) 2.72E-4 1.25E-3 360 RBCW Div. A 3.09E-4 3.24E-4 4.85 RBCW Div. B 3.09E-4 3.243-4

• 4.85 RBCW Div. C 3, . 09 E- 4 3.24E-4 4.85

• CCFs within that system The effects of component CCFs on system unavailability are significant.

. The most significant effect is on the core flooding mode of RMR, where the system unavailability with component CCFs is over in times the system unavailability without component CCFs. The largest contributors to RHR CCF are connon-cause f ailure of the RHR pumps to start, and common-cause failure of the pump room air conditioners. Common-cause failure of the 3

, injection valves to open is also a significant contributor to RNR CCF.

For the HPCF system, the most significant CCF contributors are common-cause failure of the pumps to start, and mispositioning (closed) of manual valve F005. The reason for the large CCF of the manual valve is because of a very high assigned random failure probability (1.0E-2) as taken from WASH 1400. This is an unreasonably conservative value.

The individual divisions of the RBCW system Vere not significantly affected by component CCFs. However, the interdivisional CCFs have a measureable effect on core damage frequency, as discussed below.

m

,(

Effect on CDFt Svataq CDF Increams M CIAAAA HPcr 5.6E-9/yr 3.6 RHR (flood) 4.9E-9/yr 3.1 RHR (cool) 5.4E-10/yr 0.3 RBCW A,5,4 C 1 90E-8/yr 12.2 TOTAL 3.00E-8/yr 19.2 The most significant effect on CDF is due to the CCTs between all 3 divisions of RBCW. This is primarily due to the failure of both HPCF and RHR Core Flooding, gi'en v loss of all RBCW divisions. All other CCTs have very little Sffect on CDF.

The common-cause failure of all three divisions of RBCW is balanced among common-cause failure of heat exchangers and common-cause failure of pumps (failing to run). Plugged strainers and temperature control valves

, failing closed also contribute to the RBCW interdivisional CCFs.

et .

O N

1 3

..... ...... ... -e .. . e, _

- ...s u m m >