ML20206G541

Forwards Draft Environ Assessment Re Proposed Certification of AP600 Std Plant Design.Environ Assessment Will Be Used as Basis for NRC Finding of No Significant Environ Impact Resulting from Certification of AP600 Design

ML20206G541
Person / Time
Site:	05200003
Issue date:	05/05/1999
From:	Joshua Wilson NRC (Affiliation Not Assigned)
To:	Mcintyre B WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
References
NUDOCS 9905100050
Download: ML20206G541 (27)

Text

o

' May 5,1999 Mr. Brian A. McIntyre, Manager

- Advanced Plant Safety and Ucensing f

l Energy Systems Business Unit -

Westinghouse Electric, Company P.O. Box 355 Pittsburgh, PA 15230-0355

SUBJECT:

AP600 ENVIRONMENTAL ASSESSMENT

Dear Mr. McIntyre:

The U. S. Nuclear Regulatory Commission staff has prepared the enclosed draft . j i

environmental assessment relating to the proposed certification of the AP600 standard plant design. This environmental assessment will be used as the basis for the NRC's finding of no I

significant environmentalimpact resulting from certification of the AP600 design. If you have any questions on this matter, you may contact Dino Scaletti at 301-4151104.

4 Sincerely, WWN Jerry N. Wilson, Senior Policy Analyst License Renewal Project Directorate Division of Regulatory Improvements Programs Office of Nuclear Reactor Regulation Docket No.52-003

Enclosure:

As stated cc w/ encl: See next page Distribution w/ enclosure: tDocket FileYPDLR R/F PUBLIC ,

Distribution w/o enclosure:

SCollins,0-5 E7 DMatthews -

r CG'imes h

JNWilson DScaletti GMlzuno,0-15 B18

. &%L79. Wb i Document Name: G:/ WILSON /SOO-LTR2.WPD ,

f_

OFFICE: BC:PDLR: DRIP PDLR , , , PDLR: DRIP PGdjrpFkP NAME: CIGrimes MI '

DScalattih JNWi RA ht(( "

DATE: 5/2.9/88 3 //788 3 //7/88" 7888' OFFICIAL RECORD COPY 9905100050 990505 PDR 0 b tik A ADOCK 05200003 PDR

Mr. Brian A. McIntyre Docket No.52-003 Westinghouse Electric Corporation AP600 ec: Mr. H. A. Sepp Advanced Plant Safety & Licensing Westinghouse Electric Corporation '

Energy Systems Business Unit

' P.O. Box 355 Pittsburgh, PA 15230 Ms. Susan Fanto Advanced Plant Safety & Ucensing Westinghouse Electric Corporation Energy Systems Business Unit -

P.O. Box 355 Pittsburgh, PA 15230 Ms. Lynn Connor DOC Search Associates Post Offic6 Box 34

• Cabin John, MD 20818 Barton Z. Cowan, Esq.

Eckeri Seamans Cherin & Mellett P00 Grant Street 42nd Floor Pittsburgh, PA 15219 Mr. Charles Thompson, Program Manager AP600 Certification NE 50 .

19901 Germantown Road Germantown, MD 20874 Mr. Ed Rodwell, Manager PWR Design Certification Electric Power Research Institute 3412 Hillview Avenue Palo Alto, CA 94303 '

3

r-b

.y

.3 b,g' 3 2;tg gn ,,

-.. d. /'7 ENVIRONMENTAL ASSESSMENT ,/ Mf A 2 BY THE OFFICE OF NUCLEAR REACTOR REGULATION'3MNII[,'.-_,[,

U.S. NUCLEAR REGULATORY COMMISSION t'OQ/'

%%nik, RELATING TO THE CERTIFICATION OF THE AP600 STANDARD PLANT DEt$1GN 's DOCKET No.52-003 MM ..Isz,.,;ure-

. cw.

N S s:)k?f?_

.y xn a'~ s, , -

~

E ,p'

' $ ', fs'y., . l,

,ue

~

• lij%Q 1$f/

w; 7

,f' -Q,;!/

~

,/

.,  ? ll

• .,Y 4 a i d: ,5 e5

+ y. 3

, k Dg ,, , ,

. j, .v. ;_ g (fe f) ,p$ <

'b ,5; .;,,?

1 r2 ;., ,

y].

'.) ef:, .y , j s .i

> 4t; ,) = ,

v g) ~, c'p;  :, ,

, y.4 [.3, i ~, S- i

,s' ,I

. \

,. a.g: ,

y ,1 . ! ,$^s% .?}

9e

s. .* )

?  ;)?  !

n eo a

.?

S,Q.. e%I! .

.s)a)9.Y

.g i. ,, .t + n lr.

..y - i Q'<T *,.y' MgN"&'?'

w..

)

Enclosure

TABLE OF CONTENTS 1.0 I NTR O D UCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.0 THE NEED FOR THE PROPOSED ACTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

.. .......... 4 3.0 ALTERNATIVES 3.1 Severe Accident Mitigation TODesign THEAltematives PROPOSED AMD ACTION . . . . . . . . . . . . .

3.2 Potential SAMDAs identified by Westinghous  % .... d h ........... 6 ~

3.3 Staff Evaluation . . . . . . . . . . . . . . . . . . . . . i....... M!

3.4 Risk Reduction Potential of SAMDAs . . . . f......

3.4.1 Westinghouse Evaluation . . . . . 1. . . .

.. k.ht i? WiLh-........

........H f L ...d.10

... 210 3.4.2 Staff Evaluation . . . . . . . . . . . . . b , . . . . . . . . . . . .MsY. . 12 3.5 Cost impacts of Candidate SAMDAs . . . . W M...................... 12 3.5.1 Westinghouse Evaluation . . . . . . . . LED. . . . . . . . . . . . . . . . . . . . 12 r

3.5.2 Staff Evaluation . . . . . . . . . . . . .  !

3.6 Cost Benefit Comparison . . . . . . . . *i2i .....

. . . .$. .f..............

. .' S$8fh...14 . . . . . . . . . {.

3.6.1 Westinghouse Evaluation . . . . . . . . . . . . . M & ............. 14 '

3.6.2 Staff Evaluation . . . . . . . #fi k....... "'QGT.............. 15 3.7 Further Considerations . . . . . . ................ 17 l 3.7.1 Uncertainties in Co z..%.

msg .'.M( uenc R.M?L . . . .p ' ident-Related

. . . . $yMand Acc Exposures . . . . ;.... M................... 17 3.7.2 Reassensment of esign rnative t Benefit Relationshi in Light of Uncedainties M . . . . . . l . . . . . . . . . . . . . . ........ . . .ps 19 3.7.3 Further Evaluation of gn Alternatives With Potentially Favorable Cost-Benefit Factor .J............................21 3.8 Conclusions . t. .k. . . . . . . $i .. . .. h gdf.............................

. 24 jfiW {W$V 4.0 TH ENVIRONME AL IMPACT OF E PROPOSED ACTION . . . . . . . . . . . . . . . . 25

.e ,

wn 5.0 AGENC AND PERSO CONSULTED, AND SOURCES USED . . . . . . . . . . . . . . 26 Table 1 Comp

,N $1#

a o%@f Estimated 15en'efits from Averted Offsite Exposure . .

/<

Table 2 Key Differen

• pig'oen Betw the Westinghouse Approach and NUREG/BR-0058 ... 16 l > 'V ' l k

'3 Key Paramete :Used by FORECAST in Evaluating Maximum SAMDA Benefits . . 19 i I

4 SAMDA nefits Accounting for Uncertainties and Extemal Events Effects

$.3;,. .

(Benefits, 1 996$) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 5 ' Maximum Benefit from Individual SAMDAs . . . . . . . . . . . . . . . . . . . . . . 22 m{5Qnp"y 2

1.0 INTRODUCTION

The U.S. Nuclear Regulatory Commission (NRC or the Commission) has issued a design certification for the Advanced Passive 600 (AP600) design in response to an application submitted on June 16,1992, by the Westinghouse Electric Corporation (hereinafter referred to as Westinghouse). A design certification is a rulemaking that amends Title 10, Part 52 of the ,

Code of FederalRegulations (10 CFR Part 52).

{

/ k This report presents the environmental assessment (EA) for this rulem h the NRC  !

has prepared in accordance with 10 CFR Part 51 and the r ulrome of .g. .. National  !

Environmental Policy Act of 1969 (NEPA), as amended. 'EA a ~~the environmental impacts of issuing a design certification. In addition, thig Irt addre f sh5rex accident [

mitigation design attematives (SAMDAs), which the NRp has decid to r as part ofWs clat(f final EA for the AP600 design. This report does not ac dress the ironme constructing and operating a facility which references l ie AP _ _ besign certifi at'a2' particular site; such impacts will be evaluated as part o cation (s) for siti ,#

constructing, and operating such a facility. j As detailed in Section 4.0 of this report, the NRC det Ined" this design certification does not constitute a major Federal action significan "affecting the quality of the human environment. This finding of no significa(impact is based ~the fact that the design certification would not independently author lz6the siting, constructio !br operation of an AP600 reactor design. Rather, the certification would erely e AP600 design in a rule that could be referenced in a construction permit (CP),(early site ' permit (ESP) operating license (COL), or operating license (O .) applical,lon. Further, because the certification is a rule, it does not in'volve o any res)urces that would have attemative us Therefore, the NRC has decided to pre re'an envi " mental impact statement (EIS) in 1 y

connection with this action.

A hf In additio pursuant to NEPA, the NRC a fi N.[f reviewed Westinghouse's evaluation of SAMDAs that gen (n,frically apply tohe AP600 design. n that basis, the NRC found that the evaluation provides a reasonable assurance that certifying the AP600 design will not exclude SAMDAs for a future facility that would pro coat beneficial had they been considered as part of the original design certificah application. Issues are considered resolved for the AP600 design certificatio p ^ %g pWA 2.0 T , NEED FOR E PROPOSED ACTION T RC has loc.g sou safety benefits of commercial nuclear power plant ardization, as w .as the early resolution of design issues and the finality of these tions. The NF36 plans to achieve these benefits by certifying standard plant designs.

B to 10ffR Part 52 allows for certification in the form of rulemaking of an essentially o' nuclear plant design.

yff The sed action would amend 10 CFR Part 52 to certify the AP600 design. The amendment would allow prospective licensee's to reference the certified AP600 design as part of an ESP or a COL application under 10 CFR Part 52 or for a CP application under 10 CFR Part 50. Those portions of the AP600 design included in the scope of the certification 3

rulemaking would not be subject to further regulatory review or approval. In addition, the amendment would eliminate the need to consider SAMDAs for any future facilities that reference the certified AP600 design.

3.0 ALTERNATIVES TO THE PROPOSED ACTION The NRC had two altematives to certifying the AP600 design in an amendment t 0CFR Part 52. Specifically, the NRC could (1) take no action to approve the design, ~ issue a final design approval (FDA) without certifying the design. In and of th ,

ese altamatives would not have a significant impact on the qual' of the vironment because they would not authorize the siting, construction, operati la'fscility.

In the first case, the NRC would not approve the desig ) Therefor . fa $ built usi the AP600 design would require licensing under 10 CF Part 50 jio CFR . 52,' Sub%rt C, as a custom plant application. Moreover, all design w o. have to be red as part of each application to construct and operate an AP "at a particular site. ' ai result, this alternative would not achieve the benefits of standa , provide early resolution of design issues, or permit finality of design issue resol "

'st1 In the second case, the NRC would issue an FD .Appen ~ 0,t610 CFR Part 52, but would not certify the design in a rulemaking. ..

_ , ' although t 'NRC would have approved the design, the design could be rg, ed and thus would rpquire reevaluation as part of each application to construct and oper an AP ,facilityat;a particular site. This issue 't " not achieve the benefits of alternative standardization would permit or establish early resolutiog of design li.s 'ue re M ion.

Mi The NRC sees no advantage in i~ alter 'e compared to the design certification rulemaking proposed for the,AP,800 desig : Although 6either the alternatives nor the proposed design certification rulemaking would signjticantly affect the quality of the human environment in and of th,emselves, thepe r i' making achie ~' standardization, permits early resolution of design issues, and finality of resolutions for_ design asues (including SAMDAs) that are within the scope of the design certification. Therefore, the NRC concludes that the alternatives to rulemaking would not achieve objectives the Commission intended by certifying the AP600 design pursuan)to'10 CFR Part Subp' art B.

3.1 Severe /81-@tdA AccidentMitioation Desian Alternatives (SAMDAs)

// 1 t%%W Condstent with its objec Ive@f standardization and early resolution of design issues, the ission decided t waluate SAMDAs as part of the design certification for the AP600 d n. In a 1985 po ' statement, the Commission defined the term " severe accident" as an pont that is 'beyon,d a substantial coverage of design-basis events," including events in i

bich there is subetantial damage to the reactor core (whether or not there are serious offsite Design-basis events are considered to be those analyzed in accordance with

$ NRC's'Stan'dard Review Plan (NUREG-0800) and documented in Chapter 15 of the AP600 Desig'n~Co'ntrol Document (DCD).

As part of its design certification application, Westinghouse performed a probabilistic risk assessment (PRA) for the AP600 design to achieve the following objectives:

4

~

Identify the dominant severe accident sequences and associated source terms for the design.

Modify tive design, on the bases of PRA insights, to prevent or mitigate severe accidents and reduce the risk of severe accidents.

Provide a basis for concluding that all reasonable steps have been taken 19 reduce the chances of occurrence, and to mitigate the consequences, of severe accid,ents.

Westinghouse's PRA analysis is presented in Chapter 19 eAP6 , Safety Analysis Report (SSAR). ]jQg in addition to considering attematives to the rulemaking ocess as "U i lon 3

' applicants for reactor design certification or cps must consideriltemat ~ ' design featu'res for severe accidents consistent with the requirements (10 CF6Part 50, as w .~a court ruling related to NEPA. These requirements can be s ' marizeid as follows:

WWh,

• 10 CFR 50.34(f)(1)(i) requires the applicant to rfi 'a plant / site specific probabilistic risk assessment, the aim of which is to seek a ' h imp monts in the reliability of core and containment heat removal systems as . e,significant ' practical and do not impact excessively on the plant. ;T 4

• The U.S. Court of Appeals decisio ,in n v. NRC,869 F.2d 719 (3rd Cir.1989), effectively requirkthe N in the environmentalimpact view perf d und(ude' consideration

~ Section 102(2)(c) of NEPA as partof certain S of the OL application. [,'

/

Although these two requiremen ar's not daredily related, they share a common purpose to l consider alternatives to the , to evaluate potential attematives improvements in '

the plant,detign which i safety ' proposed pe des

nee'during severe accidents, and to prevent viable allematives from foreclosed. 'shbuld be noted that the Commission is not required'foconsider attem a to the design in this EA on the rulemaking; however, as a matter of dis'cretion, the Commigsion has determined that considering SAMDAs is consistent with the intent oTIO CFR Part 521gr earfy resolution of issues, finality of design issue res(A tion, ardenhancing the benents of standardization.

In its on in Lime Action v. NRC, the Court of Appeals for the Third Circuit expJ4ssed its opinion th it would likely be difficult to evaluate SAMDAs for NEPA purposes on l

a eric basis. Howe , the NRC has determined that generic evaluation of SAMDAs for the standard desig is warranted for two significant reasons. First, the design and uction of all plants referencing the certified AP600 design will be govemed by the rule ng a singledesign. Second, the site parameters specified in the rule and the AP600 estatWelfthe consequences for a reasonable set of SAMDAs for the AP600 design. The I ~ residual risk'of the AP600 and limited potential for further risk reductions provides high co dence'that additional cost beneficial SAMDAs would not be found. Should the actual parameters for a particular site exceed those assumed in the rule and the SSAR, SAMDAs would have to be reevaluated in the site-specific environmental report and the EIS.

5

l

~

3.2 Potential SAMDAs identified by Westinchouse To identify candidato design altomatives, Westinghouse reviewed the design altamatives for other plants including Limerick, Comanche Peak, and the Combustion Engineering (CE)

System 80+ design. Westinghouse also reviewed the results of the AP600 PRA and design altamatives suggested by AP600 design personnel.

Appendix 1B of the SSAR does not explicitly state whether Westinghou 's evaldtionincluded

. plant improvements considered as part of the NRC's Containment Per na .trnprovement (CPI) program (NUREG/CR 5562 -5567 -5575, and -5630) Howey .W ghouse stated I that the types of design changes identified in the CPI pr ftAave ~ ~ "tisen incorporated '

l Into the AP600 design or have been considered as desi 3ttismative a ~~im provements ,

identified in the CPI program were also evaluated in o 'docume revi *y/

b .

Westinghouse, including the CE System 80+ design a 'mativey IUations. g / l Westinghouse eliminated certain SAMDAs from furthe i I ration on$b the/!I basi t they are i already incorporated in the AP600 design. Such feature " include'the following:

. hydrogen ignition system . 1%

reactor cavity flooding system reactor coolant pump seal cooling (AP , . ~ canned mot '

s)

• - reactor coolant system depressuriza ' %f ' ^ '

e external reactor vessel cooling i  %

e nonsafety-grade containment ,' F, On the basis of the screening, Wes[nghousey ained 1 potential SAMDAs for further consideration.Thess SAMDAs, ^ bed i Section 1 of the SSAR, are summarized below:

V lh ' l (1) Upgrade the Che dal ' Volum . tem (CVCS) for Small Loss-of-Coolant A,ccid' ents (LOC jThe CVCS is ontly' capable of maintaining the reactor cooling system (RCS)I for LOCAs #ttfeffective break sizes up to 0.97 cm (% in.) in diameter. .This de emative would extend the capabHity of the CVCS so that it could the RC ' during small and intermediate LOCAs up to an effecti size of 1 ~ '(6in.)in diameter. Implementation of this design i alte, would, require in . ation of in-containment refueling water storage tank lRWST)/ co inment recirculation connections to the CVCS, as well as the addition of second line fr 'the CVCS pumps to the RCS. Westinghouse estimated that

,' implementing thi *de%n altomative would reduce plant risk by at most 5.5E 04 person rom /yr.

Fittered Con ment Vent: This design altamative would involve the installation of a

1. s)h filtered c ment vent, including all associated piping and penetrations. This

$(goverpressu~re modshca failures as well as a filtering ca would provide a means to vent the containment to prevent catastrophic ered vent would reduce the risk associated with late containment failures that might occur after failure of the passive containment cooling system (PCS). However, even if the PCS falls, air cooling would be expected to limit the containment pressure to less 6

than the ultimate pressure. Westinghouse estimated that implementing this design altamative would reduce plant risk by at most 1.0E-03 person rem / reactor-year.

(3) Self Actuating Containment isolation Valvos: Self actuating containment isolation valves could increase the likelihood of successful containment isolation during a severe accident. This design attemative would involve adding a self-actuating valve or enhancing the existing containment isolation valves on containment penet ,tions that are normally-open. (Specifically, penetrations that provide norm fly op thways to the environment during power and normal shutdown conditions 4Thi '

Id permit automatic self actuation in the event that containme ' conditi e a severe accident. Closed systems inside and outside con ' nt, s 'ais residual heat removal system and component cooling, would _ ciuded fr [tiiisliesign altemativl' Westinghouse estimated that implementing thi sign alte tive reduce plard/ l risk by at most 7.4E-04 person-rem /yr. ' '

p j (4) Passive Containment Sprays: Installing a p . safe,ty-related containm spray system could result in the following risk benefits: $$ ,

(a) Scrub fission products, primarily for gtor (b) Provide an attemative means to fl thevessel re%. (inntisolation failure.

vessel retention).

l l

(c) Control containment pressure for; cases in whihthe PCS has failed.

l Westinghouse estimated that implemeyting this design alterytive would reduce plant l risk by at most 6.9E-03 person-rem /yr, whichwould represe t eliminating all release  !

V4y*

categories except containment b s.

(5) Active High Pressure Safety) tion S em: Ad%g"a. Eafety-related, active high- i pressure safety injection ystem wouly enable the/ reactor to prevent a core melt in all

, eventa except excessive UDCA' and anticipated transient without scram. Note, however, that this design alternativs' is not coffsistent wittithe AP600 design objectives, in that it

~

would change the AP606 from a p " with oidy passive systems to a plant with both passive and active' systems. Wes estimated that implementing this design aftemative woulfroduce plant risk by at most 6.1 E-03 person-rem /yr.  !

ya4 $.A )

(6) Steam prator Shell-%ide Heat Removal System: This design attemative would 4 involve installing a passivY ~ faty related heat removal system to the secondary side of the,steiam generators This hancement would provide closed loop secondary system l

I cooling via thb of natural circulation and stored water cooling, thereby preventing ss of the prima y heat sink given loss of startup feedwater and the passive residual

heat removal hed,e$ hanger. Westinghouse estimated that implementing this design 31 alternative wou reduce plant risk by at most 5.3E-04 person-rem /yr.

4 4 Direct Ste nerator Safety and Reliet Valve Flow to the IRWST: To prevent or

$$ reduce fisslun~ product release from bypassing containment during an steam generator

%ptubeTufture (SGTR) event, flow from the steam generator safety and relief valves could M'be ~ directed to the IRWST. An attemative, lower cost approach to this design attemative uld be to redirect the flow only from the first stage safety valve to the IRWST. I Westinghouse estimated that implementing this design altemative would reduce plant risk by at mest 4.2E-04 person-rem /yr.

7

.,..m.mw - m., .e+-.

I

~

(8) Increased Steam Generator Pressure Capability: In lieu of design attemative (7) above, fission product release bypassing containment could be prevented or reduced by I increasing the steam generator secondary side and safety valve set point to a level high l enough so that an SGTR will not cause the secondary system safety valve to open.

Although detal'ed analyses have not been performed, it is estimated that the secondary side design pressure would have to be increased by several hundred psi to make this l altemative effective. Westinghouse estimated that implementing this desi alternative I would reduce plant risk by at most 4.2E-04 person-rem /yr.

(9) Secondanj Containment Filtered Ventilation: This design alte ve ti uld involve installing a passive charcoal and high efficiency p , , te a er system for the i middle and lewer annulus region cf the seconda concrete co inment(below //

Elevation 135'-3"). Drawing a partal vacuum o ~e middle 'ul ' an eductorwith motive power from compressed gas tanks wou ' operate he filter sy QThis design altemative would reduce particulate fission product relea)sifrom any fa ed contain penetrations. Westinghouse estimated that im tenting this design alteli stfve would l reduce plant risk by at most 7.4E-04 person-re ~ ' p* i (10) Diverse IRWST Injection Valves: In the curr esig ~hh squib valve in series with a check valve isolates each of the four IRWSThection p  ;;To provide diversity, the design could be modified so that a differe'nt vendor providehhe valves in two of the lines. This enhancement would reducithehelihood of comrEion cause failures of the four IRWST injection paths. Wes}in'ghousej stimated that irdplementing this design alternative would reduce plant risk. by at mjet 5.3F-03 3 person-rem / reactor year, which would represent eliminating all',oore damage sequences,resulting from a failure of IRWST injection (3BE sequinces), f /

/P~ .J /

(11) Diverse Containment Rebirculation/falves: ,In e current design, a squib valve isolates each of the four containr'nent recirculation paths. In two of the four paths, each of the sgulb valves is irgs# $

eries with a cheI valve.' in the remaining two paths, each squib valve is in series with a motor-operated valve (MOV). To provide diversity, the design cduld be modified si t a different vendor provides the squib valves in two lines. This enha'noement would r ~ het likelihood of common cause failures of the four containhient recirculatio thskWestinghouse estimated that implementing this design attemative wo'uld reduce plant risk by at most 1.5E 04 person-rem / reactor-year, which would repress elirninating all core damage sequences resulting from a failure of ntainment re tion (3BL sequences).

(1 O! Ex Vessel Core r: This design attemative would inhibit core-concrete interaction n (CCl), even in . es where the debris bed dries out. The enhancement would involve 0 structure in the containment cavity or using a special concrete or coating,

[h; designing The curr . . P600 of design incorporates a wet cavity design in which ex-vessel cooling is i hhused to maintain core debris within the vessel. In cases where reactor vessel flooding I Qhas failed, the PRA assumes that containment failure occurs from an ex vessel steam l explosion or CCI. Westinghouse estimated that implementing this design attemative would reduce plant risk by at mest 6.1E-03 person rem / reactor year.

l I

1 8 1 l

l i

J

~

(13) High Pressure Containment Design: The proposed high pressure containment design would have a design pressure of approximately 300 pai, and would include a passive cooling feature similar to the existing containment design. This design would reduce the likelihood containment failures from severe accident phenomena such as steam

. explosions and hydrogen detonation. However, this alternative would not reduce the frequency or magnitude of releases from an unisolated containment. Westinghouse estimated that impismenting this design attemative would reduce plant risk y at most _

, 6.1E-03 person-rem / reactor-year.

(14) Increased Reliability of the Diverse Actuation Syst DAS): gn alternative would involve improving the reliability of the DAS. '_DAS 'riorisafety system that can automatically trip the reactor and turbine ang tece ~ ed safety features (ESF) equipment if the protection and gafety monit is unable to[f perform these functions. In addition, the DAS r r$ ides dj 'e mon "'of seloded plant parameters to guide manual operation an confirm reactor trip an f.SFj#

actuations. Wtstinghouse estimated that impi ~ ~ '~this design altema%efwould ,

reduce plant risk by at most 2.2E-04 person-re r. l y s. ,

3.3 Staff Evaluation i The staff reviewed the set of potential SAMDAs- ntified by Westi VM<

e and found it to be reasonably complete. The activity was acco,6pii.hed by reviewing design attematives associated with the following plants: Limerick'(NUREG 9974),LComanche Peak (NUREG-0775),

CE System 80+ (NUREG 1462), Watts,Bar (NUREG 0498), and the ABWR (NUREG-1503).

Also surveyed were accident management strategies (NUREG/CR-5474), and attematives identified through the Containment Performance improvement (CPI) Program (NUREG/CR 5567,-5575, 5630 / a'hd;556 9 7 h W $

The results.of the staff's assfossment are 'mtriarized in Appendix A to the " Review of Severe Accidentjaltigation DesjprfAltematives ( . DAs)for the Westinghouse AP600 Design"(SEA 97-2708410-A;1) prepared,by Science an ' Engineering Associates Inc., and dated August 29, 1997. Tlia(appendix briefly~eummarizes each of the design attematives identified in the foregoing references.' Also inElyded are the Westinghouse AP600 design attematives, which are discussed ih%ppendix 1B of to -SSAR. In all, the staff reviewed more than 120 possible design alternativesinclucAng most 'rsprovements identified as part of the NRC's CPI program.

Specific niprovemoritsiconsidered applimble to the AP600 included a filtered containment vent and aJio)oded rubble bAfcore-retention device, two improvements specifica NUpEG 0660 for evaluatior(as~part of Three Mile Island (TMI) Item II.B.6. The list of 120 also inglisded potential SAM As 6riented toward reducing the risk from major contributors to risk for

, including SGT , events.

A .

the We house analysis did not consider several design attematives, in most

' the 6cluded attematives are either (1) already included in the AP600 design, or (2) in~ terms of risk reduction by one or more of the design attematives that were included in t 'Wesitinghouse analysis. In other cases, the design attematives were pertinent only to boiling water reactors. The stcff's preliminary review did not reveal any additional design altematives that obviously should have been considered by Westinghouse. Also, 9

~

Westinghouse considered some potential design attematives to be considerations for accident management strategies rather than design alternatives.

The staff noted that the set of SAMDAs reviewed by Westinghouse is not allinclusive, in that additional (perhaps less-expensive) SAMDAs could be postulated. However, the benefits offered by any additional modifications would not likely exceed those for the modifications evaluated, and the costs of altamative improvements are not expected to be less than those of the least expensive improvements evaluated, when the subsidiary costs associaldfd with maintenance, procedures, and training are considered.

The discussions in Appendix 1B of the SSAR do not spech bas process that Westinghouse used to screen the many possible designJheri1atives ~a t the finallist of7 14 selected for further evaluation. Similarly, Westingho6se's res a tb _ ~' staff's requesif for additionalinformation (RAls) provided few additionalinsightsJdo'the pr .' Nonethieless, as noted above, the staff's review of the more than 12 candidate designs did 'iot identify any r new alternatives likely to be more cost-beneficial than included in Westinghouse's evaluation of AP600 design alternatives. On this basis, ataff concludes that the set of potential SAMDAs identified by Westinghouse is acce leM,g 3.4 Risk Reduction Potentialof SAMDAs Wig 3.4.1 Westinghouse Evaluation

.,!RMb W

'a  %

4%

in its evaluation, Westinghouse assum that ea des Taltamative would work perfectly to completely eliminate the respective accident se maximizes the benefit of each design altemative,q6ences.

, which is measured on the basis Jhis' of risk assumption is conservat reduction. (For example, the risl[ reduction pfssigned tofssive containment sprays assumes that ali release categories except containment bypass are eliminated.) In each case, -

Westinghouse used analytical models andresults contained in the AP600 PRA to estimate the terms of[whole body person-rem per year (rebeived by the total population w 80.5 km (50 rn!.) of the AP600 plant cite. Each of the 14 design alternatives was evaluated separately. ' 77 ,

gi$% gp&g 9

Table 1 of this'EA aummarizes Westinghouse's risk reduction estimates by comparing the benefits these es,of averting timates are pr offpein' section exposure using each potential design alternative. The bases for 1.B.7 of the SSAR.

M m  ;'

N Qg 10 l

{1)

"s ,

i f

t e0o0m f

n0w$

e5r 9 3 6 9 0 7

2 5 8 4 8 2

7 3

9 2 2 9 2 5 7 2 1 1

8 2

8 5 2 1 5

5 1

4 4 3 B$s1o9 4 4 a Oep f r f

t S -

"s0/ m t

i f

e e0r$

n0 6 71 0 2 7 3 6 1

3 3 2 9 5 3 3 7 2 e2n9 3 2 0 8 1 1 1 2 5 8 8 2 2

B$s1o9 1 1 1 1 m

a Oep f r f

e e t r

= S n

+ o s

c u $ r p o e x h

• 5 E g s 0 e nei t 7 0 n h sf e 4 6 5 4 9 3 3 3 5 4 1 9 9 2 4 s s n 4 3 3 3 0 e 3 1 W Be f

f O ,

d r f e

e kp

,e i l

t r

e s t n f e

v Rm r i

5 5

0 0 4 7

0 9

0 1

3 5

2 4

2 4

4 7

0 5 0 0 2 4 a i l

A dr e a 3 l 1 1 1 2 p t an 0 1 0 6 6 0 0 0 0 3 5 0 6 6 0 m teyn e 7

0 0

0 0

0 0

0 0

0 0

0 0 0 0 0 0 0 0 0 yr l o r o 0 0 0 0 0 0 0 0 0 m - p f

r es vr 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 0 3 y r-t s Ae p e,0 i

f e t a 6 n $ 0 r e

e dt ,

0 0 0 0 0 0 0 0 t t B es 0 0 0 0 0 0 0 0 0 0 0 0

0 0

0 n u ar 1 1

d aCo t 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 o t t

e n 0 0 0 0 0, 0 0 0 0 0 0 0 0, 0 sc u n a tw a l 0, 0, 0 0, 0 0, 0 0, 0, 0 0 0, 0 0 i m sd 0 0 0, 0 0 0 0, 0 0 0, 0, 0 0 0, d oc h Ep 0 0 3 0 0, 0 0 0 0 0 0 6 0, 0 e s s a 5, 0, 3 2 7 5 6, 0 7 i hv d 9, 0 3, 2, 2, E C 1 5 3 2 6 5 1 4 c veu 1 8 2 1 5 f

o l f e

n r

o a v

f e ce o e f o

m e y

i s i t

v T. v t i m %f 7 e f

a r

a s m e a l l i

e p n Gy s y a e R V b a n

. 5 %

m r

e e v r t s t p o s y 1 7

o t S p y a le a i e S ,

C l

A l

a S e b C t

a v n s s ,

V d S l

n o H e l l a o t

c c t

e i

r 1 g t n n r .

i e R r t

n V i

t f e e o e ;tu u f i

o v d e n a ~ f f e s i s n s V n u e e e a s. c i

l e C t s o v b

a D  ;

r. n e s a e d s ta i s c v e e 2 t

t e l o j a y r e e e A a s s 6i t

i t s

T n m s " u n y i P e t P r v u l D e f 3

o u I

e fa e n f f f

e o o n l

i t l o tn d C- fte o

d r

o l t

a c t n

s r r a i i V c r e r r o S- Sr F t i g V e n a i

a n e e v t

l o of h i

i M

t n t d e Se d V w- I l l

e to r

e n

t n

e i o R h

r e m n D i

A n

f l

y y ec a t n n S l

a n d d r e i c tn c i

a f g t

a u h r s e e o i o o n n l a e r S e G n i

j t

a t n y s e t e c m i

u G s r n  ;

n m C o ;i n t n

i n

d n C y s o e a D u u sos m I C  ;;,

e T n o o i t i

c m a g i r

t a G a n i e l e i c e t e P r e O S ta r o e a v c c n tn fa t

A e m C r i c c csu t

n u a a i

h i n S W C l

e t e

r C Co a S %i u  ;

g e a e d y r R C

o l e

s s R c

e s s e i i G S t e a I s e e f t i si der Ac eve 'er t f i t e d eee e f

v s d 6 s r s E f i

e S y y- e e sa tc f

- s v i e

m c a e t a e

r n

o c

5-5 V

e P ea i

i

% n e n e S t F t

l i

l S P A t

S D r

i c n e t h x- v cnr  ; 0 0 B B I S O D E 'r i i 1 jtl j j'

,l :l: i

3.4.2 Staff Evaluation i I

The staff reviewed Westinghouse's bases for estimating the risk reduction associated with the various SAMDAs, and concluded that Westinghouse used a reasonable, and generally j conservative, rationale and assumptions as the bases for the risk reduction estimates regarding j aach design attemative.  !

The level of risk reduction estimated for the various SAMDAs is driven .two rlying assumptions in the methodology. Specifically, Westinghouse's risk re n%stimates reflect i d the . ~ estimate (mean) only the values contribution without consideration offrom intemal uncertainties in coreevents da initiated at powerf'frequen CDF) o 1(%t,'or consequences. Although this is consistent with the a ch taken pr ~ design altemative evaluations, further consideration of these

//

tly higher ors cou fead to j risk reduction vclues, given the extremely small CDF risk atos in the _ line PRA for {

intemal events.

/ W In assessing the risk reduction potential of SAMDAs for '

Westinghouse's risk reduction estimates tt'ematives;infor the variousi conjunction with APb des supplementary parametric analyses to evaluate the pote'ntial !?iQact of extemal events and l '

uncertainties. These analyses are further discussed in Section 2. of this'EA.

gjy ~ -

37 i 3.5 Cost Imoacts of Candidate SAMDAs /g. M y *A

^*

3.5.1 Westinghouse Evaluation g j

/

Sections 1BA.2,18.4.3, and 18.8 of the

./. ~

Q 7

f discuss the capital cost estimates for the AP600 1

i i

design a er' natives evaluded by Westinghouse, and Table 1B.8-1 of the SSAR presents the results cost evalu'ations.: Specificallyhor each design attemative, Table 1 B.81 lists the l potential reduction, the'cqpital benefit (assuming the design alternative was highly effective in reducing acqident risks), the'p cost, and the not capital benefit. Notably, )

id not account for factors such as design engineering, I Westinghouse'sunt testing, anyanintenance~associaevaluations (ted with each design alternative. If included, th '

would increase the o'verall costs and decrease the capital benefits of each altemative. Thus, i thisa  !

j ' ach is conse' lg.W )

x

.2 Staff Evaluation .

N j%.

gauge the .onableness of the cost estimates that Westinghouse presented in the SSAR, .

staff compared the capital costs for the AP600 design alternatives with those evaluated for i the BWR and CE System 80+ designs. However, there is not an exact match in the design altematives among the reactor designs, so only broad comparisons are possible.

12

For example, the AP600 active high-pressure safety injection system, which is estimated to cost $20 million, adds an active high-pressure safety injection pump and associated piping, valves, and supports, thus adding an entire new safety-related system to the AP600 design.

This altemative can be compared to the attemative high-pressure safety injection for the CE ,

System 80+ design, which is estimated to cost $2.2 million. However, the design attemative for the CE System 80+ design simply adds parallel piping and valves to an existing system, which would be expected to cost only a fraction of the total system price.

Similarly, the filtered containment vent for the AP600 de pan be 'to systems with similar functions for the ABWR and the CE System 80+ sna. T 1 design included a filtered containment vent and all associated piping and rations. _'ABWR design added' an ex-containment filter system to an existing venting em. The 'te ' ~ Idesign included a filtered containment vent similar to the multi venturl scrubbing emsim tedirrsome Europe an plants. The estimated costs for the three veliting sy ms- $5 milli "$3 million,  ;

and $10 million, respectively - reasonably agree with a other given the differe' Tin the j designs. M '

The costs for the non safety-grade containment spray for AP ' "n; which was evaluated in an earlier version of SSAR Section 1B before1t was incorporat 'ipto the AP600 design, can be compared to the reactor building sprays for the'ABWR design andWe attemative containment spray for CE System 80+ de,sig' n. This/ app design fttemative involves adding piping and spray heWars inside containment, and6onnects.to an existing fire water system.

Similarly, for the ABWR, the existing ir/c~ontainrpent fire shy system would be modified to provide sprays in areas vulnerable,to fission pjoduct relepse. The ABWR modification would thus be limited to providing sprays ont to selected areas of containment. For the CE System i 80+ design, this alternative invopes adding' piping tojdn~nect to the existing in-containment l spray systeyn, together witjinew pumps tg supply tiie water. Estimated costs for these three spray systems were $41f,000 for the AP6QO design, $100,000 for the ABWR design, and ,

$1.5 million for the CE System'80+ design.7n light of the scope differences among these

. l design aile tives, the estT tes fo' rthe AP600 spray system appear to be reasonable.

l 5%

These com indicate that ths;c6st estimates for several of the AP600 design a alternatives reason ' agree'with the costs for roughly sirnilar design alternatives evaluated for l the ABWR and the CE 80+ designs.

? y j l

fdrther assess the sonableness of the AP600 design attemative cost estimates, the staff edindepe cost estimates for one particular design altemative, the active, non ety-related coplakiment spray system. (This analysis was performed before the j ty-gradispray system was incorporated into the AP600 design and deleted from SSAR i 18.)Mie assessment assumed the addition of fire protection system grade spray he ers and supply piping inside containment (carbon steel), and the addition of control valves and piping outside containment and connected to the existing fire water supply system. The resulting costs for the containment spray system ranged from about $300,000 to $350,000  ;

(1996 dollars), depending on the assumptions made regarding the required pipe size. These 13  !

l t

m-.._.__ _ . . . _ . , , _ . , . - . . _ -_ _

independent estimates did not include design engineering; first-of-a-kind costs; or allowances for associated personnel training, procedure development, or recurring operations and maintenance costs. This approach is similar to that used in Westinghouse's cost estimation.

Thus, the Westinghouse estimate of $415,000 for this design attemative reasonably agrees ,

with the independent estimate. In addition, the staff developed an independent cost estimate {

for a containment spray system similar to that described above, but with increased pumping capacity. (The increased pumping capacity is needed because Westinghouse's lett.or of March 13,1997, indicated that the currently designed fire water supply ystem ipfitpable of delivering less than 1.89 kUmin (500 gpm) to the proposed containme spray 4ystem.) The system evaluated for this altemative would increase the fir ater pu '[ capacity so that each pump would be capable of delivering 11.36 kUmin (3000 ,to th talriinent sprays against a containment pressure of 310.3 kPa (30 psig). Iping u to"siupply, fire waterjt"  ;

the containment in the current design would be increass In size to ucY flow resistandel This modification to the AP600 design was estimated t'fcost a . $'370, $996, dollars): As with the foregoing estimate, no allowance was made f fpersonnel training, pr development, or recurring operations and maintena .V 7 j

1. 5 ~

h On the basis of this audit, the staff viewed Westingho '.?e's appto mate cost estimates as estimates, and the level of adequate, givengiven precision necessary thethe uncertainties greater uncertalsurrounding t[ inherent on thethe t underlying cost'Ie6efit side, w!

costs were compared.

'? 9g%gg f lK fj 3.6 Cost Benefit Comoarison .,

nt -

Eh  ;

3.6.1 Westinghouse Evaluation [ [ ..j /

s . N.k h After considering the ris toduction potentia and cost impact of the various SAMDAs, Westingliouse performed c'ost-benefit comparison to determine whether any of the potential severe accidertt mitigation deIsign features would be justified. To do so Westinghouse assessed the bdnefits of each design alternative in terms of potential risk reduction, which was defined as the' reduction in whole b6dy person-rem per year received by the total population within aJadius 80.5 k (50 mi.) of the AP600 plant site. Westinghouse then assigned a value l of $1,000 to each per 1-rem lof averted offsite exposure, which was assumed to account for property damage. This value was treated as the annual levelized bot fiealth b t for averted effects risk. and @y

$determine thy ximum expenditure justified by a given reduction in risk (" maximum capital

, Westl~nghouse divided the annual levelized benefit by the annual levelized fixed rate 4The annual levelized fixed charge rate was determined to be 15.7 percent in current U.S. dollars on the basis of factors and methods provided in documents developed by the Electric Power Research Institute (EPRI P 6587-L) and the U.S. Department of Energy (DOE /NE-0095). Westinghouse calculated the fixed charge rate using a component " book life" ,

of 30 years. The use of a high charge rate tends to minimize the capital benefit associated with j 14

each design altemative. Nevertheless, the 30-year life used in the calculations makes little difference in the economic benefit compared to the more typical 60-year life, particularly when the high levelized annual fixed charge rate of 15.7 percent in used.

4 The Westinghouse approach for calculating the benefits or reduced risk from each individual design altamative also does not give credit for averted onsite property damage a ,

replacement energy costs which are realized through a reduction in acc dont fr ncy. The onsite property damage and replacement anergy costs may have bee [1eglepted because the estimated CDF is very low. However, as indicated below, t e onst conside' rations can

^

substantially add to the benefits that may be achieved u _ ign d ".

Table 1 of this EA reports Westinghouse's cost-benefi '~timates each SAMDA' using a screening criterion of $1,000/ person rem-ave toide/vifywhetherany [ thief i SAMDAs could be cost effective. As shown in Table 1 'highe.st capital benefit ' iculated by Westinghouse for any design alternative is about $50, , the capital cost for the least expensive design alternative is $33,000. On this basisjWestinghouse. concluded that no additional modifications to the AP600 design are warrhnted. 18/>

3.6.2 Staff Evaluation y ' %g WAk b

y \

p ,

g, The NRC recently updated its recominonded a ch for exposures. Previous guidance specifiedof thalgperson-reg,the exposure should be valued at trionetary c

$1,000. This conversion factor fpr offsite doses was intended to account for both health effects and offsite property damage, and exposurdincurred in future years were not to be discounted.

The recent guidance giveniri this NRC's r latory' analysis guidelines (NUREG/BR-0058, Revision ), recommends'using $2,000 p ' person-rem of exposure as the monetary conversio)n factor. In addition for assessin values and impacts, future exposure discounte'd to arrive at thel?ptesent worth. Offsite property damage from nuclear accidents is .

to be separa' valued, and if ~ part of the $2,000 per person rom value. l I

7. ibm hj/  !

l w^

Evaluations ricently .

by Brookhaven National Laboratory for the NRC assessed total I costs, associated with effects (NUREG/CR-634)(; Costs were assessed for each of the five NUREG-1150l

( " d Gulf, Peach Bo cim,3equoyah, Surry and Zion). The results indicated that overall associated with its releases of radioactive materials, presented on a cost per  ;

rem of e ' to the public, ranged from about $2,000 to more than $5,000 per rom, de , ~g on factors such as the assumed interdiction criteria. A criterion of 000 per -rem averted was added to account for offsite property damage and other 4 r ed costs foisevere accidents. Thus, the Westinghouse cost benefit evaluation approach used for AP600 design attematives is not consistent with the approach recommended in .

NUREG/BR-0058 Revision 2. The key differences are summarized in Table 2 of this EA, and j the staff's independent evaluation is found below.

l l

1 15 l

--e .-e...n. .xem --

- - - ' ~ ~ ~ ~ * ^ " ' ' ' ' ' - " ' - ' * " ' * '

Table 2 Key Differences Between the Westinghouse Approach and NUREG/BR-0058 Westinghouse's SAMDA Approach NUREG/BR-0058 Recommended Approach

$2,000 per person-rem averted to account for

$1,000 per person rem averted (for valuing health effects, plus $3,000 per -re m risk reduction) averted to account forpther e effects and related costs A 15.7% discount rate 7% dis ate <

1 Jf?f No accounting for benefits of averted onsite cleanup and decontamination costs Consi Q}iven ration g for benefits erderted onsite ' up and decontamination costs

[N$Nh

%$$ib No accounting for benefits of averted f ration giv ~forb'enefits of averted replacement energy costs g replacement, energy sts

/ h@;gpA g y u

~

a reduction. offsite risk, the staff applied the To recommendedarrive approach at a baselinein NUREGKiRpotential 0058, ber[ fit from tpRevision 2h the des i the AP600 design. This EA used a discour}t'ra'te of 7; percent and assumed a reactor life of  !

60 years. The averted riskJdr each desigr[altemative was taken from Table 18.8-1 of the i SSAR. In addition, the staff used two monetary conversion factors for radiation exposures. The l first is the($2,000/ person'irem recommended in NUREG/BR 0058, Revision 2. The second, j

$5,000/ person-rom, is inten'ded to account for offsite property damage and health effects. The l results for ea'chs design alternativq are shown in columns 5 and 6 of Table 1. For comparison  ;

purposes, Westinghouse's estimales of the capital cost, averted risk, and capital benefit for each l design alternative are also presente'd (columns 2,3, and 4 of Table 1). A 100% effective design '

l alternative Gould redu'ce the CDF and/or offsite releases to zero. Estimated benefits from a 100% effective design mative'are also shown for each of the attemative cost bases (last row of T 7

1). y i

A

'results shown in J le 1 Indicate that the benefits calculated using a 7-percent discount r ",'a.60 year play We, and a $2,000/ person-rem conversion factor is about a factor of four r thanMose calculated by Westinghouse. TDe benefits calculated using

~rson-rern are about a factor of 10 higher than those estimated by Westinghouse. The hi ^ ' dspital' benefit shown in Table 1 amounts to less than $500, while the capital cost for the least expensive design alternative is $33,000. Thus, even with the highest benefit basis

($5,000/ person-rem,7-percent discount rate,60-year life), the calculated benefits are almost two orders of magnitude too small to justify the addition of any of the design altematives listed, it should be noted, however, that this assessment neglected the benefits from averted onsite 16

r costs, which are relevant for design altamatives that reduce core damage frequency. Dollar savings derived from averted onsite costs are treated as an offset or reduction in the capital cost of the design attemative in the staff's analysis. Averted onsite costs are significant for certain

. design attematives and are further considered below.

. 3.7 Further Considerations The estimates of potential design altomative benefits listed in Table 1 of is Westinghouse's estimates of averted risk and neglect the benefits fr rt site costs. As mentioned in Section 3.2 of this EA, Westinghouse's risk e atos icoou, nt for uncertainties either in the CDF or in the offsite radiation a ' es r ~from a core damage event. The' uncertainties in both of these key elements a fairlylarge , ma key safety

~

/

q features of the AP600 design are unique, and their relia has be av through / l analysis and testing programs rather than operatin Hence. kddition, " estimates

"~~

of i CDF and offsite exposures do not account for the a risk ~fr" "extemal eve as' l

. earthquakes. #

To further explore these areas, the staff screened the date SAMDAs to determine whether any of the design attematives could be cost-beneficial 6 hen t " inefit analysis incorporates uncertainties, added risk from extemalevents, and rted onsite costs. The staff then performed a more detailed assessment for.atiose' design alte having potentially favorable cost-benefit factors under these m lim considerationarThese analyses are 4

^

discussed in Sections 3.7.1 - 3.7.3 below. [

?

3.7.1 Uncertainties in Core Damagefrequen Acci elated Exposures

/@  ? l Revision 8 to the PRA discussed he uncerta in the e ated CDF for the AP600 design. -

Specifically, the CDF uncertainty)dist' ributior/was characterized by an error factor (

n6rmal distribu about median, 5.7.- and also Assuming the ratio a log'of the' median to{l6n, the GF is the ratio of the 95th perce jhe 5th percentile. Thus, the CDF for internal events could be Afactor of six higher or lower than ass'umed in the analysis discussed above.

%my Myg.

Additional facitors that could su ~ ~ increase the estimated CDF for the AP600 plant include the cont'rlbutions from events and' accident sequences that have not yet been identified, as well as those ideritified sisquencds.that have not yet been analyzed in the PRA. Examples of <

the latter jriclude extem' gl events such as fires and earthquakes. Notably, the CDF base estimate #of 1.7E-07/ read pyear'does not include the contribution of extemal events. In the PRAjWestinghouse indi cattd_ that extemal events, in particular intemal fires, are estimated to inc{ ease the CDF by ab9 Jt a factor of four. However, the PRA available for this study did not the potential contrbutions from seismic events, which could readily increase the CDF by

' order of magnit " more. These external events can also degrade the containment nee, so the releases from containment may also be higher than for accidents inal events .

The tal increases in CDF attributed to accident sequences that have not yet been identified is very difficult to estimate. Presumably, the contributions from such sequences should be small if Westinghouse performs the PRA in a thorough and systematic manner. For the purposes of 17

.e.. ~ v. u.v.n e ~<-e.~e.o.mm-s _~=v *y w~m='m**

_ 'y ' ' " '

• the present analysis, the effects of these sequences are assumed to be captured by the potential increase in CDF attributed extemal events.

Section 18.6 of the SSAR presented Westinghouse's estimates of offsite exposures for the major release categories (RCs) defined for the AP600 design. On the basis of the CDF reference value of 1.7E-07/ reactor year and the total risk of 7.3E-03 person-rom / reactor-year,-

Westinghouse estimated that the " average" offsite exposure is of the order of 50, . person-rom per core damage event. However, Westinghouse's documentation did ot indi the uncertainty in the estimated releases, y The average offsite exposure of 50,000 person-rem per A core Westinghouse is a factor of 2.7 lower than the average p

" %ent estimated bg

^

i for the fivq, current-generation nuclear plants addressed in NUREG 150"exposur [if the NUREG-}160 (after 's ~

plant releases to that of a 600-MWe plant). The better " orma sf the A 'deeigninay be attributed, in part, to methods and assumptions for defi ao " terms, as 'the.high likelihood of successful RCS depressurization and in ve rete'ntion of damaged in the AP600 design.

> }yiIg% !f*1 Uncertainties in the offsite exposure estimates for the 600 aresignificant. As described in Section 19.1.3.3.3 of NUREG-1512, tha AP600 Fina ,.

Evaluation Report (FSER), the AP600 risk profile is shaped by th major ass s regarding containment failure modes and release cha g@'

• conservative assumptions regarding early an failure from ex-vessel phenomena e optimistic assumptions that exterhal react 8 vessel i will always prevent reactor pressure vessel (RPV) breacRC / s e substantial credit for addit dnal ser removali GTR events If early containment failure ' avoided (as f }  !.

by deterministic calculations performed I subsequept to the PRA) reactor pres re vessel breach instead results in a more benign release (e.g., s a containm failure in the intiermediate time frame), overall risk for internal events would be reduced by about of two. By contrast,if credit for external reactor vessel ,

cooling (ERVC reduced or ted, containment failure frequency would increase proportionally, nos;all RPV brea ~

are' assumed to lead to early containment failure in the baseline P ~ the most limi ng assumption that ERVC always fails and leads to early containmen't failure, tII txintainment failure frequency would approach the core melt frequency '

and risk $ould increase ' la. factor of 20 (to about 0.16 person-rem /yr). Similarly, offsite risk can be icantly impacted Ke design falls to realize the decontamination factor (DF) of 100 l ap to aerosol relea 1 fractions for SGTR events predicted by the modular accident analysis p l ram (MAAP) to a 'nt for fission product removal by impaction on steam generator tubes. ,

V this credit for a ' I removal, the risk contribution from a containment bypass is minimal

'(I $mroentof the . ' Without this credit, overall risk for infomal events would increase by a i fa ' of seven%nd would be dominated by containment bypass releases. Finally, the PRA did i the' Impact of the non safety-related containment spray system on fission product j rele@asssFContainment sprays could significantly reduce the estimated rij (by perhaps a factor of 2), since the sprays would be effective in reducing the source terms in j the risk-dominant RCs such as early containment failure (CFE) and containment isolation failure j (Cl). However, sprays would not impact releases attributed to SGTR events.  ;

i 18 l

I l

i

_ . _ __ _. . . _ . . _ 1

e- - - -

In summary, the actual offsite exposure could range from a factor of two lower to an order of magnitude higher than the Westinghouse estimate, given the uncertainties in the underlying

. analyses of containment performance. This uncertainty range was factored into the staff's reassessment discussed below.

3.7.2 Reassessment of Design Altemative Cost-Benefit Relationships in Light of Uncertainties

~

The staff-performed analyses reassess the benefits of potential AP600 design ,,

tives taking into account the uncertainties in estimated CDF, offsite releases of radi iv terials given a severe accident, and effects of extemal events. For these a ~yses, l liiff iitimated the maximum benefits that can be achieved with AP600 desig iaissuining that a altemative can either completely eliminate all core dama sients or ~tely' eliminate offsis l releases of radioactive materials if a severe accident " occur. "Ic ' these estimated l benefits, the staff used the FORECAST code (NUREG/ 5595, ision 1, RECASTP Regulatory Effects Cost Analysis Software Manual, Ve 4.1A cience and fNff%f Associates, Inc., July 1996). FORECAST allows the us [diri thes parameters and providos a means to combine uncertaint

~

~

provides a distribution for the bottom line costs or benefitsfa thus presents a picture of the uncertainty in the " bottom line" figures. Table 3 of thistA pres the key parameters used in evaluating the maximum potential benefit, g )

Pnh,~ w Table 3 Key Parameters Used by FOR, Evaluating Maxmum SAMDA Benefits Parameter [ [ %d%F Value Reference AP600 core damage %ency[ 1.7E[/ reactor-year (EF=5.7)

Average public radiation expgdure per fy" 4a)ssumed error factor:4 ,200 person-rem (ro accident: . 77 fi - 5)

Plant Igetime g >M $$6@ ~60 years

%h Discount rate' 7%

Conversio'fijactork  %@4 $5000/ person rom Replacemedenptgy costs V $277,000/ day of downtime j

8 Ave.rted cleanup'an%@ination costs $1,690,000,000 / major accident j$erted replacementjinhy costs * $20,200,000,000/ major accident

, on NUREG/CR ,-eccounts for both offsite health effects and offsite property damage effects 8

on guidance in NUREG/BR-0184 (not adjusted for AP600-specific features) i on average nt energy costs for pressurized water reactors in the 600 - 1000 MWe range  !

F Yh% x.f

, purposes of estimating the maximum potential benefit from AP600 design attematives, the assumed that extemal events and accident sequences not yet accounted for in the PRA would increase the reference CDF by two orders of magnitude, (i.e., a factor of 100), with an EF of six used for this higher CDF. The staff then evaluated cases assuming the reference value of 50,000 person rom per accident. Table 4 of this EA presents the results of this analysis. j I

19 I

i

Table 4 SAMDA Benefits Accounting for Uncertainties and Extemal Events Effects (Benefits,1996$)

" 5% 95 %

Description Confidence Mean Confidence I av.i A

uvei F

Base CDF (1.7E-07/yr) and reference offsite release (50,000 person-rem); design

$26,600 alternatives which reduce the accident frequency to zero p1, ,

y g ,

/

6

., L -

Base CDFincreased by factor of 100 to account for extemal events and other g.g'{g accident sequences not yet accounted for; 2 /

.500 $647,000 $2,257,000 other factors same as Case 1; design ,

.g w yh 4

attematives which reduce the accident frequency to zero 7 ?{h Base CDFincreased by factor account for external events; ot factors ,

hi 3 same as Case 1; design alternatives wpich [$1,700 $49,000 $223,000 reduce the offsite releaseykzero, b do not t' change the accident frequency { i

, (N x]Y f JYY:R Y2WW The entries in Table 4 inacate that design attematives which prevent accidents (reduce the accident ' frequency to zero)'are'much more cost-effective than design alternatives which reduce 1 or eliminate offsite releases bdt' ve no effect on accident frequency. This is because of the I fairlylarge ben 5 fits associated averted onsite cleanup and decontamination costs, and I avoided rep'lacoinent energy costs, neither of which are assurned to be impacted by design I altamatives which do reduos,accillent frequency.

/

Casej1is the reference

lkW utilizing the base CDF and Westinghouse-estimated offsite m

e 'res. In this case eistimated benefits are considerably higher than those cited in T 1 of this EA, p ' ly because they include averted onsite cleanup and decontamination

, as well as ave replacement energy costs.  !

% ,' I

'2 and.a the effects of the higher CDF associated with extemal events, but do not i .the effects of possible higher releases from containment attributed to such events. (In .

otheiwords, these cases retain the base offsite exposure of 50,000 person-rem / event.) These cases may be used as the basic benefits including extemal events and assuming that extemal events would not impact containment performance. Case 2 shows the potential benefit range for a design altamative which could reduce the accident frequency to zero. Case 3 applies to a 20 i

l

design alternative which would eliminate all offsite releases, but which would not impact the CDF.

Table 5 of this EA combines the information in Tables 1 and 4 to estimate the total benefit possible from specific design attematives. The design attematives are divided between those that impact the CDF and those that impact containment performance but not the CDF. Benefits have been estimated by taking the fractional reduction in risk for each design alterpative (compared to the AP600 baseline risk as defined by Westinghouse) and pplyingest fraction to the mean benefits displayed in Table 4. Design alternatives that reduc e Cghere applied to the Case 2 mean benefit, while those that only effect contain nent per noe were applied to the Case 3 mean benefit. ) 4 M hA, [

The values shown in Columns 4 through 7 of Table 5 r t benefits Icu ed using the aan values. By contrast values shown in Co!umns 8 throug 11 werepilculated g theb #

95*-percentile values, in other words, there is only a 5 will be greater than the values shown in Columns 8-11. rcentdhance that the'g' benefits j Pi 1 Nik The use of the maximum benefits typically improves the steenefit ratio by a factor of l

l approximately five, but does not alter any of the overalkonclu about design alternatives that have acceptable cost benefit ratios.

t/N~2N,h..

"iQWS-y?

3.7.3 Further Evaluation of Design Alternatives W Potentially Favorable Cost Benefit Factors Design alternatives that are within a decad[e #f0of m8 criteria of ngt , cost-benefit

$5,000/ person-rem were subjected to.further pr, bilistic and deterministic considerations, including a qualitative assessment the following: f yr p .g the impact of additional nefits that could accrue for the design alternative if it would be effective in reducing risk from certah extemal events, as well as internal events

/ effects ofg:improvements the k jp.y already(made at the plant y.

any operational disa%g: h with the potential design alternative dvantage associated yp:t , Rps 1 None of the,designattematives havea cost-benefit ratio of less than $5,000/ person rem.

However,,the only de standard are the dive %n attematives that come within a decade of the $5,000/ pers rsejRWST valves at $19,800/ person-rem and the self actuating containment isolation val "sgt $33,700/ person-rem, as described in the following sections.

M "

&w, j

%ik 4dMef

' rippy 21

l E. E E. R R R il B  !. R I. H. B. 1

] !  !.  ! : i ! 18  !! 1 1

1 1  !

' ' I' il i .l.i ! ! ! .l i ,l l l i l ,l ii 1

. t j

11 l;i l!

J l 111 i l! ! !  !

lif :  ! 11!

l i )), j l l i l 8 5 l- 1 ! ! ! H i l l Il8l i

t 11 3 3 a iIl n I' i ! ! l !  ! ! !  !  ! I l l 1 1  !  ! l

! b l

i i  ! ! ! !

j

! ! i ! s i i i l!

!lil!

s j> !1 1

i 3

h jl l j 1 1 j

1 i

J l

1 1

lI 8

t i

l l ' # '

i b l  !

l l l l l si l h!

ly l i l l l ! a ,j il,!

i! n u l 1 ,li i l l i _

3.7.3.1 Diverse IRWST Injection Valves in the current AP600 design, a squib valve in series with a check valve isolates each of four

• lRWST injection paths. This design altomative would reduce the likelihood of common cause failures of IRWST injection to the reactor by utilizing diverse valves in two of the four paths. If it functioned perfectly, this design alternative could potentially reduce the CDF by about 72-percent.

When taking into account external events, other accident sequences not yet includpd in the AP600 PRA, and other uncertainties, this design altamative is estimated to be h hly cost-effective. In the absence of a comprehensive extemal events PRA for the AP600 plant, , . d' to estimate the effectiveness of this design altemative in reducing the risk frqm such owever,it appears likely that failure to inject coolant to the reactor would remaufa' prom' ' con;tnbutor to the CDF

~"

from external events, in which case, diversity in the IRW injection v elp to redufe the risk from both extemal and intamal events. ,

/ '

Alternative vendors are available for the check valves, er, is questionabEwNether^ check valves from different vendors would be sufficiently diffe lto'be considered diversiiunless the type cf check valve was changed from the current swing check valve to another type. The swing disk type is preferred for this application, and oth ~are considered less reliable.

WJh Adding diversity to the injection line squib valves would require a ' I spares at the plant, and some additional training for plant operations an,d Maintenance staff, would not appear to add significantly to the operational aspects of th . ,

r, a greater issue concems the availability and costs of acquiring diverse es fro asecond vendor. Squib valves are specialized valve designs for which therg a few _ ilor's Westinghouse claims that a vendor d

may considering not that theybewouldwilling only suppfyto design, two valves per plan qualify,,and build aJ aso squib valve for this AP600 application alternative assumes that a seconql squib valvsfvendor exdts and that the vendor would provide only the two diverse IRWST squi6 valves. }fie^ cost irrgfact does not include the additional first- l time engineering and qualificiition testing that would~be incurred by the second vendor. l (Westinghodse estimategidiat those costsfpuid i be more than a million dollars.) As a result, l Westingh'ouse concluded 1 hat this design attemative would not be practicable because of the i uncertaint'y10 availability of a'e'ocond squib valve design / vendor and because of the uncertainty in I reliability of a'nother type of check valve. The staff considers the rationale set forth by Westinghouse rjgarding the pote'rgal reductions in reliability and high costs associated with obtaining diyerse to be reasonqible. On the bases of these arguments, the staff concludes that this gn altern need not be further pursued.

~

3.7.3 rSelf-Actuating C ~

&&W ent Isolation Valves

t .

T design alternative uld reduce the likelihood of containment isolation failure by adding uating valves , enhancing the existing containment isolation valves for automatic closure containment ditions indicate that a severe accident has occurred. Conceptually, the ti would tielther an independent valve cr an appendage to an existing fail-closed valve that

" respond to post-accident containment conditions within containment. For example, a fusible link would inelt in response to elevated ambient temperatures, thereby providing the self actuating function to vent the air operator of a fall-closed valve. This design alternative is estimated to impact releases from containment by only 10-percent. It has a cost-benefit ratio of

$33,000/ person-rem, and achieves this ratio primarily because of its low capital costs.

23

This improvement to the containment isolation capability would appear to be effective in reducing offsite releases for accidents involving either extemal or intamal events. Also, the effectiveness of this design altamative would not be affected by the design changes made as a result of the AP600 PRA.

The addition of this design attemative would impose minor operational disadvantages to the plant in that the operations and maintenance staff would require some additional trainin In addition, these autorriatic features would require periodic testing to ensure that th are f ," ning Property.

Perhaps the biggest question regarding this design alte is wh ciiin be implemented for a cost of only $33,000. The cost estimate does not a ' r~to incl and qualification testing that would be required to dem ate that 9a$first-time Tould perform)~tf enginee6ng intended function in a timely and reliable manner. The ts a 'ed with ~~ . testing and maintenance also do not appear to have been included The statif believes that ' actual hosts of this design altamative would be substantially higher tha estinghouse's estimate ~'r haps by a factor of 10) when all related costs are realistically consi UOn the basis of the unfavorable cost-benefit ratio, and the expectation that actual costsy,lo' u Idgeven higher than estimated by Westinghouse, the staff concludes that this design alt (mative a cost-beneficial and need not be further evaluated. T 3.8 Conclusions ' ' k/. t,

~

As discussed in Section 19.1 of AP600 FAER, We 'g ~ ^ 'ex^ tensively used the PRA results to arrive at a final AP600 design. As a fr s6ft, the ' ated (bF-and risk calculated for the AP600 plant are very low both relatiVo t6 oper g plants " d in absolute terms. Moreover, the low CDF and risk for the AP600 d[esign~refle . estin e's efforts to systematically minimize the effect of initiators / sequences 1 hat have mn impo t contributors to CDF in previous PWR PRAs. Westinghouse has lehd this o $ctive tafgely by incorporating a number of hardware improvementsin the AP low CDF risk for the design. Thes .other AP600 design features which contribute to design are scussed in Section 19.1 of the AP600 FSER.

)

lW $%

Because the . design al Tciontains numerous plant features oriented toward reducing ,

CDF and risk, , benefits and ri on potential of additional plant improvements are significantlyj w=t is true fo@' rboth intamally and externally initiated events.

the features already i ~ f _in the AP600 design, the ability to estimate CDF and risk approacks the limitatio probabilistic techniques. Specifically, when CDFs of 1 in 100,000 or 1, years are estir. , in a PRA, it is the area of the PRA where modeling is least co te, or supporting sta a sparse or even nonexistent, that could actually be the more i ant contributors fisk. Areas not modeled or incompletely modeled include human

) r ty, sabotage initiating events, construction or design errors, and system interactions.

h impr a in the modeling of these areas may introduce additional contributors to and .

. staff does not expect that additional contributions would change anything in a 'tery; i

24 1

The staff concurs with Westinghouse's conclusion that none of the potential design modifications f evaluated are justified on the basis of cost benefit considerationc. The staff further concludes that i it is unlikely that any other design changes would be justified on the basis of person-rem exposure considerations, because the estimated CDFs would remain very low on an absolute scale.

4.0 THE ENVIRONMENTAL lMPACT OF THE PROPOSED ACTION lasuing an amendment to 10 CFR Part 52 certifying the AP600 design uld not titute a significant environmental lmpact. The amendment would merely codify re of the NRC's review and approval of the AP600 design as defined in the' F ER, dat ; Sip "ber 1998 (NUREG 1512). Further, because the amendment is a rul would have alternative uses. " e ar " resources involved that As described in Section 3 of this EA, the NRC reviewed emativ , to the de ce,hwi .

rulemaking and alternative design features related to p ting and mitigating ' accidents.

Consideration of alternatives under NEPA was necessa l(1) to'show that the de ~Tertification rule is the appropriate course of action, and (2) to ensure thidesign codified in the certification rule would not exclude any cost beneficial posig, ns ghanges related to the prevention and mitigation of severe accidents. The NRC concludes'that tnialtematives to design  !

certification did not provide for resolution of issues an dd the pro design certification I mismaking. ~* '

This design certification rulemaking is in k.

..[ f Q@

y Ing wi m 's intent in the  !

" Standardization and Severe Accident Ppicy Statepe'nt 10 CFR Part 52, to make future l plants safer than the current generation plants, toischieve rty risolution of licensing issues, and to enhance the safety benefits of standardizatiofi.# Throug its own independent analysis, the NRC also concludes that Westinghouse #a'dequateft " considered $n appropriate set of SAMDAs, and none were found to be cost-beneficial. Althdugh no deelgn changes resulted from reviewing the basis of the"PRA results.y 3.2 i of th[s EA p~ resents examples of these design features relate to sever a a ccident prevention and mitigation, but were not considered in the i SAMDA evaluation becauseMs were already part of the design. See FSER Section 19.1.6, "Use of PRA in the' Design . Process.

@;#/ w

'1 g 3r Finally, the certi>fiestion rule by itself spuld not authorize the siting, co AP600 desfgn'nuclea7 power plant. The issuance of a CP, ESP, COL, or OL for the AP600 design will re ufre a prospectivkapplicant to address the environmentalimpacts of construction and oper n at a specific sit i.y that time, the NRC will evaluate the environmental impacts and iss 'an EIS in accorda s with NEPA. The SAMDA analysis for the AP600, however, has been

, pleted as part of th ,EA and will not need to be reevaluated as part of an EIS related to siting, ion, or oper . n.

Y h

&%g:.Y 25

)

1 5.0 AGENCIES AND PERSONS CONSULTED, AND SOURCES USED The sources for this EA include Westinghouse's "AP600 Standard Safety Analysis Report," as amended, August 19,1996; and the NRC's

• Final Safety Evaluation Report Related to the  ;

Certification of the AP600 Standard Design" (NUREG-1512, Volumes 1,2 and 3), j September 1998. 4 l

The Director, Office of Nuclear Reactor Regulation (NRR), has determin d unde . ' National Environmental Policy Act of 1969, as amended, and the NRC's regulat _ in CFR Part 51, Subpart A, that this rule is not a major Federal action signif tly a quality of the human environment, and therefore, an EIS not required. . basis f determination, as documented in this final EA, is that the amendment to 1 ~Part 52 siting, construction, or operation of a facility using the A desi iiot authorize only codify the th[e .4 AP600 design in a rule. Therefore, the NRC staff did n issue t for coq bp Fede/ ral, State, and local agencies. However, the NRC's findinggno.si cant environ impact was published in the Federal Register on XXX XX.1999, toggther,with the proposed A design certification rule and there were no comments received rehted to'this EA. The NRC will evaluate the environmental impacts and issue an EIS as appropjr'afe' ce with NEPA as part of the application (s) for the siting, construction, or operatfoh of a ' ".& l The Director of NRR finds that Westinghouse's tion provides ent basis to conclude that there is reasonable assurance that an arnendment'to 10 CFR P 52 certifying the AP600 i design will not exclude a severe accident daIsign alte/6ative for~a facdtty referencing the certified  !

design that would have been cost beneficfal had it)ieen hansidered as part of the original design l certification application. The evaluation'of these Jes'ues unber NEPA is considered resolved for i the AP600 design. W 7 '

5

'g

$ J gM /

l j  ;

-w

Q f.; y g&ml%

w 1. nsb. t

[Mg  !

$[!(TM. I g~dgN d  %,f ngyg$'Y$ 4 m

4

l. [

A y}., ,

WA /,

$ h 26

_ _ _ _ _ _ _ . . _ .