ML20127C607

Draft Evaluation of Unisolated LOCA Outside Drywell in Shoreham Nuclear Power Station

ML20127C607
Person / Time
Site:	Shoreham File:Long Island Lighting Company icon.png
Issue date:	02/28/1985
From:	Hanan N, Ilberg D BROOKHAVEN NATIONAL LABORATORY
To:	NRC
Shared Package
ML20127A367	List: ML20127A363 ML20127A500 ML20127A968 ML20127B005 ML20127B638 ML20127B641 ML20127B709 ML20127B780 ML20127C153 ML20127C469 ML20127C607 ML20128P302 ML20128P388 ML20128P824 ML20128P861 ML20128P871 ML20128P892 ML20128P901 ML20128P921 ML20128P941 ML20128P963 ML20128P984 ML20128P988 ML20128P994 ML20128Q012 ML20128Q031 ML20128Q051 ML20128Q062 ML20128Q089 ML20128Q101 ML20128Q225 ML20128Q285 ML20128Q299 ML20128Q313 ML20128Q366 ML20128Q388 ML20128Q409 ML20128Q435 ML20128Q448 ML20128Q467 ML20128Q477 ML20128Q487 ML20128Q492 ML20128Q502 ML20128Q508 ML20128Q520 ML20128Q524 ML20128Q535 ML20128Q846
References
FOIA-85-199 NUDOCS 8503290251
Download: ML20127C607 (100)

Text

{{#Wiki_filter:p-. / \\ An Evaluation of Unisolated LOCA Outside Drywell in the Shoreham Nuclear Power Station D. Ilberg N. Hanan Draft (for comments only) Risk Evaluation Group Department of Nuclear Energy Brookhaven National Laboratory Upton, New York 11973 February 1985 Prepared for U.S. Nuclear Regulatory Commission Washington, D.C. 20555 m 25c3rio251 X&

~. es 9 CONTENTS Page LIST OF FIGURES........................................................ iii LIST OF TABLES......................................................... iv PREFACE................................................................ v ACKNOWLEDGMENTS........................................................ vi 1. INTR 000CTION...................................................... 1-1 1.1 Background................................................... 1-1 1.2 0 bj e c t i v e s................................................... 1 - 1 1.3 Scope........................................................ 1-2 1.4 General Description of the Problem Evaluated................. 1-3 2. EVALUATION OF PIPE BREAK FREQUENCIES.............................. 2-1 3. ASSESSMENT OF MITIGATION CAPABILITY............................... 3-1 3.1 Reactor Buil di ng Inf ormati on................................. 3-1 3.1.1 Ins t rumentation f or Diagnos ti cs....................... 3-1 3.1.2 Sump Pumps and Fl ooding Buildup Vol umes............... 3-2 3.1.3 Containment Atmosphere................................ 3-3 3.1.4 Procedures............................................ 3-6 3.2 A Small LOCA Outside Drywell (< 1-1/2" Break Size)........... 3-6 3.2.1 Acci dent Condi ti ons and Al a rms........................ 3-6 3.2.2 Rea ctor Buil ding Envi ronment.......................... 3-7 3.2.3 Operator Response..................................... 3-8 3.2.4 Es timati on of Core Damage Frequency................... 3-9 3.3 A la rge Loca Outs i de Drywell (> 6" Break Si ze )............... 3-11 3.4 A Medi um LOCA Outs i de Drywell T2" <, < 4").................. 3-12 3.4.1 Accident Conditions Alarms and Opertor Response....... 3-12 3.4.2 Es ti mati on of Core Dama ge Frequencies................. 3-15

4.

SUMMARY

4-1

5. REFERENCES........................................................

5-1 APPENDIX A: PI PES AND VALVES FAILURE RATES............................ A-1 A.1 Pi p e Ru p t u re.................................................. A-1 A.2 Valve Failure Rates........................................... A-2 A.3 Comp a ris on wi t h LOC A Frequ enci es.............................. A-3 APPENDIX B: LINES CONNECTING REACTOR PRESSURE VESSEL TO R EAC TOR B U I L D I NG................................................... B - 1 APPENDIX C-IDEN IFICATION OF PIPE SECT FOR BRfAK FRE UENCY ESTIMATION....... IONS AND DISCONTINUITIES .............................. C-1 ii

-6 LIST OF FIGURES Figure Page 1 General Description of SNPS Reactor Building Elevations (with Emphas is on HPCI Steam Line Routing)................. 2a Lines from Raactor Pressure Vessel to Reactor Building....... 2b Tip Drive Guide Tubes Connections to Reactor Pressure Vessel. 3 Event Tree Diagram for Sequences Following Small LOCA Ou t s i d e D rywe l l............................................ 4 Event Tree Diagram for Sequences Following large LOCA Ou ts i de P rywe l l............................................ 5 Event Tree Diagram for Sequences Following Medium LOCA Ou ts i de D rywe l l............................................ O iii

c' i. O LIST OF TABLES Table ~Page 1 Summary of Failure and Unavailability Data for Pipes and Va1ves................................................. l 2 Estimated Frequencies of Breaks Outside Containment.......... 3 . Summa ry of Frequenci es of LOCA Outs i de D rywe11............... 4 Reactor Building Temperatures at Several Elevations ' Res ul ti ng f rom a 40,000 l b. Dis charge..................... 3-4 5 Core Damage Frequencies for Unisolated LOCA Outside Drywell. 4-1 i O iv

( 0 PREFACE This work was prepared for the NRC which requested it within a one man month time frame. This dictated the use of all readily available information and refraining from physical analyses. Some of the phenomenological assump-tions are approximate; hence, more accurate analysis may result in a somewhat different contribution to the core damage frequency for the medium LOCA out-side the drywell (the major contribution to core damage frequency is this study). Nevertheless, tGe identification of the relative hierarchy of con-tributors is believed to -be reasonable. O l V

A. s ACKNOWLEDGEMENT The authors wish to thank Kenneth Perkins, Kelvin Shiu, and Robert Young-blood - for their helpful comments. Cheryl Conrad is much appreciated for typing this. document to meet a tightly imposed dealine. e 9 vi

t 6 i 1-1 1. INTRODUCTION

1.1 Background

The SNPS-PRA(1) considered LOCA outside the Drywell (LOCA in the Reactor Building) in two ways: a) Interfacing System LOCAs: Appendix F of the SNPS-PRA estimates the initi-ator frequency and the core damage frequency for this case. The BNL review (2) of the Shoreham PRA re-evaluated the initiator frequency as well as the core damage frequency, and found an increase about an order of magnitude of the core damage frequency.. This result is included in the present study, and for more details see Appendix C of Reference 2. b) High energy line breaks inside the Reactor Building: The SNPS-PRA included in its analysis only pipes larger than 6 in. in diameter, on the premise that, if not automatically isolated ample time is available to isolate breaks in smaller lines before they cause adverse containment conditions. The freiuency of unisolated line breaks downstream of the l outboard isolation valve was calculated to be relatively small. The BNL review of this part agreed with the SNPS-PRA, as discussed in Appendix C of Reference 2. In the SNPS-PRA and the BNL review, all~ the isolation valves were assumed to be capable of operating under a postulated break and the resulting break-flow conditions; random failure to operate was used in both studies. It is shown in Reference 2 that interfacing system LOCAs are the major contributor to LOCAs outside the drywell. 1.2 Objectives This study is a special consideration of case (b) above stemming from the assunption of the failure of the corresponding isolation valves in the case of a line break outside drywell. NRC requested BNL to re-evaluate the core damage frequency from high energy line-breaks inside the Reactor Building (same as case (b) above) under the assumption that most of the isolation valves are not qualified to close under break-flow conditions, i.e., assuming the failure of most of the isolation valves. Under this assumption, there is

~ ~ ~ 4 6 1-2 a need to examine the rupture of any pipe (regardless of diameter) opening a path'that leads from the Reactor Pressure Vessel (RPV) to the Reactor Build-ing, fcc potential adverse environment or flood effects. This assunption obviously increases the contribution of the high energy line breaks to core damage frequency and requires consideration of other lines connected to the RPV of diameter < 6 in. This study considers the following questions: (a) What would be the increase in core damage frequency due to the assumption stated before, i.e., the failure of isolation valves to perform their function? (b) What would be the contribution to core damage frequency from each pipe connecting the RPV and the Reactor Building? (c) What isolation valves would be important for mitigating the outside drywell LOCAs? (d) What is the characteristic time available for operator action? 1.3 Scope The scope of the BNL study was defined to cover the following: (a) To identify any significant(*) high energy lines leading from the RPV to the Reactor Building with a potential for affecting safety systems, if an unisolated break were postulated. (b) To estimate the change in SNPS core damage frequency relative to the SNPS-PRA(1) and BNL review (2) due to the following assumptions on the - operation of isolation valves following the occurrence of a line break: (*)The contribution from downstream moderate energy lines of a system was neglected if it was smaller than the contribution of the lines upstream.

o s 1-3 (1) The Main Steam Isolation Valves (Inboard and Outboard) on all four main steam lines will isolate in all the cases considered, having the failure rates shown in Table 2 (discussed in Appendix A). (2) All check valves will close on reverse flow as designed with the failure rates shown in Table 2 (discussed in Appendix A). (3) All other isolation valves will fail to close when receiving their signal to close. No partial closure is assuned for these valves. (4) Manual valves are assuned to be available for isolation if access-ible by the operator. (5) Remote operated valves that do not receive automatic closure signals upon sensing break conditions are identified. However, no credit is given for them in this study. (c) To provide the list of the more important isolation valves from the standpoint of reducing the core damage frequency. (d) To provide some crude insights on the time available for the operator to respond to such accidents. 1.4 General Description of the Problem Evaluated The Shoreham Reactor Building surrounds the MARK II containment structure (the drywell). At its lowest elevation (referred to here as Elevation 8), the building is an open cylindrical compartment, i.e., there are no barriers in Elevation 8 compartments. This open area presents the possibility that excessive water released into the compartment may adversely affect the ECCS equipment in Elevation 8. The SNPS Reactor Building has openings between its floors, and a line break at a high elevation will affect the entire reactor building (see section 3.1 for more details). Figure 1 provides a general description of the SNPS Reactor Building Elevations. l

m = + +.... - '4 n C a [ uc-44 43 O YOP OF Ok _ na.t 1 itsno Tom POLAn CaAng ....A f I L,..n*

ps CAS ~,',-

s e Ett. SMAFT OJ ,r'8 OECONTAtRNATION &#EA gg gygg,9e i,- lwn! I I.*.

WWL, i

1. E ACTOR g.4,

,Pd2M*JA Ftf N CLOSED Loop e 8 Mf-089 [ C00.deG WATER s 'Lw=d VST a s <n exj.n l c4 + l l

• s EL '50\\9' IkasER &CuYE 4 s

SLLLows SE AL ') w 548Y L Qul0 ,[ CONTROL STS. vthTILATiose nEActag STANEF 00 ComTwo. [ T =C. vt,,, F 7 ,a QuC = 6 j e e , ACTIVE ' { f." fbfL ZDNE '} ' / C i f SMsELD MLL',.'L. 4...J o t, O - ) WI l!. '1 /,-f, :, y mm, s% , oEm ,3 w e,.Att .r 'Ul ' ' ', * ^ PEpsomhEL Ca0 EIOD [b" '\\ afpaan %) = E0..E.T s i-6 PLATR mas 4 ETfAM C ' ' J. ' vi-: MA $.4"'" " CS'#3Td.q* p. ? j CE

• /

.7Y* ELL SEAL % CR0 MASTER s k' i M M I'.* k . i,[ ,[, ; .,s,g.,:, a .. n,, i i i t cLoSt * ~ "t.'/ sTEAMu e ~" W' -c.a.Ast ro o..at. l C ,AC,C,E,ss --- r. .-anst AACE l j i li Cm0 Puupw I "* g84E,", [J, d j,! @ T2[( l fj l PLATF = fg n.e 0-r

1

~ m,i f .' CLOSED LOOP l nC gyr C..w..n.T E.0,,M E "AMEC49Pr r , MPCI TungimE s' M00ULE = 4 =-2 kemi ssoN I@,0* h PC OL i s l + tW .l, l l [ l at:~i ?.t'.. : : T. (.,.- m,..c 41c' Fig. 1 General Description of SNPS Reactor Building r==r ',','e'."T,'. Elevations (with Emphasis on HPCI Steam Line Routing) From: Shoreham Nuclear Power Station - Unit 1 i Final Safety Analysis Report PEvisioN #4 - CECEMBER 1978 .,-n ... - - -,.,,, - - - -~

1-5 Figures 2Mnd Nshow lines that connect the RPV to the Reactor Building and provide a potential path from the RPV to the volume of the Reactor Build-ing in the event of a break with a failure of the pertinent isolation valves to close. These figures do not show all isolation valves, but only those that are designated as containment isolation valves. In some cases, the most important being the RWCU, other valves are available to the operator for remote line isolation from the control room; these valves are not shown in Figures 2 and 3. A list of the lines emerging from the RPV and some additional information associated with these lines (size, type of isolation valves, and process or standby line) is given in Table B.1 of Appendix B (reproduced from the SNPS-FSAR(3)), e +-

c_______ s cco cca [ ~ (MSIV.LE AKAGE CONTROL SYSTEM

cc, ces

~ TYPICAL ARRANGEMENT FOR cci. cce 4 OUTBOARD MSIV'S) i ~ _i_ "d XID MAIN STEAM Xil RHR-HE AD SPR AY LINE TO R PV E 4.5 culsce h, 8 a._. ces 1 M AIN STE AM LINE DRAIN 1 1 i, d cce X2A FEEDWATER LINE ] :, b 8 MAIN STE AM LINE DR AIN XIC MAIN STEAM cci s X28 FEF -> 'TER LINE r: a d X20A CORE SPRAY PUMP DISCHARGE TO RPV " b XIB M AIN STEAM cce ~ "1 2 [ MAIN STEAM LINE DRAIN " 0*, cce cce [., b.] XtA MAIN STEAM X308 CORE SPRAY PUMP DISCHARGE TORPV U cce na ]PPRE ~ MAIN STEAM LINE DRAIN c y W u h-X36 STANDBY LIQUID COOLANT TO RPV [ Q[ l J 8 - X3 MAIN STE AM LINE DRAIN REACTOR hT c'c c'4 cc. ne'

1. VESSEL 1 FJ XI2 HPCI TURBINE STEAM e

X30 SAMPLE COOLANT FROM RPV c8 X16 RCIC TURBINE STEAM "* 2g cc'kc, INLET LINE 50A RHR INJECTION LINE TO REClRCUL ATION SYSTEM RETURN LINE & I2 E X68 RHR INJECTION LINE TO L W - RECIRCULATION SYSTEM RETURN Llf cce ecs X4 RWCU LINE FROM RPV N _"r E } '"N" cce en g fj X5 RHR SHUTDOWN e e \\ c;.' ~ COOLING LINE FROM RPV cc, Fil RECIRCUL ATION PUMP ~- 1 I .J s s SEAL INJECTION g7 ,C ~ FIO RECIRCULATION PUMP RCULAT SEAL INJECTION cc%gia 99 pyMps C2 PASS REACTOR SAMPLE X JE T PUMP FLOW INSTRUMENT LINE Yh.'FC

:f

( 37 1 DRYWELL WATER SUPPRESSION CH AMBER LEVEL \\ AIR SPACE yyyy lill - ~.- SUPPRESSION POOL ~,~ Fig. 2A Lines from Reactor Pressure Vessel to Reactor Building.

s i e h f 3 i i .t REACTOR PRESSURE VESSEL \\ s ces scs X22A-R8CLCW-TO RECIRCULATION PUMP 5 L 8 AND MOTOR COOLERS L T X228-RBCLCW-TO RECIRCUL ATION PUMP l AND MOTOR COOLERS ~ j ~ ces s cco I X234-R8CLCW-FROM RECIRCULAT10N PUMP 1 AND MOTOR COOLERS Q f I X23B-RBCLCW-FROM RECIRCULAT10N PUMP AND MOTOR COOLERS REACTOR g RECIRCULATION X38 NITROGEN / AIR PURGE FOR TIP x Z PUMPS h X37A TIP DRIVE GUIDE TU8ES b: b X378 TIP ORIVE GulOE TU8ES h k X37C TIP DRIVE GUIDE TU8ES b X370. TIP DRIVE GUIDE TU8ES y DRYWELL i SUPPRESSION CHAMBER E } XCS-DRTWELL FLOOR SEAL PRESSURIZATION < WATER LEV L ~ SUPPRESSION POOL i Fig.43 Tip Drive Guide Tubes Connections to Reactor Pressure Vessel 1 i 1 l i t

o a 2-1 2. EVALUATION OF PIPE BREAK FREQUENCIES This section covers the evaluation of the frequencies of high and mod-erate energy pipe breaks excluding interfacing LOCAs. The interfacing LOCAS are addressed in Appendix C of Reference 2 and the results are included in Tables 2 and 3. The pipes considered in this BNL study are listed in Appendix B. All lines which are associated with General Design Criterion (GDC) 55 are analyzed in this BNL study.(*) In addition, the Transversing Incore Probe (TIP) Drive Guide Tubes (GDC-57) are considered. All other lines referred to in Table B-1 as GDC-56 or 57 are not connected to the RPV; they are mainly connected to the Suppression Pool (the routing was rechecked). The SNPS-FSAR(3) was the main source for determining the number of pipe sections and valves or other discontinuities on each line. The isometric drawings of pipe routing in the Reactor Building from Appendix 3C of the SNAPS-FSAR were used. They were compared with the system-specific drawings given in the other FSAR chapters. The summary of this task is presented in Appendix C of this report. The evaluation of pipe break frequencies was made with the failure and unavailability data summarized in Table 1. The bases for the values shown in this table are further discussed in Appendix A. The failure and unavailabil-ity data were used with the nunber of sections and valves or discontinuities identified for each line, to compute the frequency of line breaks. The summary of this task is presented in Table 2. An example of this computation is shown in Appendix C. The results of Table 2 were next grouped into seven different cases: (a) Large Interfacing LOCAs (Liquid discharge through break) (b) Large LOCAs outside Drywell: (1) steam and (2) liquid discharge (c) Medium LOCAs outside Drywell: (1) steam and (2) liquid discharge (d) Small LOCAs outside Drywell: (1) steam and (2) liquid discharge. (*)ln References 1 and 2 consideration was given to lari,e 5reak LOCA outside the drywell, i.e., lines which are 6 in. in diameter or more.

a Table 1 Summary of Failure and Unavailability Data for Pipes and Valves Failure Rate (Mean) Break Non-Break Component Failure Mode Exclusion Exclusion Pipes > 3" Rupture 8.6x10-ll/hr 8.6x10 10/hr (persection) Pipes < 3" Rupture 8.6x10-10/hr 8.6x10 9/hr (per section) Check Valves Severe Internal 3.3x10-3/yr 1.eakage Rupture 1.5x10-10/hr 1.5x10-9/hr Motor Operated Failure to 8x10 3/d Valves (MOV) Operate (w/ comman,d faults) Failure to 6x10-3/d Operate (w/o command) faults) Two MOVs (CMF) 2x10-3/d Rupture 1.5x10-10/hr 1.5x10-9/hr l L_

TABLE 2 Estimated Frequencies of Breaks Outside Containment NtNBER OF: IN!ilAL BREAK L S V ISOLATION VALVES BREAK ESTIMATE 0 DESCRIPTION OF THE CASE ANALYZED .d ^ BREAK LOCATION SIZE I E A ASSUMED FLOW: FREQUENCY N C L VALVES FAILURE STEAM OF BREAK CASE E T V DESIGNATORS PROBABILITY OR LIQUID OCCURRENCE S I E O 5 N 5 (*) Main St?am IB21-A0V081 Break exclusion section and valve between Reactor Line 1 24" 4 1 1 Inboard MSiv 6.0E-3 steam 5.0E-8 Building penetration and the outboard HSIV. (Elevation 78). 11 24" 4 1 0 Inboard and 2.0E-3 steam 6.0E-9 Break exclusion section from outboard MSIV up to Outboard the.'et-Impingement Barrier. (Elevation 78). MSIV 1821-A0V082 Main Feed-Break exclusion section and testable check valve water Line i 18" 2 1 1 Check Valve between reactor building penetration and the F002 A/B 3.3E-3 steam 1.4E-8 testable checkvalve. (Elevation 78). Testable C.V. Break exclusion sections and IB21-MOV035A/B from II 18" 2 3 1 IB21-A0V036 [3.3E-3]2 steam 7.8E-11 testable check valve up to the Jet-Impingement A/B and C.V. Barrier (Elevation 78). F002 A/B High Pressure Break exclusion section and valve between Reactor Coolant injec-1 10" 1 1 1 IE41-MOV041 1.0 steam 2.1E-6 Building penetration and the outboard isolation tion (HPCI) valve IE41-MOV042. (Elevation 66). Steam Line II 10" 1 6 6 IE41-MOV041 1.0 steam 1.4E-6 Non break exclusion sections and valves from and IE41-outboard isolation valve up to HPCI turbine. Four MOV042 openings (24 hrs each) per year of valve MOV-042 are assumed. (Elevation 66 down to elevation 17). III 1" 1 17 17 IE41-MOV048 1.0 steam 1.0E-3 Non Break exclusion sections and valves from and Reactor Building penetrations up to the 1-1/2" 1E41-MOV047 HPCI/RCIC drain line to condenser. Normally open path. (Elevation 66 down to elevation 11). 4 This includes all discontinuities, f.e.: valves, pumps, reducers and heat exchangers (see Appendix A). 4 9

~ t TABLE 2 Estimated Frequencies of Breaks Dutside Containment Cont'd. NU4BER OF: INITIAL BREAK L S V ISOLATION VALVES BREAK ESTIMATED DESCRIPT!0f; 0F THE CASE ANALYZED BREAK LOCATION SIZE I E A ASSUMED FLOW: FREQUENCY N C L VALVES FAILURE STEAM OF BREAK CASE E T V DESIGNATORS PROBABILITY OR LIQU10 OCCURRENCE S I E O S N (*) S Re'cter Core Isslation Cooling I 4" 1 1 1 IE51-MOV041 1.0 steam 2.lE-6 Break exclusion section and valve between Reactor (RCIC) Steam Line Building penetration and the outboard MOV 042. (Elevation 87). II 3" 1 6 6 IE510iOV041 1.0 steam 5.8E-6 Non break exclusion sections and valves from out-and MOV042 board isolation valve up to RCIC turbine. Four openings per year of valve -042 are assumed (All elevation below elevation 87 down to elevation 8). 1" 1 14 14 IE51-MOV048 1.0 steam 1.2E-3 Non break exclusion section and valves from Reac-and tor Building Penetration up to the 1-1/2" HPCI/ IE51440V047 RCIC drain line to condenser. Normally open (Elevation 87 down to Elevation 8). RCIC/HPCI Steam IE51/IE41 Section between HPCI and RCIC drain lines connec-Drain Line I 1-1/2 1 1 0 A0V-081 or 1.0 steam 5.E-5 tion and the penetration to the main steam tunnel. ADV-082 (Between elevation 11 and 70). Re:ctor Water MOV033 (F001) First section in Reactor Building. It is break Cirenup System I 6" 1 1 1 and MOVF102 1.0 Liquid 2.1E-6 exclusion and normal operating. (Elevation 112) (RWCU) Supply anJ MOVF100 Line F106 II 6" 1 1 0 The above and

1. 0 Liquid 7.5E-6 Section from outboard isolation vlave to the 6x3" IG33410V034 reducer. Non break exclusion (Elevation 112).

or(F004) III 3" 2 3 3 same as above 1.0 Liquid 5.4E-4 Section and valves fran reducer up to RWCU pumps. (Elevation 112) 4 h k

/* TABLE 2 Estimated Frequencies of Breaks Outside Containment Cont'd. NUMBER OF: INITIAL BREAK L V IS0tATION VALVES BREAK ESTIMATED DESCRIPTION OF THE CASE ANALYZED BREAK LOCATION SIZE I E A ASSUMED FLOW: FREQUENCY N C L VALVES FAILURE STEAM OF BREAK CASE E T V DESIGMATORS PROBABILITY OR LIQUID OCCURRENCE 1 E O S N (*) R::ctor Water IV 2 1 1 The above + C1sinup System 2 manual 1.0 Liquid 1.8E-4 Sections and valves from 2x3" reducers to 3x4" (RWCU) Supply valves reducers. (Elevation 112). Lina V 3" 2 2 2 Same as above 1.0 Liquid 3.5E-4 Sections and valves from 2x3" reducers to 3x4" reducer. (Elevation 112). VI 4" 1 5 2 Same as above 1.0 Liquid i. + Manual 4.0E-4 Sections and valves from 3x4" reducer up to the valves discharge of the non-regenerative Heat Exchanger and up to the normally closed IG33410V035. (Elevation 126). M319 Steam Line 1 3" 1 1 1 IB21440V031 1.0 steam 8.8E-6 Normally open - break exclusion section and valve Drein (Inboard) between reactor building penetration and the outboard valve IB21-M0V032. (Elevation 76). II 3" 1 1 1 IB21-MOV031 1.0 steam 8.8E-5 Normally open -non break exclusion section and and MOV032 valve f rom MOV032 up to the Jet-Impingement Barrier. (Elevation 76). Main Steam Line I 2" 4 2 4 Inboard MSIV 6x10 3 steam 5.lE-7 Break exclusion sections and valves between main Droin (Outboard) IB21-A0v081 steam line connection and IE32-MOV021. IB21-M0v061 end MSIV leakage MOV062. M0V063. MOV064 (Elevation 76). Contrsl (inboard) MSIV Leakage I 2-3" 3 1 2 Inboard and 2x10-3 steam 6.0E-8 Break exclusions sections and valves between main Contr:1 (out-Outboard MSIV steam line connection and IB21-MOV034. IE32-MOV024 bo:rd) IB21-A0V082 IE32440V026. (Elevation 76). i a b

r 7 TABLE 2 Estimated Frequencies of Breaks Outside Containment Cont'd. NUMBER OF: INITIAL ~ BREAK L V ISOLATION VAL VES BREAK ESTIMATED DESCRIPTION OF THE CASE ANALYZED BREAK LOCATION SIZE I E A ASSUMED FLOW: FREQUENCY N C L VALVES FAILURE STEAM OF BREAK CASE E T U DESIGNATORS PROBABILITY OR LIQUID OCCURRENCE S I E O S N (*) S lit:rfccing LOCA: - RHR Snutdown I 20" 1 2 Liquid Cooling All four interfacing LOCA cases estimated on the - RHR Head Spray II 4" 1 2 Liquid basis of 0.02 for testable check valve unavail-Line 2.0E-6 ability times 10-3 for spurious MOV opening and - RHR/LPCI Injec. III 24" 2 2 Liquid 0.1 for probability of low pressure piping to fall Line to Recirc. ~ before isolation. See detail in reference 2 Lines IV 10" 2 2 Liquid (Elevation - 8 up to elevation - 87). - LPCS Injection LO-F008 and Stcroby Liquid 1 1-1/2 1 1 1 ' Inboard C.V. 3.3E-3 Liquid 1.5E-8 Break exclusion section of the SLC (Elevation 112) Contr:1 (StC) II l-1/2 1 1 1 F007 [3.3E-3]2 Liquid 1.0E-9 Non break exclusion section of SLC (Elevation The above 112). and Outboard C.V. F006 I Contral Rod Driva (CRD) I l-1.0 Liquid 1.0E-4 Scram Oischarge Volume (SDV) header rupture. (Non 1-1/2 break exclusion). The pipe break frequency is taken from fluREG-0803. (Elevation 78, 63 and 40).

l (~ 4 l .l' i l I TABLE 2 Estimated Frequencies of Breaks outside Containment Cont 'd. l NUMBER OF: INITIAL BREAK L 5 V ISOLATION VALVES BREAK ESTIMATED DESCRIPTION OF THE CASE ANALYZED l BREAK LOCATION SIZE I E A ASSUMED FLOW: FREQUENCY l N C L VALVES FAILURE STEAM OF BREAK CASE E T V DESIGNATORS PROBA8ILITY OR LIQUID OCCURRENCE l S I E O S l N (*) S Ricirc. Pump Seal Injection I 3/4 2 2 2 1.0 Liquid 2.0E-7 Other 3/4" lines 1 3/4 1 20 20 Valves of the 1.0 steam 1.8E-3 Branches from various sys-and j system shown in tem shown in Liquid (All elevation) this table this table l1 Sample Coolant i from RPV I 3/4 1 2 2 1.0 Liquid 1.8E-4 1' Reacter Post i Accident Samp1-I 3/4 1 2 2 1.0 Liquid 1.8E-4 I fng system (PASS) i TIP Drive Guide i Tubes I 3/8 4 2 2 8411 valve 1.0 Liquid 1.0E-5 (Elevation 60). and shear 1 valve 1 l t e g k I L

o 2-3 The combined frequency in each group is shown in Table 3. Note that the LOCA frequencies of the large and medium breaks groups are dominated by the line breaks of a single system. For the liquid breaks, it is the RWCU, and for the steam breaks, it is HPCI and MSL drain systems. In the latter case, the 10-in. HPCI line break has a frequency of 3.5x10-6, while all other line breaks which contribute to the large LOCA steam line break have a frequency of 3% of that of HPCI. Similarly, in the case of the Main Steam Line (MSL) drain break, its frequency is 92% while the RCIC break frequency is only about 8%. Therefore, in the rest of this study, when discussing large or medium breaks, only the line breaks of the dominating systems are included; namely, the HPCI 10-in. line break, the RWCU 6-in. and 3-in. line breaks, and the MSL drain 3-in. line break. The small steam line breaks are mainly due to HPCI and RCIC bypass line breaks (it is the case of a blowd'own limited by the 1-in. bypass line). This will be referred to as the 1-in. line break even though the lines may be larger in diameter. The small liquid line breaks are represented in this BNL study by the RWCU 3/4-in. branches, and by the CRD SOV header piping rupture (reproduced from NUREG-0803(4)) which are about 1-1/2 in. equivalent dia-meter. Table 3 also includes, for each of the LOCA-outside-Drywell groups, the liquid or steam break discharge flow rate at two different times: (1) Initially, when the break occurs and flow rates are at their peak values, and (2) At about 30 minutes later afte,r coolant injection is established, depres-surization of the RPV is completed and operator takes control of the injection according to procedures, keeping the core covered. These flow rates values should be taken as crude estimates. They were obtained from NED0-24708(5) for the purpose of providing some indication of the time available for operator diagnosis and response. The NE00-24708 report provides this information for the entire spectrum of break size under con-sideration in this study.

~, _ _.- t 3-1 i e 3. ASSESSMENT OF MITIGATION CAPABILITY In this section, the effects of LOCA outside the drywell are discussed according to the three different groups: small, medium, and large pipe breaks (see Table 3). Based on these effects, some insight on the time available for mitigation is presented. s 'he first subsection provides general information on alarms available for diagnostics, containment sumps capacity and flooding data, and some crude informat' ion on the containment atmosphere temperature increase due to steam..or caturated liquid discharges. The next subsections describe the mitigation conditions for small, large, and medium LOCAs outside the drywell. 3.1 Reactor Building Information 3.1.1 Instrumentation for Diagnostics The following instrun entation and alarms are availabh to alert the i operator in the case of a pipe break in the Reactor Building: Reactor Building ventilation iso htion alarm Reactor Building equipment sump level alarm in the vicinity of the break Reactor Building floor drain sump level ala'rm Reactor Building flooding alarm at elevation 8 (see ' additional descrip-tion below) Area radiation monitor alarms ) Reactor Building Standby Ventilation Exhaust high-radiation alarms Area high-temperature alarms on elevation 8 and on the floor'where the break occurs Specific systeds nave their own break detec. tion instrumentation such as the RWCU, MSL drain, HPCI, and RCIC. Reactor Building low dif ferential pressure alarms. Most of these alarms are also sensitive to a small break LOCA of about 3/4-in. diameter but some set points Wil only be reached after about half an hour. i e'

3-2 The Reactor Building (RB) water level at elevation 8 is detected by two RB level monitors installed on the RB floor. The flood alarms are activated by the monitors when the water level is more than 0.5 inch above the floor. The sump alarms will be activated when the water level reaches the sump alarm setpoints installed at a level just below the level that activates the RB flood alarms. Sump alarm sensors are installed at various locations in the RB. The area high temperature alarms include the following: RCIC and HPCI turbine steam line space high temperature (7 sensors each). Isolation signal setpoint at 155*F (elevation 8) RHR space high temperature alarm (6 sensors) with setpoint at 175'F (ele-vation 8) RWCU space high temperature 118 sensors) isolation signal at 155 F (ele-vation 112) Main Steam line space high temperature (4 sensors per line) isolation signal at 200"F (elevation 78). Main steam tunnel containment penetration area high temperature (4 sen-sors) located in the area of MSL drain lines. Isolation signal at 140 F. 3.1.2 Sump Punps and Flooding Buildup Volumes The open area of the elevation 8 floor is approximately 5,500 sq. ft. This area is the total floor area minus the area occupied by equipment founda-tions, columns, drain tanks, etc. Based on this area, flood buildup on eleva-tion 8 is 3400 gal /in. The drainage capabilities at SNPS are: Reactor Building Floor Sumps - 2490-gal capacity Reactor Building Equipment Sumps - 1660-gal capacity Reactor Building Porous Concrete Sumps - 500-gal capacity. O

i l Table 3 l Summary of Frequencies of LOCA Outside Drywell Break Flow Conditions (*) Initiator Initial After 30 Minutes Break Location Frequency Initiator Stm/ Liq, Ib/sec Stm/ Liq lb/sec (Main Contributor) (Event /yr) Large Size Breaks -Steam 1400 Liquid 1200 HPCI(**) 3.6E-6 l. (elevation 8') 4 1 6" Liquid 1200 Liquid 700 RWCU 9.6E-6 (elevation 112') Total 4 1 6 1.3E-5 l 1 Large Interfacing Liquid 1200 Liquid 700 LPCI/LPCS 2.0E-6 r LOCAs elevations 87' down to 4 > 6" 8' Medium Size Breaks Steam 120 Steam 60 MSL Drain 1.0E-4 2" 1 4 1 4.3" Liquid 400 Liquid 250 RWCU 1.5E-3 i (elevations 112'-126') Total 2 1 4 1 4.3" 1.6E-3 i Small Size Breaks Steam 10 Steam 5 HPCI/RCIC(**) -3.0E-3 J (elevation 8') 4 < 2" Liquid 25 Liquid 12 RWCU Branches -1.5E-3 I (elevations 112'-150') j Total 4 < 2" -4.5E-3 l \\ (*) Approximate crude estimates of steam or liquid discharge through break from NED0-24708. 4 (**} Break can occur between elevation 66 and 8, but the other break locations discharge through a pipe chase to elevation 8. I

,..L 3-3 These systems have a total sump capacity of 4650 gallons. The total sump pump capacity is 640 gpm, as follows: Four 50 gpm equipment drain sump pumps (elevation (*) 9 ft) Six 50 gpm floor drain sump pumps (elevation (*) 9 ft) Two 20 gpm porous concrete sump pumps (elevation (*) 9 ft) One 100 gpm leakage return pump (elevation (*) 12 ft). The leakage return pump is designed to process radioactive water. If the floor drain sump pump indicators register radioactive material, all sump pumps will isolate. The leakage return pump can then be manually activated by the operator. In addition, only the leakage return pump is powered from onsite AC power. It can be inferred that if flooding is not arrested before it reaches the 1 ft level above the elevation 8 floor (elevation 9), the sump pump capacity may drop from 600 gpm to 100 gpm. This corresponds to accumulation of about 42,000 gallons. Furthermore, since this study considers primary water release, it is assumed that only the leakage return pump would be operating (other sump pumps would be isolated). RCIC, HPCI, LPCI/RHR, and LPCS are all located at elevation 8. It is assumed that they become disabled when water reaches 4 ft (about 160,000 gallons) as stated in SNPS-PRA.(1) 3.1.3 Containment Atmosphere The SNPS-FSAR(3) includes in Appendix 3C a few calculations of Reactor ~ Building temperatures for water and steam line breaks. Table 4 shows the re-sults of one calculation for the discharge of 40,000 lb of ' saturated water at RPV normal power conditions out of a 4-in, line break at elevation 112 ft of (*)lf water reaches this elevation, the pump is assumed to fail.

3 Table 4 Reactor Building Temperatures at Several Elevations Resulting from a 40,000 lb. Discharge Equilibrian Initial Maximum Reactor Building Temperature (*) Temperatures Elevation ['F3 [*F] Comments 8'-0" 104 < 140 40'-0" 148 6 3' -0" 183 78'-7" 194 112'-9" 217 Break location at 112'. Outside the pump room temp is 177 F 150'-9" 148 175'-9" < 132 (*) Reactor Building humidity changed from 50% initially to 100%. m

3-5 the Reactor Building. In this deterministic analysis, the break was assumed to be isolated by an RWCU isolation signal at 40 sec after initiation of the break. This break results in less than 5,000 gal at elevation 8 or less than 1-1/2-in. water accumulation on that floor. It is seen from Table 4 that a break of this size is rapidly affecting Reactor Building atmosphere condi-tions. The other calculations reported in Appendix 3C of SNPS-FSAR(3) are similar and lead to the assumption that conditions of 212 F in the Reactor Building elevation 8 will occur under the following circumstances: (1) A RWCU line break discharging more than 500,000 lb. This is approximate-ly the amount discharged from a RWCU 3-in. line break in 15 minutes (5 minutes for a 6-in. line). (2) A MSL drain line discharging.more than 100,000 lb of steam at RPV normal power conditions. For a 3-in. MSL drain line break this will occur in approximately 10 minutes. (3) A RCIC/HPCI 1-in. line discharging more than 15,000 lb of steam at RPV normal power conditions directly to elevation 8(*). For a 1" line break, this will occur in more than 25 minutes, and therefore 212 F con-ditions at elevation 8 from these line breaks are not expected to occur (**). Temperatures higher than 140*F in elevation 8 can result when steam is discharged directly to this elevation from a 1-in. RCIC or HPCI line continu- ,ously. (*l RCIC and HPCI steam lines are enclosed in piping chase which protects higher elevation against a steam line break in these systems.

However, for most steam line breaks in higher elevations, steam will exit at eleva-tion 8.

(**)The 15,000 lb discharge would cause the saturation conditions only if dis-charged during a very short time, which is not the case here. r -e :- ,y,- +y-, y, ,,m.m. ,.--,,,.--.,.-,u-,- ,,,.,-,r,----,.-.-.,v- -. - -

o 3-6 3.1.4 Procedures Given. a LOCA outside containment, the *NPS procedures dictate rapid manual depressurization of the RPV by the ADS. This action substantially reduces the flow rate through the break. If low pressure injection is pro-vided.at about 200 psi, break flow may become only about one-half of the initial break flow. Given an RB flooding alarm, the operator is required to: Monitor RB level to determine the approximate leak rate, and to ascertain the approximate location of the break (using additional sump alarms and high area temperature alarm) Monitor parameters such as line pressures and flow rate of the safety systems, as a leak may affect these system parameters If required and plant conditions permit, dispatch an operator to the RB - floor to visually locate the source of leakage. Isolate the break using the appropriate system procedure (HPCI, RCIC, RHR,others). 3.2 A Small LOCA Outside Drywell (< 1-1/2" Break Size) 3.2.1 Accident Conditions and Alarms The description that follows is based on an analysis by NRC staff of a pipe break equivalent to a 1.2-in. line break. This is discussed in detail in NUREG-0803.(4) The description in thi s section applies to small line breaks, in general, and applies to the SNPS. It does not, in particular, apply to SDV header pipe breaks to which the original discussion refers. The break. described is a water line break discharging 550 gpm (- 70 lb/s) initially. This is equivalent to a l'.2-in. line break discharging from the RPV at 1032 psi conditions. Several alarms are available to the operator as described in section 3.1.1 above. The most expected early alarms are from the Reactor Building radiation monitors and from local area high temperature alarms.

'~ o 3-7 NUREG-0803 cites a calculation for a typical BWR Reactor Building that shows a temperature rise to 110 F in 10 minutes and 140*F in 30 minutes for a discharge of 550 gpm at RPV conditions. (This amounts to about 130,000 lb over 30 minutes.) It may activate high temperature alarms if the set points .is 120 F, but it will not isolate HPCI or RCIC systems. The SNPS sumps and flooding setpoints are low (see section 3.1.1), i.e., .at 1/2-in. above floor level which corresponds to 2000 or 4000 gallons of water accumulation. Therefore, the water accumulation at the 550 gpm flow rate will cause Reactor Building sump and flood alarms to actuate within 5 to 10 minutes (assuming 35% flashing' into steam, travel time through stairwells and floors, and partial. accumulation in equipment sumps (up to 2,000 gallons). 3.2.2 Reactor Building Environment The water released from the break will exceed the local drain sump capac-ity, and some will flow to lower elevations through stairwells. Assuming that only the leakage return pump is available,(*) the accumulation of water at elevation 8 would be less than 0.13 in./ min i.e., it would take six hours to reach the level that threatens ECCS equipment availability. Thus, ample time is available for the operator to recognize the need to depressurize the reactor and reduce break flow. Note that Appendix 3C in the SNPS-FSAR states that equipment along stairwells is protected against dripping of 212 F water. During the initial blowdown, temperatures in the nearest area to the break can reach 212 F. The Reactor Building temperature is expected to rise significantly as shown in-Table 4 for a discharge of 40,000 lbs of saturated water at elevation 112 ft. This is a 10 minute discharge from the 1.2-in. line break described here. While it may result in high Reactor Building temperatures when discharged over a short period of time, it results in 110*F in the Reactor Building if discharged during about 10 minutes (see section 3.2.1). However, the temperature in containment will continue to rise due to the continued discharge through the break and may reach the 155*F RCIC/HPCI isolation temperature after about one hour. The Reactor Building Standby Ventilation System (RBSVS) of SNPS has a heat removal capability of less than (*)Radwaste system tanks capacity allows for about one day accumulations of untreated water at a 100 gpm pumping rate. n.

3-8 5% of the heat discharged by a 1.2-in. line break, before reactor is depressurized and the break flow is reduced. 3.2.3 Operator Response At.Shoreham, the operator will have a flooding alarm and high Reactor Building radiation alarm at about 10 minutes as-discussed in the previous section. For a small LOCA outside drywell, with feedwater still operating when the LOCA occurs, scram may not always occur immediately. Following the scram, the operator will try to keep the normal feedwater injection and therefore keep MSIV open. If the MSIV remains open (which is the more probable case), it may take a while before the operator will notice the abnormally high feedwater flow rate. It appears that the flooding and high reactor building radiation alarms will indicate that a small LOCA have occurred, and the increased feedwater injection flow may be used for verification. Therefore, it is expected that the operator will recognize a small break LOCA in the reactor building within about 30 minutes after scram. Unless the operator perceives a LOCA, he will depressurize the reactor at a rate of only 100*F per hour. In such a case it will take 4 hours to depressurize the reactor to 100 psi. and reduce break flow by about a factor of 10. As seen in section 3.2.2, four hours are available at SNPS, without flooding to elevation 12. However, in this case, the temperature in Reactor Building may reach 155*F or higher (*) between 1 and 2 hours and trip HPCI and RCIC, and most probably require depressurization for low pressure injection. These events would lead the operator to recognize the small LOCA outside containment with very high probability, if he failed to recognize it during the first half hour. It should be noted that unlike the generic analysis on NUREG-0803, the authors believe that recognition of a small break LOCA outside drywell at SNPS would be a high probability event mainly because of the improved arrangement for flooding detection at elevation 8 (relative to the arrangement assumed in NUREG-0803). High radiation and high temperature conditions in the reactor N A GE analysis estimates that the maximtn bulk temperature in the reactor building would reach about 140*F (See NUREG-0803l4)). h _. _. ~

= i 3-9 2 building will enhance the probability of recognition. This BNL study assumed that it is most probable that manual depressurization of RPV to reduce flow i and : enthalpy discharge through the break would take place after about 30 minutes to 1 hour into the accident. e The depressurization of the RPV may reduce flow rate and enthalpy of the water discharged through the break to a level accommodated by the sump pumps, and may reverse the conditions in reactor building 1.e., conditions may start to improve. It is indicated in NUREG-0803 that rupture of blowdown panels may be required to establish a path for leakage of hot humid air to outside con-tainment.(which is larger than the " natural" 100% per day leakage rate from reactor building), in order to improve the reactor building atmosphere condi-tions and to allow safe operator entry. As shown in NUREG-0803, depressuriza-tion reduces significantly the dose received by an operator entering the reac-l tor building. If an operator is re' quired to enter the reactor building to isolate a i break, it can be dona for a 1.2-in. line break with early depressurization (and low primary water activity). It would be possible to stay for an hour, and this seems to be sufficient for isolation purposes. Appendix 3C of SNPS-i FSAR considers 30 minutes to be sufficient time to walk through all SNPS ele-vations, locate a break, and isolate it. 3.2.4. Estimation of Core Damage Frequency The description of the event and the reactor building conditions follow-ing a small break-LOCA outside drywell were discussed in the previous sec-tions. These are now sunmarized in the form of an event tree in Figure 3, and quantified. Feedwater-and high pressure coolant injection are in general available under the circumstances of small LOCA. ADS, LPCI and LPCS have very low unavailabilities. The values for their quantification are taken from Reference 2. The events that are differently quantified are: (1) the proba-bility-- that at 30-60 minutes the operators take actions and complete rapid f manual depressurization, (X ), and (2) the probability of controlling the H condensate flow if required (V). The X =0.01 is taken basically from H l NUREG-0803 where 5x10-2 is used. The difference between NUREG-0803 and BNL values is due to the SNPS improved early flooding alarms which increase the probability that the operator recognizes the LOCA outside the drywell and follows the required depressurization procedure. e-r, ,9 ---.w-.-.,_.m- .,.,.--..,,,,_.,..-7..-% y..,,y_.,.._, 9 -m. ,ry..,y m --,, e n - r ,,.. w ,.-,.,-,,._..,..~,--m_.y,,._.,

5,..ti Feed

• lev H Pc#/ Lo17 0perot**

4PCI/ cal==ste Con 4cc h *"5 A ' A D3 F.Hw g pcy Pen p.mp feae a.8A,p.1 4<.<t.W Pr**' A**' t A.it.de **P Fre y er Rent R u. X Xu y '. v a via c f a s sz. 4 xf o-4 09 0/ .2. 4 E -7 s 0.ol / 5 E-3 h . o. o 2 g, g g _) G x o-'t i 2.! c-% i

0. g o.01

1. 1 90g-3 i

i i 0'G'N j O.01 l 2.7 t-t o j E E'# 36s-9 To tal= l.1 E-6 F.sn< - 3: E v~s m 9- ~ l-s r ~ ~s FL9 %R ' L och 1 h3< A e f%. w ra 0

s e 3-11 The V=0.1 is the common value used by BNL in Reference 2 for controlling condensate injection if suf ficient time is available to the operator (in our case 30 to 60 minutes). The V=0.02 includes a factor of 0.2 for.the possibility that no damage to LPC1/LPCS will occur even under the circumstances that the operator does not depressurize the reactor early, but rather depressurizes it at 100*F per hour rate, for 4 hours or more. In such a case NUREG-0803 indicates that entry to the reactor building may be delayed -for up to 20 hours. The LPCI/LPCS may survive the adverse environment in the reactor building for such a period, because they are qualified to sustain these conditions for at least several hours. - A factor of 0.2 (compare V ill on Fig.- 3) for the LPCI/LPCS availability apparently underestimate their availability. The event tree quantification yields a core damage frequency of about 1.1x10-6 per year for small LOCA outside the drywell, when it is asstned that the motor operated isolation valves are failed. Note that no distinction was made between steam and liquid breaks in the case of the small LOCA. The calculated core damage frequency would not change much if a distinction between liquid and steam break were made and apparently the flow out of a steam line break would be smaller after depressurization. 3.3 A large LOCA Outside Drywell (>,6" Break Size) This case was treated in the BNL SNPS-review (2). However, the assump-tion in the present study is that HPCI and RWCU isolation val.ves would fail to ' close. Only HPCI lines were treated in Reference 2, and a LOCA frequency of 2.7x10-s/ year was obtained. If we postulate that the isolation valves fail upon demand, a LOCA frequency of 3.5x10-8/ year is obtained for the 10-in. HPCI line break (see Table 2). The 6-in. diameter RWCU line has three isolation valves inside the drywell. Only one of them close automatically on sensing line break conditions in the RWCU lines. In Table 3 when no ' credit is given to these valves a break frequency of 9.6x10-6/yr is obtained as derived in Appendix C of this. report. In Reference 2, the three valves were considered (having

3-12 different isolation signals and one of them is of a different design), and, it was estimated that their failure upon demand would Le less than 2x10 /d, and the frequency of the 61n. RWCU line break would be about 10-8/ year. Thus, in Ref. 2 it was not further considered because the frequency of interfacing system LOCAs, was calculated to be 2x10 6/ year (see Tables 2 and 3 of this report for results and Ref. 2 for more details). The. interfacing LOCA frequency is also estimated in Reference 2. The results are reproduced in Tables 2 and 3. This LOCA frequency does not change under the specific assumptions of this report. The total frequency of large LOCA outside the drywell assuming isolation failure, and including interf acing LOCA becomes 1.5x10-5/ yea r. When this is used with the event tree of Ref. 2 (see Fig. 4), a core damage frequency of 3.0x10-8/yr is found. The 0.2 factor is the probability of operator failure to control the condensate system pumps' flow to the RPV in the short time available (about 10-15 minutes). In the case of a large LOCA outside drywell, the discharge to containment is about 1200 lb/s and saturation conditions in the bulk atmosphere of the reactor building are reached within 5 to 10 minutes. The ECCS equipment at elevation 8 would be flooded in about 15 to 25 minutes (the latter. number corresponds to 35% flashing). Thus, it is obvious that no isolation is possible, as it was.also assumed in the SNPS-PRA and the BNL review for large LOCA discharging saturated water or steam into the reactor building. This core damage frequency of 3x10 6/yr is 7 times larger than that given in Reference 2. This is because in Reference 2 the interfacing LOCAs were the dominant contributors. They are dominant when credit to isolation valve closure is considered. 3.4 A Medium LOCA Outside Drywell (2"< 0 < 4") 3.4.1 Accident Conditions Alarms and Operator Response The most dominant case of the medium LOCA is the 3-in. RWCU line break as shown in Table 2. The frequency of a RCIC 4-in. line break is small compared j to the total medium LOCA frequency of 1.6x10 3/yr; the RWCU 4-in. If ne break j i m --_m

4 ~ 0 a t' m e m m m o-e8 - e a a m m y u u u t SE,b u u u u m, o o o o w o a 8 c: Elb W cc e = "S 2 m g a w u g_ b b b b + + + MO 6 6 5 5 5 5 5 o .c O C Y " TW m e3 8" g kb 6 m% o = 9 w Ei 3. f CJ i c C" i : e-. = .o (A <3 3 b c C Ci se o mM Le Cl m mM YE m Z O =lC =C w E Z N 6Q ou >= a C +J CJ

e. c ci N

Ywa e_ W< 2 t a s' g d M ? l U PM, h 5 5 o $C m M m h >2 W w O 8 M 5 g- = ~

3-14 frequency is significant but the sections considered are relatively downstream and estimated to be 1/4 of.the total RWCU break frequency, whereas the other 3/4 are for 3-in. line break or less. Thus, our discussion in this section refers to a 3-in. RWCU line break. The RWCU is located at elevation 112 ft to 150 ft. At 150 ft the demin-eralizers are located, which process water at low pressure and at about 125'F and, therefore, not considered. Thus, the break location of significance can occur at the 112 ft or 126 ft elevations. On these elevations, the line are enclosed within cencrete shields providing physical separation from all safety related equipment (see App. 3C of the SNPS-FSAR). Table 4 present the approximate temperatures in the reactor building fol-lowing a RWCU 4-in. line break at elevation 112 ft in the RWCU pumps room. It is estimated that about 10 times the amount discharged in that case, i.e. 500,000 lb, would result in saturation conditions in the reactor building. This will take about 20 minutes if the flow rate of Table 3 (400 lb/s) is assumed. It apparently will take longer because of the decrease expected in the break flow due to depressurization after a few minutes (up to 10 minutes). It is expected that the blowdown from the break will cause immediate MSIV closure and loss of the feedwater system. In about 10 minutes or less, the temperature at elevation 8 will reach 155'F and trip the RCIC and HPCI, which started a few minutes before that on low level (L2). Therefore, in this case, it is immaterial whether the operator depressurizes the RPV, because early automatic ADS actuation is expected for this case. The water discharged during the first 10 minutes would flash ('35%) and the remainder (about 20,000 gallons) will cascade through the stairwells to elevation 8. Appendix 3C of the SNPS-FSAR considers this effects and states that no safety system would be affected. This accumulation is equivalent to 0.5 ft and will result in flooding alarm in the control room. The radiation and temperature alarms are expected to be on in many areas of the reactor building. Therefore, it is believed that the situation of LOCA outside drywell and the reactor building adverse conditions and flooding would be recognized with high probability within the first 10 minutes. Earlier recognition of the LOCA and depressurization of the RPV would not change

3-15 much the progress of this accident sequence. However, if operators fail to recognize the event and follow the procedures (which call for keeping RPV at low pressure and controlling the injection flow), then the reactor building conditions may severely deteriorate. The depressurization would apparently happen at about 10 minutes.. Then the LPCI, LPCS and condensate pumps, may all inject water to the PRV, and dis-charge a large amount of hot water through the break. While this hot water would have less enthalpy than the saturated water discharged during the first 10 minutes, it has flooding potential because of its high flow rate. Flooding may occur in an additional 30 minutes if the flow rate to the RPV is not reduced by keeping it at the lowest possible pressure without uncovering the core. This is the operator action specifically required for the case of medium LOCA outside the drywell. In such a case LPCI/LPCS may maintain core cooling for long period and condensate would not be needed until several hours into the accident. 3.4.2 Estimatinn of Core Damage Frequencies The estimation of core damage frequency for the case of a medium LOCA outside drywell is shown in the event tree in Figure 5. The initiating event does not distinguish between water or steam line breaks. They are considered similar because even though the steam discharge through the break is smaller, the impact on containment atmosphere temperature and pressure is about 5 times higher for a steam line break than for the case of a similar size water line break. In the long run, after the RPV is depressurized, the flow out of a steam break may be significantly smaller if the core is not flooded so that water is discharged through the break. If the water level is kept below level 8 (L8), then the steam flow out of the break is expected to be relatively small. Thus, it may not be sufficient to create flooding sufficient to damage the ECCS equipment. The liquid line break is therefore the dominating case. Thus, the event tree starts with the medium LOCA frequency from Table 3. The feedwater and RCIC/HPCI are assumed to be unavailable. Depressurization by ADS is

-4 4 a r. ._..m+..Ja _:.-_ _~. ...A- -._.-i+-*..J-..--*--_ei -e-y s e E* a le s l .I. ~ ,, %, -4l 3 a n k vu Y $8 h4 4 ? n u m v e u e o J e 0 e 4 4 4 W 4 mg w a W u! 73 dis = w h m % e.

• s w

o b A 2 4 o s N 1.I 4 ! =1 =3$ b ~ u t w 4 e pg *e o 5 C + $, 9$ J._h g w A N t > d 4 K h a t w 3 s 1 es 4 hh K W N vg s NK UA o e, \\q U d M U d o =<j a \\ b o e 4y 6 0 W 4 0 0 e LL<d e4 g 98 5 4.i.e $. i e= (s o = . - - - ~.

~ o 3-17 considered to occur at about 10 minutes into the sequence. The low pressure injection systems will start to flood the core. Therefore, operator action to control the injection flow rate is needed to reduce the impact on the reactor building and gain time before the condensate system would be required. If the operator recognizes the need to control the injection, then the condensate system pumps will also be controlled at a later time with a higher reliabil-ity. If the operator fails to control the injection, less time will be avail-able to control the condensate pumps injection because they may be needed at about 10 minutes into the accident. The values used for the probability of successful operator action are thought to be on the conservative side given the time estimated to be avail-able. Therefore, the core damage frequency for mediun LOCA outside drywell may be smaller than 1.4x10-5 for the case that no credit is given for RWCU isolation valves. On the other

• hand, the phenomenological assumptions used may not be realistic and may underestimate the break-flow and Reactor Building conditions, so that less time will be available for operator corrective action than assumed above.

4-1 4.

SUMMARY

The BNL review (2) of SNPS-PRA estimated a core damage frequency of 4.2x10-7 for LOCA outside the drywell in the SNPS; this is mainly due to interfacing system LOCAs. In this study, an additional assumption was introduced at NRC request: namely, that isolation valves would be treated as failing to close upon demand. The only exceptions to this assumption are the MSIVs and check valves. The effect of this assumption is shown in Table 5. It is seen that the core damage frequency increased by a factor of almost 50. The leading contribution comes from medium LOCA outside the drywell; in par-ticular, the RWCU 3-in. line break is seen to be the most important (see Table 3). Table 5 Core Damage Frequencies for Unisolated LOCA Outside Drywell Class V Core Damage Frequency Isolation Valves Isolation Valves Assumed Assumed to Fail to to Close on Demand Close on Demand Initiator (from BNL Reference 2) (from this analysis) Interfacing LOCA 4.0 E-7 4.0 E-7 Large LOCA Outside Drywell 2.0 E-8 2.6 E-6 Medium LOCA Outside Drywell 1.4 E-5 Small Loca Outside Drywell 1.1 E-6 Total 4.2 E-7 1.8 E-5 Table 2 provides the informatkon on the most important isolation valves whose failures contribute to the results of Table 5. RWCU isolation valves are the most important. Next,'but by far less important, are HPCI and MSL drain isolation valves.

q-4-2 Tables 3 and 5 show that under the assumptions used in this study, the core damage frequency from LOCA outside drywell is dominated by the RWCU medium LOCA breaks. Also, the large LOCA contribution comes mainly from the RWCU system. Therefore, it should be noted that beside the inboard and out-board containment isolation valves, the RWCU also has two additional isolation valves that do not receive an automatic signal to close when a line break occurs and are available for timely remote closure. This action can be taken half an hour 'after initiation of the accident when the reactor is depressu-rized, and before the loss of low pressure injection. 4 4 e

A 5-1 5. References 1. "Probabilistic Risk Assessment Shoreham Nuclear Power Station Long Island Lighting Company, Final Report," Science Application, Inc., June 24, 1983. 2. D. Ilberg, K. Shiu, N. Hanan, and E. Anavim, "A Review of the Shoreham Nuclear Power Station Probabilistic Risk Ass essment," NUREG/CR-4050, BNL-NUREG-51836, December 1984. 3. " Final Safety Analysis Report Shoreham Nuclear Power Station Long Island Lighting Company," SNPS-1 FSAR (Revision 31, August 1983). 4. " Generic Safety Evaluation Report Regarding Integrity of BWR, Scram System Piping," NUREG-0803, August, 1981. 5. " Additional Information Required for NRC Staff Generic Report on Boiling Water Reactors," GE Report NEDO-24708, December 1980. 6. Reactor Safety Study- "An Assessment of Accident Risks in U.S. Commerical * ] Nuclear Power Plants," WASH-1400, NUREG/74-014, October 1975. 7. S.L. Basin and E. T. Burns, " Characteristics of Pipe System Failures in Light Water Reactors," EPRI-NP-438, August 1977. 8. W. H. Hubble and C. F. Miller, " Data Summaries of LERs on Valves at U.S. commercial Nuclear Power Plants," NUREG/CR-1363, EGG-EA-5125, May 1980. 9. Oconee PRA, A Probabilistic Risk Assessment of Oconee Unit 3, NSAC/60, June 1984. i f e 4 f 1

se A-1 Appendix A Pipes and Valves Failure Rates A.1 Pipe Rupture The nain data sources used for probability of pipe ruptures were the Reactor Sa fety Study (6) (RSS) and the EPRI-NP-438 report (7). In the Reactor. Safety Study, pipe rupture rates are based on the large amount of data prior to 1973. The EPRI report includes data for an additional two years. Even though it does not change the RSS results on pipe break rates, it provides more insights on the failure mechanisms leading to pipe breaks, mainly vibrations and pressure surges. It also points out that expansion joints and reducers may be at locations more susceptible to breaks. In the BNL study, reducers and valves were considered as rupture locations, in addition to pipe sections. l The SNPS-PRA(1) uses the RSS data for pipe breaks. However, it distin-l guishes between pipe sections which are " Break Exclusion," 1.e., are designed to criteria provided in Appendix 3C of SNPS-FSAR(3), which basically allow l for larger design margins and higher quality control of these sections. These increased margins are assumed by SNPS to reduce the failure rate of these sec-tions by a factor of 10. BNL accepted this assumption, and the basic values used in the study are similar to the SNPS-PRA and are summarized in Table A.1 below. l l l 1 .i 4 m -

1 st i. Table A.1 i Pipe Rupture Rates Computational Mean Assessed Range (non break-exclusion Computational Break Non-Break Component pipes) Median Exclusion Exclusion Pipes > 3" dia. 3x10 3x10-'/hr 1x10-18/hr 8.6x10-II/hr 8.6x10-18/hr per section I Pipes 4 3" dia. 3x10 11 - 3x10-8/hr lx10-s/hr 8.6x10-18/hr 8.6x10-8/hr ~ per section i 9 i f 4 l'

A-2 The pipe rupture data of the RSS is applied section by section, where a section is defined (RSS, page III-41) as follows: A section is an average length between major discontinuities such as valves, pumps, etc. (approximately 10 to 100 ft). Each section can include several welds, elbows, and flanges. In this study, piping was also divided into sections where discontinuities were considered to be: -- Valves -- Reducers -- Pumps -- Heat Exchangers Appendix C presents the details of the pipings and their division into sections. A.2 Valve Failure Rates The main sources used for valve rupture or excessive leakage failure rates were the Reactor Safety Study (6) and NUREG/CR-1363 report (8). The values of the NUREG/CR-1363 evaluation are about a factor of three higher than those in the RSS (see Table A.2 for comparison). However, the NUREG evaluation includes also 'small leakages such as from packing failure. Similarly, the internal leakage rate of check valves given in the NUREG evaluation includes many small leakages which are just violations of the Technical Specifications limits, and too small to be considered in this study. The NUREG/CR-1363 evaluation reports about 130 LERS under the title of " External Leakage / Rupture." However, no case of valve external rupture has SNPS-PRA(I) estimated from this list that a value of 1/18 may be occurred. used to modify the RSS rupture rate to better represent severe rupture of valves. This value of 1/18 is also used in this study. Based on the above, the BNL study essentially adopted the SNPS-PRA approach, i.e.:

t Table A.2 Valve Rupture or Excessive Leakage Rates l Computational Mean Assessed Break Non-Break Component Source Failure Mode Range Exclusion Exclusion [hr-13 [hr-1 3 [hr-13 Check RSS Internal Leak-10 10-6 3.8x10 7 Valves age (Severe) NUREG/CR-Internal Leak-1x10-6 1363 age (all sizes) Check RSS Rupture 10 10-7 2.7x10-9 2.7x10-8 Valves and Motor NUREG/CR-External 7x10-9 7x10-8 Operated 1363 Leakagg/ Valves Rupture O l l i-

9 Table A.3 Motor Operated Valves Failure Rates Value Used in Component Source Failure Mode Assessed Range Mean Value BNL Study RSS Failure to 3x10 4 - 3x10 3/d 1.3x10 3/d operate Motor (include command) Operated Valves NUREG/CR-Failure to 8x10-3/d 8x10-3/d 1363 opeaate (MOV) (forBWRs)' (irclude command) NUREG/CR-Failure to 6x10-3/d 6x10-3/d 1363 operate (for BWRs) (w/o command) i Command Failure 2x10-3/d 2x10-3/d Failure of of Inboard and Both MOVs Outboard 110Vs (Inboard and Outboard) Oconee MOV 10-3/y l PRA(9) Spurious Opening l O k

T ~ r Table A-4 A Comparison of Frequencies of Loss of Coolant Accidents ~ i RSS EPRI-NP-438 SNPS-PRA LOCA I Pipe Break Mean All Sensitive Mean This Study: Diameter 90% LOCA Pipes P eS(*) LOCA LOCA Outside (Inch) Range Frequencies (Mean) Frequencies Drywell g n l l Smal1 LOCA 1x10 4 - 1x10 2 2.7x10 3 -10 2 8x10 3 8x10 3 5x10-3 1/2" - 2" Medium LOCA 3x10 5 - 3x10 3 8x10 4 3x10 3 3x10-3 1.6x10-3 2" - 6" l large LOCA 1x10 s - 1x10-3 2.7x10 " -1x10-3 7x10-4 7x10-4 3.5x10-5I**) > 6" I*I lt is assumed that 10% of plant piping are LOCA sensitive pipes.(Ref.1) (**)The large diameter pipes are " break-exclusion" and are assumed to have 1/10 of the RSS rupture rate. I

u A-3 (1) Use of RSS failure rates for valves. ~ (2) Apply a modifying factor of 1/18 to the RSS valve rupture data. (3) Distinguish between valves which are in the break exclusion zone and those which are not. A factor of 1/10 is applied to the rupture rates of the break-exclusion valves, similarly to the factor applied to the pipe section they are located on. To summarize, the value used for valve failure rates were: check valve internal leakage: 3.8x10-7 x 8760 = 3.3x10-3/ year valve rupture (break exclusion): 2.7x10-9 x 8760 x 1/18 = 1.3x10-6/ year (non-break exclusion): 2.7x10-8 x 8760 x 1/18 = 1.3x10-5/ year. For simplification of the analysis, the valve rupture rates were also used with other discontinuities between pipe sections, such as reducers or pumps; this may be a conservative assumption. In addition to valve rupture and internal leakage, other failure modes of motor-operated valves were needed in this study. The additional failure modes and failure rates used are summarized in Table A.3. A.3 Comparison with LOCA Frequencies The analysis in the main part of this report involves a large number of pipe sections and valves. In general, more pipe sections and valves are located outside the drywell. Thus, the frequency of LOCA outside containment should be a large fraction of the plant's LOCA frequencies. Table A-4 compares the results of the LOCA frequencies from this BNL study with the RSS results (table III-6-9 of RSS), the EPRI-NP-438(7) results, and those of the SNPS-PRA. 5

i, ~. Table B-1 PROCESS PIPELINES PENETRATING PRIMARY CONTAllMENT o (Numibers in parentheses are keyed to notes on pages B-6 and B-7; signal codes are listed on page B-8)., o. [ >C <= WW SKall65 DALW ANG/OR retunaf intaleMist emets gatW5 meMient htLAllW IS SM kAlet POWER POWER C105 tug monnmL g-ef PER FIPE Sitt Ft Matt lytt leONu 10 CIO5E 850tAllen IIME(SEC) SIAIUS las ItAs tersi sluts 15d8.. Alt.s (22)- -. _. _EaC tlets llut (Is.) Cinelatsget (6 ___ _ 22. ). __ _ _.6) (5.6) SlQeAL (to) (s 9) stM4aK5 3 (5 m I-D.O.C.S nele Ste.e 55 4 8 24 leside AS Elebe alt /AC/9C Alr/ Spring B.C 9.f.P.S.T.Ist 3-5 Open (1) in 1 24 entside A4 Glebe Air /AC/DC Air / Spring 8 C.B f P.R.I Ist I-5 Open (3) ,{ s Mete sce.e Iles Brete and luity-55 4 .I 2 beside le Clase K AC t.C.S.E.P.t.l.nd 0 4 - Deen testage Castrel Systee 55 4 3 2 1/2 Atside le Glete AC AC Closed (64 s 3 3-24 lead ster 55 l I la Beside Chest flew pewerse flees Reverse flew II/A S em [ t I le 0.aside flC flew se,erse lie d be esse lle=/f, m/4 , opea (l* s Air /5pring E fst en a-2e feed.eter 15 1 ~1 It Beside Check flaie meverse fles Sewese flew N/A Open M le outstJe WIC time me,ers,lle.1 neverse fleis/f G. h/A D en (11) g, i i Alr/5prisg hM n et a-I Male Steen llee trate 55 1 1 3 leside se Sete AC AC 0.C.S.f.P.R.T.tM A 16 Open O s' - 1 3 beside IG Este DC GC 8.C.S.f.P.R.I.pM.0 16 Opee 7 u,, g E-4 kW(a tsee frem N W SS' 1 "I 6 Inside Mis Gate AC AC 4.J.WI.A 30 Open m

• _5 s

1 6 &stside 64 Cete DC DC 4.J.W.V.Ree,8 30 Open y \\ C \\ 35 aus 56,14.m Coelleg from SPW 55 3 8 20 loslJe le mate AC AC A.F.U.9M 28 (lated fD .g I 20 btslJe IIS E**e SC OC A.f.U.tel 23 Closed ) \\[ 3-&a.a edtlejectiestiesto$strce-55 2 1 24 leside WIC flee neversetid sieverse fles m/4 Closed (I) lettee Systen betare 1 2 lastJe 86 Sete AC AC A.les IS Closed fD i t 24 0.astJe le Sete AC AC aut 24 Closed (52) r+

s-PA.s se. Centelasret Spray tryhell 56 2

I le -entside De Sete AC AC f.G.Ist St Closed (2) O 1 le 0.tside se Ansie AC AC f.s aN la Closed (2) o m On l E-04.4 esen. Centelament Spray 56 2 3 '4., deutside se Elebe AC AC f.G. Ret 25 Closed (2) O l Simpy.essi.e Ch eer 1 86 thstslde le Sete AC AC f G.let Il Closed (2) <1 16 Outside le Elabe AC AC f.G.tn 29 Cleted (2) "5 l l CD E SA.S.C.S em Pump Section 56 4 3 20 beside le Sete AC AC lol, 106 O =en (II) C,, t n a ta f I

t. Table B-1 (continued) PROCESSPIPELINESPENETRAT!bGPRIMARYCONTAINMENT I wew leC4tten yetw Amares i \\ telnRag Inomie Wat W S WIIIDEAL NLAllW IS Of1hABOR PindCA PCMCR CIS5tes IIDAMAt stelAleggg i et Fts PIM SIN thil;Aag lytt le SPts IS C105C 150tAfl0E IlfE (SEC) SIAlui nu n4:loes laws 35at4]En (223 i seC timC5 lauf (Is.) taalAlwiat ('.22) (5.6) (5.6) SICaAt pe) (a.s) n Man 5 -304 me test lies a,8we u i 1 m Alside Ma sieme AC AC f.s.mi 7e Closed (2) to klysesslee Chad er, $=pprestles Feel Cle aw hetwo. I 2 6 Gutside IS Este AC AC 4.f.We i + he Steen C=adenslag Discharge. I I 4 Swaside MS Gate AC AC f.E.est 26 Closed 31 Closed he Mieles flaw. I l 4 Setside MD Cele AC AC we 20 Opee (16) i Case ipeep lest time, and 8 I le Atside MG Elebe AC AC f.G.ast 67 Clased Cwe sp<ey neelme flem I l 3,

a..ide-me sete AC AC ist is Open (16) s-les lhe test time netwo to 56 3

3 96 Atside IS Glebe AC AC f.E.BM g 79 Closed (2) I hspeestles thenber. KK n'alaus flow, l 1 2 tutside 88 Glebe BC K est 58 Closed (M) l' IsPCI Memlem flew. I 1 4 butside le Glebe K K* Mt i me Ste Ca lcastag D6scherge. I 1 4 8.tside le Sete AC AC t.E.918 j 26 Closed i 29 Closed (16) 4 h e Minesun flow. I 1 4 A tslae se Gate AC Care Spray test Slee.. I 1 M Asside 98 Clebe AC AC ist to C en (M) i Co.e Speep estates flew. and 1 I 3 8.tside Ms Cete AC AC tot 4 Opee (M) AC f.G.RII 67 Cleted Sellef Valve 94steerge free AMA l 1 2 Atside hellef yelse Nigh Blffer-Spring m/6 N/A Cleted i. 5=pply to ACIC Famp bctles entist Peesswa 3-11 tese - Iseed Spaar llae le arv 55 I I 4 Ins 6de see Cete AC AC 4.f.u.est 20 Closed 'i 1 4 Atside 88 Globe DC OC A.I.M. All Il Cleted I-82 hPCI Imeblae Steen lelet lese 55 1 1 le lostde le Sete AC AC E.Ift II Opee (7[ I I laside MS Globe AC AC s.mt 12 l le outside 80 Cete K pc E.fot 45 Cleted i 7 l D es l'7 i i I I Gutside Mb stehe K OC s.#4 12 Open i 7,i I-ll serCI Imeblae faheest l 1 18 Outside le Este SC 0F wt 102 Open 4 2 IS Atstde Check f)ew neverse flew ucverse flee N/A Closed 5-14 5 pere Ili) a-IS NPCI Pus,Sectlen M l 1 M Atside. se sete BC K t.Ist 78 Closed 1 E-M BCIC Iwbies Steam Inlet llae 55 I I 3 Inside se sete AC AC t ast n Spei ')l I I .Beside 840 Elebe AC AC t.ht 12 Opee 7l i t I Atside See Gate K OC st.081 16 Closed - 7 i l 1 0 4 side se fil he DC OC E.let 12 'opee . I i ? i I I t

s TableB-1(continued) l l PROCESS PIPELINES PENETRATING PRIMARY CONTAINMENT- ,-i l l l 1 PalltAar VAs W IOCAlleg WAt W ANh/08 i 198485 VAI WS hoftlNAL DilAllW 14

CPikAIOR, P0utt POWLR (IAalatuit sel Of Pt R PIPt Siti f aleiAtf Iff(

to Orts 10 (t0tt $581Aff84 Ilt( (ifC) 51 AIMS fl055NG 80ee14L Pi rat :6AllOrt5 t illf 5 IS0t Alf B (22) CaC t lut1 tibt (IN-) (0881AlWit hi (6.22) (5.6) (5.6) SICatAL (10) (s.9) ettenas5 i j E-li DCit ist ine Taliaust 1 1 8 Outside IIB Gate K 3C sut 3r Open ) { 2 8 Outside theck fles Reverse flew Deverse flew s/A (leged (53) I-IS SCIC Vacuus Pump pischarge $4 1 1 2 Setside MS Step theck fleu/DC Aes. Flew /DC hve. flow /ase 13 (losed (13.21) t t 2 Outside Check flow Reverse flew Reverse flow h/A Closed I 19 M IC Pump Secties 1 1 6 Outside MS Cate BC SC Ast 35 Closed I 204.8 fare Spray P.se Discharge te 55 2 1 le laside VIC fles Severse flew Deverse flew m/A (lesed (3) RPy t 8 2 Inside MS Clebe AC AC fut 14 (lesed I 1 la Outside M Cete AC AC tot 43 (lesed (13) a 234.8 tese Spray Pisap Sucties 56 2 I I4 Outside le Gate AC AC 888 36 Open i s-220.B es(t(W to hecirc. Pump med 57 2 l 4 Outside MS Cete AC AC 8 80 23 Opre Nter foolers I-234.s ' Watt (W frun Decirc. Pump and il 2 3 4 Outside le Cate AC AC ase 23 One Ntor (coles s I-244 to N SpCLCW te Orywell taill toelers 8 l 3 laside Check flew Reverse flew Reverse flew t/A Open 1 3 Outside le Sete AC AC f.G.I. Ret 16 aa r X FaA.D as(ICW fram Drywell hit Coolers be 2 3 4 laslee MS Case AC AC f.C,2. Int 20 ope i 4 Outside te Cate AC AC f G.I Itt 20 C es l E-26 Purge Air to Orywell M l 1 18 Inside AG butterfly Af/Afr Spring t.let 5 (lesed (17) I 18 OutsIJe A0 Butterfly AC/Alr Sptlag B 1W0 5 (lesed (17) s.27 Purge Air fraan Dryuell 56 8 l 18 Inside A0 Butterfly A(/ Air Spring 8.lut 5 (lesed (17). I IS Outside AG Sutterfly AC/Alr Spring t.let 5 Closed (li) h i I f e k e

1 Table B-1 (continued) PROCESS PIPELINES PENETRATING PRIMARY CONTAINMENT l'atteet 9AtW l0(Alles VatK AmeMR lasele VAI WS 90peltAL Dil Allt( IG (#f R450R Peuft POWER Cteilet menMAL (tmIAlwumi Of fit Plf5 SIM thiraaf lirt 34 Orf 5 le C105C 350tAllem IIME (5ft) SIAlus Piht thalloN5 Ilut5IsetA1(8(77) Eac 18:45 lint (ie.) C4mI Alesu al (6,22) (5,6) (b.6) SICmAL (le) (e.g) steinants 3-it Purge Air to Suppressten Ch.neer M i 2 le outside A0 Butterfly AC/Alr Spring I. 881 5 Closed (l)) ~ 3 29 Purge Air free Suppression M l 2 83 Outside As Butterfly AC/Alr Spring L, aft 5 Closed (17) Ctamber .a B-30 54eele feeleal wa APW 55 l 1 3/4 Inside A0 &Ie6e AC/Alr Spelag S.C.Ast 15 Open 8 3/4 Outside A0 Glebe AC/Afr Spring 8.C,hel 15 Ogu s 38 Iquipment Drains fran Drywell M i 2 3 Outside MB Cate AC AC A.f,000 16 Opee I-32 fleer Orains free Orywell 56 5 2 4 Outside 80 Cate AC AC A,f.IWE 16 Opes I ~ I-33 Spare ~ (15) 'I-34 Spare (15) ~ l I-35 Spare (15) 3 36 Standby tiquid Coelaat to RPW 55 1 8 1 1/2 leside (beck flew Severse fles Severse flew N/A CIssed I I I/2 outside Cheth flew Deverse flow Reverse Elem N/A Closed 2 I l/2 Outside taplesive AC h/4 Int lastantly Closed i { 3-37A Nitrogen /Aer Purge for IIP 57 8 3 3/8 Outside Check flew Reverse flew Beverse flew N/A Open f. 1 5-318.C,9 flP Orive Eulde Tulses il 3 5 3/8 Outside Ball AC 5pring 8.5 Closed (14) l 3/8 Outside implosive N/A SC aft lastantly Open (14 1 Shear I-38 IIP Delve Cuter fut.es 57 l 1 3/s Outside ball AC 5pring 0.5 Closed (14) I 3/8 Outside implestve N/A DC tel lastantly Open (141 j Shear ~ t I-394.8 lastriseent Air to suppresslen M 2 3 I Outside Check flew Beverse flew Reverse flew N/A Open g Chamt.er 1 1 Outside No Glebe AC AC f.G,aN 5 Closed l l i f i I f h j

Table B-1 (cofitinued) PROCESS PIPEl.INES PENETRATING PRIMARY CONTAINMENT. VAI W 10 Call 0E VAtW WelV0R FNittav Issett Wat W5 NonINAL B(L AIIW 10 OrtkAIOR Poult POWS (1951NG. NinetAL @ lAlwuel of PE R PIPC SIM PelliARf IIPC 10 OPCs le Cle'4 150tAIION IIMC (5fC) SIAlr5 I-[Id ikAlltwas llNES ISOLAlto (22) GOC llN(5 llNC (IN.) CONIAlesENI (6.22) (5.6) (5.6) SIGNAL (10) (s.3) htMAaES 40 Spar. (is) i 1-41 HPCI Vacenan Steater M i I 2 Outside se Globe K K F and I. M 13 opee

• 2 IS Butside towet flew newge se flem Aw,er}t,l lin __ NJA

[]gged j 1-42 NCIC Vacuum Breater M l l i 1/2 Outside PC Globe SC DC f and I. tit M Spee 2 0 Outside Check f les. Severse fluw Reverse flow N/A Closed I 43 kodt Relief Valve DIscherpe M I N/A N/A N/A N/A N/A N/A N/A N/A N/A vacuus Breater. Ridt Neat iachanger Vent. 2 2 1 Atside le Globe AC AC M 30 Closed t Sult met Cachanger, and 2 I I Outside hellef Valve. high Pressura Spring N/A N/A (lesed Is'Cl Ste.a Supply to SNt Heat fathanger 2 3 6 Outside Relief Valve Nigh Pressura Spring N/A N/A Closed E-44 Contalanent Atenspheric Centrol M i I 6 Outside MD Cate AC AC RN 31 Closed 4 from Suppressten Chas&er and I 4 Outside MD Cate AC AC Ret 20 Closed i Deruell floor Seal pressurlas. 57 I I t/2 Outside M0 Globe AC AC Eli 6 Open tien 45 Contalmuent Atmospheric Central M i I 6 Outside MD Gate AC AC BM 31 Closed from imperestlee Ches&er. and I 4 outside MG Gate AC AC BM 20 Closed Drywell liner Seal Presserlaa. 57 I I 1/2 Outside MG GImbe AC AC kN Open alca E 46 Containment Atenspheric Centrol M i I 6 Inside M0 Gate AC AC kN 31 Closed feen Dryvell l 6 Outside PC Eate AC AC eM 16 Closed 6 E-47 Centelnuent Atmospluric Coatsel M I I 6 laside HD Gate AC AC 8 86 31 Cleted l from Drywell I 6 OutslJe

1. 6 Este AC AC het 16 Closed CRD Is. sert and Withdraw times 55 137 3

3/4 Setside GIcbe manal mamal N/A N/A Open 120) 137 1 1 Outside Globe manal Naval N/A N/A Open (20) a5 3 Spese (15) 45-2 Spare l15) i l i l I t 't

w it i s .s e 5 Ka e t ) ) ) ) ) ) A i 5 5 5 S 5 h ( 1 I )) 5 1 t 1 ( ( ( (( ( ( 1 44 1 L5) d mpg d dd d d d dd e etl. e ee e e e ees gAe s ss s s s sk o nn nn oI( o oo o o o o<l ee ee NS l ll l l lC Ce OO oO lnC pp pp C CC C C ) EE C NS) t(OI A A AA 50E( / 0/ t 2 2 t// 44 44 tN N 1N S 3 3 iN 11 11 Cl l e w N o 0L l lA f lN e AGI M s tem Ml p 40S R. r l R.

t. h.

e 5 A A f. vA G. G. G. G. 1 i N RN A l S DN fi ff e M e/ / N/ n uo t l s f Ro) El6 e T WC g g sl O 5 n n ra N. P0( i i vn eu E 1 r r M p Cp C C C ea CC CC S AS A A A RM AA AA N I L AT e N r r N tE) u u O rP6 s s l uO s s a C o 5 he he wu Pa( gr gr lon l iPCiP C C C a CC CC ) Y h AH A A A fM AA AA d R e A u M 8 e e v v 0 n I /R l l a a R 0O ) W ev ee ee i nIi2 t P AAt2 b e e e bt bb a f ot t t t oo o-Rl n fi6 e e a a a k ll lI o G WP i lGl C G C ce Gs GG ( l l et c N tC e ea 0 l e ha ee O0 A i S lB M t l CC tl NM I V ( T 1 A l R o0 I i T l1 N B l f M E AEnP CVAN e N 1 Olel e ee e e e e e e E Illa d dd d d d ed ed ed l AaI is ss i i i is is is ii di di di P lPN s ts s b Wi O t tt t t st st st tB C u uu u u u au nu au a S A O OO O O O lO IO lD T E V N 1 I L2 L A1) le5i // // t. 22 22 EP iMCl 11 11 0P( e I NI 6 11 l 6 6 Ii iI 'lI P P 5 S Waif t l S li 3 21 2 2 2 II li 1I P E A t V CO R R 5 ( ( P tfl e sOi 2 22 1 - 1 3 l i I a l t .W 5 5 ~- M M M 6 6 6 5 l l o o t r r e n a d Ic n t t e o n n I a e o l i t t / C C a a ) L p i d 2 t u cr cr d i H 2( I n n ie i e. a a e a rb rt R m Rm 0 o V e ee en e lt e s r lt f t l h e ha i l s l s rr C ph pi l y ee sC sC A ey ey At gg l o o w s. u s. 0 lp nn o tmn tmn e yl yl 5 p aa o o o c ru e u u hh P Ai Ai i DS bS 1 S.cc n ts t s r t g t g s s v a n 5 erfE on ne ne e nn nn ( u er er S ei =i N se isu s p laS s a y nu tnu a a e gt t mp np t naa sD ap l emr r s=o ll o S ee e e u u w t e e hHn r, laS e e l lai la6 e e t pm t t r t n n r s l c r r f s88 pe no no a a a p p PE11 uP ot ot p p r oM oM p p S S H 99 S C f S S D f f S 5 5 t1 0 n0 fil AHl ANA 0 8 2 3 4 5 lMIk 1 4 5 6 2 8 9 1 1 1 1 1 1 Ai alf 5 5 5 5 5 5 5 5 5 5 5 5 5 Pmn 3 5 1 1 1 1 1 3 1 1 1 3 1 ti (P f

.as am. an. a.= a a.m. m s.p. .a. esu. em .su. s .a .a .n a.n. a g O O O =O O O O O O OO O gC~

== h;24

• 1 I.t

..j 23 3* da* dai so 55, WY ww ww ww: u WM- --a E Y" C: 24 22 <e =.

== 24 24 22 22 me 5 3 3 33 33 W 25 5 gr t t tt E.= I.r .E.r

..=

E E il i l s x x =a-:*; FF w s. O* WW WW .g.A O a D D JW JW JJ JJ ,.c k-a w a *. o '." 45 3 I II 33 ~- I" WW Wit WW W W

== . 1 o ~ e

s a

22 c Q-5- D t: IEM. 22 55 4 g f Eg"* JJ JJ JJ

• U

$3 OS it y I 22 22 22 N9 52 55 55 u '~ m M, b Eo = z M 5 q-

• w n w

6s3 43 43 43 43 45 45 45 e m x=-5 g g a

• 2d
• J 3d
• J 2J "J

"J s i Y:. w

== .a ,= r S.t. - aa aa n. m'l Eit i I QI w l. 8 ~~ a C. m Rgg a I ! I ws 2 3 3 l.l. l. 3 l. .f m s i i w w ~ .e ~-. v v s s s. a a 2 2 a 5 .t. t

8 s

-E .!! !..!-1 9. .1 1 a s 2. i s 1 w w w = = 5 a Ew= 531-e. ce = e o a. e e u e e e A a e a. e s s s s e c. m-- 'A 'O 'O E h 3.,,,.A l# '.O .I 'A A s e R. '.A. '.O A a' E = = l 4 -o y-r- -.a. 7 m-- -5 --_r.g- +q y 9 -._wn-% gw -y y

m, TableB-1(continued) PROCESS PIPELINES PENETRATING PRD1ARY CONTAINMENT These notes are teyed by nuncer to corres:ono to nuncers in carentresas. 1. watn steam isolation valves re utre tnat sots solenoid pilots te eetnertized to close valves. Accunulatar att pressure plus spring set t:getner close valves.nen cota atlets are :eenergtred. Voltage fativre at only one ottat atti not cause valve closure. The valves are sat to fully close to less taan $ seconcs. 2. Containsent spray to drywell and sucoression c3amoer and RNR test if ne return t3 sugeression caamcer tsalation valves.611 nave the capaattity to Be manually reccened after aut: mat 1C closure. Thts setuo ell) ;ermit containment saray for hign crywell pressure emnatttcas and/or sucoression water enoling..nen aut:matic signals are not present, taese valves may se esenes for test or coerating convenience. 3. Testaole caect valves are cesigned for remote scentng wita zero differential sressure across tre valve seat. 'he valves stil close on reverse flow even tnougn tne test sett:::es may se positioned for coen. 'he valves esti ocen.nen suno disenarge pressure exceeds reactar pressure even tnougn tne test set tes say se positioned for close. 4 This Itne is only neeced curing matatenance. Service air supply is disconnected during giant ooeratton by acatatstrative control. AC motor ocerated valves eveutred for isolation functions are powered frem ene emergency AC power tuses. CC operated isolation valves are powered from the station tattertes. 6. All motor coerated isolation valves atti re utn in the last costtion uoon fatture of valve power. All atr-ocerates Isolation valves util close upon str fatture. F. $tgnal 5 ocens. signal c overrices to close. 8. Power coersted valve can De opened 'or closee ey remote manual suiten for acerating convenience during any mode of reac*.ar operation except onen automatic signal is present (see sete 2). 9. sorsal status position of valve (ocen or closed) is the position during normal power oceration of tne reactor. 10. The speciffed closure rates are as required for cantalrunent isolation only. 11. $cettal air testacle enect valves vita a posttive clostng feature are cesigned for renote testtag curing normal oceration to assu e amenanical oceracility of tne valve afsc. The reacte testing r feature =t11 cause only a cartial envement of tne etsc into tne flow stream. ett'l caly a sinor . ef fect on flow. Usen recetst of an isolation signal, the actuator spring force sell ettner cause a s11gnt reduction in flou =nen tse fetewater systeg is availaole or cause tne valve to close, prevteing a costtive closure differential pressure on the seated sisc. nen tne feeewater flow ' ts not availaole. 12. This valve will oces wnen tota a low reactor pressure vessel pressure and an accicent signal are present. 13. The motor coerstar of tats valve is key locked open during normal operating conattions. 14. Traverst.ng In-Core P oce (TIP) Systems When tne it7 systes cante is inserted. tne tall valve of t.*e selected tuce scens aut:matica11y to that ene proce and caole cay aavance. A maatstem of four valves may se ocened at any one time to concuct tne callaration, ane any one guide tuce is used, at most. a few nours ser year. If closure of tne line is reeutred during calibratton, as indicated by a containment isolation sfgnal, tne caele is autamatically retracted and tne tall valve closes aut:matically after c:m. pletion of caole ettsdrs=al. To ensure isolation cacastlity. If a TIP caele fails ta ettneraw or a tall valve fails to close, an exolestve. snear valve is installed in esca 1tne. t; con receipt of a remote manual signal. tais azolosive valve =111 sneer tne flP caele and seal tRe gutoe tace. 11. All unused penetrationsicesignated *! care *) are capped and seal eleed. 16. Valve elli close on system nign flow. 17. Isolat*on stqnals A or F t11 f netiate tse reactor building stancey ventilation system =nten in turn isolates tne surge air tsolation valves. 18. This valve ullt open wnen nota a low differential 3ressure across tne valve and an acctcent signal are present. 19. Pressure sensors and sensing stese line cressure are used for interlocs control to 3revent inadvertent valve ocening at htga staan Itne pressures (acove 25 astg). .--m y

w. TableB-1(continued) PROCESS PIPELINES PENETRATING PRIf1ARY CONTAINMENT Notes (Continued) 20. Control Rod Crive (CRO) Insert and Withdraw Lines: Criteria $5 concerns those finas of the reactor coolant pressure boundary penetrating the primary reactor containment. The CRD insert and withdraw lines are not part of the reac*ar coolant pressure boundary. The classification of the insert and withdraw ifnes is Quality Group 8. and therefore designed in accorcance with ASME Section !!!. Class 2. The basis to wnich the CAD lines are cesigned is coeunensurate wita the safety importance of isolating these lines. Since enese Ifnes are vital to tne scram. function, their operaallity is of uomost concern. In the oesign of this system, it has been accepted practice to omit automatic valves for isolation purposes as this introduces a possible failure mecnanism. As a means of provtdtng positive ac*uation, manual shutof f valves are used. In the event of a creak on these Ifnes, the manual valves may be closed to ensure isolation. In addition, a ball chect valve located in the insert line insf ee the CAD is designed to autcmatically seal this line in the event of a break. 21. This !C stop check valve is nor-ally in a closed positica due to its check valve feature, but its M0 is in the open position. The to provices a backup' to close the valve to provide additional hign leak tight integrity. 22. Abbreviations used in table: AO - Air Operated M3 - Motor Coerated VTC - Pneumatic Testable Check Valve RHR - Residual Heat Removal Systes RPY neactor Pressurt Vessel RC!C - Reactor Core Isolation Cooling System RWCU - Reactor Water Cleanuo HPCI - High Pressure Coolant Injection GCC - General Cesign Criterion RSC.CW - Reactor Sullding Closed Loco Cooling Water TIP - Transversing Incore Proce CR0 - Control Rod Crive M5!V Main Steam ! solation Valve ~

Table B-1 (continued) PROCESS PIPELINES PENETRATING PRIMARY CONTAINMENT ISOLATION SIGNAL NOTES $1GNAL DESC21pTICM A* Reactor vessel Icw water level 3 - (A scram will occur at this level) 5' Reacter vessel icw water level 2 - (The reactor core isolation ecoling system and the hign pressure coolant injecticn system will be initiates at tais level, and recirculation cu::es are tripced) C' High radiatico - main steam line O' Line break - main steam line (high steam flew) E* Line break - main steam line,(steam ifne tunnel nign tes:cerature) F= Mign drywell pressure G Reactor vessel lew water level 1 - (The core scray systems and tne low pressure core injection ecce of RHR systems will te tattiated at tais level) J' Line break in reactor water cleanuo systes - hign space temperature, hign differential flow hign differential temcerature Li$e break in steam line tc/frem turnine (hign steam line scace temcerature, K* high steam ficw. low steam line pressure or nign tursine exnaust diagnragm pressure) L 2eactor hilding standby ventilation system initiatica M nign radiation signal downstream of primary contatrtnent purge filter train 0 High as:cient temcerature in main steam tunnel penetradon arc; (MSTM) P' Low nain Steam line pressure at inlet to turnine (Rt:N moce only) R Low condenser vacuum T High temperature in Turcirte Sullding U 'Hign reacter vessel pressure W' Hign temperature at outlet of cleanuo system nonregenerative heat excnanger I Low steam pressure y Stancey liquid control system actuated I Low level in ASCLCW head tans RM' Remote manual switc3 frcm main control roca

• These are tne isolation functions of tne primary containment and reacter vessel isolation control systemt otner functions are given for infor:natica only.

C-1 Appendix C Identification of Pipe Sections and Discontinuities for Break Frequency Estimation Main Steam Lines All sections of the four lines in the Reactor Building are break exclusion. Two sections are considered: one to the outboard MSIV; one from the outboard MSIV. Main Feedwater Lines All sections of the two lines in the Reactor Building are break exclusion. They include check valve inboard and testable check valve outboard. Their failure rate is assumed to be similar. . High Pressure Coolant Injection (HPCI)

Reference:

FSAR and LILC0 drawings no. M10121-17 and M10122-14.

== Description:== 10 in.: one section and valve to the outboard valve (MOV-041). Break exclusion. Under normal RPV pressure conditions because inboard valve is open. 10 'in.: six nonbreak exclusion sections (4 challenges per year of 24 hrs' each are assumed in these sections): To reducer Branch SHP-171 + valve MOV-049 Reducer / valve F001 To steam _ turbine stop valve To turbine admission valve The turbine assumed to be equivalent to one section 1 in.: two bypass sections and a valve. Six sections downstream to the RCIC/HPCI drain line. Two branches. All nonbreak exclusion. Normally open.

g, C-2 i RCIC

Reference:

FSAR and LILC0 Drawings No. M10116-16 and ii10117-13

== Description:== 4 in.: open M0V inside drywell to the outboard M0V. It has a bypass line of 1 in., normally open. Break-exclusion. six sections and discontinuities: 3 in.: to 3x6 reducer to drain pot and 3x6 reducer - to steam turbine stop-valve to steam admission valve to steam turbine governing valve the turbine treated as one section. Following the turbine, low energy assumed. 1 in.: Bypass is 2 sections - Drain lines. from drain pot to RCIC/HPCI drain line are con-sidered six sections. Branches: two or more 3/4 in, branches. Quantification: 4 in.: [8.6(-11) + 1.5(-10)]

• 8760 = 2.1(-6) 3 in.:

[8.6(-9) + 1.5(-9)]

• 6
• 4 (times per year) x 24 (hrs) = 5.8(-6) 1 in.:

[6 + 6 + 2] * [8.6(-9) + 1.5(-9)]

• 8760 = 1.2(-3).

Reactor Water Cleanup System (RWCU) Supply Line

Reference:

FSAR Figures 3C-4-15A,B,C and Figure 5.5.8-1,2,3 . Des cription: 6 in.: One break exclusion section and valve 6 in.: One section nonbreak exclusion to reducer 3 in.: Two lines (having three sections each), two valves each and one pump each. 2 in.: Two lines with section and reducer / check valve. 3 in.: Two line with section, valve, section, reducer

C-3 4 in.: One section and two valves. One of these valves is normally closed. Another line with section, HX, section HX. The heat exchanger (HX) considered as one section in our approximation. Beyond the second heat exchanger, temperature is less than 125'F and not considered to be high energy, and will not result in a large environmental effect. The high energy part of the RWCU on the return line from the regenerative HX to the feedwater line is not considered a significant additional contributor, compared to the part already included. Standby Liquid Control (SLC)

Reference:

Figure 4.2.3-11 of FSAR and LILCO Drawing M10115-16

== Description:== 1-1/2 in.-line; 2 check vlaves one inside and the othe routside drywell designated F006 and F007 r.espectively. Sections : up to CV-F006 is break exclusion section; from F006 to the two normally closed explosive valves is nonbreak exclusion section. Branches: four 3/4 in.-branches from the main 1-1/2 in.-line. Quantification: [8.6(-10) + 1.5(-10)] * (8760/2)

• 3.3(-3) = 1.5(-8)

[8.6(-9) + 2 x 1.5(-9)] * (8760/2) * [3.3(-3)]2 = 1.0(-9) Control Rod Drive

Reference:

NUREG-0803 The contribution comes from the Scram Discharge Header rupture as explained in NUREG-0803. The value of the rupture frequency of 10 4 is derived from that report. w---

....- ~ C-4 Recirculation Pump Seal Injection

Reference:

FSAR

== Description:== Two 3/4 i n. -li nes ; 2 check valves one inside and the other outside drywell. Apparently, it is not break exclusion pipe. Quantification: Similar to SLC but not break exclusion -- 2.0(-7)' Sample Coolant From RPV

Reference:

FSAR

== Description:== 3/4 i n.-l i ne ; one normally open inboard air-operated globe valve. One normally open outboard air operated globe valve. Assumed to have one line, two sections, and two valves in reactor building. Nonbreak exclusion. Quantification: 2*[8.6(-9)+1.*5(-9)]*8760=1.8(-4) Reactor Post Accident Sampling System (PASS)

Reference:

FSAR

== Description:== 3/4-in. line. One manually operated globe val ve outboard, normally open. Two solenoid operated globe valves, normally closed, downstream. Quantification: same as above. TIP Drive Guide Tubes

Reference:

FSAR

== Description:== four lines of 3/8 in. The tubes are normally with nitrogen. In order to cause LOCA, all the following must occur:

s -e-- + - - ~ - - ~ ~ ~ - - + - - ~ - - ~ - ~ ' C-5 One tube rupture inside RPV Nitrogen system alarm fail to alert the operator Operator error in using the system, failing to operate the shear valve. (The TIP is assumed to be used 4 times per year.) a Quantification: 4

• 4x10 2 x 10 1 x 2.5 x 10 3
• 10 4 'x 10 6 Other 3/4-in. Lines It is estimated that there are about 20 sections of 3/4 in., test lines, and other lines branching from the systems listed in this table. Many of them are in the RWCU and are potential " liquid" break location. Other branch out of HPCI, RCIC, and other steam lines, and are potential " steam" break location.

o 4 w - ,---.n -g,----

m._._. ?7 "'I jy BROOKHAVEN NATIONAL LABORATORY ]~ ASSOCIATED UNIVERSITIES, INC. l ~.. .a Upton, Long Island, New York 11973 (516) 282s Department of Nuclear Energy FTS 666/ 2435 December 6, 1983 -. / Mr. Ed Chow Reliability and Risk Assessment Branch Division of Safety Technology Office of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission Washington, DC 20555

Dear Ed:

Enclosed please find a copy of a memorandum to I. A. Papazoglou from Y. H. Sun, E. Anavim, and K. Shiu, containing a number of questions on the Shoreham PRA. These questions have been generated in the course of our review of the PRA. This submittal satisfies the contractual requirement of Task 2 of FIN A-3740. If you have any questions please do not hesitate to contact me or the authors of the memorandum. Sincerely,ak L. I. A. Papazoglou, w. 2 Group Leader Risk Evaluation Group IAP/dm Enc. cc: A.- Busiik w/o enc. A. Thadani R. Hall w/ enc. R. Bari L 4,. - -. ...,n n

7.-.' gg.- ..,s. BROOKHAVEN NATIONAL LABORATORY MEMORANDUM DATE: December 6, 1983 To: Ioannis A. Papazoglou , E. Anavim, K.' S khW' <hh~. FROM: Y. H. ! itt u

SUBJECT:

Questions on Shoreham PRA Attached is a list of questions generated in the' course of the Shoreham PRA review to date. Additional questions may arise subsequent to the submittal of these questions and the reviewers would appreciate opportunities t~o discuss them with the authors of the PRA. YSH/EA/KS:dm i_m

7 o -D 3 . j... ~

1) Does failure of M function (SRVs open) result in large, intermediate or small-LOCA's?

2) How many valves would be involved in the P functions (SRVs reclose)? (p.

3-38)

3) What type of "LOCA" is associated with the failure to reclose more than one SRV7 What is the success criteria for a transient event given (1) 1 SORV and (2) two or more SORVs? Provide basis support by documentation of your answer.

4) In the transient and ATWS event trees, how is the repetitive opening and closing of the safety relief valves modeled?

5) The PCS system is assumed in the Shoreham study to be sufficient to remove decay heat from containment given a failure of the P function. What is the success criterion for the PCS with regard to one, two, or more than 2 SORVs? Provide a basis to support the answer.

6) On page 3-38, it is stated that "only a fraction of the eleven valves are required to open to be succes What is the basis for this statement?

How many of the eleven SRVs a,sful." re involved? Would the same number of SRVs be included in other kinds of transients?

7) On p. 3-42, "GE a'nalysis for Limerick-PRA indicates that once the MSIVs have closed, the ADS and low-pressure system injection initiation can be delayed for times in-the range of 30 minutes without fuel clad temperature exceeding 2200 F."

Is this assumption valid for Shoreham? Provide a basis to support this assumption.

8) Explain why the same failure probability (5x10-3) is being used for PCS unavailability for turbine trip and loss of feedwater (pps. 3-35 and 3-81).

Provide the basis on how the value of 5x10-3 is derived.

9) Provide the basis of arriving at a probability of 70% for the MSIV re-opened and feedwater recovered for the MSIV closure initiator, Figure 3.4-3.

10) Provide the basis of assuming 98% probability for the " Condenser Pump

-Injection" availability. Does this probability reflect the fact that the initiator of the event is' loss of condenser vacuum. (p. 3-8, Table 3.4-5)

11) As stated on p. 3-89, the severity of the resulting condenser vacuum transient is directly dependent upon the rate at which the vacuum pressure is lost.~ How is the analysis on a loss of condenser vacuum performed to '. -

ensure that the most severe scenario is included?

• 3'.'

h i.

12) How do the success criteria differ for the 110RV and 2 or more 10RV cas-

.es from those of 1 SORV and 2 or more 50RVs? In the present analysis it is stated that the 10RV acts initially as a small LOCA. Is this a valid assumption regardless of the number of IORVs involved? (p.1-30)

13) Once the reactor is shut down, the 10RV event tree is similar to the turbine trip transient event tree (p. 3-134).

If so, why does the feedwater recovery probability differ from each other in the 10RV and turbine tri~p event tree analysis? (Figures 3.4-1 and 3.4-7)

14) How does the additional heat that is discharged into the suppression pool during the initial period of a SORV affect the definition of success criteria? Would the failure probability of 5x10-3 for PCS still hold if there were more than one 10RV involved? (p. 3-134 and Figure 3.4-7)

15) What is the basis of using 8.6x10-Il per hour /section as the pipe rup-ture failure rate for the break _ exclusion pipe? What is the basis for using 1.5x10-10 per hour as the valve external leak / rupture rate for

~ break exclusion valves? (P.A-24)

16) LOCA frequencies listed in Table A.1-7 (initiating frequency for a large LOCA in main steam lines within the reactor building) cannot be re-produced from the data suppl'ied in Section A.1.3.3.

Clarify the evalua-tion process. (p. A-27).

17) LOCA frequencies (2.1E-6) listed in Table A.1-8 for HPCI u'nisolated pipe cannot be reproduced from the supplied data. Clarify the calculation.

(p. A-28).

18) Clarify the reference leg design of the reactor water level system.

It is indicated that there are 51 instruments connected to the reference legs at the Shoreham plant. Clarify potential common mode failure of instruments due to failures at reference legs. (p. A-40)

19) Provide a discussion on the failurh of the reactor water level system re-ference leg that will cause the level instruments to indicate a low level.

(p. A-40) 20) It is stated in the Shoreham PRA that the surveillance of the reactor water level instruments at most operating BWRs is performed each time the plant is manually shut down (4.3/ reactor year). At Shoreham plant, it will be performed only at refueling outage (1/ reactor year). It was thus assumed (in the PRA) that the initiator frequency of the reference leg failure at the Shoreham plant is 4.3 times lower than most of the other BWRs, because the maintenance error at Shoreham is reduced due to less frequent surveillance actions. However, the Shoreham PRA didn't clarify why a less frequent surveillance is enough for detecting potential instrument failures and so to optimize the reliability of the reference leg instruments. (pps. A-47 and A-48).

c ?.' ?_ _. _ _ _. p.

21) The initiator frequency of the high drywell temperature was determined as 0.0093/ reactor year. This was determined from LERs of the period from 1971 to 1981. Only two events were found in this period of time. How does this initiator frequency compare with the failure rate (hardware failure and other potential common mode failure) of the drywell cooler systems?- (p. A-51).

22) The frequency of a single DC bus has been estimated as 3x10-3 per bus year or 6x10-3 per reactor year. However, based on NUREG-0666 the operators' failure to recover from an unavailable DC bus has been as-signed a conditional probability of 0.5.

That is, the frequency of 3x10-3 per reactor year has been assumed for Shoreham PRA for single bus failure (p. A-34)..Is the 30 minutes recovery assumption consistent with conditional probability of 0.5 used in deriving the loss of single DC bus initiator frequency?

23) A discussion is presented on pages A-34 and A-35 about the initiator fre-quency of multiple DC bus failures. NUREG-0666 reported the failure of be6.0x10gsiontobe6.0x10-3 Provide details on how the value 2.0x10-4 per one DC div and the failure of two DC divisions to for conditional probability of losing the second division was derived.

24) Table 3.2.3 (p. 3-315) given* a failure frequency of 2.5x10-3fe,y (1/1400) for the service water system. A service water frequency of 1/215 or 4.6x10-3/r.y was used, however, for sequences following a loss of service water initiator, (Figure 3.4-56, p. 3-316). This inconsist-ancy appears to originate from assuming the total BWR reactor-years of operation as 400 years (p. A-52) rather than 215 years (p. A-45, Table A.1-16).

Clarify this difference.

25) What is the basis for a scram failure of 10-5 under large and medium LOCA? It is stated that the common mode failures of all control rods to insert were considered. However, it is not clear which common mode failures were considered and how the common mode failures were modeled in the quantitative evaluation.

26) What is the basis for assuming that the failure of the vapor suppression mechanism is dominated by the drywell floor seal and other penetration seal failures, instead of failures such as downcomer vent pipe failures?

Why is this assumption, which was used in WASH-1400 and the Limerick PRA, ~ applicable _to the Shoreham case? Provide a discussion on other mechanisms, such as the vibration failure (environmental stress) of downcomer vent pipe under failed SRVs or large LOCA conditions, that were considered in the PRA? (p. 3-146).

27) The heat removal capability of the PCS was not. considered in the large e

LOCA event tree, and the reason was stated to be the prevention of the release of fission products inside the primary system. Why was this con-sideration not applied to the medium and the small LOCA cases?

,_~.. t o' Z f. C3 2_L7 z d;

28) The "LOCA outside containment" event includes only large LOCA but not medium and small LOCAs. While it might be generally true'that the con-sequences of medium or small LOCAs could be considerably less, the poten-tial for system interactions resulting from medium or small LOCA owing to spray in the vicinity of vital equipments in the reactor building seems to have been neglected. Please comment.

(p. 3-156, 3-231).

29) The potential water sources (p. 3-232, Table 3.4-21) for excessive water at elevation 8 should include reactor coolant water that is released to elevation 8 during a LOCA.

This event has not been included in Section 3.4.2.4 (Loss of Coolant Accident outside Containment) because, in that section, ECCS are considered to be available at the beginning of the ev-ent. 30) It is assumed that most pumps, turbines, and electrical panels are dis-abled if water level is high enough to contact electrical features on the equipment. However, equipment malfunctions may occur under water spray or high moisture conditions, which may happen before 'the water level is high enough to submerge the electrical components. What judgement has been made for conditions other than high water level situation? (p. 3-231).

31) It is assumed that the poteritial initiators for the internal flooding should have a water leakage rate of greater than the capacity of sump pumps (640 gpm). What assumption was made concerning the sump pump failure, and the loss of AC power or DC control power to sump pumps? (p.

G-4),

32) It is stated that 3'-10" is the critical height for_ the core vulnerable analyses under internal flood condition (pps. 3-237, G-5).

However, Table G-3.2 shows that many ECCS control components are arranged at levels lower than 3'-10", such as HPCI vacuum pump, HPCI condensate pump, RCIC instrument racks, HPIC instrument racks. How is the failure of the-se components with water level below 3'-10" been considered in the an-alysis?

33) Frequencies of the internal flooding (initiated by pipe failures) listed in Table 3.41-23 and Table G.4-5 are not consistent. Please clarify which frequencies were used in the analysis.

34) Please clarify why the manual error of opening the isolation valves dur-ing major maintenance action is judged to be insignificant.

The opening of the isolation valves during maintenance action will result in internal flooding' in the reactor building (p. 3-246).

35) What is the basis for assuming a negligible probability of S/R valves v

failing to open under the condition of loss of one division 125V DC bus? (p. 3-277). How does the loss of one DC division affect the success of the ADS function and what manual actions are required if the ADS is not sufficient to depressurize? L_

fo -a 9- .l-36) It is stated that the loss of one DC bus initiator has been considered as the bounding calculation. Please clarify the reason why loss of multiple DC bus incidents (for example, loss of two DC bus divisions) will lead to a lower conditional probability for the vulnerable core. (p. 3-278),

37) Based on the Shoreham Emergency procedures, the primary backup for the s

loss of reactor building service water is the turbine building service water. Although the motive forces for the two service water systems are separated, the two systems use the same water source. As indicated in Shoreham PRA, the operation experience with the loss of reactor buildup service water is mainly originated from problems of water intakes (con-tamination of intakes by sand, ice, and/or clam growth, or by the loss of instrument air). There is, therefore, a significant possibility that the turbine building service water may be in the same trouble when the re-actor building service water is in trouble. Was this common mode failure considered in the analysis? (p. 3-315). / 38) With the loss of reactor building service water or the loss of AC power, the room coolers would be unavailable and the RBNVS and RBSVS would be isolated. Please clarify the effect of these events and indicate the possibility that the reactor building temperature would rise to the point that HPCI and RCIC will be igolated from operation. (p. 3-314).

39) On p.1-29 in the discussion of success criteria for large LOCA, it is stated that "at some locations, breaks toward the lower end of the size range may require operation of one or two safety relief valves", provide the rationale on why these LOCAs are not included in the medium LOCA ev-ent.

40) It is stated on p.3-32 that "...the turbine trip event is that closure of the turbine stop valve from high power will drastically reduce steam flow from the reactor vessel while the feedwater flow will continue. These two functions result in a water level swell in the reactor vessel to Level 8...and may trip the feedwater pumps on Level 8 trip set point". If analysis indicates that there would be a Level 8 trip given a turbine trip event, provide the basis why all turbine trip events are not treated as loss of feedwater events. 41) In the turbine trip event tree, Figure 3.4-1, two feedwater related functions are identified. Provide more detailed discussion on these two feedwater functions and how their conditional probabilities are derived. p 42) In the' support system screening event trees, the transient induced loss of offsite power function is assigned a value of 3.1x10 4 Provide the the WASH-1400 value of 1.0x10 g was selected for this analysis instead of basis on how and why this valu 6 et - * -t y e 1r re-gy..t e -e v. s-,- y e +,,e--w.-,s.ie--w we v. -- w-g -w m-v-w---- -e--w-- - - = - - -

. ~. - -. - o' y i 7 .'..p.

43) The Shoreham PRA states that the HPCI system is completely independent of AC power. However, within the HPCI system, the loop level pump and the k d'i inboard isolation valve (F002) are both powered from AC 480V bus.

It is not clear whether the unavailability of these two components, which are vital to HPCI operation, as a result of the loss of the corresponding AC 480V bus has been considered in the analysis. (p.B-1).

44) The steam supply to the HPCI pump turbine could be. diversed from the steam supply line through the opening of valves F028 (Figure B.1-1,'.

Clarify the functions of these valves and the importance of loss of steam supply to HPCI pump turbine due to the opening of these valves. (p. B-2).

45) Loss of HPCI flow will occur when the following two automatic operations fail :

reclose of the mini flow line af ter HPCI pump reaches the rated speed. transfer of the water source from CST to suppression pool after CST is depleted. Clarify the importance (to the loss of HPCI flow) of the relay logic failures which f ail the two automatic operations stated above.

46) The steam supply to the RCIC pump turbine may be lost due to the opening of valves'F025 and F026 (see Figure B.1-2).

Clarify the function of the-se valves and the consequences when those valves are at open position. 47) It is stated in Shoreham PRA that fault trees of the diesel generator system and the scram system were not used for the quantification of these systems, (p. 1, Vol. IV). How are the unavailabilities of these systems calculated? What is the purpose for the development of these system fault trees contained in Volume 47

48) Clarify the possible failures of diesel generators and off-site power line due to synchronization error during loaded testing of diesel generators.

Since the automatic synchronization mechanism is not instal-led at Shoreham plant, the synchronization error may occur because of hu-man error or because of synchronization check rela ~y malfunction (p. B-110).

49) Clarify the importance of the maintenance unavailability and/or the com-mon mode failure between diesel generator fuel oil transfer pumps (to supply fuel from storage tank to day tanks). These failure modes are not included in the fault tree for the diesel generator systems.

1 l V l

J'},W 3 /a t

50) If the reactor. pressure is above 403 psig, the.LPCI injection valves are closed and the LPCI flow is bypassed to the suppression pool via the minimum flow line. Clarify potential RHR pump failures owing to inadver-tent closing of the minimum flow line, which will cause RHR pump over-heating when LPCI injection valves are closed.

(p.B-38). This question is also applicable to the closing of the minimum flow line in the core spray systems. V 51) On p. 3-174, the Shoreham PRA stated that in the event of an ATWS, the high pressure systems are used to mitigate the accident and depres- .surization of the reactor vessel is to be avoided. This decision, ac-cording to the PRA, reduces the potential for problems with controlling water level during and after depressurization. Discuss in more detail the basis 'why depressurization and low pressure systems are not con-sidered for ATWS mitigation. Provide any thermal hydraulic analysis that was performed to justify the problems associated with water level con-trol.

52) On p. 3-178, the PRA discussed the requir'ments for the operator to properly control high pressure injection given a turbine trip ATWS.

Re- /, actor water level is to be maintained above Level I with the HPCI or RCIC A" 4 " " system. _ How is this requirement reconciles with the emergency procedure M guide that states that during an ATWS, the RPV water level is to be P ; {* maintained at TAF?

53) In the discussion of HPCI reliability, p. 3-197, the Shoreham PRA presented that the HPCI can remain functional in an ATWS event with high suppression pool temperature for 40-50 minutes. At that time, hot shut-down condition is achieved and the RCIC will be adequate for coolant injection. Why is the RCIC lube oil system not subjected to the effects.

of high suppression pool temperature just as the HPCI? If they are similar, what is the basis for assuming that the RCIC is available sub-sequent to the HPCI failure due to high suppression pool temperature.

54) Provide.the basis for assuming that at suppression pool temperature below 240 F, the added unreliability of HPCI owing to lube oil cooling is small and that the 50% increase in HPCI unreliability is conservative, p.

3-197.

55) According to the information provided in the Emergency Procedure Guide (SP#29.024.01, Rev. 0), no action is required of the operator to inhibit ML-ADS prior to termination of all injection except CRD, RCIC and HPCI and 3

maintenance water level at TAF. Provide clarification on why this is not included in the procedures and on any other detailed procedures that the operator has to follow to inhibit ADS.

56) On p. 3-198, it is stated that for some sequences, continued high pres-sure system operation could actually increase public risk since it would delay core melting until after containment breach. Such a statement implies that termination of high pressure system operation could indeed be beneficial to the reduction of risk for certain ATWS sequences.

Provide the rationale and justification to support such a statement. , e M ?.,n' L;if t ,N i ,~ 3

-a, 4 d Fo Pun a n o r n c ( $t An Evaluation of Unisolated LOCA Outside Drywell in the Shoreham Nuclear Power Station D. Ilberg N. Hanan Draft (for comments only) ~ Risk Evaluation Group Department of Nuclear Energy Brookhaven National Laboratory Upton, New York' 11973 February 1985 Prepared for U.S. Nuclear Regulatory Commission Washington, D.C. 20555 w+ F; m

s 1 1-1 1. Introduction

1.1 Background

fM The SNPS-PRA(1) considered LOCA outside Drywell (LOCA in the Reactor g Building) in two ways: a) Interfacing System LOCAs: Appendix F of the SNPS-PRA' estimates the initi-ator frequen y and the core damage frequency for this case.N BNL review (2) of gShoreham PRA re-evaluated the initiator (r*1tAtsty ras well as the core damage frequency and found an increase M about an order of magnitude of the core damage fraquency. This result is included in the present study; ant! for more details see Appendix C of Ref. 2. b) High energy line breaks inside Reactor Building: The SNPS-PRA included in its analysis only pipes larger than 6" 'in diameter on the premise that ample time is available to isolate breaks in smaller lines before they cause adveje, ginment conditions if not automatically isolated. The frequency of,line breaks downstream of the outboard isolation valve was calculated to be relatively small. The BNL review of this part-agreed with the SNPS-PRA as discussed in Appendix C.of Ref. 2. In thg SpPS-PRA and the g g g,,a,lgthe isolation valves were assumed to perat under O / p breakkflow conditions. It is shown in Ref. g that interfacing system LOCAs are the major contributor to LOCAs outside drywell. z 1.2 Objectives This study is a special consideration of case (b) above stemming from the assumption of the failure of the co? responding isolation valves in the case of a line break outside drywell. NRC requested BNL to re-evaluate the core damage frequency from high energy line-breaks inside the Reactor Building (same as case (b) above) under the assumption that most of the isolation valves qualified to close under break-flow condition %, i.e., as5unL ng the are not s i e s

c. v e <

w .-e failure of most of the isolation valves. Under this assumption,gthe rupture of any. pipe (regardless of diameter) opening a path that leads from the Reactor Pressure Vessel (RPV) to the Reactor Building, n:-"

• L n m 1*ed for potential adverse environment or flood effectsg regardlett cf hcw =11-

.t he pr- @ h414tv of nino etintirro nr 4 e 0] a t,

• ga 44fineo ie m

v

i i 1-2 This assumption obviously increases the contribution of the high energy line breaks to core damage frequency and requires consideration of other lines { connected to the PPV of diameter < 6". These iine>..ey Lewme sisnificoni. t :::ta i+ 4 act: -is e umed t.h et one ei ;r.cre 15eletien vel,e5 will f 11 te i Olete 07. i.iie..d 0"00 th0 bre;k hai cCCuii66. This study considers the following questions: ~ (a) What would be the increase in core damage frequency due to the assumption. stated before, i.e., the failure of th6 isolation valvef to perform their function? Pts 20:emptic.; rcquire: the *;;1u:10n cf ether 'ine aat -cen:idered in the ^M S Doa ar ia the ML.criew.- (b) What would be the contribution to core damage frequency from each pipe connecting the RPV and the Reactor Building? (c) What would be the more important isolation valves to mitigate the outside drywell LOCAs? c.[neube rr s Ica-(d) What is the c-der Of J.;wi;t.u a cf time available for operator action? 1.3 Scope The scope of the BNL study was defined to cover the following: (a) To identify any significant(*) high energy lines leading from the RPV to the Reactor Building with a potential for affecting safety systems, if w ere. an unisolated break would be postulated. (b) To estimate the change in SNPS core damage frequency relative to the SNPS-PRA(1) and BNL review (2) due to the following assumptions on the oparateow ava":bility-of isolation va.1 ves following the occurrence of a line break: (*)The contribution from downstream moderate energy lin s of a system was neglected if it was smaller than the contribution of the lines upstream. s e 4-+--,r .~-,...-....,,.v

_s i 1-3 ~ (1) The. Main Steam Isolation Valves (Inboard and Outboard) on all four main steam lines will isolate in. all the cases considered, having the failure rates shown in Table 2 (discussed in Appendix A). (2) All check valves will close on reverse flow as designed with the failure rates shown in Table 2 (discussed in Appendix A). faL k (3) All other isolation valves will act, close when receiving their signal to close. No partial closure is assumed for these valves. (4) Manual valves are assumed to be available for isolation if access-ible by the operator. (5) Remote operated valves that do not receive automatic closure signals' upon sensing break conditions are identified. However, no credit is given k them in this study. (c) To provide the list of the more important is:...-lon valves from the standpoint of reducing the core damage frequency. (d) To provide some crude insights on the time available for the operator to respond to such accidents. 1.4 General Description of the Problem Evaluated The Shoreham Reactor Building surrounds the MARX II containment structure (the drywell). At its lowest elevation (referred to here as Eflevation 8), the budding is an open cylindrical compartment, i.e., there are no barriers in the Elevation 8 compartment. Mr.;ria This open area p~ resents the possibility that excessive water released into the compartment may adversely affect the ECCS equipment in Elevation 8. The SNPS Reactor Building has opening 5 between its floors, and a line break at a high elevation will affect the entire reactor building (see section 3.1 for more 'detailg). Fig.1 provideja general description of SNPS Reactor Building Elevat[ons. t f M i 1._ _. m o n.__ a = n. _,.. - - -

's 1 c v OA Qo' 9 tasno Tom PoLan enant Pf. - T y Q.a,6 ja.;../.: L-caa ~** / b ELL SMSF, g g. l j CUme - oFCONTaemesaTeoff atta n gyya.,* .r i! M, j,'!(ls,T - 8

I

f..*"

ia dl N co?st?.E." E r m.o,, at L 'c&tm.ss , f \\.. ~.. wa rx.in c1 g': . I.. g#;gi bc gy@ + i i {. ,,w.. ourt. etuows stat g w a SMaE>eY Lloue IL C NTAob STS. o h Vt ML sMfEST LML CoNTRO 9.+- setnaus v,, r e----, =- .v d-e ~ I activt ; 7 :9 l rt v4W f.; ,futt zow: sans man' ,j } },,; 'a s +a, .~ um ,\\ z 7* msg j_

.{g ransonut 2

sh cou r.ot 6 ca o \\ G '/,. "00' $sD+r I meraam E. touipu t nT I 1 Awioumik *La7E I naase strau Tumwt etatination ' V.,

4R;

=a gr.t. o.. a-cco w s uns r s p j P n ..g'cen 4 -eno waster N ': # ,*r.y e e m.,.,,

g., g -,

.5o, . _~ ' S t.J i

• i n..

r. ;i,

E - caort rea necs 2*P '

• q

,r ?J 4 ;p ' .t */ tmetos to met ,cets:{ 5 LJ f..,; i f, j: p srcamunt LoCE J -Inst mace 5 i 4 4 [ e.o =7 / J ,tavr - oo = it.es.t, t.o.o,, I b .o-t / u. .......t IP45* t ost l ccyte .m-mtre y273-. e ' M i q="& AP# w L _'I L.*.' '. a -o-ng~:. t..:: r .. a 4

1. u'-

• ?d.1 Fig. 1 General Description of SNPS Reactor Building.

r=:r ',',',',",u','.,,,, Elevations (with Emphasis on HPCI Steam Line Routing) j Fron: Shoreham Nuclear Power Station - Unit 1 Final Safety Analysis Report REVtsio*4 te. DECEUBER 1978 ? i

I i 1-5 Figs'. 2 and 3 show lines that connect the RPV to the Reactor Building and provide a potential path f om the RPV to the volume of the Reactor Building Mi> fh Mi ef< r = E. u# o[ a break with a failure of ig isolation valves to close.would bc posi.uioi.;d.- These figures do not show all isolation valves, but only those that are designated as containment isolation valves. In some cases, the most important being the RWCU, more valves are available to the operator to manually initiate a line isolation from the control room;lAn, y A pr 4.re ut sk.a a tw Fgs. 2,*A3 A list of the lines emerging from the RPV and some additional information associated with these lines (size, type of isolation valves, and process or y standby line) is given in Table 8.1 of Appendix B (reproduced from b SNPS-FSAR(3)},

• -o 9

9 e W e r-

1 e I T su ne ces.cas MSly. LEAKAGE CONTROL SYSTEM - (TYPICAL ARRANGEMENT FOR p ""e cce. ere 4 OUT80ARD MSIV'S) Xil RHR-HE AD SPR AY LINE TO RPV 5 J; f A 2l XID MAIN STEAM ,j najus 8.1L d" [. h 8 MAIN STE AM LINE DRAIN l d cci cce g g' M : I X2A FEE 0 WATER LINE X!C MAIN STEAM "L.1** 1 .E N AIN STEAM LINE DR AIN b X2B FEEDWATER LINE b M : XIB M AIN STE AM S ] MAIN STEAM LINE DRAIN

X2!A COIE SPR AY PUMPDISCHARGE TORPV ece ai

• il '"

' Z cce cce . X20B CORE SPR AY PUMP DISCHARGE TORPV E ) XtA MAIN STEAM E' "' ..[ cee, ecs r MAIN STEAM LINE DRAIN SUPPRESSION Q M - X36 STANDBY LIQUID COOLANT TORPVf [- ('} 9 ' VESSEL .I PJ X12 HPCI TURBINE STEAM X3 MAIN STE AM LINE DRAIN T ~ REACTOR hT ~ ec gy,, s cce -cce INLET LINE T-X33 SAMPLE COOLANT FROM RPV N? o so na X16 RCIC TUR8INE STE AM "W ip'. } c[ge, INLET LINE RECIRCULATION SYSTE M RETURN LINE ~ M l 9. '~ .]1 M X69 RHR INJECTION LINE TO X6 A F,HR INJECTION LINE TO I r L 6 " RECIRCULATION SYSTEM RETURN Lif cci cce X4 RWCU LINE FROM RPV E E .r M ~ c'.' XS RHR SHUTDOWN eg ry ce, COOLING LINE FROM RPV Fil RECIRCUL ATION PUMP ~' fr ,C _ FIO RECIRCULATION PUMP .I ~ SE AL INJECTION ~ s RC AT SEAL INJECTION ?':81*, Ji. e W PUMPS C2 PASS REACTOR SAMPLE M i l Y W CRD INSERT ANDWITHORAW LINES JET PUMP FLOW INSTRUMENT LINE c :: LI f P(YYP. FORIST UN8TS) DRYWELL WATER SUPPRES$f0N CHAMBER LEVELN AIR SPACE yyyy lill-SUPPRESSION POOL _~ Fig. 2 Lines from Reactor Pressure Vessel to Reactor Building.

v o i l: ~ L 1 i l E a REACTOR PRESSURE VESSEL i X22 A-RBCLCW-TO RECIRCULATION PUMP 3 X 228 -RBCLCW-TO RECIRCULATION PUMP AND MOTOR COOLERS .L 8 AND MOTOR COOLERS ces s ces X23 A-RBCLCW-FROM RECtRCULATION PUMP Q f I AND MOTOR COOLERS X238-R8CLCW-FROM RECtRCULATION PUMP AND MOTOR COOLERS REACTOR X38 NITROGEN / AIR PURGE FOR TIP J: RECIRCULATION M PUMPS X37A TIP DRISE GutDE TUBES ^ h b b b X378 TIP DRIVE GUIDE TU8ES b X37C TIP ORIVE GUIDE TU8ES X 80 M370.TIP DRIVE GUIDE TUBES DRYWELL ~ q X4S-DRYWELL FLOOR SEAL PRES $URIZATION f E " SUPPRESSION CHAMBER SPACE E X44-DRYWELL FLOOR SE AL PRESSURIZATION WATER LEY SUPPRESSION POOL Fig. 3 Tip Drive Guide Tubes Connections to Reactor Pressure Vessel ' v. r.n.:n. G:-.. ~ _;g I= g e e

) ~ i s 2-1 ^ 2. Evaluation of Pipe Break Frequencies l This section covers the evaluation of' the frequencies of high and mod-erate ' energy pipe breaks excluding interfacing LOCAs. The interfacing LOCAS are addressed in Appendix C of Ref. 2 and the results are included in Tables 2 and 3. 3 ~ The pipes considered in the BNL study are listed in Appendix B. All i lines which are associated with General Design Criterion (GDC) 55 are analyzed. in the BNL study. In addition, the Transversing Incore Probe (TIP) Drive Guide Tubes (GDC-57) are considered. All other lines of Table B-1 referred to as GDC-56 or 57 are not connected to the RPV; they are mainly connected to the. Suppression Pool (the routing was rechecked). The SNPS-FSAR(3) was the main source for determining the number of sec-2 I tions and valves or other discontinuities on each line. The isometric draw-ings of pipe routing in the Reactor Building from Appendix 3C were used. They were compared with the system-specific drawings given in the other FSAR chap-2 ters. The summary of this task is presented in Appendix C to this report. The evaluation of pipe break frequencies.fas made with the failure and unavailability data summarized in Table 1. The bases for the values shown in this table are further discussed in Appendix A. The failure and unavailabil-ity data were used with the number of sections and the. rie Of valves or discontinuities identified for each line, to compute the frequency of line breaks. The summary of this task is presented in Table 2. An example of this computation is shown in Appendix C. i The results of Table 2 were, next grouped into seven different cases; W _"r (a) Large Interfacing LOCAs (Liquid discharge through break) i (b)- Large LOCAs outside Drywell: (1) steam and (2) liquid discharge (c) Medium LOCAs outside Drywell: (1) steam and (2) liquid discharge (d) Small LOCAs outside Drywell: (1) steam and (2) liquid discharge. 9 L .,-r---. ,.-ow..e,g=... -,a,o e ,ew,- .---.s,e., .,-,,,,.--,,--y,m,.r, ..,,-,,,.e r- ~-.--,,4-

s + & he l: $ u w msty Oh I "'

• S
  • Y"'

DA

n PcP'3 6' V' M 3 f

MP i Fache F ac tur-Aah (A) b Wf 0hth g, gg (3reak

y. m - Bre.k j

Exc (.u tres Ge tus.,u l-P pa 5 >3 stu p.ture. g,gme,-"/Ay 8 6 *10- //iv (Persecfrh) Pepes 4 3" Itu.p tart

8. 6

• io'"/A v F6*frYl*

(Perle (f.ow) i Stv tr 8-33"/y l d. A v t..e T**=r d L <. f m s Auptuye

l. 5=to- [4 r I F *le-9/Ar t

i A.fa oPeJA F t r* E g,,, _3/j O 'rA v.1 o ( Nev) P (w /g-AJ FuW k 6

• lo
• YA op<rste (WIo g.J)

Tuanov,

,,,, sg (CHF) ~Ro pb ure Jgwto*/Ar 1 Ex fo/Ar l

i d 9 e 4 ._. E TABLE -2 Estimated Frequencies of Breaks Outside Containment NUMBER OF: INITIAL BREAK h3 of ISOLATION,, VALVES BREAK ESTIMATED DESCRIPTION OF THE CASE ANALYZED BREAK LOCATION SIZE ( E A ASSUMED FLOW: FREQUENCY pr c VALVES FAILURE STEAM OF BREAK p I CASE 6 y DESIGNATORS PROBABILITY OR LIQUID DCCURRENCE j e Y Emp Main Steam IB21-A0V081 Break exclusion section and valve between Reactor Line I 24*' 4 1 1 Inboard MSIV 6.0E-3 steam 5.0E-8 Building penetration and the outboard HSIV. (Elevation 78). II. 24" 4 1 0 Inboard and 2.0E-3 steam 6.0E-9 Break exclusion section from outboard MSIV up to Outboard the Jet-Impingement Barrier. (Elevation 78). MSIV 1821-A0V082 Phin Feed-Break exclusion section and testable check valve water Line I 18" 2 1 1 Check Valve between reactor building penetration and the F002 A/B 3.3E-3 steam 1.4E-8 testable checkvalve. (Elevation 78). Testable C.V. Break exclusion sections and IB21-NOV035A/B from 11 18" 2 3 1 IB21-A0V036 [3.3E-3]a steam 7.8E-11 testable check valve up to the Jet-Impingement A/B and C.V. Barrier (Elevation 78). F002 A/B High Pressure Break exclusion section and valve between Reactor Coolant Injec-1 10" 1 1 1 IEC1-H0V041 1.0 steam 2.1E-6 Building penetration and the outboard isolation tion (HPCI) valve IE41-MOV042. (Elevation 66). Steam Line II 10" 1 6 6 IE41-H0V041 1.0 steam 1.4E-6 Non break exclusion sections and valves from and IE41-outboard isolation valve up to HPCI turbine. Four H0V042 openings (24 hrs each) per' year of valve MOV-042 are assumed. (Elevation 66 down to elevation 17). III l' 1 17 17 IE41-MOV048 1.0 steam 1.0E-3 Non Break exclusion sections and valves from and Reactor Building penetrations up to the 1-1/2" IE41-MOV047 HPCl/RCIC drain line to condenser. Normally open path. (Elevation 66 down to elevation 11). This includes all discontinuities, i.e.: valves, pumps, reducers and heat exchangers (see Appendix A). .-e-t y 9

,r 1 r ) - l 3" o t l8c ce ) t url a/ en k2 x p .s c. ouAneI nn a1 6. m a a2 o(ORC nu e1 ) u e4 mF i Pn ot r e2 p R0 o dt m c bn h1 l r. m v r "H e D eao p m o t1 U i E nV f o sa si C Z eO euef2 ee i t on W Y eM snsl /y nt a to R L w eises1l A td vba el l I e et o is t v i N er l r ov1 a n l va t A ba auetl m. ni E av o vtr aer) ia .( l e p E eb anvho8 am g vl u S vt dC w tN r n E A l u nI2od n de i. n( r ,t C ao aC4dno.io h. dg o e v R0 at Ct) l n in c E e s 7 rt I 0 H dh no 8npea Co7 ut ai d ii to u T nt ot e ousv Rt Ba l s e F d t pl otnel dnn re sl a j vni ne d r ou r O nn cuaicodE noa op ic m oa e vt ein ai t o x o N i se ast oo t1 c de r O tn vfv act I a1 al r f I co nl oenr Cr ea ak T ei oa l ot on Pt n R m oa s P st ivreietw Heo r be e I a.s a sn o ni no tr v R nr unewueed net in ub l) C ot)l oyolPn epa o a2 S i e7ci l c i7 e v nd n v1 d E sn8xt r e x gl 8 wee on mo 1 D ue eaeben thl ia oN d t l pn l p inn et e t r nn n c oko nkdf o b cn f ao o C x gi a s s o a l ai dn eo i e n t e i gi e i r t nne si nr nt iar nt ruda oae s oe oa kdvbdiabB v i w tu ic i v al e rnv Ce tnt sl t u te t eil naeenrIl coe rc cd cl n r u E o o pl ooCE eiB ix ee eE em BB(NboeNtR( St( Fe Sr S( n i E a DY C t ECKN 6 6 3 6 4 n TNAE 5 6 o AEER E' E E E E C MURR IQBU 1 8 2 E. 1 E 4 e TE C 2 5 1 5 2 5 d SRFC 5 i EFOO 7 s tu O O I s L U k A Q a IK:MI m m m m i i d d e TAWAL a a a a u u r IEOE e e e e q q B NRLTR t t t t i i IBFSO s s s s L L fo s V 6' e T i I c L n DE1 'e ER8 u SMUA q EUL8 0 0 0 0 0 0 e V SI0 r L SAR 1 1 1 1 F A AFP - 1 1 V de t ) d e a S 1 1 8 7 m N R 4 42 d 4 r 0000 3 b 1 n v 0206a4 o i O O 0 04-0 0 1 o F111 eo a t I T V V0 V V 4 (FFFvV)

t.

E A SN M MO MnM I88 3OO bM0a s T A O OV OdO E12 VV oO4s L EG 0M - a- /00 3MM a0 O VI 1 1 1 1 1 - - 0 3Fe S LS 5 5d 5 5 5VV Vdd e3(m 2 I AE E En E E E00 Onn hGra VD 1 1 a I I IAA Maa TI os EL w 4 0 3 B F # A$6s ( 1 6 1 0 1 A O T R 4 5C$$as E 1 6 1 1 1 1 3 B M UN, FEs 1 1 1 1 1 1 2 I: K AE 2 EZ 3" "6 / RI 4 1 1 6 3 BS 1 E S I A I I I C I I I I I 1 I 1 I ge nn N ii m O lL a m - I o e rey T eom t etl A rCa S t sp C o e e ayp O Cnt I n WS u L oS Ci S ri PL rp ot) H ou) K t aC /n t nU A cl E aoCI Ci caCe I a aeWn R esR Cr elRi B RI( RD RC(L <'l .I' 1 ) 4 j i

N ~ e3

== u -o 28 ~* I* I$ E -= e aa ,2:*. Es e-S "WJ 8,- 2 8 "t 3-. .ta ~ gwm u.2 8 8 8 t. & 2".2 2.- W .e u E 3 3 32g tS4"~ .3 EG - - = ~ 4 4 4.R

• 85 t*8

?x

24" w-5 e

ee 8.= = ;3; s 4 >~. .. =. n ,N. E 444 ~. 438 "1422.. 8E .2 m n u e u ww-

• ~

N o n 8 5:55 8g" EFEt. 7 8" 8,. 888 6 6- .uuw -o ~ 5

• E ' 8 - 8. 8' ;E12 Y A a 8

88

0. 8 %.

84 ; 8 3 v.> t A228~;

• tf "t*

~ ~52 ;284 eu.?S: 8 3- "85 ~ 8 M e->eE-8m. 8 w .N E i~15 145~

• *;[8 5 s*E 1"J E

"8 .&.8 t*

EE ;88 c.e.u.8.u..

-.a u - - u w owu u .-E. N W ; d" s t. =" : EtogIt" %.58 184 ovu u u-u .e -A*At AE85 4282:2 A"E A"E I ~ =>WWW$ i T i 7 g = n w-ww w w w w m -8.5 w g >w v -n -M B 8 -5 o W tw3w"g.g- = g e e 2 2 8 =" w"~5 ~. m 3 O i C C ~ ue ow-a 8 Es8 52 = o E w o t dQ52 ~ M 3 ! E 2 r %.6. .8 E 5" + -.- --. 2

2. _.

R EN G~ ,E *" 4 E8 o - on e >o .32 u o. .c. o-wo A..S..- w3 E o 64 T.u o W 4 M 4 a w . C.. E. e a up a Q p-gg w w m gam M aM .am a EE-N NV N OWN >m hN M S+$ N N. E-NS- 'm M N N m e g M N W M m N N N M - m w E kw w~ m E-N M T M M. N - N I .M 4 u . O.@ D = u 6 m x 0 I a, ao.oea 4 w.a O 4.cm3 $MA .d>O)E oa o9-M~ M *G o * ~ ~ w .e s. e un U C= ..e x u 4 u MDW u .E go go.uzuo m oMOO =v-a 9 1 ,~. , - + ~.,., -, -,, - - -, -, v,,,,,,- r-

i Ni: i @.g TABLE -2 Estimated Frequencies of Breaks Outside Containment Cont'd. NUMBER OF: INITIAL BREAK k 5 v ISOLATION VALVES BREAK ESTIMATED DESCRIPTION OF THE CASE ANALYZED - BREAK LOCATION SIZE / 8 A ASSUMED FLOW: FREQUENCY I k VALVES FAILURE STEAN OF BREAK { f CASE 5 DESIGNATORS PROBASILITY OR LIQUID OCCURRENCE 5 (d Interfacing LOCA: - RHR Shutdown I 20" 1 2 Liquid Cooling - RHR Head Spray 11 4" 1 2 All four interfacing LOCA cases estimated on the Line Liquid basis of 0.02 for testable check valve unavall - 2.0E-6 ability times 10-3 for spurious MOV opening and - RHR/LPCI Injec. III 24" 2 2 Line to Recirc. Liquid 0.1 for probability of low pressure pfptng to fall before isolation. See detail in reference 2 Lines IV 10" 2 2 - LPCS Injection Liquid (Elevation - 8 up to elevation - 87). LO-F008 and Standby Liquid I 1-1/2 1 1 1 Inboard C.V. 3.3E-3 Liquid 1.5E-8 Break exclusion section of the SLC (Elevation 112) Control (SLC) II 1-1/2 1 1 1 F007 [3.3E-332 Liquid 5.0E-10 Non break exclusion section of SLC (Elevation The above 112). and Outboard C.V. F006 Control Rod f.o Drive (CRO) I 1-1.0 Liquid /.0E-4 Scram Discharge Volume (SDV) header rupture. (Non 1-1/2 break exclusion). The pipe break frequency is taken from NUREG-0803. (Elevation 78. 63 and 40). 4 1 I >f i

i;.,

. I., { $.'

a 3...

l 9 8 h l- + ?- = TABLE -2 Estimated Frequencies of Breaks Outside Containment Cont'd. l t NUMBER OF: VALVES BREAK ESTIMATED DESCRIPTION OF THE CASE ANALYZED I INITIAL BREAK *L { ISOLATION V BREAK LOCATION SIZE I ASSUMED FLOW: FREQUENCY T I VALVES FAILURE STEAM OF BREAK ( 5., CASE 1 DESIGNATORS PROBABILITY OR LIQUID ' OCCURRENCE Recirc. PermF Seal Injection I 3/4 2 2 2 1.0 Liquid 1.8E-4 5 4 Other 3/4" lines 3/4 1 20 20 Valves of the 1.0 steam 2E-3 Branches from various sys-and system shown in tem shown in Liquid 3"~S (All elevation) this table this table Sample Coolant i l from RPV I 3/4 1 2 2 1.0 Liquid 9.E-5 Reactor Post I.f-4 Accident Samp1-3/4 1 2 2 1.0 Liquid ing system (PASS) TIP Orive Guide Tubes I 3/8 4 2 2 Ball valve 1.0 Liquid 1.0E-4 (Elevation 60). and shear - valve a 9 4 e t e

s 2-3 /48 The combined frequency in each group is shown in Table 3. It. ven be- -not4ced that the LOCA frequencies of the large and medium breaks groups are dominated by the line breaks of a single system. For the liquid breaks, it is the RWCU and for the steam breaks,it is HPCI and MSL drain systems. In the latter case, the 10" HPCI line break has a frequency of 3.5x10-6 while all other line breaks which contribute to the large LOCA steam line break have a frequency of 3% of that of HPCI. Similarly ig ge case of the Main Steam Line (MSL) drain break, its frequency is 92% and RCIC break frequency is only g about 8%. Therefore, in the rest of this study, when discussing large or medium breaks, only the dominating systems' line breaks are included; namely, the HPCI 10" line break, the RWCU 6" and 3" line breaks, and the MSL drain 3" line break. The small steam line breaks are mainly due to HPCI and RCIC bypass line breaks (it is a case of a blowdo*wn. limited by the 1" bypass). This will be referred to as the 1" line break even though the lines may be larger in dia-meter. The small liquid line breaks are represented in the BNL study by the RWCU 3/4" branches, and by the CRD SDV header piping rupture (reproduced from NUREG-0803(4)) which are about 1-1/2" equivalent diameter. Table 3 also includes, for each of the LOCA outside Drywell groups, the liquid or steam break discharge flow rate at two different times: (1) Initially, when the break occurs and flow rates are at their peak values [ (2) At about 30 minutes lat,er after coolant injection.is established, depres- .surization of the RPV completed and operator takes control of the injec-g tion according to procedures, keeping the core covered. (Lv raths estc4 These4 values should be taken as crude app =Loetion. They were obtained from NE00-24708(5) for the purpose of providing some indication of ime available for operator diagnosis and response. The NED0-24708 report he the efet:g: cf. providh this information for the entire spectrum of break size under consideration in this study.

Table 3 ,e i' Summary of Frequencies of LOCA Outside Drywell Break Flow Conditions (*) Initial After 30 Minutes Break Locat' ion Initiator i Initiator Stm/ Liq lb/sec Stm/ Liq lb/sec (Main Contributor) F{egyngy Large Size Breaks Steam 1400 Liquid 1200 HPCI(**)- 3.6E-6' (elevation 8') + > 6" Liquid 1200 Liquid 700 RWCU 9.6E-6 t (elevation 112') Total + > 6 1.3E-5 4 Large Interfacing Liquid 1200 Liquid 700 LPCI/LPCS 2.0E-6 LOCAs elevations 87' down + > 6" 8' Medium Size Breaks Steam 120 Steam 60 MSL Drain 1.0E-4 2" < + ~< 4.3" Liquid 400 Liquid 250 RWCU 1.5E-3 ~ (elevations 112'-126') Total 2 < + f 4.3" 1.6E-3 Small Size Breaks Steam 10 Steam 5 HPCI/RCIC(**) ~3.0E-3 (elevation 8') + < 2" Liquid 25 Liquid (elevations 112'-150') 12 RWCU Branches ~1.5E-3 Total + < 2" ~4.5E-3 (*) Approximate crude estimates of steam or liquid discharge through break from NED0-24708. (**) Break can occur between elevation 6 hand 8, but the 'other break locations discharge through a pipe chase to elevation 8.

3-1 }. MRicaHu Assessment Ok N.'ke) ~ 7 gre In this section the effects of the LOCA outside drywell /s discussed s according to the three different groups:'pf small, medium, and large pipe breaks (see Table 3). Based on these effects, some insight on time available for mitigation is presented. The first subsection provides general informa-tion on alarms available for diagnostics, containment sumps capacity and flooding data, and some crude information onhontainment a/ atmosphere temper-ature increase due steam or saturated liquid discharges. The next subsec-Te5 pede tions describe the tigation conditions for small, large, and medium LOCAs g outsidefdrywell in this order. 3.1 Reactor Building Information 3.1.1 Instrumentation for Diagnostics The following instrumentation and alarms are available to alert the operator in the case of a pipe break in the Reactor Building: Reactor Building ventilation isolation alarm 4 (le Reactor Building equipment sump level alarm in((vicinity of break 4 Reactor Building floor drain sump level alarm Reactor Building flooding alarm at elevation 8 (see additional descrip-tion below) ~ Area radiation monitor alarms _ Reactor Building Standby Ventilation Exhaust high-radiation alarms Area high-temperature alarms on elevation 8 and on the floor where the break occurs Specific systems have their own break detection instrumentation such as the RWCU, MSL drain, HPCI, and RCIC. Reactor Building low differential pressure alarms. Most of these alarms are also sensitive to a small break LOCA of about 3/4" diameter but with(longer delay time; some set points will be reached s 6e uen after half hocr. g g

~ 3-2 v'b The Reactor Building (RB), level at elevation 8 is detected by two RB level monitors installed on the RB floor. The flood alarms are activated by the monitors when the water level is more than 0.5 inch above the floor. The sump alarms will be activatec w1enkwater level reaches the sump alarm set-points installed at a level dh below the level that activates the RB flood alarms. Sump alarm sensors are installed at various locations in the RB. The area high temperature alarms include the following: RCIC and HPCI turbine steam line space high temperature (7 sensors each). Isolation signal setpoint at 155'F (elevation 8) RHR space high temperature alarm (6 sensors) with setpoint at 175*F. (elevation 8) RWCU space high temperature (18 sensors) isolation signal at 155'F (elevation 112) Main Steam line space high temperature (4 sensors per line) isolation signal at 200*F (elevation 78). Main steam tunnel containment penetration area high temperature (4 sen-sors ) located in the area of MSL drain l'ines. Isolation ' signal at 140*F. 3.1.2 Sump Pumps and Flooding Buildup Volumes The open area elevation 8 floor is approximately 5,500 sq. ft. This area is the total floor area minus the area occupied by equipment foundations, columns, drain tanks, etc. Based on this area, flood buildup on elevation 8 is 3400 gal / inch. The drainage capabilities at SNPS are: Reactor Building Floor Sumps - 2490 gal, capacity Reactor Building Equipment Suiaps - 1660 gal. capacity Reactor Building Porous Concrete Sumps - 500 gal. capacity i p 4 v. .. -..,,. - _., - ~ -,,. -, w,

r 3-3 4 These systems have a total sump capacity of 4650 gallons. The total sump pumps capacity is 640 gpm, as follows: Four 50 gpm equipment drain sump pumps (elevation (*) 9') Six 50 gpm floor drain sump pumps (elevation (*) 9') Two 20 gpm porous concrete sump pumps (elevation (*) 9') One 100 gpm leak' age greturn pump (elevation (*)12'). The leakage return pump is designed to process radioactive water. If the floor drain sump pump indicators register radioactive material, all sump pumps will isolate. Th,e leakage return pump can then be manually activated by the operator. In ad$1on, only the leakage return pump is powered from onsite AC .i fesAM f It can be inferred that if flooding is not arrested before,the l' level th abovegelevation 8 floor (elevation 9),Mtthe sump pump capacity may drop from Cwery '=t to accumulation of about 42,000 600 gpm to 100 gpm o This it -~! N2 gall ons. Furthermore, since this study considers primary water release, it is assumed tha't only the leakage return pump woul;d be operating (other sump pumps will be isolated). RCIC, HPCI, LPCI/RHR, and LPCS are all located at elevation 8. It is assumed that they become disabled when water reaches 4' (about 160,000 gallons) as stated in SNPS-PRA.0 3.1.3 Containment Atmosphere ' The SNPS-FSAR includes in Appendix 3C a few calculations of Reactor Building temperatures for water and steam line breaks. Table 4 shows the results' of one calculation for the. discharge of 40,000 lb of saturated water at RPV normal power conditions out of a 4" line break at elevation 112' of the D If water reaches this elevation, the pump is assumed to fail. O e

l l t 3-4 Reactor Building. In this deterministic analysis, the break was assumed to be . isolated by an RWCU iso 1~ation signal at 40 sec after initiation of the break. This break results in less than 5,000 gallons at elevation 8 or less than 1-1/2" water accumulation on that floor. It is seen from Table (4 that a break ~ of this size is rapidly affecting Reactor Building atmosphere conditions. The other calculations reported in Appendix 3C of SNPS-FSA are similar. and lead to the c n"that conditions of 212*F in the Reactor Building elevation 8 will occur under the following circumstances: (1) A RWCU line break discharging more than 500,000 lb. This is approximate-ly the amount discharged from a RWCU 3" line b'reak in 15 minutes (5 minutes for a 6" line). (2) A MSL drain line discharging more than 100,000 lb. of steam at RPV normal power conditions. For a 3" MSL drain line break this will occur in approximately 10 minutes. (3) A RCIC/HPCI 1" line discharging more than 15,000 lb. of steam at RPV normal. power conditions directly to elevation 8(*). For a 1" line break, this will occur in more than 25 minutes, and therefore 212*F con-ditions at elevation 8 from these line breaks are not expected to l occur (**). Temperatures higher than 140*F in elevation 8 can result when steam is discharged directly to this elevation from a 1" RCIC or HPCI line continu-ously. (*)RCIC and HPCI steam ifnes are enclosed in piping chase' which protects higher elevation against a steam line break in these systems.

However, for most steam line breaks in higher elevations, steam will exit at eleva-tion 8.

I**)The 1[000 lb. discharge would cause the saturation conditions only if discharged during a very short time, which is not the case here. e s

. ~.. - -. - --~. l 3-5 3.1.4 Procedures Given h MeMM: LOCA outside containment, the SNPS procedures. dictate se,a rapi adepressurization of the RPV by the ADS. This action substantially reduces the break flow rate. If low pressure injection is provided at about 200 psi, break flow may become only about one-half of the initial break flow. e GivengRB flooding alarm, the operator is required to: ~S l Monitor RB level to determinegapproximate leak rate, and to ascertain the approximate location of the break (using additional sump alarms and high l area'temperhurealarm)- 3 Monitor parameters such as line pressures and flow rate of the safety systems, as a leak may affect these system parameters If required and plant conditions permit, dispatch an operator to the RB floor to visually locate the source of leakage. Isolate the break using the appropriate system procedure (HPCI, RCIC, RHR, others} 3.2 A Small LOCA Outside Drywell (< 1-1/2" Break Size) { 3.2.1 Accident Conditions and Alarms The description that follows is based en an analysis by NRC staff of a pipe break equivalent to a 1.2" line break. This is discussed in detail in NUREG-0803Y The description in this section applies to small line breaks, in t L general, and applies to the SNPS. It does not, in particular, apply to SDV header pipe breaks to which the original. discussion refers.- t The break described is a water line break discharging 550 gpm (- 70 lb/s) initially. This is equivalent to a 1.2" line break discharging from the RPV at 1032 psi conditions. Several alare.: are available to the operator as described in section 3.1.1 above. The nost expected early alarms are from the Reactor Building radiation monitors and from local area high temperkure alarms. A 3 1 ___~,..~...._,.,....._...m..

3-6 eks. NUREG-0803 referentt% a calculation for a typical BWR Reactor Building that shows a temperature rise to 110*F in 10 minutes and 140*F in 30 minutes ~ for a discharge of 550 gpm at RPV conditions. (This amounts to abouty30g0 lb over 30 minutes.) It may activate high temperature alarms if set,4e n 3 120*Fj g{t will not isolate HPCI or RCIC systems. s eJk The SNPS sumps and flooding setpoints are low (see33.1.2), i.e.. ' at 1/2" above floor level which corresponds to 2000 or 4000 gallons of water accumula-tion. Reactor Building sump and flood alarms within 5 o{r 10 minutesT (assuming 35% x flashing into steam, travel time through stairwells and floors, and partial accumulation in equipment sumps (upto 2000 gallons). 3.2.2 Reactor Building Environment The water released from the break will exceed the local drain sump cgc-ity, and some will flow to lower elevations through stairwells. Assuming only g the leakage rhurn pump is available, the,ayugtfon of water at elevation 8 would be less than 0.13 inch / min af six hours to reach the level that threat-ens ECCS equipment availability. Thus, ample time is available for the opera-tor to recognize the need to depressurize the reactor and reduce break flow. p.k N Appendix 3C inhNPS-FSAR states that equipment along stairwells is protected against dripping of 212*F water. During the initial blowdown, temperatures in the nearest area to the break can reach 212*F. The Reactor Building temperature is expected to rise significantly as shown in Table 4 for a discharge of 40,000 lbs of saturated water at elevation 112'. This is a 10 minute discharge from the 1.2" line break described here. While it may result in high Reactor Building temper-atures when discharged over a short period of time, it results in 110*F in the Reactor Building if discharged during about 10 minutes (see section 3.2.1). However, the temperature in containment will continue to rise due to the con-tinued discharge through the break ang regh typj 5*yCgHPgQsolation g temperature after about one hour. ThepBSYS) of SNPS has a heat removal capa-bility of less than 5% of the heat discharged by a 1.2" line break, before I reactor is depressurized and the break flow is reduced.

~ .-.-u ,_..m, 3-7 Table 4 Reactor Building Temperatures at Several Elevations Resulting from a 40,000 lb.' Discharge Equilibrium Initial Maximum Reactor' Building Temperature (*) Temperatures Elevation ["F] [*F] Comments 8'-0" 104 < 140 40'-0" 148 63'-0" 183 -78'-7" 194 112'-9" 217 Break location at 112'. Outside the pump room temp is 177"F 150'-9" 148 175'-9" < 132 (*) Reactor Building humidity changed from 50% initially to 100%.' i 3.2.3 Operator Response 'bt The small LOCA will result in immediate scram. 3(perator will try to keep the normal feedwater injection and therefore keep MSIV open. If4MSIV remains open (which is t'he more frequent case);the level M not be able to remain normal with RCIC injection alone. Feedwater or HPCI will also be needed. Thus, within one hour the operator would recognize the loss of inventory from the mismatch in the injection flow,' or from the need to use HPCI or feedwater, If,MSIV close, sir rather than RCIC alone. , operator would expect to observe level recovery within the first 20-25 mihf RCIC @ CRD work. In the case of a 1.2" break, it will take longer to recover level and that would be recognized by the operator after about 1/2 hour. m 9 ,y Os,.._ ,,__-..c._...-W_a ,,,__r.____--,,.y,,, ..e.,w,_,. --r

~ 3-8 4 In addition, the operator will have a flooding alarm and high Reactor Building radiation alarm at about 10 minutes as discussed in the previous section. Therefore, it is expected that the operator will recognize a small break LOCA in the reactor building within about 30 minutes after scram. Unless the operator perceiveha LOCA, he will depressurize the reactor at a rate of[180*F per hour.ody". In such a ' case. it-will-take--4 hours to depressurize the reac-tor to 100 psi and reduce break flow by about a factor of 10. As seen in sec-tion 3.2.2, four hours are available at SNPS, without flooding) elevation 12. However, in this case, atemperature in Reactor Building may reach 155*F or Ltptr mope (*) between 1 and 2 hours and trip HPCI and RCIC, and most probably require depressurization for. low pressure injection. These events would lead th rator ~ to recognize the small LOCA outside containment with31gh y, if he failed to recognize it during the first half hour. & 2a><ftL~hw D a)Las t It should be noted that unlike)NUREG-0803, the BNL re"hnrs believe that recognition of a small break LOCA outside drywell at SNPS would be a high probability event mainly because of the improved arrangement for flooding detection at elevation 8f (relative to the arrangement assumed in NUREG-0803). High rgigi,gr) anp jig,hgemperature conditions in the reactor building will enhance theptRPfe. BNL assumed that it is most probable that manual depressurization of RPY to reduce flow and enthalpy discharge through the break would take place after'about-30 minutes to I hour into the accident. The depressurization of the RPV may reduce flow rate and enthalpy of the water discharged through the break to a level accommodated by++the sump pumps, 4. W. r and may reverse the conditions in. reactor builgng end,thEy may start to im-g prove. It is indicated in NUREG-0803, that^ blowdown panels =*"re. may be k required to establish a pa+ss for leakage of hot humid air to outside contain-ment (which is larger than the " natural" 100% per day leakage rate from reac-tor building), in order to improve the reactor building atmosphere conditions (*)A GE analysis estimates that the maximum bulk temperature in the reactor building would reach about 140*FQsg.p c -o to 3 W)

_. ~ 3-9 and to allow safe operat'or entry. As shown in NUREG-0803, depressurization reouces significantly the dose received by an operator entering the reactor building. u If noperator is required to enter the reactor building to isolate a break, it can b_e done for a 1.2" line break with early depressurization (and low pri-mary water activity). It would be possible to stay for an hour, and this seem to be sgf{icient for isolation purphs. Ap 3C of SNPS-FSAR considers 30 minutes y sufficient time to walk through all SNPS elevations, locate a break and isolate it. 3.2.4 Estimation of Core Damage Frequency The description of the event and the reactor building conditions follow-ing a small break LOCA outside drywell were discussed in the previous sec-tions. These are now summarized.in the form of an'eventbee in Fig. 3 and quantified. Feedwater and high pressure coolant injection are in general available undegtgrcumstances of small LOCA. ADS, LPCI and LPCS have very low f pwavut. k r ; g e b o b i l :t. k s. The values for their quantification are taken from Ref. 2 (r:e&d Offt. The events that are differently quantified are: (1)the probabilitht 30-60 minutes h operator.5 takek ' actions and complete { rapid manual depressurization, (X ); and (2) the probability of controlling the H condensate flow if required (V). The X =0.01 is taken basically from H NUREG-0803 where 5x10 2 is used. The difference between N nd BNL values tg A { to the SNPS improved early flooding alarms da4 increasef the proba-La

• =~ ' a t.

l bility thatfoperator recognizes the LOCA outsidefdrywell ~ and follows the required depressurization procedure. The V=0.1 is the common va'lue used by BNL in Ref. 2 for controlling condensate injection if sufficient time is available to the operator (in our case 30 to 60 minutes). The V=0.02 includes a factor of 0.2 for the possi-bilig that no fagage: to LPC1/LPCS will occur even under theggumstances that Aoperator wi+1 not depressurize the reactor early, but hem i-l depres-surize it. at 100'F per hour rate, for 4 hours or more. In such a case NUREG-0803

.%.m*-- nas e e IW 3 11 O + t*.4 Ts h i e3t 7 e i A e s - n w s ky7o W 8 m u a p $d 7 S N o ~1 w = n e a n M v k N -2 L 5 a ss s 7 w %7 + 4v4 T 'o 4 E'; r E t 494 > e e U a.ia. x= n -e g*kg 4 o 6 O' g e p Wg> g 4 0 X d b.te a W DE 5 Q Of 7 a. e E 2-a 5N Y 5 w 13 d e 6 e it 4, u, a v s u, y .3

3-11 indicates that entry to the reactor building may be delayed for sever-al,up to 20 hours. The LPCI/LPC3may survive the adverse environment in the reactor building for such a period, because they are(coveport y tilqualified to sustai)n these condi-on F e) i hk tions for at least several hours. A fac' tor of 0.2 the LPCI/LPCS availability g apparently underestimate their availability. The result ef the eventltree quantificationf t core damage frequency o[ M q,m,4 jNvfor small LOCA outside drywell, a=cunte u, rcughly 1.1x10 6 per y :r, when no A if u n y motor operated isolation valves h gina, arf fu hl It should be noted that no distinction was made between steam and liquid breaks in the case of the small LOCA. The calculated core frequency would not change much if[ distinction between liquid and steam break e!d h5-vare made and apparently the flow out of a steam line break would be smaller after depressurization. 3.3 A large LOCA Outside Drywell (>,6" Break Size) This case was treated in the BNL SNPS-review (2),.However, the assump-tion in the present study is that HPCI and RWCU isolation valves would fail to close. n jpt("*" Le ct) OKyHPCI.wef treated in Ref. 2 and a cccc d;m:ge frequency of 2.7x10-8/ year f was obtained. If we postulate that the isolation gives fail upon demand,,$he a. /-dfr*tNc: ult: Ybd forY10" HPCI line 5 0= 4 Teb!e--3, of 3.5x10-6/ year ;;E2[t: e . break.U N g%4tr The4RWCU line t :9 ef 5" df: meter has three isolation valve @beerd) ins,4 the drywell. Only one of them close automatically on sensing line break con-ditions in the RWCU lines. In Tabled 3 when no credit is given to these valves ~ ~ a break frequency of 9.6x10 6ps as derived in Appendix C of this report. In Ref. 2, the three valves were considered (having different isola-M W ub.M tion signals and one of them is of a different design)j he, it was assumcd t their failure upon demand would be less thah 2x10 "/)d and the frequency t Thus& M.s of6"RWCUlinebreakwouldgeabout108/ year. t it was not further con-sidered tpegMYts eONES interfaci As was calculated to k have a i.equer.c., cf. 2x10 6/ year { $ u TM g 143 e f tk ry,t for rMtp-y w n uan. 9 s

d e 3-12 4 l The interfacing LOCA frequency is also estimated in Ref. 2. The results are reproduced in Tabld 2 and 3. This LOCA frequency does not change under the specific assumptions of this report. fle The total frequency of large LOCA outside, drywell assuming isolation failure, and whe/n including interfacing LOCA np4 become 1.5x10-5/ year. When 2 this is used with the event tree of Ref. 2 (see Fig 4), a core damage fre-quency of 3.0x10 8fs found. The 0.2 factor is the probability of,arr~ operator [ dre te fML to control the condensate system pumps flow to the RPV in the short time available (about 10-15 minutes). In the case of a large LOCA outside drywell, the discharge to containment l 1s about 1200 lb/s and saturation conditions in the bulk atmosphere of the reactor building are reached within 5 to 10 minutes. The ECCS equipment at elevation 8 would be flooded in, about 15 to 25 minutes (the lat,te aber possible, as was,in% flashing). corresponds to 35 Thus, it is obvious that no. e rt:p is e assumed in the SNPS-PRA and the BNL review for large LOCA discharging saturat'ed water or steam into the reactor building. me W heet d m.- %. .t.,. LM

• Thei resett,[s ser t. 5:

34 7 times h4sde.r than,in Ref. 2. This is / because in Ref. 2 the interfacing LOCA were the ' dominant contributors. They are dominant when credit to isolation valve closure is considered. 3.4' A Medium LOCA Outside Drywell (2"4 9 4 4") 3.4.1 Accident Conditions Alarms and hperator hesponse The most dominant case of the medium LOCA in T& is the 3" RWCU line as th e u Taloi) breair(. Ine f requency of a RCIC 4" line break is small compared to the total medium LOCA frequency of 1.6x10

. the RWCU 4" line break frequency is signi-ficant 'but the sections considered are relatively downstream and estimated to be 1/4 of the total RWCU break frequency,.wherkthe other 3/4 are for 3" line break or less.

Thus, our discussion in this section refers to a 3" RWCU line break. 4 4 4 i ,.....,,,,.~n ..,,.-.,n..w_n-,,,n_., ,. - _, ~ - -,,,,, -,..,., -, .n~...-_-.

e I f l I V V o b ea s s s s r r s oe s s s lac n a a a C l l l l u C C C V ) Y r C Y f N K K K K E x U R O O O O E Q { E r R e gF F P( R E O V C T N A E N V V V U G T l l ll l l l QI l T T T T T T 0 l l l l l E S l 0 0 0 0 0 0 S E D A ^ ^ ^ A A A gn iW uof e O S = l w C w l P oy L f r 1 A sf VO 0 e M E c R n r. e c. T uq x. AE t T e! I C S : E e t R r-I w y D o f C; 5 me 0 ad ri gs at E i u T A DO S s N A eA e E e eC D v N N rO O TL C t e

ng N

er 4va O I EL TC E J t7 N l I C a s: P y T yr: L N 0 A L 1 O f'; O C s T 0 ~ S s C v H A 4O C S 5 -0 1x l { A T C N O E E g L DI t 7 4 A E C S l A, t O i A G M t. uI g cN ( O C

I

o 3-14 A4 !!ca ter, kt 150)are The RWCU is located at elevation 112' to 150'. located,(t'fIe deminerah'which process water at low pressure and at about 125*F and, therefore, not considered. Thus, the break location of signif t-cance can occur at the 112' or 126' elevations. On these elevations, the line are enclosed within concrete shields roviding physical separation from all safety related equipment (see App. 3

1. 1. )

Table-4 present the approximate temperatures in the reactor building fol-lowing a RWCU 4" line b?eak at eleisti6W 112'in the RWCU ~ pumps room. It is estimated that about 10 times the amount discharged in that case, i.e. 500,000 lb.,would result in saturation conditions in the reactor building. This will take about 20 minutes if the flow rate of Table 3 (400 lb/s) is' assumed. It' apparently will take longer because of the decrease expected in the break f afterfewminute{upto10 minutes e to depressurizatioQ It is expected that the blowdown from the break will cause immediate MSIV closure and loss offfeedwate(. Nn about 10 minutes or less, the temperature a elevation 8 will reach 155'F and trip the RCIC and HPCI, which started a few minutes before that on low level (L2). Therefo , in this case it is 'immate-j rial whether the operator depressurizef the RPV, Marly automatic ADS actua-tion is expected for this case. The water discharged during the first 10 minutes would flash (357.) and Tt%.ht d te the _sast (a hnut 20.000_ gall ons).wi l l cas.cada_thenugh._the. s tai rwells_. t o_e.l evat i on 8. Appendix 3C of the SNPS-FSAR considers this effects and states that no safety system would be affected. This accumulation is equivalent to 0.5ft and will result in flooding alarm in the control room. The radiation and temperature' alarms are expected to be on in many areas of the reactor building. Therefore, it is believed that the situation of LOCA outside drywell and the reactor building adverse conditions and flooding would be recognized with high cMb Dwithin the first 10 minutes. ty Earlier recog-nition of the LOCA and depressuriz he RPV would not the pro-gress of this accident sequence. However if operator 34&l fail to recognize ' he event and b( follow the procedures ( call for keeping RPV at low pres-t sure and to controlc)the injection flow) then the reactor building conditions may severely deteriorate.

3-15 The depreslurization would apparently happen at about 10 minutes. Then the LPCI, LPCS and condensate pumpf may all inject water to the [V, and dis-charge a large amount of hot water through the break. While this hot water would have less enthalpy than the saturated water discharged during the first 10 minutes, it ts: Loss flooding potential because of its high flow rate., Flood-ing may occur in[ additional 30 minuteQf,jhe flow rate to the RPV N not ,be reduced by keeping it at the lowest. pressure without uncovering the core. This is the operator action specifically required for the case of medium LOCA outside tdrywell. In such a case LPCI/LPCS ma maintain core cooling for long uw e 4 period and condensate would not be needed several hours into the acci-dent. 3.4.2 Estimation of Core Damage Frequencies The estimation of core damage frequency for the case of a medium LOCA outside drywell is shown in the event tree in FigureC5. The initiating event does not distinguish between water or steam line breakj They are considered similar because even though the steam discharge through the break is smaller, the impact on containment atmospohere tempera-ture-and pressure is about 5 times higher for{ steam line break than in--%e. [o case of a similar size water line break, t4 In the long run after,RPV is depressurized, the flow out of a steam break may be significantly smaller ifyhe core is not flooded so that water is dis-charged through the break. If water lev'el is kept below level 8 (L8), then the steam flow out of the break is expected to be relatively small. Thus, it h fir c. e.6 may not be sufficient to create X flooding Ms oms to the ECCS equipment. The liquid line break is therefore the dominating case. Thus, the event tree starts with the medium LOCA frequency from Table 3.. The feedwater and RCIC/HPCI are assumed to be unavailable. Depressurization by ADS is con- ~ 10 minutes into the sequence. The low pressure sidered gccur at about injection will start to flood the core. Thereforehoperator action to control g 9

I g S O \\ i 1 , Lie 4 u 7 ~ ll t /. ' - m. .b w l ', 3 i = h ) 1 \\ \\- a w i. 'f k g W 4 4 3 A I "U (. o . Q, i og E e t o ( '.- e. e o e Sw m V W u e w e .c t' J o-4 ,y d J[ W W \\. 4 lwv 't I \\ \\ W G K 1 1+e Y. 1 + .s = m r =% A

• g 1.

D y Q 8

o

( a d 4 o d g w h d ( 814 u, u ( 4 Sw e W ef =P k o d ( (3, 1 4, x * ".s w + a O w I w 3 y L ** L.U e $4t A (, w e p F g s i T W l A 3 n i =e i 4 \\ fD K I W "U + \\ Q j w '/ O (; N 6 CYJ 3 M o a., u o "C, Y 'f'I e) 1 >4 5 n -1 x o o 7 w 2 1 3 cr e 1 LL, d s i, j W ct 5 w o4Ti t 2 9 i pa N a ~ ..--- - -..-t -

3-16 -{;k tiede$ 4 fl 4444 ow rate is needed to reduce the impact on, reactor building and gain time beforekondensatesystem'wouldberequired. If operator recognize the need to vsv atto control the injection then the condensate system pumpg ge 'jgr gg .,4 controlled at a later time with a higher reliability. If. net,3 ess time would 1 t he wr8 be available fer tha ep:. ;ter to control and -li n condensate pumps injection, i because they may betope needed at about 10 minutes into the accident. tksusLk b 4 The values used for operator action are -- 14c+4e!y on the conservative side. Therefore, the core damage frequency for medium LOCA outside drywell' is expected to be smaller than k 8'x'10 5 for the case that no credit is 2110:::d f eets for RWCU isolation valves. O e 4 O O 9 6

f 4-1 4. Summary ~E.t f c eba) a con key h s.s e,

• l-4. 2. s to * /o >-

The BNL review (2) of SNPS-PRA fuend t-het LOCA outsidhdrywell cor.tri - A we e t bM A -2dO 7 tcrsee dem ge %?'eacy in the SNPS;4mainly due to intergcing systemL{ In this study an additional assumption was introduced, g NRC request;sth ' isola ion valve would be failing to close upon a sWs ~ reated as u .s o ve th e,L v a u-demand. % ffec of this assumption is shown in Table 5. It is een that the core damage frequency increased by a factor of almost 50. The con-th tribution comes from medium LOCA outside drywell,gn particular,the RWCU 4 3" line break is seen to be the most important (see Table k). Table 5: Core Damage Frequencies for Unisolated LOCA Outside Drywell Class V Core Damage Frequency Initiator Isolation Valvet Assumed Isolatior Nsfumed toWlose"on Demarp to Fail to Close

1. " M UE b ETl Interfacing LOCA

'4.0 E-7 [ g Deg{nd h,A uc) 4.0 E-7 Large LOCA Outside Drywell }M q8 2.6 E-6 Medium LOCA Outside Drywell 1.4 E-5 Small Loca Outside Drywell T-1.1 E-6 Total 4.2 E-7 1.8E-g gle 2 provides the information on the most important isolation valves thet --the4r failureycontribute to. the results of Tablef5. RWCU isolation valves are the most important. Next, but by far less important, c.SM HPCI and MSL drain isolation valves (see Tde 2). -r M ) 3815 L hJ u hr & o.y v p i.. o. c H e el n'u &L1 1Lb a 4L de .-p h a c M t: - m t v an te 50 dominated by the RWCU medium LOCA breaks. NP-8. < 3 Also, the large LOCA contribution comes mainly from the RWCU system. There-N, coJan,,, w L u cp. fore, it should be noted that beside the inboard and outboard isolation o id s,.4 valves,&,RWCU has two additional isolation valves that do not get, automatic ,4p% A. >-ece g 3 sisnal to close when a line break occurs and are available for timely c[^*M+'f-I c L C C+M -act4ee. This action can be taken half an hour after initiation of the acci-V dent when the reactor is depressurized, and before the loss of low pressure injection.

k.* 4 4.h ULJ .E... %ohbs'.bkr.c, heI h...b.41LUhtb.~...Sk o-ob&Av.cAsr...Ao? & s kr.'m !* y... T'b'S 0.Y ey Ser t s ta._. A H saleNn,_ k., $ psy G n. O _ 8 u.r4 . A 4 5 b wA_.43, I 9 E 4 - ,2. Q.IfleM.,.fl..fh..a.,.f._.Hnna_ .k ay, w, _ A ke. g,. u. o..l___ llt Skaye)a w Yu._tSw P.auv SO_a fe.s Eo!.A,d des. A tl<- A iresi n.,} ". EloFcRLeg Nofo kVL - Hu A G G U1).}_,Je c. I 9% 4. " Fe a J Q h, A,A, r,i 12*f.rt SL. L 4 L 1 - St.)...>

Lad, 0

Tr As ) L,, U. ~, Qu, " SuPc-1 Ft AIt,. ( R racse.. 21 k.. tera). 4. 6 wer e Sh_7 b bd. 3

• R (p. r 4 Ag.$r49 Idynh._,/

8up S er y c << s k f:p_cq *. Atyi_t e. G.- o.t_o3_,_ky t 9 t t. SM efe.J _Id.t_m.!r.m_Aq.w3L__h Etc .(GM_.61_.s.uc. Aspar4 5-e c,w AsAcq_.k sIu Auds r s.*>_.6L Rets.r k N.U o-2h 7 U T. Dt.r._f.3.to. C dy lfab.. " b A.CLeU N s} bC N Ae1Its ed. S. (, - Lt5v m y .%.s u.tc..0 N' b.~ f.oy _P.0+ ]^c',.1x.19 s H -IL/o o ; +1vya *l7t+ -o I (s. eef. /1vs. c SLL 1.x,_aJ G.. % 9sarnr, c[ u_,e}eer,fe1 e}_ftya S4 e % f a.L ea 7. /.)L1 %Io R__*Je r.r ", ._E_p A 1 - # P - Id I. $uy./977. W. K 1 E.h2b.b.,..s,)c.F.,b}br, ' p)s_$ay_m_3.rgr s _of l R As a w I4bve 2 a$ l V S. % wa..a

nusAs, P,< s PCs (.

Uut c.6/ce-12(3. cc e-sn-rtis_. AA.a.) __ t 9 C 0 - 4. OCCPEG PR A - A fr,Q,b 4,, A s/e $rle/J md o} Oc o_nn &$,

1. > 'Tiaaf_14.tM

A-1 Appendix A Pipes and Valves Failure Rates A.1 Pipe Rupture The~ main data sources used for probability of pipe ruptures were the Reactor Safety Study (6) (RSS) and EPRI-NP-438 report (7). In the Reactor Safety Study, pipe rupture rates are based on the large amount of data prior to 1973. The EPRI report includes data for an additional two years. Even though it does not change the RSS results on pipe break rates, it provides more insights on the failure mechangIt also highli,,,gding to pipe brea le and pressure surges. ts that expansion joints and reducers may be at locations more susceptible to breaks. In the BNL study, reducers and valves were considered a'st, rupture locations, in addition to pipe sections. The SNPS-PRA(1) usek the RSS data for pipe breaks. However, it distin-guishes between pipe sections which are " Break Exclusion," 1.e., are designed to criteria provided in Appendix 3C of SNPS-FSAR(5), which basically allow for larger design margins and higher quality control of these sections. These increased margins are assumed by SNPS to reduce the failure rate of these sec-tions by a factor of 10. BNL accepted this assumption, and the basic values used in the study are similar to the SNPS-PRA and are summarized in Table A.1 below. O

i l Table A.1 -l Pipe Rupture Rates 'i Computational Mean Assessed Range (non break-exclusion Computational Break Non-Break l Component pipes) Median Exclusion Exclusion j Pipes > 3" dia. 3x10 3x10-8/hr-1x10 18/hr 8.6x10-II/hr 8.6x10-18/hr per section [ Pipes 4 3" dia. 3x10 Il - 3x10-s/hr 1x10 8/hr' 8.6x10-18/hr 8.6x10 8/hr per section ) 4 O 6 t t 1 t .o se

4 A-2 The pipe rupture data of the RSS is applied section by section, where a section is defined (RSS,' page III-41) as follows: A section is an average length between major discontinuities such as valves, pumps, etc. (approximately 10 to 100 feet). - Each section can include several welds, elbows, and flanges. In this study, the pipings were also divided into sections rehere discon-tinuities were considered to be: -- Valves -- Reducers -- Pumps -- Heat Exchangers Appendix C presents the details of the pipings and their div.ision into sections. A.2 Valve Failure Rates The main sources used for valve rupture or excessive leakage failure rates were the-Reactor Safety Study (6) and the h0 REG /CR-1363 report (8). The values of the NUREG/CR-1363 evaluation are about a factor of three higher than those in the RSS (see Table A.2 for comparison). However, the NUREG evaluation includes also small leakages such as from packing failure. Similarly, the internal leakage rate of check valves given in the NUREG evaluation includes Tech *Yecifications many small leakages. which are just violations of the ' limits, and too small to be considered in this study. The NUREG/CR-1363 evaluation reports about 130 LERs artiunder the title of " External Leakage / Rupture." ' Howev'er, no case of valve external rupture has SNPS-PRA(1) estimated from-this list that a value of,1/18 may be occurred. used to modify the RSS rupture rate to better represent severe rupture of valves..This value of 1/18 is also used in this study. s Based on.the above, the BNL study practically adopted the SNPS-PRA approach, i.e.: ..y 4- . ~., 29ye,,,_,,9.,.,g~o. ,,%7y,,..,

" ^ '

• p

.... CL.. M_ h 1 ' .fak.vt-ku,f*_b_lA.t..4. o r.. [ x e 4.sso.v_.e._ L < k s s....~.A..s..It..s.. ~.. . c.w. p t t.co. 4. O t$<21e S.. .... AAa u b"'I"........._.. - _ _ " _ _ '. _... _. . -.. (owp ok G..-.$u.v..c..-.. A ** .. J.e Sr..k.._._ y,..g,,,! a. H, ~ E==kst=a Endu u.t ...d..e, .. [. Q+.. \\ [ka-'3 C M-* 1 ^' ..80.._..lo'g. 'y.- -. c Ne el< 0, S S... CJuns).Islc)a. 3.s '~ to' ' ~- s asvera .Vafves...-.... ~ A/OltES/CR-y "*' y F .f.x..,.,~.6 13(, 3 ( n scp,y. I c we.k ...V e. -- t...-

a. A h,t.r dpe,hi

. MURE 6 CA-Emb tlaL 3y '. -.. -.......

1..= to

. 7 = go.. T -7 vabs 136 3 /Aptre l ...............--..j 8 .e g h gggm e. _-.9'1-

r A-3 (1) Use of RSS failure rates for valves. (2) Apply a modifying factor of 1/18 to the RSS valve rupture data. (3) Distinguish between valves which are in the break exclusion zone and those that are not. A factor of 1/10 is applied to the rupture rates of the break-exclusion valves, similarly to the pipe section they are located on. To summarize, the value used for valve failure rates were: check valve internal leakage: 3.8x10 7 x 8760 = 3.3x10 3/ year .g th 81 x 8760 (= 1.3x10-8 valve rupture (break exclusion): 2.7x10-9 / year 3 (non-break exclusion): 2.7x10-8 x 8760,l'ht)= 1.3x10-s/ year.

• For simplification of the analysis, the valve rupture rates were also used with other discontinuities between pipe sections, such as reducers or pumps; this may be a conservative assumption.-

In addition to valve rupture and internal leakage, other failure modes of motor-operated valves were needed in this study. The additional failure modes '~ and failure rates used are summarized in Table A.3. A.3 Comparison with LOCA Frequencies The analysis in the main part of this report involves a large number of pipe sections and valves. In general, more pipe sections and valves are located outside drywell. Thus, the frequencies of LOCA outside containment should be a large fraction of the plant's LOCA frequenc.ies. Table A-4 compares the results of the LOCA frequencies from this BNL st'udy with the RSS results (table III-6-9 of RSS), the EPRI-NP-438(7) results, and those of the SNPS-PRA. k w D l

l) l 7 i a. 4 j / u 'I s y y 1y j A 4 a f ,O = l 3 e e b g 6 a S ( v" A c d L N~ y y g j u ,s g e F ud 3 g 4 7 eV l l 3/4 a r a V 7 = 3 l b. 'o r t e = p O 3 ro e L ) )6 f L. h n t. a. u. h {* e. F. .2 e' e. a e j* a. 3,, ,. v-p .d s i o r f F> N u h ** /a u,f t p a. u v ,Q o< a w u rQn. u o oc FPhc. F o (w a.r F e e r e .o 3

v. u.

e. I )f c ) t

u. y r

a s ) c R u g

a. %.

n u / / n e o 5 <3 H s$ w B l f 0 o g a, R su a H r. t c F w u. g t d P n a uI o s ov } p G f u d n a ) f w rds )v e b uE v o o, o N I C P O V ( i: 4'\\'!i!i!l1: .;il,\\ i!!: i

• '^~""-*4

_m _e_ w w

• ,~,-

m, s e O m

• =

f gdt wh A T a o o e, 5 9 o E K j y ~ 4 *& X w u e D 4 M hM de w h y ~ .f W C. 4 b T W. b h 8 I L $ g d s a e y u .a a d Y N

• O el g

b 4 i 3. 5%> rn iE. ~ ) m v Q, W-w e m 4 W 4 k*7 wa s s w R *d b s e o w g 4 L 4 m a ;;, I M N N Of 4 o$P 'E w *- 4J ? s N pt a. j b 1* 2. g 'u d we 7, I i 8 [ 2 o 4 he ~. m w 1 z O e LA .d C "" y 4 W ~ 5 n. .A 4 + 0 E* s p 1 7 l i e 4 5g h d 8 7 d ,8,,, Q IL a y o. x A L fo9 'E 4 't a w 4 d e 7 Q . A ?s ~ 4 I f$ -I 2? h g el e 8::r & d. 'g 4' \\g O 7 p 45 o e f 7 d ,d 4 h 3, ~ 'e g, emmS D C 4 4 A 0 D b O z m + 6 8 s o o $ + Hd .'o 9 Ik M h z = v 4 5 4 g i w o 3 n p O e E N E k M g O-. 3 ~ln N E M ~

~ a

i

-Table B 1 s: PROCESS PIPELINES PENETRA ING PRIMARY CONTAINMENT- .R o (Numliers in parentheses are keyed to notes on pages B-6 and B-7; signal codes are listed on page B-8).. g a y C. 3 VARVC 19CA1104 vat W 400/05 ,;Palnnaf IRNSER VALVE 5 NOMiuAt tilAllW 10 CPERAIOR POWE R POWER (105tnG N0Rrint AI Al q gl 0F Pts PIPE slM PRIM 4af ITPE 10 Gpts 10 0105E 15084110s. IIHC (5(C) 51Alus g ins inAlien tint 5 15ciAils (22) uc tim:5 tint (Is.) ConiAltdual (6.22) (5.6) (5.6). SlasAL (10) (a.s) atMana5 s .I IA.S.C.3 ble Slese 55 4 3 24 leside A0 Globe Air /AC/DC Air / Spring - B.C.S.C.P.R.T.lut 3-5 opea (I) 1 24, Outside Ao Chl.e Air /AC/DC Alr/ spring B.C.B.C.P.S.I.181 35 Open (1)

• O N ln Steen line Drale and Milt-55 4

1 2 beside m Eint.e AC AC 3.C.D.E.P.R.I. int.0 4 Osea 'O leatage Central Systes 55 4 3 2 t/2 Outside MS Clet.e AC AC 6 Closed (59) a i r0 t ' 2A f eed ater 55 I I le leside Check flew lieverse flew Beverse flow g/A Osen '[ s- .O t is outside VIC flew Deverse flow / Revesse flew /f. N/A Opea (II) s Alr/ Spring G.AM ,u3 t te feee.ater 55 1 il Is InstJe Ched flew severse fl.w a verse flew N/A Open (ll) ~;.f: n/a Open M 1 IS Outside VIC flee gewerse flow / Reverse flow /f.G. Alr/ Spring. kN n ct t3 Mala 5tese tlae Drain 55 l I 3 Inside m Cate AC AC e.C.O.C.P.R.T. AnA 16 Open 0 1 3 Outside MG Gate DC OC 8.C.B.C.P.R.T.nM.0 16 Orca 7 ~ O {.

I-4 WCW Line free RPV 55 1

1 6 laside AC AC A.J.lul.0 30 Open E ,m Gate i 6 Ouiside e G.i. = OC 4.2.v.v..n m D e. g j i !-5 aim 5heiduwe Coellag free RPW 55 5 1 20 laside MD Cate AC AC A.F.5,9ft 23 Closed c I j I 20 Outslee le Gate IC OC 4.f.U.let 23 Closed 6 tD t-64.0

• • lajectise llae to Secircu.

55 2 1 24 leside WIC flew neverse flew. Neverse flee N/A Closed (3) { ~ lallen System Setura 1 2 laside to Cate AC AC A.let lg Closed CD 1 24 ' Guistde HD Gate AC AC el 24 Closed (12) r+ t 2A.B 48 - Centalament Spray Drywell 56 2 l le . thetside M3 Cate AC AC F.C nit St Closed (2) U I le. Outside MG Angle AC AC F.G.AM le Closed (2) M to tu

8.sA s isla - Conf alneral Spray 58 2

1 8 .I Sutside m slet,e AC AC f.C.hpl 11 Closed L2 1 8" j Suppsession Chead.er I 16 Outside MS Cate AC AC F.C lut 21 Closed L2 l o j l 16 Outside to Globa AC AC f.G.R81 79 Closed 1 2 1 5 i 03

a $4.s.t.o em Pump section 56 4

I re .htside se 1. ate AC AC RM, 106 Osen (13) C C1. i e 1 1 g i

,4 Table B-1 (continued) PROCESS PIPELINES PENETRATING PRIMARY CONTAINMENT vat W $0(A11018 VASW A180/04 k g;IpAng leemfa vat W5 sposelssat R!I AllW 10 OrtkAIOR Pon8f8 Pourt (105thC NomIAL I t(21 Allerg all Of Pit fif( 587( talIJ4f 11Pt 10 CPlit 10 CIO5E 150tAlloll llME (SEC) SIAIUS ~i li.td.ihAlloi45 tilif5 150tAlf0 (22) CDC lin(5 lissE (III.) WalAlldful -(6,22) (5.6) (5.6) SICatAL (10) (8.9)

• N HAasi s.10A mal test line 8etwa 56 8

t 86 Outside M0 Clebe AC AC f C.M 79 Closed (2) [ to Suly ession tha=d,er. 5=ppresslen Pool Cleanup hetern, t 2 6 Outside MS Cate AC AC A.F.M Suit 5 team Condenslag Olsclearge. I 1 4 Atside M0 Cate AC AC f.C.M 20 Eleted 38 Closed } DHR Minimum flow. I I 4 Outside MG Cate AC AC fut to Open (16) Cove Speay last t ime, and I l le Atside M0 Globe AC AC f.C.AM 67 Closed 't j j (see Spray Mietsman flow I l 3, outside to Cate AC AC tel 16 Open (16] a s let hua lest time Resure to 54 1 1 16 Detside MS Clobe AC AC f.C.aN 73 Closed (2) Suppression (hamber. RCit Mininue flow. I 1 2 Outside te Clebe DC DC WM 18 Closed ItPfl Minimum flow. 3 1 1 4 Outside to Clot,e DC DC

• lot 29 Closed s t&)

l16 Wut Stese Con.h nsing Olscharge, i I 4 Outside te Cate AC AC l'.C.RR 20 Closed B IblR Mintaman flow, l 1 4 Atsice le Cate AC fore Spray test line. I I to Outside to Globe AC AC f.C.ast 41 Closed 4 AC m 20 Open. (16) Cose Speay Mints.se flow, and I l 3 Outside tu Cate AC AC kN 16 Osen (16) Bellei Walve Olschange from Seat i 1 2 Atside Relief Valve High Offfer. Spring N/A N/A Closed 5=l+1y to E(IC Pump Sectlen ential Pressure i s.ll lete liesd Spray Line to Ret 55 l 1 4 Inside Mo Cate AC AC A.T.W.lWe 20 Closed l i 1 4 Outsise MD Clebe DC DC A.I.U.8N 11 Closed I 3-l2 HPCI Iwebine Stean Inlet line 55 I I le Inside Mo Cate AC AC s.m It I I Inside se Globe AC AC E.RH 12 O en i t 7) I le outside te Cate DC DC E.tM 41 Closed i G en l l II l' l Attide M0 Clebe DC DC a.an 12 Open 17] l I) i I-il HP(5 Iwelaf sse (shavst 56 1 3 58 Alside le Cate DC OC itM 102 Open 2 le Outside Check flew severse flow iteverse flew N/A ' Closed l l s-84 Spare (15) j i f s.ll ear (I Pune Section 56 8 8 15 outside. De Cate DC DC s.let FI Closed \\ j s.16 p(IC Imbine Steam inlet line 55 I I 3 Inside to Cate AC AC

c. tit 16 Open L7i l

I Inside le Globe AC AC t.RN 82 Open 7 1 I l cutside MG Cate DC OC E tM 16 C)esed 7 i j i Outside M0 Clube DC DC K let 12 (h en , I i i [ l } I 1 )

~ Table B-1 (continued) 1 PROCESS PIPELINES PENETRATING PRIMARY CONTAINMENT d reemet v= w 60CA 10m vAtw Mm/Da usata v= ws moninAt sil Allw is OrtnAita tours touta CiOSinG n0mm t(AI AlW1(sil Of Plt titC 5178 f pillAilf IVPC 10 DPfu 10 (tosC 1501 Alloll Illt (5fC) 51Alus P(84t l6AllOr 5 tlhil 150tAllD (22) C0C l ias[5 LINC (let,) C(mIAIMilhi (6.22) (5.6) (5.6) SIGNAL. (10) (s.9) ItMass s ll SCIC Iwel.ine (al.aust 56 I I S Outside M Gate sc Ot ist ja apen 2 8 Outside Check flew Aeverse flew Severse flew N/A Closed (l}) 3 38 BCIC Vacuus Pump Discharge 56 1 1 2 Ntside MD 5tep Check flew /DC Sev. Flow /DC hev. Flow / M 13 Closed (13.21) t 2 Outside Check flew Reverse flow Severse flow h/A Closed s 19 BCIC Pump Sectlen 56 I I - 6 Outside oc Gate DC aC M 31 Closed s 204.s Ceee Spray P.sgo Olscharge to 55 2 I le. leside VIC flem Aeverse flew Deverse flow N/A Closed (3) { RPy 1 2 Inside M Globe AC AC M le Cleted e ^l le Dietside M Gate AC AC M 43 Closed (Is) j s 25A.D Cese Spray Fings Sectlen 56 2 l i 14 btside IO Gate AC AC M 76 Open s.22A.5 kB(tCW to RectrC. Pump and 57 2 1 4 btside M Cete AC AC get 33 gpeg Ibter Caelers 8 3 234.8 knCICW fram Pecirc. Pungs eed il 2 1 I 4 Outside te Cate AC AC M 23 Open poter Coolees s.244 se H 98CLCW to Osywell Unit Coolers 56 8 1 5 laside Check flew Reverse flew 9everse flew N/A ogen 8 3 Delside le Sete AC AC f.G.I.0ft 16 Open s.257.8 ksCLCW fram trywell Iktt Coolers M I l 4 lhside M0 Gate AC AC f.G.2.M 20 Osen i e I ) 4-Ov! side 10 Gale AC AC I.G.I.M to Os=n 4 a 26 Purge Air to Drywell 56 1 4 g 18 leside - A0 Butterfly AC/Alr Spring L'.m 5 Closed (17) ,1 - le out slJe, A0 Betterfly AC/Alr Spring L.M 5 Closed (17) I.2/ Purge Air frue Drywell M I I l 18 Inside A0 Butlerfly AC/Alr Spring d.M t t

• It outside 40 Butterfly AC/Alr Spring t.M S

Closed (til 5 Cleted (17). h l

e-t i Table B-1 (continued) PROCESS PIPELINES PENETRATING PRIMARY CONTAINMENT l'altiAnt fALVC 10CAlle VAtM Ame/0R lasWR vat W 5 N0 HINE hilAliW 10 Or(A4100 PotER POWR Ct05tNG NonHE t twel Alwt at af MR titt $1H thirAtt IIM 10 OrtN 10 C105C 150LAll0E Illt (5CC) 51Alui ri ul.l.kA l.l_(W5 TINE 5 150LAllD (22) G0C tlN(5 (INC (IN.) CalAlmut (4,22) (5,6) (5.6) SIENAL (le) (e.g) atsmaE5 -Pe Purge Air to Saleresslea Chea6er M i 2 10 Outside A0 Sutterfly AC/Alr Sprlag L,0ft 5 Closed (17) ~ siI 5 Erge Air from Suppression 2 56 1 2 10 Outtlee A0 Butterfly AC/Alr Spring L.RM 5 Closed (17) { Chaes.cr 3-30 Sangle Cootent free BPV 55 l 1 3/4 leside AG Elebe AC/Alr Spring s,C,ast I5 i o ca 8 3/4 outside A0 Globe AC/Alr Spring 8,C,lel 15 Open I r 1-31 (quisment Drains from Drywell M l 2 3 Ousside MS Cate AC AC A,f, Ret 16 Open I-32 fleer Drales from Drywll M 8, 2 4 Outside MS Cate AC ' AC A,f,Rst 16 Open i 3 38 Spar's (15) 'I 34 Spare 3 35 Spare (15) (15) s.35 Stemney t I,.14 Coelaat to SPV 55 1 1 1 l/2 Inside Check flew Reverse flew Reverse fluu N/A Closed 1 I I I/2 eetside Chett lla. Deverse flew Reverse flew N/A Closed 2 1 1/2 Outside Caplesive AC N/A RM. Instantly Closed' s-314 Natrosen/Asr Purge for llP $7 1 1 '3/0 Outslee Cheth flow Peverse flew Reverse flew N/A Open 3 3th,C,0 flP Orive Guide lut,es 57 3 1 3/0 Outside Ball AC 5pring 0.5 Closed (14) l 3/8, OutstJe laplesive N/A SC Rst lastsatly Open (14l Sliest I-33 IIP Drive Guide tut,es 57 l 1 3/0 Outslee Ball AC Spring 0.5 Closed (14) } I 3/4 Outside b lesive N/A OC till lastaally O en (14) l 5 ear I 39A.B lastrument Air to 5mperesslen M 2 1 1 outstde theck flew severse flew severse flew N/A Open Che=&er I I outside le Clele AC AC f,G,aN 5 Closed 3 i I

. Table B-1 continued) L 4. PROCESS PIPELINES PENETRATING PRIMARY CONTAINMENT s' VAsW 10Calles WAtW ANtVOR rumnet smessa vat w1 NOMINAt oftalla le ortmalen route r0wn Ct05tNs wasut itAll Aligg ql CC Pt R PlFC Silt reif34v IvrC 10 DMN 10 Cio5C ISOL Alloll llHC (SEC) 51AIUS s alu tkallem a lw5150t Alf a (22) seC timC5 tlw (lu-) CouiAlmust (6,rr). (5.6) (5.6) 51sNAL (10) (s.s) atMAac5 te Spa,e gis) s.41 wCI Vacia.a steener u I I 2 a.t:Ide se sle6. eC oC r end r. sM 13 open *' 2 it Ntside Chect flew Reverse Flan. legdle.w JA Clgted 2 5 42 DCIC Vacanan treaner M .l i i 1/2 Outside te Glebe DC DC F and I. AM 16 .l t Closa;ld Spes 2 4 entstee Cheth flew Severse flea. Reverse (Iow N/A l I.43 StW 9ellef Valve Distharge M 3 N/A N/A N/A N/A N/A N/A N/A N/A N/A Vuimme Beester. ala h at isthanger Vent. 2-2 Ie amtside le Glebe AC, AC BM le Closed i pim Nat inchenpr. and 2 1 li Butside tellef Valve, high rressure 5pring N/A N/A Cloself le'Cl Ste.a 5 ppuy to Sam i l W at isthanger ,2 3 6* Outside Bellef Valee Nigh Pressura Spring N/A N/A Clos e l 8 44 Contal.a ent Ateespheric Centrol 56 'l I 6 8.tside le Cate AC AC SM 31 Closei free Sunpression Chant.er. and 1 4 A ttlee le Este AC AC SM 20 Closel Dry ell floor Sean Pressurise-57 l I t/2, ihrtside le Globe AC AC BM Open alen 1 s.45 Contale ent Atmospheric Centrol 56 I t 6 Outside te Gate AC AC RM 31 Closei i f ra.m Suppresslen Chead.er, and i 4-Outside le Gate AC AC RM 20 CloseI Dry ell iloor Seal Presserlas. 57 'I I. 1/2 C<,tsIJe PC Globe AC AC RM 6 Open live s/46 Centalancat Atenspheric Centrol 56 1 l 6 laside 96 Sete AC AC BM 31 C1 sed ee-or,en i 6 Omitide m s.t. AC x mi i6 a.ses a-47 Coatalancat Ataesplieric Centrol 56 I I-6 laside e Gate AC AC SM 38 Closed free Dry ell 6 htside te Gate AC AC bel 16 Closed CAD lasera and Withdraw tlaet 55 117 8 3/4 blside Enebe Manuel Manual N/A .N/A Open (20) i 137 I I outside Globe Manual Manual 11 / 4 N/A Open (70) [ } a5 1 Sp.ee (I5) (5 2 Spare l (15) ,l t

r e C E = = = = = = a = = = =

=

= =

==T gt; w. = mEd 2: s=

== Eg-E* 11 w ww w w w w., os ao O 20: --s 8W-2 22 3 2 2 22 C: 8 NW 5 I i EE EE 5-e ss es 2 Ek e E E 22 J.*

• ~-

8 3 a8-NcM Z zoo 3 3

h. *5 w.

A WA W W W 22 WW WW x .x E m-t I o w & *. U feO in 1* II m s E*WE' W W W 02 WW WW m cc E I h: c 5 2 E g t:~. I3 cc II gg e +. m o e,3 g--e

3

-J J 4 t. ss 33 c 8 u z g 2 2: E E E .3 EE EE w s M, Bo 4

l-m W

awg: 5:.E C 44 4 4 o 5 %='f a :: -4 44 44 44 .a 'O m 2" J JJ J J J "J "J "J E a gas. I,,, _ m =~

== m. n== m w ,e m O CC 5 si. Y m 51. l -l, 1. I,.l.l. 3 3 3 .I T 7 a. 2 2 2 1 3 2' 1' 's -e 2 5 -1 2. 2, u E i H 11 si EI II = , 22 a-z- = a: z: y i.il s si si } si 5 sh:

]

15 11

== = 1: I q1 -:aa s x5 ,~ g. 3. I gj

J i

a usi

==-

, -i ~ ? ? ? ? ? .?. .?. T,. .T T .T T = - - e ' - ' - i ' 5 " i, i i a ' " - ~

e Table B-1 (continued) PROCESS PIPELINES PENETRATING PRIMARY CONTAINMENT 5 vat W 10CA110il VatW Ah0/0A ralf9Af lassia Vat W5 20filhAL pitallvt 10 CPikAlet P0ute Nutt CROSING N0klM gy pgg pgpg Sl;g InjstAtt 11N 10 OffN 10 (105C 150t A11011 IIHC (SEC) SIAIUS id 11 45 tild5150sAlle(22) EdC t lkf 5 8 lift (sN-) (UNI Alldll"I I022) (5.6) (5.6) SIGs AL (10) ,(e.g) StetMrsi a ai 16 Spare (IS)

• s5-87 5pers (IS) 31 10 Spare (15) 35 19 Spare (15) a5-20 Centalement Atmospheric Central 56 I

I 6 Inside te Gate AC AC SM 32 Closed to Drywell 8 6 0=tside HD Gate AC AC Set 32 Closed ab.21 Contalament Ataespheric Centret 56 8 3 6 Ir. side H0 Gate AC AC RH 32 Closed to Oeywell I 6 Outside MG & ate AC AC 8 88 32 Clased I 35-22 Centain= rat Vest to selff5 56 'l l 6 laside. A0 Outterfly AC/Alr 5pring 1.aN 5 Closed (17) ~i l 6 Outside A0 Butterfly AC/Afr Spelag t.RM S Closed (17) a5-23 Spare (avseeved for SPV Internal laspectlea) s5-24 Spare (15) (15) 1-25 Spare s5 26 5 pere gi53 ( (15) a5-27 Spare (15) E5 25 Spare (15} e5 29-5 pere (15)- j .5 2a Sp.ee (15) d2 lastrument Air to Drywell 56 I I I 1/2 laside check flew Beverse flow severse flew N/A Open a l I/2 Outside re Globe AC AC 884 5 F.5 Open l a-5 Instrument Air to Drywell 56 I I i 1/2 Instda Cleth flow Neverse flew 8everse flow N/A Osen l I i 1/2

Outsid, 80 Globe AC AC RM 2.5 Open 6-11 Recirc. Pu, Seal lajectlen 55 I

I 3/43 Inside check flew Reverse flew severse flew N/A Open f-Il arcarc. Pua, Seal lajection 1 3/4 Setste. (tect flew Reverse flew Deverse flew N/A. Oswa 55 I l 3/4 Inside (heck flew Beverse flee Reverse flow N/A Open 1 3/4 0stside Check flow Reverse flew severse flow N/A Open a e--e I i

l? 6' ' TableB-1(continued) PROCESS PIPELINES PENETRATING PRIMARY CONTAINMENT i +1 These notes are seyee er nunoer to correscend to nunoers in parentheses. 1. Mata steam isolation valves recutre that tota solenoid pilots be 4eenergtted to c!ase valus. Accumulator air pressure plus spring set togetner close valves unen octa pilots are caenergized. Weltage fattur: at only one pilot will not cause valve closure. The valves are set to fully close te less than 5 seconds. 2. . Containsent spray to drywell and suooression enameer and pMR test Itne return ta sucoressten enameer isolation valves stil nave the cacaotitty to be manually recoenes after automatic closure. This setus util perutt containment spray for hign crywell pressure comettfons and/or swearession water cooltag. Wnen automatic signals are not present, these valves may be openes for test or coerattag conventence. 3. Testaele enect valves are ces19ned for resete ooening =tta zero differential pressure across the valve seat. The valves wl:1 close on reverse flow enn snougn the test settenes may te postttonee for coen. The valves stil ocen anon puso discaarge pressure enceecs reactor pressure even tnouga tne test :=tten may be posittoned for close. 4 This ifne is only needed during maintenance. Service air supply is disconnectee durtag slant operation by acataistrattve control. 5. AC motor ocerated valves reevired for isolation functions are powered from the emergency AC poner tuses. DC operated Isolation valves are powered from sne station batterier. 6. All motor ooerated isolation valves =111 remain in the last posttion woon fatture of valve pe=er. A11 atr.coersted Isolation valves wtl1 close upon air failure. 7. Signal 3 opens, signal c overrides to close. 8. poner coersted valve can De ooened or closes by remote manual switen for ooerattag convenience euring any mese of reactor operation oncept unen automatic signal is present (see note 2). g. mornal status positten of valve (open or closed) is the position during normal power coeration af Oe reactor. 10. The saecified closure rates are as reevired for contairement isolation only.

• '~

11. Soettal air testante enect valves vita a positive closing feature are oesigned for remote testing euring normal oceratton to assure meenanical osereolitty of tne valve etsc. The remote testtag feature stil cause only a partial savement of sne etsc into the flow stream. wita caly a staor . ef fect on flow. coon enetot of an isolation signal, tne actuator spring force util ettrer cause a sitgnt reduction te rio= -nen ene feeewater system is avallatie or cause tne valve ta close. provising a positive cissure esffersattal pressure on tne seates af sc. unee the (cec =atar flew l . ts not availaole. This vein o'111 open =nen boca a law reactor pressure vessel pressure and an acciant signet are 12. Smen t. 13. The motor onrator of trts valve is tay locaed opeg euring normal operating condttions. 14. Tranratag In. Core emne (TIF) Systems When tne TIP systes caole is inserted. the ball valve of tne selected tune ocess automatically se snat tne prone and :aote may aevance. A maatsum of four valves may se ooenee et any cae time ta conouct ce calibration. ano any one gutee tune is used at most. a few nours ser year. If closure of tne Ifne is reevired during calteration, as indicated by a contatammt isolation signal. tne cable is automatically retracted and tne 3411 valve closes automatically after com. pletten of catie witnera=al. To ensure isolation casant11ty, if a TIP caole fails ta =ttnere. or a ball valve fat 1s to close an emolestve, snear valve is installed in esca Ilme. tloon recetst of a remote manual signal. tais enslestre valve will snear tne T!7 caole and seal tne 94400 tune. ts. All unused penetrations (destgnated 'Soare") are capped and sea! weleed. 16. Yalve stil close on system alga flow. 17. ! solation signals A or F 411 initiate tne reactor tutiding stanany ventilation system =nten in turn isolates tne purge air tsolation valves. This valve =111 open =nen Deth a low differential pressure across tne valve ane an acetcent signal 18. are present. , tg. pressure sensors and seistag stese Itne cressure are used for Interlocs control ts prevent taaevertent valve opemag at htgn steam Itne pressures (acove 25 astg). 1 I e .a-m.~ ~ "~

r~ e Table B-1 (continued) PROCESS PIPELINES PENETRATING RRIl%RY CONTAINMENT Notes (Continued) 20. Control Rod Drive (CRO) Insert and Withdraw Cinest Criteria 55 concerns those lines of the reactor coolant pressure boundary penetrating the primary reactor containment. The CR0 insert and withdraw lines are not part of the reactor coolant pressure boundary. The classification of the insert and withdraw Ifnes is Quality Group 8. and therefore designed in accorcance with ASME Section !!! Class 2. The basis to which the CR0 lines are oesigned is commensurate wita the safety importance of isolating these Ifnes. Stace enese lines are vttal to the scram. function, their opersollity is of 6pmost concern. In the oesign of this system. It has been accepted practice to omit automatic valves for isolation purposes as this introeuces a possible failure mechantsa. As a means of providtag positive actuation, manual snutoff valves are used. In the event of a creak on these If nes, the manual valves may be

• 10 sed to ensure isolation. In addition. a ball check valve located in the insert line inside the CR0 is designes to automatically seal this line in the event of a break.

~ 21. This N0 stop check valve is normally in a closed position due to its check valve feature, but its M is. in the open oosttion. The M0 provices a backup to close the valve to provide aeditional hign leak tight integrity. 22. Abbreviations used in table: A0 Air Operated M Motor Operated VTC Pneumatic Testable Check Valve RHR Residual Heat Removal System RPY Reactor Pressure vesset Reactor Core' Isolation Cooling System .RCIC RWCU Heacter Watar Cleanup MPCI Migh Pressure Coolant InjectioS = GDC General Cnsign Criterion RSCLCW Reactor Sullding Closed f. coo Ccoling Water T!P Transversing Incore Pet.be CR0 Control Rod Crtve M5!V Main Steae Isolation Valve 4 8

n s' : i Table B-1 (continued) PROCESS PIPEl.INES PENETRATING PRIMARY CONTAINMENT ISOLATION SIGNAL NOTES SIGNAL' DESCRIPTTCN A' lleactor vessel low water level 3 - (A scram will occur at this level) 8* Reactor vessel low water level 2 - (The reactor core isolation ecoling systen and the nign pressure coolant injection system util te initiates at this level, and recirculation pumos are tripped) C' High radiation main steam line De Line break. main steam line (high steam flow) (* Line break - main steam line,(steam line tunnel hign tesserature) F* Migh drywell pressure G Reactor vessel low water level 1 - (The core spray systems and tae low pressure core injection sece of RHR systems util be initiates at tais level) J' Line break in reactor water cleanuo systee high space tencerature. hign differential flow, hign differential temcerature K' Line break in steam line to/from turoine (hign steam line scace temcerature, hign steam flow. Iow steam line pressure or nign turoine ennaust disonrage pressure) L 2eactor building standby wentilation system initiation M High radiation signal dowistream of primary cantainment purge filter train 0 High assier.t tercerature in main steam tunnel penetrosion arca (MSIPA) P' Law main steam line pressure at inlet to turoine (AUN rece only) R Low condenser vacuum T High temperature in Turbine Su11 ding tl High reactor vessel pressure V' High temperature at outlet of cleanco systes nonregenerative heat escnan;er I Low steam pressure 7 stancey liquid control system actuated Z Low level in R8CLCW head tank AM' 2 emote manual switca from main control room

• These are one isolation functions of the primary containrient and reactor vessel isolation control systant etner functions are given for infomation only, e

l

n, m ,a,4 .a A-A e++-- w m--e=- ,.h* ,,,n,m%_,m. ._a .c i e e l \\ l G G f y w ,c,.}}