ML20105A904

From kanterella
Jump to navigation Jump to search
PRA for North Anna Power Station Svc Water Preservation Project,Part 1, Supplemental Rept
ML20105A904
Person / Time
Site: North Anna  Dominion icon.png
Issue date: 09/08/1992
From: Afzali A, Donovan M
HALLIBURTON NUS ENVIRONMENTAL CORP.
To:
Shared Package
ML20105A897 List:
References
PRA-920908, NUDOCS 9209180041
Download: ML20105A904 (39)
v • d • e


Text

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

j 0 1 i 1 2 1 3 1 4 i 5

";*'"* ' ' ' ^ "

SEAv!CE WATER SYSTEM

  1. ~"'#"'

NAPS UNIT 1 320MAF.N 13 SW2 i4 0  :,5.g i a c ,,,,., ,,  ; . ..w . ., , , ,

w ,.,..

-s E As*

I l l I

-q := , ag .t =J att 5.* j ra ,5 ;4t= t-se-a.,g j

, . r. . , a r a h l 13 RJ4

$ i'se;mv.f0-25o10 15mPAI-FR 15W10 3 44 .;.4 7 9 R 034 24 2 .

l 3

mese 5

l 6 1

7

.C

  • 8 1

l

~

4 -

3 3 9 E

5 9209180041 920911 C-44 PDR ADOCK 05000338 P PDR

~ , ,-.

0 l 1 I 2 1 3 1 d I 5 SERVICE MATER SYSTEM NAPS UNIT i .?MAF N.13 SW2 15 i 0

. .. , n .. j ma, aw - &,, ca 3a .

L0is > O!S>a4:4 l

&a'a 8 3 *4 arte a u t *eO f st af 4a )

G$a23$t2 I

N t 5t %t 4f P a !. #E se at%

7 T4 SPGar M atta 4 (31 19

&&a f Ga'a5 ma l*,t t gag;E 2 GSs2tt22 ILwPlP-UM M KJ 2 S4 402 3

1 I rahx y ea;,pg y la e5025 aQ 96 18* e5 C9] aw0 94 c a ' .4 &a tivt G5s21$31 G5 mil 533 I

f 2 8e -u3s 2218 2-se-w;t i228

$3 Aiov$ tt03t0 $mA!3us CLCSED No %G e15S:34 DWING s!5513N d

2SewJv*$C-2218 2$ses0v-$C-22PB t ?!!-03 , Pa 1 ??t-scs , is

~

I t Es-m;. tija t - t o - m;v - 12 2 a Aufl1 CtOSED SPAIOLS CW40

ie!NG *1%1%

34419G ul5$1w 5 is,,oy.5:.,23, 33,,,oy.gc.322, i ric ! , as i vi-n , is 6

7 S

i w!

N J

2 9 3

L-45

s b i 1 l 2 1 5 1 d I 5

    • 2"""'**

SEAVICE WATER SYSTEM

""* SW2 16 w s usir i 3ao w .u.i3 0 G5=2sbot ....t.. .. l: q..;  : . .t 5p_,, o, , m w ,- 9 u 1

su ,a i  ? 22 j

~

l I j  :-f = i nt et < a uf A.*

i 8 &li$ fias 1-9* #* 1B j He- 2 F alu s Q atn I 15s;k v 4 0- 15s 10 15 md a I 4 R- ISud t B 3 J4 ::1  ? 3 1 -00s 24

) ,

i1 4

2 2

4 3

4 s

e name f

't m

i 1

5 1

i I

6 i.

7 I

W e 8

I i o 1

~$ -

2 d

5 9 i o

?

4 3 I C-46

0 1 1 1 2 1 3 1 4 i 5 SEAVICE WATEA SYSTEN NAPS UNIT 1 320WAF.N.13 0 , . _ ,. ,, i g 7s . .t . ,_ _,_  ;.7 , ,

s , ag

~

=

t' 55 f * .;>an:2 an's F;n d a^A4 0 ils *Cf ef a t 4 3 1 1 cssas7:2

- I

'q :w: vet n %t se a'%

i s ti >E 0F 4! 4 a .t R 9 28* 1*

tn.a v ra t-g es :N'INaw; I 2 r Leit 722 t i ep lP-Lee-<f@~

2 M ;G7

)

fTI I l I F a :L Al ;*

as:.M tv 16 '- e3 -ts) a%C 61 16*-e%-082 ama 83 pa r 3 &&fe5 3 GSe21731 G5m21733 I

t- se sav-1226 1-So-asv-1218 r# A 0M CL;5ED SP/13US CLO5ED CJ!% =!i5:0% Ount% e:St!]N d 15anov-SC-1228 15stsov-SC-1218 i et 0 9 , is i vtqu , as I _

t s o -=c v -2 22 a 2 5e-*;v JJia SP.R!M CLOSEO SPA 1;uS CLOS (0-w:% als5!DN 3 alNG *Iss; w 5 23s,,0y.sc.2224 2s=wov-sc-2214 i at aci , 2e t af E5 , Je c

6 7

t f .

I o

i a

$ 9 i i

1 l

+

l C-47

o I t I 2 i 3 4 l 1 5 SEAv1CE WATER SYSTEM euss uun i 32 m r.u.i3 SW2 18 1 asat sst da l = nt a t t ,N C a r 'a t* " q ,;5;w . ; a- s I

ra;LLNg ty a,a ;.  !

sit 44 804 4a24

  • {

I G5mdtBt3 l

~~

l I I se a'N Fa!L>( y r AIL.ng 0F M aDE A & 131 14 6 -Sd-=C v - tic 4 10 2-Se-uCv-Pha f0 i ma ;%'E %AW E M4 03'1 4 2 iSwFIP-UM-4 RA GSs21822 GSu21824

~

l I ex cn TI A 'D W AL yt No ELECT A1C PCaf.2 uGTDA TEEfD vatvf No ELEC1EIC Katu j Fall A CLCSf 3 493 i mCC tal-1 Fa!LS CLCSED 480v mCC Pt 1 1 $a-unv 1204 1 -f D-*C- 13 2-So 80v-2204 2 f 8"#C- 10 3 35wasov-FC-120A GClH612 2S wano v-F C-220 A GE2H512 1 CX -004 t 1"00 1 0 A 002 3 Lmo e

i 1

1 i

,i 4

Fa tLW y AM Su ate .()n Foe ga%4 5

6 cs,,,,63 i

~

I I I Su atu ratLM y t a!' vAE y 4 &OEA 8 141 14 1 La-*0s-12C8 2

  • 5e-w0v-22Mt na tuttmanct tSmPIP-t>_4 7EI 65m21872 GSu21874 2 S4E 062 3

i

~

l I motDA '1PERfD vaLvg 40 (LLCialt PCwE A =0f04 UPERfD vatvt W ELittaIC E n(&

! f a!LS CLCSED ago y act twt t f att$ CLOSED 490v eCC Nt t O 1 to 804-12C8 t -(P at-10 2-Se-uCv-22C8 2-tD-at to j 8

m al ismiov-rc-tron cttHela ti ce assuov-sc-220e i onm2 GE 2H512 exo 3 3 muu n

9 s

3 C-48

l 0 1 1 i a i 3 1 4 6 5 -

BEAAING COOLING SYSTEM 0

n u.t inte BCS i asa,,st sa i :ng a r t:ee ;a rt :6- t v-92 atelstan 06 17-62 _ _ _

1 tmaC(Aart K satts FLCs 7014 (13 CHILLIAS MCSt:3 1 I I Es in malNf twaNCE SCS falls 3410 watut f arts e4N Ef23 Dup!=G ause aC/Ji StaAf 34 e opgaatton f attust 3

18CS.- *D-ma !NT GSCS133 G4CS134 *(4- 1 -4C--CPE N g c) g .g.

mas;at vatet e&Nuat va(st Fa!LS CLOSED 6FatLS palls CLQED irstts TO came 70 (P'85 d

100rv--#C-10CA 10Carv-FC stC8 s M ea t Mea l

1-SC-e ta Fa!L5 70 t-9C 8-It Fa!LS 10 etm t'aat 5 cacstst aestsi e

6 7

I I

=0 ALI. PJ@ ..LDSS 7 0F75171 4 AL1 RAF g

Fatts to staat FatLS 10 atse 0 10CPaf-FS-tGCit t -EM WP-24 tKPaf451-48 Cit

= i mei a twee s sees ies e -

I 3

q att mpp L;55 0F 7751ft pengpt .

3 Fatts to sus .

9 t0C7afs m-secta s-( M ooP te

{ s s4*> ies g a .iw es

-C _ 2_am mmmm _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ = _ __

Figure C-4 Loss of Fire Protection System Fault Tree i o I t I a t 3 1 4 3 3 FIAE PROTECTION System l r w mEE s,,, FPS i  ;

o

. a. .e .a m a .1 m. , .,,.: 4. :, m . .. ,, g j 1

%a
:4xart 'Cs to.

'4 mag 19GPt*

1(as c j pgggy I I I a # Fa!L W l eaMuat watet optmafon Fa!LA 10 FIpt Imptus( tes f.(

sa!L5 CL:20 #attl C0w(C1 f 4 FPS 70 se 40 Ovfdt sain 70 DPet f.( CM-8 CDOLins

~

3 West) tF pery--d C tF owva H P- t -8 0--0Pt se W P5 t J4 c) . ., .. g ,_,g c) i I i i

=0 f 32 DRivtm *.se 3:E!JL calvin abe stat apeg Fat 3 ativna o to Mu patL M ratLM tw T4 ag. a C:># 10 that rLos

'O OVf $ttX Fa!LS d

7P5140 GFP51st tsuptp-stp-aL180G 1516(56-aa-03f l G t 114. S E 41 5 repsis4

r. m r 1 43 I I L355 & OF711ft Cit 3EL Fa1.A 80s(R 1

6 3.tp.taty.g. l g esieg a iaea 1

C; tit. Gt4 Aafosi OttstL 2 4aaf04 Fa!LS to staaf FatLS 70 sue Gt4RIC SG10A CM.T 7 1(GtC4-FA- W if GEDG-FS-FPS 31 *d 6 2 attaa O!fSti DR! vim 8W uC%fPW pontA FatLW

~

Fa!LS TO Illpe Fa!LS TO Sf aat O tFAP&f =#f4-tFP2 IFAD4f-f$ tfp2 7 PS154

2 iise as ime

~,._

i i 4 att * # LO55 QF QFF$1ft @ A f kN 90 mon 3 rafts 70 cue Fa!LS f3 Staat

9 tr epar-s hir e s <pwoce-24 trasar es-ifat i sites as a iotes t mei C-50

s ruoanotees ter I 00 2ne s se-es woes e o mes sa, Laatir:rseisa tate 6-tB-47 1 30 the ? ?'at

  • ms e #
  • w .1M s.h.

1:

th 4 i ::-

e d-

~~

A C *(

5 58

'.1 .A '{i

  • 1, i

ti b

T j -

km  ;

$3 .

M~f

~'

I g '80,3

= b 1.<$.5 I k.

l O SE a -g .g .g e i

%? L' s' 0

at t' r r

,. t m .

x 2 -

-. g5 av

~ g+4 .-

4.

U g  ; {'3'in e.

o IIbc, i

3 6f92 r yi**

  • 1 wo *o ===
  • .- ) o-O M s ]. .

o I

-4 -

m 0

o 1

e 8 ,

i

<3 -,3 .i3 _

$ 3 3

  • i i i i 3 !s t

t

.cr -

- -. -s g a a s --

x 4; I  !),2 y ,a '

. n i A M .

E p

. *s:le m -

g-e e.

c.

e s

e

e. s. s s

f- I s. s s~ s .ame j

I I 2

n. r f i I E I I I A I- A X f-g

)% 2s  ; >

s > > > ~ -

e a 3 g

- 5 } 3 -

4

- v -

$g a.

  • y

% 9 2 a v i 2 s s a a  ; s e 2 2 s s s . moi s - o s s 3 ., s s s - .- ,

!3 $

!  ! !3 3 $ $ 3 h I h I k

8 3 3 3 3 3 3 3 3 3 3 9 m

A 4

4 l

]

i i

1 1

1 1

PROBABILISTIC RISK ASSESSMFNI' FOR NORTII ANNA POWER STATION l SERVICE WATER PRESERVATION PROJECT l I

i .

I, PART1 ,

j Supplemental Report

]

I. >

I i

i

Prepared for.

Virginia Electric Power Company f

1

by Amir Afzall 3

i  :

3 September 8,1992 i

i a

Approved by: M N-

! Michael D. Donovan

! Assistant General Manager Risk and Reliability Division ..

i IIALLIBURTON NUS Environmental Corporation 910 Clopper Road i Gaithersburg, Maryland 20878 3

.i l

W W

-y9ywr-7y r Y- s y p- e+,rt'e m we-%p e,---se *

  • 'py,,--n-e
  • t r @ .sh-$-w+--* -w imer *bn-i-.{ t1-' %r-- i y a-e4 - F # w-
  • w'

4 ABSTRACT d

j This is a supplemental report to the Probabilistic Risk Assessment for Part 1 of the North Anna Power Station Service Water Preservation Project. Part one of this supplemental report l

! represents the assessment of the change in failure probability of the unit 2 Air Conditioning system while its backup chiller is being supported by the Bearing Cooling (BC) system. Part 2 presents the justification for using the methodology (Log-Linear Model) which was utilized

! for quantification of the service water pipe rupture frequency.

The results of the analysis for Part One indicate that the most significant change in failure
probability of the Unit 2 AC system is a result of the possibility of a design basis accident
(DBE). The change in the Unit 2 AC system failure probability, during the period when the 3

bearing cooling system is used for the Unit one chillers, will be in the range of 1.4E-6 to 6.8E-6.

The conclusion of Part 2 is that the Log-Lineu Model is a conservative method and is appropriate for the condition of the North Anna Plant Service Water piping, i

1 i

a J

G l '

i

TABLE OF CONTENTS PART 1: Assessment of the Unit] Air Conditionine System Reliability Durine Comnietion of Service Water Preservation Protect 1.1 Introduction 1-1.2 Objectives 2 1.3 Scope 2 1.4 Methodology 3 1.5 Results 4 1.6 Concluxons 5 PART 2: SW Pine Rupture Freauency References Appendix A: Quantification of Failure Scenarios Appendix B: Pipe Rupture Frequency Quantification ii

}

PART1 j j

ASSESSMENT OF TIIE UNIT 2 AIR CONDITIONING SYSTEM RELIABILITY '

{

DURING COMPLETION OF SERVICE WATER PRESERVATION PROJECT

1.1 INTRODUCTION

I i

] I i This is an assessment of the reliability of the North Anna Power Station's (NAPS) Unit 2 Air

! Conditioning (AC) System during performance of the SWPP activities associated- with i replacement of service water piping to the Unit l's components. The change in the reliability I

j will be due to the change in the operational configuration of the backup chillers for the Unit 2 j control room and emergency switchgear room (CR/ESGR) AC system. Unit 1 is not evaluated in this study, since this unit will not be in operation during the isolation of the SW supply to the f

j unit's chillers.

,] The environmental qualifier. tion basis _ for the Unit 2 CR/ESGR AC chillers is the b'ackup j function provided by the Unit I chillers. The normal supply of water for the AC chillers is

{ provided by the service water (SW)_ system. The SW system is a safety related system. -Its l components are seismically qualified and are supported by the emergency power supply,' During .

performance of the service water preservation project (SWPP), the SW supply to the unit 1

chillers will be isolated and the Bearing Cooling (BC) system will be utilized to supply water l to at least one of !he unit I chillers. The BC system is not a safety related system and is nnt l seismically qualified or supported by the emergency power supply. This change in the source i of water to the Unit I chillers lessens the reliability of tlie Unit I chillers and therefore the r

backup chiller for the Unit 2 AC system. Virginia Power intends to submit to the NRC an exemption request from 10 CFR 50.49 for environmental qualification basis for the Unit 2 Control Room A/C chillers for the period while Unit 1 is shutdown and its AC chillers are being supported by the BC system. This reliability assessment of the consequence of this change in the configuration of the supply of water to the unit 1 chillers is performed to support the _

exemption request, I

i l

1.2 DJJECTIVES 1

The objective of this probabilistic assessment is to quantify the change in failure probability of the Unit 2's CR/ESGR AC system due to the change in the configuration of the backup system to the Unit 2 AC chillers.

i 1.3 S_ COPE

The source of the harsh environmental stress for the Unit 2's chillers is the main steam line
(MSL) rupture in the Unit 2 Turbine Building (TB). Steam released from the ruptured line can propagate to the Unit 2 Air Conditioning (AC) Room via a louvered wall interconnecting the Turbine Building and the AC room. The environ' mental qualification of the AC chillers is based on the availability of chillers in the other unit to provide chilled air. The major difference l

between the regular configuration of the AC system and the configuration of ths. AC system j during the SWPP activiti ;is the source of water supply to the Unit 1 AC system chillers. The following cases are analyzed to evaluate the stated objective of this study:

Case 1. A main steam line rupture in the Unit 2 Turbine Building is coincidental with failure of the BC system supply to the Unit 1 Chillers. Tnis case is analyzed under the following set of accident scenarios:

Accident Scenario 1- BC system fails to provide water to the Unit 1

chillers and before recovery of the BC system the Unit 2 AC chillers fail due ta the MSL 'iure in the Unit 2 Turbine Building (MSLRTB);

Accident Scenario 2- A Loss of Off-Site Power (LOSP) event occurs and before recovery of off-site power, the Unit 2 AC chillers fail due to the MSLRTB; i

2 4

e, -- - .e..y , ,, - - - - ,

Accident Scenario 3- The Unit 2 AC chillers fail due to the MSLRTB and before recovery of the Unit 2 chillers, the supply of water from the BC l

system to the Unit I chillers fails.

Case 2. A Design Basis Earthquake (DBE) resulting in the main steam line rupture in the Turbine Building and failure of the BC system. This case is analyzed under the following accident scenario:

Accident Scenario 4- A design basis seismic event occurs causing the main l steam line rupture in the Turbine Building together with failure of the BC System.

l

! It is important to note that this study does not condder failure of the Unit 2's AC system due to any cther failure mechanism other than the harsh environmental conditions for the Unit's chillers caused by MSL rupture in the Unit 2 Turbine Building. Also, no evaluation of the l effects of a design basis earthquake (DBE) on the failure probability of the plant's components and systems is included. The analysis assuc!ws, in an event of a DBE, all components which I

are seismically qualified will remain unaffected and those not qualified will fail.

1.4 METiiODOLOGY The change in the failure probability of the Unit 2 CR/ESGR AC system induced by the change in the configuration of the backup system for the Unit 2 AC chillers is given by:

Ug =[{ P3 Where U4e is the change in' failure probability of the Unit 2 AC during the performance of the SWPP and 3

P (i= 1,2,...)is the probability of occurrence of an independent damage inducing event.

Evaluation of the P,is carried out by:

1. Identification of the interdependencies, if any, between each postulated damage inducing event. That is, the consequence of occurrence of each postulated damage inducing event is evaluated to identify the systems / components potentially at risk.
2. Evaluation of the probability for occurrence of postulated damage inducing events associated with accident scenarios mentioned above.

The North Anna Power Station IPE [ Virginia Power,1992] data and models will be used to quantify all failure probabilities and the consequence evaluations, g 1.5 RESULTS This section reports the results of the analysis carried out to quantify the contribution of cases I and 2 to the change in failure probability. Details of the analysis is presented in Appendix A.

The contribution from each case is shown in Table 1.1.

For Case 1, the change in failure probability of the unit 2 AC system during the construction period of concern will be negligible and no further evaluation of the consequence is considered necessary.

The corresponding change in failure probability for Case 2 has a wide range. The range is from 1.0E-6 for 90 days construction period and using EPRI seismic hazard curves to 6.8E-6 for 120 days construction period and using the LLNL seismic hazard curves. The reasons for this 4

l

- - - _ _ _ - - - _ - - - - -_w

variance in the failure probability include:

1. The range of postulated exposure times of the plant to the possibility of occurrence of a DBE event. This variance is induced as a result of uncertainty in the length of time required to complete the SWPP activities.
2. The differences in the numerical value for frequency of occurrence of a DBE, using the EPRI and LLNL seismic hazard curves.

As can be seen from Table 1.1, the second reason has much greater effect on the value of the .

failure probability. -

The change in probability of failure for the unit 2 AC system is not equivalent to the change in:

core damage probability. There are several contingencies which would prevent core damage -

given failure of the AC system'in the scenarios analyzed. For the case of the DBE,'the core damage probabilit,' cannot be quantified without performing a seismic PRA of the plant. The change in AC failure probability can be considered a conservative upper-bound for the change in risk..

1.6 CONCLUSION

S The result of this analysis indicate that for Case 1, the change in failure probability of the unit 2 AC system during .the construction. period of. concern will be negligible and no further evaluation of the consequence is considered necessary.

The calculated change in failure probability;for Case 2 has a wide range, depending on the length of the construction period and more importantly on the difference in the probability of ,

occurrence of a DBE, obtainable from EPRI and LLNL seismic hazard curves.

FI +-

5

This change in failure probability of the unit 2 AC system is not equivalent to the change in the Core Damage Probability (CDP). There are many accident mitigating systems and measures which can be utilized to prevent a core damage event even if the AC system is inoperable.

However, even if the most conservative value for the change in failure probability of the unit 2 AC system (6.8E-6)is considered as the upper bound value for the change in CDP, the change is not significant.

Additionally, the results of the SWPP PRA indicate that, if the BC system is used as a source of water to the unit 2 AC system backup chillers, the change in the CDP, due to operation of the SW system in one header configuration, will be reduced from 5.lE-6 to 3.0E-6. This change in CDP is mainly due to providing ar. independent sou ce of water to the chillers which in turn introduces defense against common cause failure of water supply to the chillers.

Considering the above factors, it can be concluded that providing the BC system to the unit I chillers while SW system to the chillers is isolated is an acceptabie measure.

1 i

=

6

TAllLE 1.1 Contribution of Study Cases To The Failure Probability of The Unit 2 AC System .

CASE ACCIDENT SCENARIO Uw I - DC System Failure 1.8E-9 MSL Rupture before Recovery of (2.5 E-9)

BC

. Failure to isolate the MSL Rupture LOSP 2.9E 10 MSL Rupture (3.8E.10)  ;

Failure to isolate the MSL Rupture t

- MSL Rupture 3.3E.i t

- Failure to isolate the MSL Rupture (4.4E.ll)

-LOSP CASE 1 TOTAL 2. lE-9 (2.9E-9) 2 DBE Event causing 5.lE4' BC Failure (6.8E4)*

LOSP MSL ttupture 1.0E4'

- Failure to Isoiste '.he MSL Rupture (1.4E4)'

The change in failure probability values in parentheses assume a constniction period of 120 days, others assume a construction period of 90 days, a Using LLNL 85th Percentile Seismic Hazard Curve b Using EPRI 85th Percentile Seismic Hazard Curve 7

l l

--a -: mm

PART 2 SW PIPE RUPTURE FREQUENCY A PRA analysis was performed to deu 'nine the increase ir core damage probability resulting from r?;nting with one service wr *ader out-of-service. The principal contributor to increased risk is the possibility of pine ,

e in the operating header.

The frequency of non-ixluabb rupture of the Service Water (SW) piping was quantified using a piping rupture frequency mode! which was developed for the internal flooding analysis of the Surry Power Stahn [ Virginia Power,1991). This pipe niptum frequency modelis a system specific model based on the observed rupture events in the US Nuclear Power Plants with an estimated 1463 plant ." ears experience. This model is identificd as the Img-Linear Model.

Amener approach which has been used in other studies to calculate piping rupture frequency is Wash-1400 data [NRC,1975). To calculate rupture frequency using Wash-1400 data, the median pipe rupture frequency reported for a piping section (Table III 4-1) can be utilized.

Pipe rupture probabuity can also be evaluated from empirical correlations derived by Thomas,

[ Thomas,1981) which takes into account the following effects:

1. Historical failure data
2. Pipe and weld geometric factors
3. Penalty factor for plant-age
4. Penalty factor for weld material

~

In short, this model contains parameters which addresses the pipe characteristics including the pipe thickness.

A comparison of t'1e tw tre frequencies which can be obtained from the above mentioned models was carried out using the evaluation of the rupture frequency for the 10-inch diameter 8

SW piping upstream of the isolation MOV for the return header from the SW supply to the Component Cooling Water (CCW) Fuel Pit Cooler. This piping section has been ie ntified as the major contributor to the loss of service water initiating event during one header operation.

The calculations and justifications for Thomas Correlations and WASH-1400, are presented in Appendix B.

Table 2-1 Comparison of Service Water Pipe Rupture Frequency Using Alternative Methods Rupture Frequency Medtl Events / Year

1. Log-Linear Model 1. lE-5
2. WASH-1400 Model' 9.5 E-6
3. Thomas Correlation
  • 5.lE-6
a. The median rupture rate for piping sections with diameter greater than 3" was used
b. Plant specific data, including the piping inspection data from a January 1992 inspection were used A comparison of the pipe rupture frequency values derived using cach of these methods is presented in Table 2.1.

s The results indicate the Log-Linear model provides a conservative estimate. Although it does not explicitly address the piping characteristics, it does not under estimate the rupture frequency.

The values obtained from this log-linear model were used in the PRA analysis to ensure conservatism as well as to provide consistency with the IPE data. Since these data are obtained from actual SW operating experience and have been shown to be more conservative than data obtained by alternative calculations, they are appropriate for use in the PRA analysis.

9

i 1

j 1

4 i

l REFERENCES

1. Virginia diectric Power Company,1992, Probabilistic Risk Assessment. North 4

Anna Nuclear Power Station Units 1 and 2 for the Individual Plant Examination, Richmond, Virginia, (To be published).

l j 2. Virginia Power Electric Power Company,1992, Probabilistic Risk Assessment.

j Surry Nuclear Power Station Units 1 and 2 for the Individual Plant Examination, Richmond, Virginia, August.

3. Nuclear Regulatory Commission,1975, WASH-1400. NUREG-75/014. Reactor
Safety Studv
An Assessment of Accident Risks in U.S. Commercial Nuclear 1

Power Plants, USNRC, Octetar.

4

4. Thomas, H.M. , 1981, Pine and Vessel Failure Probability. Reliability Engineering, Volume 2, pps83-124.

I

)

l 10 l

2 APPENDIX A QUANTIFICATION OF FAILURE SCENARIOS 4

h a

_- _ __ -_-_-_-__am____-_-_ _-__ --- _mm_s-- --- -- J

i i i

4 4

APPENDIX A Qttantification of Failure Scenarios s

This appendix presents the detail calculation of the change in failure probability of the Unit 2 1

j AC system for the postulated accident scenarios. The construction activities associated with this phase of the SWPP is considered to be ..ompleted within 90 days. However, due to unforeseen circumstances the construction activities may continue for up to 120 days. The analysis for both j construction durations are performed. The nume? I values presented in parentheses are for 120

. days exposure time.

)

i 4

e i

i i

4 i

i t

i A-1

1 A.I. Accident 1 Failure Scenario The change in failure probability of the Unit 2 AC system, U,c, is obtained by performing the following analysis:

1. Identification of the interdependency, if any, between the BC system failure event, MSL rupture failure event, and valve closure failure event.
2. Quantification of the probability of BC system failure in 90 (120) days (ie. the period of time that BC system will be supporting the Unit I chillers), Pi ;
3. Quantification of the probability of Main Steam Line Rupture in the Unit 2 Turbine Building after occurrence of the event outlined in number one and before recovery from such an event, P2 . In this analysis, it is assumed that the BC system can be recovered within 30 days after its failure.
4. Q .antification of the probability of failure to isolate the ruptured header from the main steam valve house, P3 ;

The result of item one analysis is shown in Table A-1. No interdependency between the postulated failure scenarios is noted. Thus the change in failure prebability of Unit 2 CR\ESGR ccoling is given by:

U3 e' = Pi

  • P
  • P 2 3 where U3 c' = Contribution of the accident I failure scenario to the change in failure probability of the Unit 2 AC system.

Pi is quantified by construction of a fault tree model(See Figure A-1) for the failure of the BC

D m

A-2 i

system. The model is constructed based on the simplified flow diagram of the BC system supply to the Unit I chillers shown in Figure A-2 and under the following assumptions:

1. 1-BC-P-1 A is the running pump
2. 1-BC-P-1B is the standby pump
3. Failure data for motor driven alternating pump as described in appendix C of the North Anna IPE is applicable to these pumps
4. Failure of the normally open manual valves to spuriously close is not considered as significant
5. Probability of BC pipe rupture is conservatively assumed as 5.0E-5 for the 90 days titre period
6. The sta dby BC system's pump will b2 taken out for maintenance approximately for 43 (60) days, during the 90 (120) days time period P is calculated by multiplying the frequency of the main steam line break, quantified in the NAPS IPE as the mean frequency of initiating event A, by 30/365, where 30 days is conservatively assumed to be the mean time to repair the BC system. However not all the main steam lines are located in the Turbine Building. In this :--lysis, it is assumed that approximately 75 percent of the lines are in the Turbine Building. Thus the probability of main steam line rupture is given by:

P=

2 [ Frequency of "A"] * [30/365]

  • 0.75 From Appendix B of NAPS IPE, mean frequency of "A" is 5.0E-4; thus P=

2 5.0E-4 * [30/365]

  • 0.75 = 3.lE-5 P3 is quantified by the construction of a simple fault tree, as shown in Figure A-3. This fault tree is constructed 'oased on a simplified main steam line diagram as shown in Figure A-4. A MSL rupture in the Turbine Building can be isolated by closure of the main steam Trip Valves A-3 4

(TV) (hts-TV-101 A,B,C) which are seismically qualified and are located in the hiain Steam Valve House (htSVH). These valves can be closed manually or automatically. Per NAPS steam system training manual (NCRDOP-23-NA, page 18-19) a main line trip valve shuts when any-of the following conditions exist:

1. either the train A or B pushbuttons on the Safeguards Panels is depressed,
2. intermediate high-high containment pressure (17.8 psig on two of three channels),
3. either train A or B safety injection signal on high steam line flow coincident with low-low T.,, or low steam line rupture,
4. Control Room App. R isolation switch in EMERGENCY CLOSE, or
5. Emergency Switchgear Room App. R isolation switch in EMERGENCY CLOSE.

In this analysis condition 3 above is taken credit for initiation of the automatic isolation.

Additionally MSL rupture can be isolated by closure of the motorized non-return check valves (NRV) which are also located in the MSVH. The assumptions used in the construction and quantification of the fault tree include:

1. Failure of any one of the three TVs or NRVs to close is considered as the failure ta isolate the MSL rupture in the Turbine Building.
2. Probability of the failure of the actuation signal is the same as that evaluated for t'r probability of "NO TRIP SIGNAL TO M AIN STEAM TRIP VALVE" in the NAP 3 IPE steam generator fault tree model (Page 3, gate SG1344);
3. Probability of the operator failure to isolate the main steam line before significant steam is released in the Turbine ' Building is assumed to be 0.1. This assumption is made since deterministic evaluation'of the time available to the operator to isolate the main steam line rupture, before chillers are damaged, is not available presently. However, the following existing emergency procedures are applicable:

1-E-0, " REACTOR TRIP OR SAFETY INJECTION, REV 9,12-14-91, A-4

1-E-3, " STEAM GENERATOR TUBE RUPTURE, REV 4, 12-27-89".

4. Probability of loss of power to the NRVs is not modeled since review of the IPE model for 480 V buses (eg review of EP1, page 9, gate 12) indicate that probability of random failure of the bus (excluding LOSP which will be addressed -

separately) is insignificant compared with the operator failure probability value (0.1) assumed in here.

Thus from quantification of the fault tree shown in Figure A-3, the probability of failure to isolate is P3 = 1.17E-3.

Therefore the change in the failure probability of the CR\ESGR cooling for 90 days project duration is given by:

Uny' = Pi

  • P 2* P =3 4.99E-2
  • 3. lE-5
  • 1.17E-3 Uny' = 1,8E-9 and for 120 days project duration is Usv' = 2.5E-9.

This change in failure probability of the CR\ESGR-cooling is considered negligible and this scenario is not considered any further.

i A-5

A.2. Accident 2 Failure Scenario The change in failure probability of the Unit 2 AC system for this scenario is obtained by performing the following analysis:

1. Identincation of the interdependency, if any, between the BC system failure event, MSL rupture failure event, and valve closure failure event.
2. Quantification of the Loss of Off-Site Power (LOSP) probability, P , in a 90 (120) day time period. In this analysis it is assumed that the LOSP will result in failure of the BC system.
3. Quanti 5 cation of the probability, P 3 , of MSL rupture before recovery of LOSP.

In conventional PRAs the time to recover from a LOSP event is 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />. In this analysis, it is conservatively assumed that OFF-Site Power and therefore BC system will be recovered in 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Additionally, very conservatively it is assumed that Unit 2 is not affected by the LOSP event. That is Unit 2 is not tripped.

4. Quantification of the probability, P., of failure to isolate the main steam line rupture in the Turbine Building.

Thus Uc2=

3 p,.p,,p, P., is given by multiplying the annual frequency of LOSP event, as reported in the NAPS IPE, by the time period of in',erest, which in this failure scenario is 90 days. Thus Pc = 1.14E-1 * [90/365) =2.8E-2 (3.7E-2)

A-6

4 4

where 1.lE-1 is the annual frequency of LOSP event (from Figure 3.1.3-1 of the NAPS IPE).

j Also probability of the MSL rupture is quantified as described for failure scenario one but the exposure time is only one day (24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />). Thus,-

P=

3 [ Frequency of "A"] * [l/365]

  • 0.75 i

P=

3 5.0E-4 * [l/365]

  • 0.75 = 1.0E-6 Finally, from Table A-1 it is noted that the loss of off-site power will fail the source of power to the motorized NRVs but will not affect the main steam TVs, re-quantification of the isolation J

fault tree (Figure A-3), given that failure of the' NRV branch is 1, gives:

i i

P. = 1.03E-2 Therefore, the change in failure probability of the Unit 2 AC system for 90 days construction period is given by:

e 1

! Uc2 = 2.8E-2

  • 1.0E-6
  • 1.0;4E-2 = 2.9E-10 3

3 and for 120 days construction period is:

i i

Uc2 4 = 3.8E-10 Again, the change in failure probability of the Unit 2 AC system for this scenario is considered to be too small to merit further analysis.

4 1

A-7

.n-, - , , ,- , , . - , ,- , , . . - - ,

A.3. Accident 3 Failure Scenario The change in failure probability of the Unit 2 AC system for this accident scenario is obtained by performing the following analysis:

1. Identification of the interdepe..dency, if any, between the BC system failure event, MSL rupture failure event, and valve closure failure event.
2. Quantification of the probability, P,, of MSL rupture in a 90 (120) day time period. This probability can be obtained using the same approach as described for accident i failure scenario.
2. Quantification of the LOSP probability, P., after the MSL rupture event in the Unit 2 Turbine Building. Since after a MSL rupture the unit will be tripped, using the conventional PRA approach, the exposure time for the LOSP event is considered to be 1 day (24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />)
3. Quantification of the probability, P,, of failure to isolate the main steam line rupture in the Turbine Building.

Therefore:

Ux/ = P

  • P
  • P, 7

and P, = 5.0E-4 * [90/365]

  • 0.75 = 9.2E-5 (1.2E-4)

P, = 1.14E-1 * [l/365] = 3.1E-4 P, = 1.17E-3 (assuming LOSP does not occur simultaneously with MSL rupture).

Uxc3 = 3. lE-4

  • 9.2E-5
  • 1.17E-3 = 3.3E-11 (4.4E-11)

Again the probability of this plant damage state is considered too small to merit further consideration.

A-8

A.4. Accident 4 Failure Scenario Review of Table A-1 indicates that this is the most sever accident scenario considered in this analysis, since occurrence of a design basis earthquake (DBE) is assumed to cause rupture of the main steam line in the Unit 2 Turbine Building, failure of the Bearing Cooling system (neither the MS piping in the Turbine Building nor the BC system is seismically qualified) and LOSP However, all other seismically qualified systems are assumed not to be degraded as a result of a DBE. Additionally no other significant damage is assumed to be induced by a DBE.

These assumptions are made because no probabilistic evaluation of a seismic event at the NAPS has been performed up to date and performance of such an evaluation is out of the scop of this study. To quantify the change in failure probability of the Unit 2 AC system for this accident scenario, the following analysis are performed:'

l. Quantification of the probability, P i o, of a DBE in a 90 (120) day time period.
2. Quantification of the probability, Pu, of failure to isolate the main steam line rupture in the Turbine Building, i Pm is obtained by multiplying the annual frequency of a DBE by [90/365]. Using the 85th percentile curves from both LLNL and EPRI seismic hazard curves (per recommendation of the Generic Letter 88-20, Supplement 4), the annual frequency of DBE (0.18g)is 2.0E-3 to 4.0E-4.

Thus the change in failure probability of the Unit 2 AC system is given U3/ = Pm

  • Pn = [2.0E-3] * [90/365]
  • 1.03E-2 where 1.03E-2 is the probability of the failure to isolate, given that the DBE has lead to LOSP and failure of the main steam line NRVs. The value of the Pn reported here also assumes that the DBE has not degraded the relays that send the auto trip signals, the mechanical integrity of the TVs and put additional stress on the operator. Thus U3 / = 5.lE-6 (6.8E-6)

If 4.0E-4 is used as the frequency of a DBE, then Ux/ = 1.0E-6 (1.4E-6)

A-9

Table A-1 Interdependency Matrix Between Consequence of Failure Events ACCIDENT SCENARIO POSTULATED FAILURE EVENT POTENTIALLY DAktAGED COMPONENTS OR SYSTEMS I UI DC SYSTEM Ut CHILLERS MSL RUPTURE + ISOLATION U2 htAIN Fw PUMPS CAPABILITY U2 CONDENSATE PUMPS U2 CW CONDENSER MOVs 2 AND 3 LOPS BC SYSTEM MSL ISOLATION NRVs MSL RUPTURE + ISOLATION U2 MAIN FW PUMPS CAPABILTTY U2 CONDENSATE PUMPS U2 CW CONDDdSER MOVs 4

DESIGN BASIS EARTHQUAKE BC SYSTEM LOSP MSL RUPTURE MSL ISOLAT10N NRVs Ul= Unit 1 U2= Unit 2 A-10

i o [ t ( 2 1 3 1 4 5

BEAAING COOLING SYSTEM 0 n u.r ract raa svic BC2 1 r.. . u i : .:a :: :n v : ,- i s-;. 2e :5 : :,, ,-

1

sa:tha'i 5c ea ga 2 ;s ' O '*4 .* :T 1

_._ M ..fas G6C2:23 I

~

()  !

-4csas:..

  • 7 #: Sc3 sa:,5 M to ragg W ;r es: , :.

2:2?%0 : w:mE%= as at/Os staar a

. a.
  • a( :s 4c : set

' a h/(

t%t 90 3 an sta 3

iBCF!8-AU-900ays

  • hx'%

G8c2i33 Gec2i15

  • m ' 'i i

00nMON Cau st IT(DENCEnt F a h /(

taltiFE F 14 CF r4 Sf astgal 4 178a14a5 38CsCM-pL iS00CF GOC 2:a5 5 Nei 2tto t -S -o- t a ra:.3 3 t ac-a-13 r a hs 70 LA staaf 08C21$i 00Ch:33 staatge a ragts sinalida t rahl 6

secscx-A -secssi secsch-st-incss2 5 79E42 ilta s N @J 254 7

~

f I

=0 att As* MO ati %mp 4 ALT &

j nares to staat s:.o aisit rarts to u B

incpar-<s-sects iscpai-w-inces secpai-n -aacia t w@s s act@i i ia @a 4:eo

- i aC A ' *.=0

$ Wa!.s f 0 A,4e L

iSCPaI-fA iSC &

> ta @a aisc Figure A-1 BC System Fault Tree Model

Page 1 NUPRA 2.0 FILE : BCS200.FTP HNUS Env Minimum Cut Set Solution for fault tree BCS200 Performed :

, Serial no.= 3 15:50 31 AUG 1992 Cut Set Equation produced is : BCS200 EQN BEARING COOLING SYSTEM FAULT TREE FOR HVAC Top event: GBC2123 Top event unavailability (r.ev. appr)=

Cutoff value used =

4.99E-02 1.00E-10

( M sse'o" II'N /p/p0 )

Number of Boolean Indicated Cut Sets = 6 Number of MCS listed = 6 MINIMAL CUT SETS SORTED BY UNAVAILABILITY

1. 3.56E-02 1BCPAT-UM-1BC1B 1BCPAT-FR-1BC1A
2. 5.75E-03 1BCSCN-PL-1BCCCF
3. 5.07E-03 1BCPAT-FR-ABC1B 1BCPAT-FR-1BC1A
4. 3.31E-03 1BCSCN-PL-1BCSS1 1BCSCN-PL-1BCSS2
5. 1.41E-04 1BCPAT-FR-1BC1A 1BCPAT-FS-1BC1B
6. 5.00E-05 1BCPIP-RU-90 DAYS

- ^

NUPRA 2.0 Page 1 FILE : BCS2002.FTP HNUS Env Minimum Cut Set Solution for fault tree BCS200 , Serial no.= 4 Performed : 16:09 31 AUG 1992 Cut Set Equation produced is : BCS2002.EQN BEARING COOLING SYSTEM FAULT TREE FOR IfVAC Top event: GBC2123 Top event unavailability (r.ev, appr)= 7.02E-02 Cutoff value used = 1.00E-10 ( dhed /Jm/fah>\

Number of Boolean Indicated Cut Sets = 6 Number of MCS listed = 6 MINIMAL CUT SETS SORTED BY UNAVAILABILITY

1. 4.74E-02 1BCPAT-UM-1BC1B 1BCPAT-FR-1BC1A
2. 9.01E-03 1BCPAT-FR-ABC1B 1BCPAT-FR-1BC1A
3. 7.67E-03 1BCSCN-PL -1BCCCF
4. 5.88E-03 1BCSCN-PL-1BCSS1 1BCSCN-PL-1BCSS2
5. 1.88E-04 1BCPAT-FR-1BC1A 1BCPAT-FS-1BC1B
6. 5.00E-05 1BCPIP-RU-90 DAYS e-+

m 4

__m____.

1 l

1aom Other 1.umsh a

d Pump Diwharge M'"'

Strainer I'"I- E qr Valve f -

s T To

} _

Other IIcaring Ilc.P. Ill * -

(yLeads Cooling y Tower A __

s nia.. g M O V-11C-122 g , ,, , , p. , ,, ,

_ Valve Con arset N " " "' I Pump Diuharge BC-P.l A 3,,,,,,,

ES(;R Chillers if To Control Room ES(;M Chillers Figure A-2: Simpilified Flow Diagram of The Hearing Cooling Supply To The Control Hoe.1 Emergency Switchgear Room Chillers l

l

9 0  !
  • I 2 1 3 1 4 ,

5 ISCLATION OF T4 MAlh STEru 4 0

LINE AuptupE FAULT TAEE MS2 1 c.4. r sa j : ster::% :4?g Je.3 t.;a at < s .

  • 1 3 4 . .at * ; 5L a *g

=a 1% i'Ea* -. %E 3 s'.cf :% '4 'O Gass 1122 1

I sa:s sitae salm 5'taa .

g #..agtgem sauvtl it!# va5vtl f alls talsi t.0SE f3 CLCSE 3 i Gae5i131 - - G885 133 i

40 49 70 cm maaoat C4Cs vaust ra;.jt aCf!0m to CLOSE 4 v a.vt

@st I43 GaeS t 143

~

i i

  • C at: *a:s $t3na s Finates Fatttst

'O '4 T es is tutf[ aft e4 CLOSURE I

tee 5-aCT Ittota8C ( DatE3GO-38tt a ser ** i sami I I I '

w;tcs M at; .aset sofoa OPtaf0 vAet sofca DPt4fD estet optaa fps Fa!LJag ~(

rai 3 OpEm FAILS PE4 tais5 GPile . fp lutflatt mrv

'74!.5 *C Otr 'FatiS TO CLO1El :Falc5 'O CLOSE) CLO5vnt 6 t erSIs0v-FO-+dav t o t t asses 0v-70-evlo t ta6sev4 0-Svtot #48-tE380-36tt i :u -::2 i suma i suma e ami 7 .

~

I I C4Cu vaid , . - ci(ca val 4 C4:n vas et .

2 'alL1 Pt* (Falk$ f alL$ File IFalL$ FatLS 08Em ' Fat.5 f0 CLO5El TO CLOSE1 TO CLOSU

f. luGOlv-F M TV10ta tes00tv40-f v to tt les50tv to tt 2 a<m2 i *= . .

3 44 .

1 -

y 9 -

8 G

Figure A-3 -Isolation of the Main Stream Line. Rupture Fault Tree

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

t ,

i-i NUPRA.2.O FILE : .MS100.FTP- HNUS Env i~

Minimum Cut Set Solution for fault. tree MS100 , Serial no.=- 3 j Performed : _ '16:50_ 31-AUG 1992'

! Cut Set Equation produced is : MS100.EQN i

. ISOLATION OF THE MAIN STEAM LINE RUPTURE FAULT TREE l

Top event: GMS1122 l Top event unavailability.(r.ev. appr)= 1.17E-03 P Cutoff value used = 1.00E-10

! Number'of Boolean Indicated Cu't' Sets = 8- -

{ -Number of MCS listed = 7 MINIMAL CUT SETS SORTED BY UNAVAILABILITY l

i- 1. . 3.4 E-04 1MSCKV-FO-TV101A HEP-1Es 3-3&11 l 2. 3.44E-04 HEP-1E3&O-3&11 -1MSCKV-FO-TV101B f

3.. 3.44E-04 rIEP-1E3&O-3&11 1MSCKV-FO-101C'

4. 3.75E-05 1MSCKV-FO-TV101A 1MSMOV-FO-NRV101' 1 5. 3.75E-05' 1MSMOV-FO-NRV101 1MSCKV-FO-TV101B l 6. 3.75E-05 1MSMOV-FO-NRV101 1MSCKV-FO-101C j- 7. 2.66E-0.5- .-1MS-ACT-TV101ABC_ HEP-1E3&O-3&1l i

i i

i 5

4 i'

t l-f-

t i

4 k

4 k, .

i~

[

.i

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

NUPRA 2.0 Page 1 FILE : MS1133.FTP HNUS Env Minimum Cut Set Solution for fault tree MS100 Performed : , Serial no.= 3 3:00 1 SEP 1992 Cut Set Equation produced is : MS1133.EQN ISOLATION OF THE MAIN STEAM LINE RUPTURE FAULT TREE Top event: GMS1133 Top event unavailability (r.ev. appr) = 1.03E-02 s Pr,b of Fai8 <e to Cutoff value used =

1.00E-10 '

Number of Boolean Indicated Cut Sets = 4 g , X t c g ,, , L o 3 p' Number of MCS listed = 4 MINIMAL CUT SETS SORTED BY UNAVAILABILITY

! 1. 3.44E-03 1MSCKV-FO-TV1018

2. 3.44E-03 1MSCKV-FO-TV101A
3. 3.44E-03 1MSCKV-FO-101C
4. 2.66E-05 1MS-ACT-TV101ABC HEP-1E3&O-3&11 e

4 s

I i

i f

+

4 4

4 5

e .

, - e

7 7

  • m

. m o M

L h h } _h }

C L >

a INSIDE REACTOR CONTAINMENT -

".i.:. OUTSIDE REACTOR CONTINMENT 7:

c, n

E E i

w  ? E

- A  ?'  ?
. 8 -

+

3.

<\: < < \N

1. ' \ i O l.

m - O e O 3 o.

m. >

E' k= $M

< .! M !M

- .e., E .

v.. .' E E

- v.. v
  • o.

a .

a a.

g . . . . .

. . . . .n .. .

2 INSIDE TURBINE BUILDING Ch I 7 l s- Ec -

= m T

  • g -

m

%* 2.

L. ee

= EE2.

n~

M C 3

1 I- 1 I V V n>H e ae v+

r:

M&v R-m ~p =

Q L*

eo

$7u <

h-

.E .5' I m

  1. 3 .5 Ev# =
4) (

APPENDIX B PIPE RUPTURE FREQUENCY QUANTIFICATION ll '

. . . . . . . I

i i

i

1 APPENDIX B PIPE RUPTURE FREOUENCY OUANTIFICATION f

B.I. Frequency of SW Piping Rupture Using Thomas Correlation The Thomas Correlation (Published in 1981) can be used for prediction of pipe rupture frequencies. This methodology was used in the Oconee Study [NSAC,1984] and is an empirical correlation based on actual service failure statistics. In this methodology :he actual pipe

, thickness can be used to predict failure frequency.

I i

l The general approach for evaluating failure frequency was similar to that adopted in the Indian Point Safety Study (1982).

1. Compare the generic data on pipe failure mechanisms with the North Anna Power Station JN attributes to assess the relative frequency failure at NAPS.

t

2. Determine frequency of pipe ruptures in the SW system using Thomas Correlation and correct for specific attributes of NAPS SW System.

I B.I.1 SW Pinine Runture 4

i The data upon which the Thomas Correlation is derived comes predominantly from high pressure systems where the ratio of design pressure to system working pressure is about 1.1 to 1.5. The ratio of the North Anna SW design pressure,25 psig, to operating pressure of the return header pipe (the major contributors of T6 IE) is considerably higher. This higher ratio is a measure

, of additional safety margin for general causes of failure.

A representative list (extracted from Thomas,1981) of pipe failure causes and their relative contributions (fraction of failures by each cause) is:

f B-1 3

. , s __ _. - - _ ,

Percent Generic Leaks.

Manufacture and Fabrication 21.4 Material Selection 28.8 Fatigue - Vibration 4.3 Low Cycle 7.8 Expar'sion/ Flexibility 2.7 Corrosion / Erosion 24.6 Mal Oper* . 2.1 Thermal / Mechanical Shoc!: l.3 Miscellaneous 7.0 100.0 The SW system at North Anna is examined below against possible causes of failure in order to justify a frequency reduction factor for:

Manufacture. Fabrication and Material Selection Errors - The SW system at North Anna has been in operation as long as the plant has been in commercial

, operation (approximately 15 yeais) and all major components are several years old. These types of causes are generally revealed early in the life and should have already been detected at North Anna if they exist. Howaver, errors may still occur during repair work. The reduction in failure frequency isjudged to be at least 90%.

Fatieue The SW Systems operate at low temperature and do not experience wide-temperature fluctuation and is therefore not susceptible to thermal fatigue.

However, conservatively, no reduction in failure frequency is taken credit for.

B-2 L

Exoansion and Flexibility Such problems may arise due to design not adequately considering and allowing for pipe expansion and flexing caused by changes in temperature, pressure or other types of loads. It is Ekely that such problems in the design would have already been revealed and the failure frequency is judged to be at least 95% for these categories.

Corrosion / Erosion The SW lines have been or are being inspected and repaired in response to corrosion problems. However, as a conservative measure this factor has not been reduced.

Mal 09erations The potential for maintenance / operational errors leading to the system being left open prior to reflooding or inadvertently opened prior to isolation is separately addressed in the internal flooding analysis of North Anna Power Station. The opportunity for mal operation of the system leading to internal stresses which caused a component failure is negligible. However conservatively, this parameter is not reduced.

Thermal and Mechanical Shock The SW system does not experience wide temperature fluctuations' and thus the potential for thermal shock is negligiMe.

The only internal mechanism for exerting mechanical shock would be water hammer. The reduction in failure frequency is at least-50% due to these categories.

Miscellaneous The contribution from this category has not been reduced.

B-3 u.

Table B-1 Revised Frequency Calculation

, 1 2 3 Pipe Failure Cause  % Generic Leaks PC of Generic Ixaks Applicable to North Applicable to North Anna (After Applying Anna Above) PC/PL Manufacture and 2.14 0.08 .017 Fabrication Material Selected 2.88 0.03 0.086 Fatigue

-Vibration 2.15 .20 0.43

~

-Low Cycle 7.8 0.03 0.234 Expansion and Flexibility 0.14 0.10 0.14 Corrosion /

Erosion 24.6 0.02 0.49 Mal Operation 2.1 0.02 0.945 Thermal and Mechanical Shock 0.65 0.20 0.13 Miscellaneous 7 0.04 0.28 Total 51.61 3.2 The revised frequency contribution from each category is shown in Table B-1, Column 1. Also shown in the above table (Column 2) are the ratios of ruptures to leaks (PC/PL) for each type of failure mechanism. These ratios are based on information provided in Table 3 of Thomas, 1981. Column 3 is simply maltiplication of Column 1 and Column 2 and shows the frequency of rupture due to various mechanisms as a percentage of the leak frequency determined from generic data. Overall this percentage is 3.2Fe (i.e., a ratio of .0032)

The frequency of pipe and valve rupture can be calculated using the Thomas Correlation as l

B-4 l

discussed above. An evaluation of SW piping upstream of the isolated MOV for the return headec from the SW supply to the CCW Fuel Pit Cooler is given as an example. Rupturing this piping section in the major contributor to the loss of SW initiating event while in one header operation.

Frequency of leakage (uncorrected)

F=P;L P P+nA ' "}

c, c' Dp = OD = 10.75" diameter Lp = Length = 120" tp = pipe thickness = 0.283" includes corrosion (Measured January 1992)

A = penalty factor for weld - 50 Lw = weld length ~ 1.75 x tw tw = weld thickness tw ~ tp P1 = lx 10~8 yr 4 (leak rate)

Dw =

weld diameter Dw = Dp n =

number of welds = 4 B-5 l

,f F = 1x10' x (10.75 x 120 + (50 x 4) 10.75 x 0.50) 0.08 0.08 F = 2.94 E-4/yr Multiplying this leak frequency by the North Anna specific rupture to leak ratio 0.0032 developed above gives a pipe rupture frequency of 9.5E-6.

B.2. Frequency of SW Piping Rupture Using WASH-1400 Data Assuming 4 discontinuallies (i.e.,4 piping sections) and using probability of rupture as 1.0E-10 (Table III 4-1 of WASH-1400), then:

F = 4 (1.0E-10)(365)(24) = 3.5E-6 4

B-6

. . . . . _ . . . . .

Retrieved from "https://kanterella.com/w/index.php?title=ML20105A904&oldid=2799094"