ML20037B782

Ltr to All Pending OL Applicants of NSSS Designed by B&W Requesting Info Re Auxiliary Feedwater Sys

ML20037B782
Person / Time
Site:	05000000, Bellefonte
Issue date:	04/24/1980
From:	Ross D Office of Nuclear Reactor Regulation
To:	AFFILIATION NOT ASSIGNED
Shared Package
ML110830264	List: ML20037B782
References
NUDOCS 8012010476
Download: ML20037B782 (52)

Text

{{#Wiki_filter:_ ~ i ENCLOSURE 2 lQS C90y% UNITED STATES 51 NUCLEAR REGULATORY COMMisslo,a; [., b.. -'- j [ W ASHING 70rd. D. C. 205!5 g VQ,/ f 't, '

• j'"

April 24,1980 TO ALL PENDING OPERATING LICENSE APPLICANTS OF NUCLEAR STEAM SUPPLY SYSTE".S DESIGNED BY BABC0CK AND WILC0X

SUBJECT:

ACTIONS REQUIRED FROM OPERATING LICENSE APPLICANTS OF NUCLEAR STEAM SUPPLY SYSTEMS DESIGNED BY BAECOCK AND WILCOX RESULTING FROM THE NRC BULLETINS AND ORDERS TASK FCRCE REVIEW REGARDING THE THREE MILE ISLAND UNIT 2 ACCIDENT !n our letter of September 27, 1979 to all pending coerating license applicants concerning followup actions resulting from our reviews regard-ing the Three Mile Island Unit 2 accicent, we indicated that eacn appli-cant would receive additional guidance from the NRR Bulletins and Orders Task Force. This guidance would be related to (1) Auxiliary Feedwater (AFW) systems, and (2) analyses for small break loss-of-coolant acticents and in-adequate core cooling, including guidelines for emergency operating procedures. The purpose of this letter is to advise you of the information we require related to Auxiliary Feedwater systems. The requirements w re identified during the course of the NRR Bulletins and Orders Task Force review of pressurized water reactor nuclear steam supply systems in light of the Three Mile Island Unit 2 accident. Auxiliary Feedwater (AFW) Systems ~ The Three Mile Island Vait 2 accident and subsequent

vestigations and studias highlighted the importance of the AFW system en the mitigation of transients and accidents. Following the Three Mile Island 2 accident, operating plants having nuclear steam supply systems designed by Babcock and Wilcox were shutdown. During these shutdowns short-term actions were taken to improve the reliability of the AFW system. As part of the long-term requirements of the shutdown of these plants, more systematic relia-bility analysis of the AFW system has been conducted by the licensees.

l The staff is currently evaluating these analyses. Al a as part of its assessment of the Three Mile Island Unit 2 accident a.ed related ir.p?ications for operating plants, the ttaff evaluated the reliability of AFW systems for all operating plants ha <ing nuclear steam supply systems designed by Westingnouse and Combustion Engineering. l The objectives of the staff's study, related to coeratian of ;ressurized water nuclear steam supply systems was (a) to identify nececary changes ir AFW system design or related cperating procecures at c;erating plants in orcer to assure continued safe oceration of these plants, anc (b) to identify cther system characteristics in the AFW system design of these l Plants which on a long-term basis : pay require system mcdifications to improve reliability. l l on*** s 47C ,...~_.7 _y__ __-._-.m _..,..__...,,.._..my

r . April 24, 1980 To accommodate these objectives the staff reviewed plant-specific AFW systen cesigns in lignt of current recuirements, and assessed the relative reliability of the various AFW systems under various loss-of-feecwater transients, one of wnich was the initiating event at Three Mile Island Unit 2, and other postulated potential failure conditions by determining the potential for AFW system failure due to common causes, single point vulnerabilities and human error. It should be noted that, because of time and personnel limitations, our evaluation of AFW systers was limited to operating reactors at the tire these studies were initiated. Our review of these systems resulted in the identification of changes required for current AFW systems. Some of these changes were generic and others were plant specific. A summary of the techniques used in our evaluation is provided in Enclosure 1.* With respect to operating license applications such as yours, we will recuire that you (a) provide an evaluation which shows how your AFW system meets ea-h TE;-.c_ cr.; ia 5;d.-a u Review Pica 10.4.3 and Brancn iecnnicai Position A55-10-1, to) perform a reilaciitty evaluation simiiar in metnoc to tnat cescricec in enclosure a :nat was pertorrea for operating clants and suomit it for staff review, (c) tactor tne recommencations of Enciosure i into your piant cesign, anc (c) responc to tnclosure 2, wnica recuests tne information necessary Lc cetermine Ine cesign oasis for your AFW system flow reculrements anc to vertry tnat your Arn system wiin meet tnese recuirements. You are required to provide the information discussed above as an amencment to your apolication. Your schecule for submittal should take into consideration our requirerent that both staff approval of your AFW design and actual accomplishment of any necessary modificatiens be completed orier to issuance of a full power license, as reflected in item II.E.1.1 of the TMI Action Plan (NUREG-0660). -(L) J' ./'. - -Since rely,- i a* Y } f. t [j. q - D. -F. Ross, Jr., Acting Director Division of Project Management Office of Nucletr Reactor Regulation

Enclosures:

As stated cc: Service Lists

• Enclosure 1 is Appendix III to the staff generic evaluation of Westinghouse designed plants, NUREG-0611. The techniques described and recommendations are applicable to Combustion Engineering and Babcock and Wilcox designed plants.

.<v. - -, - -+- --. g_-w, 7,.-y

---

.ww-n-,

-.,u.- ,p_.,,,,,,...w.y.,.,,,p..y_,..y ww,-y, y- .,,,_,-----m -ow+--- g-.---+--

f ENCLOSURE 1 APPENDIX I!! Aux!LIARY FEE 7 WATER SYSTEP.5 1. IhTRO UCTICN The Three Mile Islane Unit 2 (TMI-2) ac:icent anc sucsecuent investigations anc stucies nignligntee the ' cor ance of :ne auxiliary feecwater (AFW) system in t e sitigation of transients and sc:i:ents. As :ar of its assesse.ent of ne TMI-2 ac:1 cent anc related i=clications for ::erating slants, ne staff evaluaten the AFW systems for all ccerating plants naving nuclear steam sucaly systems (N515) cesignec Oy w stingscuse (3) (25 units) e or Cosoustion E gineering (CE) (S units). (See note below.) The :cjectives of this stu:y ere to: (1) icentify necessary enanges in AFV systes cesign or relatec prececures at these plants in orcer to assure tne continuec safe c:eration of these plants, anc (2) to icentify otner sysm cnaracteristic: in :esign of the AFW system for these plants nica, on a 1cng ters basis, say recuire systas socifica.tions. To ac: m-plisa these cojec-ives. -e: (1) Reviewec piant specific AFW system cesigns in lignt of current regulatory requirements, anc (2) Assessac the relative relf amility of the various AFV systems uneer various less of fetcwater transients (cne of wnica was tne initiating event at 7MI-2) anc :tner ;;sta-latec ;ctential failure c:ncitions by cetermining the ;otential f:r Ar"W system f ailure cue to c:mmon causes, single point vulneracilities anc numan errer. As part of our evaluation, we performed a stancarc ceterministic type of safety review, using as principal guicance the ac:::tance criteria sce:ifisc in Section 10.4.9 af the Stancarc Review 8kan (SRP) 51) In conjunction with tais caterministic review, we useo event tree anc fault tree logic tecnnicues, as part of a reif acility analysis is cetermine ccminant failure ncoes and assess AFV systes cc::arative relianility levels uncer specifiec ty:es of transients. When the recommencaticns icentifiec in inis review are imolemantec, tne reliamility of the AFW systems for esca ocerating plant snoulo se im:rovec, with the cegree of imorovement cepencent u en wnetter the AFV systers were initially enaracteri:ec as naving relatively nign or law reliasilities (see Section 4.5 of this a:pencix fer catails). The time anc :erscnnel limitatiens i= cosec on tais stuey :reclu ed a ::molete anc extensive review of eacn AFW system. The aeview.as asec =rimarily u:en information :revicea 'y esca licensee at a four-ncur meeting with tne staff review team (c:mcosed af a systems engineer anc a reliacility engineer) to review the as-:uilt AFV system :esign anc Ocerat on. C:nsecuently, the results saoule :e viewed in terms of tne general c:nclusions tna insignts, anc not as an assolute relianility analysis of ;eneric or plant-s ecift: AFW systems uson i iCUcies f the APd systams at 3acc:cx anc Vilc:x (E&W) cesigrec-:cerating plants ere asi : I sumjects of sacarate C:mmission orcers anc Other wort ;erfor ec ty tne NAC staff. !!*-1 1

i I i .nica the accestasility of these AFW systes ces1gns may :e ;ue;ec. This eltacility essess-eent resulted in the cevelcoment of generic anc plant scecific rec:mmencatters to tm: rove AFV systes reliacilify. It was recogniced that it would se very tifficult anc suajec to lar;e uncertainty if an atteest.as sace to cuant1fy tne reliacility trarovement inneren througn isolesentatics of the recsamencec actions. It.as cac1 cec :nat sucn an effort.as outsica the secoe of tnis stucy. Some AFW systess in 0:erating nuclear power plants cs not meet all cur tnt staff licensing critaria containec in the Stancarc Review 81an. The cegree of ::nformance varies witn :ne age anc 1:ecific plant cesign or tne 33 anits accressec in this stucy. For azaeole. 10 arenitect/engtreering organt:at1ons were involvec in the olant cesign anc construe:1:n of these 33 units. A specific ocjective of ints 'stucy.as to ceterstne wnetter the lact of c:nformance witn any of these later requirements representec ;ctential safety ;rceless, consicering the THI-2 experience. The recommencations (centifiec in tais stucy reflect areas of potential weatnesses wnere enanges to improve AFW systes reliamility snoula :e icolementec.

• he results of the AFW systes casign review and the evaluation of THI-2 accicent imslica-tions.ere jucgec to retutre consicerat:ce fer c:rrective action if any one of the following conciticas.as icentifiec:

(1) Caeson.,cce failures (particularly those relatac *.s numan error). (2) Single point failures, or (3) Any cceinant causes of AFV system unre11asility. Our limited review focused on the implications of the THI-2 accident, particularly human errors, anc thus we cia not reevaluate the cesi;n : asis for eaca AFV system, nor cic.e focus uten all posst31e systes interactions that could affect AFV system reliamility. =owever, if the inferr.ation suggestec a ;otential for loss of AFW frem saca causes, this ;otant1al.as notec curing the specific plant reviews, wita followuo evaluations recommencea :s cetermine the need for acciticnal actions. In cetermining wnich safety issues recuirec snort-term licensing action versus tacse inat c:ulc te ceferfec for furtner evaluation, we usec simplifiec engineering tvaluations anc cualitative jucqment of the safety significance of the various issues.

• n tais *egarc..e rec:nmenced actions if their ieclementation.oulc provice suostantial, accitional r:tection recuired for the sunlic nealtn anc safety. The recommencec actions. eat scecif'c anc saf ety-significant in their cnarac er, c:ula :e i=cler. entec in a timely sanner, anc oulo not likely se overturnec er contracietec ey continuing stucies or investigations. Icee cf them say eventually :e cisclaced, however, cy more ccamrenensive lon;-term changes in nuclear sewer plant regulation. In some cases, basec on information sr analysis :eveiccec to : ate.

it is not clear that a basis for a cecision is availaole. In suca cases. e ' ave juc;ec the ites to te of sufficient safety significance to recuire an early c:mmitment ts get s: ctes I:!-2 r w.

---

,-m,--

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

( cr testing unCer way la OrCVice a basis fer resolu*icn of ine isset. As Pecu1 rec, int ree mmennec ac ton is to c:tain a c:amitment for a 1cnger-tara =cetittation, stu=y, :r test

y affectec Itcensees.

2. PURPCSE AND 015 XIPTION OF AUXILIARY FEED' ATER SYSTEM The AFV systee funcitons as an emergency system for tne removal of. neat fr:e tne :rizary system wnen the main feecwater system is not availasle. It also plays an iscortant role in zitigating the effects of some :esign = asis events; for emarole, sese small unream 1 css-of-ccoiant ac 1:ents (LCCAs). The AFW system is esignec : nel: tr.e Olant at het stan::y, or to cool ::=n the primary system to tercerature anc ressure levels at.nten tne Icw pressure cecay neat renoval systes can coerate. The AFV system can also :e asec curing normal plant startuo anc snut==wn concitions. AFW systems usually c:nsist of a c:==1 nation of steam tur:ine-criven anc electric motor-criven zu==s. The AFW system can =reWice. -it any one cuma cut of servica, enougn ater to tne steas generators for cecay neat rescval following icss of sain feeewater flow. Tacle !!I-1 Orevices a sunmary of the :umo ::ssinations, f1c= ratings anc soces of initiation for tne 4FW system fer esca =iant rewte.ec. A::encia 1 previces specific AFW system cescit:tions, a sinclifiec ficw sheet for ea:n f ::erating plant, and an evaluation wita c:rres encing *ec mmencations. 3. CETERMINISTI EVALUATICN 2.1 n acreveu.-c inc Ob4eetives In our review of current a:alications for ::nstru:-ion :ermits anc ::erating licanses for pressuri:ec water reacters,.e evaluate the AFW systes ta assure that tne cesign :caforms s the a:alicasle General Oesign Cr1tarta (0 C) in A= enet A to 13 074 Part 50. The eners' Oesign Criteria icentified in Section 10.4.9 of the Stancare teview Plan 4:alicaole to the AFV systes cesigs are listec telew: (1) 30C 2. "Cesign 3ases for Protection Against Natural 7menomena.* as relatec to struc-tures housing tne system, anc tne system itself Oetn; ca:asle of itnstancing LSe effects of natural pnenemena suen as eartacuases, tornacces, nur?1 canes, anc ft ccs. (1).G;C 4, "Envircemental anc Missile Oesign aases," witn res ect to structures mcusing the system and the system itself teing ca:aale of withstancing the affec s of exteraal missiles and internally generatec =issiles, stee.nts anc ;st imotngement forcas ( associatec with pipe :reans. (3) CO: 5, *! haring of Structures, Systems, anc Csm:enents." as relatec to tne ca:acility of snarec systems anc c:eeenants ie:criant to safety to :erfers recairtc safety { funC* ions. .(4) 000.*9,

• antrol Acca
• as relatec :: :ne :esign :a:a:ility :f syste= instrumentaticn and c:ntrols f r 3r =ct not snutscwn af ne react:r, anc :ctantial :a:acility 'sa su sequent colc snuteewn.

I l

-2 1

i s =. 3,c .eg. .g. .Ux.....Y....... EYg..yg - :.... ; 3. Nu... c., ... = .:5..N...,5c. 3., ... u e.. -. : 2. + t. ...A.

• s...

,AN...a 1 i Ah; I'."SUSI CN ENG'NE.TAING i Wettingscuse-Cesignec Plants

• c. of Pumes/

AF=5 *cce plaat T<:e e' 0**we ra:aettv e* feet atten 0:, neats fe...e 1-5 teas : riven Staan: 700 ;;m 1 Aut::4 tic Valley. 2556 ft. 2-acter Ortvan

• ter-350 ;;= 2 i

(eaca) 2556 it O. C. ~c.s 1-5 team Oriven* Stea=: 800..;;m 7 Ju".:satt: 7tr anit ,.. ft et:r : aces i 1-*ct:r Oriven* Tctor: 450 ;;m J ss :1y : eta 'i 2714 *: .n1 ts 1 i Farley 1 1-Steas Oriven S ta a : 7CO ;;s 3 a t:matic u 1:53 :st;

-*otar Ortven

• t:rt 250 ;;s J (ea a) 1*SS :sig

~tena 1-Steam Oriven Stea=: .00.. ;;m a Aut aatic 44 :sig 2-a:::r Oriven

• r:

00 ;;a 4 (normal AF=5) (eacn) 1114 :st; 2-=:ter Ortwen met:r: (*0 ;;a sanual (stanc:y AFV5 naccas v ct -Steam Ortven Steam: e 450.;;.s 7 =anual .nn. ,sta M. 3. A::1nsen 1-stets ' riven Steam: 50,0 ;;m ) aut:satic ....~,.si 2-act:r 3rtwen .=:t:r: 200 ;;= 2 (eaca) 1300 :si !ncian Pt. 1-5taan Driven' Steam: 300 ;;s i Awtssatic

er. nit 2&3 1250 :st;

-*oter Oriven' M:ter: 400 ;;s 7 (eaca) 1250 :sig (e.aunee 1-steam Oriven Stea=: 210 ;;m 7 Awt:casic ..o.e, 2-et:r Ortven .=:t:r:

0 ;;s J

_ (eaca) 2150

• Norta Anna 1 1-Steam Oriven S te am: 700 ;:n f Ast:satic a.n0..

-=ot:r ; riven

• c to r: 250 ;;m )

- (eacn) 2500 ft 2 air e 1-5taam ; riven

• Ste am: 2:0 ; = 7 Autema 1c 8er enit I

!slan: 1&2

• 2:0 :s q

r :u==

1-oter riven' = t:r:

;;m )

r ally 1:00 :ss; feecs :::cstte l wnt steam

ecerators 1

1 l 8t. fenca -Ste am Ort en' S te am: :00 ;;s J Aussmatic 8er.nt: 1L:

• 152 :sig eter :w==

1-act:r " riven * =ctor: 200 ;;.9 ) i;; lies (eaca) 1152 :si;

tn.nitz

= sc:e: 2ee

--eats et.wan

... 4 y_ _..,_.. _ _ _ _ _ _ _. ___ _y.,.--, _-.,.y .c_ ,y,,.,m,, .~ .w.. y,,. _. - -,-m. ,,,,,.,w ,______e,

( 7ABLI ::-1 (0:ntinuec) 'estingnouse-Designes Plants No. of 7cmes/ AF.5 *: e 81 ant Tvee of Orive Oa:acity e< '-'t atate

-*eets Sales 1 1-Steam Oriven Staas: 3R0 ;;s 7 Aut::atic 1150 :si 2-Motor Driven set:r-140 ; m J (eaca) 1200 :si San Cnofre 1 1-Steam Driven Steam: 200.,;;i 7 sanual

_.__..si 1-wotor Driven a t:r: :25 ;;m ! c 1025 :si Sur y 1 & 2 1-Steam 3rtwen' Steam: 700 ;;m a Aut:matic ne :u o 27:0 ft esca ATV 2-actor Driven' setor: 250 g:m i systes :an (eaca) :720 ft feec :::ostte untt irojan 1-Steaa Driven iteam: 960 ;;m a Aut:matic 2400 ft 1-Oiesel Driven Diesel: 960 ;;m J 2:00 f Turtey Pt. 3-Steam Driven' (eaca) SCO ;:n 9 Aut:matic One :uco 3&4 for Octa units 2775 ft er ally su: lies eaca unit - Irt :uma is

acxu: for e t.ta unit Yansee Rowe 1-5 team Criven*

S team: 90 ; s @ Manual

• .a ; P.g sne 1*00 :s1 sa f e t,.

fe;e:- -ion syste?s serve ss

a:x*.c Zion 1 & 2 1-Steam Driven' 5ta am: 900 ;;s 7 Automa ti:

er snit 3099 ft sotor Driven' seter: 450 g;m 9 (eacn) 3099 ft "sota: see 0:ments coluan h

l ...?. a.

7A3Li !!: l (constncec) AUXIL*ARY FEII 4713 SYSTEMS

mcust*.cn Engineering-Cesignec 71 ants Plant No. of pum:s/

Casacity AF=5 woce Type of 3 rive of : nit atica 4NO 1-Steam 3 riven Steam: 575 ;;m J a tomatte u 2!00 ft. 1-dotot 3 riven Mc La.m 575 ;;m 9

201 ft.

Calvert

Iiffs 2-Staas Oriven 700 g;a #

Manual 112 per unit 1100 psia esca steam: 250 ;;a f Ft. Calhoun 1 1-5 teas Driven 24C0 ft. Semi-automatic actor-criven 1- % tor Driven Motor: 250 ;;n 3

w manually connectea to 2400 ft.

ciesel generator Steam: 500 ;;m #

• ain vannee 1-staan Driven 1100 psig "anual 2-dotar Driven
• cto r: 1500 ;;m 7 (eaca) 1100 asig Steam: 600 ;;m 3
• anual Milistone :

1-Steam 3 riven 2437 ft. 2-*otor Driven .v tor: 300 ;;m 7 o _. 2437 ft. (eacn) 5te am: 415 ;;m ) Palisaces 1-Steam Dr1ven 2733 .anual 1-*ctor Driven =otor: 415 ;;m 7 2733 ft. S team: 300 ;;m 7 St. '.ucie 1 1-Steam 2 riven 1:00 asi 2-actor : riven

• otor: 250 ;;m 1 (eaca) 1:00 ;st

! i

.t

4. *::aling.ater.* :s assure : e : sea:tlity, s transf er.e:t ':a:s '-:s :*e (1) a reat sins.n:er all ::erating'c:ncsticas, re:wncancy of ::-:: rents res:::r sys:as ::

1ve ccm : rent fat 1.re. inat 3e safety func-ica can :e perdermec ass 6stng a single a:: s anc tre ca acility is isslate ::scenents :P st ing, if recatrec, sc nat tne sys es safety fun::icn =1*'. :e Saintainec.

$. *!ns

e: isn af :oling '.ater Systtrc.* as relatec :: cesign ;revtsions sa e :

(5) :: ecut:2ent.

ersts :erincic inser:1:e ins:e: ten :f systes ::::cnents an:

est;n =revisiens ta:e 1 00C a6. " Testing :f :alin;.atar iza.es." as relatec ::

(7) assure a::re:riate 'unc enal testing :f e systes ans : recnents : se a-1: systes 1:

era:ility anc :er*:r ance of ::e:cnents', anc =acastlity :f ine integrate:

fucction as intance: sring all c:erattng :enettices. In cetermining = netter tre AFW system cesigns for sa:n a::liest:ces zeet taese erers! Review Plan an: iranca esign riteria : e staff uses Section 10.4.9 of -te Stan=ar: 453 10-1) as gu:can:e. These Tecanical 7csttien ASE 10-1 (Bereaf ter referrec to as

r::ecures 1 :e 2:cuments ::ntain tne s::e: Lance criteria for :ne AFW sysses anc ine review

• f t e staff c:ncluces ore eet.

asec Oy Se staff La :etermine if tasse a::e:tance :riter14 tren it is also a:1e to c:nc1w:e tnat tre re uirements tras tne a :e: ance c-t'eria are set, satisfiec. of tre a: lica 1e enerai Oasign Ortteria art

any of nese at :ne cutset of inis assessment af ::erating plants : a:

• .as re::gni:::

levi ew 8143 inc1w=te;

14nts :o no =eet each cf tne acceJtance critaria of : e 5:ancarc 731s situation esists

e AFd systes :esign.

Se cit en

0. 4. 3 anc A55 : --1,.ni ca < el ate ::

se zu licatt:n tecause tne :cerating licenses for -any of taese ;iants -ere isswec :rter ::

the :s:11:ation :f tne Plan anc for seee of trese :lants. ; rice :: of One Stancare teview ene-al esign : rite *14 in Fe:evary iS?1." The 5:ancar: Review Plan.as ariginally isswec in Novem:er 1975 an revist: in 1978. ". hen 13e staf f issues ar= cr *evisec regulat:ry recuire ents anc gatcante it a::resses s ::erating clants.

P revisec re:viretents or ;4fcance sncule :e :actfi++e:

This :ecisien is ;uicec y Sectica 10.109 of onetner One new as -ell as alants uncer;:ing licensing review. 7his requistion states 13at. if a fincing 10 *?4 Port 10 of the :mmissi:n's regulations. recutrements :revice sL:stancial, acettional :retaciten.nt:n is ta:e is tne effect ina

e :acifistec :n ;iants -t:n ::erating rew

• s -etwirec *cr :ueli: *ealin arc safety. tney are :::een tace f:e sever 31 recetrt*ents licenses ( :erating iants'.

iwcn a fincing Nas -:: in 32P Se: tion ".0.2.i..nien ;;;11es : AFs systass. Consecuent:y. as note: 5::e ::erating ;ia*. s := {,ot :eet ai* Of tr.e e:wirements ':::se:

ntaire:

~ 1::ve, t?e AFV systems 1:

n ;ater :esigns.

'~ 1 :ne :as:: safety as :ar: :f 'O ?; 2 art i; ' A. n:stn ;re J;; ert :t:=uigate:

nsiceration esecciec in : e

• 0 nac :een in ;enersi.se f* m ine tar:y ;950's.

l .ee e I ese s

h i ,7-d '.*t secuic te mote: inat AFV system :esigns.nien :o not neet ne :rtierta ti :ne 1:ancar: Revie. 31an are 30: necessari y in ::nflic.1:n :ne ;ene si Oesign Or te-a.

evia: :ss fr:2 ine itan:ar: teview 81an may :e justifiec (even on sew ;14nts) :r:vi:ec :nat an ac:::t-a=le level of :rctection is provicec in the overall :lant :esign. Price :: our assesszent, s;ecific cocumentatten of ceviatisas

• roe tne Stancare leview Plan, See:1cn 10.4.3, anc A33 10-1 nac not :een recaireo for all : erating plants.

It is a;ains: :nts eacxgrounc :nat.e :evelocea ;uicelines for :ne ceter:1ntstic review af AFV system cesigns for ccerating plants. These guicelines are provicec :elcw: ~ (1) Oster91ne tne extent to.nics ne AFV system cesigns sett the ch taria of tne :ar ent Stancarc Review Plan. (2) '*here AFV system cesigns co not meet the Stancar Review 71an criteria, cetermine . netter :nanges can ce icentifiec that will significantly u: grace tne auxiliary feec- .ater system in c:erating plants to came tnem less sus:ecticle to single point failurts, numan errors, anc conson soce failures. (3) tec:mmenc areas of the AFV systes :esign to ce evaluatec for longer-ters im:r:vements in the reliacility of AFV systes casigns. 4 REL!A3!LITY EVALUATICN 4.1 !acace une anc Obieetives The General Oesign Oriteria (COC) :ntainec in A::encia A to 10 C7R Par 50 pr:vice sinimum recutrements ta :e satisfied in the cesign of nuclear po=er plants, as suca, the 00C ;revice the tasis for the staff's ceterministic review of One cesign features of nuclear :cwer plants, inclucing those of the auxiliary feeewater (AFW) systems. "o. ares :nis enc, tne Stancar Revi ew Plan provices criteria anc sucolesental guicance to tne staff fer assuring ::nfor-sance vita :ne : C, including those a::licaole to AFV systt:s. ec=ever in :rovi:e a::1:1:nal ins 1;nt regarcing tne potential for failurts of the AFV system not :verec Oy :ne Stancar Review Plan, tne reliasility assessments ciscussac :elow ere erformec. l The THI-2 ac:icent cemonstrated that numan errers of :enmission er omissien can lese :: r l failures of recuncant anc diverse AFV systas ecui ment to :erf rm as :est;nes. ~nus. ne r TWI-2 es:Ar,ence tencs is ::nfirs past stacies*) incicating numan er rs are ::minant III 'act:rs in reacter a::icents. Carrently, a variety of AFV system ces1gns are seing use in ne 23 ::erating :iants usigg 1 d-anc OE-cestgree reactors. *his factor gives rise to a sariety f nareware :ecencencies anc possi le vulnera:111 ties :rougns a:out y 9uman interaction vi:n :ne :esign, :r :o.ssi:1y scoe ciner c:mmon influences that ::uic affect AFV systes :cerati:n. 3sst sta:tes") Save

reviced useful engineering insignts int : nose areas :f system :esign.eere 9eman 'nter-I actions ::uls significantly affect one availaoility of stanc:y safety sys t=s.

• te af:re-

-5 l

l

~ ( tentterec :as s:tcies ave also :rovice: se:::1:nai *est;nts for ne. ore :re:asie transie-events tr.at :anc :J ::sinate tse cemanc *:r suc:assful ::eration of :..e 2F'e systaes. The relianility assessment a: rcaca usec anc the :rinci:41 insignts se: esults are sunmar-1:ec :elow. ~he c:m:arative rellazilities of tne AFW systes for tne 33 3 anc !! tesigne: c:ersting react:rs were evaluatec for :nree ciffertnt ints:ation events ano are sac n in Figure III-4 Figure !!!-i Ortsents :ne c:m arative reliantlities :f tne AFW systass fer eacn of sne 15 t-:esignec ::erating reactors. The results snown in Figures !!I-a anc II-5 incicate anat the reif a:ilities of the existing AFV systes esign vary :y at least an orcer of sagnituce. The cominant c:ntricutors 1: inis varia 111 y in relia:ility =ere. in ;erera numan errors anc single Octnt vulnera:ilities as :escr1:ec later. Plant-s ectfic :etails :n Laese AFV system casigns are :roviceo in Ap en'cix X. 4.2 telia:ility sssess ent at:reacn enc se ee Reliacility tecnnicues anc insignts =ere usec in :nis assessment = su: lement tne sore tracitional :eterministic type of safety review. The rinci:al tecnnicues usec in inis assessment inclucec the event-tree anc fault-tree logic tecnnicues. These tecnnicues are ~ consicereca'-) to re: resent an ac:e:tanle a::rcaca in esta:lisning the priorities for the C resolution of generic safety issues. These tecnnicues, anc the insignis cerivec ty use of suca logic, nave teen er loyee recently to perform a risa-asec categori:ing ano screening of the various generic safety issues.(6) Acc:rcingly, the staff usec tne aferementienec tecnnicues to f:cus on nese otential failures :nat c:ula ecminate tne anrelia:111:y of AFs systams :uring tne follo ing transients. 4.2.1 L:ss of Mafn Feecwater This transient involves ne interrustien of :ne sain fece.ater flow anc the sucsecuent tri :ing of :ne reactor. React:r ex:erience so;;ests nat scou :nree interau:tionf of ne main feeewater systes ay :e ex:erience: from a num er :f :suses esta reactor year." 4.2.2 Less of Main Feeewatar Cue = anc Loss of Offsite 7c-er This transient is initiatec :y ine loss f :ffsite :cwer.nt:n, in turn. :suses ne inter-Pucticn of :ne :ain fetcwater system anc ne trie:1ng :f :ne react:r. React:r e2:erience sug;ests inat :ne main feeewater systes may Oe interru::ac :y tais ;ranstent a::reximately 0.2 :: 0.3 times :er tect:r year

• inis av er iay 4:: ear :3

nflis'..>ti :ne infor at:en :resentec fn 'a le

• -a fn A::encia **. As.as noteu in Section 4 ' :nat ac: enc 1x, :ne events 'istec n Ta 1e !!-a rt: resent a minimum feteuency of 'ess of 'terwater events. since :tter int. sting twents

.nica resultes in a loss of feecwater say Set aave :een ' clucec.

a.2.3 L:ss c? watn Feee ater ana L:ss f All Alternating Osrrent Ac.er (Staticn a e 31acxcut). This event is initiatec sy ne Icss of offsite zower, as is the ;revious t.:rstent ciscussec in 4.2.2 a:cve, exce:t that tne onsite emergency alternating current (ac) ower sources are also ; cst. Thus, inis scenario re resents a significantly segracea case ecmcarec : inat

escritec in 4.2.2 amove. Newever, since tais event recuires the c:ncur ent loss of all onst ta ac newer sources (e.;., usually two ciesel engine generators), its likelincoc cf cccurrence snould se orcers of magnituce less. This low procacility of c:urrence notwitn-stancing, reacter ex:eriences nave revealec some crecurscr1 to tais scenario. For exacole, in a f ew instances, all ac : ewer nas seen Ics: for :ericos less than five sinutes, tneremy encroacning on the time to toil the steam gene'rators cry. In anctner instance, only one of tne ecuncant ansita emergency ac =cwer sources as availacle for a ; erica of accut 50 minutes. Thus, if for scre reason tne e:eratirq ac source =ert : fail curing inis 50-stnuta intarval, inis scenario c:ule nave Occurrec.

The asility to c::e ith tais event.as not a licensing recuirement fcr ine earlier licensec plants. However, the more recently licensec plants, as well as those currently uncargeing constructicn ;ernst or caerating license reviews, have :een recuirec to ;rovice AFW system

esigns ca:asle of functioning woon the loss of all ac power. Accitionally, the cecisien wnetner or not this transient shoulc se a plant casign masts, anc for nat merioc of time this c:nciticn would te assumed to exist, is teing reviewec as an unresolvec safety issue uncer Generic Tass Action Plan A-'A II). Secause of the a:ove consicerations, tne a:ility of eacn ::erating : Tant's AFW system to c::e vita this transient was inclucec in :nis as sess ment.

4.3 *Geeeeic Event Trees The incuctive logic usec in-evaluating the-relative reliacilities of the varicus AFW systems involvec the use of generit event trees. The cominant failures affecting the availa:ility of ne various AFV systes cesigns for eacn of the three transients were assessac sn a c:nciti:nal tasis ratner than on an overall prc atility : asis; f.e., tne reliacility of eacn AFW systes was calculatec, given nat tne a:alicaele transients cescritec a:cve nac ccurrec. Figurt I*I-1 illustrates an event tree a :licacle to sary' of the current AFV system :esigns. Altacugn :nis event tree cces not c:ntain all the various systems that say tec:se involvec ver ne c:urse of the transient, it illustrates ossible ac:icent secuences anc outc::es nat c:ulc result. The meavily snaced secuence illu.tratas :ne of tne loss of main fete-l .ater transients cescritec amove that cesancs successful ::erati:n of tne aFV system. ine .1me interval of interest f:r all tas transient events c:nsicerec is the snavaila:ility :f AFs systems curing the merica of time to toil the steam generators :ry. Beyen: tai s interval. ;rimary c:olant.oulc se cisenargec via ;rtssuricer relief anc/cr safety valves anc there:y te lost from the ;rimary coolant system. Vitnout the satisf act:ry ::eratten :f

rimary c olant saaeus systems (e.g., nign pressure injection systems), LNt reactor c:re l

c:ule :e ancaverec and eventually camagec. Further, as tais :cil-cry tire is a: roacnec, the acility to crive tne steam tur=ine-criven tumes AFV c:ulc :e Icst.

• f ine AFW systan

!!!-10 - - - - - - - - - - - - - - - - - - - - - - - - - - ~ - - - - ^ - ^ ^ ^ ^ ^ ^ ~ ^ ' ' " ^ ' ^ ' ^ ^ ' ~ ^ ~ ' ^ ^ ~ ^' ~~~ ~~~^^"'

e 4 5 a I 2 ~ a s 2 s = c li w I 5 =* i 2 2 2 E a a a s e e 66 i .2 7 a 3 i .. s. - g i II lg3 I I l: I I I g 3 ! I 6 i. dji I l 's i II IgI I gl 1 I i I I g I g 3 l i I i I, I

g i I l

,e i 1 g g g I I YI I g g = h !l Y I Y! E I l ! ) m{3 I I I i i I I I I I i 1 s I i i i I I ! 1 1 I I I l ~3 = - ~- 1 I I I Ig 0 g :h = I I i ! I I l 1 5 ~ LJ ! LJ I L.J LJ -.J ..J ! I i I a > ;t I' 1 I a I I l I I s

• • d 1

i I 1 g g ( ~ ~,i I _r l I I I g I l 12 c c>-l I i I i I 14l ~. 1 ~53 e i I l 1 I t13 .s 'gG 1 l l 1 12 2 J l _J l i 1 II I I l cl 5SI I I I~l 5 a = =a5 I I I I zl c I l .J 3 i ~5 I l' ij 2 1 '>

• l 253 I

I I L 6 J &~ l I al r _J fI 4.35 's 'l <j i e I 3

1. ;.

I "5 1 E' 9 d 0

• III l

j If ' N:j<si4i i 4 ~' i;5 j l ll ! Il l I 13 -:s 1 ~1.3 I s i e = -= = 3 ,Y n 1 v ~ .12.

est;n ::ntains :miy stea= tur:1re-cr*ven :ur:s, or if ine transient is s.:n inat :nly inis AFW s.:systaa is avalla:le. and if Lne ::ii-cry tire is a::rea: ec, inen :ne liaallace: Of init14 nq aFW systes : eration -culc te re ace: significantly.

4. 4 8avit ~~** '_::1e :::rea:n

~?e :ac*: ;.e Ic;1: usec in evaluatin; tne relattve reliamilities Of tne varicus AFW systems .as :ase: :n the 10: lean ic;1: ass:::atec =ita fault trees. A sinclifiec or recuer: fault-tree a: r:4:n.as usec :: est:nate tne unavallasility :f AFW systass := a :ar.anc. !n nts assess ent, unavaiia:ility.as taaen as. :e ng synencocus =tta tne unreliasility. This a::rea:n tiiac :n tne engineering insignis ava114:le tareu;n a::11:ations :f :ne systes f ault ; ves in e Reacter Safety Stu:y (WA5rt-lC'('), anc :n sa:secuen =ces uncer aaen

n 4::iti:nal lignt.ater reacter casigns. This lat ar wart was :asec :n a systra survey aralysi s iet. nicae(*).nere eeuce:

8 an: si::lifie: fault tree ':;1:.as usec :: est ate an: tne ::=1nant system failures an everali system unavailt:111ty. ~igures ::!-2 an: :: -3 illustrata tre sir:lifia: fault tree Icq1: ?:P an AFW system :esign. Each fault tret icentifies tne rin:::a) failures ex:ectec to P. ave tne ecs; influence :n tre unavaiia:fiity :f LPe aFW system for the s:e::fic transient event i:entifie: in tne figure. Ic assist in : arar:1ri: ng the tlia:ility of tse varteus AFW systes casi;ns an: ; nel; in icentifying the e=rt likely f ailurts that ::ul: af fect the various cesigns :Lan-itative estimates -ert mace f r:s tne fault tree logic structure. Tewart ints en:, a s:e::fic :ata

ase.as ::==ile: an: usec : ;eaerate test est: sate f ailure ;r::asilities anc nu.an er cr

cten Tais ::nsi:erec :: :e a::li:::le :: tn:se ::::enents an: numan intara :: ens ever Lne

• arge f tne AF. sys:t3 :esigns.' a :r nci:al reasen for ::=:iling tnt: res estimate ti e :f cata :ase was to assure that the cuanttiative estinates :f relia ility :erivec f r:e the f ault trees :wis te usec :: ::: art the rela:1ve relta:ilities of the <arious AFW system :esigns. This.as ::ssible ecause tne cata vert ::nsistently use: Over the range Of tne aF systas :est;ns Oy all :f ine rtita:111:y engineers =n were invcive: in :ne fault analyses. The :sta :ase an: its use in a fault tree ic;1: str.cture are cescrite: rtefly

elcw.

4.5 lats !ase a*: ::liestien Ta:1e :::-2 :rtsents cata ::::ilec ste:ifically fer ::ncu::in; inis AF systes ;ereri: assesseent. ~re ::r: nent failure an nu 4n err:rs ;r::a:iitties :resente: in 7acie :*:-2 e: resent : r tr.: :es; estimates. Ite :t==:nent failure rates.ere :erivec fr:m several r-s sources, in:1.:ing Lne tes:::r 54f ety ita y C. AIM "3)' an: :ng:ing %R* :ata assessment

r:grt:s. The eart:us numan errer :r::a:ilities.ert :erive: ft:s ::tn LSe ita: :P I4fety "Eva..atin; :ne var a:il;;y in AF.5 cesigns.as ine :r're::41 aim tn ints assesseen:

atner : nan evaluat nq variasility in sata 10 :e a::11e t: a s:e::fic :es1;n. -,.i., tee.-

O ( O e Ja-21 $15 i t 5 !.j ~ ~ 2, =i C il ~ !si-))xr l8-IM I: I.e n.11-- -al-gg i

==- a. .!.i.E { }$ ,3 } ~ = _1 e}t i i .i.s. J 1., l.: l 2 B 1:2 \\ h E! l } bE h ~ v ..i _= f. m-y g. 3 - -e e A a s :J si V t a 3 i O e ? I ,9{} ...t i > si l-c I r 1 3 } ].4. :3 F .{ e *. J *. l $ 3XI~ ~ ?_ 1 ".I .yp g J V T 5 a32 !} -=- 4.n I l =l la a ~ /- -E .,I II l e he 1 .2 w I$ Jt 24 21 13 ~ it ] N O l!'II.i 4 .~ -{)=I Id! r-- f4 I,m. 1 l p* _- ( 911i!"..it ~ suit' O 9 9 e __.,_v _g .y., ...-,v...,--..--

l w apWR Sese to 1 et 3 Sam.mo h so w' s ? e==. t owe. 3 g LCFw CaevCC + l 1 I p wan Tome t & _; m n a Ar* m=g asem Tes== 2 T e & h=== 4m 7, ( o.- \\ / L .s . na sa 0 +*e 3 1 l ( l c-e v... w se I v.-. c2 v c4 A c: a o.-. a wc o t ae a.- s " Ace s Ana:=uman vam.s ves=me ty *.w I an.g fy fe.un 2 en.e 6mm*89 C.W'.8 b.m8fW OWm.ms Figure 1113 Srnshfiec Fswit Logic Struc ure - LOFW Transient. Cnsy CC Powr Availamie. l l [

( TABLE lII-2 iASIC OATA USED FOR PURPCSES OF CONDUCT:NG A CCMPaRATIVE A55E55NENT OF E3:57.NG AFWS ES!aNS & rnE A POTENTIAL REL;A5:L* TIES

oint value Estimate

f 8recasility of" ailure en Ocean:

1. Cereceent. d a re a -* ' s i l era Cats a a. <aives: =anual valves (pluggec) ~1 x ' 3,, Che:x Vaives 4 si a 10 Motor 0:eratec valves Meenanical Ccm=onents si 13,* Plugging Contricution -1 x 10 Control Circuit (local to Valve) ,2 w/cuarterly tests sd x 10 : w/contaly tests

• 2 s 10"'

Piston Actuatec Valves McV-aecnanical Cocoonents $1 x 13, 50V-aecnanical Comeonents s1 x, 3 Control Cir:utt (Note: Use MOV c'" Failure Rate if Valve is not Fail Safe), c. Pu es: (1 Puro) 3 Meenanical Comoonents

• 1 x 13 Control Circuit (Local to Puro -

acolies to Electr1 cal Pusos) w/ Quarterly tests

• 7x'3.j n/ Monthly tests

%4 x 10 c. Ac aation tocic (Assumes at least 1 of 2 logic) ~7 x 10 */ train "irror f actors of 3-10 (to and cown) aceut suca valves are not unexoectec for sasic cata uncertainties. resresents a nuncer so small in magnituce that it may be neglectec f:r tasis of this stuc/. ea e S

1I-15 x.--w-

-r<,-%v- ,,.r -c.., .c, ,. ~, - -. - - - - g-.-,.,c,--- y---

d TABLE !!!-2 (Continued) "I*T

• w&~NiiNaNCE OUT10E 7:NTc'sUT!CNS:

a. saacusationas -car:ac 1. Test Cutage i Q feMrs/ test) (etests/vear) IEST ' ~ vnes/ year 2. Mair.tenance Cutage 0.22.sa-s/ aint. n e-1 MAINT. i;] 3. Data Tasles fo-Test i Maint. Outages' it==& Y OF TEST ACT CURATION Calculated Range en Test Mean Test Ac-C:==enen Ac-Ouration Time, nr Ouratten Ti e t, nr 3 Su ses 0.25 - &

1. 4 Valves 0.25 - 2 0.86 Dieseis 0.*5 - 4

1. 4 Instra=entation 3.25 - 4

1. 4 LOG-*CRMAL NCELD MAINTENANCE ACT OURAi!CN Range On

• ean Act 0:meonent Duration Time, hr Oeratten Time, nr Vumos

.t2 = 14 1/2 - 72 13 Valves 1/2 - 24 7 Dieseis 2 - 72 21 Inst w ntation 1/4 - 24 6 These cata ta:1es were taren frte the teactor Safety Stucy (kASH-1400)I2) sete: for ;urposes of this AFV system assessment. *shere the ;1 ant tecnnical toecifications slacac liett? on the cutage curatien(s) allcwec for 4FW system trains, this teen spec limt =as used tc esttr. ate the. wean curatten times for saintenance AC?S. In general, it was fcunc nat ne cutaces s11c=ec for saintenance eminatec :nese centricutions to AFW system vnavailacility frs> outages cue to testing. w f 6 eee ee ....s

r. - -----, ~ ~ ~ ' ~ ~ - ~ ' ~ ~ ~ '"

F IAntf lll-2 (Continued) lit ihmin Act s & t e s ors - f alluie flata: E st imat.:.I II. man f rror/f alluro Prr.l abilllies Ho.lifying factors & Situations With Valve Position With local Walk-Aroun<l & W/0 [Ilhcr Indication in Control Room Doub le. Che( k Ps uc ediar e s e i Point Value [st Est. on Point Value ist ist. on Point Value ist On Irror frror [ stimate [rror factor factor factor Arts & fraors of A Pre-Accident Nature a. C ' 94ldi'AliE5ilil5 net Mil 5i'Ieil7Haint (a) Specific Single Valve Wrongly Selecteil out of A Population i of Valves During Conduct of a lest or Halutenance Act IE No. I

• 10-2 " 1 I " 10-2
• l 10-2 " I of Valves les ruleselatlost at Cliolce) 25 2

20 2 R

• 10 R

10 (b) Inadvestently leaves Corsect Valve Isi Wa ung Punillon N5 m 10, .j .y 20 lib = 10 10 NIO 10 2. Hose thain one valve is affected NL a 10' 20 Hi a 10~ 10 N) m 10' 10 (coupleil es s on s) 3. litscalibeation of Sensuss/ Electrical kelays (4) One Sensur/Helay Affecte.1 145 m 10' ID N 10' 10 (h) H.are than une Senson/ Relay Affected l - Ni a 10 ) 3 10 - N) a 20 10 6

. - = j f i I Allit til-2 (Cont inued) lime Actu.ition Hecdod istim.ite1 failure Istle sicil f allusi: Ovceall l a t ism et e I i P ute. for PrImasy Psub. of atfier istin 1e Ireor fatIor Operatur to (Battisp) Contsol of f ailure esi necrall Actuate AfW5 lim. Operatur to Probability P o s.l...I,i l i t y Actuate AfW5 b. Act s & l'e rus s of a Post-An lilent flature l !) 8. Hanual Actisations of AlW system from Control j ltuum j (a) Considering "Dedic.sted Operatur .(H5 min. N2 m to H2-m 10 in .tn Actuate AlW bystesi and Possible HIS min. NI m 10 4 No.5 (mod. dep.) H5 2 10 10 i ( Dackup Actu.stion af AlWs .' N10 mlu. N5 m 10 H.25 (Iow dep.) Niu 10 4 ? l (a) ConsiderinD "flun-Dc. fluted" H5 min. HIm10,f N5 x 10[ 10[j N5 x 10 Opce at us-to Ac tuate.*J W sysicai Nih min. 3 No.5 (m<ut.

• p.)

HS y/o 10 arid rustible Battiep N10 mise. N5 m 10 H.25 (luw elep. ) N!n 10 Atutation ut AIW syste.e A 7 I l i l i 1 Y l i l 9 i l 9 1 i i

( Stucy anc frem discussiens vita rec:gni:ec u erts in ne fiela f nu=an :enavior anc retta-

llity at Sancia Lascrat: ries. The Sancia excerts are sresently -orung wiu %2C's Office of Nuclear Regulatory Researen on aw.an f actors-relatec r grams.

The test estinate cata in Tasle !!I-2 are suojec. to censicersole uncer*.ainty, and say nave er cr screacs of an orcer of e.agnituce on eit& sica of :ne cata. Hence, aluougn un cata

ase r.ay be usec to catain relative reliamilities., care must ze taaen in ascribing a nign cegree of nuerical precision to mese values N cr to results cerivec free their use.

Because of tnis, any relative values for AN systes reliacility presentec nerein snould not

e inter:retec s have a nign cegree of :recision. The cata from Tamle III-2 were acollec" ts the fault logic structure in orcer to Octain relative essoarisons of :eliamility of tne various A W systes cesigns. It was founc that'tne various AW system casigns cic emniett c:nsiceracle variacility vita regard to cesign a: reaca anc in their at.r.an influences. For u a=c i e, some aW syses incluce tarte feecwater punos (t-o electric motor-criven anc one steam turoine-criven), are autsaatec, anc no single :cint vulneracility was icentifies in our review. In c:ntrast, some AW systems have two Our::s anc are not auttaatec, uereoy having a strong cecencence on human influences for their :erformance. In accition, some of nese cesigns also have single point vulneracilities that coula ctantially negata the two train AW system recuncancy (e.g., a single manual valve). Clearly, one signt reasonably ex:ect to find a significant variance in reliasility tetween sucn casigns without having availacle an acuncance of cata of great precision.

4.5 kam-r of elissilitv-?ased Results (Gener'e) Figurt I!!-4 illt.strates ce results of the generic AW system relia ility assessment. as can :e seen from Figure I*:-4, prelfainary assessments of ne reliasilities of the AN system cesigns range frem nign to low. On a more quantitative casis, uis range ceciets cifferences in reltamility of the existing AW systams of = ore than an orcer of sagnite.de for each of the three transients consicerec in tais assessment. Eaca calumn in Figure !!!-4

ecicts the relative reliacility of the various system cesigns for a : articular transient.

Plant-scecific anc eneric rec mmencations to im: rove on and strenginen AN system rellasility were cavele:ec as ;4rt of this overall stucy ano are ;resentec in Section i of tais a::encix. These rec:rvnencations reflect the engineering insignts serivec tres this relianility evaluation as. ell as those cerivec frca the ceterministic evaluation. The

• ne cata -as a oliec to the various icentifiec faults in ne fault logic 'st ucture anc a :oint value estimate as ceterminec for the t o fault event (i.e., AN System unavaila-ility).

Suen an accreaca is consicered acecuate ta ;ain those engineering anc reliacility insignis sougnt for nts AW Systas reassessment. As notec, no attem:t was sace to cased intr:cuce ce somewnat time consuming, calculational elegance. associatec.itn tne =rocess sf er ar =rceagation i.mto this assessment (e. g., wante :arlo). Prior ex erience nita such a taiculational recess nas revealed a sciewnat ;recictasie outccme tiat, even.ita me very recuncant systems, c:ula te sligntly nigner inan the motnt value solution (e.g..

• ac. r of noor:xt: ate!v inree times nigner than ne :oint salue anc asually ;ess). ihoulc meet exist a t1early overwnelming fault in a systems :esign. een ne r: cess of tr?or

r::agation would De ex ectec *.o se merely one of nigner elegance anc it -culc yield no ic:ortant change to un cuantitative solution.

!!I-13

1 E. t-ol 1 i !

i t

1 i i, t ; j -= I 21 4 1 I i I i i j p ii il i e'=- ..e I I.'

j. g j-i 1

I i t ., ( i l ll. !

i !

i : j j 6 i - it: j 5-E 4 ' I. l: i 1 i i i e i - - - -.i, j i e i. . :, ! I I I i 1 i l I l l i, ; j l .j 7 l f I ' i ' i ' i E ! l4 3 E I l j i i i l l 3 1 - i I l' i g I i i i i : ! ! 4 i, s

i.,

i i g - _b i 2 s = I i i i i i i i i ie s eieie, ei e,ej : i ; j ,1 4 ji j,* r 21'i! I I I I i ! i i i ! l ieiei ; ;, ; ; i = a L. I ' i ' ' t i ! I I i t i i 1; jire a i i i ; i ; i j} ! i I I ; I i i i i ! i ; ieiei i ! i i : i, ? ,l .- - = a d' I t I I I I IJ ! I 'e' i ! i i : > 1 ! i ! jj j - i il 4 i,,,. i, ii. -? I i 1: 4 .- I i i i, I i i i i i 1j; l- -2 ..eio ....e.~i P 1 ii i ! I I i ; ! I i t i : i t i }} d i =. a i ! I IIiili I I I I I l l i I il i i i i i 2 i s

2 L

s v = ,e!..eie.eieeif*5 f [ ll I I I i i l I i l l l l l t i i h .f, I_ ~!! ! I I l I i l l l l l l l l*j l l i ; l l l l l ! l l 1 l l l l i ! l l ! l i I * ? j m l ! 1 1 I i ! l l l ? l'; ', i, i i i l [ s-7 i i! I i ' ret feie,ei ; - ' i l 2 l 4 l j I l ..o; e i i e t r 2 ei ei e t e.o.e e. I i, t ja i i e j l ! ! ! l I l l l l j l t i s > I l k 3 ,~; :-

-i

~i-.m

. I I. j l s I== l t m I 6 i ji a e ms-r 6 i 1 -l...f.t-,,.g 8 8 I I T ] ~.a:-.e. i- ,,,4 g l I i l l 4 .I ,i i, i t E u E-M' M M_ m

4 ( ec:mencations :erive:

• m the elia:ility evaluation ;enerally terc to recuce %.an er :-

otentials anc ctner

::inant failures, anc are :re:osec for all AFs systes :esigns as a: licamle fer all of ne AFV systes cesigns regarcless of netner uese esigns are enaracteritec as raving a tlatively nign er low reliasility. The taree transients usec in the assessment are cescr1:ec in :stail :elow.

a.5.1 Loss of *ain Feecwater (First Column, Figure III-4) Accrex1mately eign units were icentifiec as having AFV systems wita elatively Icw reliatility for tats transient. These AFV system esigns ;enerally *ecuire manual actuation anc incluce t'.o ;u: s in their cesign. Some.ere founc to nave single co1nt vulneranslities sucn as a single manual valve tnrougn nica al'1 AFW flow : asses (typically a saintenanca valve), wners human er or ;cssimility was generally f:unc ts :e tne ::sinant ::m:n soce failure ::ntribut:r. In some :ases, ceficiencies in Tecnnical S:ecificatices art ce

rinci a1 c

ntributer :: unavaila:ility; e.g., limits were not fe:csea on tse allowec autage inte-val for an AFV system train. Were sucn a ceficiency.as icentifiec, tne relianility of tne AFV systes c:ulc te acversely affectec if :ne of tne trains was t: Se allowe: u te inceertole for an extencec cerice of time. In general, fer tne eign; units Cnaractari:ec as naving a relatively low AFV system relia:111ty for tais transient, ne c:minant failure is tne f ailure to sanwally initiate the Are systen, riants r*cuiring =anual AFs systes initiatien art Ortsently recuirec, by recent II iulletins(9**"*) to pr:vice a :ecicatec incivicual.s manually acaata tne AFd tysten u en loss of sain feee ater. The results presentec in Figurt III-a consicer enly tne relia:ility of tais cecicatec incivicual to ac aate the AFd systes. It is likely inat, in :na event inat tais :ecicatec incivicual fails to perfor-a the AFs system actuation. :actu:.cule se provicec :y licensec react:r ccerators in the c:ntrol roca. Discussions =tta em:erts on numan rellasility incicate inat

~ the cnance of f ailing to actuata me AW system frca tha control room signt :e recucec :y a facter of t o to four by the tactu: c;erator, ce:encing on :ne ti e vinc:= availa:1e (see cata ta les). N :nis ;stential for it:r:ve numan reliacility.ere talte factorec into the Figure III-a results, taen etner ;otential f ailures, sucn as Me single valve vulnera:ilities, c:uld secome tne ::minant c:ntributors to the unavailacility of AFd systans. Inerefore, the net tenefit in AFV systes unavailacility =igat :e limitec to the sferementienec factar of two, unless the next level of cominant f ailure ceces.ere to :e in: revec u:cn. Ine :egree to wnica suca sue:essiva im:revements signt further (moreve AF. system reif a:111ty.as teyonc f,ie resent wort sc::e. Newever, e::.r encations art sace in ie:L cn i sf this a::encix that snould ir:reve taase next levels of ::sinating faults, sucn as tne single nanual valve. Inose AFd systas casigns inat c:Wic be cnaracteri:ec as eing of sectu:: relia ility ;ecerally were aut:matica11y actuatac ita sanual :acxu. no ever, single :ctnt vulners:111 ties =ere icentifiec wnien ouic list the eliacility. Ctner f actors, sucn as ne lact, af s:ecific list ations.:n allowec AFV train :utage time anc limitations sn AF. '1cw

• ate to tre steas ;enerators tecause af water-namer ::ncerns, ::ulc nave an acverse ef'ect n :ne AFw system *ecuncancy anc inus limit tne acaleva:le reliacility.

=:rovements in taese areas.ould serve to furtner i. creve reliantlity f inese AFd system :es1gns. I:!-2; E

r ...dk 4 A: roximately 15 units were enaracterited as naving nign AFW system reliamility. The'se AFW system cesigns were ;enerally of nign recuncancy anc nac no coservec single :otet vulneracilities. Consecuently, tne reliacility of taese cesigns would to excectec to se limitec by numan interactions that could acversely affect the installed nareware recuncancy. For examcle, some periccic tests of aFW systems are c:ncacted in ways that c:uld invalicate AFW system recuncancy. Usually, suca tests are not staggerec (i.e., esca recuncant c:econent of tne AFV system is tested my the same personnel anc on the sace snif t) suen :nat if icentical human errors were to ce mace on ricuncant com enents ine entire AFW systas c ulc se sace ineffective. The net result s that tne effect of these numan errers coula :erstst until the next test interval, at =nica time the errors snoule se :etectaJ. To recuce suca

otential vulnerasilities, rec:nmencations were sace for streng nenec acministrative c:ntrols (e.g., increvec valve locsing procacures) anc *consicerations are seing given for stag;ering tests of the incivicual AFW system trains, suca tnat only one train woule :e testec or, any given shift. acciticnal insignts cerived frem this evaluatien suggest inat the cuality Of periccic testing, as -ell as of the AFW system cesign, sncula ce imcrevec. For anacole, testing r: grams that inca:acitate more tnan ene train of the AFW systas should be revisec so that the :ericcie tests cemonstrate availasility of flew pata to the staam generators rather taan nenate the flow ;ata.

a.6.2 L.s of Main Feeewater Que ta Lass of Offsite Power (Second Colur.n Figurt III-a) The relianilities of the various AFV system cesigns for this transient =ere generally founc to te cuite similar to those for tne previous transient, i.e., loss of sain fetcwater. Onsite ac ;cwer sourtes were c:nricerec anc tne pctential incact of cegracing taese power scurces (e.g., the loss of one of tne two amargency ciesel generators) an the AFW system relianility =as estimatec. Cepencing on the AFW system cesign a.ic on tne ac :ower cecencencies identifiec, varitole -impacts were estimated.~ Mowever, these variations generally =ert not cosinant failure soces, and were similar to those previcusly cascritec for the loss of main feeewater transient. - - - - - - - - - - - - - 4.6.3 Loss of Pain Feecwater and Loss of All Alternating Current Power (Station ac Slacrout), (Thir Column, Figurt 4) this assessment carriec postulatec cegracation of the ac ;ower sources one stem furtner inan tne loss of main fetcwater ai; the loss of offsite power. All ac :ower sources were assumec l unavailasle, and the se cecencencies of the AFW system were exolorec. In generai, tne steam turnine-criven ;u=o of the aFW system =as tne only ;otentially :cerzole system for inis Some of the AFW system :esigns have only steam turoine-criven ;umos.

erefore, scenarlo.

these cesigns ;otentially nave greater availacle recuncancy for inis scenario. The relati,ve i reliacility of the various AFW systes cesigns variec by more taan an orcer of magnituce for tais transient. Seven reacter units =ert enaractert:ec as naving relatively low reliaciitties for tais transient. These particular AFW system cesigns cic not necessarily folicw taetr

rior cnaracteri:ations in Columns 1 and 2 of Figure !!!-4 This :1fference is cue to tte strong ac cecencencies =nica exist in the steam t*.. cine-triven tratn of tneir AFW system.

All seven units ceoenc on ac power to provice luce :i1 c:oling for tne steam turnine-tr ven III-22 b

~ (

ue

. Witnout :nis lu:e :i1 ::aling, it.as assurec ::a: :ne :c:a.ouis :ver ee: sne even-tually 'ali cue to snaf / seal /cearing 'ailures.

Est: mates :n.nen : o '31*cre si;n: :::.r vary, ut it as assu=ec to c:Or in a relatively sner time interval. As a ::nsecuence, those AFW systems naving this ac :ower cemencency.ere fuegec is nave a 1cw elia:ility *:r inis event. Mowever, it snculc :e notec nat :rel'minary esults 'rts a su:secuent tast 4: an ::erating :lant incicated that the effect of Icsing luce 011 ::oling ay act :n as

c as assumec in :nts evaluation.

. cst of tne turnine-criven ;uzes :f tne AFW systams use AFV flow to c:ol tne luce oil. Several AFW system :esigns nave valves tnat :ecenc an ac =c=er for ::eratien.

• n ::ntrast ts those AFV systems nay1ng ne luce o11 c:oling ac :ower cemencencies cescricec a:ove, these AFW systems can te suc:essfully oceratec':y sanually scaning tne <alves. Generally, inese AFV system :esigns are charactari:ec in Figure III-4 as naving a low-to-Tecium reli-amtlity. The nature of the valves' ce:encencies an'ic :o.er vartec :etween :nc :esigns.

For exam:le, certain cesigns.ere f unc to have ac ::eratec steam acmission valves cesignec to fail closec on loss of air su cly is the valves. Since, =1 loss of ac c-er, tne air su::ly to taese valves c:ule te cepletec in a: cut one-nalf neu.. ne c:erator oulc te recuirec to tame accitional manual actions to ree:en are maintain the acmissica valves ::en, until ac power anc/or an air su: ply caule :e resterec. Otner :lants also nave AFW system cesigns cnaracteri:ec as naving low-t -secium relia:ilities. 5 :n :lants ;enerally inc!.ce valves that are ce:encent en ac :ower. However, tne ac:ess ::ncitions are suen na :ney recuce tne likelihocc of sue:essful local sanual ac: tons ceing taren. Scme casigns -ere also enaracteri:ec in inis low-t secium relia:il':y range :acause no s:ecific limitatiens existec :n the allowec train cutage times, a f ac r that Ft: resents an ' cetant c:ntr?:ut:r to the AFV system unavailasility. These AFV system cesigns that were enarac*ert:ec as naving a relatively algn reltactlity for this transient generally nac no icentifia:1e ac :o-er ce:encencies anc -ere auto-zatically actuatec. For taese cesigns, ne :cminant 'ault c:ntri utors.ere ~ tnose associatec with nareware f ailures,.nica coulc not te rectifiec in a :1mely.ay :y tanual act;ons. a.7 telia:ility Charic eaitatiens of ArW Svttees *n 31 ants Ustae ~ es:- e euse-Oes e e: Geset:rs (: tant-! ee* *4:) Figure !!I-5 naracteri:ss the results :f LMis relia:11ity assess =ent of :ne AFW system

esigns in c:erating :lanta using 1-:est;nec react:rs.

he ::eratir; 11st:Py :f :nese

lants e: resents a cu=ulative ex ertence f a: cut !!O eact:r fears. As as :een :iscussec revicusly, the AFV systam :esigns with Icw reite:t*'ttes #:r :ne

• css f ain 'ee: water trsnstant.ere ;enerally ::sinatec :y numan er :rs 'n :ne tanual ac:ca::en :f :ne AFk sys*es.

or :y errers associatec alta single sanual <aives 'n tne systes. Those :' 11gner t H 1:'I'*y .ere also generally =cainatec ey human influences,.n:ca ::wic afd e: :re ecuncant is:ect

f the ATV system :esign.

n ;eneral :ne sa:cr* y :f inese aF% system :es;;ns are f a

nf t;uration nat incluces tnree AFW v==1ng systars.ni:n are :: swine ssie :: single

oint ' allures. Thertfore, iney.ere assessac is :e :f e:ativety it;n -ti'!:il' y.

".'.'.1 1

l 1 I I l I I I I l 3 i I f 1 I i l i I _11 1 I i l i I 6 1 I I i i ! ! I 1 I i 3 ~ Y i ! i ! !"i I,i ! I I I"i i ! ! I ',i.l 2 I I I i l 1 1 I I I I i i l i ! - ! lll 3 I ti i I i ! i 1 I i l I - I = e ,ll a g. 3; i 1 - i i iii E. 1 i iii i l i i i i i 1 i i s I j l i i l l i I i i i ! l i ! i i I 5 w j? ! ! I i ! 6 t l 1 1 I i e i ! I s i = 4 i i i ! ! 1 ! I I I i ! i i i 3 5 g j i l i l I i..._ i i i i f*i*lelete!*ieiI :. jil ; ! ! I I i 1 i i i*l*I ! I i ! ! ! {! 2 3 i i l i I l I ! l.i i ! I i ! I i !.5 i. E E 3 I i i i t letel i l i l I i i l i l{ - C - ] }l i l i Ie! I I i i ! l I I l l 5 lh i ) a ? .._,i.... i i i ; .e i i i i lg ; g g

• • i ' ' t i i i e i l i t I i ! l l3 J 3

l, jl l,!. 3-r a i : I I I I I I i i l i 1 i l 0 il i ' i i i I i i l i l i i l ! I I ! }l j j i-

1 a i : ' t i I a

i_t 18:*i*:*: *!*t*: ] ~ii _idI ! l l ! I i ! l'i 'l'l t I l l l 8 .i. i_ (,l ? $ i i i l I i I I i i t i i l i l i i i ! ! I ; I-d = _,y ! i i i l i i"i ! I j !l l l 1 !ei ,e. ! ! I I l ! I i i ' I g a 1 i : I i!,i i i ; e I i i e I 3 i j ; e: 8! I i ! i l I i ! l ! ! i I = l _l ij i i i i ! i l i i ! I i i i i ! I f '4

i 1 l

i l i i I I i i ! i } 5" i l ~ I I l? 1. j ~ 1 a I e. i i a 1, 2 c, =. 6 a r 2 a j i l = ~ m o o e S e -,,------,m---e en- ~., - - - - -, ,---ee_m ,-,--,,,,,-we ,en,, ,-,n_,,-w,-n-,,-,,,-,-m-n.n,-n,n----,v--,--,,n.,w---

t In c:nnection sita tne ?:ss :f sain fecewater : snstent inc :ne *:ss :f at: tc :c e. re*e is a significant varia 1:n in ne reliacilny f e var'ous AFw systam :est;ns. Inis variation is at*ributac to cifferences in :ne cesiq1 of De ATV systems :rev1cusly ::scussec. Those plants naving an ac ;cwer :ecencency (e.g., luce oil c: cling is :ne steam tur:tre-criven AFW pumo) nave ne lowest ATW system eliacility for u1s transient. Ac::rcingly. ree:mencations are sace to eliminata :ower cemencenet es.nlen c:uic result in :ame f atiure .itain a sacrt time interval. Our assessment of eaca :f ne lants listec in Ngure I:!-5 is cescrites in A :encia L - The reliasility assessment a: reaca usec anc tne :rinc::al insignis anc resub are summart:ec :elow. The results snewn in Figures : I-a anc !!!-5 incicate :nat ne relia:ilities of Oe existing AFW system cesigns vary :y at least an Or:er of magnituce. The cominant c:ntri:utcrs in :nis varia:ility in eif acility -ere, in generai, numan er ers anc single point vulnera ilities, as cascritec later in tnis 4:penc1x. 5. SumARY OF REIULTS AnD REC;WE.CATICMS This section arisents in sumery fers the results of :ne AFV system review anc rec:m-sencations :nat snowla :e teclecentec to incrave :ne serf:r.ance anc reliatility of the AF= systems of ne varicus ycesignec coerating slants. A :encix X provices a se:arata AFW system cescr1 tion, evaluation, anc rec:mencaticas f:r esca incivicuai :lant. 5.1 :ee:mencatien 04 tee: ries The rec:mencations are categert:ec as ;eneric anc :iant-s:ecific, as.e17 as snce -ter 2 anc 1cng-tarn. The generic rec:mencations (cesignattens 33 anc ;L refer :: generic snort anc iong-ters, res ectively) are a result of similarities in AFs system ;ctantia: ;r:clems tet een plants anc are ar.clica 1e to more : an one plant.

• be generic rec:mmencations anc ce :sn-co.s whica lec t: :nese rec:mencations are :escr1:e: in uis sec::en. There are alsc ; tant-t s:ecific ree:nmecca:1cas cat are unicae is a ;1ven :iant's AF= system. The slant-s:ecific l

l recomencations are accressac : ort fully in ne incisicual ;lant evaluations in AC:encia L Se snor.-term rec:mencations e:rtsent acticos :s :: r:ve ATV systam eliasility nat 1 sneuia te imolementec ey January 1,1980, or as scon :nereaf ter as is ;racticacie.

• .n general, taey involve u: gracing of Tecnnicai Icecificatices :r estastisning :recacares is avoic se sitigate :otential system :r : erator *111ures. The icng-tarm rec:: enca:1:ns involve systes cesign evaluations anc/or iccifications :: ' ce:ve AFw' system eita 11t*y anc

( remresent ac-ions t3at snould :e tmslementec ey January 1. ;M1, :r as socn wereaftar as is rscticaole. This implementatica scnecule is intancea :: :e ::nsistent v nn ce sc.t w ie for implementation of tne. eevirements s:ecifiec in NUREG-0578 ().

f ::nflic s secula

~ arise; :ne senecule s:ecifiec 1n NUREG-0578 taxes stece ence. t I There are t.c significant limitations of :ne AFW system *eview anc evaluatt:n -r.1en saoul:

e 9ctec. as. ell as neir affect :n :ne te:mencattens.

.g.. P L

( *. ;.9t?e :ur review c:verec ine classification anc civistonal r-cancancy of :cwer 5:urces 'cr 4FW system acui: ment anc instrumentation anc c:ntrols and tne ty;e sf instru=entation anc ::nt.ols =rovicec for tne overall 4FV system, we cic act atterat ta review cetailec Icgic anc c:ntrol ciagrams. This exclains in part tne c:nservative acar aca =e usec in acolying to all plants the snor anc long-tars generic *ecommencations GI-7 and 3L-5, nica coal witn non-recuncant anc n n-C! ass li circuitry for AFW systes automatic initiation systems. (*) The review is not c:nsicerec ts :e a c:poiete evaluatien of :cstulates sign energy 01:e =rearsetnat ::ule affect tne 4FW system. since pi ing isometric anc slant arrangement crawings wert not revie=ec. newever. =nere system flow sneets revealec

tential pine treats tnat coule cause :tal loss of 4FW system ca:asility, nese

rcolas areas nave teen f

entifiec anc incluces in :ne long-ters rec mmencattens f:r furtner evaluation.

5.2 Sher e-m Geme-se :ee:r-encations

5. 2. 2 Tecanical Specification Time Limit on AFW System Train Outage CONCP*9 Sever?1 of :ne plants reviewec nave Tecnnical Scecifications tnat :ersit one :f tne 4F'=

system trains :: De out of service for an incefinite time :ertec. Incefinite cutage :f one train recuces the cefense-in-cestn provicec Oy multiple AFV system trains. ee:meeaestien 03 The licensee snoulc promose accifications to the Tecnnicai 5:ecifica-tions to li:1 ne time anat one 4FV systes Dumo 'anc 7ts associatec flow train anc essential instrumentation can as incoeracle. The outage time limit and sucsecuent action time sneule te as recuirec in current Stancarc Tecnnical 5:ecifications; i.e., 72 nours anc.*2 nours, res:ectively. 1.2.2 Te:nnical Soecification Ac=inistrative Centrols on Manual valves ':cx an1 ver1*y Pesition ~ ~ ~~ ~- 0: ace-, - Ieveral of the plants reviewed use a single manual valve or multiale valves in series in :ne ::nmon suction at:tng :etween the rimary.ater source ano the a Fs' system :u o At s me plants tne valves are !ccrec omen. -nile at stners, they are tot ':cnec in sucticn.

ssitt:n.

f ne valves are ina:vertently left closec, tne 4FV system.ouic :e 'accerscie.

ecause tne.ater sucoly to tne :u=:s would te isolatec. Since there is F.o tmote selve,

=csttien incicatico for these valves, tse cocratcr nas no immeciate means of :etermining dalve :Os1 tion. Curtner, ine Tecnnical 5:ecificattens for slants with Iccmec-::en tanual valves :s not recuire :er*::tc ins:ection is verify nat tne valves are locxec anc in :ne ::r ect

-25

{ For sost slants =nere tne valves art o; locsec c:en, valve ;ositten :s serifiec sosition.

n some periccic masts.

Kee:nrencatien GS-? - The licensee snoule locx ocen single valves or multisle valves in sortes in the AFW system pumo suc-ton piping an: Iccx caen otner single valves or sult:cle valves in series tnat c:ula interruct all AFW flow. Monthly inspections snoulc ce oerformec These ins:ections snoule to verify inat taese valves are lecxec enc in tne c:en ;osition.

e crecased for inc:r: oration into the surveillance recuirements of the plant Tecnnicsi See ieconnencation GL-2 for tne foncer-term resolution of snis ::ncern.

Soecifications.

5. 2. 3 AFV $ystem Flow Throttling-water nammer

- Several of the plants reviewed a::arently tarottle cown :ne AFW systes tnitial Concer9 fic= to eliminate or recuce tne ;otential for water na.9mer. In such cases, ine overail relia:ility of tne AFW systas can te acversely aff ectac. tectaneecatien GS The licensee nas statec inat it inrettles AFW systes flow to avoic The licensee snould reexamine the ;ractice of tarettling AFW system ficw ta =atar nammer. avoic water hammer. The licensee snould verify inat the AFW systes will su:oly on cesanc sufficient initial flow to the necessary steam ;enerators is assure acecuate cecay neat removal fc11cwing loss of eain feecwater flow anc a reactor tric f rom ICC: =c er. In cases =rere inis eeva*wation results in an increase in inttial AFW system flew, the licensee snoulc provice sufficient information to :emonstrate that the recuired initial AFW system flow vill act result in plant camage cue to =ater massner.

5. 2. 4 Emergency Prececures for Initiating Backup water Su: plies

- Most of the plants co not nave written procecures for transfer-ing to alternate

Ceneer, Witnout s:ecific sources of AFV sucaly if the primary su;;1y is unavailasie or exnaustec.

criteria anc ;rocacures for an c erator to follcw to transfer to alternate ater sources. One primary sucaly ::ula :e axnaustec anc result in :umo :amage or a long intarruction of AFW flow. Gee:--cacation 05 Esergency :roceoures for transfer-ing to alternate sources cf AFW These procacures snouic incluce :rt er a su ciy snoulc se avallasle to the plant : erat:rs. .nen, anc in.nat or.er. the rsnsfer to alterr. ate -ater scurcr s is inf orm the coerater: sncuic take place. The following cases snovic :e c:verec y the procacurts: The The case in.nten the :r' ary ater su:=iy is 9ot initially availaole. (1)

retect

rocecures for this

sse snoulo 'ncluce any ::erstcr actions recutree t:

tne AFW system pum:s against self-:amage :ef:re.ater flow is.nttiatac.

-27

(2) 're :sse in.nica ine primary water sucaly is eing :eoletec. The :rececure f:r ints :ase snoulo :rovice for transfer to the alternata =ater scur:es =rter ::

raining f tne rimary =ater su:aly.

5.2.5 ~:er;ency 7Pececures f or Initiatino AFW Flow Foilcwing a C:zolete Less of Alternating Current 7 ewer - iome ::erating alants :e= enc en ac :ower for all sources of AFW system su:;1y, 0: Pee-9 inclucing ine turcine-criven pu=m train. In the event of loss of offsite anc snsite ac

cwer, ac-te:encent luce oil sucoly or luce oil cooling for ine :urc will st:c, anc/cr manual actions are recuitec to initiate AFW flow f rem tBe tur:1ne-criven ;un: ty manually

ening tse tur:in. steam acmission valve anc/or AFW system flow c:ntrol valves. There are no procacures availacie to the plant c erators for AFW system initiatien anc c:ntrol uncer taese c:ncitices. This : uld result in a consiceracle tiee ceiay for AFw system initiation, since tne :: erat:rs woulc not :e quicec ey ;rucecures :ealing.ita tais event.

e:09meaca-te9 05 The as-tuilt plant shoulc be ca:amle of provicing tne recuirec AFW a

If ft:w fe' at least ; o nours frem one AFV pump train, ince:encent of any ac power scur:e. sanual AFW system initiation or flow c:ntrol is recuired following a c:molete loss of ac

c=er, emea;ency crececures snculc te estan11snee for manually initiating anc

:ntrolling tre system >ncer inese c:ncitions. Since One water for cooling of ine luce oil for tne tur:1ne-criven pu== tearings may be ce=encent on ac power, cesign or procecural enanges snall :e sace to eliminate tnis ee:encency as soon as gracticasle. Until this is cone, the t'e ;ercy rececures snould provice for an incivicual to :e stationec at tne tur:ine-: riven

ura in :ne event of tne loss of all ac :ower to pontter puso usaring anc/or lume oil If necessary, nis operat:r -culd ocerate tne tur:ine-criven ;uma in an taeceratures.

on-off moce until ac ; ewer is restorec. Aceouate lignting :owerec y cirect current (c:) =c=er sources and communications at local stations snould also De provicec if manual initiation anc c:ntrol'ef the AFW system is neecec.- (See Recommencation GL-3 for tne lenger ters resolutien of nis concern.)

3. 2. 6 AFW Systes Ficw Patn verification

acer9 - Periccic tas-ing of the aFW system is ac::molisnec y testing of incivicual

roonents =f :ne flow train (;ericcic :umo recirculation flow test or aut matic valve actuation), inus altering ine normal AFW system flow pata(s). The flo. :1:acility of the entire AF= system, Or at least one integral AFW system train, is :nly :emonstratec on systam

emano foilcwing a transient, or if tne AFW system is usec f ar normai plant startuo or snutcown.

t tecent Licensee ~ vent ie: orts inoicate a need to ir:reve tne :uality of system testing anc maintenance. Soecifically, periccic testing and saintenance grocecures inacvertently rtsult in (1) : ore than one 2FW system flow train ceing unavailaole :uring the tast, or (2) tne 47W system flow train oncer test not teing stocerly restorec is its oceraole ::ncition foitowing III-28 ,.--~,-_n- - ~,,. - _ - -,. .,,_,-.y,,. .-,,,,,_a, ,y,.,_...n,n.,_.. ,,_.g.,y,..

tre test or aintenan:e. ort. 'he Of*ica of :ns:ecticn anc nf:r:seen as tamen actt:n ::

r-ect Itas (1); ce ec:anencation :eio. is sace u ::r ect has ( :.

Kee:-reacatien 03 The licensee sacula :snfirm f'ow :ata availasility :f an AFW system flow rain :nat nas :een out of servi:e is :erform :ertecie testing or :alntenance as ' follows: (1) P-ecacures sncule e imslementec to -ecu:re an ::e tur u :etemine that tne AT= system valves are :recerly alf nec ano a sec:no scerator to :nce:encently vert *y inat ne valves are recerly alignec.. (2) The licensee saoule =recose Tecnnica'l Sce: fications t: assure cat. :rior s

lant startuo *silewing an extencec uld snutcown, s

• low tast.cule te erf ormec to vert *y ue orsal ficw =ata f re.m ue =rimary AFW system.ater scur:: : ce The *1cw test snoule :e ::ncuc ec.ita AF4 systas valves in steam generat:rs.

neir ncrsal alignment. 5.2.7 men-Safety Grace, Non-4ecuncant AFW System Au*. matic *nitlatian Signals Cancem - Some :lants ita an automatically initiatec ATd system utili:e sc.* initiation signals tnat are n:t safety grace, co not meet me single failure criterien, anc are not rN uirec :y ne Tecnnical icecifications to :e testec : era:cically. !nis can fesult in recuced reliatility of the AFW system. see:-rem::::ien "I The licensee snoulc verify tnat ce aut:-at:: sur.1F4 system signals and associata: :ircuitry are safety ; race.

• f tais :annet te veriffe. tee A system automatic initiattenl ys em snowlc ne socifiec in ne sacrt-term : teet n e ft,ne:1:nal recuirements listac : ele. For the longer-ters, ce suu=atic intitation signals anc circuits snoula :,e 1: graced to meet safety grace recuirements, as inc1:atec in Rec:mencation GL-5.

(1) The cesign snould provice fer tne aut:mati: initiati:n of me F. system **cw. (2) The aut:=.atic initiation signals anc :ir:uits snoule :e :esi;ne: so eat a single l f ailure will not result in tne loss of AFd system function. l l (3) Testa:ility of me initiation signals anc :treuits snail :e

• enture :f ne casign.

(a) The inti.iation signals anc :irtuits snoule :e :cwerec *r:m n e erer;tecy :uses. (-). Manual ca:asility a initiate ine AF systam ' em ue ::nt :1 :cm secula :e retainec and snoula e imolementec so taat a single *sikre n ne manua; tirtutts .411 not result in tne loss of system *une-ton. ...,a. ..a :

t (5) The ac mot:--: riven :umos anc valves in tne AF. system saoulc te inchcec in ce sut:matic actuation (simultaneous anc/cr secuential) of tne loacs to me emergency

uses.

(7) The aut:matic initiation signals anc cir:uits snall :e sesignec so tnat their f ailure will not result in tne loss of manual ca:acility to initiate the AF4 system frem tne ::ntrol roca.

5. 2. 3 Automatic initiat'ier. of AFV Systems Cerce 9 For plants vita a manually initiatec AFV system, there is tne notential for failure of tne coerat:r tc sanually actuate tne systes' following a transient in time to maintain tne steam generator water level nign enougn to assure reactor cecay ne tt removal via tne steam generator (s). ' nile II fulletin 7c-C6A ecuires a :ecicatac incivicual for rcesignec coerating clants =ttn a sanually initiatac ATV system, furt:er action shoula :e taten in tne This c:ncern is icentical to Itam 2.1.7a of NUREG-0578.C) sncrt-tars.

3ee:vecation GS The licensee snould install a system to automatically initiate AFJ system flow. This system need not te safety grace; however, in :ne snort-term, it snould ) For teet tne criterla listec telow,.nien are similar to Item 2.1.7.a of NUREM578. tne longer-term, the aut matic initiation signals anc circuits snould :e u graced to meet safety grace recuire:nents, as indicated in Rec:meencation R-2. (1) The oesign snould provice for the autczatic initiation of the aFd systes flow. (2) The automatic initiation signals anc circuia snoule :e cesigned so inat a single f ailure will not result in the loss of AFd system function.~ ~ ~ (3) Testaoility of the ' initiating s~ignals~ and cir:uits snoulc ce a ~ feature"ot tne ~~ ~ ~ cesign. (4) The initiating signals anc circuits snovic te cowered frem the eeergency :uses. (5) Manuar ca:acility to initiate the AFw system frem tne control rece snoulc te retainec anc snowic :e imolementec so tnat a single failure in tne manual circuits '~- will not result in tne loss of system function. (6) The ac mot:r-criven :umos anc valves in the AFs system snoule :e inc!uced in the l automatic ac.uation (simultanecus anc/or sequential) of me loacs *: ine emergeocy I tuses. j l ( The aut:matic initiation signals anc cirtuits snoule :e :esignec so inat Seir t (7) failure -ill not result in tne loss of nanual :acamility to initiate ne ATV system fr m ine ::ntrol reos, t I i !!*-10 .--.-.y---- .,,_.,--.-_,._.o,

1

5. 3 :ect :?:r al I.*e-t

• - :ee:.eacatteas he 'tilcwing acciti:nal sner -tarm ree:mmencaticas resultec ' rem :ne staf*'s tassens

' earnec Tasa F r:e review ac,c ne !ulletins & Orcers Tass Force review of AFV systees at la:c:ct & Wilc:x-:esignec erating plants, su:secuent to cur review of tne AFW system

estges at h-anc *-i- :crignec : erating :lants. They 9 ave not seen examinec f:r s:ecific 4: lica:ility c incivt:ual.-Inc 22-tesignec ::erating plants.

1. 3.1 Pri=ary AFV ' ester Sour:e tew Level Alarm

• -ce 01 ants.nica :s not have level incication anc alarm fer :nv primary -atL source say not ;rovice One o: erat:r.ith sufficient information to procerly c:erate tne AFW system.

see:,mencatfen - The iicansee sneule :revice ecuncan: level ircies::en anc lo. m.el alarms in the : ntrol r:cm for ne AFW system artmary water su: iy, :s allcw ne ::eratst to antic:: ate the neec :: name e:.ater or transfer to an alternate.ater su:31y anc revent a Ic= u=c suction ;ressurm c:ncitten f rom =c:urring. The ic= level alarm set:cin: saculo 411:w at least 23 minutes ':r ;cerator ac-ion, assuming tnat ne largest :2:act:y AFW pue: is :: erat'ng. 5.3.2 AFW Puzu Encurance 7;st .- Since it may te necessary to rely :n tne ATW system to remeve :ecay Seat for .ece*9 extencec cerices of time, it sncula te cemonstratec :na: :ne AFW :ue:s have tne ca:a:ility for ::ntinuous c:artti:n Over an extencec time :eric witncut failure. leg,:- e-cati:n - The licensee saculd :erform a 72 neur encurance test on all ArW system 'f sucn a test or ::ntinuous perice of oceration nas not teen ac::::lisnec ts cate. ue s Following ne 72,nour~;uco run, tan sum s sneule te snut cowe anc c:oiec : wn anc **en restartec anc ri,n for one t. car. Test ac:e:tance criteria snoule incluce cemens:rati g nat

ne :ue:s rematn =itnin :esign limits wita res;ect is nearing / tearing i1 tee:eratures 2nd i

vitrati:n anc inat ;um: P:cs 1-:ient :encitions (tem erature, ng=icity) :o not exceec environ- \\ Sental cualificatica limits f:r safety-relatec e:uiement in :ne reem. 1.3.3 Incication of arv ;iew to tne Steam Generators C:-c e a, Incication f ;F'= 'icw :s ne stats ;enerst:rs is ::nsicerec trecriant is the j manual enquiation f AF= '!:n :: saintain :ne e:ui ec steam ;enerater.ater level. This <e j

ncern is icentics7 ts :ta= 2.;.7.: cf wuREG-Oi78. --)

• ee: cencat* n - The licensee snoule ::: event tne ' llcwing tcuirements as stect'iec Oy

tam 2.1.7.: en : age a-32 f-903EG-0573:E~2I (1) Safety grace f ret:sti:n =f AFV **.e.

esca staam ;enerat:r secuic :e :r:vicec in the ::n to: :cs.

9 w-- ..y -~__~v,,.,. .-.~.._e ..e e,,- _ ~

{2) The aFW flew instrument enannels snould te :owereo frem the emergency :uses

nsistent ita sattsfying the emergency ;ower iversity recuirements for tne AFW systes set forth in Auxiliaqr Systems Branca Tecnnica> Position 10-1 of the Stancarc 2eview Plan, Section 10.4.9.

5.3.a AFW Systee Availacility Ouring Periccic Surveillance Testing Ocnee-n - Some :lants require local manual realignment of valves to concuct periccic ;u=o surveillance tests on one AFV system train. 'enen suca plants are in this test sace and there is only ene retaining AFW system train availaole to ressono to a camanc for initiation of AFW systas ::eraiton, the AFV systas recuncancy and acility to vitastand a single failure are lost. trf ee-cencatien '.icensees vita plants =nica recuire 1ccal sanual realignmer.t of valves to c:nouct periccic tes s on ene AFV system train anc wnica have only One remaining AFW t.atn availacle f:r coeraticn shoulc preoose Tecnnical Soecifications to :rovice tnat a decicatec l incivicual wno is in c:neanication with tne control rocm te stationec at the manual valves. Usen instruct;on f rem the control roca, tais cocrator -ould re-align :ne valves in :ne AFV system frca the test noce to its coerational alignment. 5.4 tere-Team "eae-s e Recenmencatiens 5.4.1 aut: matte initiation of arw sy,te,s C:neen - This concern is the same as snort-term generic recenmencation G3-!: namely, failure of an c:erator to actuate a manual start AFV system in time to maintain steam generator =ater level nign enougn to assure reactor cecay neat removal via the steam generator (s). 3ecenreacation GL For plants with a manual starting AFV system, the licensee snould install a system to autcmatically initiate tne AFW system flow. This system anc associated aut:matic initiation signals should te designec and installed to meet safety grace recuire-ments. Panual AFV system start anc c:ntrol ca:acility snoula :e retainec with manual start serving as sackus to aut:matic AFW system initiation.

5. 4. 2 single Valves in the AFW System Flow Patn

rcern - inis c:ncern is tne same as snort-term generic ree:nmencation '3-2 -- namely, AFV system inoceraciitty cue to an inaJvertently closec manual valve that could interrust all AFW system flew.

ee:m=eneat'en IL Licensees eita clant designs in =nica all (;rimary and alternata)

=ater su:alies to the AFV systems pass tarcugn valves in a single flow =atn snould install recuncant :arallel flow patas (pising and valves), i III-32 i

• .icensees.ita :lant :esigns in.nten :ne :rimary AFa system.ater su::iy : asses inrecqn valves in a single fles :ata, :ut the alternate aFW systet -ater su::ites ::nnect : tne a r-systas :umo suction piping :cwnstream f tne scove valve (s), secula 'nstall recuncant va!<es

arallel to the acove valve (s) or provice aut:matic coening of ne valve (s) fr:s tne al*ernate =ater su aly 6:en low :umo suction pressure.

The licensee sacula precose Tecnnical Soecifications to tec:r: orate a::recriate pericesc ins:ections to verify tne valve ;ositions ints tne surveillance recutrements. 5.4.3 Eliminatien of AFV System Cesencency on Alternating 0.rrent 2:=er Following a 0:molete Loss of Alternating Carrent 2ower. C:neeen - This concern is the same as short-ters generic rec:emencation 35 namely, celay in initiatica of AFW system =ceration er maintaining AFV system ::eration f:11cwing a

cstulatec less of ensite anc offsite ac pewer; i.e., ac pewe? 31a

1:ut.

Ree:mmencation OL At least one AFW systes ;umo anc its assectatec fic- ;ata anc essential instrumentation snoald autceatically initiate AFW system flew anc te cacaole of teing coeratec ince encently of any ac power sou*ce for at least two nours. 0:nvertica of Oc ;ower ta ac pewer is ac:e::Acle. 1.4.4 prevention of Multiple pumo Casage Ove to L:ss of Suetion Resulting Frem Natural stencmene Conce-n - In many of tne coerating plants, Me normal water su::1y to the AFW system sum:s (inclucing the interconnectec pising) is not protectec fr:m eartacuases or tornacces. Jay natural ;nenemenon severe enou;n to result in a loss of tne =ater sa: ply ::uic also se severt enow 1 to cause a icss of :ffsite sewer with loss af sain feecwater, resulting in an aut:matic tr.itiaticn stqnal to start the AFV system zumes. The ;umns ecule start witncut any suction neac, leacing to cavitation anc multiple :um: :amage in a sacrt ;erioc of time,

cssi

1y t:o snort for the scerat:rs is taxe action tnat.oule :retect tne :umes. This may ieac

• .s ana::::trole ::nsecuences f:r some plants, cue t: a c mslete icss of fecewater (natn and auxiliary),

tec:-maacatien - L-a - Licensees 9aving slants vita uner tectac normal AFW system.ater su:clies snoule evaluate tne cesign of ineir AFV systems :: :etermine if autcmatic :rstec-sten of tne ;umes is r.ecessary fs11: wing a seismic event :r a :r acs. The t;se availaole

efere :umo camage, the alarms anc irc1:stions avatiaale :: ine ::ntrei :cm c erat:r, ano tne time ecessary for assessing ine :rcoles anc taning actica secuic :e ::nstcerta in

etermining =netner scerat:r action :an :e reliec sn is ;revent :uco :amage. C nsiceration saculc de given o srevicing :umo :r:tection :y seans sucn as aut:mati: swit:never :f ine

umo suctions to tne alternate safety grace source of.ater, aut

matic :wmo tr*:s sn ic=

suction pressure, or u gracing tne normal scurce of ater is teet seismte :atagery ! anc tarmaco :retection recuirteents. 1-er e ,-r y. e, y-wy- ,,,.,rer.----,m,e,-,.-,,-< ..wm em,mv. w yw--e, .-----v --%v-,,y -%w y+. - -.+ -

A* s 5.5.5 Non-Safety Orace. Non-tecuncant AF4 System Automatic :nitiation Signals Sneem - This c:ncern is tne same as snort-term generic rec:mencatica 03 namely. recucac AFW system relianility as a result of use of non ssfety grace. non ncuncant signals, wnica are not ericcically testad, to automatically initiate tne AFW system. seewencatien R-! - The licensee snoulc ungrace tne AFW systas autssatic initiation signals anc circuits to meet safety gract recuirements.

5. 5 81 ant 5:ecific ars syste. :ee:: wnestiens The sacrt-tem anc long-tem ;:lant s:ecific reccescencations acclicaole to tne AFW systems for eacn plant are icentifiec anc ciscussec in'Accencim X.

5. 6 Su:r arv of are Svstes tee:-rseneatiens 'er ' estimeneuse-Cesienec he-niec 31 ants facle III-3 teicv stasarites the sacet-tarm anc icng-tam generic anc ::lant-specific rec:w:encat1cas for the AFW system at eact reasignec ::erating reactor. The accitional generic snort-ters recoarnencations discussec in 5.3 of tais a :encia are net inclucec in Tacle I!!-3.

newever. tnese recorrnencations are inclucec in tne incivicual plant AF4 systes evaluations centainec in Accencix x. .g m e.ee e s 34 4%,.- - - y. -.4 y 3,_.. -... y ,,y, -,_---%_,yr_+.

i f i t .m ,/ 1 I431e * !-3 (attached) Nw ] [. 9 i e I k j 4 h G h i yL e 1 I ) J I. r m 0 0 = = - m h

• • T.*t 464..

e m ew-r--- www w - e ,,+we-see-aweew,,,.we-e.v.__ pm eg, ....,..,,my-,y wpgyem*yg m-a - + -r m ww i.ws7% -_m---w,wmy g ,ww w srwwy-,- yvy-vtv

A h, 4, h j ( ~ ~ -l? 4EFERENCES i 1. Stancar Review Plan (NUREG-75/CS7), C.Santar 10.4.3, ano 3ranen 7ect.nical Position ASB 10-1 2. React:r Safety Stucy (kASH-1400), Cc ooer 1975. 3. .memorancus, 5. Levine to M. Lewis, cated June 27, 1978. 4 tisx Assessier.t Resiew Grows Aeoor*. to the U.S. Nuclear lequlatorf Ca vaission" (the " Lewis Cemittre Report"). NUAEC/C2-04CO. } 5. Comission Policy;oli Risk Assessment. i. lcentification of Uaresolvec Safety Issues telating to Nuclear Dower Plants, % KEG-0510), s Pacer,.Iresentec at American Nuclear Society sy S. Asselin, J. Micx: san, et al,1978. 3. 3. IE Sulletin 79-C,6A.- 10.

E '3ulletin 79-C68.

a f t I I I t ~ - 35 ..~-...,r,..- y ,c.m-y c ,wy,,-.7%---,-,,m.,,,,,,_y,,,--.,e m,-y,,m.,,,,%.-,,,p_, ,y-.-.,,_ ,c-.w._, 1-wr- -n

Table 111-3 (IO Cee eric itero acn.f at ions 1 E Plant Specific Recommemlations E [ c 1. 2 ? 7 b b' O 1; 8 5 1; c I a t 37, 0.. '. - 2; d.p "i ! 7.

a Y

. r. a a r .r y-c PC Ea t A p *t: O .,A 'i

• g %>, p

?. 'n

8. **

x., a

1. L D3 5 24 Y I: 3*.' t#

l' 2

A; a c> r> e-

L y

.; Et ah ;u 2 L* ;'

u "

~ a> a: s~ an'g'I e, L u d3 e; :: d !r r '.~. t ; i' ?. f: ?: r, ?.

5. 1 5. *

?. * ?. i' ':! a 5. Ts ?f ;' jl 7 f '<' ? TJ " '. 7t ?f ?a f.y '.(j f 3 3 3 3;3 3 3 3 J d d, 4 d Plant Short lere tong Tern Short fera long Term Seaver Valley el I E N N 2 elect pumps N -Hodify Tech $pect to require periodically -Evaluate postulated AfW5 distharge 1-turtelne pim.p testing and verification of position of Automatic manual valves from Alver Water System to pipe break concurrent ulth single. Inillation AfWS. active failure to determine AfW Inlllation -Implement leproved procedures for loc 61ng systee mo<llfications or procedures necessary or descrit.e f nw plant can manual valves and steeger testing of AfWS t.e safely shutdown I,y use of oti.er pep s. available systems. -Re evaluate ell nment of AfW pump discharge 0 - block valve to assure required Af W flod for normal, transient, and accident plant conditions. D.C. Cook l&2 E K E X X X X X I elect piap* I tusblue pint * -Complete licensee proposed maJlf1-per unit cellens to eliminate turbine piap 8 (elec t p eps train depesulence on AC power, s h a r'ed) AutemetIc Init at1 n Ierley #1 X X u a u x 2 elect pmaps I tur ble.e pump Auto ntic In i t'l al lan s I t e i I l

tL s t s I t e e y n n l f r rc o t e n - i i 1 au en dtd i v o aal t n t t e e nnl e nof icrl la gl e s ana ad e w a e !. 1 t o t a ar wn nal roat n ,c t f e aoi s rye le=t pe nl n ep ef ot pt d m netl I. )l ae d e l i j i l I' t as as ia al c m pl ( awclini u i a I. ng v s If i s 1 er f io i. m ar o l r e g e. lei i a r ot f el di os af s s. l. r u onnf nwt e no of ee p o. ae a 1 t npd sif t ~ l nl gl od n coa n ol e ,l w a t s m ol ot t rt na emca pr r e w t nadt 5 no5k rt w lesc r ti i a eio a t m oue eid t l t e T y ai eiaWl p e cwa e o aa ar f t s umt vrf al p f e ep laht s uumu g i r l r soeA sraw oi i i nA r. et eg os n l u ae rn t eaC o f h e e G8 t e s o vt Jpe) e t d . s nA vt ue l la cay 5 np

i. k neeni er as eapr emn a

p p2. rill m(Ii a rt t rl a p loe iucio rer e creS

s. uaed-.

s s 5 plil i od e W il ee S .f l ow ms ct n e f da p a s ef oeosis a s J. d eHnf eo.n.eli d n l, a t t egt tue o. t gt t et ea o ef l ay eur at)l i et t wt ae dne ulCl apmt i id s ees wl snil n t l pA p s eoa( s o ili hl pnel p w. nopo e.) d ap e idt l s o vovtt od vunoo unfil es eic e i e eoil .l n n I soCf aaiomf RiR waftls tC p1f a 8 u mo cca c e i s il f p mt i o u o. c t ps

p s

3hd 1tl L, - ou .b o. n l a i 5f isr r o e e 4 w e rh1 o r ut - p e . d a 1. l T _ ep c 5a t ow r roer o pl r o h f ut s t dc s gea ) W ence ti or ( k r 1 y c ple l o l l h l, ti l sw t1 e n I oetd e mt e i aym l enl r l siho a T imtf vinr tel oe i eMp l i .;I,. 't 3

• 3 % f,#<

t T4.o ' # * : L 2 ;;.,. l a _7 {c ;' ! 713 :t

• t'.

r h .:52 C!'

• 2 *' 2 ;.,2 a

r c T.b. t 7/310o fe T .t V .4' ! 7 4.,o X g X n l ; 311 o l ia g' * ' c : 4.,, X s_ y <,o n, 5. V.53Sf 2.,2 o i t I ,h-y<. s e# i 53S!;.3 5. X e. . ;[,. 'l0" 3%!#4 t 14.o

• 2!3. L! c.3 X

n t e E c" %2 .d* !: ;.,2 X . X c i

% ;;.i 3.,<

r b3tS ;., m R X .* 3. t ,.n;c#% o - t r a e i C. g 3a

s. E g. 3. 7 1,.

t X r .).' t%2 o t*

7.,. ", a2 ' 0 4 4.1..

t K S 3; 3Ea .* & c._ c' I 3' 5 5. n.,s. X .. I!, I r-Oi. K' f.

r. 3 X

.y2* c s. Is s p p a m pi p u . p nN im k p n n c n 5 p e c oA p o o W le e i niiTt i l i t n Altl t t t e l e cb s aE c t a ml t a es e u o tMr l si eai ai n.l n nnl nt

l. uut a

a i et t i aeni dL ni l l a unt an a - a n P GM2IAIi2HI 11 2HI ,1: ,i!4 ,ji;!)i 3j.j 2!ij 4]2!l!}j{j4 ] j;j!l1;}4 } 3 ?ltl. 1 42i i .'e l

=.- ~ _ _ ._.:.-..~, - y 3 m h ao 1 e an e e e .e. e. =- e: h e o g =ce-T e 3-a e ee -., e .e . 4 e* p =* 2 v e e.=. =w r ee e ee. o w - 3 c. s: v o.u. e

== = ee =< 1 -.* e t,.* c,. v we C a "E e.o = .o e..eo = 3 e e E

== m== c ** * .E o =* 6 = C u e3 e3 e eu EeweC g -o eoh o .e. 4.en.e.e e e==* e .h. heoe een m a .=

• e e es q,

e -e 3 J m e CP h.a =* 4 c

• =

== m -C e

• c*

== e c4 eee .e.e 3 =e= 3 e e h 0 .3== o== a >-

c. a e

N e w e e== C n. e

e. 5 4 e e. a e.

= 13 e e .-.e cC

w.m

== 3 e e - e t, t. e eeeo ce e .e. = - v ce oe .e. e m :D c. c o 3 we e== a .o. e.T e o h.e h = m e e.a * *=.h w e 3 = = = 3 3

===m e c.e.n a a e s - c. 3 =J

== e gpe ea.gsg .6 we & af. e o j g aetw e = e.ns W D e me w

== e an e I a i, e -. a e .h a.w. wh a =- i e 3. e c

== -=- .a .o e e. e T e cm = m= c w' ee .e e ew s e e e e e e

a. e. e a.

c. m t., --h.. =.4 a e.h. en c.e, 3 e a.ae oe e e== k .a .s a a a w c e 3 c 9 % e .s= oe 6 =.s c..m= .e. e eo oe ee 3 w T

t. c.

g e ee ee - e

e. e. e

= > m se m w a m

=

eo e L e

e. : e

- 2 2 c' e e" *5 c

===oNw

e. T = a. e o a n em ac

.e e e wo e= e e== e e e a oe e e aw - -_ o a.e 4 -- h c. =, - p w - as

g.

ew oe

e. c 4o w h.c.

w

== w o T e e. aw-oe s er 3 c Awa m one c an .* 1 3 .e. ve a wa e 4oa4

=

• o.e.

w== ve

• -e. e -

e = c e- =ee =h=.=.- e e.= - w e

== mA h.a 3.e e a s 3 h e-as .=

== e

== 3 = e

• 3 e e

.e - E n =-3= ee wc w g== > e o.- o o o e e== e c.- e e. ,zc2xe em o a w c. stewe65 L.8;5 avse.ctny l" l" tyy a:ea? 4:sse$.uow:c.*5 awaee..we ta nsen op;;egr g g ,= l atreve carn 4 eidtt e nw:,.79 e wgaa; awg go rawaowaceg. l l .aaes ?v eavutzt!?:r.'S a l. f IdA(I% (InufM o l g l .utef =ong at6utt:?. S I" l" en s.=a y 3 0 .c. unt entry :::veetnwit. S l l l y 57J7 40 1 1 uotten :v :stes=:ny:=.c+- i l i

• !j s t ewa 65 -er%5 36%'*o:37 l

en g 4:vaS 4:aarc.uow:1.ca l l" l" w. m ad l e uotst:i;taan v:ss aosg:s.ee l" l" g*

nca

rts..mo, ;y l

l .aan=asc.a 4:uaSaaw:::.c.a a = aa gi saggc;n$.alem snuseg f 'd l* l C8 .aanca:o. 4:ua6aag::t.c! . iw e e,. astm l" l .Suttsscavi ac!j em.y:*.e5 i l s*at'A t enweae l" l" i 5

utry.o!! a! ute.:act u:st:7.ca i

ste>2.e,2 a*ti n utrai *- :acs u:si:t.ca f" i l e e c. c, C 3

c..g

.g e e i-o _ c. .e

• s,g e

e _g :. e :. = i e = c - e a.u o -o as wo seue .3 = -- e. c--- e.s =c s =.e e - d as me det.e== w.3..{ e es 'I e E.3. v e e c e e C e.3 e e er e= 4 4 et { .e a W ** 3 ed .e S. 3 a=e.0 .=8 - c==.2 h.3 .2 g .e = e =*8 4 e r "c e .e 3 3 e er er * = - s e== e.p 3 es.=*.2 e e e

. = =

ea 2 c e. E== ==N Neme at =* ar. N

• 4==

I g.r.---e- -wwm w*w, e--7.. m-ww-fem *-e'*-- M-N---- "'"'"-*--*w t~~**" '~^~~P-' ^ * - ' ' " * ~ ~ ' " " ' ' ' '

. _ _ _ _ -. ~. _.-. - _ ~ _ - - i i f 4, l J i i I e t i i o 1 e t o l 1 ) E t ~ m C. 1 M w .u 4' er e i .=. e. e 8 k i 6. .o= m 3 w I m, s. .2* l Sun i ste.06$ t.ess a6;ews.nv i g e.gv ace 25 X:asse.uou:c.79 l" awa.owawa te.n;ei,wc p;asaag g" aSesse cang ae dttinw:,.75 f f wta.i awo jo aawao.acag O g, amaag !v ateutet t!:r.95 a i c saaten genwrn o e i .utre aceg ai$wtS:7. S l .i 5727 ao j j -jf votitat:v :stemetsv:t.99 I $#av 40 l i ]l uctient:w :ttacciny:s.tg l I ( ste6065 Lae:S 36;e*0;ny 1" 3 cafy actari h airs.ucw:s.e5 I l' " E ucate:ts taan utv4 mot j:o.cp r

nca

etg.. mag ;y h

.&.nca:cJ4 Cuaiaseg :.en l= a f-sattcans.aise chaseg 4 l" GI j .aan=a:cJ4 4.cuaSaas::,.cm 4.savow.aseg g l .Suttticau! moij esyy:r.g3 il l ) saater tenue. i utry aos.3 ai$utt sact u:st 7.g3 l" lft nn uteJe

6a n ulA i

~ a ta1 :aos u:ei:.ew l= d a e e :.

g..

g..(

n.-

• y e

e 5

a. n.". 2 m.

uf E = ., a. - u o wa .c. c -.* =. ..* - c c. - -* e 3waa== - s u..a s <w-e- u= sE e-

• .3 6 0.*

3b = 3 .e. e es o c.e.= w t. e. g e . e .s L= =t w asr g; e. 3 N== e 4== { em.e.e se e E e we <.=. t ',e~~-v- -r--,,--,+---,---nn,--mm,.em w.n .w -,--m--~c,, .,,-.-w ,-.w-g ,---.-.-w-, -.e ,~,,r,- -r-,e,---

..... _ _ ~ t 1 I J = 4 a f ? ( .a o I C 1 o i r V e c .o., 4 1 .2 .iIe l + f E .u. 1 g l a. 4 c-w I i .*e 4 e s } )

• o a

w m. t T .2 o i l i s tewol5 1**;S 2 61**01AV euy aosan Ata;e$.uowit.*S ' = i= i I i swesowawe ge.ntew ontaato., a$emen er.eu asettenw:,.3m --- ~ * -- t -- w = wba.i ewy go A3b4DweceQ m= g ataca av ateutett;:t.95 o I e seagen genwen .o m -useg egg as6utt:7 75 a' SNT 40 .e. l vatten.3y sttewo.my:t.in t I q! SNV eo l s =} uotten ey sitewetny:o.cn l stewsgs tar;; 262ewotny emp aceJS Atagee.aow:!.e5 L I aI m 3 l uottest;taan uteg aos j:o.e5 lx u 1 i. 1 nciaelg.eaos 07 6 e w i ye .aancesca. S. ua$aeart:c.ca sastGen5 **2em sn u eg-- .aanca:naa. Aswa6aac :e.rn. eengew.42em o 1 [ Su t t tteau! =ot g en.y:t.TS 3 I* l t seateA tenuen i bieg aetj stSuts.sact v:st:7.e5 l" 16*61 **L1 heetnD utta! c1 :act u:at:1.ca i t e ( a.t.

c. t

~. e. g a. m E e s.5 - _E 1 =..u. E u o c- . u.a. a i g as== u

.c ...a w.. s., = =. s..e u. o. - -..=.o.- ..2 4 e o : d

• J...*

w 4 .C. & w9 W em.E.C. e be

• e e

~g-----<w,<,-wwe--wwc.-,ye w-, ,--n.,'e m e r, m ary, w.,- w w e mers, w w % -, w mv.,,..myp-mw _. ,-w,-w, ,, wwww,--m,,--+-,n

-.u..- _x-axs +. -a. s e 4 e %a

== e g b .a =8 b 3 =*J c

• = =

&N D 9== g f b a'*gegw2 1Ye= O

• • 4 4

e C ob .ef== ** 5. b o %

4 m.O, a= **e en e em g a-e r, J.e a=

m en er (== 4' e*

• WG

==* 9 : =.;3 6

== 9 c2 ) C C'e= h

==w no J J e os.n.a.% ** =

==.an e, I eew - e=

==. EJe

• S79 h

W==: =. g== 4 e 3 : O . = = = g==, en : ee we Ce -. asm g 3s ..lb a ee-w o-2..g c, ~ C :n.:. - - w s o me z -e e T u en 2h 2 =a. s 3 4 e= C C== G. e h e( g. **5-F== g

• 1 )==

w as w"=.C *E* G. 43 . 3 5t c en g >h =9 ee u.

• e= & O 3 e e.e 0 ea 3

en 4

== m C E e J *T - C *== i

== J m.s e3) e >.* 9 ""== em C. > .e. i

==C.C O o me e =.=. o.e R 3 3

=== e e E 4 "3 e.i w e h e. =O.O - 30 13 0

== = en ea== f e*

C=

ep 2 e ens a== e a=

• . *
• 1 -

4 ed.* E e =* em J wt h a

== 9.e eG

==4 e e e er em 93

• 4e an

=wb w

• in

.e e.r e es wwh e J.es== c en v.=e3== m = e w me

==.e m== g es en s e eJ b e c ** *== e es====== - g en g as =.e= t .== e w 2 2D e.S - C e== =.:==== C 5 e5 w j

0 m

e a. en.==.* C.o..e.

e = a g

e. = o'e e = E C 5 m e e e s.

h E te.s*= w e 4 C**3.a = a e - - e :. 3 m w

ewwJ & en e D 9=

g J i e i 44 e %r 'l 3 a eE j em .e e.m w -.em. e 3 a= C e e ee se w 8 l e, C (3 'i c.i e 95

===4 C e Se u* .e. 95 o

==& e *= 4. m u O h e.

== h b g w== e n. a= m e he o. es.

== 4 q

== e e l .e e0 e w aC w o m a, es co e s. 2 O C tm e =5 ..w..e. e t .e. =, e.a.e o

a e-h-eec

- e > m.a., - c. eoe~. w cm v ~.o.a ~~ .,.e.= o e -a ..c+ w e2 e i s - o.6 e ww l e. stewa65 ide'S 36;'* Gin, engy ameJ9 4:ajte.uou:c.aa swamowawe taanseg og :a:oa, a$eurtS cang asetitnw:,.at .e = usesA eso go A;wao6,acag .e = N Jaace ?y attutstit:t.99 ac sangen genwen .o, m .uttf notj ate #.ute:?.9 pl 19 7 eo .e l uo ttvat:v st rwe ny: t.-a = 4 .:q 5747 40 ,E : uottynt:w site c.=v :.c.a. t } e_ s tewa ts we:S ;6;**Giov 3l e=_w seest (tasse.seu:t.ea w. e "4 uot:r:t;taan utng e nj:n.en w r.cueL6 M c Of .aanca:cas 43uaSaarn:c r! .e '= ec se6Lccn3.atem enn y e. e L; x f .aar ca:c;q 4:ua$aaq:=.e5 l .*etw.atem .g

• utttso;ui no!j ingy:t.en l%

l i j seateA t e% en 2 'uttg ae s j a s *u t t.:act t::a t y.ca laf et*;* u n eJ I 36861

• • 6.
    • ! ace 'J}4g

} I e

e. _i

-= i a f egs = a - :. e a 1 c 8'

==.*a.*

.

O 8 3 w J *= e j .e oe w e== e

== 3 : =* 9 C e =e 2-a e 9 3 l a. we==== Z== a l ~. r.-.- .m. ..m_,~,,--.wc-m,_,,,_.,-, , ~. - y,, -- -,..wm.,,-.-.,...,.,__, w- ~,,~

= en W i ..O. 4 .s In

== b T g

2. :s "2. 6 O **

4. C ep : --g a l

=J =.

=b 4 I 4 c w-e - w I 3 ~3 = ~* e a da-i I

4

-2= E: l w: T e== g E 3C er 3 C "S h C 7 "2 J== m

== 1 = e ee

• w &.3 e.C. e e==

s= e w me

• 3 as : e ee er S

C7 = b74 > C @ ea C== 0 4 e== 0 3 E== a=

== w== es =. C CB 3 w 4; === W "en 5 g ee3 hC .ef ma me n. en me 4 er

• 2 3 eC C & w==

l e ot l

== b es e3 C's 3 w { O.e lw 3 to . e at e m w =. C e& O

== > "3

m. *O 3 es e,====.

X .== g, a e ma e >> =. & C C** m g e = E.== Gr f '9 > Q.3 - e as as C. Cw ew E e wi ed um as J

==.4 C .e, m== d e l e C e e g - e we, -C s

e. a C.

4 e h O w'y as b & 4 C ed G== eW 8 G - 9 O Cf w C (". ei e -. e b.e o 3 o 4 V ed

==== 4 9= 0 E. e e e eh & ea CoOw % -- C. =- GT a . wC e s w .o e..w==.=.w.. m e e.e s.o em. D D

• s z

h a= h w w.* e - - C, - ei

-ee =

C eoc e w -.e a a - a, ee -= e a a s. e, C C &c. w w e o e, e - .a wo -.e E?e o = m c. # v -s 2-

• w-a

<C 3 e e. s steur.65 a:s :6:rectny i I cagy a:taa 4:aste.acw:e.-a i l* . na.aw*we t a.nte % :p;an.e t g i E j g= a5v-va== 9 asettenw:,.*? a wha.; + c,c 4:weowacao i i .a.:s y e.ruimte?:t.-a pI l" i saataa ten *'a 3i r i i .r.v.==vy aq$ute:?.-P. l l" .l SNy,o i 3l uet:t ::w :t ts:try ti. a I t . =., ~,j SNf so 1 i e uotst.:v :tteweinv:n.ca l l .i 5L8 365 3:5 apamicing i c 3l en.. a:rsa 4:a a re-ucw : g.ee l f i w f tj 1 4 851

otte

t.t.ae u r., acij:o.r?

l* in: :rgs...oe ;;y g" .a. :a::. 6 a b aq ::.e? 8 y s+,pe.as 4 sm.a:eg i S:t .aan=a:u s X:ua5;a= ,.ca =, l" . a.s.e.atem o Sustt:N ui =ce! gasv:t.ca

f, l

j { seateg tenwww ( j hag mor! at$ure :aeg u:ai: ca i t t 16861 dwta I sStino att;! -- :acq u:at:s.ca e & C.

• 6

.o _ c.

e. (

L -+ m e. g. s a. C e L as G. e.* 2 ""- -.o a U = [ eve e.9.a.= 2 et e-a= 2==== te O = * = *. tr===*** m % u.3 2 O e. es e e # 2e4 e e .a h : e= 6== l g is e== j 2. m-2 h =# O ee Q es -

• ====2 3 as j

l -lw a s.e !3" 9 & T C

== w - = s e, e - = = =

==

es me 4 4. s wt ** ** a ww O 3 ar.== }==

• s'* aC==

1 e w-m,,. ,g,,-,-,_-,,.--,n,----.,,.w ,,,-.n,.y++-~, 7.,,,n,_,,,,,,wmnypym mn,ygm ,,n n m e.,,,n _g.4,.. ,=.m_gmany,.w-_ww ,o,a.e-w-,-wy

~ u a e e h 6 6 e e e en ef* I =C.e g en h 4 e** b EC.h .c ew a -er g

==h .e e d.e 'C'9 l er e =w 6 e aoe a =*e J4 6

o c

e

==o.o -- rE.c.,.T - e e =. e==ee* *. = = O-

0wO w==.==

• O. m -

C O W % 9 p 3 e ea 34 F = ee "e=e.s .w.as e.* J. =E a.e== e en

=e

e

a. f e*C 1

g & w =. = 0 0eeb EOe = e .a O.e e 6.e e .&, Q.== b w J O, a,r== I en C. G. = e g = g, 0 e *R w c *. c t

e. e w m ===

{ et J* O *D = G. *= b h 4 ime a

• 2,**

33 2m a== me 2 e.= a= E '== ., e.h -m e *eo-os - anc.n= = a a..o g e ::.. =. e 6 e e a e E & *3.=== e = = 'e.ee 0 3 3e C (== ===.ss==.O w 0 3ea8 S e c a = = = Q 4 =* J 3 ee e e w- .e.= ww3-oe e 3 er a.= eO 3 w.= =.

• = = = e s w.M* an.e

.

• w h

>oe# = .c.h. w c. 3.w .=e e e e. aC- *.2. =4. c e l G. e =3 = e = I e o ea e =J es w < o== -e is w me c..= = 3 e== o w C 1 E g J. w e e e==.a. w== = e = e o =

c. = =.e e.e =

.e,.= C -J e e.3. o -.=e

== e- . h e===.h e as a

m C. - =t ew = = = b===

g e -e e 3.p g e. =e > > *e .o. - === a = o-j es da e a e = < e = =. e l' 3 g g e e oe e e >4. ., 4

.

=< = w em ew3=es. wo-a-- =. e i i we E w g imi w.

• f I

a c91 c. .S e.e. to o

r w

e== 6 .W ..es t .a o Strwe)$ 6613 :Gewc ng g g e i enfy 3 taa Agaget wCW:.*9 l e.a.c aw, t e.n;aw..o c ;a::. 8 ads:Tri Pms aieti n.:t. r g wts.1 a c jc a;waowace; .e. 1 .aacq ry attutati??. r ci ; l saatta traws. 5 4 I .etts =ci a a sia te :7 9a

• l imat 4G

~ _El us ti?". w : t t et:et v : t.*a 6 l .21 SA.sv so

d..!

w: Lit n* :V :litectny:=.ca 3l q stewet$ Lt;$ s,tano;ny f engw a:taa Isaatg.aow:f.ca wi l "' I' 6 l ucitt:t;t.ae vity mo s. ::.ra l I u ...ca:ets.aaoet l p .&==a: 2s 4:uaS as;::.ea 8 i T,i sati::.as.asam saa:aE i i '8! .a.r':a:cas 4:uaSaaz :t.ea l .a aw.asam 5 I .*ut it:Lui nos j engy:=.ca f, luttemoifa#5ute.:aceu:a!:7.ca saatta genus. laSe mS ut ta e 16861 **tA l ara : ace u:at :i.ra =. osa e.= p n :..* {. . aeo=_ a a e :.. e .* e - w [ 4.==.l=. *L * *7. =..=, . s e i w.=a5e em== <s e .e l l C .e,. a :h .e es-

-.

CL a w e, =..=.==.a e. .e=, e O C = C

== l E. l e= s==== 0 en 3 am a== i 1. ( l._-,__,

e e e 9 C oh

• .G W 4

.e . = =..P T, G r. e 3*O so w e 1.= C

==.C*. .8

• b e "5

=.9 .am w a..a gt w 3 w3 g.

<w

. c. " ="" . 3==.. o e. w c== o a g L. a 4 w - - =3 e n.: e-z a n-s - O s e C "= oeaf .& =*= e e,e - c ( = = * .o e.3 o.* C e.

== o wt 5. "A Ce e S

e. e.

6 e w7 - to.g c.e a.=..e.s 8 w= 4 .= .= 3 .r oo

• .e o o e.: T.

e

• === G. w=

C**C w e E.=* S . e G W

t..=

.o w a - e, Ec3 .s..e e .e = es C 7 8 =.=* 0 < Q C. =E==* .O. C 3 O

== 5* e e.

==.3 g .e, w e .= e.== a = e.e.h. o e.o, c. c.

• C o*

a., 3 e SL. - - e,. - .m .a 3e .w,.= -m ..m. g v a o = eo.a e - e e.. - = =3 en o c .e ..o,.=. = mC a e3 w o. c <..a. c. e : c o . xa -c .s es-e e e. e s o e e. o. e o e.0. =: o.

o. c. w -

a.=a i, w a w = .a==-aw. .t = w. E w e o.a .=. c 2 w m< .G. n .e k . ee Paa c - w., eo e a e e.* =c. b 3 c.w a o = a e o. c eoe o a i. e .a n ..o..* 3 ~ v T. x uc

r. 6

-..x eo- -..e w E 7as o.,. .m. at c. I ste6365 a;5 Ob;rwo Pr ) easy a:tJR AtajP$.uCW:*.9a enia.ouawe t e.nsee.. won ate, a$twe? zng asetstnw,. y wie.4 ag,o A:wacwacac " ~ aa c4 ;v strutati?:t.-n u. santen tenues a; .airg motg at$ute:7 99 I q Sr f so j a .o TJt."PF".:V :t*PmC%nV!t.*a_ 3 41 Saar,3 t ,e l vette'ts:v OttPtetPy:2.ca_ l gj ste6065 a;5 25;r.asny [ 3 ingy a:ran Ata2:5.uois : r.ca I w**! 1 ue ttt:t; t.an v=.vg nei g :e.ca ' w -e

ncz:aiji.a.oey y

.aansa:o44 4:ua$aas=::.e3_ a = s.6tcr.ns. 2:n :na:rg Si .aar:a:c24 X uaSaaur::,.cr._ m .m eg.alem o .Sut t tioJul =os a cagy:e.ra l saatra ten 6rw luteg motg an5ute.:act u:st y.en l $ tin $ uttJi

6* G **61

'a ap ace u:ai t.ca ~ gm ..2 7

• J, =

eG = e E 3 CL e a es C 3.a =. * .C= E.C = = = w ea3c = ~ 3.h e .e ai.

• .e.

t.* a s 2 .es - J - es C =.

• e c.3 s

I e = g== 4'l.E 4.o. C -

a. f l

3=.e e e.* K.C. 1 e '8 w t.f

.~a =. s. a. e e 9 e 4 .em C' G C . C* o

• o-

-e. .= 5 .= m-w 6 ~3 % .a e C h @C. o w e.a, 2 l=. { -.gca. ~ .u-oe o.a = w -

• • e.6. =. 6 w a w.=eu,.= 3 "5

aew e F C 3. e. w.C b o .o 9s w - e eca e oa - m.- C =- 6 g a S ".J e *" s,..a. w, b c meo,. .o. cwcw s..e eew. V e.., e c.a e m. c e w.o,.z.o =. =. o wo . 6.: % c..=. c

• c. - o= =

=

• ** ** * { w "I
  • G

.o. em =.c.= e. = m w m .m-o.-e.2.we - = e-6

• * * * * * =.c 4 db.

c h.e * *= ** u.p.E o a u 6= w ) Z 3 u e2 ~n .,1 ) e e.a B e o- %~ 6 .a e e= w ee ea w we o 4 4 -..e. m 1 u* a

.

= m m, -.a. c am-w .e. e e cs t' &.e e 4 .=o - w o. e as ~ => c o .n o. 1 .c 6 l stewags.Jr:$ : pene my ^ I em y a:eJa Ata;ee.aew::.9n j"" l ewsw3w664 te.ngew s p;agc.* 4.nT:;lT M d=ng afdt;[nw:,.9n .a m .e. uge44 awQ 30 A;uapwagag I = I .aacg Oy attutate :r..ta_ o c SanteA tenwag a!" [ 4174 ao!! at6ute*?.*9 ol 57.2y,G l 'I UCLOPP*V 3 t

• enc
• nw ! t.*g -

-f l SMV 40 g u0 t * ?Pt:y O t'. E.'A.? ? ny !3.cn i e, ]l StenCl$ W:15,h",twogny .e, l cary aceJp /* 33 awagg:g.cn et cui x u0l*T2tJLJa4 ut e / mo t t :3.cn. y u. -- }

nca:etg.aaco Of 8

ye .aJ'ica?ca4 4. ua$sast ::.em j SabtGGD$.alem Cf.a;eg-w e. .34 n 34 30.44 X2Wa 6a ter2 : n.en .e w l .maniTM.aleg o l .',U t (

• t OJ u l mot! ca.ry!7.en N

l SanteA Lenweg utte =otj affutt.:act u*ai:7.ca l 16msi dell 46etne ;tvJi a09 Sace u:atit.en R a %

• i

-i IL e C. o.* u o As as.*.C. .e e C. - w.2 e.e. .a w. t e e-=6 3 e .= .o e, L.2.. = e e i

a..

ao ce== e E e -rwm-w-c ~~ -~ m w x ewsc+es-o---vw,-,-y..e,*,----m---warm-pme---ww--o - y y,- - -,--w w w w-m m r,p,pv e,w w w, we g w *w r, y y w or m g p-e-y-nw-v ~.-ew*

---

ev-,e

1 E:: CLOSURE 2 s 4 Basis for Auxiliary Feecwater System Flow Reouirements As a result of recent staff reviews of operating plant Auxiliary Feed-water Systems (AFWS), the staff concludes that the design bases and criteria provided by licensees for establishing AFWS recuirements for i ficw to the steam generator (s). to assure adecuate removal of reactor decay h2at are not well defined or documented. i f We require that you provide the folicwing AFUS flow d3 sign basis infor-mation as applicable to the design basis transients and accident con-ditions for your plant. I 1. a. Identify the plant transient and accident concitions considered i l in establishing AFWS flow requirements, including the following l events:

1) Loss of Main Feed (LMFW)

2) LMFW w/ loss of offsite AC power

3) LMFW w/ loss of onsite and offsite AC cower i

j

4) Plant cooldown l

5) Turbine trio with and without bypass i

o) Main steam isolation valve closure (

7) Main feed line break

8) Main steam _l_ine break-"

9) Small break LOCA

10) Other transient or accident conditions not listed above 4

k 4 k i

.~ 2-b. Describe tne plant protection acceotance criteria and corres-conding tecnnical bases used for each initiating event identified above. The acceptance criteria should address plant limits such as: - Maximum RCS pressure (PORV or safety valve actuation) - Fuel temperature or damage limits (DNS, PCT, maximum fuel central temperature) - RCS cooling rate limit to avoid excessive coolant shrinkage - Minimum steam generator level to assure sufficient steam generator 'ieat transfer surface to remove decay heat and/or cool down the primary system. 2. Describe the analyses and assumptions and corresponding technical j justification used with plant condition considered in 1.a. above includinc a. Maximum reactor power (including instrument error allowance) at the time of the initiating transient or accident. c. Time delay from initiating event to reactor trip. Plant parameter (s) which initiates AFWS ficw and time delay c. between initiating event and introduction of AFWS flow into steam generator (s). j d. Minimum steam generator water level wnen initiating event ' occurs. I 3 l l r .ys~m..--%-w,y w<ww e e v, e .-----w wm me~t-mex--v-"T--**"*-'"e--~" ~#'""

O . l 9 e. Initial steam generator water inventory and cepletion rate before. and after AFWS flow commences - identify reactor decay heat rate l used. f. Maximum pressure at which steam is released from generator (s) and against which the AFW pump mu:t develop sufficient head. 1 g. Minimum number of steam generators that must receive AFW flow; j e.g.1 out of 2?, 2 out of 4? h. RC flow condition - continued operation of RC pumps or natural-circulation.

i. Maximum AFW inlet temperature.

j. Folicwing a postulated steam or feed line break, time delay assumed to i clata break ano direct AFW flow to. intact steam j

generator (s). 'AF.i pump flow capacity allowance to accoccodate the time delay a6d maintain minimum steam generator water level. Also identify credit taken for primary system heat removal due to blowdown. i. Volume and maximum temperature of water in main feed lines i between steam generator (s) and AFWS connection to main feed line.' 1. Operating condition of steam generator normal blowdown following initiating aver.t. f 4 ,,m.,_._._.._.._c._,. .. ~.. ,....-._..-.-.m.,...-~,

O 9 4 _4 m. Primary anc seconcary system water and metai sensible heat used fo r cooldown and AFW flow sizing. Time at hot standby and time to cocidown RCS to RHR system cut n. ia temperature to size AFW water source inventory. 3. Verify that ne AFW pum:s in your plant will sucoly the necessary flew to the steam generator (s) as cetermined by items 1 and 2 above considering a single failure. Icentify the margin in sizing the cump ficw to allow for pump recirculation ficw, seal leakage and pump wear. .}}