ML20028B815

Summary of 820903 Meeting W/Util,Ge,Fauske & Associates,Inc, Bnl,Nus,Science Applications,Inc,Commonwealth of PA, Limerick Ecology Action & Applicant Counsel in Bethesda,Md Re PRA

ML20028B815
Person / Time
Site:	Limerick
Issue date:	11/29/1982
From:	Abelson H Office of Nuclear Reactor Regulation
To:	Office of Nuclear Reactor Regulation
References
NUDOCS 8212060393
Download: ML20028B815 (47)

Text

{{#Wiki_filter:. ,:. + DISTRIBUTION: NOV 2 S 19E See next page VDecket1NostfR5035$3533 APPLICANT: Philadelphia Electric Company FACILITY: Limerick Generating Station

SUMMARY

OF SEPTEMBER 3,1982 PRA MEETING A meeting was h' eld with Philadelphia Electric Company on September 3,1982 at NRC Headquarters in Bethesda, Maryland as part of the technical review of the Limerick Probabilistic Risk Assessment (PRA). A list of attendees and the meeting agendas _are provided in En' closures 1.and 2, respectiv' ly. e Copies of the visual aids used in the PECO presentations are provided j in Enclosure 3. Highlights of the meeting are presented below. R. E. Henry of Fanske & Associates (consultant ~ to PECO) provided a detailed discussion of a revised core melt phenomenology, including: a) The effects of a molten coresslurry in the lower plenum and its relation to vessel failure modes, b) The relation of the stored energy in the core debris to the rate of concrete attack and flow into the pool, c) The layout of the diaphram floor with locations of the floor drains and equipment drains. d) The contribution of the static head of the core debris retained l in the pedestal region to the pressurization of the drywell. e) The particle size of the core debris and its coolability n inside the pool. f) The effectiveness of the pool in spurging the fission products released from the core debris. g) The effect of containment spray on core debris coolability. Dr. Henry stated that, given a vessel failure, major portions of the released core debris would likely go into the suppression pool rather than be retained on the diaphram floor. He stated that the containment response assumed in the Limerick PRA was very conservative. 8212060393 821129 l PDR ADOCK 05000325 A PDR OFFICE) .a.= a m.aa ~ a.. ..~~a..... a..~..= ~* a a a a ~ ~ * = * *"a ~~~~~aanaa** i SURNAME) ........a.am...... .. u...a a m...= " a an.* * *. =

• m.m

........... ~.... ..a ."...a

~a

" a a a *~ a *"a =* "* a.a ** *

• a a a

."an***"a"" DATE) NRC F00M 318 (10-80) NRCM 024o OFFICIAL RECORD COPY um. mie

/+- , The SAI presentation regarding release fractions included: a) tbdeling of the effect of the suppression pool bypass on the source term. b) Treatment of LOCA and in-vessel steam explosion sequences as potential pool bypass paths in the source term analysis. c) Assessment of the impact of the slow leakage of SGTS on the sourse term. d) The impact of the containment spray system on the assessment of decontamination factors for saturated and unsaturated pool l conditions. L e) Treatment of the pool behavior during overpressure containment failure modes concerning the amount and the form of fission l products released to both the containment and then the environment. SAI stated that their treatments of release fractions is very conservative. The SAI presentation on consequence analysis included: a) The effect of the formation of fog on the ground contr.mination. l b) The effect of the size of the fog particles on the internal dose conversion factors and the range of the distance over which radiological consequences occur. I c) The effect of evaporation of the fog on the airborne concentration. d) The potential impact of incorporating a delay time on the evacuation model on the estimated early fatalities. e) The method of combining CRAC results of the five release l categories (each category run on individual basis), using a special i computer code, "RESULTS". i In response to the staff's concerns, SAI stated that: a) The evacuation model used in the Limerick PRA (Rev. 4) did not include delay time in order to be consistent with the WASH-1400 evacuation model. b) For consistency with WASH-1400, latent cancer effects will be received by 10 percent to account for thyroid cancers which were not previously included in the Limerick PRA. c) Estimates of the plant mid-life year population distributions are not used in the Limerick PRA, Rev. 4 and, hence, correspondino CRAC resu)ts are not ava ilabl e. omcs) sunune) oare > unc ronu ais oow nncu ono OFFICIAL RECORD COPY usa m inei-m uo

m o c:. The GE presentation on accident sequences analysis included: The PRA basis for sele ting the 80%-20% allocation of the a) c sequence to Class III (ATWS sequences in an intact containment) and Class IV (ATHS sequences in a failed containment). b) The modeling of human error to reopen the SLC system valve F036 after maintenance or testing activities. c) The modeling of human action to override high turbine exhaust pressure trips of HPCI and RCIC. d) The assumptions used to obtain the unavailability values for the station battery sets, 4KV switchgear, and all diesels (failure to start and run). a There was also a discussion on treatment of dependencies. Dr. E. Burns of SAI stated that the Limerick PRA (like WASH-1400) did not include an evaluation of cut sets of the functional fault trees. However, he indicated that the PRA has modeled and quantified some important dependencies such as AC power, DC power and HVAC dependencies (Appendix I of Limerick PRA, Rev. 4). H.' Abelson, Project Manager Licensing Branch No. 2 Division of Licensing 1 \\ J D..g.L.g: l orext> suame > .1.1../.*...f../. 8. 2........ om> unc ronu ais tio4oi nacu o24o OFFICIAL RECORD COPY usom,ui_m..

r ENCLOSURE 1 LIST OF PARTICIPANTS Name Oraanization E. S. Che11iah DST /NRC F. D. Coffman DST /NRC D. D. Yue DST /NRC B. Hardin DSI/NRC J. Meyer DSI/NRC Z, T. Men Qza SAI K. W. Holtzclaw GE-Safety & Licensing H. Abelson DL/NRC M. J. Wetterhahn Conner & Witterhahn, P.C. R. E. Henry Fauske & Associates. Inc. w,_L. G. Frederick GE S. H. Gibbon PEC0 G. F. Daebler PECO E. T. Burns SAI r A. Hodgdon NRC/0 ELD l W. F. McElroy PEC0 J. A. Dorsey LEA S. Maingir PA. D.E.R. T. Pratt BNL H. Ludewig BNL R. Newton DSI/NRC E. R. Schmidt NUS S. Acharya DSI/NRC , r. 4h I E. Chan OELD/NRC . s /hp,0' ,e>' ~,,,,,

ENCLESURE2 e t LIMERICK PRA REVIEW MEETING SEPTEMBER 3, 1982 sGENDA IIEM TIME INTERVAL APPENDIX H - COREMELT PHENOMENOLOGY 1 - 1 1/2 HR APPENDIX D - RELEASE FRACTIONS 1/2 HR 03.48, 3.53 APPENDIX E - CONSEQUENCE ANALYSIS 1 - 2 HR SECTION 3.0 - ACCIDENT SEQUENCES 1 - 2 HR 0 3.40 THRU 3.47, 0 3.49 THRU 3.52 APPENDIX I - SYSTEMS DEPENDENCIES 1/2 - 1 HR MEETING

SUMMARY

(2PM) 1/2 HR 'l

'ERCLOSURE 3 e EFFECTS OF A SLORRY IN Tile LOWER PLENUM O TIME AND LOCATION OF VESSEL FAILURE ESSENTIALLY UNAFFECTED DUE TO 3-D ATTACK OF Tile PENETRATION. O ABLATIVE ATTACK WOULD BE REDUCED THUS GIVING A SMALLER TOTAL FAILURE SIZE - THIS WOULD INCREASE THE DISCHARGE TIME INTO THE CONTAINMENT. O LESS STORED" ENERGY IN THE DEBRIS - THIS WOULD DECREASE THE CONCRETE ATTACK RATE AND/OR DECREASE THE STEAM PRO-DUCED. ~ O MORE CORIUM WOULD BE. RETAINED WITHIN THE PEDESTAL TO FLOW INTO THE SUPPRESSION POOL AFTER FAILURE OF THE FLOOR AND EQUIPMENT DRAINS.

MATERIAL DRAINAGE INTO Tile SUPPRESSION POOL O FAILURE OF THE CONCRETE DUE TO THERMAL ATTACK OF THE CORIUM WOULD REQUIRE AT LEAST SEVERAL TENS OF MINUTES. O MOST PROBABLE FAILURE LOCATIONS IF MATERIAL IS ACCUMULATED IN Tile PEDESTAL: (1) FLOOR DRAIN CONNECTION, (2) EQUIPMENT DRAIN CONNECTION. O RATE OF ENTRY INTO THE SUPPRESSION POOL WOULD BE LIMITED BY THE STEAM VENTING RATE THROUGH THE PEDESTAL DOORWAY AND PORTHOLES. O PRESSURIZATION OF THE PEDESTAL REGION WOULD BE LIMITED BY Tile DRYWELL PRESSURE PLUS THE STATIC HEAD OF THE CORIUM RETAINED IN THE CAVITY. O FRAGMENTATION OF CORIUM IN, WATER WOULD BE SUFFICIENTLY SMALL TO ENSURE A COMPLETE QUENCHING IN A COMPARATIVELY SHORT LENGTH. O VENTING OF STFAM FROM THE POOL WOULD REQUIRE SUBSTANTIAL BOILUP, THEREBY PROMOTING CIRCULATION BETWEEN THE INTER-NAL PEDESTAL REGION AND THE REMAINDER OF THE SUPPRESSION P00L.

e =_ u -t_.i a = t >.I >. u' iii i< g .d. __, -Q q-r_ 'or >'y 1 i . 7 ._p3., -r. '.. g p lj e _a 4, J. 3 ii ,i ...~m.w . ~.. .=g s'Y -' l r -.l.a.:_;.c' l'a.d ]-] l u..r.en__l e= ..a D. w -l: - f,g. .. g .( s a a Try .g s ni;gT, p 9, i up i, i s. c;r--- - ' -l -._ ~' I g 'g y .t'"L j e

• d' l A ss..w v.y s---- :3

- ]_]= .... j , u n r..- e T ,r...: ass; .. r i -.T..-.

,r.

1 l - .7- .u w ,.,._ l ',C.., l-lj .!h I,D - .M l.

i.t:-

l F'.:..F.. ; [ h .Tm.r_: rh y :j w .;. rg;g c_3-g..... :.,, 1 m e t.s. w

1

..~ ~I'-2' D..J:il=. .\\ 1

,.- * \\

. y:"'p_1.,ig,2,h. J, % A. d ... lid".l .A-S' 1 . _c .m.,. n t, j' 'x 9.--l 5q, \\,,,,, p ,1 - t-y --l(

{3

-*~wa l 9 g-g [_ : ...r l

':tl l

b, l '"* . i l y -. !h_. u.... __'./ .r .I .it s i n,---- ~ .s. p -*- - _,.,c_ _,, _n._ } m.. ;- >.e._.f[L_r_P...; *~.,. :

:::.tr :i:

':l.e :..}:^a.;."

• y:

g i l e.- ( !".-l't.2.~ '

• l i l l..'4

s.s); j l p.M; l i e '- Nl M.;'J;T-El/j NL [3N @gb'h*,.;-- eD j 1. -u cr. , q _ s_. b,i 7. j 5IT G _m .i ' *.6,.,r. - --li.

• 4.--,,1.

i a_.__.2, y. 1 l yl j i l. = j .'M i--- ?. I c

m.,p V

-.v .l I l j iI+: i [ we.8.N, T. 'T l m;..'7, -~ 7, 49*- l y g - _.C%v(i7 't. l l m 3 i . ra i li i i

-'y i y-i

_. ',.,3 f;5 -'! ;lj;< e J [l' s,., (p. I i ' C 'IJIV-O I h 7 g- %-ly*r/ _,, Q,.-

m....r-e p.

l' i.,

• f f *.' '31 t._.l E e;

~~]*_,--l, I 3 i A.: t3- ' p 2,,4 "-a*- l _u _ { i l , g. l i .m p{~' _L _,,.

• L _./.,.,

i , ts ? ' *_ ** l . L y,%,7-y,, r'*- g ._ u c --~4 3 ; l-l'*- N - It$k

• - ]M.]:

1 sectioJ A-A IT a. e a a 1 4 o s

m, % DQ

Q 4'

J g a w l ..f s .. ' A g: . es v,ec rqi 2 *Lc 1!L 6~.. * ,.g.,. . - 9. " * ~.. ..'~....L

• n.;.,.

. g... :c. 7 c... - e r,'

g

• ' 'd "g.,.7.y..,..

,I.. ** * * .p.***r ~ s4,. [* 1., NIN3Yhdh. k.h; $f:t.:-)..~_. htr a.-

• -5j /

y ~ ~= . U I U,~s *

g. o g'

.. ), f A* N I ~ g s ' f *.' Lyr-& -.rIO2I._- W I -I I a.*,.as.p u* *J t E D.:.rc I PM3 > tw< c.fty c. s;g i,g "' ~ }, e.r s r.r t r. r. v.m 1 A. Lsw: *. *> t?. r . g 3* L (*.Ni Vi K M (.s tr.: 0* p A :*4 ? ,KK~f .-[ f; L *.g;.e.} * .,s. e' ' .!.J . s.AL y. .a l

• =

W..?/ e'5f.1" N -.?.

• * ! l.

t t: 1

e..e.t

,( l a ) t ~r jg L f. f ss *u.g ag u n.m 4, w 1 3 se J \\ A s s ,s 4* Fl_ddR ' DRAIN CONNECTl'ON DETAILn,. PED Y El.(.

9. -

'4 Fic.'5-1 9, s ! e\\ s ~ k { f. s ,q 'hk

• \\

_ ) t, s. g s ~, g a e 4 mg y L, 4 x t A [, + t.'$ 1 i u % (\\ t 5 L* a f 4-g 'g i 8 3_ 4 i f . 'g - 3. y, g x 4 1 k i t s vs } '{ t ,2

• 4-g

\\ j t y. k ( N' ~ \\ d A s x.* - g-\\ 1:l ,.M.TM t p i lit 4* n P. / a hl-- tt p.* h t rs.=. cxT^.'An ad +

: Ni v.,-fMI:. -

2 k

r. o c a #-e. t :ert.

U e u e-t w rot n r r et a

• .sq' a-rar: ro..r:>r s _..

-- y . ~ I, 3

op} rj r ::f N *

~ N t; u.:rt r ce.3 <,o, l . n.cr3,s..a = ~ .=x..,, , 73..= x.=, - a g...,3.ui, w- ~ y .v, 5 p:c. e...' o . Ety. l / t

z..

.-. /-<l; a \\ .. :... : es..,.... 3 l ;* \\ v.:: rL: n v.rst> _s t s. t i + & EQUIPMENT DRAIN CONNECTION DETAIL '2 { 8;3 SCALE '$f i O a Fic. 5-2 h \\ r s ,\\ .i .s e i k

• I

,\\ ' '

i e PEDESTAL PRESSURIZATION Material Discharge Rate O Uf F DW + O gh - Pp= P F Vaporization Auh 4 = MF D pcF T=pAuh79 gg fg AU = ggg Steam Velocity 2(Pp - P;q;9) u_. y g p 9 9 9 DW + O gh - Pp= Pp - P P ;q P F p gq G k

CORIUM BREAKUP i 2(PDW - P ) p

V 29h = 4,5 m/sec Up

p 4 We = M U d f F = 12 0 I ) d= 12_U 12(0,5) .= 2(PDW - P ) 1000(4,5) p f M T F l d = 0.0013 m = 1,3 mm l l l G

DEBRIS ACCUMULATION WITHIN Tile' PEDESTAL O AT 1% OF NOMINAL POWER, ALL OF THE DEBRIS WOULD BE COOL- ~ ABLE IF THE PARTICLE SIZE WAS 11 MM OR LARGER (1-D, COUNTERFLOW BED). e 0 AT 1% OF NOMINAL POWER, 50% OF THE DEBRIS WOULD BE COOL-ABLE IF THE PARTICLE SIZE Mds 1 MM OR LARGER (1-D, COUNTERFLOW BED). 0 FOR A TWO-DIMENSIONAL BED, ALL OF THE DEBRIS WOULD BE C00LABLE IF THE PARTICLE SIZE WAS ABOUT 600 fiM. O FOR AN INTACT CONTAINMENT, DEBRIS BED DRYOUT COULD ONLY OCCUR IN THE PRESENCE OF AN OVERLYING WATER POOL. (1) THUS ONLY A PORTION OF THE BED COULD BE DRIED

OUT, (2)

CONCRETE ATTACK AND MELTING, AS WELL AS THE RE-LEASE OF NONCONDENSABLE GASES WOULD TEND TO RECONFIGURE THE BED INTO LARGER PARTICLES, THUS TEND TO MAKE THE BED MORE C00LABLE. 0 OVERLYING WATER POOL WOULD BE VERY EFFECTIVE IN SPARGING THE FISSION PRODUCTS RELEASED FROM THE DEBRIS.

FILM BEllAVIOR-FILM TilICKNESS Y cV _ (1-a) cd 6 *.A.f _ (1-a) 2 ~6(1-c) s N4nrP L SURFACE llEAT FLUX 0 Od q/A)P N4xr 2 ~ 6(~l-c)~AE P i FILM TEMPERATURE DROP t q/A)p.6 ^T7 = -- g f i l:- cd2 1 0/A)B (1-a)- AT = 36-k (1-c)2L g 4 STABLE BUBBLE AT = 4aTsat h 0 6 79 g l l 1

,. o Sicam Flow s -Wat er Film L m -P ar ticl e s /,l\\ (W %. J, d _T i e > a.) D r a g o n a l_i q ui d Film ~ L - +1 d h-M Steam 6;lO Water t \\ l l j [ O Particles V U S t e arn Flo w b) Hydrodynamic Limitation ~ Above the Bod Y t

l i P A R.Tl C L E B E D D R Y O U T H E A T F L'U X LlOUlD FLUIDlZATION 4 ~ (9, - Po) 4 / 3 Y r 9 = C y r" "2 p o a y-c 3)d U"2 4/.3, = Do Po l l I MEAT FLUX i q'/ A = My U h,o, a ( } ,'O / A = p, h'o 4/3 a p, - o q / A = h,,)p} AIS 5 Or d C o l

I PARTICLE BED DRYOUT PAODELS WATER I i 10' O BARLEON & WERLE HYDRODYNAtAIC D SQUARER ET. AL. STABILITY A TRENBERTA & STEVENS E } O ANL DATA x >i O 3 u. D I' COfAPLETE O b 10' SOLUTIONS I a 3 LifAITATION O ^ WITillN Tile DED y LAIAINAR c. o ASYlAPTOTE 8 8 [#8, ~ ~ ~~ ~ ~ ' O.1 1.0 10 100 g PARTICLE DIA?AETER, mm 1.

EFFECT OF CONTAINMENT SPRAVS ON DEBRIS'C00 LABILITY 0 DRYWELL SPRAYS COULD POTENTIALLY COOL ALL THE DEBRIS IF IT IS DISTRIBUTED ON THE C0"TA!"ME!'T FLOOR. tt#PM N O IF THE DEBRIS REMAINS IN THE PEDESTAL, ACTIVATION OF DRY-WELL SPRAYS WILL LIKELY NOT PREVENT FAILURE OF Tile FLOOR AND EQUIPMENT DRAINS, BUT WILL RETAIN.MORE DEBRIS ON T!!E. ~ CONTAINMENT FLOOR, ~ pp p fRR(rH.

+ CONCLUSIONS h t O ASSESSMENT FOR MELT PROGRESSION AND CONTAltlMENT FAILURE I IN THE L'IMERICK STUDY WERE CONSERVATIVE. O MIGRATION OF DEBRIS INTO THE SUPPRESSION POOL WOULD EX-TEND TIMES TO CONTAINMENT AND INCREASE FISSION PROD' JCT RETENTION. L dW- "W M* W w aw. O ACCIDENT PROGRESSION COULD POTENTIALLY'BE T$RMINATED ilY ACTIVATION.OF THE DRYWELL SPRAYS. n-, - --a op.~w. ---ee -.,--ac,,--..- n, -- ~ , ~~ p a-- n--- r-

H. 08-(d) Provide a discussion of the potential effects of containment sprays on the accident consequences for the following cases: (1) Prior to diaphragm f ailure (2) Af ter diaphragm f ailure Discuss the ef fects of pool temperature ranging f rom near room t emperature to saturation on the DF and the failure modes.

RESPONSE

- No credit taken for containment sprays in LGS PRA Sprays would provide DF of 100-1000 for airborne contaminants Pool temperature effects on pool DF 6 - Unsaturated Pool - DF >10 Saturated Pool - DF > 10 (100 and 10 used in LGS PRA) t 88 - For overpressure failures ( Y,Y Y ), 3 Containment sprays would unintain containment integrity

PRA D.08' QUESTION CALCULATIONS WITH B0ll INDICATE THAT FOR CERTAIN ACCIDENT SEQUENCES A SIGNIFICANT QUANTITY OF THE GASES FORMED DURING CORE HEATUP AND MELTING MAY NOT ACTUALLY BE RELEASED FROM THE PRIMARY SYSTEM VIA SAFETY RELIEF DISCHARGE INTO THE SUPPRESSION POOL. CONSEQUENTLY, SIGNIFICANT GUANTITIES OF THE FISSION PRO-DUCTS (INCLUDING VOLATILE SPECIES ASSOCIATED WITH THE MELT RE-LEASE COULD BE DISCHARGED FROM THE PRIMARY SYSTEM AFTER VESSEL FAILURE DIRECTLY INTO THE DRYWELL. BASED UPON THE ABOVE CONSI-DERATION SHOULD THE DF VALUES IN THE TABLE ON PAGE D-9 BE APPLIED TO THE ENTIRE MELT RELEASE? IDENTIFY ALL OTHER POTENTIAL SUPP-RESSION POOL BY-PASS PATHS, INCLUDING LOSS OF SRV DISCHARGE LINE, THAT HAVE BEEN CONSIDERED AND STATE YOUR CONCLUSIONS RhGARDING THEM. RESPONS$ 0 SUPPRESSION POOL BYPASS WAS EXPLI.CITLY MODELED IN THE SOURCE TERM ANALYSIS FOR ACCIDENT CLASSES 1 THROUGH IV'. 0 OTHER POTENTIAL BYPASS PATHS WOULD INCLUDE LOCA SEQUbNCbS AND IN-VESSEL STEAM EXPLOSION SEQUENCES. O SUPPRESSION POOL BYPASS OF ACCIDENT SEQUENCES WHERE THE CONTAINMENT IS INTACT DURING MELTDOWN IS NOT EXPdCTdD TO IMPACT THE CONSEQUENCES.

PRA 3.48 QUESTION WHhT CAPACITY WAS ASSUMED FOR THE SGTS? THE SAR (P. 6.5-3) STATES THAT THE SGTS FANS ARE SIZED FOR 3000 CFM EACH, BUT THE FILTER TRAIN IS SIZED FOR 11,000 CFM. WHAT OPERATOR ACTION IS REQUIRED TO PLACE BOTH SGTS EXHAUST FANS IN OPERATION GIVEN ACCUMULATION ON THE FILTER?

RESPONSE

O SGTS FILTER TRAIN CAPACITY OF 11000 CFM WAS ASSUMED. O CRITERION FOR SECONDARY CONTAINMENT INTEGRITY GIVEN A SMALL CONTAINMENT BREAK WOULD BE AFFECTED BY THE ASSUMED SGTS FLOW RATE. O SOURCE TERM AND, CONSEQUENCE ANALYSES FOR SLOW LEAKAGE CASES ARE NOT EXPECTED TO IMPACT THE RISK.

l PRA 3.53 QUE5TI N ~ QUESTION 3.04 ADDRESSED A POSSIBLE RELEASE MECHANISM DUE TO FLASHING OF THE SATURATED POOL AT AN OVERPRESSURE CONTAIUMENT FAILURE EVENT. WHAT FRACTION.0F THE SUPRESSION POOL FLASHES FOR EACH OF THE RELEVANT CONTAINMENT FAILURE MODES? WOULD PARTICULATES AS WELL AS SOLUBLES BE RELEASED DURING THIS TIME? RE5P60S O RE-RELEASE OF FISSION PRODUCTS UPON OVERPRESSURE CONTAINMENT FAILURE EVENT IS APPLICABLE TO CLASS I AND III ACCIDENT SEQUENCES. 0 15% OF RETAINED FISSION PRODUCTS DISCHARGED THROUGH THE SRV'S DURING MELTDOWN IS RELEASED BY SUPPRESSION POOL FLASHING. O PARTICULATES AS WELL AS SOLUBLES ARE CONSERVATIVELY ASSUMED TO BE RE-RELEASED. l

PRA$-08' QUESTION THE CONCLUSION THAT FORMATION OF F0G WOULD SIGNIF.ICANTLY REDUCE THE CONSEQUENCES OF A GIVEN RADI0 ACTIVE AT[,10SPERIC RELEASE AS STATED IN APPLICANT'S RESPONSE TO THE STAFF QUESTION E 02 NEEDS FURTHER CLARIFICATION AS WOULD BE APPARENT FROM THE FOLLOWING CONSIDERATIONS: WHILE, BY APPLICANT'S REASONING, FORMATION OF FOG MAY REDUCE ~ THE RANGE OF THE DISTANCE OVER WHICH RADIOLOGICAL.CONSEQUEN-CES WOULD OCCUR, THE RADIOLOGICAL DOSE WITHIN THE REDUCED RANGE WOULD LIKELY BE HIGHER DUE TO HIGHER GROUND CONCEN-TRATION FROM ENHANCED FALLOUT OF RADIONUCLIDES IN THE FOG PARTICLES. ADDITIONALLY, THE INTERNAL DOSE CONVERSION FACTORS MAY BE HIGHER DUE TO INHALATION OF MOIST AND LARGE-SIZED FOG PARTICLES CONTAINING THE RADIONUCLIDES. THEREFORE, POSSIBLE REDUCTION IN THE SPATIAL RANGE OVER WHICH EARLY HEALTH EFFECTS WOULD OCCUR MAY BE PARTI ALLY OFFSET BY IN-CREASES IN THE INTENSITY OF THESE CONSEQUENCES WITHIN THIS i REDUCED SPATIAL RANGE. FURTHER, EVAPORATION OF THE FOG WOULD l RESTORE THE STILL AIRBORNE PARTICULATES BACK TO THE FINES l WHICH WOULD THEN BE TRANSPORTED TO FARTHER-0UT REGIONS; THUS, STRETCHING THE SPATIAL RANGE OF IMPACT.

i

RESPONSE

~ 0 INCREASE IN GROUND CONTAMI!1ATION LEVELS WITHIN A REDUCED SPATIAL RANGE. O DECREASE IN AIRBORNE CONCENTRATION. OVER A WIDER SPATI AL RANGE. o COMPETING EFFECTS ON CONSEQUENCES COULD RESULT IN A REDUCED RISK DUE TO F0G FORMATION. l 1 ?

PRA E.09 Appiicant's response to the staff 'Jues t i on E.04 is ,not adequate. fhe appl ican t's response states that the CCDF of ea rl y fatality in Figure 4.13 of Rev. 3 to PRA shows the results of a 10-m i l e vs. 25-mile radius evacuat ion zone. However, the applicant has not used a del ay time before evacuation to produce the results in Figure 4.13, while the applicant's response to thd staff Question E.08(c) states tiiat the es tima ted delay time is about 5 hours for this site. Provide an assessment of the effect of the del ay time (Range of 2 to 5 hours) on the CCCF's of early f atal i ties, for both 10-mil e and 25-mi 1 e r adius evacuation zones. Perhaps the current status of emergency planning for Evacuation of the Limerick site is still in a preliminary s t a:J e and i t; may become possible that the anticipated delay time before emergency p1anning wi11 full y nature. However, the staff will review the sensitivity of early fatality predictions r epr e sen t ed by CC0F's to the delay time bef ore evacuat ion.

RESPONSE

Incorporation of a celay time in producing the results shoun on Figure 4.13 would be expected to increase the early fatalities CCOF for both the 10 and 25 mil e evacuati on zone. As stated in our response to question E.04 the anal ysi s assumpt ions used in th e PRA where purposely Selected to provide as d i r ec t a compar ison of LG5 wi th the WASH-1400 anal ys is as possible. Inclusion of delays in the present anal ys i s, which used the WAS4-1400 evacuation model, would produce incorrect and misleading resul ts s i nce the effects of evacuation delays were indirectly included in the UASH-1400 ev acuation model. E.10-(a) Juspify the use of an input of thyroid cancer fatality 10 per million thyroid pegson rem, instead of the WASil-1400 value of 1.34x10 per million thyroid rem.

RESPONSE

At the time of the Reactor Safety Study, WASil-1400, the CRAC code did not provide the capability of including the effect of thyroid cancers in the latent cancer eval-untion. Informa tion availabic at the time of the LGS-PRA indicated that the effect of thyroid cancers were not included in the WASil-1400 results. Since the CRAC code usep in the LCS-PRA includes a thyroid model, a value of 10 was used to climinate the effect of thyroid cancer and provide direct comparison to WASII-1400. It has been subsequently discovered that the WASII-1400 results for latent fatalities were increased by 10% to account for the lack of a thyroid cancer analysis. This factor has not been added into the Limerick analysis and should be added to the latent cancer effects for Limerick to provide direct comparison to WASil-1400 results. .A l s l l I

• 6P E.10-(b)

Provide the results of CRAC runs varying the year during which the meterology was taken. Limerick release categories are evaluated for meteorologies for the years 1972, 1973, 1974, 1975 and 1976; whereas, for comparison the WASil-1400 release categories BWR 1 through BWR 5 are evaluated at the Limerick site only for the year 1975. 4

RESPONSE

The question contains an erroneous ~ statement. The evaluation of the WASil 1400 BWR at the Limerick site was based on all five years of data"(averaged), as are all other results in the LGS-PRA final report. i o b i r r I l l 3 . -.. -.. - - -,,.. - ~. -.. -.,,..... -, -, -., -. -,. -. -. -, -. - -. -

s E. 10-(c) Provide the results of CRAC runs varying the eleva tions of meteorological data (30 f t and 175 f t).

RESPONSE

In the LCS-PRA, 30 f t, meteorological data was used in the CRAC runs for the C4 y, C4 y' and C4 y cases, and 175 f t. meteorological data was used for the OPREL and OXRE cases. The basis for this choice was the energy of the release - low for C4 y, C4 y'and C4 y*, and high for OPREL and OXRE. CRAC runs varying the elevation of meteorological data were not made in the LGS-PRA. s l ~ 4 "i l i I l l 1

E.10-(d) Describe the method used to combine the results from the release categories done outside the CRAC code.

RESPONSE

Each of the five accident sequences was run using CRAC for each year (72-76) on an individual basis, and upon completion of each run a formatted version of TAPE 10 file was saved onto a pennanent file. A separate program (RESULTS) was then executed which reads as input the individual CRAC files and a table of accident frequencies i derived from Table 3.5.14 to produce the total CCDF. The algorithm used in results is the same as found in CRAC routine store, entry FSUM. The program RESULTS has been verified to perform the required calculations correctly and will reproduce CRAC results to 6 significant figures.

• To be revised hy Revision 5 to the LGS PRA.

'a i 1 e 4 - O e i e TABLE 3.6.5 . RADIONUCLICE RELEASE PARAugiffi$ AND RELEASE FRACTION $ FOR OWt!hANT ACCIDEhi $EQUENCE CLA35E5 A%D ComiAlwtE4T FAfLURE fiCCES LISTED T4 TA9th 1.5.14 Elaglag C0gTAlwetti UI IhI 1:ME Of 4 OURATIon 11ME FOR ELEVATION EMERGT CRAC $r 'I Ra I III tat Re *I 1( eCM 1 f e #I I Cs 'I I I In8df RittASE OF PELEA$E EVACUATIDE OF RELEASE DELEASE 2 3 GAAP SEQUE4CE ("e) (Mr) (Hr) (feet) (106 3?J/ur) j cor i St eae C,,C, 2.0 0.s t.0 e2 > >0 n.0

0. 0 0...

0.s0 e.03 0.30 3.0.ir3 esplee see C 39.0 0.5 8.0 82 130 1.0 0.096 0 16 0.40 0.01 0.40 2.0=10"3 2 and C 2.0 0.5 1.5 82 130 1.0 0.096 0.30 0.40 0.01 0.60 2.0m10-3 4 bydrosas f.b' Belease parameters emed for cot; esplestem C,C,C.C [ 4.0

0. S 1.0 82 I?D 1.0 0.20 0.06 0.90 0.0a7 0.40 1.0 10*I l g 2 3 g m.

v.,

set.aoe,.rs.eter.. d fe, 0 erte ~I 3.0 min ' l Drywell C l7.0 2.0 6.8 82 170 1.0 0.11 0.04

0. 01 6 0.0I 0 1 enin j

3 ~3

• 3~

and eetwell C 37.0 2.0 T.0 82 1.0 1.0 0.06 0.023 0.4 4.3 10 0.069 4.7s10 3

• I 4.6 10*I 9.1 a10 everpressere C

7. 0 2.0 4.0 82 120 1.0 0.06 0.026 0.073 2.7m10 3

f a tture r toelod tas C.C 7.0 2.0 6.0 82 120 1.0 0.064 0.073 0.313 - 6.4 10*I 0.019 1.3 30 g g bydrogee C 37.0 2.0 70 82 1.0 set calculated, he smaller thee Cef g,, 2 C,iY Y pelease parameters used for C61: i IV f a11ste C l1. 5 2.0 1.0 82 4.0 1.0 0.261 0.2 02 0.4 14 0.029 0.09% S.23:10 l 4 prie, t. sett i I ( ** i' 9elease parameters used for C&#; Wetwell C l 1.5 2.0 1.0 42 1.0 1.0 0.07 0.09 0.20 0.016 0.084 6.0=1(# l g f.u.re prior to melt i CA V" v" Benease parameters used for (4f*g Lees of C l1.5 2.0 1.0 0 1.0 1.0 0.13 0.70 0.35 0.09 0.12 7.On10"I l 4 ,,re s.. pool prter to melt 4 1 Eee see v" tess of C.C 7.0 2.0 4.0 0 120 Not calculated, be mailer th C.,' g 3 Peel C 37.0 2.0 7.0 0 1.0 1.0 0.13 0.17 0.S 0.02 0.c4 6.2310*I 3 I leabs-tr. Ar se1 CTS C,C 7.0 2.0 4.0 42 120 Net calcolated, tvt ses11er thee Cs,*Et,' C2, C4 g g C '# 37.0 2.0

7. 0 42 1.0 0.7) 1.9:10'I 9.0=10 4,,,g,-2 g,g,g,-1 g,g,-3 3,,,g,-a

-2 2 & t i.

sai_,

with Scis C.C 7.0 2.0 6.0 42 120 Net cale=1sted, but mes slet thee t.f,C2 C4 g 3 C '# 37.0 2.0 7.0 82 1.0 0.73 2.7a10*I 0.8m10 g, g,3,-4 1. 6=10*I 3.3a10 S.t:10 -5

• I 2 4 (e) Inclada. Re, tr (g) Instedes la, T, fr. pb, Co. Pr 54, p Po, So, Eo, Pu p

(b) lastedeo I (elemental), le (h) The five eases see la CRAC (c) taalvdes Co, ab (t) Time gree oestdent taittanton (d) lacledes 1e, se, Sh (3) Sensible heat rate few plume else fe) tecledes Sr. Sa (f) tecludes to. Rh, 74 Ao, Te

DI! A f_2.12 P.esponse to the staff Question E.0B(a) is incomplete in tnat immographic data in C F' A C format 'for the p1 ant midiif e year, is not provided. Provide an estimate of s i t e derro jr aphi c i fit or mat i on ,~ for the midtife year. PESPONSE Est ir a tes of 's i te demograph ic information are provideti in the LGS-ERCL Section 2 1.2. Es tin ates of future populatico di str i bu t ions where not used in the anal y si s. Specific inpu t:; to CR AC were not generated. 4 l PRA 3A0 OUESTION WHAT IS THE PROBABILITY OF FAILURE OF FAST TRANSFER ON NON-VITAL BUS POl1ER FROM THE UNIT AUXILIARY TRANSFORMER TO THE STATION STARTUP BUS? HAS THIS FAILURE BEEN ACCOUtTED FOR? [SPONSE O FAILURE PROBABILITY OF FAST TRANSFORMER SWITCH IS SMALL (IN 4 RANGEOF10-3TO 10 ) RELATI'vE TO OTHER NON-VITAL SYSTEM (FEEDWATER) FAILURES. 0 OPERATING EXPERIENCE DATA FOR LOSS OF FEEDWATER EVENTS ALREADY INCLUDES SUCH POSSIBILITIES BUT IS DOMINATED BY OTHER SYSTEM FAILURES.

PRA 3.42 QUESTION ON P. 3-78, STATEMENTS ARE MADE ON THE CONTRIBUTION OF CERTAIN SEQUENCES TO BE DEGRADED CORE CONDITION CLASS Ill AND CLASS lV. PROVIDE FURTHER DISCUSSION ON THE PROCESS USED IN DETERMINING THE CONTRIBUTION TO EACH CLASS. PROVIDE THE BASIS FOR SELECT-ING THE 80-20% ALLOCATION OF CLASS Ill AND IV WHICH IS REPORTED IN ALL OF THE ATWS EVENT TREES.

RESPONSE

6 ATWS SEQUENCES WITH NO POISON INJECTION MAY RESULT IN CONTAINMENT FAILURE DUE TO OVERPRESSURIZATION IF COOLANT INJECTION CONTINUES. 8 BOTH HPCI AND RCIC HAVE PROTECTIVE TRIP SETPOINTS WHICH WILL PREVENT THIS FROM OCCURRING. 8 HOWEVER, SINCE CLASS IV CONSEQUENCES ARE MORE SEVERE THAN CLASS 111, A CONSERVATIVE ESTIMATE OF THE CONTRI-BUTION TO CLASS IV FROM THEIR SEQUENCES IS INCLUDED. o DOMINANT FAILURE MODES LEADING TO A CLASS IV CONTRI-BUTION INCLUDE: e OPERATOR PROBABILITY OF OVERRIDING THE HPCI INTERLOCK WITHOUT A WRITTEN PROCEDURE. e CONTAINMENT LEAKAGE SO LARGE AS TO LIMIT CONTAIN-MENT PRESSURE BELOW 65 PSIA - MAKING THE CONTAINMENT EXTREMELY LEAKY (I.E. EFFECTIVELY OPEN TO THE ATMOS-PHERE). e FAILURE OF THE HIGH PRESSURE TURBINE EXHAUST TRIPS.

PRA 3.43 - QUESTION ON P'. 3-78, A PARAGRAPH WAS DEVOTED TO ADDRESS THE OPERATION OF THE HPCI SYSTEM AND ITS RELATION TO THE CLASSIFICATION OF SEQUENCE. IN THE ATWS DEGRADED CORE CLASSES FOR THE T C Cpg 2 FEEDWATER EVENT TREE, (P. 3-78) THE UNAVAILABILITY OF THE HPCI SYSTEM IS NOT CONSIDERED. P,ROVIDE THE RATIONALE FOR INCLUDING HPCI WHEN THE SEQUENCE IS EVALUATED FOR ACCIDENT CLASSES BUT NOT INCLUDING,IT IN THE RELEVANT EVENT TREE.

RESPONSE

O ATWS TREES REVISED BASED UPON DELETION OF BRIDGE TREE. O METHOD CHOSEN FOR DISPLAY OF INFORMATION IS A FOOTNOTE AND TEXT RATHER THAN A MODIFICATION TO THE EVENT TREE. O INFORMATION PRE'SENTED IS EQUIVALENT FOR EITHER METHOD OF PRESENTATION.

PRA 3 44 The Linerick PRA, Rev. 4 assumes a single val ue of 0.5 for the spli t batween containment leakage failure and overpressurization failure in the containment event tr ee. Perf orm a par ametric study to determi ne the*scositivity of the accident consequences (risk) to different values of containment l eak age /c on ta i n ren t overpressure split

fraction, e.g.,

0.1 and 0.9 RESP 0dSE The reques ted paranetr ic stu:1y was performed. The pa rame ter delta was assigned a value of .9 (that no containment leak suffici'nt to prevent e overpr essur i zation occurs) in figure 3.5.6a on page 3-114 of the' PRA ( Vol u rr e 1). The containment f ailur e mode spl i ts for Class I, II, and III core mel t sequences were then recalculated. Using the existing release category groupings, ano appl y i ng' the recalculated splits resulted in an increase in the OPREL frequency from 6.93E-6 to 1.246E-5 (a factor of 1.785). The resul ts of a re-execution of the OPRE1. CRAC case shows that the total risk for latents and cost where increased by factors of 1.71 anel 1.75 r espect ivel y. Acute fatalities 'were not affected as OPREL has no contribtulon to the acute fatality CCDF. . 9

pol 504 4NJF CT'C'4 COOkA wi iN JE Cf f 04 IN ADy(R1[h? QPt* AllON Ng&TAgUQy&( L'

• • 0 E'V &f t 405 %Qf' im Qu wste CONTRCleocs t eso pa gegg gggggy gas e 4CTV4fl0 DOtlhof 80T*8 CNI

'8 08 40'Lti v V, 'g','p@ t ell 10 Tustg e($SvuI Y bVII QR DCtC AUTou &T-CghtlNyg nuns mHm If0VfwCf tgyg b leg a ag'C(Q33 SLC nas3taf LC CONitQb g agag7pm CQag pyg 8CALLY 70 auw YI4RI C0408780wt iC e t,2,1 3 iy y g@ y 9, y, C C C t v t, Cg og t,C,wg ou ,3, l l'" f,Cu#,g tJet0*' CLk3S h*v' ~,y e,'o-' CL Att lV 7 * 'O 1,C,v, 3 2='O .7 C/ 3,ye10~' ClasseV I l. 7 r,c u if) }j j '~ ' / f ,/ w s,r.,u 4 l'8 Y CwV8 ; 1J :50 CL A$$ H@Y ' l p 1 f { Y C Wu blGWGIOkt CLA38tv pg N j Tyfy 9 kgGbtGlgkl CLASSIV V ' / r i e.te _ # 7, g,w, _ g,,ig-7 tsaas us b, T ?gP 95 y e i u 4 CLAlf fitAW' l L Y,tgtwg 3 9,1g r _[*,.____ T,9yN ggggggigtl CLA3B lv I ~ y 28 T1 T,CgPg ggg(IGt3tf CLA W lV / j -' a - 0 T C py 3Jegg*f CLAN:ll l - _ _. ~ _... _ _. - - yy ICW IJ e10*I CLASBlv E8 pu

==r o ,w Y Cw gy og C p \\ F ub3*3 WCW WM " 0.1 ~ l'" T,C Cu 12*12 9.1e90 CLASSIMAv* / l jp /T .4 !.I-ICC U wgd'Ligistg 'CLASSEY TJee90** F g l2, , te90*3 T,C Cgg Nf0Lf 0is%L 1, CLASSIV 4 ,p D y l f.,_t i.- l 40?1 ' L

j Y C C,yu NEGUCf3Lt

i. CLAS$lil

,j "e,,- py ~ T,C c,g P y OE j-j j a T,C C,g $r2 NTCLlofttt CL AES l'tAv* l. ..f ~ c #A r l u 4 ) - s ~ I' W - f/,C Py ht0Lt010Lt CLASSfY / g gg g t.l s.19~3

3.,g-3 1,t C,y*O weguciety Class V j'

,3 u s.e en ,,c,ci2N Ntot'Cl*La Ctass iti d e' to _w 7,C Cq Nf CLIGISLt CLASSIV s. u i.- i _ 4 T,C C,il o.4,,-* Ct.ss W /'a g ^ 1 -W - __T,C cA n.b uo-6 cLus ing f l ~_ ALL soof ts Amt sat ?Asts 3 e 2 t-p s j 1 ), A 1 c r , ~ Figure 3.4.9b Event Tree Diagram of For,tulate'd ATilS'Alcident Sequences Following rp MSIV [I _ C 22 ~ Closure Initiator v a: i t 2 + y ~ Li ' 'e . -. ~ - r 1 - r

s FPA 3.46 q s QlfESTION j K IN THE MAY 14, 1982 MEETING WITH PECO, A STATEMENT WA3 MADE \\ TbTHEEFFECTTHATINT'HEQUANTIFICATIONOFTHESLCSYSTEM ?. ? '(FAULT TREE, CONSIDERATION WAS GIVEN TO THE POSSIBILITY OF AN i c0PERATOR FAILURE TO REOPEN VALVE F036 (P. B-66) AFTER MAIN-

I TENANCE OR TESTING OF THE SYSTEM.' REVIEW OF THE SLC FAULT

h,'

%i TREE INDICATES A PARTICULAR VALVE FOR UNAVAILABILITY. PROVIDE A DISCUSSION ON HOW HUMAN ERROR IS INCLUDED IN THIS VALUE.

RESPONSE

s 1

0 SYSTEM FAULT TREES CARRIED DOWN TO COMPONENT LEVEL. O FAILURE MODES INCLUDED IN THE COMPONENT FAILURE PROB- \\ ABILITY ARE DEFINED IN THE GENERIC C0f PONENT FAULT is TREES. t O REFERENCE THE GENERIC FAULT TREE B.9.3 (SEE ATTACHED). I O LER DATA FROM EG&G INCLUDES THE FAILURE MODES IN THE (3 GENERIC FAULT TREE, SPECIFICALLY PERSONNEL ERRORS i

REFERENCE:

NUREG/CR-1363-VOLUME 1. l- \\' i l l L

a 1 i t l e S A S = I-4 4 a ,J u a = 4 = cL = b "._*3=e eo W we d 3,,8 W 2";38g3- .E. -2 m Q ~n Y uw ................Es .. < _a ....d m u

)

3 .g_ m m ;;, c Si = ?. 2 -=. - m. 2 < s. -J 3 .3.a >- m i Ea=r=y-3 4 a n a =- >5-ou ~ unno E I J N Bt ~ ~ ~ m EE ~ a. G l 4 .= - w.c. a y,e m .a. E= a. a 4 .1

s. *$

g s. .m s_ _a (( o oo

st e

= w u p, = ?. 2 e. a u g u= t. .O m >- v c 3_ am w ~ }:: c= E a o c-a h- .. > = t _= = - g; wm x 3 3 3 5 =1 *. i 5_5 ~ 55 3 a w l 3. 1 _m.1 o 2 2 m. i, e a a := =a =_ .9 g 8 w q i + 3-75 Rev. 3, 4/82 "~ .+-4.,

7'- [3 ' 3.47 What are the high press.ure set points for the HPCI and RCIC turbine exhaust trips? What is the error tolerance on these set points? Once the high exhaust pressure set point is l reached, what are the required operator actions? Are there written procedures covering operator action to override this trip?

RESPONSE

The high turbine exhaust pressure setpoints for HPCI and RCIC are: HPCI - 65 psia RCIC - 25 psia The error tolerance on these set-points is not known at the present time, and was not used (or needed) for the PRA. The operator is not required to override the HPCI or RCIC high exhaust pressure trips. t There are no written procedures instructing the operator to override the high exhaust pressure trips. I i 4 l I

-g = r \\ i

Mb.1

[ . n; <t$ s e. g

' " 11 M2 1-I 1F g.

Go d 5e M*y*- ,. f :. E,g- -E$ abn N I _2 gd 3=%,5 -6= l

R 3.--

gsr c l f o p *e :.9 0 a a s:t -s ro g: .) l sL m m IPC tE - 5 -bil =i 88 'l-m- g ~sENFd is:EGd " ii 15 i hEEb _ N, Q =a

=

h"l s*3*p c 2 s i g r gg ~ E n as 5"O.. ' 3 g 5a

• B 4-O f - I' m

U k_ U g.=4 I bu g V ~ 9r=U 80 0.A t U <e-bc) Css l

• G_

,# *5 *, - C.5 80 rw dB n ,e_ ' E.2ff:r E r% cn e r-g _5o j i T&T o 5 % q 2E5 si1j ds.- O j$es. = L Rr= L 2 5;E c=- a-

n..

c a.,n q _, g f.,$. 55"a-() .l 8 j w; a i. e x.. L-

• g.:

SS > $

• • f.5..,

we E e G l 92 14

PRA 3.49' QUESTION ~ FOR THE ELECTRIC POWER SYSTEM FAULT TREE, IDENTIFY THE REFERENCE AND THE ASSUMPTIONS USED TO OBTAIN THE UNAVAILABILITY VALUES 0F THE FOLLOWING BASIC EVENTS: A) BATTERY SET UNAVAILABLE B) 4KV BUS OR SWITCH GEAR UNAVAILABLE C) ALL DIESELS FAIL TO START RESPONSE 3.49(A) BATTERY SET UNAVAILABLE

REFERENCE:

TABLE A.2.2 LGS PRA (WASH-1400 ) ASSUMPTION: MONTHIv TESTING _g T 3.75 x 10 720 HR. CALCULATION:.. UNAVAILABILITY = A7 HR. 2 = 1.35 x 10-3 RESPONSE 3.49(B) 4KV BUS OR SWITCHGEAR UNAVAILABLE

REFERENCE:

ANSI /IEEE STD 500 4.5.1 METAL ENCLOSED POWER SWITCHGEAR 1.3 X 10-7/iiR. 4.6.1 METAL ENCLOSED BUS 1.16x 10-7/HR. 2.46x 10-7/HR. ASSUMPTION: WEEKLY TESTING CALCULATION: U = Al = X 10-7 168 HRS. = 2.1 x 10-5 x 2 HR. 2

1 y RESPONSE '3.49(C) ALL DIESELS FAIL TO START 0 FAULT TREE PROVIDES SEVERAL MODES OF DIESEL AND MULTIPLE DIESEL FAILURES (SEE FAULT TREE). O THESE FAILURE MODES INCLUDE COMMON MODE MULTIPLE DIESEL FAILURE AS RECORDED IN LERS MAINTENANCE COUPLED WITH RANDOM FAILURES RAND 0M FAILURES SUPPORT SYSTEM FAILURES ALL DIESELS FAIL TEST 0 DOMINANT CONTRIBUTOR IS THE OPERATING EXPERIENCE DATA HISTORY OF COMMON CAUSE FAILURES OF DIESELS. O ALL DIESELS FAIL TEST IS AN ATTEMPT TO IDENTIFY OTHER POTENTIAL SMALLER CONTRIBUTORS. l l t i i r l l

PRA 3.50

• QUESTION HOW HAVE THE GENERIC COMPONENT TREES IN SECTION B.9 BEEN INCORPORATED IN THE PRA?

RESPONSE

O GENERIC COMPONENT FAULT TREES ARE USED AS EXPLAINED IN SECTION B.9. O THEIR PRINCIPAL USE IS TO PROVIDE THE REVIEWER INFOR-MATION ON THE FAILURE MODES INCLUDED IN THE COMPONENT FAILURE PROBABILITIES. l e 9 v m. .r-,,___..--

1 PRA'3.52 QUESTION FROM THE DIESEL DEPENDENCY ANALYSIS (FIG. B.5.5) THE CONDITIONAL PROBABILITY THAT DIESEL 2 FAILS GIVEN DIESEL 1 HAS FAILED IS GIVEN A PARTICULAR VALUE. HAS THIS VALUE BEEN USED IN THE LOSS OF ELECTRIC POWER ANALYSIS?

RESPONSE

0 YES. O CONDITIONAL PROBABILITIES OF P(2/1) ARE USED IN ALL SYSTEM FAULT TREE EVALUATIONS. O THE SPECIFIC PORTION OF FAULT TREE USED IS INCLUDED HERE. [^ t e 4 ~ ^ ~

• ww s y

PRA 3.52 o QUESTION ~ HOW DOES LGS-PRA ACCOUNT FOR DIESEL'S FAILURE TO RUN GIVEN START?

RESPONSE

1. SINGLE DIESEL MISSION FAILURE PROBABILITY 0 DATA USED TO CHARACTERIZE DIESEL FAILURE PROBABILITY INCLUDES FAILURE TO START AND RUN. (1.7X10-2) O AVAILABLE DATA EVALUATION INDICATES THAT THE DOMINANT CONTRIBUTORS TO DIESEL UNAVAILABILITY ARE THE FOLLOWING: FAILURE TO START (PEACH BOTTOM 7.8x10-3) FAILURE-TO PICK UP' LOAD AND RUN INITIALLY (PEACH BOTTOM 8.9x10-3) MAINTENANCE UNAVAILABILITY 0 NO ADEQUATE DATA EXISTS TO CHARACTERIZE THE LONG TEhM OPERATION OF DIESELS. O FOR REQUIRED MISSION OPERATION OF 2-4 HOURS THERE WOULD BE A VERY SMALL CONTRIBUTION TO THE SINGLE DIESEL FAILURE PROBABILITY. 2. MULTIPLE DIESEL MISSION FAILURE PROBABILITY 0 SINCE COMMON CAUSE FAILURE RATE DATA IS ALMOST EXCLUSIVELY " FAILURE TO START DATA", THE CONTRIBUTION OF RANDOM INDEPENDENT " FAILURE TO RUN" PROBABILITY WOULD BE NEGLIGIBLE IN THE QUANTIFICATION OF DOMINANT SEQUENCES.

1 DATED: E 2 9 198P MEETING

SUMMARY

DISTRIBUTION: ...Dncket Fil e _(50-352/353). # .. RC POR N Local PDR NSIC PRC" LB#2 File Habel son EHylton Hodgdon, OELD ESchelliah, DST FDCoffman, DST DDYue, DST BHardin, DSI JMeyer, DSI RNewton, DSI SAcharya, DSI EChan, OELD .}}