ML20126J456

Severe Accident Research Program Plan Update

ML20126J456
Person / Time
Issue date:	12/31/1992
From:	NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To:
References
NUREG-1365, NUREG-1365-R01, NUREG-1365-R1, NUDOCS 9301060161
Download: ML20126J456 (89)

Text

- _ --

r NU REG-1365  :

Rev.1 Severe Accicent Researca Program P an Upc ate 9

4 s

i

).S. Nuclear Regulatory Commission '

Office of Nuclear Regulatory Research l

q >? N Okr

'o,

q. 5 l t

?

93010!>0161 921231 POR NUREG t 1365 R ppg ,

t l . -

L..,,-._,.L,.,..,,.._--....-.._,.-,-,_,,.--.-,-,,,,-,,,,,,....m..., 4.,m., . _ , , - . , , . . . . .- . _ _ , . . . . - . _ . . -

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

l l

AVAILA0lLITY NOTICE i

Availability of Referones Matonals Cited in NRC Publications F

Most documents cited in NRC pubhcations will be availabio from ono of the following ,

sourcos:

a

1. Tho NRC Public Document Rocrn, 2120 L Street, NW., Lower Lovel, Washington, DC 20555

2. The Suponntendent of Documents U.S. Govornment Printing Offico, P.O. 00* 37002, t Washington, DC 20013-7082 -

1

3. The National Technical information Servico. Springfield, VA 22161 Although the listing that follows represents the majority of documents cited in NRC publica-tions, it is not intended to be exhaustNo, i Referenced documents availablo for inspection and copying for a foo from the NRC Pubhc Document Room include NRC correspondence and internal NRC memoranda: NRC bulletins, t circulars, information notices, inspection and investigation noticos licensee event reports:

vendor reports and correspondence; Commission papers; and apphcant and licensee docu.

ments and correspondonce, The following documents in the NUREG serios are avahab10 for purchaso from the GPO Sales Program: formal NRC staff and contractor reports, NRC-sponsored conference procood-ings, international agroomont reports, grant publications, and NRC booklots and brochuros.

Also availablo are regulatory Quidos, NRC regulations in the Code of Federal Regulaflons, and Nuc!r,ar Regulatory Commission Issuances. ,

Documents available from the National Technical information Service includo NUREG4eries reports and technical reports prepared by other Fodorat agencios and reports prepared by the Atomic Energy Commission, forerunner agency to the Nuclear Rogu!atory Commission, Documents availabio from pubhc and special 16chnical libraries includo all open htoratuto items, such as books, Journal articles, and transactions. Federal Register notices, Federal and Stats lo0isfation, and congressional roports can usually be obtained from these libraries.

Documents such as theses, dissertations, foreign reports and translations, and non-NRC conferenco procoodings are available for purchase from the orDenization sponsoring the publication cited.

Single copies of NRC draft reports are available free, to the ex1ent of supp!y, upon written request to the Office of Administration, Distrloution and Mail Services Section,-U.S, Nuclear Regulatory Commission, Washington, DC 20555,

- Copios of industry codes and standards used in a substantive msnner in the NRC regulatory process are maintained at the NRC Library, 7920 Norfolk Avenue, Bethesda, Maryland, for use by the public. Codes and standards are usually copyrighted and may be purchased from the originating organization or, if they are American National Standards, from the _

American National Standards Institute,14 0 Broadway. New York, NY 10018.

gs pr- gii33 p**'D-Wgg-hs.av-^-gm*4-<. p w ar 'y- +r.,ury,-t<=--ywraa-te+- 4c ver w -gf emv v + P-w7ytg m e-gemeur e 'w-e -- e g-es te uv yfAWare qs - -- v mi= r,'e7#

> a

.i.

• hl: .

a .,-

'g_ 1

'b]

4;

~

. NUREG--1365 Rev. 1 .

1 3

I 8**=****" - h e w -- te-' .. . essa m '- ,

, -1

~

7 Severe Accident Re' search '

Program Plan Update, '

4:

~i i

Hj a

t, ,

1J

~

iU.S. NuclearLRegulatory Commission -

.>.. e t (Office of Nuclear Regulatory Research ' '

P i g i y

~

Yf"*?% ;r ' ?

n .

9. -

un r,.

, 9301060161 921231 PDR

.e NUREG ,

3365 R PDR <, a".

~%:

,:a y, - -.,

q. , .

t y

< ' ] iN' "

,. 1_ , , ,

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

,~ ,

AVAILABILITY NDTICE Availability of Reference Materials Cited in NRC Publical;ons Most documonts cited in NRC publications witi be evallable from one of the fo!!owing SourCos'

1. The NRC Public Document Room. 2120 L Stroot, NW., Lower Lovoi, Washington, DC 20555 2, The Superintendent of Documents, U.S. Governmont Printing Offico, P.O. Dox 37082.

Washington, DC 20013 7082

3. The National Technical information Serv!co, Springfiold, VA 22101 Although the Hsting that follows represents the rnejority of documents cited in NRC publica.

tions, it is not intended to be exhaustive.

Referenced documents aval!able for inspection and copying for a foo from the NRC Public Document Room includo NRC correspondence and internal NRC memoranda; NRC bulletins, circulars, information noticos, inspection and investigation noticos; licenson event reports; vondor reports and correspondence; Commission papors; and apphcant and licensoo docu+

monts and correspondence.

Tho following documents in the NUREG series are available for purchase from the GPO Sales Program: formal NRC staff and contractor reports, NRC-sponsored conferenco proceed-ings, international agrooment reports, grant publications, and NRC booklets and brochures.

Also available are regulatory guides, NRC regulations in the Code of federal flegulaflons, nnd Nuclear Rogulatory Commission Issuances.

DW Jnts availab!o from the National Technical Information Service include NUREG serios reports and technical reports prepared by other Federal agoncles and reports preparod by the Atomic Energy Commission, forerunnor agency to the Nuclear Regula*ory Commission.

Documents avaliablo from public and spot 'l technical librarios include all open literature items, such as books, journal articles, and ransactions. Federal Register noticos. Fodotal and Stato log lslation, and congresslonal reports can usually be obtained from those

!!Drarios.

Documents such as thoses, dissertations, foreign reports and translations, and non-NRC conference procoodings are available for purchase from the organization sponsoring the publication cited,

~

Single copios of NRC draft reports are available freo, to the extent of supply, upon written request to the Office of Administration, Distribution and Mail Services Section U.S; Nuctoar Regulatory Commission, Wat.hington, DC . 20555.

Copios of industry codos and standards used in a substantive manner in the NRC regulatory process are maintained at the NRC Library, 7920 Norfolk Avenue, Bethesda, Maryland, for use by the public. Codes and standards are usually copyrighted and may be purchased from the originating orgonization or, if they are American National Standards, from the American National Standards Instituto,1430 Droadway, New York, NY 10018, o -

-wa- J ,-

E L

NUltEG-1365 llev.1 4

Severe Accident Researc1 Program Plan Update Manuscript Completed: October 1992 Date l'oblished: December !!E)2 Division of Systems itesearch Office of Nuclear llegulatory itescarch U.S. Nuclear llegulatory Commissiou Washington, DC 20555

("<%

N....d.

• I

AllSTRACT in August 1989, the staff published NUltl!G-1365."Ite- during a severe accident. These mclude Mark I liner fail.

vised Severe Au;ident f(esearch I'rogram Plan." Since ure, severe accident scaling methmlology, source term 1989, significant progtess has been made in severe ucci- iwuck, cote concrete interactions, hydrogen transport dent research to warrant an update to NUltliG-1365. and combustion,TMI-2 Vessellnvestigation Project.and The staff has prepared this sal (P plan Update to: direct containment heating.'the ty>or t also desenbes the (1)ldentify those inues that have been closed or are near major planned activities under the sal (P over the next completion,(2) Desenhe the progtess in our understand- several years. These activities will focus on two ing of irnportant severe accident phenomena,(3) Define phenomenologicad issues (core melt progression, and the long-terrn research that is duccted at trnproving our iuel coolant interactions and debris coolability) that have understandmg of severe accident phenomena and devel-opmg improved methmls for assessing core melt progres- significant uncertainties that imEici our understanding sion, direct containment heating, and fuel. coolant inter- and ability to predict severe acendent phenomena and actions, and (4)1(cflect the growing emphasis in two ad- their effect on containment perlotmance,'the sal (P will -

ditonal areas -ndvanced light water reactors, and support also focus on severe accident code development, assess-for the assessentnt of criteria for containment perform. Inent and validation. As the staff completes the research ance during severe accidents, on sever e accident issues that : elate to cun ent generation reactors, continued research will focus on elforts to inde.

'the report desenbes recent major accomplishments in pendently evaluate the capability of new advanced light understanding the undeilying phenomena that can occur water reactor designs to withstand severe accidents.

' iii NUltliG-1365, llev. I W '

E

CONTENTS l' age Abstract . . ...... ...... .. ...... ........................................................... hi lixecutive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........ .. ................................. 11- 1 1 Severe Accident flescarch Plan . . . . . . . . . . . . . . . . ................... ....................... I 1.1 llackground........................................................................ I 1.2 Discussion............................................................................ I 1.3 liene fits of Additional R esearch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1,4 Criteria f or Termination of it escarch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.5 Organization of the Iteport . . . . . . . . . . . 4 . . . . . ... ......................................... 3 2 It escarch Plan . . . . . . . . . . . . . . . . . . . . . .. . ........ ...................................... 5 2.1 Dit ect Con tain m e n t l l ea ting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.1.1 ResearchNeeds................................................................. 6 2.1.2 Direct Contait, ment Ileating-Current itescarch Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

• 2.13 A n t ici pa t ed R e s ul t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.2 Core Melt Progression llesearch . . . . . . . . . . . . . . . . . . ....... .............................. 14 2.2.1 R esear ch Needs . . . . . . . . . . . . . . . . . . . . . . . . . . .. ....... ...................... 14 2.2.2 Cu r r en t R esearc h Pmgnim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

2. 2 3 A n t ici pa t ed R es u l t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.3 1 uel. coolant Interactions and Debris Coolability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 23.1 Research Needs . . . . . . . . . . . . .... ... .... ....... ............................... 26 2 3.2 Cu rre n t R esea rch Prog ra m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 23 3 A n t ici pa t ed it es ul t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - 30 2.4 Severc Accident Codes . . . . . . . . . . . . ........................................... ......... 30 2.4.1 SCDAP/RF.I .AP5 . . . . . . . . ... ....... ......................................... 31 2 .4 . 2. CO NTA I N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 2.4 3 M li l .C O R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 2.4.4 COMMIX........................................................................ 40 2.4 .5 V I C I O R I A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 2.4.6 Integrated Fuel Coolant int eraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 2.5 IlW R Mark 1 Containment I iner Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 2.5.1 Cu rr e nt R esea rch Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

2. 5. 2 F u t u r e Pl a n s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 2.6 I lyd rogen Combu stion a nd Resea rch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 2.6.1 Cu rrent R esearch Progra m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 2.7 SourceTerm.......................................................................... 45 2.7.1 Status......... .................................................................. 45

2. 7.2 P! I l iH U S - FP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 2.73 Oth er R esearch Activit ies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 v NURIIO-1365. Rev.1 w n +- c- , - , - , - ,

CONTENTS (continued)

I'aye 3 Research Plan for Advanced 1.ight Water Reactors .... . . ............... ... . . 49 3.1 Introduction . . . . . . . . . ... . . .. . . . . . .. . . .... 49 3.2 In. Vessel Severe Accident Phenomena . . . . . .. . .. . . .. . 49 3.2.1 Core hielt Progression and Reactor Vessel lower Head Failure .. . . ... 49 3.2.2 Fission Product Transport and Ilehavior . . . . . . . .. . 50 3.3 lix Vessel Severe Accident Phenomena . . . . .. . .... ... . . ... ... .. . 50 3.3.1 Research Approach . .. ..... .. . . .. . .. ...... .... .. . 50 3.3.2 Discussion . ... . . . . .. ... . . .. . .. . ... . . 50 3.4 Summary . .. . . ... . . . . . . .. .. 52 APPENDICES A Severe Accident issues, Status, and Progress to Date . . . . ... .. . . ... . ...... A-1 Il Clasure Of Severe Accident issues . .. . . . .. . . ....... .... 11 - 1 FIGURES 2.1.1 lintrainment Versus Pressure Ratio For Steam lilowdown. . .. ...... .... . .. . . 7 2.1.2 initial hiaterial Conditions For Direct Containment Heating. . . . . . . . ...... .. . 9 2.2.1 Core hielt Progression Sequence Showing Illockage or Drainage Paths. . . . .. . . . . .. 15 2.2.2 Axial Temperature Profiles for the Control lilade, Channel Box and Fuel Rods, Indicating Maximum lateral Temperature Variations . . . . .. . ... . . .. 18 2.2.3 XR1 and XR2 Test llundle Cross Sections. . . .. . . . . ....... . . , 19 -

2.2.4 hielt Drainage Pathways Through the Fuel Canisters Which Ilypass the lower Core Plate. .. 20 2.2.5 Test Ilundle Cross-Section for the NRU IlWR Test . . ..., . .. .. . .. . . . 22 2.2.6 h1P-2 Test Section Showing hiajor Components and Thermocouple Incations. . . . . . , . . 24 2Al NRC Sesere Accident Codes . . .. . . . . ...... . . . ..... . .. . .. ..... . 32 2.4.2 CONTAIN Code Validati m and Assessment . . . . .. .. . . . . 35

- 2.4.3 Ml!LCOR Code Validation and Assessment . . .. . . . . . . . . 38 A.2.1 TMI-2 Cote End State Configuration . . . . . . . . . . . A-5 TABLES 2.2.1 Test Matrix for the Ex-Reactor lixperiments un Metallic Melt Relocation and Blockage Formation

. Under llWR Dry Core Accident Conditions . . .. . ... . . . . . . .. . 21 A.2.1 Sources of Current Integral Experimental Information on Melt Progression . . A-4 A.2.2 Core Melt Progression: Status of Current Understanding. , . . . A-6 NUR EG-1365. Rev. I vi

EXECUTIVE

SUMMARY

in May 1988, the staff prepared an " Integration Plan for The focus of the SARP remains on rm iry research to Closure of Severe Accident issues" (SECY-88-147) for address technical issues of conce; 1mg contain-the Conmiission.Manyof theelementsof thisintegration rnent performance arid release of fis poducts in the plan require a basic ur.derstanding of severe accident event of containment failure.

phenomena. In fact, one of the major elements of the plan is a Severe Accident Research Program (SARP). 'the pnority for severe accident research depends on 1 ollowing the issuance of SliCY-88-147, the NRC staff, w hether there is a high likehhood that adJitional research with input from DOli laboratories and consultants from will result in a significant reduction in uncertainty and on universities and industry, identified and prioritiicd sig. the knowledge necessary for the conclusions to t)e widely nificant technical issues in order to focus the research accepted in the technical community. For some issues it efforts needed to close severe accident issues. may not be necessary to reduce uncertainties further, since otner issues may dominate overall uncertainty. For in August 1989, the staff published NURliG-1365,"Re. these and other issues for which it may not be practical to -

vised Sew;re Accident Research Program Plan

• This attempt to reduce uncertainties further, the staff recog-plan was organized in two parts. One was oriented toward nizes that some regulatory decisions or conclusions will the short-term resolution of issues related to early con. have to be made with full awareness of existing uncertain-tainment failure (i.e., direct containment heating, ll%R ties.

Mark I liner melt-through) and methodologies to evalu-ate and analpe these phenomena (i.e., seiding analysis, Table I provides a brief description of the program, severe accident codes). The second part of the S ARP Plan st tus, and major ..Wones for each of Il major severe was oriented toward providing long term confirmatory accident issues i

• s:w SARP. One of the goals of the information to support the assessments of a broad spec- S AR P is to com[Ete all the major severe accident experi-trum of severe accident phenomenology associated with inental programs v,hin the next 2 to 3 years. Assuming a the major elements of SliCY-88-147. relatively constant level of funding for the S ARP, closure of all severe accident issues is anticipated in 4 years.

Tne overall near-term goals of the plan were to provide the technical bases for assessing containment perform- Major Accomplishments ance over the range of risk-sipntficant core melt events and to develop the capability to evaluate the efficacy of Considerable progress has been made in recent years in genericcontainment performance criteria.The long-term understanding the underlying physical and chemical phe-goals of the plan were to provide an improved under- nornena that can occur in a severe accident. The staff standing of the range of phenomena expected during developed programs consistent with the Integration Plan _

severe acenients and to develop improved methods for for Closure of Severe Accident issues (SECY-88-147)to _

assessing fission product behavior and release in the obtain key data and information needed to m 16e an in-event of containment failure during severe accidents. formed decision on issues and phenomenological uncer-Since 1989, significant progress has been made in severe tainties associated with accident sequences that could accident research to warrant an update to NU REG-1365. potentially lead to early containment failure. This infor-The staff has prepared this SARP Plan Update to: mation is essential for assi'ssing potential safety improve-ments and for making decisions on whether or not par-

1. Identify those issues that have been closed or are ticular improvements are warranted. As pointed out in near completion, the Commission's Severe Accident Policy Statement, such decisions should be based on a combination of engi-

2. Describe the progress in our understandmg of im. neering judgment (i.e., a deterministic method of setting portant severe accident phenomena, and assessing safety margins) and the application of probabilistic risk assessment techniques to evaluate the

3. Define the long-term research that is directed at likelihooi l the occurrence of low probability events.

improving our understanding of severe accident phenomena and developing improved methods for A summary of the recent major accomplishments and assessing core melt progression, direct containment future ciforts under the SARI is given below.

heating, and fuel-coolant interacticns, and

1. Mark I Liner Failure

4. Reflect the growing emphasis in two additional ar- Completion of the initi:d study of IlWR Mark I con-cas -advanced light water reactors, and support for tainment shell failure was documented in NUREG/

the assesmmt of criteria for containment perform- CR-5423, "The Probability of 1.iner Failure in a ance during svere accidents. Mark-l Containment" (July 1991). An extensive E-l NUREG-1365, Rev. I

i lixecutive Summary peer review indicated that the methodology em- dine) or til (hydrogen iodide). Once in the contain-played to resolve the Mark I liner issue was sound ment, Csl is expected to deposit onto surfaces and an 1 no major deficiencies or problems were identi- dissolve in water pools forming I- (iodide)in solu-ficJ that would invalidate the results. 'Ihe peer re- tion. Subsequently, iodine behavior within the con-view als) identified the need for confirmatory evalu- tainment depends on the time and pil of the water ation of a number of subjects.nat follow-on work is solution. If p1I ccmtrol is available and maintained, under way in fiscal year 1992. little of the dissolved iodine will be converted to ele-mental iodine.

2 Snere Accident Scalmg Methodolog One remaining area to complete the test series is re.

A severe accident scaling methodology (S ASM) was lated to severe accident situations where air in.

developed to guide the formulation of experimental gressed into the core either by natural circulation or programs and analytical methods. Documentation from the residual heat retrayal system will react ex-of the SASM and application of the methods to di- othermically with the cladding producing high tem-rect containment heating (DCil) was addressed in perature in the fuel, large vaporization of ruthe-NURl!G/CR-5809, "An Integrated Structure and nium, tellurium, and molybdenum will occur.

Scaling Methodology for Severe Accidents Techni-cal issue Resolution" (Draft for comment, Novem- 4. Corr-Conc.cte Interactions ber 1991). Work on this issue is now complete. Ap- .Ihe NRC has conducted an extensive program of plication of the results of SASM to scalmg and analytic and experiment:0 research to obtain an im-operation of experimental facilities and modding proved understanding of core-concrete interactions, deselopment for teactor analysis are pursued under in FY91, an interim version (CORCON MOD 3)of the direct containment heating (DCI1) research pro- the core-concrete interaction code, CORCON, was gram described in Section 2.1.

released. The final version of CORCON MOD 3 is expected to be released by fall 1992. In addition to

-] J Source Term Issues impmved thermal. hydraulic modeling, CORCON MOD 3 incorporates the V AN ES A model of acmsol ne NRC has sponsored numerous experimental and analytical research projects on fission pmduct generation and mdionuclide release during core-concrete mteractions. Also, the NRC's nperimen.

release and transport. Early experiments and ana, lytical work tended to focus on release fmm fuct ma. tal program on core-concrete interactions has been terial under severe accident conditions. later, ex. completed. The remammg work on this issue is to validate CORCON MOD 3 against the experimental perimental data were obtained on the behavior of ,

acmsols in reactor coolant systems and contain- data Thts validation is currently underway, ment. These data were used to develop acmsol i de 'Treport and Combustion deposition and transmodels to analyze fission prod-uct behavior in the reactor coolant system (RCS)and Computer codes on hydrogen transport and com-containment. Currently, fully integrated mo&l bustion have been developed to evaluate the statie (VICI'ORIA)are in the pmcess of being completed or dynamic pressure loads from hydmgen combus-to analyze in vessel release from fuel and retention tion and detonation in containment. Results of these in RCS.The CONTAIN code is being developed to evaluations should enable the staff to make regula-analyze for the ex vessel source term, including the tory decisions to assess the potential threat to con-transport of fission products, condensation of va- tainment integrity. Since combustion pmcesses are pors, agglomeration and setthng of acrosols, and complex, many aspects are still not well understood, chemical reactions in the containment. l'or example, for a premixture of hydrogen air-steam in containment at elevated temperatures, but Technical evaluations to support the revision to cur- below the auto-ignition temperature, flame accel-rert methods for specifying the chemical form of the cration and high speed combustion may lead to a iodine entering the containment following a severe transition to detonation. Research on some of these accident (DD-14844," Calculation of Distance Fac- phenomena is continumg. Although the current re-tors for Power and Test Reactor Sites," March 1962) search is intended to reduce uncertainties, reduced -

were completed. The Oak Ridge National labora- uncertainties are not required in some cases to make tory performed analyses using a chemical kinetic regulatory decisions. A joint agreement was reached model to determine the equilibrium distribution of for a cooperative program with the Ministry of Inter-the iodine, cesium, hydrogen, and steam species en- national Trade and Industry of Japan and the tering the containment.The results indicate that io- Nuclear Power Engineering Center. Under this pm-dine entering the containment is at least 95% Csl gram, high-temperature, high speed hydrogen com-(Cesium lodide) with the remainder l(elemental io- bustion research will be conducted for next 4 years.

NUREO-1365, Rev.1 11- 2

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

f I!xecutive Summary

6. TM/-2 Vessellmntigation Project standing and ability to predict severe accident phenom

. ena and their effect on containment performance. Severe i The objectives of the , nil-2 vessel investigation accidents codes will continue to be developed to reflect  !'

program are to mvestigate the condition and proper-the current understanding of severe accident phenomena -

4 ties of matenal extracted from the lower head of the and will be validated against experimental data. Decisions TMI-2 reactor pressure vessel, to determine the ex-

' on when code development is completed require a bal.

tent of damage to the lower head, and to determme ance between the level of precision needed and the level the marg,m of structural integrity that remained m of uncertainty or variability in models of severe accident

~

the pressure vessel. Significant progress has been phenomena that are acceptable for regulatory decision made on these overall objectives. lixaminations of making.

the vessel samples from the TMI-2 lower head indi-mted that a small region of the lower head (approxi- 1. Core Melt Progression mately 2 feet in diameter) experienced inner surface temperatures of about 1350*K which exceeded the In-vessel core melt progression desen,bes the state transition temperature of the steel. 'the examina, of an LWR reactor core from core uncovery up to tions also indicated that the temperature 2 inches reactor vessel melt-through in unrecovered acci.

into the wall was about 100*K lower than the inner dents, or through temperature stabilization in acci-surface temperature. (the lower head thickness was dents recovered by core reflooding. Melt progres-5 inches.) sion provides the initial conditions for assessing kiads that may threaten the integrity of the reactor

7. Direct Containment Heating containment.

Since the publication of NUREG-1365 consider. A great deal ofinformation has been obtained on the-able research on direct containment heating has processes involved in the early phase of rnell pro-been undertaken to provide new insights and an im- gression that extends through core degradation and proved data base to answer the questions, What is metallic (but not eeramic) material melting and reh>-

the nature of the DCII threat,and what mechanisms cation. This information has come from integral and configurations exist ex-vessel that will mitigate tests in the following test facilities: (1) the Power or climinate it? In order to quantify the pressure and 13 urst Facility (PilF);(2) the Annular Core Research temperature kiads generated by DCil, phenomena Reactor (ACRR); (3) the Canadian National Reac-and processes that mitigate or augment DCilloads tor Universal (NRU); (4) the Japanese Nuclear to the containment must be considered. Safety Research Reactor (NSRR): (5) the French Phebus test reactor;-(6) the loss of Fluid Test To assist the NRC and its contractors in developing (LOFO facility; and (7) the German CORA ex-

, an experimental program and interpreting and ana- reactor fuel damage test facility. Laboratory sepa-lyzing test results, the staff convened a peer review rate effects experiments have also contributed infor-

, group to evaluate the NRC's program to resolve the mation on significant phenomena. .Most of the DCII issue.The peer reviewers meet regularly to as- available information on late phase melt progression '

sess progress and ensure that the research program has come from the post. accident examination of the objectives are being met. Three Mile Island, Unit 2 (IMI-2) reactor.

Integral testing was initiated in fiscal year 1991 to in. A comprehensive draft, "Research Plan for Melt vestigate the containment hiadings resulting from . . Progression issue Resolution," was prepared in fis-DCli.The experimental program will explore inte. calyear 1991 to address the remaining core melt pro :

gral DCll phenomena at different scales for repre. gression issues. The plan was subjected tc "xpert -

sentative reactordesigns. Separate-effects testing to peer review, and is currently being revised / N plan confirm the validity of the assumptions employed in and comments of the reviewers were used as a basis the scaling analysis was initiated in PY92, Details of to develop the plan discussed in Section 2.2 of this the testing are provided in Section 2.1. update,

, Reactor pressure vessel lower head failure maps-Major Ongoing Activities have been developed for local penetration failure and for local and global creep rupture failure as a Over the next several years, the major share of NRC's function of the characteristics of the lower head de--

severe accident research program will focus on two bris and of the lower head structure.The analytical phenomonological issues and on code development, vali- basis for these maps is supported by the results of dation and assesstuent.These issues, core melt progres- the metallurgical analyses of the samples in the sion, and fuel-coolant interactions and debris coolability, TMI-2 lower head program. A draft report,-

j_ have significant uncertainties that impact our under- NURIiG/CR-5642 " Light Water Reactor lower E-3 NURiiG-1365, Rev. I1

__ ~ _ ,-- _ ___

_ ,_ _ _ _ _ _ .. _ _ _ _ . _ _ _ _ __ __ ___4 . - _

lhecutive Summary.

Ilead l'ailure Analyses," on this analysis was issued FY92; and the CONTAIN and the VICI'ORI A code for comments in March 1992 and has undergone will be undertaken in the second quarter of FY93.

peer review 'Ihe final report will be issued in the 'lhe results of the peer reviews will focus on critical later part of 1992. needs for completing code development.

'" C""'""' '"'"i""' ""d "'b"' ' "'" bib

• Advanced Light Water Reactors There are three specific topics under this issue: fuel- Research programs have been initiated to independently coolant interaction (FCI) energetics, quenching in evaluate new ALWR (AP600 and SilWR) features, in water pools, and addmg water to a degraded core.

particular the adequacy of these features to withstand llecause of the large vanability of scenanos and pa- Evere accidents. These efforts will provide the technical rameters that impact FCis and debris cootability, t,he basis for NRC's support of ALWR plant design certifica-task of closing this issue is difficult. the plan m- tion

• volves fundamental, long-range elements, as well as assessments applying these fundamentals to specific  ;

Long-Terni Phtii reactor designs and accident scenarios. Over the last 10 years, the emphasis has gradually shifted from

" energetics" aspects to "coolability" aspects reic. Even though we anticipate that closure of all major severe vant to accident management. Ilowever, the need accident issues will be accomplished in 4 years, additional for assessing containment integrity for advanced work will continue on severe accident research, although light water reactors (ALWRs) indicates that a bal. at a reduced level of effort. Residual issues will still need to be addressed, in addition, results of severe accident )

anced approach with continued research on ener.

getic aspects is still warnmted. These fundamental experiments and research that are being conducted i energetics aspects are being pursued at the Univer. world wide will be used to continue to update and validate sity of California at Santa llarbara and at the Univer. the severe accident codes. Severe accident research on -

sity of Wisconsin. Results of this research are also ALW Rs will continue as designs for these plants continue expected to contribute to the understanding of the to evolve, It is likely that additional severe accident issues coolability experiments at FARO facility in Italy that may arise in the future.'Iherefore, the NRC will continue NRC is sponsoring. to support severe accident research, at a somewhat re-duced level, in order to improve out technical under.

Two experimentai research programs addressing cx, standing and maintain the level of expertise needed to -

vessel debris coolability were initiated during FY91. address future issues in this area.

The Advanced Containment Experiment (ACE) program, conducted at Argonne National labora' Criteria for Termination of Research -

tory (AN L), is an internationally spcmsored program with NRC participation. One phase of this program, The degree to which a severe accident technical issue Melt Attack and Debris Coolability lixperiment must be resolved depends on the needs of the related (MACE), deals with melt coolability issues. So far, regulatory decisions and on the necessary knowledge for three tests, including a scoping test, have been per- the conclusions to be widely accepted in the technical formed under the MACE program. While the community. Ideally, an issue is considered closed when M ACE tests involve prototypic debris composition, the NRC can pronounce that sufficient experimental and a sepuate NRC-sponsored program WinCOR, analytical research has been completed to allow, for the conducted at Sandia National 1.aboratories (SNL), purpose of IPEs (individual plant examinations), accident supplements the MACE program and investigates management studies, or containment performance evalu-the coolability issue of metallic and oxidic core de- ations, for the prediction of reactor plant response. In bris undcr heated side wall conditions to determine addition, uncertaintics have been reduced sufficiently the limits of coolability.Two tests were conducted in that regulatory decisions are either insensitive to further FY91 and in FY92 under the WETCOR program. reductions in uncertainty or that the residuallevel of risk is considered low enough that further work is not justifi-3 Severe Accident Codes able.

The principal integrated computer codes that sup- In practice, deciding when work is comnkled requires a port the evaluations of nucicar plant responses to subjective assessment of the potential beacfits of further postulated severe accident scenarios are being inde- research. Although these judgments are not easy, two

endently reviewed for their adequacy to meet NRC different approaches are possible. 'Ihc first approach is objectives. The first such comprehensive peer re- based on developing the analytical canability for predict-view was completed in FY91 for the MELCOR ing the complete evolution of severe accidents (under code; SCDAP/RiiLAPS peer reviewr, started in given initial and boundary conditions)in some reasonable NUREG-1365, Rev.1 E-4

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

11xecutive Summary level of detail. 'lhe other is focused on particular proc- possible after the main issues have been resolved, and a esses, relevant to specific containment integrity issues, research effort that is driven by this approach tends to be assuming that these processes can be adequately charac- rather inefficient in addressing the issues themselves.

terized within a rather general assessment of the overall sequence. Accordingly, the emphasis in the first approach The second approach provides for self-correcting focus-is on computer code development at the system level (i.e., ing on specific technical issues (such that they are well4 MEl.COR. CONTAIN rSCDAP/REIAP) and associ- posed for the research program), and it provides for grad-ated code validation or verification activities. In the sec- ual synergism and eventual convergence. This ond approach, the emphasis is on addressing specific convergence constitutes closure of an issue and thereby physical pmcesses and through them identify specific provides the basis for terminating further research. We safety concerns, then focus the research to address the are expecting to use such an approach on all containment specific concern. An example of the second approach is integrity issues. ' Itis latter approach on issues relating to that taken for the assessment of Mark-l liner attack in containment integrity and will temper the approach by NURilG/CR-5423. cognizance of the issue significance m terms of its overall risk impact.

These two approaches are not, certainly, mutually exclu-sive; the two approaches will work well together. There is An important element in the issue resolution process is already a lot of momentum on the first approach; proto- the role of peer review. For all of its major programs, the typic testing has been used to validate computer models, SARP routinely incorporates the feedback from peer re-which in turn would be exercised over the range of condi- view groups both in the area of experimentation and in tions and uncertainties associated with the data and acci- code development. The breadth of experience and exper. -

dent analysis. In fact, this approath may be even neces- tise brought to bear on the problem by the review groups -

sary for a widely available capability for severe accident assists both the staff and its contractors in achieving ch>-

assessments. Ilowever, it is also true that such use is only sure, E-5 NUREG-1365, Rev.1

~ ,

2 Tab'le 1 SARP Status and Milestones =;

s C . g _- cr j Expected , Reference 2 C,2 Issue  : Description -Status Completion Section - E- . - i s .a >

G . ex ,

y Major Program y. .

Appendix A.1

e - ' Mark I Liner , Develop probabilistic method to integrate SARP results Complete - l 9 Failure into conditional liner failure probability. Results indicate (NUREG/CR-5423) Q '

- that without water, liner failure is nearly certain; with '

water,. liner failure is physically unreasonable.

ResidualIssues: 2.6

1. Iiner failure criteria Complete

2. Melt superheat Complete 3.. Melt spreading analysis Complete

4. . Recalculation of the liner failure probability Anticipated December 1992 i l

' Major Program Direct Contair- - Develop analytical models for predicting entrainment of ment Heating core debris from reactor cavity and its interaction with

',., the containment atmosphere. .

&' 1. Integral-effect tests to simulate fundamental Ongoing December 1992 2.1

. synergetic effects of high-pressure rnelt ejection. .

'2 Separate-effect tests to confirm rate or detailed Ongoing February 1994 2.1 spatially dependent information employed in the i scaling analysis (e.g., corium jet disintegration, liquid ,

. film formation, entrainment, capture of corium in compartment). .

3.. Development and validation of system-level code. . Ongoing December 1992 2.1, 2.4 j

ResidualIssues:. _

Development of generalized assessment criteria for . . Anticipated

~

' February 1994 2.1 different cavity and compartment designs employed in L%'Rs, .

. Major Program ,

Scaling .- Develop severe accident scaling methodology to guide '

' Complete ,.

Appendix B3.

.the experimental program. .( NUREG/CR-5809)'

t l . Residualissuer .;

i.

. None-l

, , , , ..m - 4 . , _ . r_ . . ...

--,_-n..

'p 4

Table 1 SARP Status and Milestones (continued)

Expected Reference -

Issue ' Description .. Statta Completion Sectron Major Progam Source Tenn . Obtain better understanding of fission product release and Complete Appendix B.1 -

transport mechamsms in LWRs under severe accident

. conditions.

ResidualIssues: .

.1. Complete analysis of VI6 fission product release test Ongoing September 1992 2.7 ...

conducted in FY92 ' -

2. Complete VICTORIA code development : Ongoing - September 1992 L

. 3. - Validate source term codes with Phebus data Anticipated 1999

4. Complete the fission product release data for severe accident assocsated with air ingression condition

= In planning . September 1994 fl;

. Major Progart Core-Concrete . Obtain better understanding of the amounts of noncon- Canplete _ Appendix B2 -

Interaction densable gas generated from the interaction between p molten core debris and the concrete and radioactive aerosols.

a Residualissue' Compare code predic.tions to test results. Ongoing January 1094 Appendix B.2 MajorProgam Hydrogen Develop data base and computer codes to assess the con . Complete Appendix A3 Combustion - tainment performance due to hydrogen combustion. Data and transport were obtained on subsome and sonic combustion modes.

Data were also obtained on hydrogen transport and mixing. -

ResidualIssuer Data obtained to date is limited to hydrogen-air-steam _- Ongoing 1996- 2.6 2 mixtures at ambient or low temperature. Obtain better -

e understanding of the combustion phenomena of hydrogen- --:

2 air-steam mixtures at high temperature represen-tative of- -

'5- severe accident conditions. E 1,1. -

y

{

- p-a

~ -g Q' ..-=

i

.4

- b pa

'E e - er T h- 7 4 tr---- *r1 ++r - ew--A%'e

• 4 w 4 W r*-'

l l , !

7e2Eo og " 4 A 2 A A e a x c d n

e n n d

n ro i e p

e p

feed-e p p 2 2 2, A 2 RS A A 2 2 2 2 2 n

o 3 di t 0 ee tl 0

1 cp e e pm n 5

9 5 6 3 9

xo u 0 W W 9 EC J 1 I I 1

) e e e d t g t g g t g g e le a le n e n u s po t p i o

m  ;

p 6 i n u mg g e g

o g

o g

t m m i

t n

a on o n n o n n CO t

o S C O O C OO

(

c s

e n

t o

s -

e l .

r e m - r i

l a :o h o e . k ce o d M r a m t e

d nct eae c.

ent oc d l bc oro o rda e c f o

sem d!a f .

n t e ar n n r s e s

m a a t s m2 m i

emo nr ci r

.e l

t e omr ha g

ush s e l

e l edw u gt 4o f - r m e t l e oI e e eif ra er .nm na mf eat cr a

ssMt hmto r f i

f t

n cs o mtnt s t

s e i n

S eiTd o o oho c o en l t ec cgoi li t l d e P s ned l t aa t ee t a t ide a gR u l c l

t t r R ehn ed emca r A

2

- pt a I of ,

r ,iWoae n

s h rbt u mt l m ele e n

l sc+

t c u a Mproe S a g s s esLtd- a e n. f e l ohe i se ere 1

Tene d da: a as f t l . o om n o stt r m rer l

e

. msa em ole y,i cni oact oi c pel cn e r

e l s e b

T a h a h 'a t

nr d lt rt eoal e r t ao il t l aa ii mn i m t

e .

dt n aec t s n n miw o ef mf ts es v erl er mt ar e eir oto e

oa r

( )e e se oc ney srh r

om cef r teh ierm et

e imafo rn fdl n nut vngo f a u eet n nee r u na l

r s .rdoel n pme7 s

n oh txe .

iaot e oe e c o v do;c h aa ii x -

e ect eru it rbt deic r o gi r r fRe f eirf rt t

e nomb pi mhme npcase c oZ r sW n rm eisel n s crw ee sc n f e i

mhia l

omer-e o o ehloo cedosynetd s.

ct iBoidteos c n e ef t

e o f  :

s f t iecl l

eo t roge npei vf e m pt e nla t t aho sdindkencr in a e s fi n s o c sadnt gei u  :

diahn e e i a ai nd e t r i

t p ytcit mm s r v. inct inan d :e ph upncpo t gholo aso he es e r

isi cnl l r er ast ac rg I a yl yaa h e n, m r r netre CauCbcv da i

r nre r e r i

l pedo c P tx vt t a ol e c s

e v e EInxe ed m M e r dor oeaot c ivif it o vEni1n2 a _

b Mer &w p vgoAoMb y D i q . . fe no l: rosoen rc . i s

e M 12 RN MPppu1 2 3

R1 2

4 l

e n e

sn so ngn ei Vag t l t

eiri oo o 2i ct Me sdt s ya r

e I

- ts eej egHe u r n s

s Mvn or orede n I TI P CPaG Z :c9OLuS :? = - -72

.ll, llI!lll}lfll  !

I Table l'SARP Status and Milestones (continued) 1l Expected Reference - d

~ Issue Description Status Completion Section j Major Program Fuel-Coolant  ; Develop models or correlations and the technical bases for Appendix A 5 Interactions those models for predicting the broad range of resultant and Debris - phenomena ofinteractions of degraded core materials with j Coolability . coolant (including steam and hydrogen generation rates) for various reactor geometries. meltdown scenarios, and timing of coolant addition.

1. FCI energetics-relevant to in-vessel interactions in PWR.

a. Premixing experiments to determine the extent.of Ongoing September 1992 23 fuel and coolant intermmng that causes the region to become pressurized.

b. Fragmentation experiment to determine the rapid Ongoing September 1992 23 increase of fuel surface area causing vaporization of more coolant and irareasing local pressurization.

c. FCI yield expenment to determine the expansion Ongoing September 1993 23 work associated with the triggering and explosion

!? of premixture.

'c

2. , Fuel melt quenching experiment to determine the Ongoing January 1994 23 conditions under which FCI leads to coolable configuration as occurred at TMI-2.

3. Reflooding-relevant to in-vessel and ex-vessel issues

' related to operator adding water to achieve stable coolable configuration. .

a. : In. vessel reflooding issues--includes recriticality Complete ' .23

and hydrogen generation

b. _ Ex-vessel reflooding to mitigate core-core Ongoing March 1993 23 interactions -

ResidualIssuer The modEls or correlations devel oped under the FCI research Anticipated December 1993 '{

7 program will be validated against available experimental data. 1 g

Perform selective plant analyses.

p.

o J_

[.

15 E'

. x 3 Q l'

n+t, +- -

a s , ,a- w w ,

J , . . . .

.y

. Table 1 SARP S%tus and Milestones (continued)

{-

3

. C; Issue Description Status Expected Completion Reference Section b

F, e

e t c.

C; v1 '

p. . Major Program ' y

. :;e Severe Accident 1. ' Develop, validate, and document comptiter codes to Ongoing December 1994 .2.4 .g--

analyze severe' accident phenomena and issues for light 'q:

7 Codes

- water reactors. The code development program is an 3 ongoing iterative process; codes are dew;1oped and  ;

acsessed against selected experimental data; model

- improvements are identified; new models for new l phenomena are implemented and assessed. ,;

. ' 2. Extensive assessment and validation against experimental . Ongoing 1999 ,

results from domestic and international programs will -

continue to increase the confidence in the codes.

Major Program . '

Advanced Light 1. - Assess the methodologies used to evaluate containment . Ongoing September 1992 3 Water Reactors : . cooling concepts, including natural circulation cooling e . and external spray cooling of containment.

m . 2. Assess the methodologies used t(evaluate the effects Ongoing December 1992 ' 3 1 h of phenomena resulting from ALWR fuel design on severe acesdents. -

3. ; Apply severe accident codes (e.g., SCDAP/RELAP, Anticipated 8-12 months following receipt 3 -

1 MELCOR, CONTAIN) to the analysis of ALWRS. of vendor's design information ResidualIssues:

l None. ,

i l'. ,

-L t

n ,

m_.... _ ___m __ _ .m_._ _ . . _ _ ~ . _ , .__ , _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ . _ _. _ _..

l SEVERE ACCIDENT RESEARCil PLAN l.1 Background ments in a broad spectrum-coverage of the whole severe accident phenomenology.The specific issues identified in in May 1988, the staff presented to the Commission an the short-term part of the plan were: Mark I liner attack, Integration Plan for Closure of Severe Accident Issues direct containment heating, severe accident codes, and (SECY-88-147).The Integration Plan consists of six ma- scaling-the last two being primarily of methodological jor cir.ments: thrust.

1. Examination of existing plants for severe accident As stated in NUREG-1365, the overall goals of the re.

vulnerabilities (mdividual plant examinations). vised SARP plan are to provide the technological base for assessing containment performance over the range of

2. Development of generic containment performance risk significant core melt events, develop the capability to improvements (CPI) with respect to severe acci. evaluate the efficacy of generic containment performance dents to be implemented if necessary for each of the improvements, provide an improved understanding of the six containment types, range of phenomena exhibited by severe accidents, and reduce the uncertainties in the source term sufficiently to

3. Upgrading of staff and industry programs to improve enable the staff to make regulatory decisions on severe plant operations. accident issues. Since the issuance of NUREG-1365, the SARP has generated a large amount of data and insights -

4. A severe accident research program (SARP). into the progression of severe accidents. According to the revised S ARP plan, the research cfforts in the past 3 years

5. A program to define how and to what extent vulner- locused on the short-term issues listed above; however, abilities to severe accidents from external events significant progress was also made'on certain elements of need to be included in the severe accident policy the long-term plan, including in particular corium-con-implementation. crete interactions and the chemical form of radiciodine in severe accidents. In addition to the improved understand-

6. A program to ensure that licensees develop and ing of severe accidents achieved in the past 3 years, the implement severe accident management programs present plan (referred to as SARP plan update) was moti-at their plants. vated by the following considerations:

All these elements are mutually supportive and interre. 1, Identify those issues that have been closed or near lated, but the severe accident research program is the completion,-

common source of information, and hence of central im-portance. Elements 1,2, and 6 particularly depend on our 2. I)escribe the progress in the maturation a' nd evolu-understanding of severe accident phenomena and phe, tion of our understanding of important severe acci-nomenological sequences, and it is the purpose of this dent phenomena, program to provide this understanding.

3. Define the long-term research that is directed at Following the issuance of SECY-88-147, the staff estab. improving our understanding of severe accident -

lished several groups of experts to help identify and de. phenomena and developing improved methods for fine the status of SARP activities (item 4 above). Each assessmg core melt progression, direct contamment group was composed of contractor personnel from DOE heating, and fuel-coolant interactions, and -

laboratories, consultants from universitics and industry, ..

and NRC staff,These groups considered detailed techni- 4. Reflect the growing emphasis m two addit.ional ar-cat issues to define their status and the need for further e s-advanced light water reactors, and support for

' research, to identify and focus research necessary for the assessment of criteria for containment perform-sound regulatory decisions to be made within the frame- ance during severe accidents.

work of the integration plan, and to prioritize the research activities needed to close severe accident issues. The staff 1.2 DISCUSSION used the experts' input as a basis for preparing the revised SARP plan that was published in August 1989 as As in NUREG-1365, the focus of the SARP remains on i NUREG-1365," Revised Severe Accident Research Pro- research efforts to address technical issues of concern gram Plan." This plan was structured in two parts: one regarding the health and safety of the public. Technical ~

oriented to the short-term resolution of certain pressing issues regardmg containment performance are of particu-J.

issues, and the other oriented to the long-term confirma- lar regulatory significance. The two issues identified in

, tory support, and perhaps refinement, of current assess. NUREG-1365 that could potentially lead to early 1 NUREG-1365, Rev. I

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

1 Severe Accident flesearch plan containment failure are the Mark I liner attack and direct charged areas of debate. 'lhe impact of improved under-containment heating (DCll). At this time, the Mark i standing in a certain area of phenomenology on risk re.

issue is near resolution; a plan to bring closure to the duction is not always readily apparent. One example is

- DCil issue will be in place by the end of CY92. With core melt progression research, which is one of the most I respect to the long term research, it is important here to complicated technical areas in severe accident phenome-achieve a definition that will produce the sought-after nology. Core melt progression impacts hydrogen genera-understanding within a reasonable time schedule. At the Lion (timing and quantity) and core debns relocation into same time, it is imgurtant that this understanding is di- the lower plenum (timing, quar tity, rate. and composition rectly relevant to safety to adequately support elements 1, of debris). The latter aspect, in turn, impacts fuel-coolant 2, 5, and 6 of the Integration Plan. Carrying out the interaction (l'CI) during and after the relocation (ener-long-term research plan requires a lot of judgment and an getic, coolability), thermal loading on the lower head -

atmosphere of open communication among all parties (mode and timing of failure), and release of core debris involved. It also requires clear communication and a de- materials to the containment. liventually there are im.

liberate approach in planning, reviewing, and conducting portant implications for ccmtainment integrity (i.e., ener-research One purpose, then, of this plan is to address the getic FCI, DCII or Mark I liner attack). As NUllEG/

methodological and practical aspects of meeting these Cit-5030,"An Assessment of Steam lixplosion-induced requirements in future research efforts. Containment Failure" (February 1989), and NUltEG/

CII-5423 have shown, conservative treatment of uncer-Consistent with this philosophy, the staff has proceeded tainties is feasible to yield Imunding and acceptable re-with the implementation of the revised sal (P, making suits for the alpha-mode containment failure and Mark i extensive use of debate and deliberation of expert peer liner attack issues, respectively. For these to issues, the -

resiew to address a good number of issues. The expert main benefit of learning more about core melt progres-reviews include the technical program group on severe sion phenomena will be related primarily to improved accident scaling methodology and review panels for core understanding of the margin of conservatism in the analy. I melt progression research, for direct containment heating ses to resolve these issues, experiments, and for severe accident codes. In addition, the Mark-1 hner attack study (NUllEG/ Cit-5423) was From a containment integrity standpoint, learning more I subjected to extensive peer review and to smaller special. about core melt progression provides the input for con- l ized panels for follow-up activities. These activities have tainment h)ading for DCI1. If DCII were amenable to the proven very fruitful in sharpening the staff's judgment same approach as the alpha-mode containment failure concerning the practical aspects of implementing the re. and containment liner attack issues (see section 2.1), the vised S AllP and will be continued in the implementation value of core melt progression research would be m im-of this sal (P update. proved knuwledge in the area of accident management when core melt progression is considered in the presence of water as it would in a recovered scenario. The payoff 1.3 lletierits of' Additioritil Resesircli - would be in improved estimates of risk. even though the accident scenario is of exceedingly low probability and the

'Ihc phenomena under investigation and their combina- risk is already low, if we can demonstrate that accidents tion into severe accident scenarios are highly complex- can be terminated even at relatively advanced states of

, usually involving multiphase or multicomponent interac- their progression. 'the excellent safety record of nuclear tions, rapid transients, and multidimensional behavior. power plants leaves only such exceedingly low probability, Since full-scale experiments are not possible, the in- high consequence events to be of concern. Again, even depth understanding presents quite a formidable task On - though the accident sequence leading to vessel failure is the other hand, experience has shown that dominant oflow probability, lack of adequate understanding of core mechanisms and evaluation methodologies can be found melt progression leads us to use conservative assumptions to provide an understanding quite adequate for the in- to provide answers to a set of rather intangible questions tended purposes. The special challenge in planning and that we are of ten asked. For example, for ex-vessel debris conducting the severe accident research plan is to identify coolability, a core debris cootability criterion of 0.02 m2/

and use effective approaches, focusing quickly and study- MWt was proposed. This criterion is intended to be used ing in depth all the essential aspects of the behavior, in in sizing the reactor cavity for ALWits.The existing data fact, a continuing focusing and refocusing process is re- base for sustained melt concrete-coolant experiments quired, as new research results provide the basis for im- suggests that debris depth and the formation of crust on proved judgments as to where to expend future efforts. the surface of the melt may inhibit water ingression into the melt, which is necessary to produce rapid cooling. If Which severe accident phenomena are studied, how con. more realistic assumptions are used, the debris bed depth clusions about these phenomena are reached and de- would be substantially smaller than is currently investi-fended, and when research on specific phenomena and gated experimentally (75 100% of totalcore).'Ihecore issues should be stopped have traditionally been highly melt progression research can provide valuable NUltEG-1365, llev. I 2 i

,- - - . - ~ ~ ~ - - - , , , , , u, -.e+-- , --v ,,n , -en-

I Severe Accident Research Plan information on the amount of molten material that can given initial and boundary conditions)in some reasonable enter the containment at the time of vessel breach. level of detail.1he other is focused on particular proc-esses, relevant to specific containment integrity issues, As wu the case when NURIiG-1365 was developed, the assuming that these processes ctm be adequately charac-staff recognizes that for some issues it may not be practi. teri/cd within a rather general assessment of the overall cal to attempt to reduce uncertainties further, and some sequence. Accordingly, the emphasis in the first approach regulatory decision or conclusion will have to be made is on computer code development at the system level (i.e.,

with full awareness of existing uncertainties. I!xpert clici. MI!!.COR, CONTAIN, SCDAP/RilAP) and associ-tation (as done in NURiiG-1150) in the I evel 11 PRA ated code validation or verification activities. In the occ-serves only to reveal rather than to resolve issues since it and approach, the emphasis is on addressing specific deals with phenomenological issues at a high level. Quite physical pmcesses and through them identifying specific often experts rely on parametric codes in which the safety concerns, then focusing the research to address the author has programmed his opinion of what happens; t al specific concern. An example of the second approach is important phenomena may never be discovered.12urther. the one taken for the assessment of Mark.1 liner attack in more, the degree of judgment, and therefore the degree NURl!G/CR-5423. -

of confidence associated with it, depends on tbe particular issue and the nature of the phenomenology involved.The 'Ihese two approaches are certainly not mutually exclu.

priority and manner in which severe accident research is sive; the two approaches work well together. 'there is conducted depend on whether there is a high likelihood already a lot of momentum on the first approach; proto-that such research will result in a significant reduction in typic testing has been used to validate computer models, risk uncertainty and on the depth necessary for the con- which in turn would be exercised over the rang of condi-clusions to he widely accepted in the technical commu. tions and uncertainties associated with the data and acci-nity, dent analysis. In fact, this approach may be necessary for a widely available capability for severe accident assess.

This research plan is structured to examine the issues in ments. However, it is also true that such use is possible depth and develop the necessary methodologica: tools only after the main issues have been resolved, and a that can soundly address these issues. Therefore, one of research effort that is driven by this approach tends to be the goals of the S ARP is to complete all the major severe rather inefficient in addressing the issues themselves, accident experimental programs within the next 2 to 3 years. Results of severe accident experiments and re. The second appmach provides for self correcting focus-search that are being conducted world-wide will be used ing on specific technical issues (such that they are well to continue to update and vahdate the severe accident posed for the research program), and it provides for grad-codes. ual synergism and eventual convergence. This conver-gence constitutes closure of an issue and thereby provides ~

the basis for terminating further research We expect to l A Critcria for Termination of use such an approach on all containment integrity issues.

Researell An important element in the issue resolution process is

%e degree to which a severe accident technical issue m peenedewh@Ns major pmgrams, k must be resolved depends on the needs'of the related "" I ** ## ** E##"#'

regulatory decisions and on the necessary depth for the view groups both m the area of expenmentation and m .

conclusions to be widely accepted m the techmcal com- code development.The breadth of experience and _exper.

munity, ideally, an issue is considered closed when the tise brought to bear on the problem by the review groups NRC can pmnounce that sufficient experimental and assists both the staff and their contractors in achieving analytical research has been completed to allow, for the purpose of IPlis, accident management studies, or con-tainment performance evaluations, for the prediction of reactor plant response. In addition, uncertainties have 1.5 Organization of the Report been reduced sufficiently that regulatory decisions are Our intent is to make this research plan scrutable; thus, insensitive to further reductions in uncertainty. when pertinent technical details cannot be given by refer-ence, they are included in appendices at the end.

In practice, deciding when work is completed requires a subjective assessment of the potential benefits of further - Chapter 2 points out the main directions for future severe research. Although these judgments are not easy, two accident research, and Chapter 3 discusses planned re-different approaches are possible. The first approach is search efforts for ALWRs. Appendices A and B of this based on developing the analytical capability for predict. report discuss all the major phenomenological areas of ing the complete evolution of severe accidents (under severe accidents and progress to date on these areas.

3 NUREG-1365, Rev.1

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

1 Severe Accident Research Plan : _

'this report has benefited from reviews by the Advisory _- search Review Committee, and the Office of Nuclear _ ,

' Committee on Reactor Safeguards, Nuclear Safety Re. Reactor Regulation.

1 E

i ,

t I

i i

f

- NURIiG-1365i Rev. I' '4 m

4 , - -

r w 3 L-~ -

y ---m-y ,r y y & 'k }

1 I

2 RESEARCil PLAN IntroduCllOH system and anest of core damage before vessel breach). 'Ihe regression analyses would not reveal

'lhe programs desenbed in this chapter of the updated this sort of insight. ~

SARP are the high priority items that need to be ad.

dressed in the next few years, namely 2. Although the regression analysis had limitations, and care should be taken in interpreting the results,

1. Core melt progression, the regression analysis showed that the following

2. Direct containment heatmg, in-vessel issues are important to the uncertainty in NUREG-1150 results: the fraction of each radionu-

3. Fuel coolant interactions and debris coolability, and clide group in the core released to the vessel before

4. Severe accident codes vessel breach, pressure rise in the containment at vessel breach (implies that pressure of the reactor

'this chapter presents the research needs, our current coolant system (RCS) at the time of vessel breach, research pmgrams, and anticipated results for each of the the mode of vessel breach, and the fraction of the above issues. Research pmgrams on other residual issues core released at vessel breach are important), in, are also discussed in this chapter. Appendices A and 11 vessel hydrogen production, and alpha-mode fail-ure, present a brief summary ofIhe current state of knowledge on the core melt progression and fuel coolant interaction issues listed above as well as other severe accident issues. 3. The probability of failure of the Surry containment is

'lhe focus of this research is to provide validated codes to low. Ilowever, within the conditional containment assess the performance of nuclear power plants with se. failure probability that does exist, the contributions vere accidents nnd to gain a more quantitative under- .

from different modes of vessel breach are extremely standing of the overall probabilities of core damage and varied-the conditional probability of containment fission pmduct releases. NUREG-il50 and other PRAs failure at vessel breach for vessel failures at high have shown that events that lead to early containment pressure is several orders of magnitude higher than failure (such as several hours after severe core damage the conditional probability of containment failure at pmgression starts) have the greatest risk significance. As vessel breach for vessel failures at low pressure.This the accident progresses through its various stages, various indicates that in-vessel issues could be entically im-kinds of containment hiads are a consequence of the portant to containment failure probability at plants prevalent material interactions. The intensity of these that are more susceptible to overpressurization than h) ads depends on the quantities, composition, and tem. Surry, pelature of the materialinvohed and the overall configu-ration of contact. 4. The magnitude of source term releases at Surry are significantly impacted by the mode of vessel breach.

While the NUREG-1150 study did not produce quantita. The risk of prompt fatalities from a high-pressure tive measures of the risk importance of specific melt pro. melt ejection mode of vessel breach and an early gression phenomena, the results show that both the mag. containment failure is several orders of magnitude nitude and uncertainty of melt pmcesses are important. higher than the risk of prompt fatalities from vessel For example, the observation that the risk significance of breach at low pressures with an early containment DCH has decreased relative to earlier studies can in part failure. 'Ihis difference is due to the high rate of be attributed to an improved understanding of melt pro. acrosol generation Imm the molten debris during a gression that led to improved DCH testing procedures. high pressure melt ejection.This tisk at Surry islow, Further exploration of the NUREG-ll50 information but could indicate a high source term sensitivity to base together with a sensitivity study has produced quan. the mode of vessel breach at the other plants, titative informat.on on the risk importance of melt pro-gression phenomena.i 5. In vessel issues are especially important from an accident management point of view.The timing of Several of the conclusions from this study are: vessel breach is important for planning operator actions. Also, during tecovery stages, the predicted

1. In vessel issues directly influence the low contain. response of the molten core to water injection would ment failure probabilities and the low risk values be useful information for the operator to have, obtained for Suny (depressurization of the primary in summary, the evidence in NUREG-1150 indicates that

'txtter ins Fred liarper. SNI., to Brian Sheron. NRC, dated Janu- in vessel melt progression issues are important to risk and ary 24.1992. the uncertainty in risk.

5 NUREG-1365, Rev. I

2- Research Plan in general, core melt progression research provides im. 2. Debris to gas heat transfer--the embedded debris portant input for all containment failure issues, such as velocities, residence time the amount, composition, and temperature of the melt released to the containment and the mode of vessel fail. 3. Debris trapping-the capere of debris within sub-ure, In addition, since accident management is considered compartment volumes or on intervening structures to be a practical measure for mitigating the consegnence of severe accidents, an understanding of degraded core 4. Ilydrogen generation-resulting from oxidation of cooling and other phenomena related to fuel coolant in- metallic debris constituents teractions is needed to assess accident management strategies. Scriions 2.2 and 2.3 provide detailed NRC 5. Hydrogen combustion-resulting from reaction of research programs to address these issues. "'.iy preextsting hydrogen m the contatnment along with hydrogen generated during the high pressure -

At the time the reYised SARp was issued in 1989, the meh eMon issues of high-pressure me!! ejection and DCH had been 6. Water vaporization-as a consequence of debris-the subject of considerable study over the preceding sev-water interactions in the cavity or other containment eral years.The results of these preymus studies increased regions.

' concerns about these issues, and it did not then appear likely that furtha research could lead to their resolution. ideally, modeling should also address parameters, proc.

Recent understanding of the core melt progression, the ess times, and spatial dependencies as well as their rela. ,

completion of the lower head failure analysis program, tionship to, in the SASM parlance, control parameters ~l the completion of the severe accident scaling methodol. (e.g., RCS pressure). l ogy, and restructuring of the DCll experimental program

. are all contributing significantly to the near-term closure While not strictly related to DCII, the precursor proc-.

of this issue. Section 2.1 provides the detatled NRC re- esses that influence melt ejection from the reactor vessel, search program to address the direct containment heating gas blow through, and hole ablation are also necessary issue. components of the overall calculation.

Section 2.4 discusses the codes being developed and Modeling of DCH must also reflect the effects of the maintained by NRC.The MELCOR code has become the . diversity of plant designs insofar as plant-specific features NRC-supported systems analyses code that replaced the influence the phenomena. There are some 18_ different source term code package. A few detailed mechanistic cavity and lower containment subcompartment designs in codes will also be maintained that are applicable to spc. use in the U.S. commercial pWR nuclear power plants, cific severe accident phenomena.

Past research has indicated that while the cavity geometry Sections 2.5. 2.6, and 2.7 discuss the BWR Mark I contain. influences the extent of dispersal from that region at ment liner, hydrogen combustion, and source term issues, elevated RCS pressure, differences m design configura7 respectively. tion (at least f.;r the designs expenmentally investigated) .

have less impact on the entramment fraction. As part of the SASM technical program group efforts to develop a 2.1 Direct Containment Heating SC"li".g methodology for DCH testing, a correlation for -

entramment fraction (,.c.,t the fraction of dehns dispersed from the reactor cavity) was developed and assessed 2.1.1 Research Needs against a range of experimental data at different scales

. (1/42,1/25,1/10 linear scales) for different designs (Surry,

.rhe ultimate needs from research on DCH are analytical S zewell, Watts Bar, Zion). This work suggests that the models or correlations and the techmcal basis for those entrainment fraction can be correlated against a number models for ;)redicting t he entra!nment of core debris from of parameters with a notable dependence on the initial the reactor cavity and its interaction with the containment Euler number, and thus, the pressure ratio defined by the .

atmosphere and blowdown gas during a high-pressure nitial RPV pressure / cavity pressure. Figure 2.1.1, taken melt ejection' from NURl!G-5809, illustrates the dependence, as pre-dicted by the correlation 'of the entrainment fraction on -

Thus, in addition to predicting the dispersal of debris the initial RpV to cavity pressure ratio for two different' from the reactor cavity, DCH modeling must also be designs, Zion and Surry, capable of evaluating or suitably accounting for:

'!he general conclusion that debris dispersal from the

1. Debris fragmentation-the surface area for heat- cavity is nearly complete at elevated RCS pressures has -

transfer and reactions, translated into a particle di- been qualitatively demonstrated in many of the past DCH ameter ' experimental prograrns, regardless of the9 direct NUREG-1365, Rev. I 6'

2 Research Plan i

L l

L Entroinment Froction, y o o o -

P "i

~ o

. ......# . . . .....i . ....it o . . . . . . . .

• o n -

RV M o g .

3 -

l -

2 .

9 t.n -

5' P g

o - il T. g l.

N .

o .

n -

Ii -

y P

Im I l

O S.

U.I o E o pi.lQ -

-  ?

n,j l o. -

- ~

u ul lu -

C p -

o ollP .-

o .

o 1.g +

+ *

. .Bl 1.3 - .

3

,1 IR .

w ngiN .,

2 -

C m- mlo lP- -

- S b! m E t:fc8glo -

--S.E 'Q SQ -@d g; .

-PP P PP --

- .oo --

em I 1 on o e i 9oi o, io I g

'o Entroinment Fraction Figure 2.1.1 Entrainment Versus Pressure Ratio for Steam lltowdown.

7 NUREG-1365. Rev.1

2 l(esearch Plan scalabihty of indindual tests. While this preliminary con- generated is consumed at the same rate at which it is clusion needs to be confirmed by additional testing and released and mned into regions containing sufficient oxy-scahng analysis, the results of work to correlate entrain- gen, no hydrogen accumulates and the resulting combus-ment fraction culminating in the plant extrapolation tion mode is a diffusion flame, either a buoyant plume or shown in l'igure 2.1.1 indicate that research is needed to jet. Similarly, another requirement for this scenario address the other major processes, e.g., subcompartment would be the presence of an ignition source, hot debris trapping and water intenictions which may has c a signifi- particles in the case of an llPMii, or the release and cant effect on DCII hodinp Furthermore, correlation of mixing of hydrogen at temperatures above an autoigniti' n entrainment fraction data mdicates that testing need not temperature. If the hydrogen is released at a rate faster represent the full range of pmsible 1(CS pressures (up to than which it can be mixed with oxidant and burned or if the power-operated relief valve setpoint), since complete there is no mechanism for initiating combustion upon dispersal out of the canty is potentially achievable at initial release, the hydrogen may accumulate and the much low er pressures. Interestingly, the work to cort clate subsequent combustion, w hich can be presumed to occur entramment fractions also suggests that relatively simple after eventual mixing of gases to combustible levels, may models, or even quantitative empirical criteria, may be take the form of detonation. _

substituted in lieu of detailed mechanistic models.

In order to provide the answers to the research needs described above and resolve the DCilissue, the NitC has While the details of the cavity configuration, for the lim, ned range of variations considered, have been shown to developed a research program composed of three major haw ess effect on dehns dispersal at elevated pressures, elements, all of which are judged necessary to achieve it is intuitive and has long been argued that compartmen- closure: (1) integral testing, (2) separate effects testing, talization and structu.es downstream of the cavity would and (3) development and validation of a systems-level trap or deentrain debris from the flow stream exiting the code (CONTAIN).

cavity on its way to the bulk contamment. Since the free volume of the lower containment compartmentalized re. As described in Appendix 11.3, " Scaling," a necessary gion of the containment is a small fraction of the total precondition for the resumption of DCll testing was the containment volume (-15% in the case of Zion) and the development of a scaling rationale to guide the design and compartmentali/cd region immediately adjacent to the operation of planned tests. It was further determined that cavity exit is yet a smaller fraction of the total volume, it is sading evaluations would be based on the evaluation of apparent that trapping of debris in those regions would conditions for a specific accident sequence and for a spe-significantly decrease the ability to rapidly heat, by debris. cific plant design in order to assure a more scrutable and gas heat transfer, the bulk atmosphere. meaningful comparison of the scaling results with reactor design and analysis.

In addition to trapping or deentrainment of dispersed The starting point for this scalmg evaluation of appropri-debris, the other first-order mitigative effect on the high- ate initial conditions was the technical program gmup

~

pressure melt ejection (llPMl!) process is the potential (TPG) review of existing core melt progression analyses interaction of debris w:th water, eithcr in the cavity or in for a station blackout at the Surry plant in order to assess the containment large quantities of water present in the the amount of core debris present in the lower head at the containment during certain severe accident sequences time of reactor pressure vessel failure.'!hree synthesized can potentially quench or cool debris and thereby reduce sets of inittal conditions for DCIi were developed; a set of contamment pressuruation. conditions associated with a penetration failure of the bottom head and two sets of conditions (for both low While direct contamment heating research has pnnei. pressure and high pressure) associated with creep failure pally been concerned with the basic mechanisms by which of the bottom head. The quantities, compositions, and molte n debns could give up itslatent and sensible' heat to amounts of material involved are summarized in Fig-the atmosphere, a significant contribution to the pressuri. ure 2.1.2., taken from NURI!G/ Cit-5809. The evalu-zation of the containment arises from the potential oxida- ation of melt conditions for a station blackout at the Suny tion of the metallic constituents of the core debris and the plant yielded the following general insights, subsequent combustion of any hydrogen generated through that reaction. Assuming roughly 15% (by mass)

• About 40% of the core mass would be ejected in of the core debris is unreacted zirculoy cladding, the co- molten form.

rium thermal energy (sensible plus latent heat)is roughly equivalent to the chemical energy, or. the order of 1 MJ/

• From 30 to 70% of the molten material could be in a kg. linergv released from the combustion of the hydrogen metallic form. If failure of the crust or flow blockage produced by oxidation of that unoxidized cladding is also from structural weakness is considered more realis-equivalent to roughly 1 MJ/kg of melt. If the hydrogen tic than crust or flow blockage melting, about 30 to NUREG-1365, Rev. I 8

2 Research Plan STATION BLACK 0UT HIGH PRESSURE HIGH PRESSURE LOW PRESSURE FUEL CANDLING, FUEL CANDLING, FUEL CANDLING, CRUST FORMATION LIMITED BLOCKAGE LIMITED BLOCKAGE AND FAILURE OF GRID FAILURE OF GRID FAILURE PLATE PLATE MOLTEN DEBRIS SOLID / MOLTEN SOLID / MOLTEN IN BOTTOM DEBRIS IN BOTTOM DEBRIS IN BOTTOM OF OF 0F VESSEL VESSEL VESSEL PENETRATION (S) CREEP CREEP i FAILURE FAILURE FAILURE INITIAL CONDITIONS TIME OF FAILURE -3 HOURS -4 HOURS -10 HOURS EJECTED KGS ~40,000 -70,000 -80,000 MOLTEN KGS ~40,000 -40,000 -40,000 OXIDIZED ZR -40% 50% 70%

EJECTED METALLIC CONTENT 40%- 60%* 70%*

TEMPERATURE ~2500 K -2500 K -2500 K

• INCREASED METALLIC CONTENT.DUE TO MOLTEN STEEL Figure 2.1.2 Initial Material Conditions for Direct Containment Heating.

9 NUREG-1365, Rev. I

2 Research Plan

$0% of the molten mattnal will be in a metallic bris gas heat transfer in scaling may lead one to distort or form. (In both instances, the majority of the metallic even neglect principal mitigative mechanisms in an inte-comp (ment is steel; Zr is 10-25% of the molten gral t est. When those conflicts arose, scaling was driven by debris.) the objective of simulating the transport of debris to the bulk containment as opposed to direct simulation of pres-

• A significant amount of solid debris (over one-half of suri/ation. As a check on the possible distortion of phe-the molten mass)will be present in the lower head nomena, including concomitant atmosphere pressunta-for scenarios in which vessel failure is delayed until tion, counterpart integral tests using the same scaling creep fatture occurs. 'the majonty of the solid debris principles are run at different scale, may be retained in the vessel, at least in the time scale related tolIPME.The massof molten material 1:ollowmg the basic objective and guidelines for integral for a creep failure scenario was virtually identiced to testing outlined above, the simulated RCS and contain-that associated with a penetration failure. trent features and volumes were geometrically (i.e.,line.

arly) scaled, the RPV aspect ratio as well as the length To proceed with testing and validation of andytical mod-cls the staff has assumed initial conditions associated pale from the ItpV to the cavity was preserved. In keep- .__

mgw our objective to simulate fundamental or global with a station blackout sequence and a penetration failure behav@ior m a consistent fashion and as a consequence of of the reactor pressure vessel (RPV) bottom head. It is tising a simulant material (iron-alumina thermite with fully recognized that there is uncertainty surrounding the chronuum metal) rather than reactor materials, the melt precise imtial conditions for DCil; nonetheless, for pre- m u employed in the tests was scaled based on the equi-senhing scaled test conditions and conducting integral tests, we conclude that the best available information Wnum mrgvolume ratio. llecause of the lower den-indicates adopting this approach. Simulation of creep fail, q d subnt Me m sm, dng on an m@

volume basis results in a distortion of the amount of melt, ure, or for that matter multiple penetration failures, can volumetrically scaled, by approximately 48% Ihe melt he addressed by variation of the hole size used for testing.

mass employed in the 1/10 scale representation of Zion 2.1.2 Direct Containment lleating-Current Research Program .the staff has convened an mdependent peer review group As summarized in section 2.1.1, the staff has outlined a to evaluate the program to resolve the DCH issue and research program consisting of (1) integral testing, specifically assess the program of mtegral testing. With (2) separate effects testing, and (3) analytical model de. the concurrence of the peer review group on our baste velopment and validation. approach, the staff has proceeded to conduct integral tests at Sandia National laboratories (SNL)in the 1/10th The objective of integnd testing is to simulate the funda- linear scale Surtsey facility and at Argonne National mental synergetic effects of high pressure melt ejection laboratory (ANL)in the 1/40th linear scale COREXIT _

through experiments that include entrainment or facility. The integral experimental tests currently under sweepout of corium from the cavity, transport and trap- way at SNL and ANL ee considerably different from the ping of debris in the lower containment subcompart. tests proposed prior to the development of the SASM.

ments, oxidation of metallie constituents in the corium, Specifically, these testt are improved because (a) the in-combustion of hydrogen produced as a result of that oxi- itial and boundary coiditions for the tests are scaled for dation of metals, and vaporization or heat transfer to the specific accident scenarios for specific p' mts, (b)impor-existing water inventory in the reactor cavity and contain- tant scaling groups have been matched or the distortions ment.'this integral testing is designed to investigate both have been minimized for those tests in which matches the mitigative processes and phenomena that lessen the cannot be act,ieved, (e) scoping tests have been conducted impact of an HPMii as well as the inherent mechanisms and instrumentation and procedures required to carry out contributing to the pressunzation by DCH. HPMIi/DCH experiments are reliable and repnx!ucible, and (d) test conditions include more prototypic conditions in terms of scaling, we sought to simulate the pressuriza- (i.e., steam-driven melts, realistic containment com-tion of the containment to the extent practicable without partmentalitation, sources of water, potential for hydro-introducing distortions that significantly impact the simu- gen combustion, eted. Integral effects tests are based on lation of the dominant phenomena influencing the funda- detailed modeling of the Zion and Surry plants, designs mental behavior. As an example, tests conducted at less that ace representative of two large classes of reactor than full scale will pnxtuce conditions in which trans- designs. Scaled models of the reactor pressure vessel, ported debris has a shorter residence time in the atmos- reactor cavity, in-core instrument tunnel, and subcom-phere; if debris vek> cities are preserved, a test at 1/10 partment structures were constructed. The subcompart-scde would result in a much shorter residence time for ment structures and equipment modeled included the debns to heat the atmosphere.Thus, preservation of de- crane wall, steam generators, reactor coolant pumps, seal NURiiG-1365, Rev. I 10

. .._.- _____m._ _ _ . _ _. _ _ _ . _ _ _ _ _ _ _ - _ _ _ _ _ _ ..

a 2 Research Plan ,

1 table room, biological shield wall, refueling canal, radial tive_ melts (instrumentation limitations) and their added beams and grating, and the operating deck. complexity, these tests do not provide the most efficient mechanism for determining rate processes or detailed On February 14 and June 1,1992, the staff reconvened spatially dependent information. Though integral tests the DCII peer review group with the purposes of(1) de. provide a posttest measurement of entrainment fraction, ,

scribing the progress and results of the initialcounterpart there is no ready means to measure entrainment rates tests at 1/10 and 1/40 scale, (2) discussing preliminary within the cavity or to determine the characteristic proc-conclusions and observations regarding swale effects, and ess by which debris is swept from the cavity, Integral tests (3) proposing the balance of our integral test program. also do not provide for measurement of the characteristic Table 2.1.1 presents that portion of our integral test pro. particle size range associated with the entrained debris in gram proposed to be carried out at SNL For complete. the cavity. Resolution of subissues related to liPMB and ness, tests that have already been conducted are listed DCil, such as the location and rate at which material is along with the data of the test. Note that Table 2.1.1 oxidized, depend in part on this information. Further-identifies tests at a third and larger 1/6th scale that are more, the examination of rate parameters provides for proposed to be conducted. These tests would be pet. the confirmation of behavior observed over a time inter-formed using the existing Containment Technology Test val, e.g., measurement of an entrainment rate allows the Facility at SNL that has been used in the past to investi. confirmation of an entramment fraction.

gate containment integrity (containment leakage and fail-In order to obtain the general information identified in ure beyond design pressure). As evident from the test matrix, the approach m integral testing is systematic; ma. the pieceding discussion, the NRC has initiated a sepa-jor synergetic effects are added m the progression of rate effects program to be conducted at Purdue Univer-sity. Six specific phenomena will be studied in detail; testing. I or example, m, order to expenmentally measure hydrogen generation directly and to measure and observe 1. Corium jet disintegration immediately after dis-the incremental effects of hydrogen combustion, the first charge integral test was conducted in an inert (99% N2) atmos- .

phere with later tests (IIIT-3,4,5,6) allowing combustion 2. L. quid film formation upon impact of the j.et

, with an oxygen bearing atmosphere, in similar fashion,- 3. lintrainment and drop formation by streaming gas the effects of cavity water inventory and containment 4. L quid film transport from pressure and shear water heat removal are included in the test program.

Proposed counterpart testing in the 1/40 scalc COREXIT 5. Liquid film ejection and disintegration facility at ANL was originally designed to mirror the 6. Capture of liquid mass in compartment Surtsey tests through IET-7. Other options under consid-eration include use of the ANL facility to explore issues Each phenomenon will first be studied experimentally -

related to the use of a simulant melt-the ANL facility using tiie preliminary scaling parameters.The proposed has the capability for conducting tests with UOrbased mechanism and model for each phenomenon will be sys-thermitic material. At this stage three tests have been tematically tested at more realistie scales than the original

- conducted in the 1/40 scale facility, counterparts to data base. A preliminary study indicates that a 1/10 linear IET-1, Iffl'-3, and Ilir -6. scale experimental facility for the separate effects tests is adequate when water and woods metal are used as

. Relative to the research needs identified in subsection simulant material. This is also attractive since this scale is 21.1, the integral tests provide direct indication of the compatible with the SURTSEY' integral test facility at- ,

fraction of debris swept from the cavity (entrainment . Sandia. Flow visualization and detailed instrumentation fraction), subcompartment trapping, hydrogen genera- are possible by using the simulant fluid pairs. Either of tion, and hydrogen combustion. Indirect information is them is quite difficult or impossible m the integral facility derived for water vaporization, debris-to-gas heat trans- usmg nue pmtotyp,ci high-temperature matenal. The fer, and debris fragmentation. In the case of debris frag- experimmtal data are then used to verify or scale-up the

- mentation, posttest analysis provides a direct measure of prop sed correlat,ons.

i For this purpose, several basic-recovered debris, however, this is not a measure of the - existing models and correlations, as well as the ones de- -

- debris particles upon inception of entrainment nor is it a veloped by previous N RC programs, will be systematically measure of characteristic debris geometry or surface area exammed. The study will be carn,ed out by a detailed:

. for melt that can not be recovered in particulate form mechanistie modeling study, together with visual observa-(because of agglomeration of molten or semi molten ma- tions thmugh high speed photography and the data from terial). the kwal instrumentation.

Overall, the separate effects testing and model assess-

- While the integral tests provide valuable information and ment is projected to be complete within 2 years, with ,

allow for a global evaluation of DCil, because of the completion of the first phase of testing, using a water-air difficulties associated with use of high-temperature reac- system, scheduled for the end of calendar year 1992.

11 NUREG-1365, Rev.1

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

2 Reccarch Plan Table 2.1.1 Test Test inillal Conditions Target Date lirr-1 Melt Mass: 43 kg Fe oxide /Al/Cr thermite 9/13/91 -

Driving pressure: 7.1 MPa (llrr-1), steam ,

liole size: Tube ejection with ablation,3.5 cm )

, Atmosphere: Inert, N, l Cavity water: condensate level,0.9 cm :l Containment water: none  ;

l

!!rr-2A Thermite temperature measurement 11/1/91 ilIT-3 Atmosphere: reactive,0.2 MPa alr/N2 12/13/91 The rest of the initial conditions are the same as llIT-1.

IITF-lR Repeat of II?r-1 except driving steam pressure of 6.2MPa 2n/92 lirr-4 Containment water: condensate level 3/20/92 The rest d the initial conditions are the same as Ilir-3. ,

liiT-211 . Repeat of 111T-2 4/28/92 lirr-5 Atmosphere: 0.2 MPa alr. COz,112 5/13/92 The rest of the initial conditions are the same as liiT-4.

IliT-6 Atmosphere: reactive,0.2 MPa air, N2, lla (scaled) 6/18/92 The rest of the initial conditions are the same as !!!!'-3.

IIIF-7 Atmosphere: reactive,0.2 MPa air, N,,112 (scaled) 7/09/92 '

The rest of the initial conditions are the same as IliT-4.

ILIT Cavity water: 1/2 fill with water 7/30/92 The rest of the initial conditions are the same as IllT-3.

Consersion to Surry geometry and subcompartment 9/92 Ilfr-9 Melt Mass: #kg (scaled)- Fe oxide /Al/Cr thermite 9/24/92.

Driving pressure: 13 MPa steam llote size: Tube ejection with ablation,3.5 cm Atmosphere: reactive,0.16 MPa, alr/ steam /lla (scaled)

Vessel without annular gap Cavity water: condensate level (scaled)

Containmerit water: condensatelevel(scaled) li1T-10 Counterpart to if1T-9 in 1/6 scale CIT facility 10/15/92.

Scale parameters accordingly.

l Ilir-11 Atmosphere: reacthe,0.16 MPa, air / steam /II2 (scaled) -11/12/92 Vessel with annular gap The rest of the initial conditions are the same as lirr-8.

Ilir-12 Counterpart to Ilsr-11 in 1/6 scale CIT facility 12/10/92 Scale parameters accordingly.

NURiiG-1365, Rev. I 12

_ _ _ _ _._ __ __ _. ._ - . _ _ - _ _ _ _ _ _ _ _ _ _ _ . -. . ._.m 2 Research Plan The third major element of our l>CII research program is ment, and document LONTAIN models for RPV and the development and validation of a systems-level code, cavity phenomena, relymg wherever possible on past ac-i.e., the CONTAIN code. The need for a systems level tivities including the SNL preparation of a DCli models code, and specifically CONTAIN to assess the conse- and correlations document and those tasks of the TPG in quences of a high pressure melt ejection, reflects several which model assessment was undertaken. Given the rela-considerations. tive abundance of modeling approaches that have been proposed in the past, no new model development is being L The diversity of reactor designs and the range of undertaken; the individual model selected will be made .

possible accident conditions leadisg to an 11PMH following expert peer review of these models, preclude the experimental confirmation of plant re-sponse for all reasonable combinations. 'iherefore, For RPV phenomena, models will be incorporated that some modeling approach is necessary to address the are consistent with those models recommended by the potentially wide range of outcomes. TPG and used by SNL in scaling the integral tests. For the cavity processes such as entrainment rate, a variety of

2. A highly complex detailed analytical treatment of models will be included that reflect the variety of prot DCII by a multidimensional finite ddference or fictd posed modeling approaches (Whalley-llewitt, Ricou-code is not practical, given the capabilities of avail. Spalding, Kataoka-ishi correlations), incorporation ~ of able models (e.g., KlVA, CONCHAS-SPRAY, more than one model 4 a process allows for a compara.

HMS. COMMIX) and the effort required to apply tive assessment of m uel adequacy without a significant these codes even in their present incomplete form, penalty since most correlations rely on similar variables Furthermore, application of such codes at this stage calculated internally in CONTAIN.The implementation 1 would introduce the additional uncertainty associ- asd initial assessment of CONTAIN DCH models is ated with general validation of the codes themselves. aiduled to be completed by the end of CY 1992, 1

3. . IIPME is only one phase in the overall progression 2.1.3 Ah!icipated Results '

of a severe accident that procads tu vessel failure at high pressure. hwi gne resulting DCil has an influ- Our planned rs earch on direct containment heating is ence on the subsequent plant response by altering directed to produe generalized modeling, incorporated containment atmosphere conditions from which into the CONTAIN wde, capable of predicting the con-later phenomena, e.g., long term pressurization or sequences of a high-preare melt ejection for the range -

core-concrete interactions, will influence contain- of applicable accident conddons.The modeling approach ment performance. Thus, calculation of the inte- will have been validated aghhtst available integral and -

grated containment response to a severe accident separat e effects tests results. Scaling issues will have been over the entire course of the event dictates t_ hat a addressed by comparison and analysis of test results for severe accident code serve as the vehicle for the experiments conducted under this program at three dif-assessmem. CONTAIN, the NRC's severe accident ferent scales, as well as experimental data generated from -

containment response code, is the logical choice for other research, Selected analysis of full-scale reactor-modeling. As a practical consideration, past efforts plant analysis will be performed to gauge the expected to model DCil have enabled the NRC to modify the results for some subset of plant and accident conditions, code with a reasonable level of effort. Plant analysis will be performed in part during the pro-gram as an iterative process in order to assess the de-

4. Interpretation and analysis of integral tests require mands for accuracy in modeling, i.e., to determine how an analytical model capable of calculating the syner- good or accurate an answer is needed and the effect on getic effects associated with DCH. Analysis of the plant response to variations in modeling.-

integral tests with the tool to be used in plant analy- -

sis is also essential to most directly address the Another expected outgrowth of this research will be the-scalability of test results and models, development of generalized assessment criteria that may -

be used in lieu of more detailed analysis for a more spc-The principal challenge in applying the CONTAIN code cific set of accident and plant parameters. Such assess-to DCH is to incorporate models that are sufficiently ment criteria can serve as the basis for evaluating the mechanistic to predict the influence of major plant and adequacy of specific calculations based on other models wWW. peaMtas, yet are consistent with the overall and can serve as the basis for the development and assess-neineering, correlational modeling used in CONTAIN. ment of simpler models adopted for use in full plant L is also necessary to allow some flexibility in modeling severe accident codes such as MELCOR and MAAP.

given present uncertainties.

While our planned research is anticipated to provide for in accordance with our program objectives, the NRC has resolution of the issues that have long been associated initiated efforts at SNL to quantitatively assess, imple- with DCH. it is important to note that our research pro-13 NUREG-1365, Rev.1 1

2 Research Plan gram will not fully resolve questions surrounding the phe- characteristics of the melt released from the core nomena related to corium water interactions that may into the lower plenum.

result from the ejection of debris at high pressure into a deep water pool that may be in the cavity in some designs 3. '!he mode of vessel failure that determines the rate in certain severe accident sequences. While our testing of melt release into the containment and the timing program includes a very limit-d investigation of the ef. of the release.

fccts of cavity water inventory for the accident conditions simulated, a complete resolution of this general issue, to 4. '!he application of experimental results in models of the extent it bears on plant performance, relies on our the governing phenomena for severe accident safety assessmentsflhese models are also to be used, usu-research program on fuel coolant interactions.

ally in simplified form, in the severe accident sys-tems codes SCDAP/RELAP5 and Millf0R.

2.2 Core Mclt Progression Researcli If during the course of this melt progression research the need arises for special research results outside the ab(we inacssel core melt progression describes the state of an areas of focus, for example, specific material properties, i AVR reactor core from core uncovery up to reactor ves-then such research will be undertaken after a peer review sel melt.through in unrecovered accidents, or through "I the need. A strong effort has been made m planning temperature stabilization in t.ccidents recovered by core du.s resead pmgram to focus on the key areas of uncer-refhnling. hielt pmgression provides the initial condi- tainty in issue resolution, and not io try and cover a large tions for assessing the loads that may threaten the integ-t M unprioritized techmcal uncertainties.

rity of the reactor containment. "Ihe significant results of melt progression are the melt mass, rate of release, com' 'the research needs in each of these areas are discussed in position, and the temperature (superheat) of the melt the remainder of this section.

released from the core and later from the reactor vessel at vessel lailure. hicit progression provides the in-vessel Whether a metallic blockage develops across the core hydrogen generation and the conditions that govern the during a severe accident, as occurred at TMI-2, or the in-vessel release of fission prtxtucts and aerosols and their metallic melt (and later the ceramic melt) orain promptly transport and retention in the primary system. Melt pm- to the lower plenum from the core (and llWR core plate) gression also provides the core conditions for assessing has a major impact upon subsequent melt progression, as accident management strategies. shown in Figure 2.2.1. 'the mass, rate of release, and other characteristics of the ceramic melt released from th te ctor core into the vessellower plenum impact the 2.2.1 Research Needs overall consequences of the accident. Analysis indicates Understanding core melt progression phenomena, in, that in llWR accident sequences with automatic depre.

ssurization, the blowdown lowers the water level below cluding an assessment of uncertaintics, is important in _

determining the margin of conservatism for those severe the core (and core plate) so that core heatup occurs under accident issues that have been resolved (e.g., llWR Mark dry core (ntgligible steam flow) conditions All the avad-Iliner failure) and in resolving the remaining severe acci- able experimental data on melt progression are for wet dent issues (e.g., DCil)Jihe objective of the research on core conditions (i.e., water m the bottom of the core). No melt progression is to provide quantitative tools for as- data currently exist for the dry core conditions that would sessing the mass and other characteristics of the melt resuh from the autematic depressurizatio,. that is called released from the core and from the reactor vessel,as well for by the emergency operating proccoures for U.S.

as hydrogen generation and the conditions for ftssion llWRs. Analysis indicates that core heat up, particularly pmduct release, transport, and retention.This research is the rapid oxtdation transient, and metalh,c melting, relo-focused on four general issues or technical areas: cation, and blockage formation are significantly different for dry core and for wet core conditions.

1. Determining whether a metallic blockage of the core 'Ihe current experimental and analytical research pro-(or core plate), similar to that which occurred at gram to determine whether core blockage occurs or does TMI-2, would occur in BWR accidents, or alterna- not occur under llWR dry core accident conditions is tively, whether the metallic melt drains from the discussed in Section 2.2.2. Analytical modeling of the core (and core plate) when formedflhe core bh>ck- pmcess of melt rehication and bk)ckage formation is age or melt dramage branch point and the different needed to provide a tool for extending the experimental melt pmgression sequences for the two different results to the full range of risk-significant severe accident pathways are shown in Figure 2.2.1. conditions for both BWRs and PWRs.

2. 'Ihc conditions for ceramic pool melt-through from The second research area concerns the characteristics of a bkicked ceN that determine the mass and other the ceramic melt that drains from the core (and BWR NURIIG-1365, Rev. I 14

_ J

2 Research Plan C88 W A*mseh Depenae hassem pWn)

Control Ro(Wede M Mets .

I l

Coos Opens Up @Wn) oxidadenTeemient Repid togen sieur Hyalregon 3-as -Pwn e enere m en - s W n 8-Poesible MCS Tr^: " , PeNure (PWR) 2ry test Nelseenen p O' k, .,

Metelao crust IIetems Men' Draining fesse o sse er o m Finse ..

P i

men Debels Bed Ceresnic Reelt Pool . CereseteIssit Draining ercoes C.Napse Pool m e i i

1P If Mo4MWeter lateseedone tdelMWanerlatermoeiens Lower Plenuen Bolte Dry Lmeer Planuse BeNo Dry .

y Vessel Fesure VW PeMure y-Figure 2.2.1 Gire Melt Progression Sequence Showing 111ockage or Drainage Paths.

15 NUREG-1365, Rev.1 -

2 Research Plan -

core plate)into the lower plenum in a bhicked core acci- aides. Section 2.4 presents the NRC research to develop dent. lhe sigmficant melt charactenstics (mau, rate of and vahdate severe accident, codes.

release, temperature (superheat), and composition (metal content) are largely determmed by the threshold While it is recognued that appropriate seiding analyses and h> cation of melt-through of the metallic core block- are important to ensure that the cxpenments are per-age (and secondary ceramic crust) by the ymwing ceramic formed under the proper conditions and in the proper melt pool. These charactenstics are Scry important in parameter range for appheation to full scale reactor acci-assessmg the impact of core-melt accident sequences on dents, it may be that a specific issue is not completely containment loads. Most of the information currently amenable to scaling analysis. In these cases, expenments available on the late (ceramic melt) phase of melt pro- will not be used to develop empirical correlations, but pression, includmg bhickage melt-through, has come rather to identify important dominant physical / chemical hom the Th11-2 core examination. This information, processes and assure that they are properly represented however, is not sufficient to validate modeling of the m the analytical models. Analytic;d models of the key melt-through threshold and location and the characteris- phenomena are also necessary for the results to be apph-tics of the released ceramic melt. cable over the range of severe accident conditions. _

Ceramic pool melt-through from a bhicked core is a 'the primary cunent development activitics in the analy-whole-core phenomena and is not amenable to direct sis of melt pmgression processes involve the DEHRIS cxperimentation. 'lherefore, experimental research porous media late-phase melt progression model and the needs to be conducted on separable key parts of the N11 IRIS model of metallic melt relocation and blockace pmolem with integration of the results by analysis and formation that is an extension of the porous media frame-mdes. For example, the surface heat flux distribution of work of DliHRIS. Rather than trymg to describe the the internally heated ceramic melt pool that results from detailed geometry changes during melting, melt reloca-natural circulation in the pool is an important phenome- tion, and the freezing and remelting of crusts, DEHRIS non in determining the meltthro'igh threshold and hica- and MERIS use a much simpler porous media framework tion in blocked. core accidents. An assessment will be that keeps track only of material relocation and state undertaken to determme whether or not correlations de- (liquid, solid, composition, temperature). DIiHRIS pm-nved from the existing data at lower Rayleigh numbers vides a 2D(r,7) mechanistic treatment of the key phenom-and in different geometry than expected in a core melt ena involv;d in late-phase (ceramic melt) melt pmgres-accident are adequate for characterizing the surface heat sion in bh>cked coie accidents (Refs.1,2). Spectfically, flux distribution of the ceramic melt pool. The research DEHRIS treats the melting dynamics of a particulate pmgram to determine the threshold and hication of melt- ceramic debris bed supported by a metallic crust across through is discussed in Secdon 2.2.2. the fuel rod stubs in the lower region of the core,lhe modeling includes ceramic crust fonnation around the The third research area concerns the mode and I; ming of melt pool and pool melt-through of the supporting crust vessel failure in core melt accidents. A primary question is system and melt drainage frorn the core into the lower -

whether, for a given reactor type and accident scenario, plenum. htERIS uses a 3D(x,y,z) formulation to treat the Sessel failure is by penetration failure or by global or hical complex HWR core geometry and a time-varying aniso-creep rupture failure.These dtfferent failure modes give tropic effective permeability to acccunt for melt reloca-very different results on the rate of melt ejection from the tion and freczing (Ref. 3). DiiHRIS and MERIS respec-vessel and also on the failure threshold.1he research on tively, have shown good agreement with the very limited lower head failure models reported in NUREG/ experimental data currently available, or ceramic and on CR-5M2, "laght Water Reactor Iower Head Failure metallic melt rekTeation and crust formation.

Analyses," is discussed in Section 2.2.2.

An assessrent is needed of the ability of the DEHRIS 1he fourth area concerns severe accident systems codes, porous-media model to predict the late-phase melt pro-such as SCDAP/REl.AP5 and Mlil,COR. These codes pression beravior in the TMl-2 accident. An assessment model the key physical pmcesses that occur dunng in- is also needed of DEHRIS with the experimental data vessel melt pmgression, often in simplified form, in order from the melt progression (MP)experimentsin ACRR on to determine the global response of the reactor during a the rehication and failure dynamics of the ceramic nnd core melt accident. Detailed phenomenological models metallic crust system that supports the growing ceramic are also used to perform analyses of the experiments in melt pool. Also needed is an assessment of the MERIS order to gain a firm understandmg of the important physi- porous media type model of metallic melt rehication and cal pmcesses. This understanding forms the foundation bhickage formation with the experimental results of the for the simphfied models in the severe accident codes. planned ex. reactor experiments on this subject and with The detailed phenomenological models and the thorough any other relevant available information.11 the assess-analyses of the experiments form the basis for assessing ments of DEBRIS and MERIS. with possible improve.

the adequacy of the simphfied models in the systems ments, are successful, their modeling. probably in simpli-NUREG-1365. Rev. I 16

2 - Research Plan ,

fied form, can be incorporated into SCDAP/RiilAP5 into the experiment test sections are obtained by analysis and Ml!LCOR. of the flow of metallic melt from the upper three-quarters of a full scale BWR core under dry core accident condi-2.2.2 Current Research Program tionsJlhis analysis is consistent with the phenomenology observed m the ACRR DF-4 and CORA HWR tests.

As described in Appendix A.2, a great deal of information 'these tests were performed, however, for HWR wet core has been obtained on the processes involved in the early conditions. For most of the tests, there is pre-oxidation of phase of melt progression that extends through core deg- the Zircaloy (and stainless steel) surfaces by furnacc rudation and metallic material melting and relocation, heatingin steam to provide oxide film thicknesses that are IIxcept for the llWR experiment in the Canadian NRU representative of operating IlWRs. The initial tempera.

reactor, which has been delayed because of a leak in the ture distribution in the core and core-plate structure is heavy water system of NRU, current NRC research ou produced by preheating the system with downward flow-melt progression is concentrated on two major uncertain- ing argon. There is compensation for radial heat losses ties or issues. 'lhe first issue is the llWR accident condi- with external heaters on the insulathn of the test assem-tions, if any, in which a metallic core blockage, similar to bly. A detailed description of these experiments and their that in the TMI-2 accident, would not be formed. 'Ihis is objectives is given in Reference 4.

the major branch point in the melt progression sequence shown in Figure 2.2.1. The second issue concerns the A cross-section of the test fuel bundle is shown in Fig-conditions for the meltthrough of the growing pool of ure 2.23. *lhis has fulbscale radial simulation of half a ceramic melt that is supported by the metallic blockage. channel box unit cell in a llWR core. "Ihis simulation Research on the mode of vessel failure in core melt acci- - includes the control blade in the gap between channel dents, which in general is plant-and accident-sequence- boxes, the roughly equal horizontal length of unbladed specific, is limited to the validation of the models devel- gap that is potentially important for melt drainage, and an oped and presented in NUREG/CR-5642. Issues related additional two rows of fuel rods outside the gap. These to severe accident core melt codes are discussea in Sec- additional rods provide a prototypic temperature distri-tion 2.4. bution in the channel box walls, in the gap, and in the control blade, and they also provide a volume for diver-2.2.2.1 Core filockage or Melt Drainage in IlWR sion of the melt flow following channel box failure.

Accidents An ex-reactor experimental program has been started to This simulation has 64 fuel rods, channel box walls, and resolve the question of metallic melt blockage or drainage bladed and unbladed sections of the gap between the under IlWR dry core accident conditions. Unlike the pre. channel box walls. In initial operation of the experiment vious ACRR and CORA BWR experiments, the ex, and for debugging, it is planned to perform two'Xill reactor experiments do not start from the initial intact rod experiments in the simplified geometry of the blade, gap,-

geometry. Instead, the ex-reactor experiments accurately and channel box walls shown in the. upper schematic reproduce the core conditions at the onset of metallic drawing of Figure 2.2.3. These experiments will also pro-melt relocation, and the system behavior under these well vide basic information on melt rek) cation flows before -

defined conditions is then measured. channel box failure and on the effects of pre-oxidation on the interaction with and failure of the channel box walls Since internal heat generation in the metallic melt is by the control blade melt.

negligible in reactor accidents, these experiments can be performed ex reactor without internal heat generation The XR2 experimental test assemblics will include a full with pours of metallic melts of prototypic composition radial-scale mockup of a corresponding half unit cell in and flow rates (dribbles).The test assemblics have proto- the core plate region below the core that includes the typic dimensions, materials, structures, heat capacities, prototypic complex flow paths, materials, and heat ca.

and temperature distributions. Because of the axial tem- pacities, as well as simulation of the interior section of the perature distribution, illustrated in Figure 2.2.2, bk)ckage IlWR channel boxes with fuel rods and of the core plate formation is only possible in the lower quarter of the core structure below the core. The complexities of the struc-and in the core plate region, The metallic melt comes tures and flow paths in the core plate region are shown in

- from the upper three-fourths of the core and the metallic Figure 2.2.4. The core plate itself occupies only 34% of pour represents that part of the core in these experi- the horizontal cross-section in the core plate region, and t ments. The ex-reactor experimental apparatus represents most of this cross-section is relatively open with complex the lower quarter of also the core and also the core-plate flow paths defn.ed by thin, low heat capacity structures of region. The pour of metallic melt represents the incoming low-melting stainless steel. Because of Ihese geometrical metallic melt from the upper three-quartersof the core to complexities, rivulet flow complexities, and the effects of give overall full-length simulation. The mass, flow rate cutectic interactions, analysis of ' the potential for (dribble), and composition of the metallic melt poured IAcckage or drainage of cutectic metallic melts of control 17 NURl!G-1365, Rev.1

2 Research Plan nh e.da eha g.e as. be. ,. 4 h. st aie4 -p r . saeed mmy +m

'o0000000' 'o oo o ooOO' 00000000 00000000 00000000 00000000 88888888 88888888 00000000 00000000 00000000 00000000 1600 <o0000000 A0000000,

,'00000000'[ 'O0000000' ,

.. ... .... ........ . . . . . p, 88888888 88888888 3400 , , 0000000 00000000 000 000 00000000 x -

~/ ~, 88'd888 8888 888

'LI 1200- - - -- - - 5 <-

. . * * *g

()S.,*. , .*. _ *_ *.,, '

yu .

1000' -

n'.

G . .-

E '

G)

>~ .. .....,.. .... ..; . . .. :... ..

809 ,

4 600 - , - ,**-

400 0 1 2 3 4 Axial Location, m liigure 2.2.2 Asial Temperature l'rofiles for the Control lilade, Chant. Ilex and 17uel Rods, Indicating Maximum lateral Temperature Variations NURiiG-1365, Rev,1 18

2 Re$earch Plan insulation ~dg 3[N m] <[):

Electrical - - P og L

Channel Connections [) :

l ., -- # Dox Wall 2

Heating l

~ ~ ~ * '

Element l

Zircaloy M Wedge 9 Deflector L[L S:.^ l.~.d$L.u. -

_p X

+

f L -

C._a__ ,] l

,~,.

c, p 7 ' "'"g o;. -

A'N Control XR1 Bundle olade

..$.U

g;

• 'T

_, ht ,;.

Insulation s ,h OO f[

Y['O', . - O[ Channel Fuel - Box O ' #"< ',

Rod Simulotors

,/gg y 4- ]Q Walls

' O O O rd 30

& 00000 O..

e'eeeeeee h";Fl"

.Y SOCOOOO M b[$ Region A

a -, .

"'1,;

e. w (tn. ,

' b; g.ygggg g Zircaloy 00O0000 _p o*,$'o,

,e

<f: 4, .) .

{ ?l

\

Control Blade Figure 2.2.3 XR1 and XR2 Test 13undle Cross Sections.

19 NUREG-1365, Rey,1

. - _ ~ _ _ - . - - - . - - - _ - - . _ __

2 1(esearth I'lan Relocation Pathway Through Canisters bypassing the s s k Lower Core Plate-

WQQ $ :.!W6DUE

90M&% 1ljr- N

,1c -lMDDD6 1lj

,ic SPACER r r GRID I {

/-- TIE PLATE

, , , t ,. c , ,. . .

,/

MQ;Qi; ';i,1;Q (i;G .;gG?;i,i;G gg,gi Qi '

NOSE PIECE V ,

V < V V f g , FUEL SUPPORT k _/

3 5 1 Th PIECE h g g,g g F$ CORE PLATE

\); [

~

1 CONTROL f

[q;T '"'

- 7~ BLADE DRIVE d

~

v

~

yv E y

@ l

/y GUIDETUBE liigure 2.2.4 Melt Drainage Pathways *lhrough the 17uci Canisters Which Ilypass the lewer Core I' late.

NUllEG-1365. !!cy 1 20

2 Research Plan

blade anatenals and Zireakry m ' v pute repon Af ter the two initial simphfied XR1 expenments to check needs guidance from experiment ,

i m pmned to out the experimental techniques and to determine the perfonn the XR2 experiments in , environment. effect of pre existing oxide f Ims on the channel box Zir-cedoy, the first of the four planned XR2 experimerits will Analysts with MF.I. Colt has indicated that two separate be performed with all the above parameters at the end of rehications of metalhe melt occur under llWR dry core their range that favors melt drainage. If core blockage conditions, the fast a melt of II,C and stainless stect does occur under these conditions,it will then have been control blade ruaterials with a melting (cutectic) point of determined that the bhicked core sequence is to be ex-atxiut 1500'K, and the second a rnett of cladding and pected in all dry core llWR accidents. 'the experimental channel lex Zircaloy with a melting point of about matrix would then be shortened to focus on t he character.

2200'K. 'lhe two simplihed gap-geometry XR1 experi- l istics of the metallic blockage formed under the best I ments will use a sing!c pour (dribble) of control blade estimate conditions.'lhis is needed for assessing the dura-melt. 'the XR2 experirnents, which include the fuel rod bilityof the bhickage with continued core heat up and also region within the channel boxes and the core plate region for assessing bkickage melt-through by the growing ec-helow the core, will use two separate pours, one with ramic melt pool in late-phase melt progression, if melt control blade material melt with some dissolved Zircahiy drainage does occur in the initial XR2 experiment, the and the other with a higher temperature Zircahiy melt- preliminary test matrix given in Table 2.2.1 will be fol.

The melt mass, composition, temperature, pour rate, and lowed step by step, increasing the potential for bhickage the time of separation between the two pours in the XR2 to see if bhicknge is achieved under these dry core condi- l cxperiments will be determinii by analysis and by the lions.

results of the dry core test soon to be performed in the Gerrnan COR A ex reactor fuel damage facility.

.the conditions for this test matrix will be reviewed as

'ihe effect of the pre-accident oxide film on the core experimental results are obtained, particularly those i

Zircahiy (and stainless steel) from normal reactor opera. from experiment XR2-1. Currently it appears that these tion in delaying material interactions and eutectic forma- experiments and corollary analyses with additional input tion will be investigated in the XR1 experiments. 'lhese from the CORA llWR tests and from the llWR F1J IT-6 results will be factored mto the XR2 experiments. Addi. test in the NRU reactor will provide sufficient data to tional data on this effect will also be also obtained from resolve the question of metallie melt blockage or drainage COR A experiments. Current plans are for nearly all the for the relevant range of IlWR dry core accident condi.  ;

XR2 experiments to be pre-oxidized. tions. 'lhe last of a series of fulllength tests on fuel damage and melt progression during coolant boildown .

The test matrix for the ex reactor experiments has been (wet core)is scheduled to be performed in the Canadian formutated to determine, with as few tests as possible. N RU reactor in late 1992.'the results will provide unique whether there are any conditions within the range of the length effect data for llWR core geometries. Two high.

governing pararneters (including uncertainties) for llWR burnup fuel rods in the 14 rod test assembly that is shown dry core accident conditions under which metallic melt in ctoss-section in Figure 2.2.5 will provide unique data on drainage, rather than core or core plate bkickage, can fission product release in the boron containing IlWR core occur. 'lhe governing parameters are the metallic melt environment. Post irradiation examination (Pili) of the ,

temperatures, the core axial temperature gradient, and test fuel bundle should also provide some information on the presence of an initial oxide film on the Zirealoy struc- the effects of high burnup fuel on early phase melt pro-ture corresponding to that in normal llWR operations, gression.

Table 2.2.1 Test Matrix for the Ex. Reactor Experiments on hietallie Melt Relocation and Ilhickage Formation Under llWR Dry Core Accident Conditions Test

• Meks' Grad T" MtitT Zry Oxide Objectives -

XRl-1 C.fl. Prototypic Prototypic Yes Oxido Film XRl-2 C.ll. Prototypic Prototypic No No Oxide Film XR2-1 C.ll., Zry liigh liigh Yes Favor Drainage XR2-2 C.II., Zry liigh low Yes Decrease Melt T XR2-3 C.II., Zry low low Yes Decrease Grad T XR3-4 C.ll., Zry 1.nw low No Delete Zry Oxide

'C.H. means control blade matenats plus some dmolved Zirealoy.

"lligh grad T meam that the AT wcurs over a short dntance. while km grad T means that the same AT wcurs over a long datance.

21 NURiiG-1365, Rev.1

2 ltescarth Plan i

. Zr-2.5% tab LOG' LNER TUDE

,3g .-. %

/ ZlRCALOY-2 PHES3URE OUTER 1UBE

/y- N It#4ER TUBE N /' SADOLE DYPASS FLOW 44tlVLUS

.- {

NSULATION (ZR02)

I; p S190U0 LINER l.

' DUNDLE C00LN4T MAKEUP UNE zz,z,,zzif g g i i -

A CHN4t1EL 00X h

• ~

- l bhN I '

c

.--- C04 TROL OLADE

" d

! TIME DO4AN REFLECTOAETER (UQUID LEVEU

,  ; ifMD

- IFilADIATED RODS 20 f. 30 kW l'igure 2.2.$ Test llundle Cross Section for the NitU llWit Test NUltliG-1%5, lley. I 22

2 Research Plan i

lixperiments in the German Colt A coreactor fuel dam- the late phase bkicked core. Nearly all the information are test facility with electrically fuel rods have provided currently availabic on these processes has come from the much of the current information base on core degradation TMI-2 core esamination. In the TMI-2 accident, pool and early (metallic melt) phase melt progression. 'Ihis meltthrough occurred at the side of the core and r eflected information includes materials interaction (eutectic) ef- the core refkoding in that accident, it is currently not fccts and the ef fects of rellooding damaged cores. 'the known whether in unrecovered accidents, meltthough i CORA results are available to NRC as part of the inter- would occur at the fide or at the bottom of the melt pool. l national Cooperative Severe Accident Research Pmgram 'lhe failure location makes about a factor of two differ- l I

(CSAlti') and will be used to further the data base for ence in the released meh mass.

code validation and assessmentJihe forthcoming CORA dry core llWR test will pmvide data on the initial condi- . .g g  ; ggg g g ,

tions of the meomm, g melt for the ex reactorexpenments on metallic melt relocation and bhickage formatton and e been designed to furnish phenomenological will complement the results of these experiments. information on the governing rnechanisms involved in the tekication and failure dynamiesof the fuel rod-supported metallic and ceramic crust system that supports the gmw-2.2.2.2 Ceramic Pool Melt.through frum a lilocked ing internally heated mostly ceramic melt pool. 'lhe melt Cme attack on the metallic crust involves both thermal attack and pissibly chemical (cutectic) interactions among the In bhicked core accidents, the primary determinants of crust materials and the Zircahry cladding and UO2 in the the mass and other characteristics of the mostly ceramic fuel md stubs. 'the results of these experiments are to be melt released into the lower plenum are the threshold applied at full scale to bkicked-core reactor accidents and huition of meltthrough of the metallic and ceramic through a mechanistic model or models of the melt-clust system that supports the growing. Inostly ceramic through pmcess. "Ile DEllRIS porous media model, melt pool. The failure threshold and kication are them- which treats the governing processes involved mechanis-selves primarily determined by the rehication and failure tically, will be the primary tool used in analysis and inter-dynamics of this complex fuel rod-supported system of pretation of the experimental results (Refs 1, 2). 'lhe _,

ceramic and metallic crusts. Uncerthinties as to the sur- simplified late phase melt progression modelmg in face heat flux distnbution of the internally heated melt SCDAP/Riil.AP5 and Ml!!.COR will also be checked pool because of natural circulation Sre also important, against these results.

particularly for the failure hvation. nn assessment will be made and peer reviewed to determine whether or not the r correlations derived from existing data at much lower A schematic drawing of the test assembly for the melt than pmtotypic Rayleigh numocrs and in different progression experiments, actually that for M P-2, is shown geometry than in those occurnng in blocked core acci. in Figure 2.2.6. Detailed descriptions of these experi-dents are adequate to characterite the surface heat flux ments are given in experiment plans for the melt progres-distribution of the ceramic meh pool. The TMI-2 melt sion experiments and for the MP-2 expenments specifi-pool had a Rayleigh number of about 10te, turbulent flow, cally (Refs. 5,6). In these experiments, a pre-cast metalhc and hemispheric geometry. Nearly all of the evailable crust with a prototypic low UO2 content (for the dissolved data are for a Rayleigh number of 10" or less, larainar or UOz m the metallic crust) bridges a 32 rod array of clad unsteady laminar flow, and thin, semi-circular geometry. fresh fuel rod stubs and supports a particulate ceramic In addition, there are a few data in thin, rectangular (UOz ,ZrO 2) debris bed. ,

geometry at Rayleigh numbers up to 3 x 100 that agree reasonably well with the low Rayleigh number data. If The purpose of the MP-l experiment was to provide data at a Rayleigh number of 1018 should prove to be base-line information on ceramic melt dynamics in the ,

needed, an experiment with water as a simulant fluid blocked core system and on the failure mechanics of the would be feasible. The time constant for establishing supporting metallic crust. MP-1 was conducted in Octo--

natural circulatien flow and heat transfer in the melt pool ber 1989, and the experiment was terminated earlier than also needs to be examined. In addition, analysis should be planned because a temperature safety limit had been performed to assess the possibility that the metal in the reached internally in the experiment package, The M P-1 mostly ceramic melt pool can significantly affect the natu- experiment provided basic information on heat itansfer, ral circulation in the pool and its surface heat flux distri- densification, and melting in the paniculate ceramic de- J bution, bris bed and on the formation and downward propagation of the secondary ceramic crust or blockage below the Very little experimental information currently exists on growing ceramic melt pool. Peak temperatures of alx)ut the melting dynamics of the ceramic melt pool and on the 3200'K were reached in the ceramic melt pool, which tekicationand failuredynamicsof thefuel rod supported gives about 400'K superheat above the UO,-ZrO, system of metallic and ceramic pool supporting crusts in cutectic temperature.

23 NUREG-1365, Rev.1

.: =

, . . , . - . , - - ,.-n.n-n-- , , . - - , _ ~ , , nv, . , - , . - . . , . , -.,,.,-,.nn., - . , - - . , ~ . , - - - , , .n.-r ,.ww .n_,na nn-,

2 1(esearch Plan UO2-ZrO2 DEBRIS WATER JACKET - nc2m STEEL JACKET EN g scris:

ZlRCONIA INSULATION TANTALUM LINE9

/ , , , , ,

meio neim OUTER THORIA LINER -

/ N"

"" 5" INNER THORIA LINER d, _ _ d ', ",,

TUNGSTEN LINER -- _ Twcas meism

" '"' N $$$

U-Zr-O-Ag-in-S.S ""

METALLIC CRUST 7J' c, y

- maim naim FUEL ROD ARRAY 3 c"2 2,cim C/ z fm$ / Zr4KD767

[ ',' / \ neem zS STEEL HEAT SINK 2RX0062

\- TSKPC3 1'igure 2.2.6 MP-2 Test Section Showing Major Components and 'lherrnocouple locations.

NURl!G-1365, Itev.1 24

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

I t

1 2 Research Plan

{

h1P-2 s to have a metallic crust of TMI-2 composition, wax carried out f or pressure s essel steel, Additional creep a which includes pWit control rod materials (l'e, Ni, Cr, rupture data acquisition is planned for the penetration Ag, In. Cd). A purpose of the A1P-2 experiment is to materials (Inconel, stainless steel, and SA105 or SA106 determine the effects of the culectics of these control. srcel). 'lhe results of this analysis have been reported in t tod ruaterials on the crust failure mechanisms and draft NURl!O/CR-5M2, "lJght Water Itcactor lower  ;

thresholds and to investigate the effects of the f uel rod llead l'ailure Analysis"(Itef. 7).  ;

stubs on the behavior of the secondary ceramic erust. It is 7

not cur r endy known whether eutectic (chemical)interac. 'lhe analytical modek in this report at e undergoing a peer

~

tionof thecontrol matenal containingmetallicetustwith review to ensure that the report is on a firm technical the fuel rod stub Zncaloy cladding su.cif is sigmficant, it is basis, hiodel validation will be undertaken when results planned to run MP-2 to metallic crust failure 'the cc. become available from the Swiss CORVIS program of. i ramic insulating shrouds and metallic shields in hip-2 integral tests on the melt pool thermal attack on the ,

have been nushfied to prevent the premature termina- vessel lower head and the head penetrations, which will tion of the experituent that occurredin h1P-1 and toallow be rnade available to NHC through the CSARP program, ternperatures in the melt pml well above UO, melting and from the cooperative research program between the (3100K) as well as the UOrZrO, cutectic (2800PK) USNitC and the I.V. Kurchatov Institute in Russia. An without reaching the safety hmit of the internal tantalum extensive series of experiments and associated analysis shield that required the shutdown of hip-1. has been started at the I,V. Kurchatov institute on the - ,

, integrity of the vessel lower head under melt pool attack.

With the results and interpretation of experiment MP-2 A major purpose of this research is to investigate the and with the results of melt progression sensitivity stud. efficacy of ex vessel water cooling in preventing vessel ies, an expert peer review group on melt progression will rneltthroughJthe research plans include large scale inte-review the need for and the nature of further research in pral tests with up to 200kg melts of UO, or molten salt, as the area of late phase melt progression. In particular, a well as smaller scale separate effects experiments and determination will be inade of the need for further melt analysis on melt pool thermal hydrnulies, progression experiments.

2.2.3 Anticipated itesults 2.2.2.3 Mode of Wssell'allute

'lhe first major area of investigation in the current melt

'the mode and timing of the reactor vessel lower head progression research program is the determination of failure have controlling effects on the subsequent con- w hether bhickage of the core by metallie melt or drainage tainment loading events in severe accidents. A program from the core and core plate occurs in llWR dry core that reviews and extends the numerous studies of reactor accidents, it is anticipated that the currently planned failure modes made during the past decade is nearing ex-reactor experiments on metallie melt rehication and completion. 'the major potential failure mechanisms in- blockage formation under ilWR dry core conditions along clude penetration tube heatup and failure, tube ejection, _with nmdeling will provide sufficient information to re-lower head global creep rupture, and hicalized therrual solve this question.The full length IlWR test in NRU and and mechanical loads on the lower head. This last in- results from recent ilWR tests in COR A should contrib-cludes nonuniform distribution of a the debris bed and ute increased confidence to the conclusion.'the results of coherent jets of molten core material that can directly these experiments wdl be used to validate models in the contact and ablate the vessel wall. SCDAP/Rl!!.AP5 and M11COR rystems codes for use in predictions of the severe accident behavior of nuclear llecause there are a large number of debris conditions. power plants, lower head designs, and accident scenarios, analytical techniques using key dimensionless parameters were - The situation regarding the second major area ofinvesti-used to develop failure maps that indicate the relative pation, ceramic pool melt-through from a bhicked core, patential for fmlure of the various males as a function of however. is far more complex. The results and in'erpreta-the dimensionless parameters.1.imited numerical finite tion of the ACRR MP-2 experiment should substantially element analyses were then utilized to benchmark the increase our understanding of the key processes involved failure maps.- in ceramic melt pool grc wth and in the dynamics of the rehication and failure of the pool suppordng metallieand An analytical method for bulging analysis (hical creep ceramic crust system, With the results and interpretation  :

rupture) was also developed. The analytical solution is of MP-2 in hand to augment the results of the TMI-2 based on an axisymmetric plastic analysis with a pseudo- core examination, sufficient information should be avail-planc strain treatment of deformation normal to a able for the expert review group on melt progression to meridional plane. This and other failure analysis methods assess the need for and nature of future work for late-require high temperature creep data.Therefore sa series phase melt progression issue resolution, Such future re-of tests to acquire high-temperature creep rupture data search may or may not include further melt progression 25 NURl!O-1365, Rev.1

=

wwd , p ~ er , yr--*"&"-ta----~ ywre a w-t 'T,-ry , e- r w,- v-v vr e -rv tr-vv v ~mvm'-%-.- w- -

m r' ww www4rwwwi -+'-W e h

• e- -- v re- y v - m r wo u '-n--'r *-- wv--e*w--- r -T " v FS*

2 Rocarch Plan ,

1  !

j  !

i experiments (t he need for an expenment on the thermal-hydraulics of mrlt luols with internal heat renerati""

2.3 I?uel Coolant iDienictiolls alul .

may be estabbshed, ahhough thn cuuently appears un- I)el)ris Coolability  !'

likely' 2.3.1 1(esearcli Needs f

'lhe report on lower head failure analpis will provide bi"" .IY #^

M ouonandu rm IY 4"""U "I"" "I " 6'C"*'

failure maps in terms of dunem.ionless parameters that e as a po% nu o ontab may be used to predict f ailure modes, in rnany cases, the rnent fadure (alpha mo&), ugruficant prortew has been n as te cid in m rent ud stub, 4 raost probable failure rmule may be predictaI>le without further analysis, if further analpis is required, only a NURliG-ll$0 where alpha. mode fadute does not f.ectn i

limited effort with imnimal uncertamty ranges should be to k a dommant contnhotor to cady contMnmendadute.

newmry. Ilowever, the progress in understandmg this area has been mainly directed at the conditions for m venel mol.

4 ten fuel muring I into a emdant pool and its likelihood of Section 2,2 lieferences causing containment failure by energetic interactions.

The duft in emphasis to accident management for a vari- ,

ciy of reactor geometries and meltdown scenarios, cou-1, S. S. Dos:mjh, " Melt Progression, Oxidation, and pied with the wide un(criainties in the NURl!G-1150

' Natural Convertson in a Severcly Damaged Reactor experI clicitations of f ucl. coolant interactions (101), kug, i Core " NUltl!G/CR-5316, SA.ND 88 3476, Sandia gest that conhonatory research is needed to learn more

! National I.aboratories Albuquerque, NM 0990). about the fundamental mechanisms of I'Cl to be able to ,

determine the conditions under which I Cl is important to severe accident risk and accident manarement. The em.

2. R. D. Gasser, S. S. Dosanjh, and R. O. Gauntt "'lhe phasis of this research is now shifted to providing the Dl!!!RIS Module An lilfective Tool for the Analy- appropriate methodological and analyucal tools for sis of Mell Progression in 1 WRs," Oghtecnt/r Irater evalur, ting major aspects of the accident sequences, in.

Reactor Safety Information Meeting, Oct. 22- 24,1990, cluding quantification of 61 cam and hydrogen production, ,

NUR!!G/CP-0114, Vol. 2, p. 74 the mode and timing of vessel (or reactor coolant system) failure, and ex.venel events of potential significance to ,

3. R. C. Schmidt and S. S. Dosanjh. " Core Structure debns coolabihty and containtnent hiadmg. Although all lleat Up and Matenal Reh>eation in a !!WR Short' t hese issues are discussed elsewhere in t his research plan. . '

Term Station lilackout Accident," Sisth Proceed ^ there are fundamental aspects of ITCis that are perrnanc to all these issues.

mgs of Nuclear Thennal llydraulics,1990 ANS Winter - Meeting, Washington, D.C., p. 42- Three specific issues that sequire additional infonnation Nov.11- 15,1990-cither from experimentation or from analysis are; 4.

1, i:01 energetics 1 R. O. Gauntt, R. C. Schmidt, and S. S. Dos:mjh, " A Program Plan for the lix Reactor Metallic Melt Re. 2. Fuel melt quenching in water pool location thperiments," letter report to NRC, June 3. Water added to a degraded core (in vessel as well as 1990. {

ex. vessel)  !

5, The FCI characteristics are briefly described below, and R. O. Gauntt and J. W. Fisk," An l.ixpenmentalPlan the experimental / analytical data base needed for more for the M.'*(Melt Progression)l,.xpenments, letter comprehensive resolution of these issues is discussed for report to NRC, September 1989.

developing the overall research to address the above three areas.

6. R. O. Gauntt and R. O, Gasser,"lixperimental Plan for the MP-2 Melt Progression Test in the SNI. 2.3.1.1 Research Needs-l'ucl Coolant Interaction ACRR," letter report to NRC, September 1990. (linergetio)

The fundamental aspects of l'Cl are the evolution of

- 7. J. l. Rempe, S. ALChavei, G. L Thinnes, C. M. liquid interfacial (fuel. coolant) area and associated heat Allison, G. I!. Korth. R. J. Witt, J. J. Sienicki, S. K. transfer during the contact. When the two liquids first Wang, C,11. Heath, and S. It Snow,"lj ght Water Reactor lower licad l aiture Anal)' sis" (Draft, ""* N"d""'

Wment of Ne #f#"I9 "'*"""H. "NCMW hd rY hlmlf AH A m US ( ommemal Nudcar Iwer 1%na,(

N URl!G/CR-5642, I!GO-2618, December 1991. NinuG75ml4 (WAsil 14nn), Ikermbet IW5 NURl!G-1365, Rev. I 26

_ _ i i_ _ _._ _ _ _ _ . _ . _ . _ _ - _ . _ _ . _ . . . ._. ._ _

2 Itescarch Plan come into contact, the molant begms to vapmi/c at the 'lhn is particularly true for the l'Cl because many of the fuel-ovlant hymd interface as a vapor film separates the fundamental mechanisms are not well understood.

two hquidt 'lhe system remams in this state for a delay period ranging from a few milliseconds up to a few sec- 2.3.2 Ctatrerit Resetirch Progratni unds. Durmg this time, the fuel and coohmt liquid inter.

mis (wnetimes called premixt ure)because of density and 'this section describes the current research prograrn for vehicity diff erences as well as vapor production. cach of the three specific issues discussed earlict-1:Cl energetics, fuel mcit quenching in a water pool, and add.

'the upor him destabdi/ation occurs next,inggermg fuel ing water to a degraded core, fragmentation. 'lhas rapidly increases the f uel surface area, vapori/ing more coolant and mercasing the local 2.3.2.1 Ileseanh Program to Address l'Cl 1:nergetics vapor pressure.'lhe vapor formation propagate ,patially 'the objectise of this tesearch is to determine under what throughout the f uel coolant mixture, causing the macro-conditions vapor explosion energetics must be considered scopic region to become pressurved. Subsequently, the and w hat are reasonable estimates for t he energeticyield.

high pressure coolant vapor expmds agamst the inertial As a first priority, we are testing the hypothesis that mix- -

wnstraint of the surroundmgs and the mixture itself.'the ing will be limited by local steam formation and high void vapor exphmon process is now complete, transforming fractions, causing water removal from within the fuel the fuel's internal encryy into the Linctic energy of the coolant mixture. With such data, the experimental results mixture and its surroundmgs. Itxperiments to validate mixing calculations and fragmentation are needed to ad- can be compared to the computational tnalels (e.g., IFCI dress macssel and to a large extent, ex-vessel fuel cool- or PM-Al I'll A).

ant interactions. To isolate the water depletion phenomenon from particle si/c distribution effects, experimental premixing simula.

2.3.1.2 llest arth Needs-l'uel. Melt Quernhing lions are carried out at the University of Cahfornia at Santa liarbara (UCSil) using clouds of hot solid particles.

'the TMI-2 accident indicated that under certain condi- A new instrument developed specifically for this purpose lions the inel melt may be quenched at the time of pour- is used to measure the local liqu d fractions in the three-ing into a water pool in the 111 V lower plenum. h eviously phase mixing wne. 'the experiments are scaled it had been assumed that a fuel pour into the itPV lower (1/8-scale), usmg numerical simulations (PM alpha) to plenum would iesult in either settling of the unquenched create water depletion regimes similar to full-scale pours tuel (and eventual 11PV wall failut c) or a vapor explosion. of fixed site 1-2 cm corium particles /these experiments -

Although there have been many integral I Cl experi- are currently in progter,s, and the results are expected to ments, there is no data available under these particular provide the first experimental confirmation of the water conditions llecause of thislack of data, the PAltO-1.WR depletion phenomena as well as a basis for assessing the experiments are planned.'they involve a prototypic fuel accuracy of its predictions in numerical simulations.

~

mass (50-150kgof UO 2 /ZrO2 /Zrat 3000'K)pouredinto saturated water at high pressures (5-50 bar for 1-3 meters Additional melt breakup will occur as a consequence of depth with a chamber diameter of 0.5 to 0,7 m). the hydrodynamic interactions in the mixing wne. Two experimental programs are currently addressing this 23.13 Researth Needs- Adding Water to a Degraded topic. The one at UCSil is examining the fundamental =

Core component of an explosion, a single melt drop, under .

conditions that cffcctively simulate the role of such a drop Severe accident management is a natural outgrowth of in a real explosion environment.This is accomp!ished by past emphasis on risk assessment. In fact, there are a using a _ hydrodynamic shock tube capable of generating number of particular issues that must be addressed when pressure pubes similar to those of an explosion (design accident management is the main objective. ' liming is pressure 20.000 psi). With dray diagnostles, the detailed important to accident rr aagement, and only recently time-history of fragmentation can be composed, and the have Pit A studies considered it in some systematic fash.

I ion; e.g., NUltt!G-1150 considered the effect of the op-results are the key input in numerical simulations of esca.

lation/ propagation.'this work is now completci a report cration of engineered safety features (containment fan will be published t>y the end of CY92.

coolers or sprays) before core heatup, before vessel fail-ure, or after fmlure. Inclusion of this " timing" behavior A second set of vapor explosion experiments is being can indicate where opportunity exists for operator inter- planned at the University of Wisconsin.The purpose of untion to help reach a stable coolable state. Since water these experiments is to produce a well controlled one.

is the primary accident management tool and the FCI can dimensional geometry in which a fuel simulant (e.g.,2-20 alter the course of the accident, it is important to investi- kg tin at 1300'K) pours into a water column, mixes with gate the benefits of adding water to the degraded core the coolant, an explosion is triggered, and the explosion with consideration of the possible adverse side effects. expansion work is measured. These experiments are 27 NUltliG-1365, Rev.1

2 Ilesearch Plan 5

i mmed at provahng benchmark data to examme the effect the fuel-melt pour and the lateral dimension of the facd. I of fuelwolant initial conditions and nuung on explosion ity. 'lhe fallo facility has the capabihty to deliver a large i energeties. 'lhe hypothesis for these tests is that fuel- mass of oxide melt under a variety of conditions.The fuel coolant mixing and explosion processes occur under con. pour rate is planned to be within reasonable ranges for  !

trolled wnditions. Data from these expesiments enable accident conditions, t.c., jet diameters 5- 10 cm and entry I cornparium among the fuel coolant mixing and explosion velocitics of a few meters per second.

Imidels used to make the case for applicabihty at reactor scale. This program is expected to be compt'eted by the Second, the FAltO cxperiments could be considered rep.  ;

end of (TD. sesentative of two types of geometn,e situations: (1) a single jet in a large water pool or (2) a unit ecil of a-It is expected that the above-desenbed work will provide inultiple jet pour into the lower plenum, in either case.

the basis for improved assessments of alpha failure as the adequacy of a l' Alto facility vessel to provide a prop-well as for several other special effects assessments as etly scaled lateral dimension depends to some extent on they arise in llWit ex-vessel sequences, where the rete. the degree of melt quenching. It is important to note that vant " sites" of the explosion and the correspondmg level the chamber cross sectional area will be varied by a factor '

of energetics rnay be considerably srnaller than those in. of two in the tests to be performed to specifically address volved in alpha fmlare considerations and where the need this point. If the results of the scoping test (50 kgof meh) t for best estimate, rather than tvunding, analyses is of and the base case experiment (150 kg of melt) indicate :

greater importance. l'or example, the sensitivity cdeula. that melt quenching is minimal (e.g., < 10% of the fuel tions in NUltlEl150 suggested that ex vessel explo. quenched during the pour), the steaming rate will be low, sions in a Mark-Il and Mark-Ill drywell may lead to level swell will be minimited, and the steam superficial drywell failure. Ilowever, more detailed analysis is velocity will be small. Under these conditions, the lateral needed to verify that this threat exists, dimension of the vessel will not be an important concern regardless of which scenario is considered. Conversely,if the resuhs of the scoping test and the base case experb 2.3.2.2 Itescatch Program for Quenching ment indicate significant melt quenching, the scalmg of To calculate quenching, the melt-coolant interface area the experiments considering the lateral dimension will be ami the constitutive laws that define the interphase trans. problematic for either scenario. Qualitatively, one would Iers (heat and momentum) must be detennined. On the still expect the experiments to be quite informative and constitutive laws, a reasonable estimate om be made by valid for reactor safety implications. llowever, quantila- 3

^

ex!cndmg 2-phase formulations. 'lhe program involving live interpretations of the tests must then account for the the determination of interf acial area requires the lateral dimensions and the likely large steaming rate, breakup history of the melt as it descends through the superficial vehicity and all other eonsequences associated coolant. Quite clearly, such experimental data are next to with substantial fuel and liquid water interactions, impossible to obtain; our approach, therefore, has to be Hased on these considerations, it seems clear that the a largely empinad. Available experimental data in this area logical sequence rs that the scopmg test arJ the base case are very limited, thus we have joined the program at the test should be performed in tim fallo-1, Wit experimen-

• fallo facihty in 1spra. tal program, then the sading rationale should be reas-

' sewd, and finally, the parameters determined for future

'Ihe objectis e of the fallo tests to be performed at the joint 1(esearch Center in Ispra is to observe the integral experiments. Currently eight experiments are planned .

for the test series.

behavior of fuebmelt quenching at high pressures under likely severe accident conditions. What makes these ex- WH d Program for lleflooding periments particularly attractisc from a technical stande point is that they can be performed with real reactor During a severe accident situation, there is littic doubt materials (UO2,ZrOc, Zr)at temperatures and pressures that the primary efforts of the operators will be directed that are prototypic of actual severe accident conditions toward achieving a stable, coolable configuration by mak-und with the proper full-scale water depths for in-vessel ing water availabic to the reactor vessel (ifloss of cooling -

accident situations. 'lhe instrumentation within the was believed to be the cause of accident). An important PAllOJI1HIMOS facility is t.ubstantial and the data col. question that must be considered is the likely conse-lected is extensive. Thus it is likely that computer code quences of these actions. Given the uncertainties in core comparisons will be made to gain modeling insights into melt phenomena and the state of the core during an fuel-melt mixing and melt quenching in water, actual severe accident, and given the intuitive drive to put water into the core in the event of an underenoting acci.

llecause these experiments are using prototypic materials dent, it is likely that the operators would inject water into l- under realistic initial and loundary conditions at the the reactor vessel should water become available during proper vertical length scales (e.g., water depth) the ques- the course of an accident. Ilowever, along with the poten-l tion of scale only becomes an issue relative to the si/c of tial benefit of achieving a stable. coolable configuration.

NUI'I t!O -1365. Itev. I 28 i

t _ - _ _ _ _ _ . _ _ _ _ - _ _ _ _ . , _ _ . . ~ _ _ . _ - . _ - _ _ . _ _ . - - _ _ _ __

2 Ilescarch Plan i

restoring water to a core that has been seserely damaged and implied permeabdities as affect ed by fuel rod disinte-can have elfects of which the operator should be aware, gration or clad relocation.

l The operator should be fully corni/ ant of possible symp-toms and responses of the plant to added watet under 2.3.2.5 liescarch Program for IMessel 1)ehris r such orcumstances (e.g., molt en core coolant micraction, Coolability increased hydrogen peneration, increased containment pressure). 'lhe following addresses reflooding research One of the most important phenomenologiod inuer. in r elating to the in vessel and ex. vessel parts of an accident. the progression of severe accidents after the reactor pres-sure vessel has failed is whether sufficient energy can bc .

removed frorn the discharged molten debris that the plant  !

2.3.2.4 Itcsurch Program for Itoflooding (inoessel) can be brought to a stable condition and the challenge to containment integrity, whether by basemat penetration

.the working hypothesis is that adding water would be '

or by containment overpressurization, is avoided 'lhe highly benehtial and hkely to terminate the accident at this stage. The fuel-niolant interactions associated with inost comrnonly available mechanism for removing en.

crgy from the ex vessel molten debris in 1,WR contain-such addition would be benign, yielding quench. with the pombility of hydrogen pmduction in a narrow time win' ments is water addition. Issues that must be resolved to '

develop debris coolability criteria are (1) the nature and dow wipen cladding is hot enouFh and still in a highly wnfiguration of debris. (2) the heat extraction mecha.

undistnbuted geomel'Y-n sm f rom debris by water, and (3) the molten debris-con-The current research approach to this issue is to review past investigations in which water (or its simulant) has in order to obtain data to support the development of .

been added to a degraded core ar'd to determine what the coolability criteria, an experimental program called adverse effects could be and what the current state of MACli was developed under the sponsorship of NI(C.

knowledge (data and analysis) suggests. An initial review  !! Pill, and several OliCD countries. 'Ihe program is in-of past experunents suggests that a few tests have been tended to determine the ability of water to cool molten performed as part of the col (A and PilF experimental core debris during MCCI and to enable characterization program as well as the 1 OIT-FP2 test. In the past, of the resulting debris foi assessment of permanent simulant tests of coohnt added to fuel debris were per- coolability, formed at llrookhaven Argonne, and UCl>A. Although limit ed in scope, these tests address water addition during in August 1989, a scoping test was conducted in the in vessel core degradation. 'the possible adverse effects M ACl! series of tests in which approximately 109 kg of should be understood and their impact minimind, Possi. UOrZrOrZr melt in a 30 cm x 30 cm x 15 cm pool ,

ble adverse effects include: interacting with limestone common simd concrete was flooded with a 50 em head of water. "Ihe melt decay heat

• llydrogen generation and fuel heatup from exother. was simulated by direct electric heating: however, opera-mic metal. water reactions, tional problems caused the input power to significantly  ;

exceed the decay heat (by a factor of about 3). The results e linergetic 1 Cls that may adversely affect attainment fmm the test indicate heat transfer rates of 2.4-3.5 of a stable coolable state. MW/m2 duting the imHal period of water interactton. lhe rates decreased to 0.6 MW/m2 and down to 150 KW/m2 The major variables for determining w hethc r water adds.- for the more quiescent period.flhe results support the tion would have adverse effects would be the rate and concept of a thickening ecust with periodicaccess of water l characterof wateraddition and the state of the fuel at the - to the melt and partial melt quenching, time it is added. To help in focusing this wor k, qualitative A second M ACl! test, conducted in November 1991, em-scenarios of the accident were developed with reference ployed approximately 430 kg of UOrZrOrZr-to water availability and its effect. On the basis of qualita-Fe2 Orca 0 malt rnixture in a test configuration of 50 cm tively different fuel-coolant contact configurations, the

- fuel states are: x 50 cm x 25 cm. As in the scoping test, the melt pool had an overlying water with 50 cm head and an underbed of i

limestone-common sand concrete. The test was termi-l 1. Initial heatup and core degradation (rods intact),

nated after 25 cm of concrete ablation. The results indi- t i

cate a heat transfer rate of I MW/m2 during the initial

2. Advanced core depradation (core rubble, melt, and period down to about 3040 KW/m2 during t he quiescent rehication). period. 'lhe results further indicate formation of a sus-pended crust thereby preventing continuous melt in order to better understand the accident management quenching. The molten material underneath the crust issues for this stage of the accident, we are collecting ablated 25 ein of concrete at which point the test was ava lable information on core degradation morphologies terminated. Itased on the post test analysis of the second

• 29 NU1111G-1365, Rev. 'l e

., -n+w----,v,.nm.c-wv.- ,an..,,-n.- n,.--,_,-, ~,-,n --

n ,e, , ,n,.,.w + , , , . , ~ , , . , - - , , , .-.,--,,,.--nn, -..en ,,,,,v

! 2 Research Plan ,

i M ACli test, it w as concluded that the initial and boundary power was cut of f to let the debris sohdify and cool. In the l conditions of the test were not prototypic. 'lhus, a third second stage, water was added befor e pow er was apphed i

test was performed m Apal 1992 with corrected mitial to the new quenthed meltpmi configuration. Power was and boundary conditions, but otherwise in an identical then increased until rnelling and concrete ablation were conbrutation. Prehminary results from the test indicate re-established. 'lhe results indicated an imtial period of '

some degree of coolability although any firm conclusion vigorous melt. water instabdity followed by a stab!c crust must await further analysis of the test data. geometry with substantiauy teduced rates of encrpy trans-ler.

Currently, three more tests are planned m the M ACli program. These tests wdi use variable depth of Results from the limited number of tests described above '

UOrZro Zrr melt mixture in twoddferent geometrical led the experts to believe that the debris coolability (a '

conf gurations (50 cm x 50 cm x 75 cm) interacting with coupled 1:Cl and MCCI phenomenon) may be an ex-hrnestone-common sand concrete and siliceous concrete. tremely complex process that demands a more careful -

Water will be added shortly after the MCCI has started. examination of various factors contributing to the proc- l

'the objective is to measure the rate of heat removal by ess. 'this pase rise to the concept of "rnorphological" water; study the melt cooling pmeess, includ ng crust testing, w hich deals with the determination of debris mor-formation, stability, and growth; and to investigate the phology as a function of experimental variables.'lhe first effect of various parameters (e.g., debris depth cavay of such morphological tests currently in the planning geometry, power densny, conet ele composition) on avail. stage involves a prototypic charpe material on a nonreat- i i

alnlity. tive (MgO) basemat, Concurrent with the MACl! program and under the 2.3.3 Anticipated itestilts sponsorship of NRC, a second experimental program on debris coolability was initiated at the Sandia National The results of this research will identify the basic vati-Iaboratoriesin February 1991Jihe goalsof this program, ables governing heat transfer and hydrodynamics of melt.

called WiflCOlt, are to complement and augment the water interaction, meluding the effect of water mjection -

on debris configuration. Models supplemented by appro-MACil program and to provide a means to support the assessment of existing as well as ALWit designs. 'Ihe priate cor relation to address all possible phases described above would be validated with available and planned ex-WiiTCOR program is designed to nddress two specific issues: (1) the comparative coolabilities of oxidic and me. perimental data. 'lhese models then can be used to assess tallie debris and (2) the limits of debris cuolabilhy in terms the potential for and consequences of fuel coolant inter-actions, assess the efficacy of accident management of debris composition and depth.

strategies, assess the effect of 1 Cl on altering accident in the first Wi!TCOR test (Wisi'COR-1) run in Septem. sequences, and provide estimates for the rate of steam ber 1991,35 kg of charge matenal 80 w/o Al 230 -20 w/o and hydrogen generation following reflooding by water.

Ca0 was heated to melting at 1850*K within a 32 em diameter tungsten annulus. The charge and the annulus 2.4 SCVere ACCi(ICill CO(ICS were surrounded by a cylindriced MgO crucible and sup-ported on a 40-cm diameter by 40 em high concrete llecause of the difficulty in performing prototypic experi-basemat. After 2 to 3 cm of concrete ablation, the rnotten ments and the vancty of scenarios possible, substantial debris was Thuled with approximately 25 cm head of. ' reliance must be placed on the development and valida.

water at a rate of 601pm. The results indicate approxi- tion of complex computer codes for analyzing severe malcly 5 to 6 cm of concrete ablation after 30 minutes of accident phenomena or planning accident management-water addition. 'lhe results further indicate formation of strategies. A number of codes (cG., SCDAp/ Rill.AP5, relatively thick and layered crust. and partial melt MiiLCOR, CONTAIN, CORCON, COMMIX, quenching with voids between crust and quenched melt. Hi!CI'R, ' MiiLPROG/filAC, VICFORI A, . and No evidence of fragmentation is apparent from the post llWRSAR) have been developed for various stages in test examination of crust and froren debris. sesete accidents, both in-vessel and ex vessel, for both PWRs and ilWRs. Ilowever, as a result of a review of A second test, c:dled Wl?TMITI', way performed in De- NRC-supported codes and associated documentation, cember 1991 to evaluate the potential for prolonged mol- the staff has terminated support for three severe accident ten debris-water interactions which might result in bulk analysis codes, lil!CMt. Ml!!, PROG, and ilWRSAR.

frec7ing or quenching of the debrisJihe charge material llWR-specific models developed under llWRS AR spon-for the test was 92 kg of 304 stainless steel and the _ sorship are being incorporated into Mill.COR and basemat was made of limestone concrete. The test was SCDAP/RiiLAP$. 'the liliClR and llMS codes were executed in two stages. In the first stage, the metal debris developed to test models of hydrogen rnixing and combus-was heated to melting at 1990k ablation began and water tion within reactor containments. Mill COR and CON-was added at 50 lpm for approximately I hour before the - TAIN have incorporated the models that were part of NU RIG 1365, Rev.1 30

-- _ __ _ m__._ _ . _ _ _ . . - _ _ ___ __ , _ _ . . _ . - . _ _ , -

_.m_ _ ._ _ _ - _ _ _ _ . _ . _ _ _ _ _ _ _ _ _ _ _ _

t i

3 2 ltesearch Plan v

tilf!R for assessmg hydrogen challenges from severe and vahdation activnics, insights and results of separate i acadents. 'the llMS ade is being documented, and no effects and integral experiments, and user needs to sup-l fur 1her descloprnent n planned. Suppori for Mlil . Pit 00 por t r esolution of various severe accident issues, and acci-dent management. Since 1984, the ode has been used to was ter minated on the tw.i', that it would be dupheative of SCDAP/iti:1.APS ar, a detmled mechanistic in-venel se- support the analysis of several major severe accident ed I vere accident c de, however, the II:Cl module will be periments such as PilF SI D l-3/1-4,1.0FI' IT-2,  :

extracted from Mlill'ItOG to be used for fuel.wolant ACitR dim, and Coll A experiments. 'lhe code has  !

in t e r actions analyses. SCD A P/Illil .A PS is a less detailed, been used for severe accident analyses, includmg natural l more flexible ude and has been validated with consider. circulation studies and the analysis of lower plenum de-able esperimental dataJihe development and documen- bris and lower head heatup. Accident management stud- i tation of the enginal integrated risk analysis e de, the ies have been perforrned which include the analysis of Sourcc Term Code Package (STCP) has heen cornpleted. strategies or phenomena that minimi/c direct amtnin-

!. 'lhe KI CP a udlection of various axles with the

, ment heating, -

M AltCil ode servmg as the cornerstone, wdl no longer receive developmental or maintenance suppmt. Systematic assenment of the newest version of the ude

[Lc., MOD 3(Hx)] wdl be performed to define modeling

.i.he development of severe accident phenomenological uncertainties in important severe accident phenornena.

mqlcis and cornputer ctdes continues to have a role si 'Ihese uncertainties, along with uncertainties in system achievmg the objectives of the severe accident research thermal hydraulics from natural circulatitm experiments progr.un. the nde.s now under NitC's sponsorship and and design basis accident (DilA) experiments, will be support are SCDAP/itlil.AP5, Mlil 0011, and CO_ N-I AIN. In addition, several other axles such as VIUl'O' propagated through phmt and accident 4pecific condi- i tions to assess the uncertainties in conditions of ItCS .

RI A ( OMMIX,and11-( lare beingdeveloped and main' failure caused by natural circulation, a damaged core, and tamed to lyerform specific functions that require detailed hnion product and hydrogen releases during a severe modeling these codes will be used to benchmarking the accident.Somelate phasecore melt relatedexperirnents system level codes discussed carfier. Also, the core con-wdl be planned and performed as discussed in Section 2.2. '

crete mteractnyn code, i.e., COltCON, wdl be incorpo' Once these experiments have been performed, models to rated into the( ONTAIN and Miil COR codes. I he rela- treat late-phase core melt progrenion that are not ade.

tionship of these codes to various severe accident quately modeled in the code will be developed, and sys-progression phenomena is shown in Figure 2A.l. Addr.

tematic assessments will be performed, tional discussion on the status of each of these codes rs ,

provided below. '

Preliminary assessment of earlier versions of SCDAP/

2,4,1 SCI)Al'/itELAl'5 Illil AP5 that began in 1991 have identified the relative strengths and weaknewes of many imdels in the code. In 2A.1.1 Ilesearch Needs general, many predicted early-phase phenomena, such as system temperature, fission pnduct release, and the m-Since the 1979 Three Mdc Island-2 (1 Mi-2) accident, itial change in the core geometry caused by bathioning

-the NitC has conducted a broad based research program and melting of core structures, have been within the ex. .

to develop an understanding of severe accident behavior, perimental uncertainties for the individual experiments. l 7

As far as practicable, much of this understanving, from Assessment results have also revealed several areas that initial core uncovery through reactor vessel failure, is will require model improvements. Specific areas include contained in the SCDAP/Ill! LAP 5 computer cale. (1) renewed bundle heating, rnelting, hydrogen pnwiuc-SCD AP/Illil AP5 is designed to model the coupled inter- tion. and fission pnduct release during refhod,(2) initial actions that occur hetween the reactor coolant system rehication of molten material as droplets and rivulets,(3) -

(itCS), the core, and the fission products during a severe interaction between fuel nd cladding and loconel grid accident; spacers, and (4) the diversion of flow from damaged fuel =i assembles. Itemoving these modeling deficiencies will-

'Ihc objective of this research program is to develop an significantly reduce the calculated phenomenological un-analytical tool (i.e., the SCDAP/Riii.AP5 code)Ihat wdl 'rtainties in the early phase core melt progression; As a provide the NRC with the capability to perform detailed ,esult, the end state ofin-vessel core melt progression can -

analpes of in-vessel core melt progression phenomcaa be more accurately determined. 'lhis in turn should con-dunng various severe accident conditions and scenarios, tribute to a better estimate of containment huding, fis-sion pmduct release, and perceived overall risk for severe 2A.l.2 Current itesearch Program - accidents. Thus, efforts devoted to removing these code deficiencies are the main focus of current SCDAP/  ;

'lhe development of SCDAP/Riil AP5 has been bae sd Hill APS research aetivities. Other research activitiesin-on a combination of factors, including model assessment clude (1) incorporation of ORNL-developed llWR model 31 N Ult!!G-1365, Itev,1 tf'61--'% -- +-- - --%.m>e+--ay*ah*y'==-aw + ty w-s--wj , eaiw-m umry z91e-+= yv--k-*-f--7w M T *M- t4de'*t-'F* Pr' ?I'-F==C**T' OT'-'-m-P1"t g' " ' * *1't97*-d'9*E T'T'*-9dae7WS " "='

"fl1 F- '""'?

2

• C 3 x x

! M 3

al ~

a 5

a

,ta 4 4 x. '

! o I

i i Tier 1: INTEGRATED CODES 4

l j

MELCOR (2nd Generation)

. .n ;

.. s.

c.

- .a ym,, , ,y. .- - . - .. - - . . , .

. = - - . , , , -

, -- x p I S .".' . . .. %[.Md of-. i*_ . - 4 N [._". k E.. . . . - E  ?. d .- A

. ., _ . . - hk.t y qwg p,' , f ' f4 -s t

5 c

t'

! E Tier 2: DETAILED MECHANISTIC CODES i z

! x I l C w g COMMIX- DEBRIS VICTORIA CORCON (MOD 3) HMS BURN u

a

a .

E c

3

!-  ; SCDAP/RELAPS (MOD)3 '

CONTAIN '

MACCS l E t

, a i t

i i

The8vstal Ceve , FP phtmase FP T premmeno q, Osy,Conemme fhamson Wome ' FP Trmepset Comeminssenet Cupemempsi or ce ene l *ererosmee, is a :

sroen meer anacs- w s Femur, timersariani ' - Fees one m to - tm.e - n- e i t

C. ._.

! PROGRESSION OF ACCIDENT PHENOMENA

?

r l

l \

l i

1 i

i r

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

2 Research Plan i improvernents for current generation llWit as well as for The CONTAIN axle rmdels intercell flow, hydrogen and Ihe advanced SilWH,(2)SCDAp/Riii.AP5 peer eview, carlxm monoxide combustion, heat and man transfer (ra-and (3) nuxlel catensitm to treat Westinghouse AlWXI diation. convection. umduction), a erosol be havior, fission  ;

core desip changes, including core structures and mate. pnsluct hebavior (decay heating, transpor1), engineered l tials that thifer from current PWR designsJlhe purposts safety systems (sprays. fan coolers, ice condensers),

of the SCDAP/Illil AP$ peer review effort are (1) to llWR4pecific systems (suppreuion pools and safety re- >

provide an independent, high quality review of the cale, lief valve diaharge), mre concrete interaction, and sim-(2) to help the NitC determine future axle development ple treatment of direct containment heating. 'the cale

ducction, cifort, and priority,(3) to provide information provides the capabihty to analy/c a wide variety of I.WR concer ning ude deficiencies and limitations that need to plants and accident scenarios be improved or corrected, and (4) to determine the dc.

gree of technical adequacy of the ude.

One of the safety issues that is currently under extensive As for the late phase of severe accidents (e.g., from the inve%adon @ect containment heating (DCll) and formation of hundle si/c bkickagesIhrough tbe growth of P* "#"U"" " I '"'I"' "" "**"I "*"8b"" D molten pools and the rehication of rootten rnaterialinto nmhen wre materilds ejected following the lower head fa e of the mactor nel unk pmsm M in.

the lower head). the core melt progression phenomena  ;

are still pxnly understo(d because of the lack of experi-m a lap army of mmp x pnwem, many o whd

""""'Y""* """ "E " " " * """" * .

mentaldata. AsIhelate phasecorernett progressiondata w no expannental data pmvloudy existed.

become available, maleis will be developed and incorpo-rated into the cale. Modeling assessments against experi- ,

mental results have been made (see Table A.2.1). Addi- Several research activities are ongoing in an attempt to tional modeling nuessment and validation efforts will quantify the effects of DCll(See Section 2.1).111stori-continue to be made to ensure that SCDAP/Rl!!.AP5 cally, the NRC DCil experimental program was guided meets the code's design objectives and targeted applica- by calculations and sensitivity studies performed with the lions. COffl'AIN code. Although thesc sensitivity studies and pretest and pistsest predictions were useful in identifying important processes, there was concern over the degree 2.4.1.3 Anticipated Hesults of confidence in the code's treatment of the DCil pnic-Uncertainties in predicting vessel and RCS failure times esses to determine the relative importance of these proc. .

and veuel failure males using SCDAP/Riii.AP5 will be esses in the tests. In FY90, a comprehensive scaling meth-substantially reduced because of the extensive data base odolo;>y was developed and implemented for the purpose (including late phase experimental results) and model of ensuring a properly focused and technically defensibic

! improvement efforts. The end pnsluct of this research direct containment heating experimental program.This prognim is to provide NRC with a computer cale that is program will guide the development of matels for incor-capable of performing (1) plant (both PWR and llWR) poration into the CONTAIN cale.

analysis for the in. vessel core melt progression phenom.

ena for various severe accident scenarios and (2) experi.

mental analysis and support for m vessel severe acadent -

Another key research need is related to c4mtainmut analyses for A1.WR designs namely AP600 end SilWR experimentt Other uses of the ade mclu<le (1) assess-plants. Containment designs are being developed for ment of the efficmy of accident management strategies, (2) Mill 10R bcnchmarking and assessment (3) I'MI-2 Al.WRs that incorporate passive cooling and decay heat '

amoval features for protection against long-term con--

acadent evaluation, and (4) suppnt for Al.WR design certification as defined in Sl!CW91-161. tainment overpresure in accident situations. 'lhe passive  ;

nature of these containment systems poses unique chal- >

lenges to containment analysis oxles for predicting am-2.4.2. CONTAIN - tainment response in both design basis and beyond design basis events. Such 611enges which require new or im-2,4.2.1 - Hesearth Needs proved malcis include natural circulation air flow in the channel outside the containment shell, behavior of an -

The CONTAIN axic is a detailed mechanistic cale for evapomtive flowing water film on the containment shell, the integrated nnalysis of containment phenomena. It has stratification of gases within the containment, potentially been developed under NHC sponsorship to provide the . unique heat tnmsfer and condensate film behavior not capability to predict the physical, chemical, and radiologi- adequately represented by existing correlations, and nu.

- cal conditions inside n nuclear reactor containment in the merical challenges arising from the need to efficiently event of a severe reactor accident, as well as fission pnd- . perform long term containment respmse calculations. >

uct releases to the environment in the event of contain- Similarly, the performance of the passive containment ment failure. c uling system (PCCS)in the SilWR will be mateled.

33 NURl!O-1365, Rev.1 i

-, - n m.n , . , n - - , .n.-.,-r-.ne.--,_n.. - - - - - - - , - . , ,.- ..,n-n -., - - - , . .

i r

2 Researth Plan 2

)

2.4.2.2 Current Research Psogram 2,4.3 MELCOR in FY92 and i T93, selected DCil models wd! be incorpo 2.4.3.1 Ilescarth Nreds rated into the CONTAIN ude and evaluated against the related experirnental data. 'lhen, actual plant cases will llesearch needs for the hil!!COR code fallinto a nutn-he performed wit h the updated CONTAIN ade to deter- ber of areas that an generally be divided into develop-mine the impact of DCII on containment pressuritation. ment or awessment. The development areas include (1) adding calculational capabilities (or mcdels) into the ,

Another primary objective of the current rescaich pro- ud, (2) mnecung or improsing existing calculational 4

gram is to develop and validate models related to the #"U"U8 5"' '* * " ' ' ( I'Y# ' N" #" '

tors, and (4) improvmg cale documentation,, "lhe as Al..WR performance for both design basis accidents and severe accidents. 'lhe CONTAIN ude will be modified ment area ndudes O) compadng cakuMonal resub ,

accordmgly, and industry experimental programs will be th experimental data or output from other cale calcula-cvaluated and used to validate the CONTAIN code. ns, ( ) vaymg input paramders m chara h sensitivity of results for specific types of calcul@atio (3) determining reasons for differences between the re.

Other research activities include efforts to update the suits of calculations and experimental data.

CONTAIN code consistent with the improvemcats in modeling already developed under other programt Cur- A comprehensive independent review of the Millf0R rently, only iodine washout by containment sprays is imst- cale was undertaken by a peer review committee. 'the eled mechanistically in CONTAIN. 'the 'llt!!NDS code objectives of this review were to (l) provide an independ- '

developed at ORNI,is able to provide mechanistic mod- ent assessment of the hil!!f0R code,(2) determine the eling of iodme chemistry and ether fission product physi- t ech nical adequacy of the cale and (3) issue a iinal report cal and chemical processes. 'lherefore, incorporation of describing the technical findings.'lhe review comtnittee this package into the CONI'A!N code would further ad- recommended improvements in five areas: (1) Ml!! f0R var ce the overall capability to predict fission product numerics, (2) inissing m(dels, (3) revisions to existing release constituents and fission product spatial heating matels, (4) cxpanded assessment, and (5) documentation.

effects. This effort will be completed in IT93. Core-con- 'lhey concluded that when these recommendations are crete interaction phenomena modeled in CONTAIN are satisfied, Miilf0R will be technically adequate for PRA based on the CORCON-MOD 2 code. Release of the applications, although the code may not always be sufft-CORCON-MOD 3 cmic, which includes the irnprove- cient for some parametric accident management studies, ments accumulated over approximately 6 years, necessi- 'the peer review committee issued its finding in 1992

~

tales the u}xlate of the corresponding CONTAIN model; (report 1.A-12240).

this effort will be completed in FY93. Also, Figure 2.4.2 illustrates past and'present validation and assessment 2A.3.2 Current Research program activities to_ compare experimental tests against cale model predictions. *lhe current focus of the Ml!!f0R research efforts are on (1) resolving time step size and machine type depend-As a result ofincorporating the various selected models in encies (numerics problems) as urged by the peer review.

CONTAIN, the code manual will be uldated. Also, to ers, (2) correcting inadequacies in some of the phenome-ensure the CONTAIN code meets its design objectives nological models that were pointed out by the peer review

• and need for regulatory application, a peer review will be committee, (3) addmg missmg matels "nto MI!!f0R, .

undenaken by independent experts. The peer review will and (4) expanding the techmcal assessment prograrn to

, j begin in CY93, and will be completed in about 12 months provide assurances t hat the code is reliable, or, if deficien. i after initiation. cies are found, to identify those deficiencies so that reme.

dial actions may be taken.

. 2A.2.3 Anticipated itesults in regard to the five areas of recommendations by the-peer reviewers the following actions are planned.

As a result of upgrading the CONTAIN cale, actual plant analyses will be able to be performed to determine the 2A.3.2.1 MEl.COR Numericr Efforts have been directed impact of DCil on containment hiading. Also, the cale toward investigating problems and sensitivities associated -

will be capable of performing ALWR containment analy- with cale numerics including time-step and machine type ses for both design basis accidents and severe accidents dependencies, identifying their underlying causes, and for the Ap600 and SilWR plants. In concert with these climinating or mitigating them. lifforts to remedy the efforts, the CONTAIN peer review will provide an over- numerics problems started in FY9! and those problems all independent assessment of the code. 'lhis assessment identified through 1991 have been settled; others are L will assist in steering future efforts on this program. being investigated in FY92 and FY93.

NURiiG-1365 Rev.1 34

-+

J

'CONTAIN Validation and Assessment Strategy 7 3 j p p;3. p e a g. ...

i

__ .e . > ,

.t.-- a 4 3 e ... .. ,m .. a . 4 ,

i siddET iga.,;;:hisF6 h em :periittents Ek a

aa%#asmai!!FM% . . .

m"d?dinei4.m..BMRsk,%,s

1. amu mmma u

' Surc SNL and ANL i 0

NTS tests and HDR T31.5 -WPCCS LACE ABCOVEi PNLIce HDR E11.2 d Beta DCH tests TM-2 analysis l and T31.6 tests LA-4 ABS & AB7 Condenser and E11.4 1 Wetcor g:

z a, .

c3 n

w w

a a

.. e_ .

E a

I CORCON DCH HECTR k Standalone ScaEng Standalone y,

'g Va&dation Models Validation

N i

?

B g 1r. _ . 1r y y V V Core Concmte Direct Contain- Hydi g Heat & Mass Two Phase Aerosol EngrSafety integral

, Interactions ment Heating . Buming Transfer T-H and Flow Behawor Features Analysis C

~

N bhifjif"  % im e2Y a w y k u;dFQQ5i$Tuy;Q4;.liUy>~a

.1 xm e;:v - , +-; ::WiL~ p% LE? y f ^&

~ a.~sd ~

r x

.m

- g;!! e } w jlyacwn .

mg .: _A; hs hx .u m,~::eg~.MCON'"TAIN:

v v .

W>M%w@;.fa:-y.m.&&

?ut.;&.:-u

..n%

1w

-t!" W w=0 E n

< y

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

- 2 Ilescarch Plan r

2.43.2.2 Mming Moddr 'lhe review committee con. h1111 full's fission product dep>sition rmdels are cluded that match, for the following phenornena should adapted from the h1A!!!(OS containment model. As be given the highest priority for incorpiration in hiEl, such, certain processes that are not generally important in ,

C011: the containment have been neglected. 'lhese include irn-paction and turbulent deposition of actosols. lixperimen-  !

e PWit primary system natural circulation in comps tal data on containment bypass kequences performed for nents with countercurrent flom , the !!!cetric Power 1(esearch Institute, as well as calcula-tions using rnore comprehensive acrosol deposition mod.

• high pressure melt ejection and DCll, els,indicroc that the neglect of these processes mayresult in a kignificant underestimate of the retention of herosols

• ice condenser containments, in toe primary system, especially for low pressure sc.

e quences in which gas vehicitics are high Cortclations nonexplosive interactions hetween dehn.s and water, which har been developed to represent these effects are i e fission prmluct vapir scrubbing, included in the VICIOltlA code and consideration will  !

be given to adopting those or simpler versions in the o additional reactor coolant system fissmn product hilif fOlt code.

deposition processes,and Chemical reactions betweer. settled acrosols and vapors o fission product reactions with surfaces. and heat sinks in ib prtnary system can greatly affect deposition (chemisorption)and revaporization rates.'the ,

hiodel development activities were initiated in FY91 for reviewers state the lack of explicit modeling in the code ,

three missing inodels: natural circulation, dit ect contain, sequences and is particularly scri-ment heating, and ice condenser containment. Work on applies ous for cesmm to all hydroxi accident,de and tellurium compound the remaining models dealing with fission pralucts is ple nmdels for these effects which capture the most im-being integrated into the current wor kscope.'!he effort to ortant effects dunng accident sequences, as deterrnined incorporate a model into hil!LCOlt for handling direct

{y the dctailed VIC IOlll A code, wdl be considered for hil!!L,Olt in the f uture.

containment heating phenomena was tmed on mateling insights derived from the DCil experimental and analyti, ggg gg l g , .the Committec cal programs developed through interactions het ween the experimental group and the code developers. Incorpora- recommended that the followinE concerns with existing tion of appropriate DCil modeling into hi!!!folt was hie. lm, . it nudels be given the highest priority.

completed in FY92, Completion of the implementation e Condensation is treated independently in the codes of a relatively fast runnmg model for calculating natural circulation that is consistent with the h1EllOlt architee. hydrodynamic behavior modeling from those calcu.

lations of acrosol particle growth and deposition in t ry and execution time requirements is scheduled for the radionuclide inodeling portion of the code.The validity of this approach should be demonstrated by i comparison with more exact rmdcls or data. i

• ihree fission product behasior matels identified as miss- ,

ing by the hiliffolt peer review committee, are being e laconsistencies in the' treatment of chemical reac. j given consideration for in&. ion in the source term calcu- tions between col (CON and VAN!!SA should be '

Imian, especially for PI(A use, 'Itcsc are described as resolved, and tinprovements should be made to the follows: Although hililCOlt has a relatively sophisti- CollCON/hiOD2 phase diagmms, cated nudet for scrubbing acrosols in a water pool,in some sequences gas temperatures in the primary system * 'lhe matel for condensation in containment (mass may be sufficiently high that volatile fission product va- tnmsfer) should be revised, pors, rather than acrosols, would be discharged into pools, h1El Colt does not represent the removal of these e' 'there seems to be a disagreement between the pool -

vapors as they condense to acrosols and attempt to pass decontamination factors computed with the current these acrosols through the pool,lixamples of sequences pool scrubbing model and those calculations made in which this process is important include low pressure using other codes. A general feeling is that there llWil sequences with discharge of vapor through the may be an implementation error, however, no'such safety relief valve lines to the suppression pool or low- crror has been located.

pressure PWit containment bypats sequences with dis-charge to a water pool in the auxiliary building. In such flhe acceptability of using the hydrmlynamics pacLage sequences, this phenomenon could dominate the calcula- water condensation / evaporation matel for calculations tion of the source term, especially when other fission of acrosol particle growth and deposition in the radionu-l prmluct removal mechanisms are weak - clide package is being addressed.

l NUl(EGil365, llev.1 - 36

= _

h-,.-- ..aiy.-g-i -..g --m- g ppe .g,--me-- e amyn e-. , - + - - 9 --+w --.-y-.yrg ig--y9p.-yn-9., p. 4g9 pr,-.-i -- -' 'r-

• r-T / i

l 2 ltestanh Plan impmvemems in the midic and rnetalhe phase durrams plant analyses. 'lhe work in llus area is intended to be are needed because the eff etts of cutectic fermation on combmed v.ith work on other plant geometties (e.g.,

debrn layer meltmg and sohdihcatmn behanor are nn. II AW PWRs. IlWRs) for i.imilar sequences m both ex-portant. Current incompahbthties m material properties perimental analyses (usmg hilS i and l'IST test f aalities, between the cavity rnodchng package and other hill from Semiscale.si/cd equivalent facilities for those other COR pac kages are being eliminated. and duphcatne and geometnes) and for full-sequence demons.tration analy-inconsistent themistry between CORCON and ses for other plant types. Demonstration calculations to VANI .SA are being chminated by the implementation of analy/c ll&W and Combustion linginecting PWits are CORCON/hiOI)3 into hil:1('OR. hemg run at Ilrookhaven National I aboratory.

'the nulel for uindensation in the presence of noncon- In addition, the NRC is in the process of initiating an densables n bemn revised to address identified deficien- international cooperatne cifort for technical assessment eies, mcludmg modeling of the film thermal resntance of the hilil COR coJe, the hilil COR Code Assessment and high mass transfcr effects. Progr am (htCAP).The objective is to uccelerate the tech-nical assessment, consistent with the peer ieviewers com-The pool scrubbing model discrepancy w di be resolved. ments by employing the expertise of rnany of the code users both mside and outside the U.S. 'ihus. many cases 2.4.3.2.4 l#anded Auc.umon: 'lhe committee also con, can be run in a shorter timeJihis will also allow expansion cluded that the abihty of hilil,COR to calculate severe of the user community and at the same time improve the accident phenomena has not been demonsteated suffi, understandmps and abilities of the users to run the hill COR code, ciently. Such a demonstration thould be based on (1) sensitivity studies, (2) benchmarking activities usmg ex-perimental data, an'a (3) code.to-code assessments. 2.4.3.2.5 Documentation 'lhe body of existing hilil .COR documentation is sigmficant. Ilowever, the Committee A plan for a more comprehensive integral assessment of fell that detailed descriptions of the models and correla-hilil.COR a under deselopment. Tne technical assess, tions were lacking in some cases, and documentation on ment planning involves a number of related actinties, the applicabihty and benchtnarking was cither inadequate mtluding, venfication, validation, and quantiheation of or rnnsmg. 'lhe Committee also recommended that a uncertainties.'Ihis prosess involves r eviewing moJels and process for collecting, documenting, and distributing user comparing analytical results to experimental data, incluJ. guidelines to the hil!!.COR user community be devel-ing small scale and full scale expenments. Only a small oped.

portion of the comprehensive plan has been accom-plished so far, it is a high priority to have at least some Irnprovements to the hilil COR documentation are validation lesults as soon as possible for each of the major planne llowever, documentation equivalent to the phenomena treated by the code. 'lhere is a need to de. TR AL

• I todels and Correlations" code document would velop a standard code package demonstration / test prob- be highly resource intensive imd somewhat duplicative of lem for many of the h1111.COR mde packages. At least information aheady available in the hilit.COR manuals.

one assessment problem for each major package or for Nevertheless, in 1 Y93 some resources are directed to each major phenomenological as ca will be prepared (e.g., upgrading reference manuals to cover the most critical the ice condenser model being added to the pencral llS needs. Further, there is a task to develop a practical user package). Also, the program is being e2uefutly designed to guidebook for hil!LCOR users.

provide more substantive assessment of the COR and RN packages. which wcre major areas of concern to the peer 2.4.3.2.6 Addniomd Rocarch reviewers A graphic dennption of the assessment pro-gram comparing analytical results and experimental data Not all of the code development and assessment work is a

• is given in liigure 2.4.3 which relate code phenomena tesult of the peer review. Other research activities in.

bemg validated to applicable experiments. Additional as- clude: (1) incorporation of ORNI-developed llWR sessment activities involve code to code comparisons and model improvements,(2) incorporation of an upgrade of analysis and evaluation of full plant transient calculations the CORSOR fission paduct release model including for various plant and accident sequences- use of the llooth model,(3) improving core debris reloca-tion moJeling to include spreading,(4) improvement of Consistent with the draft assessment plan, activities con- core melt modeling to allow simple material (eutectic) centrate on PR A risk dominant sequences, such as sta- interactions to be treated on a parametne basis, (5) fur-tion blackout, steam renerator tube rupturc, and V- ther expansion of the techrucal assessment program to sequences. 'lhe Semiscale test selected for prirnary involve other DOli laboratories engaged in severe acci-systems thermal.hydraulie and heat transfer assessment dent research for the NRC, and (6) testing of an input complement and support these selected demonstration model for AP6(XL 37 NURiiG-1365, Rev. I

2 Research I'lan g

g, f.5 5e > 5N CC 34

,W h

<y '#

M L 0 5 3I 8?

4*

ed > c. .

, g

• 2el 9R

m. i I

$5k g

$ ED A A 7 E h b$  !!

"s m

X,N a

il! ni "I E t-W $

dD$ > E li 5 y ,*

ENy l'igure 2.4.3 MIILCOR Code Validation and Assessment NURl!G-1365. Rev.1 38 l

i

= m KgO _

h R . :

n. _

M ~

$ E E CS I N 6 LOE - -

7 1

N N 1

1 6 1

j t n

e "=e  ;-

- F O r E A ' &- _

r F gI

- C $ i nS aE

. s;' 7

< t 2 _

n - n,A . _

N o _

C _

- M -- g_3 , ,M _

F E 4 -

A Mi xaf,

- C -

W A A -

- L L r 4*~ -

l

$ o o N

l s v _

^

o r h e { .,X =.

e e 7

• Y

=

,.3 _

- A S a

e, -

k E V S S 7 M.

7mg i.

.i ~. :

d

~. .

m O C B B 9 _4 1 .;

. B A A A A n

. t* ;s n

ei o

fg N

t g s o u dr b f

t. , g

.a wy x ww y m "j_.'.. .

u.

=

l' A N

O 2 H o C

":,r .-..M-

_A MF r-E D

_7W4 A$

~ .j, M

r.

v w -

H

C h'

t

'b n3

~ -

Y 2 -

D E

-[

1-.

1. g :m

=

m b C u2 B F N

v t-p' w e

l q9.% ds ., .s y .

a 3 j' -

p .

t n

e .s..

(

r t

s i mH n

o/T m.c. ,y:.

.n; ^:s.3 t t

u. - r s 9 n ue u et 1 o -

2 f

0 g ._

e kn g 1 C S., - - . 4r M=

t n t a a n5 >

ys. . .:.z, i

a B rF ix 2 w e v M t s.

w, e n r

c s y . ~m &,n-

.p a

- y ) nfio t-f:~ , Jp,:

L.

b 3 )9 o e c C o 2 2 e

-- N noi no $ - r e te 1

.dA

,s

_i l

h R 4 P P O E D r oi n H V4 (1 5s t lota

,Qs

t. C R d d C m 3.: ,.-

5 (2 O nl I* d.- -

L o cV i

t 4' V g-

$ 3 T T 1

1 C t S

~g- r ac: NuL D5C.aeE y -

2 C =, O' w7w. % o<-

2 Itestarth Plan Matenal mteractum inodels for the msore repon will be 2.4.4 COMMIX incorporated into Ml!!fOit. ()RNI has arranted the proprammmg of the llWitS All cutet tics imdel for mcor- 2.4.4.1 Itestarth Nt ds potation into hil!!TOlt This model is lunned to llWit in the past scars, research m the area of c(de develop-apphcanons and to lower plenum debus bed behavior, rnent for re' actor safety analpis was mainly focused on developing and validating one-dimensional system codes in another important developmental area, work is ongo' such as S( 1)AP/iti!! AP5 and Miil Colt. Ilowever, a ing and planned for the evolutionary light water reactor number of phenomena encountered in postulated severe (li!. Wit)and Al, Wit appheations MILifult iscurrently accidents are inherently multidimensional in nature, capable of handhng evaluations of transients in lil Wit 'lypiced examples are: natural circulation, thiw stratihca-primary systerns and containments. It is alw capable of tion, countercurrent !!ow in a pipe, hydrogen distribution treating most f eatures of Al Wits. Mill Colt will be and mixing processc% and the effect of noncondensible benchmarked aramst results of detaded codes to deter- gas distobution on local condensation and evaporation.

mine its adequacy and areas of development needed for Al . Wit applications. In IT92, an activity is under way at 'lhe unique features of the passive containment cooling Ilrookhaven to develop and test an input model for the system in the proposed Westinghouse APtdK) plant are Westinghouse APidK) design. Snmlar actmty is under way spectheally designed to prevent damage to the contain-at OltNI, for MlWit lined on the results of the input inent during design-basis events or severe accidents.'lhe model elfort and on imtial r"ns with the model, additional passn e coohng iunctions are carried out via natural circu.

itconunendations wdl be developed for further ode im- lation inside or outside the containment. 'the COMMIX provements to integrate Mlil Colt capatulities for wde (origmally developed for analysis of transient fluid Al . Wits. Ilow and heat transfer in the reactor coolant rystem)is being modified so that it can be used for the analysis and understanding of transient fluid flow and heat transfer 2.4.3.3 Antidpated itesults phenomena in the containment. 'lhe research needs for this program are to assess and evaluate the capability of Within the contest of the current research program, the COMMIX for use in analyzing the new and unique fea-results of the Ml!! Colt peer review, and the contmuin8 tures of AI. Wit plants during design basis events or se-plans for Mlitroit development and assessment, the vere accidents, and to apply the COMMIX c4de to per-antiapated results of the Mlil COR program are: form audit calculations.

1. 1)es clopment of a second-generation integrated de- 2.4.4.2 Current itesearth Program vere accident code that should (1) appropriately

,the main objectives of this program is to improve and nudel phenomena essential to the underttandmg of emnd COM rs capability and then in turn use COM-severe core damage accidents, (2) provide predie- ,

MIX to assess the adequacy of the unique passive contam-tions of the progression and consequences of severe ment cooling system designed specifically to prevent dam-coie damage accalents,(3) permit estimates of the age to the AlWicontaminent system dunngdesign basis uncertainties associated with such prediction % and events m severe acadents, (4) h:o e a stiucture that facilitates the incorporation of new or alternati e phenomenological nudels To satisfy the ebjective above, some coJe assessment based on the ongoing experimental research pro- effons ar'c bting carried out, and a numbei of new models I '" "' are being developed and implemented into COMMIX.

'lhe following is a list of new models, needed for the

2. Provide a code that includes major phenomenologi- analysis of the Westinghouse AP(dW) design:

cal developments from severe accident researth in adequate detail to address the phenomenology and 1. Multicomponent capability to compute the distribu-also has a practical runnmg time for severe accident tion of steam, air, and hydrogen throughout the analyses. reactor containment.

2. 1.iquid fdm tracking model to compute liquid film

3. Provide continuing maintenance and user >upport thickness, vehicity, and temperature on the internal for the Mlil Colt code. and external containment steel shell to calculate transient heat remosal from the /PldK) contain.

Within a few years, all the important models for severe ment. The APMM) steel shell consists of semi-accident analpes will have been added to the Mlit 001t elhptical domes at the top and bottom of a cylinder.

code and the code will have been auessed for its ade- 'the internal liquid fdm is formed as the result of quacy. condensation of steam on the liner wall of the steel N Ulti!G-1365, llev.1 40

- - . - - - - - - - - - - . - . - . - - - --~--- - - - -

l 2 Research Plan

[

r i

shellJlhe externalliquid film on the steel shcIl wall Canada (CHNL) indicate that, especially for PWRs, radi-  :

. is formed from water fimding at the top of the onuclide release from fuel retained in the vessel in an air t (k,me, ihaporation of this liquid fdtn then occurs as erwitonment will be radically different from release dur- l the buoyancy drhen air stream passes through the ing early stages of core degradation. Plant c(mfiguration annulus outside the steel shell. during shutdown situations may also lead to the possibility  !

of air ingress into the core cither by natural circulation or 4 3. lleat and mass transfer models which must be vali- from the residual heat removal systemJihe experimental dated with the Westinghouse AP600 Passive Con- data has to be obtained for the fuel release model of tainment Cooling System (PCCS) small cale and VICFORI A before one could estimate the effects on risk 1/84cale test data. associated with the radionuclide releases of these types. ,

4. Radiation model to account for heat transfer from Depmited radionuclides may be resuspended in the reae-the steel shell to the air baffle wall in the AP600 tot coolant system when the syr, tem is depressurized s design. either as a natural event of the accident or as a deliberate - -

measure to mitigate the possibility of direct containment The multicomponent capability has been implemented in heating. VICI'Oltl A has incorporated such a ruodel Cur-the COMMIX code, and a limited validation of this capa- rently, there are no suitable experimental data to validate i bility was carried out using steam blowdown data from a revaporitation model, but the PilEllUS-FP tests may full scale vessel in Germany (i.e., ilDR ISP-23). Further pr(wide some.

validation effort is planned.'the development of the lig-uid film traclJng model has been completed. Iloth the 2.4.5.2 Current Research Plan liquid film tracking model, and heat and mass transfer models implemented in COMMIX are being validated The VICI'ORIA code is developed under an interna-with the Westinghouse pCCS small scale data. 'the vali- tional collaborating efforts including representativCS dation effort with 1/8 scale test data will ram be carried from Sandia Nationallaboratories(SNL), Argonne Na-out. 'lhe need to develop a radiation model to treat the tional laboratory (ANL), Oak Ridge Nationa! labora-radiatian from the dry patch of the containment steel tory (ORNL), llattelle Columbus laboratory (IICL),

shell to the baffle wall will be made in FY93. Chalk River Nuclear laboratory (CRNL) and Winfrith Technology Centre (WTC). SNL is the principal devel-2.4.4.3 Anticipated Results oper of the code while other establishrnents are providing specific model(s)in their area of specialty.The code and a With upgrading of the COMMIX cale,_ it will be capable user's manual was first released in October 1990. An of performing containment analysis of ALWR for both updated version of the code has been completed in May design-basis events and severe accidents. Also, COM- 1992, and the revised user's manual and the code are MIX can provide multidimensionalinformation and serve scheduled for release by the end of lY92. The VICI O-as a benchmarking tool for the CONTAIN lumped pa- RIA code c4m now provide predictions adequate for reac.

rameter system code, for accident arolyses of:

~

- 2,4,5 VICTORIA 1. Radionuclide release during the early stages of core :

l degradation, l

2.4.5.1 Research Needs

2. Aerosol processes in the reactor coolant system, The analyses for NUREG-ll50 have shown that bypass 1, accidents, such as steam generator tube ruptures to be the 3. Vapor deposition in the reactor coolant system, dominant risk accidents for some classes of plants (sub-atmospheric pWRs and ice condenser PWRs). There is 4. Radionuclide release during later phases of core substantial uncertainty in the risk associated with these degradation, especially once the reactor vessel has accidents because the accident analysis tools available for failed, the NUREG-il50 analyses omit models critical to -

proper calculation of radionuclide retention and revapo- 5. Resuspension of deposited materials at the time of rization in these sequences. The accident analysis tools reactor coolant system depressurization, appear to underpredict retention in the reactor coolant _

system. The VICI ORIA code was developed to address 6. Radionuc!ide entrapment and revaporization in rup-this uncertainty related to the bypass accident. tured steam generator tube accidents and other by. ,

_. pass accidents.

Current state of the-art models of core degradation indi-cate that much of the reactorfuel will remain in the vessel Distinct classes of experimental data are being used to and the core region at the time of vessel failure. Tests in validate the various aspects of the code:

• 41 NUREG-1365, Rev.1

,.g-*-e,,,..m. y..g#g y-m,_,,,. ,ss- 9wpa w 7 ww-i.. g p .a . y -p y -astw , +7-y -,ww-,,y,.-,, w-yg,,w w w ,-my-we-,, y, -+,-9-y,- g g gg---_9 .gg. 9--aw 99 ,.,

2 f(esearch Plan

/ Rc/ rat.c Durmg Core Degradation ject, in addition, VIUl ORI A is also used for the interna-tional Standard Problem (INP) 34 exercise for the i AI -

Separate-ciftets tests such as the out+f pdc lit and VI CON fission product chemistry experiments in the under steam and hydrogen conditions tests done at Uruted Kingdom.

ORNI. and the ST-1,2 in pile tests have been used for mut h of the vahdation of VICI'ORI A.1 rench out-of pile 2.4.5.3 Anticipated itesults tests (Pill:llUS-SID) and in pile tests (PillillUS-l'P) . . .

with inadiated fuel will be examined in the future. More .lhe end product of this research is to prov de NIR, with a mtegral tests such as the PflF tests, the 1 OFlil'P test, computer code that malcis the radionuclide and non-and the l'I.Ill tests may be used for validation of the 'ddionuclide materials release, transport, and deposition release models once clearer portrayals of thermal.hy- within the reactor coolant system under severe accident drauhc processes are available for these tests. emiditions.it willbe capable of performmg(l) plant (both ,

PWit und ilWR) analysis for the in vessel fisuon product

2. Rr/ raw Dunm: Late l'hases of Core Degradation behavior for variour. severe accident scenarios, and (2) ex-perirr.cntai analysis and support for in vessel fission prod.

uct experiments. Specific uses of the code include There are no suitable test data to vahdate rnodels of ,

release once fuel has rnetted and the core geometry has (O Pill!IlUS-l'P post test samples analysis, and pre-and been lost. Integral experiments to study fission product post test analysis, and (2) FAl rON ISP-34 exercise.'the release from late phase core melt propression and rev. code is also being used by CitNI. for the lilowdown Test aporvation are not bemg planned in the United States. l~ cthty pre and post tests analysis and by Wi~C for as-Ilowever.1he Pill!IlUS-1 P project could provide some sessment of full plant behavior under severe accident of the needed experimental data. conditions.

3 Trurnport in the Hractor Coolant System 2,4,6 Integrnted Vuel Coolunt Interaction Results of the Maruken and I.ACl! tests are bcing used 2.4.6.1 Research #nds to validate the aerosol transport models in VielORIA. In the event of a ,evere accident leadmg to core melt, ltesults of the Argonne and Sandia NationalIahoratory molten fuel materials can come into contact with water, tests as well as tests now under way in the United King- producing a fuel coolant interaction (1 Cl).1:CI's can dom (WI'C)are being used to validate models of chemical occur for a varie;y of in vessel and ex-vessel conditions, processes alfecting radionuclide transpor t. More integral includmg: refhioJ of a partially molten core, melt pouring validation will be proviJed by results of the PIIl!IlUS-l:P into the lower p enum, and melt pouring out of the vessel tests.

lower head into a water fdled reactor cavity under low or high pressure. I1 these situations non-explosive or explo-1 Rcraporizatmn of Depmard Radmnuclides from the sjve i cg s may occur. The mode and resulting energetics Reactor Coolant System of I CI's depend on the complex interaction of various thermal, physical, and chemical processes, including:

These are not suitab!c data to validate the models of coarse mixing, particle fine fragmentation, and heat revaporitation. 'lhe Pill!IlUS-1 P tests may provide transfer.

some.

The goal of the Integrated Fuel Coolant In;craction 5 Release and Traraport during Shutdown or After (lI:Cl) effort is to provide a stand alone code that embod-l' enc /Fudure les an integrated models for these processes so that FCI severe accident events can be calculated for full scale Separate-effects tests done at CRNI. are available to plants. Presently the IFCI code has phenomenological vahdate some chemical aspects of models of release and models for the pre explosion mixing phase of FCI's, and transport in the reactor coolant system after vessel fail- includes multi phase, multi-dimensional, three fluid hy-ure. In addition separate-effects tests on fission pnduct dndynamic equations required to represent non explo-in the presence of air mil be performed at ORNI 'lhere sive events and the thermal detonation and expansion are no integral da ta on core degradation during t his phase phases of explosive I CI's. Mechanistic models for trig-of an accident. pers that mitiate explosive 1 C1 events are not included; however, there are provisions for representing intention-In FY92, developmental assessment will be completed ally imposed triggers, such as those in FCI experiments, and documented. A peer review of the code will be initi-ated in FY93; after peer review systematic assessment of 'lhe fundamental physics of FCI's are being experimen-the code will be carried out. Meanwhile, the code is used tally investigated as desenbed in Section 2.3 of this docu-for the planning of post-test samples analysis and pre-test ment. 'lhis work is largely focused on understandmg trig-calculations for mtegral test for the PliliitUS-FP pro- gets and how they affect l'CI's, and the development of NURlIG-1365, Rev.1 42

2 Research Plan separate effects models.'there is a need for a code that by chemical interactions with the melt pnor to reaching will emledy the models that resuh from this work and the stect's melting temperature. I urther examination of others as our understanding of 1:CFs incicase. 'this need this issue indicated the *e dissolution ~nd the data base stems from the fact that realistk modchng of I CI's re- cited by the peer reviewer are not applicable to this issue.

quir es that the (omplex interactions between the govern- 11 was also noted by several peer reviewers that the ing processes be captured as well as the processes them- strength of the steel liner would be gredy rehced at sel<cs. 't he code is intended to fulfill the need for a tool to elevated te mperatur es. 'lhis, together with pressuritation quantitatively describe non explosive FC1's, and explo- of the drywell following vessel failute might produce a sive events when the toggermg mechanism is known. creep rupture failure of the containment liner prior to Such a pr edictive tool is needed to help answer questions reaching its melting temperature. An NRC contractor is regardmg the impact of 1 CI's for present and future currently performing a structural analysis of the liner nuclear power plants, under these conditions to determine if creep-rupture fail-ure will prece 'e liner fadure by melting.

2.4.6.2 Currcut Rescanh Program 2.5.1.2 M(It Superheat ~

FY92 and FY93 ciforts will focus on making an ll Cl a stand alone predictive tool for FCI events. Presently A question was raised as to whether the NURl!GI 11 Cl is a rmdule of the inactive M11 PROG systems CR-5423 analysis of melt superheat was based on the best code.The thrust of the present pr ogram is to extract IFCI understanding of core-concrete interactions to date. San-from MFl. PROG. Only these modules needed to repre. dia National laboratories performed the necessary con-sent I cps and the interaction of the governing processes firmatory calculations using CORCON MCD3. 'the re-of l CI's wdi be retained.This ef fort willinclude testingof sults indicate that the melt superheat duration in the resalting stand-alone cale and correction of any mod. NURl!G/CR-5423 is very censervative (i.e., CORCON cls or algonthms required to obtain a robust analysis tool. MOD 3 calculation of superheat duration is one-third of

'the code will be documented in a draft NURl!G report the values used in NURI.G/CR-5423). Therefore, the and peer reviewed in FY93 prior to issuing the final re. thermal loading of the liner used in NURl!G/CR-5423 is port, deemed appropriate.

2.5.1.3 Melt Spreading Phenomena 2.4.6.3 Anticipated ltesults A concern was raised that in underwater volcanic lava

'the anticipated result of this work will be a stand alone Dows. crusts have been obse rved to form at the lava water IFCI code that c;m be used on workstation to analyte interface, forming an annuius inside of a lava crust withm non-explosive FCI events, and explosive events when the w hich molten lava could flow without interacting with the triggering mechanism is known. An operational report surroundmg water.The concern was ' hat this phenome-that demonstrates the applicability of the code to full non, should it occur, could insulate Ihe molten core mate-scale plants wi!! be provided along with the peer reviewed da Do ng mt of the pedestal region from the overlying documentation report, water pool, and it also could prelcrentially channel the flow to the liner, eliminating the benefits of spreading.

2.5 llWR Mark 1 Containinent I.iner 'this is a complex problem not amenable to simple cxperi-mem I remludon.The Mlil TSPRl!AD-1 code has been Failtire used to perform melt spreading sensitivity analyses to 2.5.1 Current 1(esearch Program As stated in Appendix A.1, the NUREG/CR-5423 peer 2.5.2 Future Plans reviewers identified three areas that warranted additional .lhe staff plans to incorporate the results of the above-research to confirm the appropriateness of the analysis m mentioned confirmatory activities into a NUREG report.

NURl!G/CR-5423. these three areas are liner failure A final peer review of the NURiiG report will then be enteria, melt superheat, and melt spreading phenomena, conducted.

'Ihe staff has initiated necessary research in each of these areas as discussed below.

2.6 Ilydrogen Cornbustion and 2.5.1.1 uner Faiiure Criteria Research

'lhe NUREG/CR-5423 analysis assumed that the liner 2.6.1 Current Research Program would fail when the temperature of the liner reached the melting point of the liner steel. A member of the peer Appendix A.3 describes the state of knowledge of hydro-review group suggested that the steel liner could dissolve gen combustion and transport and identifies processes 43 NUREG-1365, Rev. I

2 Itescarch Plan that are not w cil understood.The NI(C has recently spon- tained by conventional heat exchanger techniques. 'lhe sored two programs for expenmental investigation of is- natural heat transfer proccues withm the explosion ves-sues that heretofore have received httle attention. While sels and the timmg of combustion and mixing processes hydrogen research has considered diffusive flame behav- will be used to vary the temperatures in the gases. Dis-ior, flame acccleration, and detonation behavior, the re- persing metal or inert dusts within one vessel prior to search to date has been limited to tests involving hydro- combustion will allow the creation of particulate laden gen air 4 team nuxtures under ambient or relatively low atmospheres, which will simulate the process of high-(lbo*C) temperature condaions. In the absence of reli- preuure melt ejection during a DCII event. Pressure and able ignition devices, auto ignition of jets and plumes temperature measurements and photographie recordmgs released at high temperature:, during a severe accident will be used to determine the nature of the combustion could result in the contmuous burning of hydrogen asit is phenomena.

released into the containment. I or premixtures of hydrogen airateam in containment at elevated tempera- livaluation of hydrogen transport in reactor containmerits tures (but below the auto-igmtion temperature), flame remains a long term research issue owing to a !!mited set acceleration and high4 peed combustion may be more of experimental data and thus limited validation of exist-hkely to lead to a transition to detonation. 'thus, our ing codes used m such analyris (i.e., CONTAIN,lih15).

current and future experimental research is directed at Current and near term future activities include a tr ode:;t confirming the treatment of this behavior, program to document the development and assessment of the lih1S code. 'lhe staff also expects to benefit from To address the effects of elevated temperatures on flame DNsponsored research to upgrade the i th1S code such acceleration and detonation transition, the NitC initiated Mat the code may be readily applied to dif ferent configu-radons; the cather versm" of IthiS used in NIK.

a high temperature, high4 peed hydrogen combustion program under a joint agreement (signed in June 1991) ponsored research, an analysis of plume mixing and dif-for a cooperative program with the h1mistry of Interna. fusion flarne combustion of hlark 111 reactors, required tional Trade and Industry of Japan and the Nuclear Power code nudcadon for different geometries. At the com.

linginectmg Center Under this agreement, a high-tem- plehon of the IX)I! related activity and our sponsored perature high-speed hydrogen combustion research pro-activity to document the code, the Nitt will be better ,

gram, extending over 5 years, has been developed. Two luitioned to apply the code for reactor analysis and vah-combustion vessels will be used for this research program dation. With regard to the TON'l AIN code that utilizes at ilrookhaven National laboratory. 'the largest of the the nadibonabmuoholume tdnye b containment two vessels will be approximately 30 cm in diameter and analym, , va adon of the flow and mmng models against 20 m long. liigh temperature hydrogen combustion mix- IIDit project data is continuing with analysts of the inter-tures will be used with a pre-ignition temperature as high nahonal standard problem ISP-29.

as700 K and an initial pressure of 1 atmosphere.The test As a result of our recent cooperative agreement with 3 gases that will be used will be mixtures of hydrogen, air, ~

oxygen, mtrogen, steam, carbon dioude, and carixm mon-hpan in the area of hydrogen research, the NitC now has access to ongoing hydrogen mixing and distribution test-onde, lhe facility wil1 be able to accommodate testmg with and without ventmg. the smaller of the two vessels n the Tadotsu facility, a largc+cale mixing and distri-bution test facility that simulates at one-fourth scale a will be approyimately 10 cm in diameter and 6.7 m long. PWil 4 hiop reactor containment with compartmen-

,this vessel will permit use of the apparatus for learning the effects of high temperature on the smoked foils and talization. Additional data from hydrogen mixing and combustion testing in the Takasoga facility, which roughly testing the instrumentation. It will also provide prehmt-nary data to assess the S, n epherdf/.ND model for calcu.

s mulates a Ill!S Alt SP-90 ty pe design, will also become lating cell sire. ^available during FY92-93. Tc';t results from these facili-ties will provide a greatly expanded and improved data base for the validation of our anatytical tools. It is antici-In the low speed hydrogen combustion research program, pated that long-term cot firmatory research will focus on the aspects of dtifusion flames scalability and transient this newly available data.

high temperature combustion will be investigated. The results will be used to help resolve outstanding issues in An important supplement to our hydrogen research pro-severe accidents, i.e., hydrogen combustion aspects of gram is that work being carried out under an arrangement DCll; high-temperature combustion phenomena, and with the ilussian Academy of Sciences in collaboration detonation initiation by high-temperature steam hydro- with the 1.V. Kurchatov Institute. Under this cooperative gen particle laden jets.'lhese experiments will use a tech- progiam, the iIhtS code, which has been used to predict nique of combusting premixed rich and lean mixtures m hydrogen mixing and combustion, will be assessed and separate vessels followed by deliberate mixing to initiate compared against llussian experimental data. Under this combustion.'lhis will enable the creation of rnuch higher agreement, the NitC is also provided with the results of temperatures (1000-3000*K) than could be readily ob- Kurchatov hydrogen deflagration test data and the results NURiiG-1365, llev.1 44

. - . - - - - - . . . ~ .-- -----~ - - _ . - - - . .. .-__

2 Ilescarch Plan of a program to develop a scaling relationship for the since the release is direct to the emitonment. Ilowever, critical conditions of turbulent jet initiation of a detona- analytical tool Uke VICTOltl A has been developed spc-tionJihis research includes elements of experimentation, cifically (section 2.4.4.) to address source term for by. pass nutnerictd simulation and theoretical analysis. An im- accidents.

i- proved understandmg of these phenomena will enable a more definitive evaluation of the detonation potential to Although additional physics and chemistry research can  ;

be performed for both U.S. and Russian nuclear power be performed to reduce uncertainties in source term phe-plants. nornena, it is important to consider the need for such >

research and the potential Ihat this research could signifi-cantly improve our risk perspective on severe accidents.

• 2.7 Source Term "C"ce' '"!"'c 'escarch is oriented to assess the NRC severe accident codes and to address residual source term issues related to plant configurations during shutdown 2.7.1 Slutits situations which would lead to the possibilityof airingress into the core either by natural circulation or from the At de present time, the NitC is pursuing several regula- residual heat re.noval sptem. 'lhe air will interact ex-tory inillatives to incorporate insights from updated se- othermically with the cladding remaining on the fuel, vere accident source terms. Updated source term insights producing high ternperature in the fuel. Vapor of ruthe- t arising from the technical update of TID-14844 are e - nium and molybdenum will be produced because of the pected to be made availabic for voluntary use by existing strongly oxidized conditions. A test will be conducted at -

licensees. A revision of 10 Cl R Part 50 to incorporate OltNI. at these conditions in FY93 if the ongoing risk updated source term and severe accident insights will study sponsored by NRC indicated that accidents during then be undertaken, with a proposed rule for comment shutdown conditions are major contributor to the core expected to be issued by early CY93. Although regulatory damage frequency. 'the NRC's participation in the positions arising from updated source term insights re- PlilillUS-FP project is to obtain integral effects experi-main to be developed, kome preliminary implications can mental data to validate NRC severe accident codes.

he seen at this time. It is clear that updated source term insights indicate the need for consideration of nuclides ,

2.7.2 l'llEllUS-FP (e.g., cesium)in addition to iodine and the noble gases. In addition, revised insights on iodine chemistry call into 2,7.2.1 Objective question the need for high-efficiency charcoal absorbers .

(assuming that the pif is controlled, post accident). 'the pe objective of the PillillUS-Fl* Proj.ect is to perform kidine chemistry cem,in turn, impact such important plant integral effects experiments in an m pile test facility, un-systems as fission product cleanup systems, control room der suf,ficient prototypical conditions, on the, processes habitability, and sllowable containment leak rate. Finally, chemistry of fis-

- and most importantly, the above discussion and all recent governmg sion pmductsthe undutransport, LWR severe retention, acci and, dent condit l risk studies have shown the importance of maintaining .the pmcesses to be investigated are those taking place in

! containment integrity under severe accident conditions in the cere region, in the reactor coolant system, and in the order to ensure low risk.'This strongly suggests that the- contamment building /Ihe experiments will also study the l

' appearance of a severe accident source term within con, degradation of high burnup fuel.-typical of the later -

tainment and challenges associated with such releases phases of the accident, should be more closely linked with the temperatures, pressures, and contamment loads, rather than an arbl* 'lhe Commissariat a PEnergie Atomique (CUA) of-trary linkage with a smgle sequence such as a large break France and the Commission of the European Communi-loss-of-coolant nccident. Ihe discoancet m present prac- - ties agreed to undertake the Pile 13US-FP Project in-tice is not so much that temperatures and pressures came close collaboration, using the experimental facilities -

- from only one sequence, Lit that they came from a se- available at the Research Centre of Cadarache, France.

NRC entered into an agreement with CHA: under this

.quence that was terminated without perceptible damage to the core, since the analysis that gave those tempera- agreement fission product generation; transport and

- tures and pressures was required to show that the peak deposition generated in the PIIE13US-FP program will - -

clad temperature remained below 2200'F. the revised be made available to the NRC.

souice term, is believed to be consistent with the source term expected to release into the contamment resulted 2.7.2.2. Facility Layout and Testing i-from core melt under low pressure severe accident se- PilEllUS-FP is a loop-type test reactor with a. Iow-quence>. It also provides the basis for the evaluation of enriched driver core of 20 to 40 MW power, using fuel rod the effectiveness of containment mitigation features un- elements. Core cooling and moderation is achieved by der severe accident conditions /Ihe revised source term is demineralized light water, and light water and graphite not intended to address by-pass accident source term, are used as reflectors.

45 NUREG-1365, Rev.1

g. f-- y pr --

m y yy-nWpa ys +

-ty u.y- m nr-pp w .yp-*,9 .a-,i---

2 Research Plan A cluster of 20 fuel rods, I rn long, in a PWR conhgura- etc. An extensive pmgram for post-test examination, to tion, is inserted in a test tram and located in the central back up and complement the on tine measurements, is hole of the dnver core of the PlilillUS-FP reactor. being put in place at Cadarache with collaboration from several qualified laboratories of the European Commu-llefore a test, the test fuel from the llR3 reactor (a llel. nity.

gium reactor that use 1 m-long fuel rods)is re-irradiated in the PilEllUS-FP in-pile section for 2 weeks using the 2.7.2.4 Test MaF.s existing pressurized water hiop in order to generate a sufficient inventory of short- and medium lived fission The main objective of PilEllUS-FP is to obtain integral products.'lhe loop is then slowly blown down with simul- effects experimental data on fission product transport in taneous reduction of the reactor power, with the in-pile the RCS and in the containtnent.This implies studies on section isolated from the loop. After these steps, testing retention and revaporization of fission products in the may begin. During the test phase, the in pile section is RCS. Revapori/ation could be enhanced either by decay connected to a circuit and vessel that simulate the primary heating or by a sudden steam spike in the circuit. The circuit and containment building of a PWR. formation of gaseous species in the containment is also of a great importance, Studies in PHEllUS-FP are expected During the test phase, the in pile fuel bundle is heated by to yield information related to volatile iodme coming fission power from the driver-core at a rate typical of a from radiolysis of the sump or organic iodine fmm paints severe accident up to temperatures at which the fuelis in the containment.

damaged. 'the test bundle is pushed to conditions in which fission product release takes place, and contro! The current test matrix consists of six tests.'Ihe intent of rods and structural materials are vaporized, pmducing each test is to capture the key process s and phenomena t afficient quantities of aerosols. 'the fuel bundle will be associated with a particular severe accident sequence damaged to the extent necessary not only to release fis. (e.g., V sequence) and not to simulate a sequence in details. The first test is scheduled for 1993, and subse-ston products and aerosols, but also to study the mechani-cal behavior of the fuel during extensive degradation. quent tests are scheduled at yearly intervals.

The released fission products and acrosois are swept by a flow of steam and H2 into the circuit that simulates the 2.7.3 Other Research Activities primary cooling system up to the point of pipe break.

Then the flow enters a vessel that simulates the contain* In addition to the PHEllUS-FP pioject, the following ment buildmg- research activities are planned.

• In FY92, complete the analysis of the ORNL VI-6 2.7 2.3 Instrumatation fission product release test and a technical report on Instrumentation is used for process control, safety, and the analysis and interpretation of the fission product interpretation of the experimental results (scientific release experiments performed at ORNL Beyond analysis), During the design of the PHEBUS-FP facility, FY92, additional tests (VI-7) will be conducted to a large effort was devoted to the instrumentation for investigate fission product release at high tempera-scientific analysis. The success of the PIIEllUS-FP ex- ture for (1) fuel exposed to air under shutdown con-periments depends largely on the capability for measur- ditions or residual fuel remaining in the reactor core ing the parameters of interest. after vessel bottom melt through, and (2) high burnup (>60 mwd /kg) ano low-power-dmity The fuel bundle region will be instrumented with tem- ALWR fuel, perature, pressure, and flow sensors. The primary RCS circuit will be provided with two main instrumented sec-

• In FY92, complete model development and docu-tions: one in fmnt of the large components (steam gen- mentation for the VICTORI A code, and validate crator, pressurizer, etc.) and one just before entering the the code against test data. In FY93, complete a peer containment vessel.The main measurements of interest review of the VICTORIA code. Further code im-are (1) thermal hydraulic conditions, (2) fluid composi- provement or development will depend on the out-tion, especialty the fission product content, (3) acrosol come of the peer review.The code will also be used concentration and granulometry,(4) deposits on the pipe for PHEBUS-FP pre- and post-test analysis and for wall, and (5) composition of the gas phases, including Mr. the International Standard Problem exercise on the These will also be measured in the containment vessel at FALCON project in the Winfrith Technology Cen-several positions in order to characterize their spatial ter (WTC) U.K. Benchmarking of the CONTAIN distribution. In addition, the sump water will be moni- code will also be carried out with the PHEBUS-FP tored with respect to dor.c rate, isotope concentrations, data.

NUREO-1365 Rev.1 46

I2 Research Plan -

o. A small experimental program will be continued at and thermochemical data for a combination of mate-Ilattelle to provide the capability to utilize mass rials (fuel, reactor structural materials, etc.)..

= spectrometry under various oxidation states (reduc-ing and oxidizing) at different temperatures Appendix B.1 provides a more detailed discussion on  ;

(1000'K to 2700'K) to obtain chemical speciation source term issues.

g .

i.

---t i

1.

t

' }

}

47 NUREG-1365, Rey,1

. .= e 1

3 ItESEARCII PLAN FOlt ADVANCED LIGilT WATEllitEACTORS 3.1 Introduction 3.2.L1 Rmanh Appniath Assess the methodologies used to develop or evaluate the The role of severe accidents, despite their exceedingly effects of phenomena resulting from ALWR fuel and low probability, has been recognized for c(msideration in core designs on severe accidents, and the effect of design advanced light water reactors (ALWR). In fact, the im- features (i.e., cavity fhnling) to prevent lower head fail.

proved ALWR design featur es that reduce the likelihood ure.

and consequences of severe accidents, and the excellent safety record that continues to accumulate frem operat-mg plants, leaves only such exceedingly low-probability, M*U Di m &n high.c(msequence, events to be of concern.The objective The ALWR fuel design typically involves a lower power of the ALWR severe accident research program is to density than current reactor designs. For example, the examine the issues in depth and develop the necessary AP600 plans to employ a power density of 73 to 79 kw/

tools to address these issues. While it is recognized that liter compared to 100 to 110 kw/ liter for current PWRs. A the Westinghouse AP600 and the GE SHWR are suffi- lower power density results in an increase in the mass of ciently different from existing LWRs that different severe the active core for a given power level, and correspond-accident issues or vanations on existing issues may exist, ingly, the mass of zircaloy as well.This lower power den-much of the phenomenological understanding of accident sity provides additional thermal margin during transients, progression, containment loading, and source term analy* and there is the potential to form a relatively thick protec-sis developed in relation to existing reactors will also tive oxide layer that could delay the onset of fuel melting.

apply to ALWRs. %c readiness and applicability of the Nevertheless,if the core does melt, there is the potential severe accident codes (e.g., SCDAP/RELAP5, MEI' for greater II, generation owing to the greater amount of COR and CONTAIN) to ALWR plant-specific designs z reakiy present. The ability to evaluate the effect of this and phenomena must be addressed, importimt phenom- additional non condensable gas should be assessed, in-ena addressed in accident progression analyses include ciuding but not limited to its effect on events involymg timing of core melt and vessel breach, in vessel hydrogen steam generator tube rupture and events with assumed production, fission product transport and hehavior, fuel- credit for natural circulation in the primary system. Al-coolant interactions, release of fuel from the vessel, core- though there is no reason to expect any new or additional concrete interactions, hydrogen burns, and containment phenomena to be important to in. vessel severe accident loading. We will assess the state of knowledge for phe- scenarios as a result of this iuel design, natural circulation nomena that arise in ALWR designs, that have not had effects may need to be modeled more accurately in our importance for severe accidents in current plants. Con- analytical codes such as MELCOR and SCDAP/

tainment heat transfer, mixing, and hydrogen distribution RELAP5.11cnce, adequate assessment and documenta-in the passive containment are examples. As discussed in tion of these codes are vital to provide th_e confidence in -

Section 2.4 of this report, NRC's severe accident codes . using them for ~ALWR analyses. Other than the lower l will be modified to be capable of analyzing ALWR de- power density, there appear to be no major differences in signs, and input decks will be developed specifically for the core design between the present generation reactors the AP600 and SHWR designs. and the ALWRs. Ilowever, some differences in the core structures and caterials do exist (e.g., stainless steel re-This plan is based on an assessment of the applicability of flector rods at core periphery, flow distribution grid, ex-current knowledge of severe accident phenomena to the . tended burnup fuels, and burnable pois(m rods in ALWR designs. The plan identifies studies needed for AP600). The severe accident code SCD AP/RELAP5 will specific ALWR designs, where the reviews for current be modified to account for these design differences. Our:

reactors may not be sufficient. We do not foresee a need current understanding, with the knowledge that will be of new experimental research beyond what has been ac- acquired under the revised severe accident research plan, -

complished or is underway for current reactors, is adequate to address core melt progression for ALWRs.

The AP600 design includes the capability to iked the reactor cavity during a post ulated severe accident in order 3.2 In-Vessel Severe Acc, ident -

to cool the reactor vessel with an external water pool prior Phenoinena to core debris penetration of the vessel. The SilWR de-sign also includes a system to fkul the cavity. Analyses of potentiallower head vessel failure for ALWR designs will -

3.2.1 Core Melt Progression and Reactor need to consider the impact of accident management Vessel Lower llend Failure strategies for flooding the cavity, 49 NUREG-1365, Rev. I

3 Research Plan "3.2.2 Fission Product Transport and research program (i.e., ACl3 consortium MACII tests at Uchavior ANL and the WETCOR tests at SNL).

• lhe existing data base on fission product release should The SilWR cavity design was indicated to be similar to the -

be applicable to the new low power density fuel used in AllWR design. In this design, molten debris in the lower advanced reactor designs. When the current assessment cavity is allowed to spread to a shallow bed and then be -

and development plan for the VICI'ORIA code is com. quenched by water f rom the suppression pool.'lhe appli-pleted, the VICTORI A code will be adequate to address cability of the M ACE data to the SBWR design should be the fission product transport and behavior in the reactor evaluated.11 is recognized that there are differences be-coolant system for ALWRs. Sincc ALWRs rely heavily on Iween the experiments and the SilWR design (e.g., debris ratural circulation in the primary system for heat removal depth and composition).

during a severe accident, the effects of natural circulation on fission product transport and behavior could be as- 3.3.2.2 Direct Containment lleating sessed with codes such as SCDAP/RELAP5 and MEl-

.Itc AP600 has incorporated an automatic depressuriza-COR, provided the assessment discussed in Section 3.2.1 tion system fer the RCS and climinated instrument pene-above revealed no major deficiency.

trations in the bottom head of the reactor pressure vessel.

These design changes, however, do not completely climi.

3.3 Ex-Vessel Severe Accident na e the DCll threat because reprusurization of the RCS during core degradation, particularly as a result of I,llenoniena core debris-coolant interaction, is still possible. Current research on lower head failure mechanisms and DCH -

3.3.1 Research Approach phenomena are expected to provide information that would allow the NRC to provide a preliminary assessment Assess the methodologies that are used to evaluate design of the mitigative features of the ALWR designs. Ilow-features incorporated in ALWRs to mitigate the effects ever, final conclusions on the retentive capabilities of a of core. concrete interactions, high-pressure melt ejection specific cavity design to preclude DCH may require fur-and DCll, and hydrogen combustion, including but not ther attention depending tipon the credit required for limited to methodologies that evaluate cavities designed such features.

to mitigate the effects of DCH and methodologies that evaluate cavity floors designed to provide sufficient area For the SBWR, the low primary system pressure, the to allow cootable debris geometry.- inerted containment atmosphere, and the passive de-pressurization system should significantly reduce the threat of a high-pressure melt ejection. Again, the meth-3.3.2 Discussion odology bemg developed in the current research program Within this context, the assessment is on the methodolo, is adequate for application to SBWRs.

gics used to evaluate design features incorporated in the ALWRs to mitigate the effects of core-concrete interac. 33.2.3 Ilydrogen tion, high-pressure melt ejection and DClI, and hydrogen 33.23.1 Research Approach The methodologies that are combustion. It includes, but is not limited to, methodolo-used to evaluate acceptability of the vendors' proposed .

gies that evaluate cavities designed to mitigate the effects hydrogen concentration criteria will be assessed, includ-of DCH and methodologies that evaluate cavity floors ng evaluation of the maximum allowable 112 concentra-

. designed to provide sufficient area to allow a coolable tion and the hydrogen control features proposed within geometry, the containments.

Another area for assessment is evaluating the new con- 33.23.2 Discussion For AP600, the current design crite-tainment cooling c<mcepts, including passive natural cir- ria proposed by Westinghouse for hydrogen is to have a -

culation cooling using air and containment dome cooling containment volume large enough that the bulk average using water sprays. hydrogen concentration could not exceed about 13%,

assuming good mixing and reaction of 100% of the active

, 33.2.1 Molten Core. Concrete Interactions cladding. local accumulation of hydrogen will be con.

L trolled by de-powered igniters. Westinghouse's choice of The available knowledge for predicting molten core- 13% as a maximum allowable hydrogen concentration is concrete interactions (MCCI) appears to be both applica- most likely based on existing correlations, which indicate ble and adequate for ALWR accident analyses. 'The as- that this is the highest practical concentration for which

sessment of the criterion that core debris be spread over stable detonations could be ruled out, llowever, a cen-an arca of 0.02 m2/Mw thermal to assure coolability needs centration of 13% is too high to cyoid . combustion to be evaluated.This is part of the current severe accident altogether. If we assume the containment volume is NURliG-1365, Rev.1 50 l-

3 Research Plan -

limited to about 100'C by steam condensation on the vironment will take care of any problems. One unresolved containment dome, incomplete combustion can occur for issue for SilWR is the treatment of noncondensable gases -

hydrogen concentrations as low as 4% Flame balls can in the isolation condenser and the passive containment grow, rise to the ceiling, and quench, leaving a stratified cooling system (PCCS). Additional work may be found region with hot combustion products along the ceiling and necessary to assess the hydrogen effect on the isolation unburned hydrogen in air below Rapid, almost complete, condenser and PCCS performance.

combustion can occur for hydrogen levels above about 8%, which would leave little unburned hydrogen. For a 3.3.2.4 Containment Cooling containment atmosphere containing some water vapor, stablc detonations are highly unlikely for 13% hydrogen 3.3.2.4.1 Research Approach. He methodologies that are and below. Accelerated flames with damaging overpres. used to evaluate new containment coolmg concepts will-sures could occur inside this containment (- 10 percent or be assessed, including natural circulation cooling using air more 112 ) in cluttered regions that promote flame accel. and containment dome cooling t. sing water sprays.

eration, but when the flame transits into the region of 3.3.2.4.2 Discunion, I,or certain classes of accidents, the open containment volume,it may decelerate again. Some ,

residual shocks may strike the containment, but with far AP600 is expected to exhibit performance characteristics less energy than for a full containment detonation.The  ! hat are different from extsting PWRs. These phenomena include, NRC position taken on ALWRs is th,t the containment design should limit the hydrogen concentration to no (1) Natural circulation within the containment, greater than 10%, and that containment-wide hydrogen control should be provided to preclude the formation of (2) 11ydrogen mixing and development of high local local detonable mixtures and lessen the accumulation of concentrations, hydrogen on a global basis.

1. * # E "" '"

Although the above criterion goes a long way toward mitigating the hydrogen problem, two problems remain (4) Water cooling of the external shell of the contain, to be addressed. They are (1) loads from low-pressure ment.

combustion, and (2) kical detonation. It appears that low-pressure combustion does not generate overpressures For the external shell, flow patterns and correct heat adequate to threaten the containment. With 13% hydro- transfer correlations are needed for water cooling of the gen, overpressures of 4 to 7 bars may be expected, de- external shell. Inside the containment, the modeling of pending on the initial temperature and steam concentra- natural circulation and mixing processes over large vol-tion. Presumably, the containment can withstand this umes will pose a challenge to existing codes. For the overpressure. SBWR, the behavior of the suppression pool can be read-ily modeled. Like the AP600, for certain accidents the To address the kical detonation problem, Westinghouse SBWR may pose new challenges in modeling natural  ?

- is depending on de-powered hydrogen igniters to burn the circulation and mixing processes for existing containment hydrogen early, if the igniters are successful at triggering analysis codes.

combustion before the hydrogen concentration builds up to a locally detonable level, even local detonations can be - Natumi circulation flow and related mixing processes are avoided. This depends on the number and reliability of - key issues for both the AP600 and the SBWR designs. For the igniters as well as the strategic placement of these the AP600, this will primarily be a containment issue. For igniters. While the reliability of the igniters could be - the SBWR, it will be an issue for the passive core cooling addressed, the number and placement of them requires system and the containment long-term heat removal sys-the prediction of hydrogt.n concentration, stratification, tem, since they are closely related.

and distribution in the containment. Because the passive designs lack active mixing capability, assessment of con- Existing NRC severe accident codes CONTAIN and tainment mixing becomes a more challenging task.The MELCOR are control volume (i.e., lumped. parameter) .

existing codes used for containment analysis utilize a con- codes that do not take into account finite gas velocities or trol volume approach, and the constituents within each momentum convection in the control . volumes, or cells.

volume are assumed to be well mixed.This may be inade- Momentum is considered in the junctions, or flow paths, quate to address the issue on hand. Special flow field between cells, but this momentum.is considered to be ; __

models, e.g., HMS may be used to address this issue. completely dissipated in the downstream cell.The atmos-pheres in the control volumes are consequently assumed Current design criteria proposed by GE would inert the to be stagnant. Such modeling is appropriate for large containment region with nitrogen.Thus for the SBWR. control volumes connected by flow paths whose cross-all the hydrogen generation issues appear to have been sectional area is small compared to the cross-sectional dealt with by assuming that inerting the containment en- area of the control volume itself. In certain situations, 51 NU REG-1365. Rev.1

l 3 ltesearch Plan however, the neglect of velocities and momentum con. acrosol transpirt, deposition, and suppression pool vection within control volumes is not justified, and the scrubbing. 'the CON!'AIN code will be adequate to ad-distribution of flows can be adversely affected by this dress the source term in the containment. Given that the assumption. issues related to nat ural circulation and mixing processes are adequat ely addressed, the existing methods for source Since the abdity of the control volume codes to calculate term analysis are adequate, the di:.tnbution (i e., stratification) of hydrogen is limited in some cases, and since the principal NitC severe acci~

dent analysis codes are all contro! olume codes, a better 3*4 StirittiiriO-understanding of the limitations of these codes and pwsi- Westinghouse and Gli have not presented the NitC with bly improved modeling capabilities based on that under- details of their severe accident research program or the standmg could be important with respect to resolving extent to which vendors' research will satisfactorily ad-severe accident issues

• dress issues related to severe accidents. '!he NitC is plan-ning te hold meetings with the vendors to understand it i, pnsible to modify lumped parameters codes by im- their research programs, We will then develop a program -

plementing vehicities withm the control volumes. ihe to resolve all severe accident licensing issues. In consulta-COMMIX code (see section 2.4.4) is being mochfied to tion with NIllt, IU!S will identify the research program evaluate AI. Wit designs for mixing characteristics. that will be carried out by NitC. We do not foresce a need for new severe accident phenomena experimental pro-In addition to the general issues of mixing and natural grams. Ilowever, NitC will develop analytical capabilities circulation, the ability to evaluate Al. Wit designs will to independently assess the vendors' analysis. To accom-strongly depend on understandmg natural convection and plish this, we plan to:

steam condensation processes, particularly in the pres-ence of large quantities of noncondensablesJihe vendors 1. Complete the research dehneated in this updated have proposed testing to address the question. report.

3.3.2.5 l'ission Product 'Iransport and llehavior 2. Perform assessment or modification and documen-tation for NitC codes (Mill. Colt, SCDAP/

'the existmg knowledge on source terms is npplicable to Illil AP5, CONTAIN, COMMIX, VICTOltl A) so the AP600 and the SilWit. Clearly, source term predie- that they could be used to perform analyses to audit tions will be driven by containment behavior, particularly licensee calculations.

i N Ult!!G-1365, Itev.1 52

i t

APPENDIX A SEVERE ACCIDENT ISSUES, STATUS, AND PROGRESS TO DATE 1ntrodUCtIDn TaNe A.] Sewre Accident Research Pwgram issues During the past few years, the severe accident research Appendix pmgram has focused on generating information that can 1. Mark 1 Containment Shell Melt through A.1 narmw the bmad range of uncertainties that have been identified (e.g., NUREG-0956, " Reassessment of the 2. Core Melt Progression and liydrogen Technical llases for Estimating Source Tenns,1986," and Generation A.2 the Knuts report, NUREG/CR-4883, " Review of Re- 3. Ilydrogen Transport and Combustion A.3 search on Uncertainties in Estimates of Source Tenns from Severe Accidents in Nuclear Power Plants,1987") as 4. TMI-2 Vessel Inspection Program A.4 limiting the ability to accurately calculate source terms 5. Fuel-Coolant Interactions and and to provide an improved degree of assurance in esti- Debris Coolability A.5 mating risks from severe accidents. 'Ihe original eight

6. S,ource Terms 11.1 areas of uncertainty in NUREG-0956 were:

7. Core-Concrete Interactions B.2

1. Natural circulation in the reactor coolant system, 8. Severe Accident Scaling Methodology H.3

2. Core melt progression and hydrogen generation,

3. Steam explosions,

4. Iligh pressure melt ejection, A.1 Mark I Contairment Shell Melt tllrough (Liner Failure)

5. Core-concrete interactions,

6. Ilydrogen combustion, A,1.1 Ilackground

7. Iodine chemical form, and An accident sequence leading to early containment fail-ute has been postulated for UWR Mark I containments-

8. h, . .ssion pmduct revaporization.

This sequence involves the direct attack of the contain-ment steel liner by molten core material following vessel In the " Revised Severe Accident Research Program failure. In SECY-89-017. " Mark I Containment Per-Plan," NUREG-1365, published in August 1989, similar formance Improvement Program," the staff addressed areas of research were combined (e.g., iodine chemical the issue of severe accident challenges to the Mark I form and fission product reevaporization), whereas other containment and proposed a balanced approach utilizing areas that are either important to accident scenanos that accident management and mitigation as the optimum way might lead to early containment failure (e.g., Mark I con- to reduce overall risk in UWR plants with these contain-tainment shell melt-through and direct containment ments.

heating) or are important to assessment of accident man-agement strategies (e.g., adding water to degraded core) SECY-89-017 stated,"there is a growing consensus that were presented separately, and research needs wer e iden- water in the c<mtainment (from an alternate supply to the tified. Also in NUREG-1365, research plans addressing drywell sprays) may help mitigate risk by fission pmduct the issue of scah,ng of severe acetdent expenments were scrubbing and possibly by preventing or delaying contain-presented. ment shell melt by core debris. Research is continuing in order to confirm and help quantify these initial conclu-Appendices A.1-A.5 provide summaries of the severe sions."

accident issues listed in Table A.1, along with their cur-rent status, progress to date, and future plans. Appendi. NRC research over the past several years has addressed ces ll.1-B.3 provide the technical bases for closure of the key phenomena associated with the liner meltthrough issues related to sourcc terrre core-concrete interactions, issue, such as melt conditiorc at the time of vessel failure; and scaling methodology. Note that Items 4 and 8 are not melt spreading characteristics; thermal-hydraulic charac-

" issues" in the same sense as the other issues that repre- teristics of molten core-concrete interactions both with sent technical phenomena associated with severe acci- and without an overlying water pool; heat transfer charac-dents; however, they are major programmatic areas in the teristics at the interface of the mohen core, overlying severe accident research plan, and therefore, are listed water pool, and liner; and fission pmduct attenuation in and discussed as separate items. the presence of an overlying water pool.

A-1 NUREG-1365, Rev.1 l

.- . _ . _ _ __ _ . . _ _ _ __m .-._._.___.__m _~ _ , ,

Appendix A ~

LIntegration of_.the research information derived from For each of the three areas, a small(three or four) group i these programs into an assessment of the conditional of experts was convened, many of whom served on the probability of liner failure both with and without an over- larger peer review panel, to assist the NRC in addressing ,

lying water pool in the drywell, given a core rnelt accident the concerns. In addition, in order to ensure that the- 4 that proceeds to vessel failure, was completed. A descrip- initial melt quantity and composition were appropriate, a a tion of this ructhodology and its conclusion is provided in meeting was held to assess the adequacy of the report's NUREG/CR-5423. In summary, by developing probabil- quantification of the initial melt conditions at vessel fail-ity distributions for important parameters Ihat factor into ure. Additional analyses were performed using the the analysis from data (where available), computer analy- APRIL-MOD 3 code in which the heat-up, collapse, and ses, or other insights, developing causal relationships be- meltingof the llWR sicarn separator ordryer were explic-tween phenomena, and convoluting these distribution illy modeled.1hc basic conclusion was that the boundary functions and causal relationships, the authors of conditions used in NUREG/CR-5423 seem to be ade-NURI!G/CR-5423 obtained estimates of the likelihood quate.

ofliner failure both with and without a water pooloverly-ing the molten corium in the drywe!!, When water was assumed to overlie the molten core material as it spreads NUREG/CR-5423 identified three scenarios that could on the drywell floor toward the containment liner, it was challenge the containment integrity. Scenario 1 is based concluded Ihat the liner failure would be physically un. on an initial sudden massive core slump (50% of the core) reasonable. In the absence of water, however, the same leading to localized lower head failure. Scenario 11 is

, conservative approach led to the conclusion that failure based on initially quenched debris and a subsequent local--

would be certain, ized lower head failure owing to water depletion and remelting of the debris in the lower plenum. Scenario til ,

is similar to Scenario !!, except that the debris heats up A.1.2 Status the lower head uniformly, resulting in weakening and d

eventual creep failure. Detailed analyses were performed -

NUREG/CR-5423 has been subjected to an extensive for Scenarios I and 11. For the NUREG/CR-5423 peer and thorough peer review. Nineteen experts knowledge- review, the peer reviewers were informed that the evalu-able in the subject matter were asked to review the analy- ation of Scenario Ill would await the completion of the i'

ses, in addition to providing detailed written comments, NRC lower head failure analysis program at Idaho Nat the peer reviewers were given the opportunity to discuss tional Engineering laboratory.1his program addresses their comments _with the authors in a public workshop the likelihood of creep rupture of the lower head as well

- that was also attended by members of the Advisory Com- as other potential failure modes of the lower head.1hc '

mittee on Reactor Safeguards and the Nuclear Safety INEL program is now complete, and draft NUREG/

Research Review Committee. CR-5642 has been issued. 'lhe preliminary results indi-cate that for a depressurized reactor vessel, global vessel failure is not likely to occur. 'Iherefore, the NRC is not

,Ihe main conclusions of the peer review and the work-shop wue, planning to perform any analyses to qualify the potential for containment shell meltthrough resulting from acci.

dents that follow Scenario Illi ,,_

h 'l 'the methodology employed in NUREG/CR-5423 l was considered basically sound-no major deficien- ~

L cies or problems were identified that would invali- While there are residual issues related to de uncertainty date the results. of analysisin NUREG/CR-5423 in predictinglinerinteg .

rity in the presence of water, it is generally recognized 2 _The sensitivity study performed in NUREG/ Wat k pmm of watu wm sha@ anenuam me ,

magn tude of anosol pmduction and radionuclide re-CR-5423 showed there was no single process or parameter that had a controlling influence on the lease. Aerosol pr duction is affected by water because overall failure probability.

am a ng sesWuce@ecoMnsanachn -

concrete must sparge through the water. A recently com-picted NRC-sponsored study investigated the detailed 3 1here was also a general consensus by the peer processes involved in aerosol trapping by water pools and '

review group that three areas warranted additional developed a simple model of water effects on aerosol research to confirm the appropriateness of.the ' generation during core-debris interaction with concrete.

analysis in NUREG/CR-5423. 'Ihese three areas ' The model provides the probability distribution functions -

are liner failure criteria. melt superheat, and melt for decontamination factor (a measure of aerosol trap-spreading phenomena. The staff's plan to conduct - ping) based on the uncertainty analysis using the Monte the necessary research in each of these areas is dis- Carlo simulation technique, llounding case calculations cussed in Section 2.5. using this simple model indicate that water has a NUREG-1365, Rev.1 A-2 l

Appendix A profound mitigative effect on aerosol production and ration illustrates essentially the general melt progression radionuclide release. phenomenology thought to apply to both recovered and unrecovered blocked-core accident sequences in PWRs, and also to any llWR accidents that have blocked core A.2 Core Melt Progress,on i sequences. Differences in specific phenomenological be-havior from Thil-2, for unrecovered accidents are possi-A great deal of information has been obtained on the ble, however, particularly with regard to the question of processes involved in the early phase of melt progression sideways or downward meltthrough from the core.

in core uncovery accidents that extends through cure deg-radation and metallic (but not ceramic) material melting De similarity of the results of the many integral tests on and rekication. Ihts mformation has come from mtegral core degradation and early phase (metallic) melt progres-tests in the PHF, ALRR, NRU, NSRR, and Phebus test sion show that the overall behavior in this regime is repro-reactors, from the I.OFI FP-2 test, from tests m the ducible and is not strongly stochlastic in nature. Such CORA ex-reactor fuel-damage test facility, and from assurance is not available for the late phase melt progres-separate-effects expenments on significant phenomena, s on, howeser, with Figure A.2.1 the%11-2 core examin- -

Ihese tests have provided core degradation mformation ation providing nearly all the available information.The on fuel failu re, Zircaloy oxidation by steam with attendant processes of melt pool growth in a particular ceramic hydrogen generation, Zircahiy-clad melting and reloca.

debris and melt through of the supporting crust system, tion, and the effects of PWR Ag-In-Cd control rods' however, do not appear to be stochastic processes.

IlWR ll C 4 control blades, high burnup fuels, and the reflooding of severely damaged cores.hiost of the avail- .

able information on late-phase melt progression has hietalh.e melt relocation leaves behind free-standing come from the post accident examination of the Thil-2 cracked UO2 fuel pellets and ZrO 2oxidized cladding reactor. Despite the core reflooding that successfully ter, shards that have melting points, includm, g cutectics, m the minated the Thti-2 accident, the general late-phase melt range from 2800*K to 3100*K. During late-phase melt progression phenomenology of that accident, although progression m unrecovered blocked core accidents and in not the detailed behavior, appears to be applicable to very severe recovered blocked core accidents such as unrecovered as well as to recovered accidents and poss . TMI-2, a mostly ceramic melt pool forms and grows m the bly to some BWR accidents as well.The integral experi, mostly ceramic dehns bed. Dus, the metallte and the, ments that have provided most of the current information ceramic debns with melting points that differ by 600'K or base on melt progression and an outline of the informa, mme bicome separated in space, and, as the Th11-2 core tion obtained from these experiments are given in Table examination shows, they behave quite differently as core A.2.1. heat.up nnd melt progression continue.%e metalhe and -

the ceramic materials need to be treated separately in The results of these integral tests and'the B11-2 core accident analysis codes if melt progression is to be repte-examination have provided a consistent picture of melt sented realistically. hts ts not done m the older sunph,- _

progression. This is illustrated in the end. state configura- fied codes that treat the core melt as a single fictitious tion of the TM1-2 cme, which is shown schematically in "corium" fluid with a unique (high metallic) composition Figure A.2.1. Figure A.2.1 shows the development of a and a relatively low melting point (usually 2550'K).

debris-supporting metallic blockage above the water level in the lower portion of the core during coolant boi!down. In HWR accidents with automatic depressurization, pri-This blockage is produced by the rek> cation and freezing mary system blowdown lowers the water level below the of metallic melt, mostly from unoxidized Zircaloy clad- core and the BWR core plate. Urder these conditions, ding. Fission-product decay heating produces a growing . heat up occurs in a dry core with very low steam flow.He pool of mostly ceramic UO2 fuel and oxidized zircaloy contribution of zirconium oxidation to the core heat upis above the metallic core blockage.The growing pool melts small under these conditions, and a large fraction of the downward and radially outward through the supporting- Zircaloy cladding melts over a short period of time. It has -

metallic bk)ckage and the secondary ceramic crust that . been hypothesized that, under these conditions, the me-surrounds the pool. During [ mot growth, the crust system tallic melt (including eutectic alloys) drains from the core -

melts and reforms, relocating downward and outward. It and the BWR core plate into the water-filled lower -

appears that meltthrough occurs when the crusts do not plenum instead of freezing to form a blocked core similar reform to continue containment of the melt pool. At to that at TMI-2.

~

- Th11-2 with a reflooded core, pool melt-through was out the side of the core.The mass and other characteristics of He reflooding to the top of the core in the TMI-2 acci-the ceramic melt that drains from the core into the lower dent, however, did not prevent continued core melting plenum in blocked. core accident sequences are deter- because the water did not penetrate into the hot molten mined by the threshold and the location of the pool region of the core.The accident was only successfully meltthrough of the supporting erust.ne Thll-2 configu- terminated when hot ceramic core melt that constitute

. A-3 NURIIG-1365 Rev.1

Appendix A Tal.lc A.2.1 Sources of Current Int (gral Experimental Information on hielt Progression Experiments Key Ir. formation PilF Severe l'uct Damage Tests Integral information base on core degradation and melt progression 1

SFD-ST,1-1,1-3,1-4 Control rod and high burnup fuel effects NRU Full-length Tests Data on length effects and the absence of a cut-off to hydrogen FIJIT 1, 2, 4, 5 generation ACRR Damaged FuelTests Separate-effects data on core degradation and melt progression DF-1, -2, -3, -4 Ilasic IlWR information from DF-4 CORA lix reactor Tests and Related llasic information base on material-interaction effects and metallic Experiments melt relocation Core degradation information for IlWR and PWR geometries, including reflood effects, using electrically heated, simulated fuel rod bundles of up to 57 rods Phebus SFD Tests Core degradation and early phase melt progression phenomena NSRR Reactivity Initiated Accident Fuel failure thresholds in reactivity initiated accidents (RI A) Tests 1.0Fl' FP-2 Large-llundle Unique results on metallic melt relocation and the absence of a (101-rod) Test cutoff to hydrogen generation with a large flow-bypass area Significant results on the effects of reflood on core degradation and hydrogen generation Unique test results with fission product decay heating ACRR late Phase (Ceramic Melt) Tests Dry UO 2debris-bed thermal characterization and melting behavior _

DC-1, DC-2, h1P-1 Dynamics of pool and crust growth in particulate ceramic debris beds TMI-2 Core lixamination htajor source of significant information on late-phase melt progression Results applicable to basic phenomenology for both recovered

& unrecovered accidents NUREG-1365. Rev.1 A-4

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

Appendix A i

.;; x' 3' -

$. $ b $ b b d%

_j N '

l lk '

i i

(

2Binlet l f' 1 A inlet

_ i O

- 0 0 0 0. 0 0- 0 0

. . . 8 0  ! 0 i O l 0l 0 l 0 : 0 : 0 i** 0 0 : o ,- o  ; o; o r o z o s o k ,

Upper grid

\ ,

o ?

• MnufhEf\ 1 damage --  %' 'i gg%" 5 N l l 3 p Cavily

\

t Coating of previously. -- se core deMs h.

I molten material on bypass region interiof {, - - Crust j surfaces ,

l

) a* , ,t 5

i ' ' ~ ~ "

j l -- 7 Previously molten j l/ material j ,.

Hole in d

/

baf fle plate ,

H. ,

o p

5 7 ~ C' -

x g uCEl-

_,., =s_a

E

l_f Ablated incore M instrument guide S~ E RN .,

Lower plenum debris

-,  ;;T:; O, .GU" D

JU. b

?

ets .- -

p ., - c o .

i k

Figure A.2.1 TMI-2 Core End State Configuration A-5 NUREG-1365, Rey,1

Appendix A Table A.2.2 Core Melt Progression: Status of Current Understanding o Early (Metallic Melt) Phase, Reasonably Well Understood Phenomena:

Ciad balhioning Intact-core-peometry oxidation heating and hydrogen generation.

UO2liquefaction (dissolution) by molten Zircaloy.

Eutectic material interactions among UO2, ZrO 2, Zr, and control materials and their oxides. Rate limitations are less well understood.

Opening up the compartmentalized IlWR core early in a llWR accident by the eutectic interaction of control blade material with zircaloy channel box walls. _

Molten iircatoy rehication is a noncoherent, noncoplanar, rivulet-flow process, not a film flow process. The melt forms an incomplete blockage that does not cut off steam flow and hydrogen generation, o late (Ceramic Melt) Phase, General Understanding:

Information primarity from the TMI-2 core examination.

Results are also generally applicable to PWR unrecovered accidents.

Ceramic melt pool growth and meltthrough frorn a blocked core.

Refhloding probably stopped downward pool and crust relocation to give side melt-through at TMI-2.

Limited melt mass released from core (20% at TMI-2).

low metal content in ceramic melt pool.

During Refh>od: much hydrogen generation and strong heating of uncovered core from Zircaloy oxidation by reflood steam (LOFT FP-2 and CORA), is not well understood.

about 20% o! the core mass drained from the melt pool The question of metallic melt drainage or core blockage is into the water-filled lower plenum (an additional heat a major branch point for in vessel core melt progression, sink).The melt was cooled by the lower plenum water and and it has a large effect upon the chuacteristics of the did not fail the vessel lower head. The core reflooding, melt released from the core into the lower plenum. In the however, did stop the previous downward rek) cations of core blockage case, a large mass of mostly ceramic melt at the metallic crust (by melting and refreezing) and may about 300*K drains rapidly into the water-filled lower have been the cause of the melt through of the melt pool plenum, as happened at TMI-2. In the drainage case, out the side, rather than the bottom of the core. This layers of quenched melt are formed under the lower resulted in the drainage of only about half the melt pool plenum water in the order of their time of melting and rather than the entire pool. drainage from the core, with the low melting metals at the bottom and the ceramics at the top.These differences A major finding from all the integral tests and also from also have a majot effect on the vessel failure process and the TMI-2 core examination is that the unoxidized Zir- on the characteristics of the melt released into the con-caloy, the control rod materials, and their eutectics melt tainment upon vessel failure. In the drainage case, the before or during the rapid temperature transient from released melt has a lower temperature and a high metal steam oxidation of the core zircakiy and at temperatures content, ranging from as low as 1200'K for the cutectics up to 2200"K for the Zircaloy. Molten metal, which includes in the TMI-2 accident, the melt released on melt-some dissolved UO2 fuel, rehicates downward by gravity through from the blocked core into the water-filled lower to refreeze and form a porous partial core bkxkage, at plenum by the ceramic melt pool contained only about least in PWR accidents with water in the bottom of the 20% of the core mass and had a very low metal content.

core. The low metal content is important because metal NUREG-1365, Rev.1 A-6

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

' Appendix A:

1 provides the potential for oxidation heating and hydrogen dation rates continued after the start of molten zircaloy_

generation after vessel melt-through. A more quantita- . relocation so long as supplies of steam and unoxidized tive understanding of the key processesinvolved in deter. high temperature Zircaloy were available. In tests with mining the melt-through threshold and location of failure llWR core geometry, eutectic interactions among the ,

are needed, however, in order to generalize the applica.

control blade melt of stainless steel and NC and the tion of these relatively benign results. Acquiring this Zircaloy channel box walls failed the walls and opened up information is a major objective of the current melt pro- the geometry to the cross-flow of steam and continuing gression research. A second inajor objective is determin- hydrogen generation as in a PWR. During late-phase melt ing whether metallic, and later ceramic, melt drainage progression in unrecovered accidents, oxidation and hy-from the core without core blockage may occur in llWR drogen generation cannot be significant because of the accidents in which the blowdown from automatic depre- low metal content and the compacted geometry of the-ssurization (ADS) lowers the water level below the core, hotter region of the core around the growing ceramic melt and core heat up occurs in a " dry core" with very low pool.

steam flow.

De severely damaged fuel bundle was reflooded in the A summary of the current state of phenomenological LOFI' FP-2 test, and most of the oxidation and hydrogen understanding of melt progression is given in Table A.2.2. generation in the test was produced by steam during There is reasonable understanding of the significant phe, reflooding. A quantitative understanding of the rates in-nomena involved in fuel damage and early phase (metallic volved, however, does not yet exist. When Zircaloy+

melts) melt progression, except for the processes of me. containing melt drops into lower plenum water, there can tallic melt relocation and blockage formation. For the late also be a small amount of oxidation and hydrogen genera.

phase (ceramic melts), there is, in contrast, only a general tion from the unoxidized Zircaloy in the melt, particularly understanding that is based mostly on the B11-2 core tf a steam exp!osion occurs.

examination and analysis. Significant information on re-As core melt material relocates into Ihe lower head of the flood effects has also been obtamed.

reactor vessel, the major concern of severe accident-During a severe accident, hydrogen and heat are gener- an lysis becomes the mode and timing of lower head ated during the exothermic chemical reaction (oxidation) failure. The research program m this area includes the of steam with core Zircaloy (and possibly some stdmless analysis and examination of samples from the_TM1-2 steel). De integral severe fuel damage (SFD) tests that I wer head. Failure mode analyses have been conducted have been performed have provided data on Zircaloy to examine failure by the ejection of a vessel penetration, oxidation and hydrogen generation during severe acci- failure of a penetration ~ outside the vessel shell, and dents in addition to the data on core degradation and melt gl bal and local creep- rupture failure of the vessel shell.

progression that were discussed earlier. These tests were Ih tadure mechanisms have been evaluated for debris conducted over a wide range of conditions in both BWR and thermal-hydrauhc conditions estimated for current and PWR core geometries. In addition, separate-effects Il%,R and PWR designs. Results of these analyses are-experiments on Zirealoy oxidation in steam have fur- reported in draft NUREG/CR-5642," Light Water Reac-nished basie reaction rate data and correlations.nc rates tor Iwwer Head Failure Analysis," that was issued in of Zircaloy oxidation and hydrogen generation are well December 1991.

known as long as intact core geometry is maintained. At Typical severe accident scenarios have been used for each i lower temperatures (below about 1700'K), the reaction of the reactor types to develop lower head failure _ maps; rate is limited by oxygen diffusion through the growing The failure maps show, as a function of system pressure -

protective ZrO 2layer formed by cladding oxidation, and and vessel inner wall temperature, the mechanisms by the rate is limited by the availability of steam and Zircaloy . which a vessel is most likely to fail. The studies have at higher temperatures.The rate limit from diffusion is shown that important parameters inciude not only the -

well described by a parabolic rate law with an Arrhenius system pressure and vessel temperature, but also the (exponential) temperature dependence. Correlations of effective flow area and wall thickness of the penetrations the experimental data on oxidation rates arc incorporated and the size of the annular gap between a penetration into the MELCOR and SCDAP/RELAP5 codes as well tube and the vessel wall.

as into other severe accident c(xles.

Under conditions that result in low heat-up of the vessel Oxidation continues as the intact geometry is lost through wall, the most likely mode of failure is ejection of a pene-relocation of molten unoxidized core Zircakiy downward tration because of failure of the vessel seal weld. In this to cooler regions of the core, but the oxidation rate be- case, the friction between the penetration tube and vessel comes less well known.This is because of a lack of knowl- opening is so low that tube ejection can occur without edge of the actual geometry, the thickness of new protec- tube rupture ordamage. As the temperature of the vessel tive oxide layers, and the locally available steam flow. In wall rises sufficiently to restrain the penetration tube by all the integral severe fuel damage tests, substantial oxi- friction, failure may occur by melt material penetration of A-7 NUREG-1365, Rev. I

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

Appendix A I

! the vessel through the tube, with subsequent melt- juncthely with direct contamment heating or steam pres-through or rupture of the tube outside the vessel bound- suntation. Hydrogen combustion can aho impact the ary.'I wo modes of melt penetration have been studied for source term by altering fission product chemistry and tlus case: conduction-hmited penetration that is gov- resuspendmg hssion products in aerosol form, flydrogen cined by a freezmg annular shcIl and bulk free /ing pene- transport is a saf ety issue for operating reactors primanly tration. Conduction limited penetration has been found insofar as the mixing of hydrogen determines the nature to give the greater melt penetration in all cases. Molten of subsequent combustion. In the event hydrogen re-debris was also found to be more hkely to penetrate leased into the contaimnent during a severe accident ac.

through a tube with a large effective diameter, such as a cumulates without igniting but mixes rapidly throughout ilWR instrument tube or a HWR drain noule, rather the entire volume, the global concentrations in most in-than through smaller diameter PWR instrument tubes. stances will remain below the hmits for detonation. If mixing does not occur because of stratification or pocket-I adure of the vessellower head by creep rupture may be ing in enclosed areas, those richer mixtur es that occur, at possible when the system pressure and vessel tempera- least hically, present a greater likelihood for flame accel-ture are suffaciently high.The failure mode can be either a cration and detonation. For advanced reactor designs _

global failure of the hemispherical head, circumferen- without active mixing systems that rely on passive contidn-(tally around the vessel below the debris surface, or hical ment heat iemoval from the containment atmosphere to bulging and membrane rupture of the shell. Vessel fail- the containment shell (AP600) or to an external isolation ure analyses to date have used assumed debris conditions condenser (SilWR), the transport of hydrogen within the to drive simple thermal analyses. Dimensionless groups containment may influence the userall heat removal ca-derived from these analyses may now be used in further pability of those related safety features, detaded analyses with severe accident codes such as SCDAP/REl.AP5. An analysis for the local bulging case A32 Status has also been developed m which h>calized vessel wall heating may occur as a result of jet impingement of mol- Research conducted world-wide over the past 12 years ten core material on the vessel wall. has extensively investigated a number of issues related to hydrogen combustion and transport during severe reactor No additional research is planned on hydrogen genera- accidents. Much of the work, performed to experimen-tion during in. vessel melt progression. Ilowever, analyses tally investigate the design and evaluation basis for reac-that are under way of oxidation and hydrogen generation tor containment performance, focused on global defla-from refhaling in the 1.Ol T FP-2 test will be completed. grations of premixed volumes of hydrogen, air and steam.

Since containment analysis generally presumed global deflagrations were the mode of combustion (this assump-Plans for lower head failure analysis include a peer review tion could be contrived to produce app.opriately conser-in FY92 of draft NURl!G/CR-5642. The results of the vative h>adings), research results were instrumental in -

peer review will be incorporated into a final NUREG establishing the dominant parameters (flame speed and -

report. combustion completeness) influencing the peak pressure from volumetric deflagrations. Diffusive burning of hy-There are some significant uncertainties related to core drogen has also been the subject of experimental research melt progression phenomena, primarily in the late (ce- conducted by luth by the NRC and the industry, in a ramic melt) phase. I uture plans to address these uncer- cooperative hydrogen research program conducted at the tainties are discussed in Section 22. doi! Nevada Test Site (NTS), hydrogen and hydrogen-steam jets and plumes in ratios intended to represent severe accident blowdowns were ignited by thermal ig-A.3 Ilydrogen Combust.mn and niters to studv diffusive burning behavior. Diffusion flame Transport research has' also been carried out at Sandia National laboratories and at other research facilities, including A3.1 llackgrouud the large scale facdity at Factory Mutual Research Corpo-ration, used to investigate hydrogen mixing and combus-The safety significance of hydrogen combustion dunng a tion in a Mark Ill containment.

ses cre accident for non merted containments is that the concomitant energy release manifested as pressuritation Recogniting that combustion modes include supersonic and heating of the containment atmosphere could pose a as well as subsonic flame propagation the NRC has also threat to containment integrity or to the survival or func- sponsored considerable research on the detonation of tioning of essential safety equipment. When hydrogen hydrogen-air and hydrogen-air-steam mixtures. The in-combustion alone is insufficient to threaten containment ternational reactor safety research community has also integrity, combustion may still represent a significant con- contributed to the data base on the detonabihty of various tribution to containment loadings when considered con- gaseous mixtures.This research clearly identified the sen-N URiiG -1365. Rev.1 A-8

Appendix A .

sitivity of scale in establishing the limits for detonability of propedies of materials extracted from the lower head of mixtures as well as providing insights to the mechanisms the TMI-2 reactor pressure vessel, (2) determine the for flame acceleration and transition to detonation.'lhe extent of damage to the lower head by chemical and

- range of detonable concentrations for hydrogen air mix- thermal attack, and (3) determine the margin of st ructural tures has been shown experimentally to be much wider integrity that remained in the pressure vessel.  ;

than the classic limits of approximately 18 to 60% estab-

. lished in small-scale testing at ambient temperatures.

'Ihe range of detonable concentrations has been shown A.4.2 Status experimentally in our research to be as wide as 11.6 to 74.9% hydrogen at 20'C and 9.4 to 76.9% hydrogen at Under the VIP program,15 reactor vessel steel speci-100*C with the limits depending on scale, geometry, and mens,14 incore noules, and 2 incore guide tubes were temperature. 'the view has been that steam greatly re. successfully extracted from the lower head over a 30-day duced the mixture's detonability, but there is analytical period ending March 1,1990. The vessel steel samples evidence that increasing temperature may make steam then were decontaminated, sectioned, and distributed to mitigation less elficient. In conjunction with hydrogen the United States and seven other participating countries combustion research. the research community has also for mechanical and metallographic examinations. Also, experimentally explored the issue of hydrogen transport the noules and guide tubes were cut and distnbuted for and mixing. While the NRC has not exclusively sponsored examination to INEl, ANL. and the CEA in Saclay, significant experimental research on hydrogen mixing. France.

our joint research agreement with Germany has provided mixing test data in the complex geometries of the Since the extraction of the test specimens in 1990, sub-11attelle Frankfurt and llDR test facilities. Industry data stantial metallographic examinations of the vessel steel from the HEDL facility and the FMRC one-fourth scale samples have been completed, including microstructural facility has supplemented that data; although this data examinations and hardness measurements; Results of was specifically related to ice condenser and Mark 111 these examinTtions have provided preliminary estimates containments under certain accident conditions. To com- of temperature histories of the lower head samples.

piement the experimental or phenomenological re- 'Ihese results show that the maximum innu surface tem, search, the NRC has sponsored the development and perature of at least four samples reached 1050-1100*C application of computer codes for hydrogen mixing and for about 30 minutes /these samples were extracted from combustion analysis, most notably the IIECIR and IIMS the lower head from a small region, approximately 2 to 3 codes. 'lhe 1It!CIR code, which employs control volume feet in diameter.The examinations alsoindicated that the modeling of the containment, and the llMS code, a finite temperature 2 inches into the wall was about 100'C lower difference code, were developed to provide varying levels than the inner surface temperature. (Ihe lower head of resolution of the containment volume for mixing analy- thickness was 5 inches.) Mechanical testing is under way

, sis. HFCIR combustion modeling has been subsumed to determine tensile and creep properties of the vessel .

into the more general CONTAIN and MELCOR codes. steel at high temperatures (up to about 1100'C).

Metallographic and scanning electron microscope (SEM)

A.4 TMI-2 Vessel Investigation Project examinations of the instrument tube noules are being performed to assess the noules' axial temperature profile and interaction of the Inconel 600 noule material with A.4.1 Ilackground- molten core materials, Examinations of previously mol-ten thermocouples in the nonles may be an important Most current knowledge of in-vessel severe accident be, indicator for determining the axial temperature profile in havior has come from experiments such as the series of the nonles.

severe fuel damage tests performed in the Power Ilurst .

Facility. ACRR, NRU, LOFT FP-2, Phebus test reactor, In September 1991, the VIP Management Board decided -

CORA ex. reactor experiments, and from the extensive to amend the original agreement extending the VIP pro-post accident core examination of the Three Mile Island gram from September 30,1991, to March 31,1993. The Unit 2 (TMI-2) reactor, performed by the Department of objectives of the amended program are to (1) perform ~

Energy (DOE). In 1988, the NRC, in cooperation with 10 more detailed testing and examination of the in-core in-foreign countries under the auspices of the Organization strument tube nonic penetrations and the in- core instru-for Economic Cooperation and Development's (OECD) ment guide tubes that were extracted from the lower Nuclear Energy Agency (NEA).~ undertook a follow-on head,(2) perform additional analyses of potential reactor program to the DOE TM1-2 examinations. The objec- vessel failure modes based on data from sample examina-tives of this program, called the TMI-2 Vessel Investiga- -tions, and (3) assess the margin-to-failure of the lower tion Project (VIP), are to (1) investigate the condition and head of the reacter vessel.

A-9 NUREG-1365, Rev.1

- Appendix A A.4.3 Future Plans tions or can significantly alter the core melt progression scenaho.

1(esults of the ongoing TMI-2 lower head examinations are expected to provide adJitional information on the Certain fundamental design differences between PWRs physical properties of the specimens, temperature distri- and llWRs require different areas of emphasis. During butions in the instrument nonles, and interactions be- the later phase of core melt progression in PWRs. the tween the molten core material and the vessel. These potential exists to accumulate lar);e quantities of melt in results then will be used to perform scoping analyses of the core region " crucible." llecause of the largely open potential reactor vessel failure modes, such as penetra. lower plenum geometry, in-vessel FCis can span the tion tube failures and global or kical failure of the reactor whole range from energetic to relatively benign. For vessel lower head. More detailed analyses of the most ilWRs, on the other hand, if a metallic or ceramic bhack-likely failure mechanisms will be performed to estimate age does not occur, large melt accumulations in the core the margin to-failure of the lower head. A final project region could be excluded. (The research effort on core report, integrating the results of all the sample examina. bhickage or melt drainage in llWR severe accidents is tions and analyses, will be issued at the comp!ction of this discussed in Section 2.2.2.1 of this report.) In addition, prograrn in June 1993, June 24,1992 the lower plenum is crowded by the control rod drives.

'therefore, small energetic FCis are more likely, and the emphasis for llWRs is on coolability.

A.5 Fuel-Coolant Interactions (FCI) During the ex-vessel stage of a severe accident, water may 3Hd Debris C00 lability be added to the reactor cavity as a deliberate measure to mitigate the consequences of core debris-concrete inter.

actions, Water may be present in the cavity as a teJult of A.5.1 llackgroumi natural processes such as the condensation of steam, flow from containment sprays, or discharge from accumulators One of the fundamental phenomena in the course of a in the reactor coolant system. I or all containment geome-severe accident is the interaction of degraded core mate-rials with coolant. In the absence of coolant, the core tries, various FCI concerns exist that span the whole range rom energetic interactions to benign coolability.

materials overheat, melt, and rehicate. 'lhe opportunity

for fuel- coolant interactions arises not only as a natural consequence of this reh> cation process (into areas occu- A.5.2 Status pied by coolant), but also as a consequence of accident There are three specific issues addressed in this section

management actions tit Jd water to regions previously depleted of coolant. C sidering the variety of reactor L FCI energetics, geometries, meltdown scenarios, and timmg of coolant addition, a broad range of resultant phenomena is possi' 1 Fuel melt quenching in water pools, and -

bit

3. Adding water to degraded core (i_n-vessel and ex-

'lhe intensity of the interactions and resulting material vessel),

state and configuration depend on: (1) the geometry of the region, (2) the masses of core material and coolant A.S.2.1 Status of FCI Energetics mvolved. (3) the thermodynamic state of the materials, and (4) the mtes of pouring of molten core rnaterials into This topic is of primary relevance to in-vessel interactions a pool of coolant or the rate of flooding a degraded or in PWRs. The first quantification of the potential for molten core. For example, at one extreme, a slow pour of alpha-mode containment failure was offered in

, molten core debris in a large, deep pool of water can lead WASil 1400. Additional quantification of this failure to complete quenching and formation of a coarse particle mode can be found in the expert opinion efforts, i.e.,

debris at the pool bottom. At the other extreme, a rapid, NUREG-Ill6, " Steam Explosion Review Group A re-massive release could yield a highly energetic exph>sion view of Current Understanding of the Potential for Con-with significtmt mechanical consequences on the system tainment Failure Arising from In-Vessel Steam Explo-and its surrounding structures. Initially, the subject of sion,1985,"and NUREG-1150," Severe Accident Risks:

FCis emphasized the phenomenon of a vapor-explosion. An Assessment for Five U.S. Nuclear Power Plants, induced missde as a possible mode of containment fail- 1990." It is generally agreed that the probability of alpha-ure. With increased emphasis on accident management, mode containment failure is negligible, although the need

' interest in FCis broadened to include core degradation to improve the quantification basis for this conclusion is regimes that can be quenched by flooding with water. For generally acknowledged. NURiiG/CR-5030 presents a such scenarios, it is important to identify potential cir- probabihstic framework that arrives at the same conclu-cumstances in which FCis can Icad to coolable configura- sions.This report highlights " premixing" as the process of N UREG- 1365, Rev.1 A-10

t Appendix A i

primary impartance in limiting the magnitude of ener- He phenomena and technical issues for these configura-getic FCis. tions are quite different and are addressed separately below.

He issue of premixing arises because there is a possibility that core relocation into the lower plenum can occur in A.5.2.3.1 Adding Water to a Degraded Core -

large massive pours in PWRs.He geometry of the PWR 0"4'*D diffuser plates leads to the breakup of the melt into

, A review of past experiments suggests that a few tests several smaller j,ets. Decause of extensive steammg, large have been performed with water addition following bun-premixtures, which are expected to develop m the above die degradation, mainly, CORA, PilF, and the LOFT-circumstances, would be largely void of water, and there- FT'2 test. Also, simulant tests of a coolant added to fuel fore, the magnitude of energetics is limited.

debris have been performed at IIN1, ANL, and UCLA.

Although limited in scope, these tests address water addi.

tion during core degradation. The major variables affect-A.5.2.2 Status of Fuel. Melt Quenching mg the consequences of water addition are the rate and character of water addition and the state of the fuel at the Fuel-melt quenching in severe accident evaluations has time water is added. Reference states for the fuel are:

several important aspects. I or the in-vessel core melt portion of severe accident sequences in current genera-tion PWRs and BWRs, as well as their ALWR counter-

1. Initial heat up and early core degradation parts, quenching considerations are fundamental in pre

• His first stage of a severe accident sequence is de-dicting the mode and timmg oflower head failure and the fined as beginning with the onset of core uncovery -

resultant impact on direct contamment heatmg (DCII).

and continuing up to the clad melting (not cutectic In accident management, increasingly more considera-dissolution) and relocation. Coolant is still easily tion is given to the possibility of flooding the cavity region accessible to the degraded portion of the core be-external to the lower head of the reactor vessel m order to prevent vessel failure and retain core debns m the vessel.

cause of efficient lateral flows through the open '

PWR lattice. Key modeling difficulties include va-For the ex-vessel portion of severe accident sequenecs, por chimney effects and permeability.

quenchm, g plays an important part m determmmg the ex-vessel relocation behavior of the melt, and hence,

2. Advanced core degradation (core rubble, melt, and long-term coolability.

- relocation)

The fundamental difficulty in addressing the role of This stage of the severe accident sequence is defined ;

quenching in severe accident scenarios relates to uncer- as beginning with the first significant fuel melting taintiesin(l)the flow characteristicsof the pour (i.e., size _ and relocation and extending through core debris and number of melt jets, pour rates) and (2) the composi- relocation onto the lower plenum. The key events in

! tion of the melt (i.e., metallic vs. oxidic components)and the late stage are the formation of a molten pool, its temperature (superheat). Conservative evaluations of pool growth, and breakthrough of the crust support such uncertainties and of the quenching process itself, ing the molten pool. The behavior following watt particularly on assessment of containment integrity, are addition during this stage is governed by crust stabil possible. Ilowever, a better understanding of quenching ity, which is domina!ed by the heat transfer proper-will contribute to the depth of understanding such that ties of the (porous) medium surrounding the crusts better judgments, especially on new plant designs, can be and of the crusts themselves, and the heat load dis-made in the future, tribution on the inner surface of these crusts by natural convection of the molten core material A.S.2.3 Status of Adding Water to Degraded Core The reflooding mode refers to two different configura-tions: The continued progression of a severe accident can lead -

to the expulsion of reactor core debris into the reactor L cavity. Debris in the reactor cavity can interact with the -

Degraded core configuration in which water is sup-snetural concrete of the containment and even,in_some plied from above or below the core, or cases, directly with the pressure boundary of the contam-ment.

- 2. A relatively shallow molten corium pool (at the bot-tom of the lower plenum or on the concrete The configuratior of core debris that is expelled from the basemat) with water on top of the molten pool. reactor vessel does not ensure that the mere presence of A-11 NUREG-1365, Rev.1

r

-Appendix A water will result in cooling the debris sufficiently to climi- issues of debris c(mfiguration and issues of heat removal .

nate core concrete interat:tions. '!he core debris must be from core debris by water.

fragmented into coolable rubble, or it must spread over a large enough area that heat extracted by overlying water t will cool the debris. 'lhe NRC is participating in the MACl! test program, which is examining the effects of water on uranium-At this time, a technically justified criterion for debris diexide-rich melts interacting with concrete, in the past, coolability during the ex-vessel phases of severe reactor the NRC has sponsored tests of high temperature oxidic  !

accidents cannot be defined. Issues that must be resolved and metallic melts with water and concrete (the WIrl'.

. - to define such a criterion can be bundly categorized as COR tests) to determine limits of coolability.

i I

l-I NURl!G-1365, Rev.1 A-12

APPENDIX 11 CLOSURE OF SEVERE ACCIDENT ISSUES 11.1 Source Term ciany early failures,i.e., within a few hours from onset of an accident) or containment bypass can lead to conse-quences that are much greater than those associated with 11.1*1 IluckE round a TID-14844 release into containment when the contain-Radionuclide releases to the environment, that is, the ment is assumed to be leaking at its maximum leak rate for type, quantity, timing and energy characteristles of the its design c4mditions. Indeed, some of the most severe release of radioactive material from reactor accidents source terms anse from some contamment bypass events,

(" source terms") are deeply embedded in the regulatory such as " event V" and multiple steam generator tube policy and practices of the NRC. For almost 30 years the ruptures.

NRC's reactor site criteria (10 CFR Part 100) have re-Source term estimates under severe accident conditions quired that for licensing purposes, an accidental fission began to be of great interest shortly after the Three Mile product release from the core mio the containment be Island 2 (TMI-2) accident. The objective of a major NRC postulated to occur rmd that its radiological consequences research effort on source terms that began about 1981 is be evaluated assuming that the containment remains in-to obtain a better understanding of fission-product trans-tact but leaks at its maximum allowable leak rate.

port and release mechanisms in 1 WRs under severe acci-dent conditions. This research effort, which involved a I! valuation of the consequences is used to assess both number of national laboratories as well as nuclear mdus-plant mitigation features such as fission product cleanup try groups, has restated in the deselopment and applica-systems and the suitability of the site,'Ihe characteristics dn sana zw mmputa s (munine mm meh of the containment " source term," which must be distin-phenomena and associated source tenns. Work has also guished from a release to the environment, are described included sigruficant review efforts by peer reviewers, for-in Regulatory Guides 1.3 and 1.4, but are derived from e gn partners in NRC research programs, industry the 1962 report TID-14844 (Ref.1). The source terrn groups, and the general public. Current risk assessment consists of 100% of the core inventory of noble gases and methods, including the latest research efforts on severe 50% of the iodines (half of which are assumed to deposit accident source terms, are reflected m NURl!G-1150 on interior surfaces very rapidly). Regulatory Guides 1.3 (Ref. 3), which provides an assessment of severe accident and 1,4 also specify that the source term is instantane- fm nuclear power plants, Finally, the occur-ously available for release and has significantly affected rence of the acen; dent at Unit 4 of the Chernobyl reactorin containment isolation valve closure times. The guides the Soviet Union on April 26,1986, and the large acciden-also specify that the iodine is predominantly (91%)in tal release of fission products resulting from it has pro-t -

clemental (12 ) form. vided further impetus to understand severe accident l

so m e nus as w as to premt such munences.

In addition to plant mitigation features and site suitabil-ity, the regulatory applications of this release also estab-lish (1) the post-accident radiation environment for which

!!.l.2 Status safety-related equipment should be qualified,(2) post ac- Since shortly after the accident at TMI-A the NRC has cident habitability requirements for the cont ro', room, and sp(msored numerous experimental and analytical re-(3) post. accident sampling systems and accessibility, scarch projects on fission product release and transport.

Early experiments and analytical work tended to focus on f a contrast to a specified source term for design basis release from fuel material under high temperatures and accidents, severe accident source terms first arose in severe accident envimnments 1.ater, experimental data--.

probabilistic risk assessments (e.g., Reactor Safety Study, on the behavior of acrosols in the RCS and the c mtain.

WASil-1400, Ref. 2) in examming accident sequences ment were obtained. These data were used to develop that involved core melt and possible containments fail

• acrosol deposition and transport models to nnalyze fission ure. Severe accident source terms represent mechamst3- product behavior in the reactor coolant system and the cally determined "best estimate" releases to the environ- containment. Currently, fully integrated models are be-ment, including estimates of failures of containment ing assembled into individual codes, the VICTORIA code integrity. This is very different from the combination of (Ref. 4) and the CONTAIN code,(Ref. 5), for the analysis the nonmechanistic release to containment postulated by of in-vessel and ex. vessel source terms, respectively.

TID-14844 coupled with the assumption of very limited containment leakage used for Part 100 citing calculations The issue of revaporization of deposited radioactive ma-foi design basis accidents. The worst severe accident terials in the RCS is of particular concern because of the source terms resulting from containment failure (espe- instability of the deposited radioactive material at high 11- 1 NUREG-1365, Rev.1 -

Appendix 11 temperatures induced by decay of nuchdes. Itecent se- tions and associated radionuclide releases are modeled by vere accident calculations have predicted that, for some the stand-alone code, COllCON-N10lO (Itef 6)flhese sequences, natural circulation of gases through the reac- models will also be incorporated into t he CONTAIN code for core may also heat structures m the itCS to substan. to allow cornprehensive evaluation of containment be-tially higher temperatures than had been previously pre- havior durmg severe accidents. In addition, the CON.

dicted. The high temperatures may also result in the TAIN c de models the thetmal hydrauhes (pressure, failure of itCS piping at certam h> cations.This failure of temperatures, etc.), acrosols behavior, fission products itCS piping is of some significance since current analyses behavior and transp>rt, and hydrogen behavior in the of core degradation indicate substantial fractions of the containment. Codes such as VICI Olt! A, CONTAIN,and core could be retained wtthin the itPV after vessel fail. COltCON-MOD 3 incorporate mostly mechanistic mid-ure. lividence from Thil-2 suggests that as much as half els for severe accident phenomena, thus allowing the the core matenal may have stayed within the original examination of coraplex source term issues in a detailed confines after the rest of the ccre had melted and drained and systematic fashion. On the other hand, the hilil Colt into the lower plenum. This remaining fuel in the core code (It ef. 7) was developed as an integral tool for analysis region could be exposed to air once the plenum has been of fission product transport in the ItCS and in the contain- _

brcached. Air will react exothermically with the claddmg ment, hil!! . Colt employs simpler rmdels for severe acci-remaimng on the fuel, pnducing high temperatures in dent phenomena in order to facilitate the fast running this fuel. Once the cladding has been oxidi/ed, vapors of time requirements for the repetitious calculations used in radionuclides not usually considered highly volatile, nota. Pit A analysis.

19y ruthemum and molybdenum, will be produced be-cause of the otrongly oxidi/cd conditions. 'lhe VIC1'O- With respect to the potential release of iodine from sup-i<lA ctde has incorporated a model to address pression pools and reactor cavity water, the ACl! program revaporization caused by an increase in temperature of and the OltNI. udine chemistry research have provided the ltCS. 'lhc VICTOltl A codc also mod < ls other impor- extensive experimental data to address this issue. At tant severe accident phenomena such as the release and OltNI, the research included iodine partition coefficient transport of fission pnxlucts, condensation of vapors, tests, hydrolysis chemical kinetics tests, radiolysis chemi-aerosols behavior, and chemical reactions in the itCS. cal kinetics tests, hydrogen burn / iodine chemistry tests, and TitliNDS models development. 'these efforts were further enhanced by the ACl! program, which included in low-pressure accident scen trios in which the reactor hygroscopic acrosol/ iodine chemistry tests, and hydrogen vessel fails, high-temperature core debris may fall into burn /iodme chemistry tests. hiany iodine chemistry mod-the reactor cavity where it interacts with the concrete. At els were developed from these data bases. 'lhese models high temperatures (approximately 1,300-1,500 C), con- are now being incorporated into the CONTAIN code.

crete decomp >ses, and the ablation pnslucts commonly 1:urther validdtion of the CONTAIN code could be done mclude water vapor and carbon dioxide as well as the using the PliliBUS-FP data.

refractmy oxides Ca0 and SiO 3 The liquefied oxidic -

compments of the concrete mix with the uranium oxide fuct and metallie oxides of the debris /l)pically, the core 11,1.3 h,ource ,I,crin Uncertm,nt,es i and Present debris is imtially all or partially molten; gases released at llesearch Efforts the debris-concrete interface bubble through the debris pool reduemg some low-vapor-pressure oxides such as With res[]ect to source tenu uncertainties, NUltiL,-

1150 has identified specific source term issues as con-14 0 3to high-vapor-pressure forms such as 120. These tributors to uncertainty in risk estimates. For fission pnd-more volatile forms then vaporize into the bubble vol.

ume, thus releasing fission pnduct species that were not uct release from fuel and retention in the llCS, a key released in the vessel. Aerosols are formed when the question is, "llow significant is fission product release bubbles exit the upper surface and f ragment. Among the during the late stages of core melt vs. the early phase?

factors that influence the magnitude of the ex-vessel re-leases are the composition and temperatures of the core lhe present theoreticat model for the late-stage rubble debris. Concre composition also has a major impact on bed (sigmficant relocation and melting of ceram,cs)as-i the amount of aerosols entrained into the containment sumes that release is governcd by gas-phase mass trans- ,

atmosphere.1.imestone concrete produces larger gas port. For Ihe molten pool, the main mechanism for fission flows and is more oxidi/ing than basaltic concrete. An pnduct release is governed by diffusion and surface con-extenwve experimental data base has been obtained on vection. Iloth of these theoretical models need experi-core concrete interactions with no overlying water pool. mental data for validation. The predominant sources of If an overlying water pocol exists, a considerable amount of uncertainty in these theoretical model are:

the aerosols may be scrubbed and kept out of the contain-e geomury of the core dehns, ment atmosphere. Itesearch effort is under way to obtain expenmental data for this case. Core-concrete interac- o magnitude of gas fluxes through the debris, and NUltliG- 1365, llev.1 H -2

Appendix 11 o thermo-chemical properties of the high- pose to produce I11 or 12, there will be little or no retained temperature vapor species that vapori/c from the iodine in t he RCS to revaporize after containment iailure.

debris. If Csl is stable, substantial amounts will be retained tem.

porarily in the RCS and will be able to revaporize, creat-The VKTOltlA code incorporates models to aJdress ing an iodine source term after containment failure. Tests t hese questions. Uncertainties in ou r current understand- at Wmfrith and SNI, have shown that Cs! will react with ing of the evolution of accidents are handled in a para- boron oxides vaporized into the RCS atmosphere to yield metric fashion. cesium borate,1, and 111. Tests at SNI. have shown ther-malinstability of Cslin the RCS environment.These tests Integral experiments to study fission pnduct release have not clanfied whether I or til produced by reactions from late-phase core melt progression and revaporization of Csl can subsequently react to form other iodides such are not being planned in the United States. Ilowever, the as Nil 2(g,c). Analyses done in the U.K. suggest that high PillillUS-FP project could provide some of the needed vapor fractions of iodine (CsOli and Cs!) at the time of experimental data. RCS failure could yield acrosols in containment that do not settle rapidly and are only slightly affected by contain- _

Por late-phase revaporization of fission product from the ment sprays, llence, a source of airborne fission product RCS, the key questions are: would be available for leakage out of the containment. -

1. After the reactor vessel has been breached and air Currently, there are no suitable experimental data to ingress occurs, what are the release rates from the validate revaporization models, but the PillillUS-FP fuel remaining in the vessel? Likewise, for shutdown tests could provide some. It is likely that continued exa-accidents, what are the release rates from the fuel mination of the chemical form of I or 111 produced by exposed to air? reaction of Cs! could alter the perceptions of risk by altering the predicted amount of suspended radioactivity

2. What chemical forms are important during the in containment at the time of containment failure '!his transport and retention of aerosols and vapors? could be accomplished by sensitivity analysis using VIC.

TORI A.

As mentioned earlier,if the vesselis penetrated, air from .

the containment atmosphere will circulate over retained it is also important to utilue risk perspectives regarding fuel. The air will react exothermically with the cladding the uncertainties in late-stage core melt and late-phase remaining on the fuel, producing high temperatures in rev pori /ation. Generally, risk is mercased by the early this fuel. Once the cladding has been oxidized, vapors of release of fission products into containment and carly .

radionuclides (notably ruthenium and molytdenum, ra. containment failure or bypass. lience, additional fission dioactive species not usually thought to make major con. products released later in an accident phase will denote tributions to severe accident source terms) will be pro- Imr rek ases a3 an earlier stage. Modelmg sensitivity ,_

duced because of the strongly oxidized conditions. Tests c n be made m risk assessments that can test uncertam- _

performed at Chalk River laboratory in Canada with ties and their implications.

uncladded and cladded fuel, and at ORNL with fuel frag- A key question regarding the ex-vessel source term is, ments under highly oxidized conditions, showed that a "What effect does hydrogen combustion have on acrosols large V!-7 fraction of the ruthenium, tellurium, and mo. suspended in the containment atmosphere?"

lybdenum was released. 'the predicted release rates are dependent on a highly oxidized condition for the fuel, Acrosol materials containing Csl must be dehydrated and f lowever, it is necessary to conduct tests for radionuclide vaporized before chemical interactions of cesium iodide releases for typical 1.WR fuel, as opposed to the thinner can be expected. Tests conducted at ORNL, as part of the cladding used in the CANDU fuel. Plant configurations ACli program, found that vaporized Cs! was unstable in during shutdown situations may lead to the possibility of hydrogen flames with the iodine redistributing as iochde, air ingress into the core either by natural circulation or 1, 2 and iodate, An excess of metal cations (Cs) reduced the from the residual heat removal system.That air will react extent of 12 formation. Oxidation to 1 2was consistent with exothermically with the cladding on the fuel, producing thermal decomposition cf Cst, but iodate praluction was high temperatures in the fuel, resulting in massive vapori- related to the nonequilibrium Oli and O radical concen-zation of ruthenium, tellurium, and molybder.am. A test trations found in hydrogen / air flames. Data are believed (VI-7) will be conducted at ORNL for fission-product to be adequate to address this question. However, the release at high temperature m an air enviromnent, the presence of other acrosol species in the containment experimental data could be readily incorporated into the atmosphere, such as acrosols produced by core-concrete current VICTORIA fuel-release model. interactions, could affect the stability of Csl during hydro-gen combustion events, even though it is a small contribu-

'lhe chemical form assumed for inline affects late-phase lion. Por instance, sihca could trap cesium to form cesium revaporization. If cesium iodide or other imhdes decom- silicate so that it cannot recombine with iodine. Other 11 - 3 NU RFG-1365. Rev.1

. ._ . .. . _ _. _ _ . _ _ . _ _ - _ _ _. _._m mm Appendix 11 basic species could react with iodine produced in the gradation prior to reactor vessel failure, and finally, re-combustion to reform iodides: lease of fission products from core-concrete interactions.

Na2O(c) + 2111(g)- >2 Nal(c) + 110. 2 Additional nuclides other than the noble gases and iodine are expected to be released. For example, preliminary Currently, the TRENDS model (Ref. 8) treats this in a indications are that the fraction of core inventory of ce-conservative fashion (i.e., it assumes that lewill be formed sium released into the containment is generally compara-in a hydrogen burn). Such a model has been developed for ble to that of iodine, in addition, some tellurium and ,

inclusion in the CONTAIN code, smaller fractions of the remaining nuclides are also ex-pected for release, in conclusion, although additional physics and chemistry research to reduce uncertainties in source term phenom. A recent study on iodine chemical form and behavior on ena can be performed,it Eimportant to consider the need entering the containment from the RCS, and the subse-for such research and the potential that this research quent revolatilization of iodine from water pools in con-could significantly improve our risk perspective associ- tainment, has been completed at ORNL ORNL exam-ated with severe accidents. NUREG-il50 indicates that ined a group of severe accident sequences used in uncertainties in overall estimation of risk are largely NUREG-ll50. These accident scenarios were for both driven by uncertainties in containment performance, pri. high and low. pressure sequences that are risk significant.

marily those associated with estimation of containment For the RCS, the analysis considered the chemical kinet-loads, estimation of containment perfonnance at load ics of 20 reactions of iodine with water, hydrogen, and Icvels beyond Ihe design basis, and estimation of the prob. cesium, and determined the temperature and time when ability and location of containment bypass. chemical equilibrium was established, Once chemical equilibrium was established, an analysis determined the -

11.1.4 Regulatory Applications / implications iodine chemical forms present. In most calculations, io-dine was released from the RCS into the containment as 11.1 .4 .1 Deselopment of Updated Source Term cesium iodide (Cs!) with very smaP amounts of I or HI.

Since the ORNL's study considered a limited set of reac.

Design basis accident source terms have been used in the tions related to Csl, iodine in the form other than Csl -

United States for licensing purposes in three distinct could be released from the RCS. In order for the RCS to -

ways, namely; release iodine in the form other than Cst, a significant fraction of the Cs has to be removed within the RCS.

1. For siting evaluations as required by 10 CFR Part flowever, for the accident conditions analyzed, only a 100, small fraction of Cs was removed within the RCS. If desired, more sophisticated treatment of Csl reactions-

2. l'or defming the radiological environmental condi. could be undertaken with VICTORIA to confirm the-tions for certain plant systems, and ORNL's findings. The ORNL results indicate that the iodine entering containment is at least 95% Cs!,5% as I ,

3. For assessing the effectiveness of plant mitigation and HI, with not less than 1% as either I or HI.This is in - ;I systems, contrast to the iodine chemical form specified by the TID The NRC is presently preparing an update of the source term contained in Regulatory Guides 1.3 and 1.4, making The ORNLiodme research and the ACE program have .

use of current severe accident research insights. This provided data that address revolatilization of iodine from effort is expected to result in changes in data for the water pools. A comprehensive model to estimate the timing of the release, the composition and magnitude of revolatilization of iodine from water pools was developed

> fission product releases into the containment, and the by ORNL Once iodine enters containment,it dissolves in chemicalform of the iodine fission products. A draft of an water pools or plates out on wet surfaces as I . Subse-updated report replacing TID-14844 is expected to be quently, the iodine behavior within the containment de-issued for comment by the first half or CY 1992, pends upon time and the ph of the water solutions. lf the .

ph is maintained at a value of 7 or greater, the amount of -

Rather thim an instantaneous release into the contain. . iodine in solution that converts to la and organic Iodine -

ment, the revised formulation is expected to be stated as a later in the accident sequence will be verylow. If ph is not series of fission product releases into the containment, controlled, radiation levels in water pools are sufficient to each one associated with a particular stage of an accident convert much of the dissolved iodine to elemental iodine or group of accidents. Hence, the revised formulation is for release into the containment atmosphere. Experi-expected to begin with the release of coolant activity, mental data on the revolatilization of iodine from water followed by the release of activity in the fuci gap, the pools resulted from the cvolution of Ph and irradiation release of fission products associated with gross fuel de- under severe accident conditions are expected from the NUREG-1365, Rev. .I H-4

Appendix il Pl11illUS- 1 P project (section 2.7.2) 'lhese data will pro- l(cferences vhic confirmatory assessment of models for ialme revob-tilvation from water p>ols. (1) J. J. IhNunno et al., " Calculation of Distance Fac-tors for Power and Test Itcactor Sites," U.S. Atomic linergy Comtmssion, TID-14844, March 1962.

11.1.4.2 llegulatory implications (2) U.S. Nuclear itegulatory Commission, "Itcactor At the present time, the NitC is pursuing several regula. Safety Study- An Assessment of Accident Itisksin tory ininatives to incorpuute insights from up!ated se- U.S, Commercial Nuclear Power Plants," WASil-vere accident source ter ms. A revision of the NitC's reac- 1400 (NURiiG-75/014), October 1975, tor site criteria (10 CFil Part 100)is being carried out in pirallel with an mterim revision of 10 CFit Part 50. 'lhe (3) U.S. Nuclear llegulatory Commission, " Severe Ac-reactor site criteria will be revised to remove source term cident itisks: An Assessment for Five U.S. Nuclear and dose calculations and to add i equirements in Part 100 Power I'lants," NUltliG-1150, Decernber 1990, for culusion area sue and population density based on those from llegulatory Guide 4.7. At the same time, Ap. (4) T. J. Ileames et al., "VIUIOltIA: A Mechanistic -

pendix A to Part 1(Kl. containing site seismic critena, is Model of Itadionuclide lichavior in the Iteactor also being revised to reflect the latest unden,tanding. Coolant System under Severe Accident Condi.

Source term and dose enteria wdl continue to be impor. tions," NURl!G/ Cit-5545, October 1990.

tant for plant design; consequently, an interim revision of 10 Cl R Part 50 will be canied out in parallel and will (5) K.K. Murata et al., " User's Manual for CONTAIN contain the present source term O e., that from 1.1: A computer code for Severe Nuclear Reactor TID-14844 and Regulatory Guides 1.3 and 1.4). *lhese Accident Containment Analysis",

promised rule changes nie expected to be issued for com.

ment by early CY93. (6) D.11. liradley and D. R. Gardner, "CORCON-MOD 3: An Integrated Computer Model for Analy-sis of Molten Core-Concrete Interactions, Users Updated source term insights arising f. rom the technical l of TID-14844 are expected to be made available Manual," NURl!G/CR-5843 (draft for comment),

ub ate for voluntary use by existing hcensees A final revision of March 1992. Available from the NitC PDR.

10 Cl R Part 50 to incmpora'e updated source terrn and severe accident insights will then be undertaken, with a (7) R. Summers et al., "Ml!I. Colt i10: A computer code for Nuclear iteactor Severe Accident Source prop > sed rule for comment expected to be issued by early Term and Iti.sk Assessment Analyses," NURiiO/

I493-CR-5531, January 1991.

Although regulatory positions arising from updated (8) !!. C. Ileahm et al., " Chemistry and MassTransport [

source term insights remain to be developed, some pre, of hxhnein Containment,"pp.251-266,hoceedmgs liminary implications can be seen at this time, it is clear of the 2nd CSNI IVorbhop on todine Chemistry in that uglated source tenn insights indicate the need for Reactor Safety Toronto, Canada, June 1989.

consideration of nuclides (e.g., cesiurn) in addition to iodine and the noble gases. In addition, revised insights on imhne chemistry callinto question the need for high-effi- 11.2 Core-Concrete Interaction cieng charcoal absorbers (assuming that the Ph is um-trolled postaccident), These can, in turn, impact such 11.2.1 llackground imp >rtant plant systems as fission product cleanup sys' tems, control room habitability, and nilowable contain* Core-concrete interactions would occur during a severe accident only after penetration of the reactor vessel and ment leak rate.

110w of the core debris onto the concrete basemat. De-

~ composition of the concrete from this inter;tetion results Finally, and most importantly, the alove discussion and in the release of steam and carbon dioxide, which may be all recent risk studies have shown the imp >rtance of main. putially reduced to the combustible gases hydrogen and taining contamment integrity under severe accident con- carixm monoxide. As the gases pass through the hot mol.

ditiens in order to assure low risk. This strongly surf,ests ten debris, they spuge small but potentially important that the appearance of a severe accident source term quantities of radioactive elernents from the debriClhese within omtainment should be more closely linked with radioactive nerosols can be released to the containment, the temperatures, pressures, and containment loads and thereby adding to the accident source term. 'the major challenges associated with such releases, rather than an areas of concern associated with core-concrete interac-arbitrary linkage with a single sequence such as a large- tions during a severe accident are the complete penetra-break loss of coolant accident. tion of the basemat and the generation of radioactive IL5 NURl!G-1365, Rev. I

_ _ _ . _ . .m___ _. _ _ _ _ _ - .. __. ___ _ _ __

' Appendix B acrosols and combustible gases. Another related concern ment atmosphere. Also, the presence of metallic zirco-is the overheating of important structures inside the con. nium produces a chemically reducing erwironment that tainment. can increase the release of certain key elements.

1hc NRC has conducted an extensive program of analytic Many of the early tests of high-temperature melts inter-and experimertal research to obtain improved under- acting with concrete were of an exploratory nature that '

standing of core-mncrete interactions. The analytical were undertaken to identify phenomena or to develop research focused on the development of models for study- experimental techniques. More recent tests have been ing phenomenological aspects of core-concrete interac- conducted to explicitly validate the models of core-con-tions such as heat and mass transfer, while the experimen- crete interactions. Such tests have been heavily inst-tal research focused on conducting scaled-down rumented to obtain heat balances and data on gas genera-experiments simulating prototypic reactor accident sec- tion, gas composition, and aerosol production.

narios. %cse studies have recognized the variety of con-cretes used in nuclear power plants in the U.S. and the lhe available data base spans a broad range of conditions -

widely diverse accident scenarios that lead to core- as well as concrete types.The data base includes tests with -

concrete interactions. The effort to understand core- both oxidic and metallic debris. There are some limita-concrete interactions was also broadened to include a tions to the data base (i.e., the data base _is of a generic reassessment of the models used to predict radionuclide nature and may not address specific issues that arise at release. particular nuclear power plants). However, model predic- -

tions do not indicate that there are substantive uncertain.

He efforts to resolve the severe accident issues associ- ties or issues affecting core-concrete interactions. Some ated with core-concrete interactions have culminated in discrepancy has been traced to the material phase rela-the development of the CORCON computer code.The tionship used in the CORCON code, Measurements of predictions of the catly versions (CORCON-MODI and the melting properties of UO2 -ZrO 2concrete mixtures CORCON-MOD 2) of this computer code have been are being sponsored by the NHC, aad improved models -

validated against many large. scale tests of the interac- are being incorporated in the mort recent version of the tions of both metallic and oxidic core debris and with code CORCON-MOD 3.

concrete.

With regard to the radionuclide release during core-con-11.2.2 Status crete interactions, an effort has been made in the devel-opment of CORCON-MOD 3 to include a detailed The experimental da ta basc on core-concrete interactions mechanistic model of acrosol generation and radion-is extensive.This data base is predominantly the result of uclide release (the VANESA model). Re model is based research sponsored by the NRC and research sponsored on the assumption that acrosols are produced by the in Germany at the Kernforschungszentrum in Karlsruhe. mechanical entrainment of melt when gas bubbles burst ,

t Data on the interactions of uranium dioxide melts with at the surface of the melt and by vaporization of volatile i concrete have been obtained from the ACE program melt constituents into gases sparging through the melt. l sponsored by the Electric Power Rescarch Institute, in Predictions of the total acrosol generation rate compared -

which the NRC is a partner. Additionaldata on core con. to results of the BETA test show that the model predic--

erete interactions maycome from the ALPII A program in - tions are well within the expected accuracy limits of the Japan and studies under way in Russia. available thermodynamic data base.

l The first consideration regarding the range of conditions 11.2.3 Future Plans

that could affect core-concrete interactions during a se-

l. vere accident is the type of concrete.The data base now Refinements to the CORCON-MOD 3 code are being-

available for core concrete interactions includes tests made with regard to phase relations and models of non-with both of the general classes of concretes in use in U.S. ideal solutions. An effort to compare the code predictions -

commercial nuclear power plants (i.e., siliccous and cal- to test results will be completed. Part of the validation of careous concretes). the CORCON MOD 3 code will be carried out under an arrangement with the I.V. Kurchatov institute.The vali.

Another important consideration deals with the composi- dation will include comparis(m with data from the BLITA tion of the core debris itself, particularly the content of (KfK), SURC (SNL)and ACE program. Once validation

- metallic zirconium in the mixture.The oxidation of metal- efforts are finished, the-development of CORCON-lic zirconium in the core debris can elevate the tempera- MOD 3 will be complete as a stand-alone model of core-ture of the debris during the interaction with concrete, concrete interactions. CORCON-MOD 3 will then be which would result in an increase in the release of nor- incorporated into the CONTAIN and the MELCOR sys-mally refractory fission product elements to the contain- tem level codes for severe accidents.

NUREO-1365. Rev.1 H-6

. - - , - = - -. . - . - - ._ - _ -- ._ - - - ,

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

Appendix D LL3 Closure of Severc Accidents Element 2: Evaluation and Specification for Experi.

mmts and 7uring in w'hich the experimental objec-ScalinE MethodoloU tives are reflected in terms of scaling rationales that are necessary to ensure that both separate-effects B3.1. Background tests and integral-effect test data are applicable to full-scale reactors, and that the tests include the In many areas of severe accident research, the experimen- phenomena important to the specified accident scc-tal investigation has evolved or progressed to the point nario, that experimental programs seek to resolve very specific issues of uncertainty for particular geometries and reac. Element 3: Data Acquisition and Documentation, in for plant configurations. The aim of these programs is which the data base appropriate to issue resolution is often to produce results that can be characterized as established and documented for subsequent use, directly applicable to reactor behavior, or at the least suitable for the development of models that allow for To perform scaling analyses that satisfy the objectives of extrapolation to severe accident reactor analysis. SASM, a hierarchically based, two-tiered scaling method-ology was developed. The complex physicochemical proc-esses that characterize severe accident scenarios and their As part of the Revised Severe Accident Research Pro- associated synergetic effects mandate a hierarchical ap-gram Plan that was published as NUREG-1365 in August proach to the problem in order to make it tractabic.The 1989, the NRC identified initiation of a severe accident two-tiered scaling methodology involves a top-dowTi sys-scaling methodology (S ASM) development program as a tem scaling tier and a bottom-up or process scaling tier.-

programmatic element. Experimental investigation of se- The top-down system scaling analysis provides the basic vere accident phenomena poses a senous challenge owmg conservation equations as well as the scaling rationale and to the fact that many processes involve a complex synergy similarity groups to be preserved in experimental design.

of fluid flow across varied flow regimes, combined heat Additionally, it is through the system scaling analysis that transfer modes, high temperature material interactions, quantification of the effects of any distortions is achieved..

and chemical reactions. Development and application of The bottom-up process scaling tier focuses on the models an SASM, representing a structured methodology that is for the important processes that drive the system re-systematic, comprehensive and scrutable, provides the sponse, confidence that scaled experiments faithfully reproduce the phenomena that will occur in a nuclear power plant. In the application of the SASM to the 1)CH issue, RPV Further, application of an SASM provides the basis for conditions were examined to evaluate the initial and -

application of analytical models validated against smaller boundary conditions appropriate for experimental simu-scale experimental facilities to full scale. lation. Scaled model laws for RPV discharge phenomena and for reactor cavity phenomena were then derived us-ing the two-tiered scaling methodology. He SASM B3.2 Status evaluation of initial and boundary conditions for DCH in

( a PWR for a station blackout scenario with failure of a i To address the scaling problem, the NRC .tmplemented a lower head penetration served as the basis for the DCH SASM development program at the Ilrookhaven Na- integral-effects tests initiated in September 1991.

. tional laboratory (BNL). A technical program group (TPG) was formed by the contractor to guide the develop- The scaling of important processes indicatcs that debris ment of the SASM and to demonstrate its efficacy by dispersion or entrainment of corium is a strong function applying the methodology to the direct containment heat- of the gas. velocity after conditions for the-onset of ing problem. entrainment are satisfied. Expressed in terms of pressure, entrainment is a function of the pressure ratio (reactor -

The results of the TPG activities culminated in the docu. vessel / reactor cavity) to the 4.6 power as well as a function -

mentation of the SASM and its application in NUREG- of gas and fluid properties.The practical effect of these .

5809, "An Integrated Structure and Scaling Methodology dependencies, at least as analyzed for the Zion and Surry for Severe Accident Technical Issue Resolution," pu!>. plants, is the very rapid increase of entrainment fractions lished in November 1991. from low initial reactor pressures up to effectively com-plete dispersal at reactor pressures of approximately 400 The SASM consists of a number of steps that are grouped pst in three key c!cments:

Element I: Specification of Enperimental Require- NUREG-5809 was issued as a draft for public comment ments, in which the experimental objectives are de- in November 1991. The authors of the report acknowl-fined in terms of the technicalissues. edged that the SASM application to DCH was not 11 - 7 NUREG-1365, Revi!

Appendix 11 intended to represent cornplete technical resolution of under the DCll research program described in section the scaling questions, as some of the scaling relationships 2.1. The discipline of a SASM will be applied to other

[

~

derived are based on specific assumptions that require experimental programs in the SARP as necessary.'lhe experimental confirmation and additional analysis. Iteso- staff also intends to iricorporate comments received on lution of those technical issues as they af fect luth scaling NURl!G-$SO9 into a final version of the report, tenta-of experimental facilities and operation, as well as model. tively scheduled to be issued by December 1992.

ing development for scactor analysis, will be pursued i

\

N Uiti!G-1365, llev.1 11 - 8

_. = _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .__._ . - - _ _ _ . _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _

NRcFOnM34 U S. tA>CLL An nE OUL ATORV COMMibSION . 1, HE POnr NUMfit H -

(7 M) (Ase.g<wwi by NnC. Add Vol.,

IdicM 1107, bum . new, aN Aar3s utum Num.

WW BIBUOGRAPHIC OATA SHEET b" ' - " *"v d (6 in.inct=,. on i. ,.m.3 NURl!G-1365 L ma ma svimiu Rev. I a oue nowr ewuse Severe Accident Research Program Plan Update uoyrs y j

December 1992

.. nN on wwo NuMutn 6 Au I t uWu G A 6, I YPE O& nLWNil Tect,nical

7. n neo'o covt noo enou.. o.i..

. et nn*=n oHwu, no4 - NAix m o Ano m s .o. o.m. on . m n.gm u. a Nuc , n.go . my cmn,.....on. .no m.mno .e o o cwir.cto,, w<m n.m. .no m.mna .o,e.. Nnc.3 Division of Systems Research Office of Nuc! car Regulatory Research U.S. Nuclear Regulatory Commission Washington, DC 20555

9. flPON80HING OHGAPM A flON

• NAME AND ADoHLb8 (if NitC. typ. ' Sam. . staav.*; if contr.ctw, prowd. NHC Divism Of,me w Heym u s. Noo.u n.gue.imy commi. m and meno me... i Same as above

10. tiUPPLLML NI AH y NOIE b

n. AnsinAc t (zoo wuo. o, i...)

In August 1989, the staff published NURI!G-1365," Revised Severe Accident Research Program Plan." Since 1989, significtml progress has been made in severe accident research to warrant an update to NUREG-1365.

'lhe report describes recent major accomplishments in understanding the underlying phenomena that can occur during a severe accident.'these include Mark i liner failure, severe accident sc: ding methodology, source term issues, .

[.

core. concrete interactions, hydrogen transport and combustion, TMI-2 Vessel investigation Project, and direct containment heating. 'the re[ ort also describes the major planned activities under the SARP over the next several' years. 'these activities will focus on two phenomenological issues (core melt progression, and fuel coolant interactions and debris coo! ability) that have signific4mt tmccrtainties that impact our understanding and ability to predict severe accident phenomena and their effect on containment performance.'lhe SARP.will also focus on severe accident code development, assessment and validation. As the staff completes the research on severe accident issues that relate to current generation reactors, continued research will focus on efforts to independently evaluate the capability of new advanced light water reactor designs to withstand severe accidents.

12. KEY WOnOS/DEScfwfORS (List words or phr.s.s that win .ssist ,.s.uctw,rs in loc.itrig tn. recut ) 4 "R^hd Y 8 WN M Unlimited 14, SECtFtrf Y cLASMICATION nuclear power plants source term g n,,, ,,g severe accidents fission product transport direct containment heating Mark Iliner failure Unclassified core melt progression core. concrete interaction ('"" ""'

fuel coolant interaction TMI-2 Vessel Investigation Project Unclassified debris coolability severe accident codes is.NuMuen & eA m hydrogen combustion advanced light water reactors hydrogen generation severe accident scaling methodology ,

NAC FORM M8 (2-80)

d 4

Printed on recycled-paper Federal Recycling Program I

1i 4 lj \!1l I.

n ae aa sc ,

sats ,

t ,

cue*n r.'ssc m*sec t

&sa n t y a, '5 e

m*2

{'

7 r

5

_ ~

~

- ~

. 7

. r g

=

, - ,=,

.7 .[

I S

• 2 S0 S0 0 M-05 I

0 3

M5 E SO5 S S EC0 S E U T Y 2 E A R . %TA T

S O C. U E V DATD ,

L *J L P EL N TU UR CO I

O E FF NG U E T F OY RG N T

L RI A N

AH E S [

P LA CW U

N -

l! f'll' l,ll I\, , 1

f]))!l

" OA

• F

• S E. s o

• D E o

Fm N

w.

"* s A r e

E G w

' TA w.

• o "5 P g'

m

)

b 3 h C

/ [

3 L ~2 I.

• 4 I "

N I

• I 1 I C

1 I L

a G3 1 CE 3w 2 5- eU s

'

• as

- w 31 c_ ;aA r, h'

5 e 5 =. p C ' 1 5' - 1 0 g.57 2e{r . Aa 1 g p T c ..

N O

I 1 S0 S0 0

'M-05 I

0 3

M5 E SO5 SS EC0 2 S E U T NE AY I T S

TR J A S O C. BN t

TD L R DA ,

APR EL N  :

rO TUO I

NG T .. CUF UER G Y T

N L A

R! 4 AH S t

E E . P LA C.

f W

N' L!'IIll!Ill!I l Il!It l