IAEA-TECDOC-757

Decay heat removal and heat transfer under normal and accident conditions in gas cooled reactors

Proceedings of a specialists meeting held in Jülich, Germany, 6-8 July 1992

INTERNATIONAL ATOMIC ENERGY AGENCY The originating Sectio f thino s documen IAEe th An i t was: Nuclear Power Technology Development Section International Atomic Energy Agency Wagramerstrasse5 P.O. Box 100 A-1400 Vienna, Austria

DECAY HEAT REMOVA HEAD LAN T TRANSFER UNDER NORMAL AND ACCIDENT CONDITIONS IN GAS COOLED REACTORS IAEA, VIENNA, 1994 IAEA-TECDOC-757 ISSN 1011-4289 Printe IAEe th AustriAn i y d b a August 1994 PLEASE BE AWARE THAT ALL OF THE MISSING PAGES IN THIS DOCUMENT WERE ORIGINALLY BLANK FOREWORD

The Specialists Meeting on Decay Heat Removal and Heat Transfer under Normal and Accident Conditions in Gas Cooled Reactors was held at the KFA Research Center, Jülich, Germany, 6-8 July 1992. The meeting was convened by the International Atomic Energy Agency on the recommendation IAEA'e ofth s International Working Coole s GrouGa n dpo Reactors attendes wa t .I participant y db s from China, France, Germany, Japan, Poland Russiae th , n Federation, Switzerland e Uniteth , d Kingdo Unitee th d mdan State f Americaso meetine Th . chaires g. Kugelewa K Profy d. b d Dr . an r Prof. Dr. E. Hicken, Directors of the Institute for Safety Research and Reactor Technology of the ResearcA KF h Center covered an , followinge dth :

Design and licensing requirements for gas cooled reactors; Concepts for decay heat removal in modern gas cooled reactors; Analytical methods for predictions of thermal response, accuracy of predictions; Experimental data for validation of predictive methods: • Operational experience from gas cooled reactors, • Experimental data from test facilities.

IAEA activitie advancen si cooles dga d reactor technology developmen conductee ar t d within the IAEA's nuclear power programme. Advance s coolega d d reactor designs currently under development are predicted to achieve a high degree of safety through reliance on innovative features and passive systems. The IAEA's activities in this field are focusing on the four technical areas which provide advanced gas cooled reactors with this high degree of safety, but which must be proven. These technical areas are:

(a) The safe neutron physics behaviour of the reactor core; (b) Reliance on ceramic coated fuel particles to retain the fission products even under extreme accident conditions; abilitdesigne e th Th f y o dissipat o st ) (c e decay hea naturay b t l heat transport mechanismsd an ; (d) The safe behaviour of the fuel and reactor core under chemical attack.

investigato T thermae eth l behaviou sucf ro h advanced designs, experimental investigationd san analytical studies are ongoing in several countries to confirm the ability of the reactor designs to dissipate the decay heat from the core by natural heat transport mechanisms without reaching excessive fuel temperatures. Experiments have been performed to obtain data on heat transport phenomena (e.g. natural convection heat transport both within and outside the reactor vessel). Experienc hean ei t transport under actual reactor condition bees sha n obtaine R Magnon di AG d xan reactor Unitee th n si d Kingdom, Franc HTGRJapann d i ean d Germanan ,n si USAe th d . yFuturan e efforts will focus on further development and validation of analytical tools, identifying safety margins, understanding the effect on system performance of uncertainties in key heat transport parameters, system optimization and experimental confirmation and demonstration of the predicted performance of passive systems.

summaryn I , considerabl operatinR eGC g experience fielexiste decaf th do i sh y heat removal and heat transport advancer fo d an , d designs, experimenta analyticad an l l activitie o undet e sy ar rwa investigate key heat transport phenomena. A key conclusion of this Specialists meeting was that for advancee th cooles dga d reactor designs currently under development predictee th , d performancf eo passive systems for heat removal needs to be proven under experimental conditions representing realistic reactor conditions. EDITORIAL NOTE

In preparing this document for press, staff of the IAEA have made up the pages from the original manuscripts submittedas authors.the viewsby The expressed necessarilynot do reflect those of the governments of the nominating Member States or of the nominating organizations. The use of particular designations of countries or territories does not imply any judgement by publisher,the legalthe IAEA,to the status as of such countries territories,or of their authoritiesand institutions or of the delimitation of their boundaries. The mention of names of specific companies or products (whether or not indicated as registered) does implyintentionnot any infringeto proprietary rights, should construednor be it an as endorsement recommendationor parithe IAEA. ofon the The authors are responsible for having obtained the necessary permission for the IAEA to reproduce, translate materialuse or from sources already protected copyrights.by CONTENTS

SUMMARY OF THE SPECIALISTS MEETING ...... 7

DESIGN AND LICENSING REQUIREMENTS FOR GCRs (Session I)

Principles of decay heat removal in reactor technology — Present status and future prospects ...... 15 KugelerK. Safety criteria and provisions for the evacuation of residual heat from graphite gas cooled reactors ...... 29 R. Lheureux, A. Aguilera Bases for the MHTGR source term and containment concepts ...... 41 P.M. Williams ultimate Th e HTR-Module safetth f yo e during hypothetical accidents ...... 7 4 . G.H. Lohnen

CONCEPT DECAR SFO Y HEAT REMOVA MODERN LI N GCRs (Sessio) nII

Distribution of the decay heat in various MODUL HTRs and influence on peak fuel temperatures ...... 53 E. Teuchert, K.A. Haas, A. van Reek, P.R. Kasten Afterheat removal for HTR-10 test module under accident conditions ...... 63 Gao Zuying, Shuyan,He ThongMin Flow scheme desigd san n feature HTGf o s R residual heat removal systems ...... 2 7 . V.F. Golovko, A.I. Kiryuschin, N.G. Kuzavkov Passive deca residuad yan l heat remova MHTGRe th n i l ...... 5 7 . D.A. Billing, S.K. Chose, J.M. Berkoe, S.A. Caspersson, G.C. Bramblett Impact of increasing MHTGR power on passive heat removal ...... 83 T.D. Dünn, A.A. Schwanz, P.A. Silady Trends in safety criteria for future reactor plants ...... 89 HickenE.

ANALYTICAL METHODS FOR PREDICTIONS OF THERMAL RESPONSE, ACCURACY OF PREDICTIONS (Session III)

Analytical and experimental investigations of the passive heat transport in HTRs under severe accident conditions ...... 95 W. Rehm, H. Barthels, W. John, J. Cleveland, M. Ishihara Presentation of decay heat removal computer codes used for gas cooled reactors ...... 104 G. Carvallo, Dobremelle,M. MejaneA. Modelin analysid gan heaf so t transfer fro MHTGe mth R core throug hsteea l reactor vessel to the reactor cavity cooling system ...... Ill D.A. Dilling, J.M. Berkoe, S.K. Ghose, T.D. Dunn, S.A. Caspersson Analysi f afterheaso t removal from modular HTGRs during accidents concepa d an r , fo t f sphericao e us l fuel element LWRn si s ...... 7 11 . 7.5. Mosevitskij, A.O. Goltsev, P.V. Mikhailov, V.F. Tsibulskij, V.D. Davidenko, V.S. Popov, Yu.N. Udyanskij EXPERIMENTAL DAT VALIDATIOR AFO PREDICTIVF NO E METHODS— OPERATIONAL EXPERIENCE FROM GCRs, EXPERIMENTAL DATA FROM TEST FACILITIES (Session IV)

Heat transfe uppee th n HTTe ri r parth f Ro t pressure vessel during los forcef so d5 coolin12 . . g Y. Shiina, M. Hishida Development of an inactive heat removal system for high temperature reactors ...... 131 K. Kugeler, M. Sappok, B. Beine, L. Wolf Passive heat removal experiment advancen a r sfo d HTR-module reactor pressure vessel and cavity design ...... 139 L. Wolf, A. Kneer, R. Schulz, A. Giannikos, W. Hafner Test apparatus of cooling panel system for MHTGR ...... 147 Takada,S. Suzuki,K. Inagaki,Y. MiyamotoY. SANA experiments related to self-operating removal of decay heat ...... 151 H.F. Nießen, M. G. Lange

Lis Participantf o t s ...... 1 16 . SUMMARY OF THE SPECIALISTS MEETING

The requirements for decay heat removal are met at current reactors by active heat removal systems. While these systems have prove highle b o nt y reliable, their failur lean coro eca dt e melt accidents. Research and development activities worldwide are now showing that design of the reactor and plant for self-acting decay heat removal to avoid core melt accidents and guarantee retention of fission products insid fuee eth l elements, activl eveal f ni e heat removal systems failactualln ca , e yb realized. Self-acting decay heat remova lapproac w e elemensystemne on a e nucleao n ht i ar st r engineering which is termed 'catastrophe-free' nuclear engineering. By this it is meant that major accidents in category 7 (catastrophic accident, e.g. Chernobyl) of the INIS scale are deterministically excluded. Experimental demonstration performance prooe th th f f o fo s f self-actino e g systeme sar required.

In Japan the gas cooled reactor activities are focused on the High Temperature Engineering Test Reactor (HTTR) project. Constructio HTTe Oarae th th f nt Ro a i Research Establishmen Japae th f no t Atomic Energy Research Institute has been under way since March 1991 and first criticality is expecte 1997n di . Thi MW(th0 s3 ) reactor will produce core outlet temperature 850°f so t rateCa d operation and 950°C at high temperature test operation. It will be the first nuclear reactor in the worlconnectee b o dt higa o dt h temperature process heat utilization system reactoe Th . r wile b l utilized to establish basic technologies for advanced HTGRs, to demonstrate nuclear process heat application o servt d en irradiatio an a ,als s a o n test facilit r researcfo y n higi h h temperature technologies.

At JAERI, natural heat transfer mechanisms (e.g. natural convection, conductio thermad nan l radiation) have been investigate HTTr dfo R conditions. Thermal analysis codes have been developed and use HTTr dfo R licensing experimenn A . bees ha t n conducte determino dt naturae eth l convection heat transfer coefficient in the top hemisphere of the reactor vessel. Also, a facility has been constructe investigato dt e heat transfer throug reactoe hth r vesse wateo t l r panels surroundine gth vessel. At JAERI research on air ingress processes is being carried out for the primary-pipe and stand-pipe rupture accidents. Air ingress processes by molecular diffusion, natural convection and helium/air exchange flow during the first and second stages of the primary-pipe and stand-pipe rupture accidents are investigated. Graphite corrosion by a high-temperature air stream is also being investigated.

In the USA, commercial HTGR development activities are conducted by the Division of HTGRs Office oth f f Advanceo e d Reactor sDepartmen S PrograU e th f m o f Energy o t . These activitiee sar focused on the modular HTGR. In the reference design, a thermal power of 350 MW(th) is produced wite reactoon h r modul r electricitfo e y generation wit e steath h m generato a side-by-sid n i r e arrangement with the reactor pressure vessel. A conceptual design of a natural draft air cooling syste removar mfo decaf o l y heat fro MHTGe mth bees Rha n completed. Activitie hean si t transport are focused on establishing detailed computer models for analysis of the performance of the MHTGR during accidents under both pressurize depressurized dan d conditions. Sensitivity studies wile b l performed to determine the effects of uncertainties in various heat transport parameters on system performance. The results of these studies will be used to determine needs for heat transport experiments. Also, code verification activities will be performed by computing system performance with independent computer codes. studieA US se havth n I e been conducted leadin recommendatioe th o gt n thaMHTGe th t e Rb change d0 MW(th fro35 ma 0 MW(th) 45 desig a o )t n design e MHTGTh . R fue s relativeli l y unaffected by this increase in power, while the vessel and internal metallic parts have slightly increased temperatures during certain accidents.

The approach for the selection of the MHTGR radionuclide source term for accidents and for desige th containmenf no t concept derives si d fro sloe mth w respons MHTGe th f eo corRo t e heatup events which result prompa n si delayed an t d portio source th f no e term prompe Th . t source term involves radionuclide releases which occur earl resula s y a depressurizatiof to primare th f no y system. The prompt source term contains radionuclides circulating with the helium, but is dominated by 'liftoff of radionuclides previously deposited on surfaces within the primary system. The delayed source term develops from releases from fuel during the subsequent lengthy core heatup phase of the accident. Although uncertainties exist expectes i t i , d thaprompe th t t source ter msmals i l enough that the radionuclides could be vented from the reactor building, possibly through filters in a relief train, if necessary. Because the delayed release occurs at atmospheric pressure, a conventional high pressure, low leakage containment is not necessary as it serves no identifiable function. Demonstration of adequate retentio f fissiono n products withi fuee nth l during heat-u hydrolyzind pan g conditions is a key technology activity within the US programme focused on further developing and confirming source th MHTGRee terth r mfo .

In the Russian Federation, development activities are under way for the modular VGM gas-cooled reactor. Thi pebbla s si systed ebe m usin side-by-sidga e arrangemen reactoe th d f o t ran steam generator in steel pressure vessels. A power level up to 215 MW(th) can be achieved while maintaining passive safety wated . Botan r r hcooleai d alternative reactoe th r sfo r cavity cooling system are being considered.

Several experimental facilitie investigato t s e performanc desig componentM f eo VG n e th r sfo have been constructe Experimentae th t da l Machine Building Design Burea t Nizhnua y Novgorod, Russian Federation. Thes (maximue W includM 5 1 ema powe electrif ro c heaters) facilit testinr yfo g steam generators and helium-to-helium intermediate heat exchangers, a helium circulator test facility (full scale for VGM; i.e. 90 kg/s helium flow with a 4 MW(e) motor drive), the TIGR facility for investigating short term pressure and temperature transients in the VGM during depressurization accidents, the MASEX mass transfer test facility, and a high temperature helium facility with electrically heated pebbl graphitd ebe e spheres, which coul usee d b carro d t t water/ai you r ingress and depressurization tests. Future operatio thesf no e heat transport experiment facilitie Russiae th n si n Federation is in question due to the adverse financial situation.

In Germany, the industry has concluded that there is presently and for the nearest future not sufficient marke tplanR neeHT t dn wit a eve a hr hig nfo h degre f inhereno e t safety features. Therefor participatine eth g firms (AB Siemend Ban thein si r joint subsidiary company HTR/GmbH), in agreement with the Federal Ministry for Research and Technology (BMFT) are terminating project related activities. Within these activities intensive technology programmes on heat transport during norma accidend an l t conditions have bee pase n t Germa th carrie a ti h t dnou reactor industre th d yan KFA Research Center Jülic wels ha l yieldin largga e bas experimentaf eo l datcomputed aan r codes validated at several test facilities and especially at the AVR for normal and simulated accidental conditions.

Now research and development activities in Germany have been mainly directed to advanced reactor plants which do not have an impact on the environment and the public even hi the case of a hypothetical accident. Safety improvements are investigated which could be achieved by a consistent exploitation of inherent safety features and passive systems. This generic research work is continuously supported by the government.

Investigation Siemeny b s regardinG A s ultimate gth e HTR-MODULe safetth f o y E during hypothetical accidents have delineated the safety claims of the HTR-MODULE and concluded that maximum offsite radiation doses unde l crediblal r e accident scenario withie ar s n legal limitn i s German o need thern o plaan yt dr s publi i e fo n c sheltering, evacuatio r restrictioo n e th f no consumption of agricultural products in case of a severe accident.

Studies have beçn performed at the KFA Research Center, Jülich, of peak temperature distributions during accident designo tw r f annulao ssfo r pebble-be reactoe d on reactors r rfo d an , design with annular prismatic fuel. For all concepts and fueling options examined, the response to a

8 loss-of-coolant accident shows thareactoe th t r inherently becomes subcritical, whil decae eth y power partl ye reacto heatth partld p u san r s removei y e environmenth o t d naturay b t l heat transport mechanisms without overheating the fuel. Switzerlandn I paste th ,n i ,researc h activitie smalconceptr R sfo HT l s wit hvera y high degree of safety were conducted. Technical experience in gas cooled reactor heat transport is being maintaine Paue th t l da Scherre r Institute (PSI), Villigen, Switzerland; howeve fiele f th hea do n i rt transport, the major activities at PSI are focused on decay heat removal in other advanced reactor designs (e.g. LWR LMRs)d san . France no longer has an HTGR development programme. However, an evaluation of advanced gas cooled reactors (e.g. the MHTGR and the HTR-MODULE) is under way. France considers the HTGR to have a wide range of possible applications (electricity and process heat) and fueling options.

Currentl Saint-Laurente yth Bugee th , d Chinoy an Magno , nA3 x type CO2 coole cooles dga d reactor operation i e sar Francen i . Normal decay heat removal from these reactor accomplishes si d with active systems which have proven to be highly reliable. Only one incident occurred which require emergencn da y backup devic broughe b o et t into operation. This incident demonstratee dth efficiency of the emergency heat removal system, which operated on natural convection of the CO

coolant. 2

Current evaluation Francn si e focu containmenn o s decad an t y heat removal topics theid an , r interrelationships. Thermal analysis methods have been develope d use an do investigatt d e th e behaviour of the MHTGR under loss of cooling accidents, and results are being used to identify possible improvements which could provide better plant economics.

In the United Kingdom, for Magnox and AGR reactors, there are ongoing programmes to investigate natural heat transport during accident conditions: specifically radiatio d naturaan n l convection from debris resulting from a dropped fuel assembly accident. Thermal analysis codes modeling the reactor system have been developed and used to investigate behaviour of a degraded core during accidents.

China's HTGR R&D activities are conducted within the National High Technology Research and Development Programme sponsored by the State Science and Technology Commission (SSTC). The programme is carried out at the Institute of Nuclear Energy Technology (INET) at Tsinghua University t othea , r nuclear energy research institutes unde Chine rth a National Nuclear Company, as wel powes a l r plant equipment manufactures througd an , h cooperation with organizations outside China. The SSTC has approved a project for a 10 MW(th) test module HTR to be constructed at INE obtaiTo t n experienc plann ei t design, constructio operationd nan rangA . f applicationeo s will investigatede b r examplefo , , electricity, steam districd an , t heat generatio firse th t n ni phas d ean process heat generation for steam reforming of methane in the second stage. First criticality is planned for 1998. (INET has previously constructed a 5 MW(th) water cooled reactor which began operating in 1989 to provide steam and hot water for district heating at INET.) The conceptual design of the 10 MW(th) test HTR was carried out under a cooperative agreement between Siemens Interatom and INET in 1988. INET is now carrying out the preliminary engineering design of this test reactor. Studie beine sar g performe seleco dt cavite th t y cooling system design. Proposed system forcee sar d circulation water cooling naturad an , l draf r coolingai t .

Regarding heat transport, INE s performeTha d detailed analyse performance th f o s e th f o e HTR-10 test module usin THERMIX-KONVEe gth K code develope JiilichA KF .y db Experiment s to demonstrate passive decay heat removal are being planned.

In Poland the HTGR is considered appropriate as a heat source for district heating as it could be sited near highly populated areas. Sixty percent of the current dwellings in the larger towns and cities are connected to district heat grids. Poland is performing an evaluation of existing HTGR designs, and future discussions will consider how to accomplish licensing of the HTGR. Regarding designs for future HTGRs there were several unifying themes presented in the meeting. Most of the HTGR designs described sought to remove decay heat by passive means, without relianc operatoe th externa n n o e o r ro l power supplies designe Th . s describe decar dfo y heat removal tende simple b robusto d t ean ,rel o witpartt inheren n w yd o hfe s an t operating mechanisms suc s gravitya h , atmospheric pressure d radiativan , e heat transfer. Proper modelin f theso g e phenomena requires further attention to assure the key phenomena are adequately represented and sufficient accuracy in results is attained.

Regarding the analytical tools for predicting heat transport in reactors and related experimental facilities, and the accuracy with which predictions can be made. Key issues are the adequacy of the modelling of the important heat transport phenomena, and the accuracy to which temperatures of fuel and key components can be predicted.

Comparisons have been made of analytical predictions with experimental results from loss-of- floloss-of-cooland wan t Arbeitsgemeinschafe testth t sa t Versuchs Reaktor frod an m, experiments involving natural convection performe LUNe th Researc t A da A KF facilit e hth f Centeryo same Th . e computer codes are also used to predict the performance of the HTR module concept. Results of comparison of code predictions with experimental data show the suitability of these codes for analyses of accidents involving loss-of-forced convection and loss-of-coolant. Some of this work has been conducte cooperativn di e activities involving Germany USA,e Japath d . nan

For the French Magnox reactors, two codes have been used to analyze post-shutdown heat removal. These codes treat the short term transient response with detailed models of the main reactor components, and the longer term (days to months) response with more simplified models. Further modeling activities have been Frenc e carrie pars th a f t o th d ou activit evaluato yt modulae eth r high temperature reactor othed an , r innovative concepts. Several accident scenarios have been analyzed and studies have been performed to determine the sensitivities of the predictions of temperatures of the fuel and other reactor components to uncertainties in material properties and other input parameters f theso e e codeUs . internationan si l benchmark calculations coul usefue db verifyinn i l g codes currently used for design and safety analysis of advanced gas cooled reactors.

For the design of the US MHTGR and its reactor cavity cooling system (RCCS), complex heat transport phenomena involving 3-D thermal radiation and natural convection under pressurized and depressurized condition wels sa non-unifor s la m geometr reactof yo r cavity cooling system equipment must be properly modeled. Also thermal properties such as core conductivity and reactor vessel emissiviry mus knowe b t n with sufficient accuracy. Therefore, activities programmS withiU e nth e to confirm the adequacy of the heat removal system design focus on modeling improvements, model verificatio validatiod nan determinatiod nan heay ke t f transporno t material propertie theid san r range of uncertainty.

During the discussion of specific approaches to modeling there was general agreement among the specialists that much work remains to be done in model development and application to properly represent the complex fluid flow and heat transport phenomena and geometries associated with advance cooles dga d reactors which rel naturan yo l heat transport phenomen decar afo y heat removal. There was agreement that international cooperation in heat transport code development, verification and validation with experimental data would be beneficial to several national programmes. Certain experiments which could provided these data were reviewed in detail.

JAERI has conducted experimental activities to investigate heat transport for conditions representative of loss-of-forced cooling at the HTTR. Natural convection heat transfer was studied using thermo-sensitive liquid crystal powder a visualizatio s a s n tracer. Future experiments will investigate natural convection with thermal radiation with temperature 600°Co t p su .

JAERI has also constructed a scale model experimental mockup of the reactor cavity cooling panel system employed at the HTTR. A cooling panel system is installed at the HTTR on the surface

10 of the reactor cavity wall as a backup to the auxiliary cooling system, with a heat removal capability of 300 kW. The experimental mockup consists of a pressure vessel containing an electric heater, and cooling panels surrounding the vessel. The shape of the heat flux distribution on the heater surface controllee b n uniforme ca b o dt , cosin exponentiar eo shapen i l . Also differena n i , t mode surface th , e temperature of the heater can be controlled. Gas pressure in the vessel can also be varied. Preliminary tests have been conducte o chect d e characteristicth k d operatios an e it sfacilit th d f an o ny instrumentation. Tests are planned to investigate the heat removal from the heaters, through the vessel wall to the cooling panels.

An inactive heat removal system for high temperature reactors has been constructed and operate Siempelkampy db , Krefeld, Germany purpose facilite Th . obtaith o t f es o yi n experimental data for heat transport through a prestressed cast-iron pressure vessel to a water cooled system integrated into the cell surrounding the vessel. All experiments were accompanied by pre- and post- test computations, which were performed by a 2-D transient finite-element code, TOPAZ. Results of experimente th accompanyine th d san g analysis confirme feasibilite dth sucf yo hvessel/cavita y cell MW(th0 20 concep e th ) r pebblfo t reactord ebe providd an , databasea validatior efo computef no r codes used to analyze heat transport for advanced gas cooled reactors.

SANA-e Th I experimental facilit investigato yt e self-operating remova decaf o l y heabees ha t n constructe operated ResearcA dan KF t da h Center, Jiilich. Another experiment, SANA-II beins i , g planned and will represent a horizontal section of a reactor in full scale, including a sector of the core, reflector, vessel and primary cell with the core decay power being simulated by electric heaters. e SANA-Th I facility will provide experience with components (e.g. heating elements, insulation, instrumentation SANA-IIr )fo . SANA-II will provid opportunite eth obtaio yt n data demonstratine gth principl f self-operatineo g heat transpor decar fo t y heat remova larga n o le scale.

The Specialists meeting provided important information for subsequent discussions at the IAEA Consultanc Co-ordinatew plao yt ne na d Research Programm Hean o e t Transpor Afterhead an t t Removal for Gas Cooled Reactors under Accident Conditions (9-10 July 1992). The objective of this CRP is to establish, through international co-operation, sufficient experimental data and validated analytical tools to confirm the predicted safe thermal response of advanced GCRs during accidents.

11 DESIG LICENSIND NAN G REQUIREMENT GCRR SFO s (Session I)

Chairman

J.T. WILSON United Kingdom PRINCIPLE DECAF SO Y HEAT REMOVAL IN REACTOR TECHNOLOGY- PRESENT STATUS AND FUTURE PROSPECTS worldwide 30 ~ K. KUGELER build up Institu Sicherheitsforschunr fü t Reaktorsicherheitd gun , (GWel/a) Forschungszentrum Jülich GmbH, 20 -\ Jülich, Germany

Abstract

Reliabl safd eean decay heamaie th t nf remova o safet e on y s i lrequirement nuclean si r reactor technology. Today in all plants worldwide this requirement is fulfilled by active decay heat removal systems. If these systems fail, core melt-down accidents can occur and time large amount fissiof so n product escap n environmente sca th o et developmentw Ne . s worldwide are especially discussed toward the question of how to improve reliability in Fig. 1: Crisis of the Nuclear Industry since 3 Years no New Orders (Source: atw 3,88, p.149, decay heat removal thin I . s paper som proposalw ene describede sspecifiar e th d can , safety atw 11, 91, p.ll) questions connected with decay heat remova discussede lar shows i t I . n tha concepta f to self-acting decay heat removal which avoids core melt-down accident whicd san h sources int market)e oth . Fro perspective mth world-wids it f eo e application typw f eo ne a , guarantees the retention of the fission products inside the fuel elements even if all active nuclear engineering is required with a new quality of safety, so-called "catastrophe-free decay heat removal systems failindeen ca , realizede db . Some technical detail thif so s nuclear engineering" meanthis y i t B .si t that major accident categorINESn se i th f o - y7 principle, whic alreads hha y been realize high-temperaturr dfo e reactors describee ar , n di scale must be deterministically excluded, (see Fig. 2) this paper. The differences between active, passive and self-acting concepts are explained. Furthermore, some limitations of the concept of self-acting decay heat removal are shown; The requirement that no catastrophic events may occur must hold for reactors and all in additio concepe th o n t f thermao t l stabilit core th ef yguaranteeo self-actine th y db g decay facilities concerned with fuel element suppl disposald yan . This requirement obviously heat removal, there are requirements of neutron physical, chemical and mechanical stability include controe sl eventth al f lo s resulting from disturbances withi facilite nth severd yan e of the core in all accidents. Future work to prove the required principle is discussed at the external impacts on the facility (airplane crash, earthquake, explosions of clouds of gas). papere th en f do . Very extreme external impacts (war, serious sabotage, meteorites) will require special consideration.

1. Status and Future Developments Necessary in the Nuclear Industry . 2 Decay Heat Productio Decad nan y Heat Remova Presenn i l t Nuclear Facilities At present nuclea0 42 , r power station operation i e sar n world-wide wit totaha l power outpu abouf o t t 350,000 MWel r well-knowFo . n reasons, construction wor bees kha n Approximately 7.5 % of the energy of 200 MeV released during fission is released with a stagnating for years (see Fig. 1) and the expansion rate is virtually zero. time delay as decay heat via gamma and beta decays. Fig. 3 shows the known afterheat However nextdecadee w th fe n i , s nuclear energy will probably hav speciataka o e t n eo l curve which reveals that immediately after shutting down a reactor approx. 6 % of the worle rolth n edi energy econom numbea r yfo reasonf ro s (CÛ2 problem, long nominal power still occurs as decay heat, 1 % after one hour and even after 100 h still Oi development tim fusionr efo , long lead introductio e timeth r sfo renewablf no e energy approximately 2 per thousand of the reactor power. Reliable remova f afterheao l achieves i t d with present reactor redundana svi diversified an t d loops. Fig. 4 shows the concept currently employed for modern pressurized water reactors Federae inth l Republi f Germanco y comprising high-pressure coolant injection systems, low-pressure and afterheat removal systems as well as flooding systems, installed separately for each operating loop.

time after shut down (s) catastrophic accident Fig. 3: Decay heat generation in nuclear reactors (e.g. Chcrnobyl. USSR, 198G)

heavy accident

serious accident (Three Mile Island, USA, 1979)

accident

serious incident

2 incident (e.g. Biblis, Germany, 1987)

I failure

underneath the scale without safety importance Fig.4: Modern afterheat removaG lfacilitie R FR concep e PW th r n si fo t 1 reactor, 2 steam generator, 3 primary cooling pump, 4 storage (pressure), 5 water-pool, 6 safety injection pump, 7 cooling pump, 8 cooler, 9 containment sump, Fig. 2: The International Nuclear Event Scale (INES), (IAEO, Vienna, 1990) coolin0 1 g pump . containment11 , storag2 1 , burnr efo t fuel elements coole3 1 , r In spite of the very greatest efforts taken to achieve the reliability of these systems it must be expected that a failure rate (even if very small) will remain for the afterheat removal systems PWRe th r lead, R meltdowa Fo .failur o st reactoAH e f th eo f no r cor aboun ei 1 t largf o R e powe shows PW a a , Figh r rn i well-knowns 2 size fo i .o s 5 t A . reasone th , s hige foune fob th h ro n t di cor thi e esar powe heaw rlo t densit e capacitth weln s i ya s f ya lo the reactor.

The senous accident at Harrisburg (TMI) demonstrated that the problem of core meltdown is indeed real. Since in this case cooling only failed for a limited period only a partial melt- down occurred (se completa ed Figan ) .6 e meltdow reactoe th f no r pressure vesses lwa 1A inlet prevented by restarting the cooling system. Nevertheless, this process drew attention to the 28 inlet basic proble initiated man d world-wide effort improvint sa g reactor safety engineering- At . tention must als drawface e ob th t thao nt t ther naturalle ear y great differences betweee nth 420 facilities in operation world-wide, apparently ranging from values of 10"3 I/a to corr 10~fo e a meltdow6I/ n frequency (see als quits oi t FigI e . obviou.7) s that essential improvements mus achievee tb d futuree her desirabls th i n t ei I . completelo et y eliminate Upper grid damage the physical fact of reactor core melting or destruction by suitable planning and design.

Î.O Loose core debris

T tbittnpfcfct. H • KcnMoli« r Coating of previously K Il«Nrra< (00*0 molten materia bypasn o l s Crust t region interior surfaces 3 I.« lütter* im*C •8 4 lullKn* I4M*C \ Previously molten 1.« Nltll«r* Ctmw« t««<«~H«- material ^ X i 1- Hol bafflm e e plate X

"4t- /l --* ' ^ »,,/' ~^ -7 1 ^ p /• 4 ,— ^ t 1 y Ablated in-core Lower plenum debris \ \ I, \ ^ instrument guide \ \ N< f (. v. ^ ^ ^c ^ \ i Possible region depleted "^-' -«. uraniun i m •*• ..». i ^\ 0 SO U » M »,.| time from the beginning of accident (min)

Fig. S: Isothermes in a 3700 MWth PWR after a loss of coolant and failure of active cooling Fig. 6: State of the partially melted reactor core in the TMI nuclear power station frequency (per year) 00 10-5 10'' 10'3

Surry I _^ AI' 600 Oiblis D ———r*. Core melt possible, — ISIS Oconcc probability — SIR Limerick shall be reduced — new BWR Zion C:i!vi;rl Cliffs Arkansas N.OI . Core melt possible, Karlsruhe-concept Cryslal River fission product retention Indian Point 2 Worldwide in containment ""*• NPI-conccpt Indian Poinl 3 activities on reactors witw hne Miltslonc 1 satiety quality Core melt possible, Millslonc 3 fission product retention SizcwclB l in reactor vessel and Scabrook containment • innovative LWR Empiriscs hau 3000 Jahren I————•———I Bclricbscrfaliruns Core melt tno possible, Fig. 7: Probabilities of core meltdown according to various risk studies fission product retention in fuel elements innovativR eLW

3. World-wide Efforts at Improved Solutions in Nuclear Engineering

Numerous effort alreade ar s y unde world-widy rwa orden ei improvo rt e solution afterr sfo - Fig. 8:WorIdwide effort achievint a s solutionw gproblene e th o st aflerheamf o t removal heat remova reactorn li possiblf i d san restriceo t consequencee tth severf so e accidento st the reactor facility. Fig. 8 sketches current approaches and lists the essential concepts emerging world-wide. Two developments, the NPI concept and the Karlsruhe concept for pressurized water reac- The AP 600 may be regarded as an established representative of first direction (see Fig. 9). achievo t torsm ai , e contro retentiod corlan e th e reactof e nmelo th n i t r containment. Passive afterheat removal systems are to be achieved by combining large water reservoirs, Figs. 10 and 11 show the basic principles of the two designs. The KfK concept includes a gravity and natural circulation. The reactor containment is also to be passively cooled. core catcher with coolin catco gt coo d core han th l e melt; finally afterheadise e b th ,- o t s i t However, certain active components (piping, valves, storage tanks) will always remain sipated int environmene oth air-coolen a y b t d double-shell reactor containment. Internal necessar functionine th r yfo afterheaf go t removal. This implies that there exist residuasa l prestressed concrete structures arranged aroun reactoe dth r pressure vesse intendee ar l o dt non-availabilit afterheaf yo t removal whic admittedls hi y very low. preven reactoe th t r pressure vessel from being blown awa cashigh-pressurn a yi f eo e path. othed an r 0 AimproveP60 designR dLW s thus essentially attemp reduco t probabilite eth y Special structure envisagee sar orden di ensuro rt e tightnes reactoe th f so r containment of a core meltdown. during core melt events. a)

1 reactor pressure vessel, 2 primary cooling pump, 3 steam generator, 4 pressnrizer, 5 cooling pipe (hot), 6 cooling pipe (cold), 7 control rod drives, 8 entry for safety injection

b)

1 RPV with core, 2 hot cooling pipe, 3 cold cooling pipe, 4 pressuriier, 5 storage tank for HP-injection, (1 of 2), 6 pressurized injectio hig7 n , hsyste2) pressurf o m1 ( e decay heat cooler water-pool8 , depressurizatio9 , n system «loadin0 1 , g water-pool (1 of 2), 11 water level after Heading, 12 containment, 13 containment spray system, 4 internal containment spray system, 15 containment

Fig.9 Concept of the AP 600 (a. general outline of the primary system, b: basic principles of after heat removal system)

19 Conta innont

diccharge channel

Ventilation

Fig. 10: Development concept by KfK (Karlsruhe) for pressurized water reactors (a: overview of the primary system, b: corecatcher)

Safety condenser concept under assessment

Fue: 1 l storage pool 1 : Safety condenaer 2 : Fuel buffer storage and transfer cell 3 t D«m. water pool In-contalnmen: 3 t refuelling water storage tank 3 : Steam generator

Fig. 11: NPI concept (a: overview of the primary system, b: safety condenser for ufterheat removal)

20 A core catcher is similarly envisaged in the NPI concept, and the reactor containment is fuel particles). This then guarantees practically complete retentio fissioe th f nno productn si also intended to release the afterheat into the environment with the aid of passive heat re- fuele th . Fig show3 .1 reactoe sth r design (modular reacto MWth0 r20 als d associe )oan th - moval mechanisms. ated fuel element concept. A core of this type cannot melt and the afterheat removal is self- acting. Attention should moreover be drawn here to the fact that this property is not re- essentiae Th l question convincingle b o st y answere demonstrated dan - re w ne thesy db o etw stricte MWthpowe0 a annulao n 20 dt a f f ro I . r core geometr elsr y (o ecor a e regioe th n i actor designs are: form of a plate) is used then it is possible to increase the output to 1000 MWth. The vessel concep always i t s decisiv achievinr efo g this output level. Ca core nth e mel reliable tb y coole core th e n dcatcheri ? This concep self-actinf o t g afterheat remova demonstratede b lo t als s oha , includine gth guaranteee b t i n Ca d thaexplosion2 H t detonatior so n problem avoidede b n sca ? following points:

Will steam explosion avoidede sb ? nucleae Th r stabilit reactoe th f yo r mus demonstratee tb d even during extreme reac- tivity accidents. Are recriticalities ruled out? The thermal stability of the core and thus the principle of self-acting afterheat removal Is the high-pressure path controllable? mus demonstratede tb .

Does the reactor containment always remain sufficiently tight? chemicae Th l stabilit core th e f ymuso showe tb n even durin extremn ga e ingresf so external media. Does a passive containment cooling function under all circumstances? The mechanical stabilit reactoe th f yo r core mus demonstratede b t , i.e. even wit- hex Wit PIUe hth S reactor, afterheat remova firss li t guarantee lengthr dfo y periods (e.gr fo . treme vessel damage no deformations of the core increasing reactivity must occur or weeke on evaporatiny )b watee gth r fro mlarga e pooprestressea n li d concrete vessel. After this extreme vessel damage particula a mus y ruleb e tb t dou r desigprimare th f no y this the evaporated water would have to be replaced otherwise, a delayed core meltdown circuit containment. would occur. Finally self-actine th , g remova afterheaf o l t fro reactoe mth r containment must alse ob Apar tquestion w frofe ma s arisin thin gi s proposed design with respec shutdowo t n safety, demonstrated. the consequences of a delayed core meltdown would have to be clarified here and also whether a core meltdown in the reactor vessel itself could possibly be cooled. 4. Conditions for a catastrophe-free nuclear engineering A completely different method of realizing afterheat removal, and thus of guaranteeing the retention of fission products in the fuel elements themselves, was undertaken for the high- These condition alsn expressee sca o b requirementx fore si th f m o n d i orden si realizo rt e temperature reactor thin I . s cas afterheae eth removes i t d fro reactoe mth r cor heay eb t con- future "catastrophe-free nuclear engineering": ductio head nan t radiation alon necessarf e(i y supporte naturay db l convection)- ma o n ; chine activr so e measure necessare sar orden yi realizo rt e this afterheat removal principle. self-shut-off of the chain reaction An essential boundary condition is, however, that the maximum accident temperature of the ro fuel elements will remain below a certain temperature (currently below 1600°C for coated self-acting afterheat removal : BasiFig12 . c desigPIUe th r Sno Reuclor, diagra ufterhcar mo t removal

22 a) Primary circuit: 1 pebbled bed core, 2 side reflector, 3 reactor pressure vessel, 4 steam generator, 5 blower, 6 hot gas duct, 7 surface cooler reflecto8 , r drive b) Primary circuit arrangement in the containment: 1 reactor pressure vessel, 2 steam generator, 3 primary cell, 4 surface cooler, 5 containment building

1600-

axiCOrfsl 2>?80cm below surface pebbld be e 1200-

200 400 600 600

c) Radial temperature in the modular reactor after 100 h (loss of coolant and loss of active decay heat removal) wit1 : h surface cooler withou,2 t surface cooler d) Temperature in the modular reactor depending on time (loss of coolant and loss of active decay heat removal) with surface cooler

Fig: High-temperatur13 . e reactor concept

23 - fue l u elemenl indestructibls ta e first barrier -to- a) reador pressure vessel as indestructible second barrier

reactor containmen thirs a t d indestructible bamer

independence of barriers from each other.

conditione Th s mentioned above have apparently already been fulfille somr fo d e nuclear facilitie exampln A intermediatse th s ei e storag f speneo t fuel element air-coolen si d (natural convection), thick-walled, dense cast-iron casks (se g 14)eFi .

This concept completely fulfil above-mentionee th s d conditions (merel thire yth d barries ri dispensed with, this functio takes ni ncast-iroe oveth y rb n wall) accordancn I . e wite hth definition given in Chap. 2, the principle of catastrophe-free nuclear engineering is realized here.

Similar conclusion apparentle sar y reache considerinn di g final disposal (e.g. direct final disposal of spent ceramic fuel elements).

Comparative consideration extene th f whico st o stabilite hth y criteri principlee th r ao f so iuu • self-action are satisfied for the various reactor types are given in the next chapter. c) c SO _ ^ ~ 20 N »^ . 5 Compariso Varioue th f no s Principle Afterheaf so t Removal 3: •— — ~-\ ~ 10 \ It is generally difficult to assess the various reactor designs since the concepts of active, 4, 5 \ to -, passiv inherend ean t safety feature usee sar quit n di e different senses. Sometime tere sth m 0) 3 s ^s ^ "absolute safety alss "i o applie classificatioA d decisivele th r nfo y significant problem field •«= 2 :*, ~~~ of afterheat removal can be undertaken on the basis of Fig 15. CO 1 0 ' •CaD Analogous to a proposal by Lidsky, MIT, four stages of safety can be differentiated- 0Sn 1n ?n sn V n 7S(

Absolute safety (leveachieves i ) 0 l reactoa f di r doe contait sno fissioy nan n products and there is thus no radiological hazard potential There is no afterheat here nor any Fig. 14: Interim storage of spent fuel elements in cast-iron casks operative temperature differences. a) storage facility, b) casks, c) afterheat (after: ANS for PWR fuel elements) Natural/forced Irfvrl 0 draught cooler no fission Jt/oducts «tuoluto 1 no »fier licat ufety no temperature «filTercncci t l

Second. l.e»el I aflccltcat removal only by containment licit conduction' »nd radiation without boundary »ny machine

APCOO NaK intermediate Level 1 loop ftftcr-Ueat rejp*v* »adultey lb s P»M| »elf»cling process«* (no pump)

•cUulLWR Intermediate heat exchanger Primary containment boundary

TypicaFig: 16 . l passiv systeR emAH (decay heat rejectio pool-typa r nfo e FBR)

Fig. IS: Concepts of safety in afterheat removal comparisoA reactorf no s wit activn ha e afterheat removal syste self-actind man g afterheat removal indicate essentialle sth y different behaviou thesf facilitro o etw y types (see Self-acting safet afterhear yo t removal (leve achieves i ) facilit1 e l th f d i designes yi d Fig. 17): and constructe thay sucn di twa afterheaha removes i t d fro reactoe mth finalld ran y fro reactoe mth r containmen heay tb t conductio head nan t radiation alone withouy tan In the case of failure of the active system (left-hand side in Fig. 17) which must al- machines. way assumede sb , eve witf ni h very small probability (i.e non-availabilite .th f yo AH greateRs i r than zero) corlarge-capacite a f th ,e o y light-water reacto type th e f ro Passive safet afterhear yo t removal (leve achieves i ) l2 machinef di self-actind san g currentl operation yi n would mel fissiote withihour2 th o d nt snan 1 product s would mechanisms (evaporation, condensation, gravity, natural convection) are suitably releasee b d int primare oth y system. Afte furthea r r perio abouf do minute0 2 t e sth combine removae th r dfo afterheatf lo . Fig show6 .1 typicasa l passive system. lower hemispherical reactodome th f eo r pressure vesse fissiole meltth d ns an products are passed into the reactor containment. There are several paths by which large quan- Active safety or afterheat removal (level 3) is then achieved if essentially mechanical tities of fission products can subsequently enter into the environment. A system of this equipment is used to remove afterheat. As is well-known, this is the case for all reac- t thermalltypno s ei y stabl thud ean s catastrophe-free nuclear engineerint no s gi ro cn tors currently in operation. guaranteed. to active aflcr-lical remuval «rlf-actiiij; .ifK'i -ln--.il rciiMiv.il right-hane Th d sid Figf eo show 7 1 . correspondine sth g accident sequenc sysa r -efo 01 tem with self-acting afterheat removal. Afte failur e l activth ral f eo e coolina d gan coolant loss, the fuel temperatures initially also rise here, although after a certain pe- riod the self-acting cooling mechanisms (heat radiation, heat conduction) become fully operativ ensurd ean e coolinmaximu e coree th th f I f .g o m accident temperatur- ere mains below 1600°C the ntotaa l fractio lesf no s than abou treactoe 10~th f 5o r fission product inventory is released from the fuel elements. Damage to the reactor pressure vessel or even the reactor containment by these insignificant release processes is

heal removal by loops licat removal onl y heab y l completely impossible. conduction and radiation machines with failure Fig show8 1 . comparisoe sth n carrie hert dou e analogousl fuee th l r elementyfo d san rates physical effects (phonons and quantum) without fai- demonstrates the completely different behaviour of the first barrier fuel element. The ques- lure rates 0 N V= tion of the independence of the fission product barriers in case of failure of active cooling is comparisof discusseo y wa demonstratey d b Fig n i an 9 .1 superiorite sth e th f yo ,r,~2»«rc self-acting afterheat removal concept. It is beyond doubt clear that a fundamental difference has been achieved in the safety behaviou f reactorro s using non-melting fuel element operatioe th d s an reactorf no s with

1ht self-acting afterheat removal.

This principle of self-acting afterheat removal has been tested in AYR, so far as the pressurized reactor is concerned. The reactor was operated over a long period without any active cooling. Fig. 20 shows the measured temperature distribution. If there had been an accident withou reactoe t coolinth n i rs systemgga maximue th , m fuel temperature would have stayed below 140 . Unfortunatel0°C experimene yth t without pressurallowet no s dewa to be carried out for political reasons, although the experiment had already been licensed.

claie Ifth m"catastrophe-freo t e nuclear engineering credible b o t s "i y made then impairment self-actine th o st g afterheat removal mus controllable b t e even under very extreme conditions. 500h I 1.3 h t

total destruction no destruction Fig firs0 2 . t lists once agai requiremente nth self-actinr sfo g afterheat remova thed an ln all fission products very small amount of specifies restrictions which mus discussee b t r whic detaifo n d i d h an lproo f muse b t containmente inth , of fission products possiblye alsth n oi (< 10->) in the supplied. Major additional disturbances which could impair core integrity upon failurf eo containment, possibly also in the environment active afterheat removal are as follows:

Fig: Compariso17 . f reactorno s with activ self-actind ean g afterheat removal (assumption i Ingres f extremso e quantitie r int reactoe ai oth f so r cor bumud fuee ean th l f po b):failure of all active afterheat removal systems and complete loss of coolant element graphit structurad ean l graphite (see Fig. 21). fuol clement of a fuel element of * ra«ltlnc can non melting core

high fission product Invcn- ver fefiow ylo n produc- In t toreoatior yP< j vontor coatinr yp« g Cl/coating » i O/parttclo( 10 ( Ε ) )

Integrity of t ho coating integrity of In« coating« g 1- glvo y activb n e »flcr-lioat voy »olf-»ctiob n g after- romovml hoat nmoval

destruction of tho coatings no destruction of the coa- cue l afailurth f o falltu-oo o f o f

1SOO*C 1800'C

SOOh Fig: Behaviou18 . firsf o r t barrier fuel elemen casn i t failurf eo activf eo e coolin ) presena g, t LWR with active cooling self-actin) b , g afterheat removal (example HTR)

i <;iti'4i nt«H i'c;

jfaihtro of cooling!

•only very small «mount« of fluion producte ««r rclcA*od from tho fuol clomonti

barriers 'art cno independent

Fig. 19: Independence of barriers from each other in cose of failure of active cooling, a) presen witR h LW tactiv e cooling self-actin) b , g afterheat removal (example HTR)

27 ro CO Begjnjjccident simulation (shut off the circulators) Reflector nose-core midheight Side reflector, core midheight Inner side Middle Outer sid; O Bottom reflector cn

Reflector nose above pebble bed CD 4-3ro Reactor shroud, core midheight c o -H Inner vessel, core midheight •:-•> u a i——l——i——r 40 60 80 Test duration fh] Fig. 20: Results of an experiment at the AVR:temperature in case of loss of active cooling (at full rector pressure

: CorrosioFig22 . reactof no r material functioa s sa temperaturf no e LIMITS OF PRINCIPLE OF SELFACTING DECAY HEAT REMOVAL Ingress of large quantities of water associated with unacceptable corrosion as well as • ingress oflarge amount watef so o t r possibl increasn ya reactivitn ei changey yb d moderation (see Fig. 21). core => reactivity effect <*• solution: lower heavy metal loading of fuel elements Chang corn ei e geometry associated with change reactivitn si y cause catastrophiy db c vessel failure. • ingres largf so e amountr ai f so to core => graphite corrosion o solution: The following modification reactoe th o st r design would completely eliminate these corrosion resistant silicon carbide additional difficulties (very extreme assumptions modulae th r )fo r reactor. coated fuel elements • catastrophic failur reactof eo r vessel Application of corrosion-resistant silicon-coated fuel elements. => reactivity effects <> solution: prestressed burst protected vessels Utilizatio f fueno l elements wit hsomewhaa t reduced heavy metal content thus adequately moderatin reactoe gth r even befor watee eth r ingres tha o risss o n tn e i : ConceivablFig21 . e limitation principle th n so self-actinf eo g afterheat remova possibld an l e solution reactivit possibles yi . Application of prestressed burst-resistant reactor pressure vessels so that deformations SAFETY CRITERIA AND PROVISIONS FOR THE EVACUATION OF of the reactor core are not possible. RESIDUAL HEAT FROM GRAPHITE GAS COOLED REACTORS R. LHEUREUX After the introduction of these measures no accident configurations are conceivable which SPT, Division Calculs nucléaires, coulR HT d e cascase th i nth tf e o doub t upo concepe nth f "catastrophe-freo t e nuclear Paris - La Défense, France engineering". . AGUILERA A Centrale nucléaire de Saint-Laurent A, La Ferté-Saint-Cyr, France It becomes obvious that investigations should be carried out to determine whether a safety concept of this type can also be realized for modified suitably designed and constructed Abstract LWRs. powee Th r GGCy giveb f nRof reactor evacuates si tha CO displaces i t y db fouy db r turbo-blowers

driven by steam fro2m the steam generator.

In order to evacuate the residual power after a unit shutdown, the turbo-blowers are driven by steam fro auxiliare mth y boiler SAINT-LAURENt s(a electriBUGEy d b r an o T A c, Ymotor1) s placet da 6. Overall Evaluation the shaft end as is done at CHINON A3. evene Ith nf theso t e cooling systems failing whils e reactoth t r remains pressurized, there ear Innovative nuclear engineering with a new safety quality is possible and must be realized in emergency systems. future. With respec concepe th o t f self-actin to g afterheat removal followine th , g demonstrations of proof are required. The latter refer both to innovative high-temperature At SAINT-LAUREN shutdowTA n exchanger placee sar higa n di h position thus creating natural convection t CHINOA steae . th m3 N A generator above sar cor e eth e which also creates natural reactors and to conceivable solutions for innovative light-water reactors. convection t BUGEA . motordrivea Y1 n blower ensures tha circulatesCO t steae ;th m generatos i r

kept in operation. 2 Proof of thermal stability. At SAINT-LAUREN BUGEd an TA wheY1 reactore nth undee sar r sligh r pressurai t e four fans evacuat residuae eth l powe open a n ri circuit CHINOt A .reacto e th 3 Nkeps closerA a i n ti d circuit Experimental determinatio temperaturf no e profile corn si e region self-actinr sfo g and the turbo-blowers are driven by their electric motors. The steam generator remains supplied. afterheat removal. These systems have proved efficien s aftea t r twenty year f operatioo s n ther s onleha y beee non inciden t CHINOa t , threNA3 e days before definitive shutdown. This necessitated going oveo t r Validatio computef no r programs. natural convection which was done according to procedures without the slightest problem. 1 Introduction Proo nucleaf fo r stability r> calculation extremf so e reactivity accidents. In normal operating conditions, the heat released in the core of GGC evacuates coolan2 i R CO a ty b dflo w ascendinge whicb y hma , Proof of chemical stability •* measurements and calculations on severe corrosion CHINOat descendingr sa o , NA3 SAINT-LAURENBUGEd t a an s Y1 ,a T due to the ingress of foreign media in the reactor core, development of corrosion- A. resistant fuel element cord san e structures. In these reactors e flui ,th circulates i d y foudb r turbo-blowers (figur . These1) drivee ear y steanb m fromaie mth n circuitn I . th ed drop evenro ,f o t therefore e floth wf i ,feedin e turboth g - Proof of mechanical stability •* calculations on core destruction due to blowers becomes insufficient shoult i , ensuree db d thae th t mechanical damage; developmen burst-resistanf to t reactor pressure vessels. latter are driven by despatching steam from the auxiliary boilers; this is provided for at BUGEY 1 and SAINT-LAURENT, whilst CHINON A3 has electrical motors on shaft ends, which can CO driv e turbo-blowereth 0 rpm50 .t sa Cross-sectio1 G FI turboblowea f no r blowine Th g functio t SAINT-LAURENa n subjece th a s f o tTwa 3 Provisions for the evacuation of residual heat in the event of reliability study (1) parametere th , whicr updatee fo s har d each los turbo-blowersf o s , witreactoe th h r under pressur d filleean d year (2) to ensure that the study's effectiveness does not with CO2 deteriorate. In 1990, monitoring of these parameters showed a clear improvement in the blowing function, mainly thanks to the improved availabilit auxiliare th f o y y boilers 3.1 CHINON A3 Durin normaa g l shutdown sequence, whe fuee nth l element The level of the steam generators was determined with regard to temperature e sufficientlar s y low e reactoth , venteds i r , with that of the mid-plane of the reactor to allow natural circulation ventilator SAINT-LAURENt (a s d BUGEYTan ) replacin e turboth g - via thermosiphon, when the unit is shut down, which is sufficient blowers for the evacuation of the residual heat in open circuit. to allo e residuath w e reactol th e heaevacuated b f o o tt r . Figure 2 (enclosed) shows the drop between the exchangers and the At e reactoCHINOth , NA3 r remain closen i s d circuits ha onc t i e core been turbo-blowerventede th d ,an electridrivee e sth ar y nb c motors until the residual heat is sufficiently low to be evacuated by thermosiphon. 3.2 SAINT-LAURENT A (51 As the integrated concept (Figure 3) of the reactors of SAINT- LAURENT A comprises steam generators located below the core, Safet2 y criteri e respecte b e fueo t th al r element) fo d (4 ) (3 s another method for evacuating the residual heat had to be found: the shutdown exchanger, comprising 6 elements arranged regularly In normal operation fuee permissiblth s li f t o ,i % 1 r efo around the periphery of the reactor and in the upper part of the element havo st e cladding temperatures greater than 515*d Can reactor vessel. uranium temperatures greater than 650'C. e shutdowTh n exchanger circuit, whic s initiallhwa y designeo dt In the event of an incident, the following deviations are evacuat residuae eth reactoe l th hea f o tr after shutdowe th f no permitted: blowers, i.e. approximately 6 hours after control rod drop, was modified to allow emergency back-up of the reactor cooling and Cladding temperature2 C0 n i s maintenance of pressure vessel integrity following an accident resulting from a common mode failure causing loss of blowing, Excessive values up to 600'C are tolerated in the event of an followed immediatel controy yb dropd lro tesA .s carrie twa t dou incident. However, they must be of as short a duration as tcharacteristicw o ne chec e th k shutdowe th f so n exchangers. possibl minutes)3 < ( e e totaTh . l cumulative duratio f thesno e incidents mus t exceetno minute0 1 d eacr fo sh fuel element. The C02 is circulated by thermosiphon (approximately 310 kg/s when brought into operation) betweesourcet ho e n,th comprising There is no limit on the speed of temperature change for these th cole th cored d source,an , comprisin shutdowe gth n exchanger incidents. with a demineralised water flow of approximately 176 kg/s. Cladding temperaturer ai n i s During operation of the reactor, the shutdown exchangers are draine avoio t dy corrosion dan . The maximum cladding temperature at the moment of venting must not exceed 350'C. This temperature must not be maintained for n ordeI o fulfit r functios it l f evacuatino n e residuath g l heat, mor ehoursw thafe a n. the shutdown exchanger mus capable tb beinf eo g brought into operation from the control room of the unit or the emergency The temperature decrease shall be such that the cumulative dwell back-up building, and then controlled and monitored from the time at a temperature of around 220'C does not exceed one week. latter, independently of the units

shutdowe Ith f prolonges ni week7 o dt s beyon 6 necessar e th d y The shutdown exchanger must be brought into operation in less for normal maintenance shutdowns claddine ,th g temperature must than five minutes following tota le turbo-blower th los f o s n i s not exceed 150'C (this value being principally determined by the order to allow the maximum temperature of the CO2 in the upper CO phenomeno hydridingf no ) e reacto th pare limite b f o to rt arouno t d d 270'C thermae Th l power evacuate exchangee th y db approximatels ri x 2 y ro Ça^5 e temperaturth d an 1W 7M démineraiisef eo d water evacuatee th y db exchanger at the outlet of. the shutdown exchangers is < 90'C. In orde evacuato t ravoid an thermae dW M th e 7 1 l x powe2 f ro . loi) exceedin a gtemperatur e feef 60'eo th d C n i tank n additionaa , l demineralised water/supplementar watew ra y r coole s installeri d CÏ/JM shutdowe ath t n exchanger outlet.

) BUGE3 3. (6 1 Y SAINT-LAURENAt sa steae th m, TA generator e locatesar d beloe wth core. Another metho evacuatinf o d residuae th g l heas twa designed. The choice made for BUGEY involved the following factors: 2 circulationCO a) , wit reactoe hth r under pressure, provided by a motor blower known as the "back-up blower", exchangen a ) b r feed water circuit comprisin motorga - driven pump and its connections, diesea c) l generator providing electrical back-ue th n pi even networf to k loss. The blower and the feed water circuit are designed and sized for an intervention as close as possible to rod drop. There is a "local control room" for the operation of this equipment. This is intended, if necessary, for the fall-back of the operatin glocates i team d an ,d outsid unite eth .

The motor blower is simple, robust and sturdy, in order to best ensure its availability and ability to be brought into operation, and so that only a minimum of constraints are placed on it during operation. It must be possible to periodically test its operating condition during operatio powee th rf no statio n without the latter being disturbed in any way. The motor blower is located inside the pressure vessel, in the EAST tunnel (Figure 4) . It was designed to be capable of being dismantled in such a way as to allow access to the pressure vessel and, if necessary, to open up a passage via the EAST tunnel betwee interioe pressurne th th f ro ee vesseth d lan "maintenanc f contaminateo e d equipment" area whic s locatehi n i d the intermediate bay. In orde o fulfit r l this condition, this s machinit d ean

FIG 2 Flow diagram for CO circuit at Chmon A3 auxiliaries comprise three assemblies- 2 132.300 131.100

8 c&ble13 s

588 cibles 125.425

Not I I lurbo-touirtâiu «: t «mené«tt * d«n t plaI « n d« coup*. Ttnilon *ppliqu4* k chiqu« clbl« 186 I. «Ifc. Longutui ioi*l* d«t clbltt 191,m Sk Poid» . + dill« In). 37,8 km

FIG. 3. Vertical cross-section of the Saint-Laurent A reactor.

33 34 a) an internal part (comprising the C02 induction and 1100 MWth before rod drop, the thermosiphon was set up on four discharge equipment) pressure linketh o t d e vessel e temperatureth loop d san s were monitore hours3 r fo d. structures (skirt); this part can, if necessary, be dismantled piec piecey eb , This test allowe readjustmena d calculationse th f o t d ,an forecasts were thus made for thermosiphon operation intervening b) a moveable part (comprising the body of the blower and after: its shaft collar) e evacuatewhicb e n th hca o t d exterior en bloc, a) stable operation at 1550 MWth on 4 loops (inlet temperature = 240'C, outlet temperature = 410'C), c) a carriage device allowing the moveable part to be evacuated from the tunnel. b) stable operatio loopMWt2 0 83 n h o s t na (temperature = 240'C, outlet temperature = 410'C). e characteristicTh motoe th rf o s blowe s followsa e rar : Thes caseo etw s represen mose th tt critical situations froe mth point of view of the quantity of energy evacuated by - blower thermosiphon, taking into accoun numbee tth loopf ro n si operation and therefore the reserve of water contained in the - nominal flow 1450 kg/s main exchanger, functioning effectively. - nominap ld mba0 10 r - feed voltage 5.5 kV At the moment of tripping, the water reserve can be assessed at - electrical power absorbed 700 kW 180 tonnes. For a prior operating regime of 4 loops, 410"C and - synchronous speed 100m rp 0 1550 MWth steae th , m generato hour2 r n minutes5 driei s1 t sou , - nominal speed m rp 0 98 taking into consideration a thermosiphon on 4 loops. For a prior operating regim loops2 f e o MWth 0 , stea83 e 410' ,th d man C - cooling circuit generator dries out in 1 hour and 40 minutes, taking into consideratio nthermosiphoa loops2 n no . - source démineraiised water - cooling water flow 2 kg/s Control of reactor temperatures requires the start-up of one of - inlet temperature between 13'd 33'Can C the auxiliary feed pumps, which may be supplied with electrical - permissible temperature rise 10'C power fronetworke mth , from auxiliary generator fror so e mth emergency back-up panel, whic diesee linkes hi th o ldt generators of CHINON A2. 2 sweepinCO - g evene I th noperatiof to n prio 155o rt 0 MWth d takin,an g into - source C02 reactor requirements consideratio nthermosiphoa loops4 auxiliarn a n no f ,i y feed - flow 10 g/s pump canno broughe tb t back into operation maximue ,th m cladding - pressure 1.01 to 1.1 P and uranium temperature reaches 470'C afte hourr3 s (Figur. e5) - temperature 50'C A lack of feedwater thus presents no danger for the installation - seal and the environment, and leaves a very long time for emergency - minimum dp of actuator control 4 bar circuit broughe b o st t into line. - opening time s 0 1 <

4 Test d studiean s s initian a r Fo al) powe f 155ro 0 MWth wit loop4 h operationn i s : 4.1 CI^NON A3 (71 (81 (9) (10) if a transfer is made to 2 loops or 1 loop, the water reserve last morr sfo e tha hoursn3 mose Th .t interesting ADecembe1 1 tes s carrien o wa t t r ou d197 permio t 0 t forward case is the point at which reduction to 1 loop takes place 00 Ol calculations based on experience. With the reactor operating at (Figur. 6) e u o>

„£DF3 1550 MWTH ASSECHEMENT DES ECHANGEURS m JK£ ° Saß D3 F 155 MWT BOUCLE4 H !SA BOUCL E

J# ~ ftn»"V <3 SO C 2 n~~*~ * 5tO ' «« ta u M l'n S C...xAW •» ^ f > » t ^ /* T^ .: * Pn «4 h'r i< i CAl E *a» \ " Sto ~1^ "P — - \ •f-r — ui 1 £ 40 4 o . | —— 0 4Jn * i X' OCIP w y ! : u w — • i — u» * " o l ~ -ftfi v N f ~ •^ _35t^*^e>*. 1 — fsz* -L**v •-™ * 4J0 ^^ -i^V 1 — >, \ \ ,- V """s. * r— u ^-*'' V .W -)" ^ 1 — n j^-* r V, \ j? 4CÖ \, r — X. :: ' .« -s 34a ^. / ^f: 1 2 r- 3Äj .W 1 'O' 10* 10+S 10*« lV3~ 1x 0" 1 0* 10M llf TEMP SECONDS( ) ES (INSTANT INITIA 0 SECONDES1 • = L ) TEMPS(SECO DES) (INSTANT INITIAL « 10 SECONDES!

FIG 5 Temperature behaviour during thermosiphon test following rod drop FI GTemperatur6 e behaviour dunng transitio thermosiphoo nt n (reductio loop1 o nt ) with initial powe f 155o r 0 MWtloop4 d operation shi an n n initiaa r Fo bl )MWt 0 powe83 h f witro looph2 operationn i s : 4.2 SAINT LAURENT (111 (12) (13) (141 itransfea f loomad1 s i ro pet afte minutes5 r watee th , r reserve lasts for 2 hours 15 minutes (Figure 7) Various tests have been carried outparticulan i , qualifo t r e th y GITA code, which is used for the calculation of the temperature changes in integrated-type reactors operating with thermosiphon. These tests were carrieAprin i Juld t an ly ou d 196 SLAt 9a n O 1 29 October 1986 furthea , r tes s carriet 240'Cta wa t ,ou d BOUCL1 £OF0 HWTA 83 2 3EH following the modification of the shutdown exchanger The shutdown exchange SLAf ralss o 2wa o brought into operatiot na d the180an 'n 240*e Ce leve th (temperaturth f lo t a 2 CO f eo shutdown exchangers) on 4 and 6 June 1984. These tests confirmed thaequipmene th t s functionintwa g correctly e heaTh .t evacuated shutdowe bth y n exchanger during thesMWt0 1 d testo han etw s swa K MWt4 1 h respectively. 3 «• « •* M 3 fr m JH irt : >. IX. ei» : 5 Provision evacuatioe th r fo s residuaf no l heat froreactoe th m r tu f ' »• f Ti n ,.E ix OSftVlfJff in vented state Z ^\ J u W MI ^ f. ] T •^ t c il-itt. p/irthr T< ti • ^ U \-=J ^. CHINQ1 5. 3 NA M U y During a shutdown, the reactor, although vented, remains in Ssi X closed circuit. The steam generators are in operation and the 'S turbo-blowers continu drivee b o et y thei nb r electrical motors >, > until the residual heat is low enough to be able to be evacuated \ via thermosiphon. "\ \ ^ .1 ^ • \ > M • • « 5.2 SAINT-LAURENT A ' ^ s sv Durin gshutdowna reactoe ,th venteds open ri i nt ,bu circuit s ,a taker showai ne Figurn ni Th fro exterioe th m. e8 s ri «y • \ k conditioned to give it low hygrometry and a suitable temperature 1 t lowewhic no poin w rs hde i thate th n(correspondin cole th d o gt } ff u i u t r s, i. points which may exist in the pressure vessel). It enters via TEMPERATUR E IDEG I > 1 N . U ! » l «M. ^ the bottom of the reactor and circulates upwards (which brings > 1 W . 1 « 1C uu S( about an inversion of the flow in the core with regard to the ' i direction imposeturbo-blowers)e th y db . Having crossee dth 1 1 core, the air descends once more via the containment annulus, 1 leaves the pressure vessel via a second tunnel, and is evacuated l e stactth o k after filtration d coolintemperaturs an ,it f i g e exceeds 60*C.

o H S 5 o" i o" 10" IQ* TEMPSCSE .NOES) (INSTANT INITIA o SECONDESt • L ) circulatee b y ma ventilator4 r y db ai e Th paralleln si , whilst plugs are located on the induction of the turbo-blowers to r prevenfroai m e bypassinth t reactore th g, 10 n FigureO .d an 9 s FIG 7 Temperature behaviour during transition to thermosiphon (reduction to 1 loop) with e timth e availabl re-establishinr fo e ventilatioe th g e th n i n initial loopMWtpowe2 0 d 83 operatio n sf i han o r n event of its total loss 7 and 15 days after rod drop is shown CO 5.3 BUGE1 Y The air is taken from the exterior and conditioned. Unlike at CD SAINT-LAURENT, the ventilation allows ascending or descending During a shutdown, the reactor is vented in open circuit when the residual power reaches no circulatio e a corfunctio e th interventions ea th n i n f no s more tha maximuna MWt8 4 f mho planned below the core or in the containment annulus. This is obtaine manipulatiny db relative gth e position4 e th f so The pressure vessel has 6 tunnels in its lower part: throttling valves, controllabl meany eb f servomotorsso , with which the wind box is equipped. On leaving the tunnel, the air e turbo-blowersth r fo 4 - , is evacuated to the stack after passing over filters which can - the EAST tunnel is occupied by the motor blower, as indicated operate at a maximum permanent temperature of 125'C above, during normal operation of the power station, - the 6th tunnel, on the WEST side, is used for the inlet and e circulateb n ca ventilator4 r y b dai e Th paralleln i se th d an , outlet of the shutdown ventilation air installation of a fifth is planned.

FIG 8 Principle of ventilation during shutdown 700.0- 700,0

650.0 650,0

600.0

550.0

500.0

o

«00,0

o 350,0

S. MO-° E t> *~ 250.0

50,0

0.0 0,01.0 2.0 3,0 4,0 S.D 6,0 7.0 1.0 9,0 10.011.012,013,01*.015,016.017,018,019,0200 2 * t B I 7 1 6 J 5 1 U , U 2 1 1 1 0 .1 9 B 7 6 5 4 3 2 time in days

QVENT=40KG/S TAIR=30DC,LOSS OF 3 FANS AT 7 DAYS QVENT=40KG/S TAIR=30DC,LOS DAYFAN5 3 1 F T SSO A MAXIMUM CLADDING TEMPERATURE MAXIMUM CLADDING. TEMPERATURE

Co FIG 9 Saint-Laurent A, Unit 2 Time available to re-establish ventilation after reactor FIG 10 Saint-Laurent A, Unit 2 Time available to re-establish ventilation after reactor

Numbe triof turbo-blower3 ro f so s occurring together REFERENCES 23 11.77 Following rod drop, blowing was maintained for 46 (1) HT-013/9/85, 4 February 1985 second motoe th turbo-blowey s b r d blowean , 5 s rwa . rNo BOUISSO) (M U then starte . Twelvup d e minutes afte e incidentth r , turbo-blower wer6 d e an brough 4 's o sN t back into REACTUALISATIO L'ETUDE ND FIABILITE ED FONCTIOA L E ED N operation. SOUFFLAGE DES CENTRALES DE SAINT-LAURENT-DES-EAUX (UNGG) 7 Operating incidents which caused an emergency back-up device to (2) HT-051/91-44A, May 1991 be brought into operatio evacuatioe th r fo n f residuano l heat BALMAIN (M)

Onle incidenon y f thio t s type occurre Jun2 1 eCHINOt n a do , NA3 CALCUL DES INDICATEURS DE RETOUR D'EXPERIENCE DE LA CENTRALE 1990, three days before final shutdow unite th .f n o Following DE SAINT-LAURENT-DES EAUX (UNGG) POUR L'ANNEE 1990 rapid passag turbo-generatore th f eo o zert s o load e controth , l rods dropped and the flow of the turbo-blowers fell to zero in (3) Technical Not C 88-201SD e 5 y 198(DR)Ma 80 1 , less than a minute. It was not possible to operate the turbo- Par f programmo t e 5001-9-3 blowers by means of the de-superheating release because of the P. MILLET (SDEEC/SDC) failure of the circulation pump which led to a lack of cooling to the condenser The reactor thus transferred to thermosiphon on ELEMENTS COMBUSTIBLES ANNULAIRES DANS BUGEY 1 four loops, and then fairly rapidly (in approximately 10 minutes) CONDITIONS D'UTILISATION on one loop after intervention by the operators. (4) Technical note SDC 88-2033(DR), 16 September 1988 The thermosiphon was deliberately maintained in operation, so as Part of programme 5001-9-3 to avoid excessive coolinmoderatoe th f o g d thuran s losf o s P. MILLET (SDEEC/SDC) ELEMENTS COMBUSTIBLE GRAPHITE D E AM SEA MHTGE BASETH R R SSOURCFO E TERM CONDITIONS D'UTILISATION AND CONTAINMENT CONCEPTS (5) SAINT-LAURENT Al safety report. File B Chapter 4-7 P.M. WILLIAMS (6) BUGE safetY1 y report. Par Chapte1 t 1 r3- US Department of Energy, Washington, D.C., (7) CHINO 3 safetNA y report, Volum ChapteI eII r 3.2.1.1b United State f Americso a (8) HF-011-012/9/71 Cc-061, 28 January 1971 GOURIOU (A) - HOURTOULLE (F) - MOUNEY (H) Abstract

CHINOI NII Significant difference transienn i s t respons materiald ean f constructioso n give high temperature gas- EVACUATIO S CALORIENDE REACTEUU D S R THERMOSIPHOPA R N cooled reactor (HTGRs), the potential for alternate approaches to the key issues of selection and analysis of postulated accidents, the radionuclide source-term mechanisms, containment design, and (9) HF-011-012/16/71 Ca-061, 25 March 1971 emergency planning MHTGRe th U.Sr e sitinth e Fo .e . th , Environmenta gus goao t s i l l Protection GOURIO - HOURTOULL ) (A U - MOUNE ) - LHEUREU) (F E (H Y ) (R X Agency's protection action guidelines (PAGs) for notification, sheltering, and evacuation, rather than the NRC's doses of 300 rem to the thyroid and a 25 rem whole body. This paper discusses how the CHINON III design and inherent characteristic of the MHTGR lead to a radionuclide source term of prompt and ESSAI THERMOSIPHO DECEMBR1 1 U ND E 1970 delayed components. Option MHTGe th r sfo R containment desig discussee nar thir dfo s source term and give support to the concept of a vented, low pressure containment in comparison to the high (10) CHINO safet3 NA y report. Volum ChapteI eII r 4.1.4.2 pressure, low leakage containments characteristic of LWRs. The source term concept has been proposed to the NRC and is currently under review. (11) D 5088/84-141, 22 June 1984 BATOUFFLE. TP I. Introduction and Suanary COMPTE RENDU D'ESSA QUALIFICATIOE ID N Significant differences In transient response and Materials of MIS SERVICN E L'ECHANGEUE ED R D'ARRET construction give high temperature gas-cooled reactors (HTGRs)n ,i A 180'C PUIS 240'C LES 4 ET 6 JUIN 1984 comparison with light water reactors (LWRs), the potential for alternate TRANCHE 2 approaches to the key issues of selection and analysis of postulated accidents! the radionuclide source-term mechanisms, containment design, and emergency planning. The differences and alternate approaches are (12) D541-GT978-FVR/CM No. 594/86, 29 MAY 1986 derived from their slow response to core heat-up events because of low FEVRE (P) core-power densities, the very high temperature the ceramic-coated particle fuel can withstand before substantial fission-product release, and the chemical inertness of the hellun coolant which negates the ESSAIS CIRCUITS RAIE SLA2 possibility of fuel-coolant Interactions. For the Modular HTGR (MHTGR), additional safety characteristics result from its design for passive (13) D561/NT/A/41/87-35, 12 January 1987 reactor shutdow passivd nan e decay heat removal passive Th . e meanr sfo NOLO) (P T decay heat removal is one of the subjects of this paper and is central to the retention of practically all radionuclides within the particle fuel NOTE TECHNIQUE RELEVE D'EXECUTION D'ESSAI during a postulate core heatup accident. FONCTION RAIE TRANCHE 1 ESSAIS D'ENSEMBLE Early nuclear reactors were snail, used crudely estimated source terms, (14) SLA/BUS/RAIE/100, 27 January 1986 t havdino de containments estimatee th d ,an d consequences were judgeo dt TASSY (J.P) be mitigated by distances to populated areas. It is historically interesting wels ,a fundamentas a l ensuine th o t lg discussion o not,t e PROCEDURE D'EXECUTION D'ESSAIS thadocumene th t t entitled, "Calculatio Distancf o n e Factor r Powefo sd ran ECHANGEURS D'ARRET ESSAIS D'ENSEMBLE Test Reactors (AEC 1962)" establishe well-knowe th d n "TID Source Term" which postulate legala s , quantitative releas radionuclidef o e e th o st containmene th f o nobliodinese e X th th f I 100 o t ef f d o o %gas ,an % ,25 solids r sitinFo . givea g n powerplant D sourcTI e e th ,ter useds i m , •t» together with the reactor containment's expected demonstrable leak rate release of fission products from the fuel. Because the delayed release is N) e metrologicaanth d l conditions pertinen e siteth o calculato ,t t e th e carried out at atmospheric pressure, a high pressure, low leakage boundaries of the exclusion area and low population zone on the basis of containmen e necessarb judgeo e servet t b i t n s o no a didentifiablytca sn e allowabl y doseda radioiodinf s0 o 3 hou2 ed an r e thyroie iodinth d o t ean d function in this case. whole tth o e body dose frototae mth l radionuclide releases. Ovee rth years, refinement and conservatisms have entered into the calculations and II. Desig d Functio Decae an n th yf o nHea t Removal System work is underway at the Nuclear Regulatory Commission (NRC) and elsewhere to replace the TID source tern by a mechanistically derived value and to The nuclear island features of the MHTGR power plant are shown in account for severe accidents involving core melting and fuel-coolant e reactoFigurTh d hea. 1 ran e t transport component e housear s n i d interactions. For the MHTGR, the siting goal is to use the U.S. separate vessels connected by a concentric flow cross duct vessel, with Environmental Protection Agency's protection action guidelines (PAGs) for all vessels housed in an underground cavity or silo. The steam generator notification, sheltering, and evacuation, rather than the NRC's doses of vessel, which also contains the helium circulator and the pressure relief 300 rem to the thyroid and a 25 rem whole body. The PAG dose controlling train, and the reactor vessel are housed in separate compartments of the the MHTGR'source term and containment analysis is a 5 rem thyroid dose at cavity which, under normal conditions, do not communicate. Should an the plant site boundary. overpressure condition occur, such as could be caused by a main steamline rupture, pressure woul relievee b d d through vent e steapathth n mi s e For. VraiTh tSt subsequene nth HTGd an R t design r largfo s e gas-cooled generator portion of the reactor building. Blowout panels connect the two reactors in the United States have used Mechanistic interpretations of the TIO sourc MHTGe edepartes th terha Rd an m d entirely from this definition. This paper discusses the factors that enter into this mechanistic desige calculatio th d inheren w an n ho d tan n characteristi MHTGe th Rf co sourca lea o t d e ter prompf mo delayed an t d components. e Optionth r sfo MHTGR containment design are discussed for this source term and give support to the concept of a vented, low pressure containment in comparison to the high pressure, low leakage containments characteristic of LWRs. The source term concept has been proposed to the NRC and is currently under review. For the purposes of the present paper, a postulated accident involving steam ingress into the reactor caused by a steam generator tube failure is taken for determination of the source term magnitude and time sequence. Other accidents continu e investigatedb o et t thia st s ,i bu tim t i e believed that the steam ingress sequence is the most illustrative of the promp delayed tan d characteristic MHTGe th Rf so sourc e term thin I .s sequence, steam enters the reactor with the eventual result that the pressure relief valve opens a.id aftew reliefe ra f cycles sticks open, causing the reactor primary system to depressurize. This is followed by a full duration core heatup event in which decay heat is removed passively by the reactor cavity cooling system (RCCS), to be described in the next session. Scram occurs eithe y insertiorb absorbef o n r materiar o l passivel y negativyb e Doppler feedback. The prompt portion of the source term occurs during the depressurization e primarth f o y syste mattea n i mminutesf ro , whil delayee th e d portion occurs ove perioa r f days o prompde Th . t source term contains radionuclides circulating wit e heliums dominatei th h t bu , y "liftoffb d " of "plated-out" radionuclides previously deposited on the cooler portions of the primary system surfaces. The delayed source term develops from the failure of a very small fraction of the fuel particles and occurs during the lengthy, core heatup phase of the accident. The fuel particle design has been described elsewhere, together witfailurs it h e modes affecting source term characteristics (Inamati, et al., 1989). This delayed radionuclide release occurs at atmospheric pressure since the prompt high pressure releas s beeha en atmosphere venteth o t d e well before significant FIGURE 1: ISOMETRIC VIEW THROUGH REACTOR BUILDING compartments e desigTh . e sucs reactoi nth hn i tha ri pressura tps 0 1 f eo compartmen texceedede b coul t no d , whic sufficiens i h o protect t safety grade equipment in the reactor compartment from overpressure damage. INTAKilfXHAUST The reactor compartment contains the Reactor Cavity Cooling System (RCCS) STRUCTURE r removafo f heao l t transmitte frot uninsulatee i mth o t d d surface th f eo reactor vessel. This system, shown schematicalla Figurn s i yi , 2 e naturally convective air-cooled system of ducts and panels that is open to the environment but is closed within the reactor building. It is a safety grade structur d operateean s continuously. Wheforcel al n d reactor coolin losts w probabiliti glo ,a y event removet i , s decay heat fully passively at a rate sufficient to maintain fuel and vessel temperatures below acceptable limits. The performanc RCCe th S f witeo h respec fueo vessed t an l l temperature EXHAUST DUCT over time is shown in Figures 3 and 4 for reactor conditions of pressurized and depressurize, respectively. These curves are taken from independent calculations performe k RidgOa y eb d Nationa l Laboratorn i y support of NRC's on-going review of the MHTGR (Williams, et al., 1989). These calculation closn i e e ar sagreemen t with those performeE DO y b d Aft COOUHG PAMEIS contractors shoult I . notee b d d thamaximue th t m cor vessed ean l temperature approachee sar about a d hours0 t8 days 3 r ,o , followin sloa g w buildup. III. Source Term Characteristics Iodine-131, which has an 8 day half life, is considered controlling in the source term description applicationd san s discussed below. While final calculations will take into account the full spectrum of radionuclides, iodine-131 well illustrates the phenomena to be considered and is likely FIGUR : E2 PASSIVE REACTOR CAVITY COOLING to be confirmed as the dominant radionuclide in the containment design basis. Tabl (Inamat1 e al.t e , 1989) characterize summarized san e sth rol thif eo s isotope with respecinventorys it o tt , inventory location, tim f releaseeo d releas,an e mechanisms. Four inventory locatione ar s identified; (1) circulating with the helium coolant, (2) plated out on 1COPC primary system surface ) associate(3 s d with defective fuel particlesd ,an (4) contained within standard fuel particles. Except for the standard fuel particle inventory e inventorie,th s give nominae nar d subjecan l o t uncertainties being addresse technologe th n i d y development program. lOWC uraniue Theseth d ,man dicarbide inventory withi defective nth e fuel particles, whic releases i h d rapidly under hydrolysis conditions, fore mth prompt sourc releasee ear termattea d n man i d minute f ro s following failure of a relief valve to close. Although uncertainties exist, it is KtfC evident thae prompth t t source term wil e sufficientlb l y smalln ca tha t i t be vented froreactoe mth r building subsequenf I . t research determines thaprompe th t t source ter larges i m r than currently predictede b n ca t i , vented through a filter on the relief train to meet goal release

0 20 40 80 » 100 120 140 180 180 200

HOURS

O MAX. COKE VESSEX AVQ+ MA . L O CORE TABL . MHTGE1 R SOURCE CHARACTERIZATIO DOMINANR NFO T NUCLIDE CONTRIBUTIN THYROIO GT D DOSE FIGUR : EDEPRESSURIZE4 D CONDUCTION COOLDOWH WITH ftCCS TEMPERATURE F COR SVESSEO D EAN . TIMLVS E OMRECA REFERENCE CASE) UDIMWCLIHC urymoftY TXMXNO OF mi*» MKouiriM* MOKCC (Cl I.llll RBUASK ouxAcmiunoii can not non nottXT encapsulated in the particles during manufacture remain to contaminate eucoiT graphite regions exterior to the coating barriers. Defective fuel is that fuel which contain manufacturt sa durinr eo g operation coating fractures 1) Circulating 0.02 BlMt** _. No» or weaknesses which effectively negat coatine eth g barrier o fissiost n (•• Deprea«.) product release during core heatup. As this release is proportional to temperature inventore ,th y is released ove rperioa hourf o ddayd d san san 2> Plateoat zo.o •laut** ». No» is expecte roughlo t d y follo e corwth e temperature rises give Figuren i n s (Be D*pr**a.) 3 and 4. As releases froa the delayed source term occur effectively at Holitur* (Water Ingreia) atmospheric pressure, these release subjece sar differeno t t transport 1 Initially phenomena than those associated with high pressure, low leakage Defective containments. •article* a) Ceetaavination •J hour* - 4*7* Temperature No« IV. Phenomenologica Desigd an l n Development Needs (Lot f Ported*o • 0*pr**i.(• ) Cool ID») Laboratory testing and operating reactor experience with ceramic particle fue botn Unitee i l hth d State Germanyd san , together with modern design b) Defect« »•3 hour*l-d*jr* Temperature Plow and analytical techniques, give good confidence in the expected source (Loll of Forced (Be Depre».) term behavior and in the performance of the reactor cavity cooling system. Cooling) A program of testing and computer code verification and validation is Hoieture No» being developed. (Water Infre««) * Depr««.(I )

r RCCSFo , performance validatio coordinatioy b n w ProductionNe wite th h n ») Standard »»10** > day« Tenperature ._ Reactor (NPR) versioMHTGe th expectes i Rf o n thif i d s desig selectes i n d Partiel» (No Event for the tritium production application. If the MHTGR is not selected, the Identified) following other options being considered are ) ful:(1 l reliancn eo computer model d existinan s g correlations with confirmatio y testinb n g during start-up and power ascension, (2) a cooperative program with the PRISM liquid metal reactor which use a ssimila r passive heat removal . 1 Approximatel 1 (th 7 fractiony2 ea inventorUC th f o ) non-intacn I y t system, and (3) development of an international, IAEA cooperative research particle «ubjec» 1 « o releaat t • lim* am r n f i f »imiteeo « » under program n deca(CRPo ) y heat removal. hydrolyxing condition« encountere n rari d e KHTCR accident«. studyine War e g several design option o achievst levea e f reductioo l d an n locations givine ar r curren e ou gW . t attentio possible f th o o e t n us e attenuatio f radionuclldeo n o assurt s e thadesige th t n will meeU.Se tth . filters on the relief train, schematically illustrated in Figure 5, on the PAG goals at the site boundary with margin. These options are enlargement reactor building itself, d reductioFigur an e reacto , th e6 n i n r building of the site boundary, elevated release through a stack, and the use of leak rate from 100 to 5 per cent per day. We have not yet selected filter filter mora d e san tortuous path withi e reactoth n r buildin ventee th o dt g types, locations, and other means to address the options being considered.

FIGUR : E5 HELIU M RELIEF VALVE FILTER TRAIN DESIGN FIGUR: 6 E REACTOR BUILDING FILTER TRAIN DESIGN Ü1 OPTION OPTION As stated, we are designing so that the PAG goals can be met with margin, O) whic goos i h d engineering practic d alsean o provide r licensinfo s g uncertainties. We anticipate possible imposed conservatisms, although do wo not believe them to be necessary, to include an increase in the magnitude CUTUT watef o r ingress, increased overall margin r "defense-in-depthfo s d an " general prudence considerations, and possible reversion to more traditional, non-mechanistic view points regardin source th g e term •M.DHvc «n C definition. Overall, we believe we have an approach to the containment •CCI »UT that is robust, in keeping with a properly established source term, and should be acceptable to regulators and the public in general. . V Conclusions The source term and containment concepts for the MHTGR that have been identified in this paper are consistent with U.S. and international past HTGR reactor operations, more recent fuel development findingsd an , current desig d analysian n s studiesplannine ar e W o evaluat.t g e containment option d establisan s reactor ou h r building design ovee th r forthcoming year worr beins Ou .ki gfollowinge baseth n do : 1. mechanistia Developmen f o e us d c tan sourc HHTGRe eth terr s ,a fo m compared to use of an arbitrary source term. We believe that this is a superior approac achievo ht e reactor safet licensind yan g goals. 2. Containment concerns for the MHTGR are best addressed by including in the containment system a vented, low pressure reactor building that recognizes that the MHTGR source term has distinct prompt and delayed components. desige Th n. 3 options outlined engineeree hereib n ca n d wit o problen h m f feasibilityo , although researc developmend an h t activitiee ar s neede o establist d h marginoptimizo t d an se design eth . References AEC 1962 U.S. Atomic Energy Commission, "Calculatiof o n Distance Factor r Powe Tesfo sd ran t Reactors, Report TID-14844, 1962. S.6. Inamati, "MHTGR Radionuclide Sourcn i e Us Term r sfo A. J. Neylan, and Siting," Report GA-A-19674 (CONF-8906184-1), SiladyF. .A , 1989 San Diego, 1989. P. M. Williams, "Draft Preapplication Safety Evaluation Report T. L. King and for the Modular High Temperature Gas-Cooled Wilson. JN . , 1989 Reactor NUREG-1338, U.S. Nuclear Regulatory Commission, 1989. ACRS 1992 Advisory Committee on Reactor Safeguards, Subcommittee on Advanced Reactor Designs, FIGUR : 7 EREACTO R BUILDING LEAK RATE DESIGN OPTION Transcripts of Meeting on MHTGR Fuel, U.S. Nuclear Regulatory Commission, February 26, 1992. THE ULTIMATE SAFETHTR-MODULE TH F YO E DURING HYPOTHETICAL ACCIDENTS

G.H. LOHNERT Siemens AG, Bereich Energieerzeugung KWU, Bergisch-Gladbach, Germany

Abstract e HTR-ModulTh a powe s i er reactor with specially favourable safety features. This is evident by the fulfilment of the following claims: the maximum possible environmental damage is limited and can be quantified; the consequence l crediblal f t extremelo sbu e y hypothetical accidente ar s restricted to the plant site and have no relevant offsite effect; the maxi- mum offsite radiation doses are below the values prescribed for basic design accidents by article 28.3 of the German Radiation Protection Ordinance; in consequence there is no need to provide strategies for public shelterin r eveo g n evacuatio o nee n o prohibi t dd an ne consumptioth t f o n agricultural products. By the fulfilment of these claims the HTR-Module can e burdeneb b eo t regarde t d no wit s a dresiduaa h l risk.

The HTR-Module (see Fig. 1) is a power reactor with especially favourable safety features. This is evident because the following claims will be met.

Clai» I e uniquth o Duet e employmen f o inherent t safety propertiea f o s Uranium/Gas/Graphite system all accident transients of an HTR-MODULE are so 1 reactor pressure vessel 5 reactor cavity extremely slow that for all credible events simple accident management 2 steam generator pressure vessel 6 full protection shell afte a gracr2 day s sufficieni o est perio1 o f terminatt o td e th e 3 connecting pressure vessel 7 surface cooling system accidents without relevant release of radioactive material to the environment. 4 primary circuit circulator

The reactor building is always accessible for. any accident management since Fi: HTR-Module1 g : Cross-section throug e reactohth r building the radiation level in the reactor building is very low. e generallTh y used probabilistic metho o quantift d e residuath y l risf o k 1.) Long-term failure of decay heat removal ÖD nuclear reactor s replacei sa deterministi y b d c predictabilite th f o y maximum possible damage potential of the HTR-MODULE. e decaTh e ycor s th transportei heae f o t y heab d t conductio d heaan n t radiation to the 3 fold redundant cavity cooling system outside of the Thus, the maximum possible environmental damage of an HTR-MODULE is limited pressure vessel. and can be quantified; the HTR-MODULE is not burdened with a residual risk. Assumin e totath g e lcavit th los f yo s cooling system o design , n limitf o s any component will be exceeded in the first 15 hours. The design tempera- Claia II tur ee pressur limith f o t e vesse< 400° f o lC wil e reacheb l d , afteh 5 5 r then a depressurization should be initiated. Since the fuel element The maximum release of relevant fission products to the environment is in temperature will remain below 1600° e integritC th e fue th l f elemento y s i s the order of a few Curies: e.g. I131 < 1 Ci, Cs137 < 10 Ci. always given e supporTh . t stabilite pressurth f o y e vesse s alwayi l s assured. The maximum offsite dose of all credible accidents is below the maximum allowable doses for design basic accidents which are given in article 28.3 Thus, inherent control of long-terra failure of decay heat removal is always of the German Radiation Protection Ordinance. given. Fission product release to the environment is not larger than allowe r desigfo d n basic accidents. Thus, the maximum possible offsite radiation dose of an HTR-MODULE is < 5 rem whole body doses and ~ 15 rem thyroid dose of an infant. These values include j^-submersion, inhalation and ingestion pathes over 50 a! ) Reactivit2. y Insertion ClaiI aII The insertion of any conceivable reactivity is controlled by design: The consequences of all credible accidents of an HTR-MODULE are restricted to the plant site only and have virtually no relevant offsite effect. Withdrawal of all rods during power operation (4/ma5 2- x= There is no need to provide strategies for public sheltering or evacuation and there wil e neveb l a rsituatio n wher e consumptioth e f agriculturao n l The consequence f withdrawao s absorberl al f o l s during power operatios i n products must be prohibited. inherently controlle e negativth a vi de temperature coefficiente Th . The HTR-MODULE is able to meet these claims. That will be discussed and fission procuct e retaine ar e sfue th ln i delement s sinc e fueeth l element demonstrated for the bounding hypothetical accidents: temperature remain belor sfa w 1600°C.

- Failure of decay heat removal - Large water ingress - Reactivity insertion - Failure of circulator trip The core moderation ratio was set to NÇ/NU > 550. This reduces the - Large water ingress increas y watef reactivito an e o rt ingrese du y o valuet s s lower than - Large air ingress possible reactivity increases caused by absorber withdrawal. - Maximum possible reactivity insertion: necessary time to transform 3300 kg of steam into watergas is in the order of about 10 h. That means, that about 53 kg of hydrogen can be released The simultaneous discharge of all borated spheres (KLAK) at cold reactor inte reactoth o r buildinn igniteca e d ignitioTh an .g f thio n s amounf o t condition would causeAf = 8 % within a minimum discharge time t = 37 s. hydrogen yield pressura s e increas reactoe th n i er buildin 7 bar0. .f o g

The maximum fuel temperature woul t exceeno d d 1300°C (T]im i= 1600°Ct ) and the temperature difference between coated particle and graphite Even the ignition of watergas produceable after hypothetical water ingress matrix would not exceed 370 K (A limit** ° *)• inte corth o e wilt resulno l majon i t r damages insid e reactoth e r building. T 80

Thus, inherent control of all reactivity accidents is valid without release e maximuTh m releas f relevano e t fission products woul ^J e aboui b dC 1 t f fissioo n product environmente th o st . und 3 Ci Cs^7. This release would mainly be caused by wash off from the primary circuit surfaces. 3.) Failure of circulator trip

Ia scraf n signal doe t trie circulatono sth p e steath r m generator dries ) Larg ingres5. r eai s out in a few minutes. The temperature of the steam generator will increase by 100 K to 700°C. The temperature of the pressure vessel will increase The largest still credible accident is the break of the connecting pressure from 250° 410°o Ct approximateln Ci »imites0 y1 . vesselopen a nr .reactoFo r buildin n opea nd reactoan g r cavity this yields a natural draught of initially 0.3 kg/s of air through the core. All e failurTh f circulatoo e r trip doea consequentiat lea o no st d l damagf o e ingressing oxygen will completely react with graphite, because th e components. temperatures of the fuel elements and bottom reflector are above the ignition temperature e regioTh . f corrosioo n n migrates froe bottoth m m ) Larg4. e water ingress reflector towards the core. Because the maximal fuel element temperature is always < 1600"C, the integrity of the fuel elements is always assured. But Due to the staggered side by side arrangement of reactor core and steam caused by an inhomogeneous corrosion of the fuel elements, it has to be generator a direct ingress of water into the core after tube rupture is expected that the first coated particle will be exposed after approximately strongly impeded. Due to temperature and pressure conditions of the primary 1.5 days. Here it is important to recognize, that the SiC-coating will not circuit ingressing water cannot evaporate necessare .Th y evaporation energy react with oxygen. has to be drawn from the structures of the steam generator. The heat content of the structures is sufficient to evaporate 3300 kg of water. Due to the low fission product release of the fuel elements the reactor building is always accessible. Accident management measures to interrupt The following combination of failures has the maximum possible damage startee b flo r thn eai wca d days 2 afteo t . r1 potential: a small leak and the failure of steam generator relief system failure anth d f circulatoo e rfailure th valv d f wateo ean r separator. Thus, large air ingress into the core can be terminated by means of A simple accident management strateg yo relie t woul e pressur: th fbe d f o e accident management in a time frame of several days. The release of e primarth y circuit a depressurizatio f e I achievedb . t no e th n , ca n relevant fissio w Curiesnfe e ordea productth f .n ro i s i s

CO References

Haque H., Lohnert G. H., "Auswirkungen und Abläufe bei massiven Lufteinbruchen in den HTR-Modul", Reaktortagung 1992 , "ThLohnerH. . eG t HTR-Module a powe, r reactor wito reactivitn h y excursion f severo s e consequences" S 1988AN . ;S ProceedingAN e th f o s Topical Meetin n "Safete nexo g th t f o generatioy n power reactors", Seattle 1988 , Lohner"TechnicaH. . G t l design feature d essentiaan s l safety related propertie e HTR-Module"th f o s , Nuclear Engineerin d Desigan g1 (199012 n ) 259-275 Lohnert G. H., "The consequences of water ingress into the primary circuit n HTR-Modula f o - Froe m design basis acciden o hypotheticat t l postulates", Nuclear Engineering and Design 134 (1992) 159-176 Reutle , Lohner, "AdvantageH. rH. . G t f goino s g modula n HTRs"i r , Nuclear Engineerin Desigd an g 8 (19817 n ) 139-136 CONCEPTS FOR DECAY HEAT REMOVAL IN MODERN OCRs (Session II)

Chairman

D.A. BILLING United State f Americso a DISTRIBUTION OF THE DECAY HEAT IN VARIOUS MODUL HTRs passive decay heat removal but there arc differences between the two reactor types AND INFLUENCE ON PEAK FUEL TEMPERATURES which mus consideree b t thein di r design hcs1 s e difTcrence e assigneb e n th ca so dt generation of the dccav powci, which depends in part upon the fuel cycle E TEUCHERT HEEN HAASA K ,VA K A , heat capacities associated with the graphite moderator Forschungszentrum Mich GmbH, heat transport mechanism differene th m s t reactors Juhch, Germany temperature distribution1; at the snrl of the décident n I transient R KASTEP N Ihc analysis of these features considers the following Tennessee University, United States of America 1 I or the stud\ of the effects of employing spherical versus prismatic fuel elements e samth e rcictor core volum d geometran e s treatei y d loadean d d with reference Abstract differeno elementtw e th t f reactoso r tvpcreactoe h T s r si7e correspond e 3 Sth O o st MW Modular High lempcratur ( oole s dGa e Reactor (MIIIGR) developey b d A unique featur f modulaeo r high temperature reactors (MOD FRsI I theis I i )I r benign General Atomics 'I/ I or the thermal conditions ofa steam cycle, both fuel element f oolanOf s responstos AccidenI a o et OCAI ( treactoc Ih ) r inherently becomeb ssu types are followed to their equilibrium cycle involving the burnup of about 80

critica decae th l y power partly reactoe heatth partld p su an r removes yi enve th io dt MWd/kg., studs hi I My outline effecte sth f fueo s l element design ronment via thermal conduction and radiation, while a\oidmg overheating of the fuel Production storage, and removal of the decay heat is studied for different r bot e PrismatiTo hth 2 e PebbluelcI c th Reactod d de Be an Reacto ) R r I (PBR)(P r , MODU conceptR l H I s having annular-core design thermad an s l power f 35o s 0 MW annular core design c utilizedar s , generatin MWe O averagIS gTh ( e core power

Based on use of Low Hnnched Uranium/ Thorium (I I U/lh( ) fuel cycles m Prismatic density was about 6 KW/litcr and the same in both reactors In addition, a second I uelcd Reactor fueU I 1 i lRs) cycle(P sd ,an Pebbln si Reactord eBe s (PBRs)l fo e th , design was considered for the PBR m which the design of the outer regions of the lowing has been determined MW0 reacto20 e tbase-MODUs th rwa n do I desigINTLRA1Oe th f no M company (1 )havin R Compariso

) Introduction Ihc comparison of the cases indicated above is given in terms of the distribution of the deca removals y it e termpea hean i d th d kf an t so an ,fue l temperaturs a s e A unde ( O I r ITic most unique feature of MODUI III R s is the passive response of the reactor to the wel s undea l r normal reactor operation l discusseal n I s d case peae sth k fuel tempcratuic loss of coolant, initiated cither by regular operation or bv in accident Pollowmg loss is definitely lower than I600°C as imposed from the demand of fission product re of coolant the reactor bc< omcs subcritical, and the power level reduces to the decay tention power Initiall e decath y y powe rinternale heat th reactoe n p th e u slated f e o san rth r erg s transferrei y d outside tin, icactor vesse a thermavi l l conductio d thermaan n l radi 2 Design and Operation ition Passive removal of the deciy heat takes place without d \nger of overheating the Ihc calculational simulation of the 1 oss of ( ool ml Accident is mule for a rcprcscn fuel elemente fissioth d n an sproduct s essentially rermin confine coatee th n di d particles tatwe statue reacto th e pebbl f th o s d r reacto o lif ebe I e equilibriuc r thiIh s i s m cycle reached after about S years of operation for the PI R this is the equilibrium icfucling ( onccpts of MODIjI II l R s have been studied for reactors with prismatic fuel blocks Öl cycl a tim t ea e a burnu correspondinf o d p en perioc Ih c ju^ i e d( o gt th t f beforo 2 1/ e CO and for reactois with spherical lue) elements Both tvpts aie cipablc of achieving safe cor refueleds ei ) Cn The calculational model includes all parts of the reactor that are relevant in determining PF e annulaRth r core regio subdivides ni d int annulao3 r fuel zones each containing fuel normal operating temperature«; and I (X A transient temperatures figure I shows the elements of two ages (The initial loading of the PfR core unices different fuel loadings reacto e referencth f r o layout/dimension d e an PBR s R thin I i se P s th e stud r th f Fo yo s in each radial /one) In the calculation of burnup and decay power, the fuel clement types having different age e followear s d separately, whil r thermafo e l evaluations thee ar y "homogenized" in radial regions In the axial direction the core is subdivided into 10 zones, representing the 10 fuel elements in a column Ihe fuel element temperatures are Void ™^.^ - followed individually (Initially differen 3 PFe s th , Rha t axial loading zone eacn si h radial «wvw«^Tv-un ^*v • Carbon bricks regioe PPRth f no , wit e firshth t axial zone correspondin fivp to e e axiath o lgt fuee cl l fc Cold helium chamber ments e seconth , d zone correspondin e nex th axia3 t o gt l fuel elements e thirth dd an , 1 zone correspondin e las axiath 2 t o gt l fuel elements) -1 Control rods 1 . i Top reflector Figure 1 also applies to the reference PBR considered here, except for the introduction - 0 of a "void" directly above the fueled region The PBR fueled region is subdivided into 4 - - !- Reflecto" "I r axiaradia9 d l an lzones , witvoie hon d region located directly abov core belod eth ean w " the top reflector The pebble fuel is recycled 10 times through the core before removal V V „ 6 - Prismatic fuel elements to reduce the axial power peaking, corresponding to the standard MEDUL (MEhrfach-DUrchLauf) fuel shuffle simulation in PBRs 1" " - Core barrel « - The calculationa lVSOe tooth s i lP reactor code system which performs reactor physics and thermal-flow analyse reactoe th r sfo r from startu equilibriuo pt involvee Th / m/3 d i - - Pressure vessel spectrum codes are GAM and THERMOS, neutron diffusion is followed by CITA PION, burnup and shuffling by FFVFR, and thermal performance by TIIFRMIX f1 500 - t' Surface cooling system The VSOP heterogeneity calculation of neutron physics allows for both spherical and Concrete reactor cell prismatic fuel elements, but the thermal calculation is not capable of temperature cell 5 calculation for the compacts of the prismatic elements 11ère, only the homogeneous Graphite column power/temperature distribution ove fuea r l elemen determineds wa t , resultin cala n gi - P culated peak fuel temperature during normal operation that is lower than would actually exist During LOCAs, however e modeth , s sufficieni l r determininfo t e peath g k fuel i temperature because the local temperature gradients within the blocks for that case are smal comparisoA l f resultno s obtaine r PBRdfo s treating "heterogeneous" powes di r Hehum coolant tnbutio d "homogeneousan n " distributio e sphericath m n l fuel showe e followindth g r Bottom reflector (1) In the homogeneous treatment the core reactivity was too high by 0 35%, due to the t underestimat f fueo e l temperature Maximum fue99°y b l Cw temperaturlo o to s ewa 1000 - t - _ _ S » Hot helium chamber (2) Durin ga LOCA , however maximue th , m fue only b l ytemperaturw 5°Clo o ,to s ewa g and resulted because the LOCA starts at a "reduced" fuel temperature relative to what L= J3 £3&.Éà it should havundeR PP e ra bee norman I n l reactor conditions peae th , k fuel temper \— ature could be substantially higher than inferred by the above PBR results along with the results from "homogeneous" calculation , sincRs I e P temperature r fo s a "fresh m s " (beginnin f lifego ) fuel elemen e highear t r "averagen thaa n ni " fuel clement

100 300 500 cm Basic to correctly estimating the fuel temperatures during a 1 (X A are the functions of decay powe d thermaan r l conductivity Decay powe Germae bases i rth n do n Industrial Norm DIN 25485 /4/ The evaluation is individually made for every fuel batch of the Figure I Reactor layout/dimensions of the PI R and of tlie reference PBR, for both reactor, and it includes explicite evaluation of its proceeding power history In order to S( and d I application1; except for the PBR there is a void" region between include uncertainties the calculitional model assumes Wo "overpower" prior to the the core and the top reflector (Solid curves Areas for nuclear calculation. depressunzation and a 2 a standird deviation for the cleciy power ol the fission pro Dotted curve-; Additiona le thcima putth r iluationv sfo t, l ) duct"; I he contributions of the actmides and activation products are included by 1 Decay Power Storag Removad ean l Durin gLOCa A conserv Uively derived bounding function«: is N suggesteDI e th n di In a I (X A the caiculational model assumes immediate dcprcssun/ation A I OC A re I he core thermal conductiv ty ) is a function of tempcnture and Past neutron flucncc suits in a core temperature rise which decreases reactor criticality and the reactor power e valueh I f Jo s used hci r I'BR fo cc bisc ar s n mcisiircrncnto d 1! graphitA" f o s y b e reduces to the lc\cl ofthc decay power within 1 minutes Subsequent!) the fuel clement'; Binkclc /V Radiative heat transmission between the spheres is based on the derivations slowly heat up c g the average temperature rises h) I 54°C during the first hour of 7chncr Schlundcr /(>/ and Robold HI I igurc 2 shows the values of/i used in this study for PBRs and also for PI Rs Over the first hours of a I ()( A a redistribution of the temperature field tikes place in e cor th s showa e igurI n n i locatioe ch I 1e maximu th f no m fuel temperature gradu ills A (NUKEM/A3-3) shifts fro botto e e middlmth th o mt e Once strongci temperature gradients o built p u d W/cm'K wird« the reflectors the heal is transmitted to the inner graphite column and to the re 05- (lectors Aller abou hour0 2 t e hea th sreactoe t th flo o t wr vessel stait o becomt s g esi nificant increasing the vessel temperature which facilitates heat release via thermal V 0,0~* 1fjz'(E > 0 1 MeV) 0 4- radiatio Reactoe th o nt r Cavit y( oolm g ) locate SysteS ( ( dm (R externa reactoe th o t l r vessel

0 3- figur show4 e time th s e variatio e decath f yo n heat sourc e distributioth e f decao n y heat m the reactor system, and of fuel and pressure vessel temperatures during a I OCA 0 2- casR I igurM eThI (nex5 eC S result te sectionth sr givefo e ) nar show s similar mfor matio summarn ni y fashion compared an , valueR sPB sR witF P h e thosth r efo 6.09 0 1 - middle Inth e drawin f Figur gstorage o th , e4 e ofhea coree subsequenth e th ,m t t storage e release vessel th e reflectorth i d th radiativny n b i ean , d an s e transfe e givea ar rs a n function of time for the P\ R l he heat being stored in reactor components reaches a 500 1000 1500 2000 •c maximum of 86 MWh after about 4 days I hat amount of heat corresponds to the en ergy produce t reactoda r full power ove minute5 I r s \—effective W/cm»K In course of the transient the fuel temperature reaches its maximum of H44°C at about hour5 7 s afte beginnine th r g ofthc acciden t thaA t t tim decae eth y powc reduces i i o dt 0 5- 0 43% of the nominal reactor power Subsequently due to the decreasing decay power IrrodiQted^Cropbife Moderator Blocks with increasing tim fuee eth l temperature gradually decreases Howeve temperature th r e 0 4- e pressurth f o e vessel still slightly rises accordin e redistributioth o gt e heath tf o nove r the whole reactor 0 3- (rodiaf)

AGL-IE 1-24 02- 4 Impac Fuee th lf o tElemen t Design

- 1 0 Ihc icacto S ModulaU r e desigth f r o nHig h lempcraturc (»a s( oolc d Reactor I (MlIfCiR)/! s beeha /n applie r comparinfo d e performancgth f prismatio e c fuel blocks and spherical fuel elements I able I compares key design chaiactcristics for the Steam 500 1000 1500 2000 ( vclc Pebble Bed Reactor (S( PBR) and the Steam ( yclc Prismatic I ucl Reactor ) (AlsR I P o ( e showdesig ablI (S ar I n ei n n characteristic c alternatIh f o s e design Figur e2 Values use i cordfo e therma studieR I P sld conductivitie(than eR PB e th n si R designatePB ( S ( AnnuladS designR r MOOUPB d e referencsh an I hav R ) I eI P e figurp to e givefue R valuee l th sPU initn f so x mitcna functioa s a l f terno n mad e oftheus c same reactor configuration indicate c equilibriuI igur Ih n i d I c m cycles pcratuic uxl Pi U neutron flucncc while the lower figinc gives the effective of both reactors have been analy/ed in this study Significant differences ippcar between thcmn e thermath l d coiR conductivitm l cPB conductivit e th f o y f o y the two types in both physics pciformmcc md in transient icsponsc to dcpicssuri/ mon U) 01 griphit) c Rs block I P n si iccidcnts UHTGR Optrohon limptroturi

t ml «i«m 2 Crophii« column

* Corbon brick* 5 Coti htlium chgmb«r 6 Hot h»)ium chomb«r 7 Cor« borr«! S Prtsiur« vttt»! 9 Void 10 Air n Surioc» cooling tytltm

UHTGR LOC* 2 h

MHfGR LOC* 12

MHTGR LÛCA 7S

gute 1 Icmpcratiiic distribution«; in the S( I'I R uiulcr normal opcntopcntini g condi lions ami tlunng a 1

56 • 3

I'lVUKiMMMt 10 50 100 houri Prismatic-SC Decay Power (relative to full power)

Integral decay UWh power

Kelt released from reactor vessel

0 hour10 s O S 10 Pnsmalic-SC Whereabout e decath f o ys powe t losa rf coolano s t

1500

fuel •i elements 1000

500 pressure 4 vessel

hour0 10 1s0 90 Pnsmatic-SC Temperature transients at loss of coolant

Figur : e4 Time v:\rifitioii ofthe decay-heat, source e distributioth , n ofdccaye heath n i t rcaclor system i pressure-vessef fueo am i l am , l temperature SC-J'Fe th r sCo R during a l.CK'A

57 Ol Iaht e! Desig haractcnshc-nC Refrrcncr fo s c Designs ANernatn ,a tni r DesigR fo l PB i n Ï abl U e Keactor Physic Othed nn s ( rhnracUnstic f Iso * quilibrmm ( yclc r Kcfcrcncsfo c r DesignTo d nn s 00 an Alternate I'BR Dcsipn Inde\ , r Normnl Operafirtf; Conilitmns

( ore MIIIGR Annular MODUI ( ore MllldR Annular MODUI 1 ucl element prismatic sphere sphere 1 ucl clcinctil prismilic sphere sphere 1 heriml power M W 150 110 (RO< -.!()()

1 m Volum f activeo e core S9R SI 7 A\g fcul enrichment Nfiw'NllM vvi "„ 8 1 1 R l>i 767 Diameter of gr iplulc column cm I6S 14(1 [ no! residence tune (full power) d SS7 SI7 72« Ihickncss height o[ iniuii^r core cm f>64 ( onversion ratio 0 4SI 0401 0444 f heliuHeao p u mt (Mcim c>clc) *( 271) • WO 2SO0 -70 » Power peakin f fuego l 28 -If. 2 S 2 r> r ba Pressur f heliueo m 60 70 Neutron leakage losses "n 1 2 1 - ID1 17ft 14 7

t 1ue UO2 + lhO2 LO2 10; F issilc inventory kg/dW|h R27 - SIR 277 144

1 nrichmcnt of uranium °n 20 8 18 7f>7 U^O^ requirement kg GW

( oated particles IRIS«) 1 RISC) Mdx fuel temperature X (799)' RS| (752)" S7S Diamete f kernelo r * ftm 500 500 outlee li t temperature ma( x" 'avg 745 (."O 74 0' 69 0 787 / 700 SS1 Carbon / heavy metal in fuel elements Nr/N,,M 457 5S3 lie pressure losses ovtr the core h ir n 23 4 S 1 073

I42S [abl gi\cI I ereactoe th s r physic d othesan r characteristic f equilibriuo s m cycle r reffo s x fueMa l tem t dcprcssiiri7ahoa p n "C 1144 1569 crence designs, and (or the SC Annular MODUI design, under normal operating con Aiswnplion of homogeneous heal ^urcc tn Ihr fuel clonicnls dition e reactoh I s r physic o reactors tw feature e e predominantlth ar s f e o s th o t e ydu

different fuee differenlth choiceo t d t carbon/heavsan y metal ratio ( /N|s(N )M) Differ ratie encc th f N o n Ci s /N)(M cause difference e neutroth n i s n leakage lossese th m , fissile inventory, and in the uranium ore requirement Purthcr, the pressure drop across Table III Decay Power, 60 Hours After Shutdown for Reference the referenc corR s mucei PB e h greater tha coreR n I acrosP , e howeverR th s PB e th , core pressure drop can be decreased by modifying the design as illustrated by the core Tuet dément Prismatic (1 O( ) Spherical pressure drop given in lable II for the alternate Annular MODUI design Decay power relativ fulo el l powe% r 0471 0 3R2 In the hcatup following the loss of coolant two baste differences between the two cases lïreakdown 1 ission products % 03S9 0314 arc observed Aclmidcs % 0092 0050

f or the spherical fuel elements the decay power is lower Other captures % 0020 0018 e prismatith r o 1 c fuel clement e maximu1th ; m tcmpcratui e fue lowes th i l f co r Inlegral dccaj power (Ml h) MWh 137 9 1 I able 111 gives the decay power in reference S( designs, (SO hours aftei reactor shut down As shown the decay power is higher in the PI R than in the PBR At the same tune ablI , I showI e s tha e peath t kfollowin A fue ( lO I temperatur a g s highei e th n i r e higheh 1 r R deca resultR PB I >P powese froth o differen n mi tw r t e icasonh I ) (I s I lie higher maximum fuel temper Unie m the PBR also results Horn two diffcicnt inventor f fissioo y n product s highe i se highe th o re I'Bl r tha th e heav Rn du ni y metal reasons (I) I he heal < apautv of the PBR core was lower than lor the I'I R cote because loading per unit volume and to the assumed occurrence of the I ()( A at the end of a of the higher coolant fraction tn the PBR core As t consequence, the hull up of the Inirnu e choicp th cycl ) f thoriuo e(2 e s fcitilma c material tcsult a significan n i s t mven I'BR proceeds fnstoi in spite of ihc lower dc'.ts pnucr is illustrated in 1 igmc S (2) 1 he tory of i1( Pa with a half hd of 27 d I he decay of2" Pa considerably contributes to the effectue thcrmil conductivit\ J oftlic I'BR was lower than the \

cludee Actrth n appliei de PBR th . o t d Considering suc a hmechanis r transferrinmfo g -———— Prismatic system heat across the gaps separating PFR fuel elements would tend to increase PFR fuel — — — Pebble bed temperatures t tha t bu include, t no effec s wa dt here.

For both the PI R and the PBR, the maximum fuel temperature following a I.OCA is reached about 60 h after beginning of the transient. The heat being stored in fuel ele- reactnn menti d san r component reaches sha maximus dit m about that time d furthean , r heat produce moro n s di e tha heae th n t released fro vessele mth s showa , Figurn ni . e5

Figur t givef e e radiath s l temperature distribution e referencth n i s e SC-PBe th d Ran 10 SO SC-PFR. at the core midplanc and 60 hours after a I,OCA. Also shown arc the thermal SC-KHTCR Decay Power (relativ o fult e l power) conductivities associated with the various regions, with the x values for the I'FR. and the U*h e recognizeb n ca t I PBRe valueo dth . thatr fo s : Integral decay power e heath t- e highefluth x o t througr e decay-hea du reflectore hth R PF e highes ti sth r fo r production 175 - temperature e loweth -th o rt thermae du e R gradien lcore PB th e highes ei n th i t r fo r UO - Heat released conductivity from reactor v«aael - the gap between the core barrel and the pressure vessel is a barrier for heat transport 105 - toward vessele sth .

Total heat stored 70 - 5. Impact of PBR Design

35 - An alternate SC-PBR design has recently been studied /8/. termed "Annular MODUL" (uel elements concept of 350 MWf The design of the reactor is made in analogy to the 200 hour0 10 s 0 5 10 MW(-MODU e INTHRATOM/SIEMENS-groupth f Lo - s givedi A Tabl.n e i n th , I e SC-MHTCR Whereabout e decath f yo s powe t losa r l coolano s t ametee pressurth , f whico rcm e s équivalan i 4 hvesse 66 e diametes i th le th o t tf o r German Boiling Water Reactor of Krilmmel (versus 636 cm in the reference PBR). The desig nreflectorse datth f o a ,controe lubeth r sfo l system f heliuo e mus , upflow, carbon bricks, core barrel and vessel, and the dimensions of the two gaps fully corresponds with

1500 0 MW20 (e -MODULth e crosTh s. sectioe annulath f o n r cors beeha e n selecteo t d r.'.yii.""" -r-""-"V"—..ftr..»r..tii.m achiev a relativele y modest pressure loss acros coree th s , therefor e volume th eth f o e active core is about 40% larger than in the reference SC-PBR design considered above. 1000 The Annular-MODUL-350 concept utilizes the standard LEU pebble element with a heavy metal loading of 7 g/sphcic and with a target hurnup of 80 MWd/kg. Differences

500 in results obtained betweee Annular-MODUL-35th d e referencnth an R PB e 0 concept reflect the different designs of the two reactors.

As shown in 'fable 1. the core heat capacity of the Annular MODUL reactor is about 10 50 100 hours large% large40 e th r thao rt correference ne th thaedu f volumeR o t PU e . Consequently, SC-UHTCR Temperature transient t losa s f coolano s t the maximum fuel temperature for a LOCA might be expected to be lower than in the reference PBR. However compariny b , maximue th g , Figure7 d man fues6 l temperature Figure 5: Time variation of the decay-heat source, the distribution of decay heat in tlic e Annulaith n r MODUL reacto s M4"i r C highe c rreferenc thaih n i n e PBR. This sur- prising results occurs largely becaus desige th f eo n difference outee th n i rs regione th f o s O) reactor system f d fuepressure-vesseo an ld an , l temperature SC-PBe th r fo s R CO compared with the SC-PI'R during n I OCA reactor surface pressure th th t f A o e s e vesse e temperatureth l abouc ar ssame r th t fo e 'c , O) Corbon S1..I Ste.l brkki - o I Core = Cort j> Cop CropKil* Grophil« Cop t« crtlc Graphit» - Air - Air == 1». 1600-

Ptbbl« b«d uoo- 1 1200-1 1 I 1000 n S '">. ^. \ W/c m/C W/cm/C SOO-j \ '^ 600- 1t -0 6 -0 6 00 | o 06 -0 4 «°-lo ooo •» -0 4 a»

200-1 °°'°ooo000o EC -0 2 -0 2 Soc Î 200 500 em 100 200 500 400 500 cm

figure 7 Radial temperature distribution n alternatdesiga R n i sPU n ( eS ond that ofusing creased different tvpc f fueo s l element wely l e e (hioptimun h i lth tt I suseR ma sPU differen t core heigh d differenan t t core annulus thickness relativ o valuet e s n optimuusea n i d m Most remarkable was the variation in the thermal performance of the diffcicnt cases as desigR I P n given in I able V I hat table gives the thermal conditions as «ell as some other values assocntcd with the various S( and G I cases during normal operation tnd during a 6 Characteristics of PBRs and PFRs as Hie Reactor Ouflet Temperature is I Of A 1 list of all undei normal operation the helium pressure diop icross the PUR Increased core is relatively high showing that the shape of the core c mty is not optimil for a operatine th r pebblo d g ebe prcssui c shoul e inucascb d d (core pressure drop vanem s vcrscl) «ilh the si|imc of th* prcssuic) As illusli itcd in the Annul ir MODI I concept In gas turbine c>clc (G I ) apphcitions of IIIRs ihc tempo iturc of the helium circuit 1 mus e selecteb t d considerably highc e itcit th thi r mfo n cycl) Proces ( e(S s heat s IIR I above incrcTsmg the innulir thickness of the fuel icgion sigmdcintlv clem iscs UK coic would ilso have \ciy high rciclor outlet tcmpci Unies (c g ~ 9W( ) In this stud\ the pressuie diop T able IV Physics < hör» it rustics of thr Various S< and (. I < ascs f«r Reference Designs I abl alsV e o indicate«; thae peath t k outlet helium temperature icl.itn e averagth o ct e outlet temperatur s slightli e \ s ie sloivcPBRhi th I n i rs R rclati\ I P e o thost cth n i e Purl element prismadc sphere radiae th o t l componene du t oTthc helium flow throng) pebble th ) d ebt SC 1 l (, (,12 S( C.I 1 C,I2 t Mcitiip or ilic helium "( 270-690 515-850 592-" >5() 270-690 592-950 Another remarkable feature associated wit n increasa h ic

Avg fcctl enrichmen N ^t *^|IKJ wt "* 8 11 8 11 8 11 8 18 8 52 8 59 s turbinga e th e r case o abous i I peiccnt 5 si ) nominae Syste0 S t th C f C o t m(R l reactor po«cr while it is only about 0 14"« Tor the S( cases I he 0 S% value coircsponds to l residenc1uc e lime (full power) C) 2X-I 6 27-16 2 (<-6 1 25 23 21 I igurc 8 gives the time variation of the distribution ol dccav heat in the reactor sjstem, Neutron Icikigc losses % in 1 — 12 l inn-.ii s 99-118 176 172 1 7 1 f d fuepressuro an ld e an th e , vesseR I P l ( temperatureS e th r fo A s durinK ( I a g 827-51« 842—547 277 299 1 issilc inventorv kg C'W[h 8Î8 >S19 292 e C512-Pth I igurd an R Igivec9 , R I sP similaI I G r informatio e correspondinth r fo n g 267 .OU n lecjutrcmcnl kg/(>Wd.h 101 311 114 260 264 PBRs I or both the PI Rs and the PBRs, the peak temperature behaviors ofthc fuel and of the pressure vessel during a I (K A arc similar In the SC cases, the decay heat first heats up the reactor, with increased heat rcicasc from the \csscl taking place with some delay I ull release of the decay power is reached after about ^ days I or the Ci f cases, the transient starts at higher fuel and vessel temperatures Here, the maximum amount of decay heat stored m the reactor is less than 1/2 ofthat in the SC case, because heat Tabl eV Therma Othe( aied r Referenc an l onditums fo r< C Varinue d th an f e«o C Designs«S , during Normal Operation andA durinOC ! ga is released fro e vesse ma th relativel t a l y high rate already fro e beginninmth e th f o g LOTA

Fuel element pnsmabc sphere I he above behavior is reflected in the maximum fuel temperature reached over the SC Gil C.I 2 S( GTI OF2 transient For the G 12-PI R case the maximum temperature is just about I80°0 higher 535-850 592 -«950 1 lealup of the helium "C 270->690 535-.8SO 592->950 270-690 e SC-Pth tha r caseFR fo n , althoug e averagth h e fuel temperatur e beginninth t ea s gi

Operation highe abouy b r t 280T further differene th , t way f distributioo s n ofth- re ce heath n i t actor cause considerable difference time th en i swhe maximue nth reachems i d Avg temperature of fuel X 559' 751' 819* 599 797 882 MTX temperature of fuel T 799' 915 • 1058 " 851 964 1065 In the PBR cases shown in F igurc 9, the tendencies observed aie about the same One Max 1 le nullet temperature °C 745 892 1009 740 886 991 basic difference is associated with the lower decay heat in the PUR as discussed in Sec-

I le pressure losses over the core bar 023 049 040 1 54 339 289 tio n4 Anothe lowee rth differenco rt heae du t s capaciti e core th ef yo becaus e th f eo higher coolant volum a consequence s A e core th , e hcatup temperatur highes e ei th n i r 1 (eat losses towards surface cooler M W 047 2 6 1 9 19 048 7 16 206 PBRs durin transienA firse X ( g th 1 t phase t th althoug f eo integrae hth l heat storen di the core is somewhat lower As a result, the peak fuel temperature during the I O( A 1 oss of C ooHnt tends to be higher in PBRs relative to that in PI Rs, however that difference decreases

Integral decah MW y he'l l h ove 0 6 r 117 117 137 119 119 119 as the reactor outlet temperature incicascs, as shown in I able V I or the peak vessel temperature valuetransienR A f s( P durinO s e I remai th te gth n somewhat higher than x decaMi y licit storeh dMW lolll 86 40 10 19 29 20 e PIÎth Re difference valueth t bu s s also decreas e icactoth s ea r outlet temperaturn i e store fuen di h l dcmenfMW s 211 18 17 17 15 14 creases Max lempcralnrc of fuel "( 1344 1507 I5f,| 1425 1546 1592

time of max fuel t'-mperilnrc h 75 56 52 48 Î9 36

x lemMi f pressuro p e C vessel 504 517 S48 474 515 53n

\cssc e x temth limh f ma lo p f eo 100 811 70 80 64 56

Assumption of homogeneous heat source in Ihc fuel clement* o> (O MWh Integral decay MWh power

• CT2-MCR (59 - 2950'C ) Integral decay power CT1-MC— — —— R (53 - 5850'C ) 175 - • CT1-WCR (53 - 850*C5 ) • SC-MHTCR (270 - 690'C)

hour0 10 s 0 5 10 10 50 ÎOÔ'hVurt——•

Prismatic »hereabouts of the decav power at loss of coolant Pebbl d Whereaboutbe e e decath f yo s powe t losa rf coolano s t

1500

1000

pressure vessel

10 50 10hour0 010 hours s 0 5 10

Prismatic Temperature transient t losa s f coolano s t Pebble bed Temperature transients at loss of coolant

Figur c \atiattoehm t Kdistributioh t ' f o n f decao n \e icncto th hc1a] A I i [ ShcnoA n v Modular High Icmpcratuie das ( oolcd Reactor (Mil IGR) Status . Intcrsoucty Zuying GAO, Shuyan HE, Min ZHANG I ncrg y( on\ersioi >I ngmccnn g( onfetenc c Philadelphia. l'A, Augus M 198 0 71 t Institute of Nuclear Energy Technology, I RentieI ] Ï [ r C,,\\ ohnerI t Tsinghua University, I Itc Modulai High temperature Rcictor Nuclear Icclinologv Vol (i2 fluK Beijmg, China I'>S^) pg 22 M) [ "scram accidents a , the core temperature raising, the HTR-10TM is shutdown automatically in 300 seconds owing to the negative reactivity feedback of the fuel Dopple effects. The peak temperature raising doesn't Undeexcee. e *C depressurizatio0 th r5 d n LOHS accident afterheae th , e removee passiv b th abls i t o a t e dvi e mechanisms like natural convection, heat conduction and radiation from the core,side reflector,reactor vessel to the reactor cavity cooling system (RCCS), the peak fuel temperature is only 921*C, far away from the safety limit of 1600*c. As the key way to remove the reactor afterheat under accidents, the RCCS is of great importance to the HTR-10TM and its high reliability should be reached. In addition, as a test reactor heatu,a p tes plannes i t simulato dt e afterheat removal of 200MW HTR-Module under the depressurization LOHS accident, Chimney stronger cooling abilit RCCf yo S shoul e designedb de th n I . preliminary design of HTR-10TM, RCCS is arranged in the concrete reactor cavity with completely passive heat transfer o type.Tw s of RCCS i.e. wate RCC r rconsiderede ai Sar RCC d San e simpl.Th e and reliable passive RCCS improves the safety and economy of the HTR-10TM activo .N e component emergencd san y diesel enginee ar s Air Cooler needed. In this paper briea , f descriptio RCCe th S f designo HTRn i n - t ChanneHo l Regulating Tank 10TM is presented. The heat removal characteristics of the both wate RCCr r ai RCCS d undeSan heatue rth p tesaccidentd tan e ar s 1 described. Theanalyzee THERMIyar e th y db X code. -Reactor Cavity.- •'.• \'

II, SYSTEM DESIGNS AND ANALYSIS MODELS cross-sectioe th Fig. s i l HTR-10TMe th f no cole .Th d helium enter reactoe th s r bottom throug outee hth ra coaxiasid f eo l tube, flows up from the coolant holes at the outside of the side reflector to the core top, then flows down through the reactor core to the helium mixing structure at the bottom reflector. The helium which absorb reactoe sth r heat the transportes i n e th o dt steam generator, which is located side by side with the reactor vessel. The reactor vessel is located at the centre of a sealed concrete reactor cavity, the RCCS is placed closing to the inner reactosurface th f o er cavity briee .th Tablfs i parameter1 e s of the HTR-10TM.

Table 1, Brief Parameters of HTR-10TM Reactor Power (MW) 10.0 Power Density (MW/M3) 2.0 8 1. Core Diamete ) (M r Number of Fuel Elements 27500 Pressure of Primary Circuit(MPa) 3.0 Inlet/Outlet Temperature of Helium(*C) 250/700 Thicknes0 1. Sidf so e Reflector(M) Diamete1 4. Reactof ro r Vessel(M) Diameter of Reactor Cavity(M) 7.4 Thickness of the Concrete Reactor Cavity(M) 1.5

Generally, afterheat removal of HTR-10TM would be provided by the heat transfer system via the steam generator during normal operation e RCC useTh s .i So remov t d e afterheat unde e LOHrth S accident and reactor power under the heatup test. It is also use o e cooreactot dth l r concrete cavity durin e normath g l Fig.l Cross-section of HTR-10TM Reactor and operation. Water RCCS Two alternative types of RCCS are designed:

1) Water RCCS As shown on Fig.l, two independent and parallel RCCSs of 205 90 125KW heat removal in each one are designed. The system consists -964 cavite th of y cooler insid reactoe eth r cavity r cooleai , n o r the chimney channel and associated tubes. The cavity cooler is (-2964 for tube5 8 madf so etha t arranged parallell innee th o ryt surface Air RCCS) e reactoth of e tubebottod rth e an ar f scavityp mo to e .Th circulao tw connectee th r y plenumb d o tube.Tw s connece th t risincoolert r ho ai plenum e e g .Th th tub o connectes st ei d from uppee th r plenum, acros reactoe sth r ai cavitinlee th f to o yt cooler. The cold downcoming tube is connected from the outlet of the air cooler, across the regulating water tank and reactor cavity to. the lower plenum of the cavity cooler. Finned air coole arranges ri channee th chimne a t a df lo heighn i y 20Mf to . e afterheatTh reactoe th f o "r cor s transferreei y botb d h naturar ai wate d lan r circulations watee .Th r enter e tubesth s of cavity cooler, absorbs the heat transferred from the reactor vessel, raises its temperature and reduces its density, then it r cooleai y e mean b rflowtubt th f ho o o p se t alonu s e th g buoyancy. After coole 'aiy db r cooler watee ,th r flows dowo t n the lower plenum of cavity cooler and a natural circulation of wate formeds ri cole flowr ,th dai s througcooler ai e rhth from + 197 intake th d exhaustean environmene th o st t throug chimneye hth , then a air natural circulation is formed. Table 2 is the brief parameters of the water RCCS. The freezing problems in air cooler is considered in the design. Regulated screens is placed at the air intake to control flows. +305 Fig.2 is the analysis model by using THERMIX code, outside the core, reflector and reactor vessel, a water circuit is added. 2} Air RCCS s showA Fig.3n o n , closin reactoe th o gt r cavity surface, three steel panels in thickness of 10mm are parallelly placed with gaps of 40mm and consist of two flow channels. Steel

Table 2, Brief Parameters of Water RCCS and Air RCCS

Water RCCS Air RCCS 1Cor: e t Channe5Ho : l Heat Removal Power(KW) 250 250 Sid: 2 e Reflector : Col6 d Channel Heat Transfer Aref o a 240 240 3: Reactor Vessel Cooler Ai : r7 (Water RCCS) Cavity Cooler (M2) 4: Reactor Cavity Chimney (Air RCCS) Height of Circuit(M) 14 34 Inlet/Outlet Temperature of the 52/83 30/83 Cavity Cooler (*C) Oi Flo we Circuit(KG/s th Rat f eo ) 1.8 6.2 Fig.2 THERMIX Calculating Grids Oi O) supportin gpanele th web weldee d sar san d together e syste.Th m O) includes two separate inlet/outlet structures, with several interconnected parallel flow blockages. Suc cavita h y cooles i r connecte environmene th o t d t directly through chimney e col.Th d Chimney air from the intake flows down along the outer channel of panels and then upward alon e innegth r channel after absorbin e heagth t froreactoe mth r vesselr flo enhanceai s i w e .Th chimney b d d an y exhausted to the environment. The heat insulating is considered to avoid the radiation between the hot and cold air channels. As comparing witwatee th h RCCr mors ai RCCSi S e e ,th simpl e and reliable freezino N . g problem r flowinexistai e sgTh too. through the cavity cooler is probably to be radioactived. However, it is estimated that the gas radioactivity is far smaller thaenvironmene nth t release limits. Tabl 2 containe s the brief parameters of this system. ,' Reactor Cavity \ ; '.', '. . Cold Channel Ill, REACTOR HEAT REMOVAL UNDER HEATUP TEST a Asmals l reactor e fue,th l peak temperatur f HTR-10To e M doesn't exceed 1000 "C under LOHS accident, however, as a comparison, the peak temperature of a 200MW HTR-Module could reach 1600 "C. In order to study the high temperature effects of fuel elements of the HTR-Module under accidents, it is planned to perform a heatup test in HTR-10TM. The test is to be carried undet ou r lower powe O.lMPd ran a helium pressur e reactoth n ei r without forced circulation. e d peaaveragth Fig.an k s i 4 e temperature profiles of the core versus reactor power under heatup test with water RCCS. The preliminary calculations shows Panels -J_' '. ' that the core average temperature is 1042 "C and fuel peak temperatur 141s i e8 *C durin e 250Kth g W heatup test negative .Th e

• Side Reflector

Reactor Vessel

Cavity Cooler

50 100 ISO 200 250 300

Fig.4 Temperature HTR-10Tf so M Reactor versus Core Power Fig.3 Cross-section of HTR-10TM Reactor and Air RCCS for Water RCCS reactivity core owinth e o temperaturgt e raisin s compensatei g d by control rods. Table 3 is the reactivity variation under the 250Kn c rtvr.WT heatuI« .«.n 4-1ipn tes4- .«.«4t -

Tabl Reactivit, 3 e y Variation unde e 250Krth W Heatup Test Fuel Temperature Raising -0.033 Reactivit Xenor fo y n Poison > 0.02 Control Rods Lifts 0.013

watee th rr RCCFo S heatup test temperaturee ,th d naturasan l circulation givee flo ar Fig.4—7n wi n e result.Th s indicate that the reactor heat of HTR-10TM under the heatup test is able removete environmene ob th o t d t passivel usiny b y g water RCCS. The peak temperature of the reactor vessel and side reflector 50 100 15O JOO 250 300 350 are 329*C and 890-C respectively at 250KW. The results of air RCCS are shown on Fig.8—11. It is found that under 250KW outler powerai e t, th temperatur lowes i e r than Fig.6 Reactor Temperature Profile Water sfo r RCCS 83*C masr ,ai snaturae floth f wo l circulatio KG/s2 6. e s Th .i n syste s abli mremovo et reactoe eth r heat durin e heatuth g p test. However, since the weak heat transfer ability of air comparing with water, the peak temperatures of core, side reflecto reactod ran r vessel wit RCCr highee hai Sar r than those with water RCCS under the same reactor power. It means that the power in the heatup test with air RCCS should be lower slightly than 250kw.

I

11n r i t i i 1l 11 m 11l 1 1n i n n 1i i n 1 m l n l 1111 n 11 l u n 1n H n i t i -t t 0 35 0 30 0 25 0 20 0 15 0 10 50 O 35 O 30 0 25 0 20 0 15 0 10 SO POWEHlkw)

0> Fig.5 Temperatures of HTR-IOTM Cavity Cooler versus Core Fig.7 Coolant Flow in Cavity Cooler versus Core Power Powe Water fo r r RCCS for Water RCCS O) CD

„liiiiiiitliiiiiiiiiliiiiiiiiiliiiiiiiiilniiiiiiilmiMmt __ 0 35 0 1030 0 0 25 150 0 20 0 40 0 35 0 O 20 30 250 O IS 0 10 O S 0 POWERikw)

Fig.8 Temperature HTR-IOTf so M Reactor versus Core Power Fig.10 Reactor Temperature ProfileRCCr Ai S r sfo RCCfor rAi S

8.O

75

70 8.5

6.0 s.s

50

«5

4 0

5 3

50 100 150 200 250 300 350 100 150 200 250 300 350

POVVEHIkwl

Fig.9 Temperature f Cavitso y Cooler versus Core Fig.11 Coolant Flow in Cavity Cooler versus Core Power Power for Air RCCS foRCCr rAi S IV, ANALYSIS OF AFTERHEAT REMOVAL UNDER ACCIDENTS The objects of afterheat removal analysis under accidents are to study whethe e reactorth r afterhea removee b abls ti o y et b d e usincompletelth g y passive mechanisms like natural circulation, heat conduction and radiation from the core to the environment—the ultimated heat sink. Cavity CooUr Outlet The analysis results of the following accidents are described: cavity Cooler Inlet — Depressurization LOHS — LOHS without scram — Depressurization LOHS with failure of RCCS 1) Depressurization LOHS The accident _is a severe hypothetical accident with very 5 10 15 5 3 «2O0 0 3 25 small possibility". Becaus primar e breaf eth o n i ky circuite th , TIMEIHRI helium coolant loses and depressurizes to O.lMPa, the helium blowe closeds ri reactoe ,th r the heatups ni . Fig.13 Temperatures of Cavity Cooler for Water RCCS under Depressurization LOHS Accident Wite wateth h r RCCS fuee ,th l peak, core average, side reflector and reactor vessel temperatures under the accident are show Fign no . sinc .HTR-10T12 e eth smala s Mi l test reactore ,th reactor thermal inertia and heat transfer surface versus the unit powe largee rar r than thos 200Mn ei W HTR-Module e fue.Th l peak temperatur exceedinw C fe thae * onla s th 3 ni ef f yo go steady state e inlet/outle.Th t temperature watef so n caviti r y coole showe rar Fig.13n no . Fig.1 Fig.1d 4an 5 give e resultsth s of natural circulation flow of water and heat removal in the water RCCS.

2.5

H I II II I I I I M I I I 11 11 I I UN It 111! UM l IM I I mull I llmmlMll II IHlllllHIII 10 15 20 25 30 35 5 10 15 20 25 30 35 «0 TIMEIHRI

O) Fig.12 Reactor Temperatures for Water RCCS under Depressurization Fig.14 Coolant Flow of Cavity Cooler for Water RCCS CO LOHS Accident under Depressurization LOHS Accident C»vity Cooler Outl*t

Heat R«J»OV*1

Cavity cooler Inl»t

0 « 5 3 0 3 5 2 0 2 5 1 0 1 5 15 20 25 30 35 40 TlMEIHni

Fig.15 Heat Removal of Cavity Cooler for Water RCCS Fig.17 Temperatures of Cavity Cooler for Air RCCS under under Depressurization LOHS Accident Depressurization LOHS Accident

l 111n illiintii mi 11n n 111111 1n i i 1n n m min1n n li i in i i iI 10 15 20 25 00 35 40

Fig.16 Reactor Temperature RCCr Ai Sr undefo s r Depressurization Fig.18 Coolant Flow of Cavity Cooler for Air RCCS LOHS Accident under Depressurization LOHS Accident Cor« Average

0 70 0 60 0 50 0 40 0 30 0 20 0 10 0 TIME! SECI

Fig.19 Heat Removal of Cavity Cooler for Air RCCS Fig.20 Temperature HTR-10Tf so M under LOHS under Depressurization LOHS Accident Without Scram Accident

With the air RCCS, the similar figures are shown on Fig.16— 19. Froe result th analysise m th expectes f si o t ,i d thae th t afterheat of HTR-10TM is able to be removed safely and reliably both using the water RCCS and the air RCCS under the depressurization LOHS accident. The temperatures of the all components in the reactor are far smaller than design limits. 2) LOHS without Scram Fig.2fuee th l s 0peaki , core average temperature Fig.2d san 1 reactoe arth e r powe HTR-10Tf ro M unde LOHe rth S without scram accident. After 300 seconds of the accident starting, the core average temperature increases to 648 *C, the reactor power decreases to 4% of the normal power due to the negative reactivity feedback of the fuel Dopple effect. The reactor is in sub-critical state. 3) Depressurization LOHS with Failure of RCCS passive th Evef i en RCC HTR-10Tf So f o s expecte i Me b o t d highly reliability, a very hypothretical accident 0 70 of 0 60 0 50 0 «0 0 30 O 20 100 depressurization LOHS with failure of RCCS is analyzed. Fig.22 are the fuel peak, core average, side reflector peak, reactor TIME! SEC) vessel peak and reactor cavity peak temperatures under the accident. It seems that the afterheat is able to be removed via concrete th e reactor environmene cavitl peath al o kd yt an o tto Fig.21 Reactor Power of HTR-10TM under LOHS temperature componente th f so withie sar e designth n limits. Without Scram Accident ro FLOW SCHEME DESIGD SAN N FEATUREF SO HTGR RESIDUAL HEAT REMOVAL SYSTEMS

V.F. GOLOVKO, A.I. KIRYUSCHIN, N.G. KUZAVKOV OKBM, Nizhny Novgorod, Russian Federation

Abstract HTGR intendee ar slocatee b o dt d near industrial plant residentiad san l areas, therefore they must possess enhanced safety characteristics. One of the most important safety aspects associated with reactor operation is reliability of residual heat removal system (RHRS). This paper describes design option r heafo s t removal system r smalfo s l pebbl d reactobe e r which provid a hige h degref eo reliability for residual heat removal.

HTGR e intendes ar e locateb o t d d near industrial plantd an s Fig.22 Temperature HTR-10Tf o s M under Depressurization LOHS with Failure of RCCS Accident residential areas, therefore they must be possessed of enhanced safety characteristics mose th t f importano e .On t safety aspects associated with reactor operation is reliability of residual REFERENCES heat removal syste systee Th RHR( m. m) S shoul designee b d o t d 1) Wang Dazhong, Xu Yuanhui, The Status of HTGR Research and meet the following requirements: Developmen Chinan ti , Nucl. Engins. Des. 132(1991) 95—97 - to ensure heat removal without exceeding the design limits 2) Gregor Tuylen Va . ,yJ etc., Examinin Inherene gth t Safetf o y PRISM, SAFR, AND THE MHTGR, Nuclear Technology, Vol.91, Aug. for fuel elements and reactor pressure vessel temperatures 1990, pp!85—214 during necessary time; realizo t - passiva e e principl heaf o e t removal without using active elements, operator actions, power consumption and auxiliary systems , - to keep availability if single failure would occur,! e RHRS must consist of some ( no less than two } independent parts

of equal efficiency, - to keep availability under external impacts such as flooding, earthquake r cras ai e d ,blastneares th an h t a s t industrial plants and other objects. RESIDUAL HEAT REMOVAL SYSTEMS PARAMETERS AND CHARACTERISTICS air 4 MWt38 5 'C.l -3 .N- t inlet Flow Air outlet Wall vessel Surface of Surfacf eo Total Height of Height schemes temperature of and surface water/water water/air surface vapour izat ion of water/air heat cooler heat heat of heat tank exhaust exchanger.' C temperature,' C exchanger, m1 exchanger' ,m exchangers' ,m arrangemenm t, stack, m a) 70 765 410 6300 7750 34 7 50 65 050/65 410 5000 6450 70 65 /60 230 7690 8960 70 b) 75 —— 5800 6840 34 7 40 70 < 350/70 —— 4330 5370 50 65 —— 3460 4500 70 d) 75 —— 3800 6840 40 40 70 O50/70 —— 4330 5370 50

65 —— 3460 4500 70 C) 70 /65 700 6300 8050 40 50 65 O50/65 700 5000 6750 70 65 /60 350 7690 9090 70 f) 70 <450/<150 —— —— 754 —— 80

eliminato t - e common cause failuree partl th al f so f so e transferreb Hea n tca atmosphere th o t d meany eb f so

system simultaneously intermediate circuit reactoe th f so r plant (including thosr fo e - mass and size characteristics of RHRS. volumes of relevant normal operation ) depending on the core type and its design rooms insid d outsidan e a containmene t shoul minimae b d l features and design limits for reactor pressure vessel and fuel Wate e ratmospher th pond e used b s ultimatan s a d n ca e e heat element temperatures. sink hear fo st remova e atmospherTh l preferabls i e e because The reactor of ultimate safety has to enable heat removing using water is associated with difficulties of reactor plant by radiation and natural convection from reactor vessel sitin d ecologicagan l problems r Thiai surfacse y HTGR' b th requiremen t o t me se e witb n h tca pebble bed core of small power (about 200 MWt). Nevertheless even for these reactors heat shoul e removeb d d outsid reactoe th e r compartment n connectioI . n with this, different flow schemef so residual heat removal from reacto whicr ai r he vesseth o t l flows as a result of natural convection have been considered. The schems are arranged depending on the extent of realazation of passive principles and influence of failures of single elements systee oth n m functioning e systemTh s wito intermediattw h e water circuits, natural convectio n the'circuii n f surfaco t e coole d forcean r d convection in the intermediate circuit are marked as a) and b). The difference of these schems is in arrangement of circuit independenc e independencTh e f intermediato e e circuie th f o t schems carriei ) t a ewit e helou d hf th heao p t exchanger inside vapourization tank. This excludes the loss of water out of surface cooler circuit under intermediate circuit depressurization. c) The cheme marked as c) has two independent intermediate water circuits with water natural convection. This enhances the exten f realazatioo t f passivo n e principles needes i t o i t d t bu , ft increase the height of water-air heat exchanger position. 1n1 H«.l HOf SCHEMES OF RESIDUAL HEAT REMOVAL SYSTEMS Independent circuits introduced by means of additional heat 1 ) witintermediatao Xb tw h e circuit*f theo s me ha on , natural convection of water, another one (transferring exchangers raise temperatur ee circui leveth n i f surfaclo t e - heat to the atmosphere) forced convection of water under normal conation d vaponrizalioan s f wateo n r cooler under equal other conditions Heat removal system marked II in the intermediate circuit under ijccident condition, <*-*• c) with two intermediate circuits am natural con- as d) has one circuit with natural convection of water. The cheme vection of water; d) with one Intermediate circuit and natural convectio f watero n ) heaf ; t removal direct is more simpl s comparea e dd make witan possiblt ) i sc h o havt e e r circuiai e ttth o with natural convection. 1- reactor; 2- surface cooler. 3 - vapourtzatioa tank; w leve lo f temperaturo l e circuith n i f surfaceo t e cooler. - exhaus4 t »tack - relie5 : ^cf valveU " water/wate- 6 , r heal exchanger; 7- valve. 8- pump; 9- expansion tank, However the cheme d) has more higher probability of water loss owing 10- water/air heal exchanger; 11- air coadecser to depressurïzation and excludes using vapourazation of water for PASSIVE DECA RESIDUAD YAN L HEAT REMOVAL IN THE MHTGR* heat removing. The cheme f) having direct heat transfer from reactor vessel D.A. DILLING, S.K. GHOSE, J.M. BERKOE Bechtel National, Inc., to the air is the most reliable and simple. San Francisco, California Some result f desigo s d calculatioan n n analysi e givear s n i n S.A. CASPERSSON the table. The results show that heat removing (2.8 kWt/m2) is ABB Combustion Engineering Nuclear Power, Windsor, Connecticut limitee siz th f reacto eo y b d r vessel surfac d possibilitean f o y G.C. BRAMBLETT placin surfacf o g e reactoe cooleth n o rr cavity wall. Thereforo et General Atomics, increase heat removin necessars i t i g raiso yt e reacto th e r vessel Diegon Sa , California temperature or to reduce the surface cooler temperature. United State Americf so a

Residua le cas heath f usin eo n i t g perlitic steer fo l Abstract reactor vessel (temperature permitted < 350 oC ) may be removed The modular high temperature gas-cooled The primary objectiv passive th f eo e RCCo t Ss i reactor (MHTGR) being developed in the maintain the structural integrity of the reactor through intermediate water circuits e temperatur.Th f surfaceo e United States has the capability to reject decay vesse completela n i l y passive manney rb and residual heat passivel y conductionb y , limiting the peak vessel temperature to 427'C cooler tubes shouldn't exceed 100 oC. natural convection radiatiod an , n froe mth (800"F) in an off-normal situation when the reacto ultimatre corth o et e heate sinth k- vessel is maintained at its operating pressure Residual heat in the case of using stainless steel for atmosphere - when normal heat rejection e vesseth If s depressunzedi l e limitinth , g system unavailablee sar methoe Th . d proposed temperatur (900T)C 2 48 s .i e These limits reactor removee vesseb y lma d directl circuir ai o t yt accordingly e U.Sfoth r . MHTGe s th Rreferrei s a o t d metallurg e basee th ar n vesselde o th d f yo an , Reactor Cavity Cooling System (RCCS) and they are consistent with fuel temperature uninsulaten a f makeo e us sd reactor vessel damage limit of 1,6WC (2,900'F) In the to the scheme f). The temperature of surface cooler tubes is not located in an enclosed cavity. The cavity walls off-normal situations, when no other more then 150 oC. are lined with cooling panels that allow outside metho heaf do t remova s availablei l heae th , t air to flow in and out in an open once-through s removei passive th y db e RCC seconA S d The cheme d) with intermediate circuit and the cheme f) natural circulation loop thin I .s arrangement, objective of the system is to protect the the heat removal takes place firsy b t concrete wall of the cavity. In order for u to transferring heat fro e reactomth re corth o et accomplis s objectiveit h s assuredlye th , are more preferable taking account of their simplicity and extent vessel by conduction, radiation and natural system has been designed with adequate convection. Second, heat is transferred from margin (excess capacity d appropriatan ) e of passive principle realazation. the uninsulated e coolinvesseth o t l g panels configurational features. The RCCS panels primarily by radiation from the hot vessel completely enclos e reactoreth thed an ,y form Thus, the following conclusion can be made on the basis of surfac relativele th o et y cold panel surfaced an , radiation cavities that trap essentially all the partl naturay b y l convectio e encloseth f o n d incident radiation thus enablin effective gth e cavity air. Third, heat is transported and emissivity of the panels to approach that of a the analyeis performed: the most rational chemes for HTGR of power rejected to the atmosphere by buoyancy-driven "black body e physicaTh " l attnbutee th f o s natural circulatio f outsidno througr eai e hth f simpleo syste e basee us m, ar n robusdo t reactof i e cbem th , e d) r ar vesse t MW mads i l0 f perliti o e20 c cooling panele systeTh sm functionl al t a s metallic parts and include multiple How times and does not rely on any active path redundand san t heat removal capacity steel and the cheme f).if reactor vessel is made of stainless steel componen operator o t r action e operatoTh . r cannot start, operate r terminato , e heath e t removal process

Ul »Work sponsored by the US DOE under contract No DE-AC03-89SF1788 I. INTRODUCTION

MHTGR Reference Design and Status e For uset th Vrai S t n di n desig active Th ne e objectiv Th e MHTGth f o e R prograe ih ms i core occupies an annular region surrounded developmen a safe f o ,t economic niuUar by inner, outer, top, and bottom graphite power e optioreferencTh n e corKipt 1 reflector elements Gravity-assisted control proposed to meet this objective i - the d mechanismro s locatee vesseth n lo d head standard MHTGR plant, consisting of finir 4O provid a emean o insert so grouptw t f o s MWt reactor modules, couple o t foud r control innee rodTh sr rods, which entee th r individual turbine generator sets producinga core near the interface between the fueled net station electricae MW l outpu0 69 f o t e inneregioth d r nan reflector e user ar , dfo The selected design features of the MHTGR cold shutdow e outeTh nr rod e user sar dfo have been arrive t a afted r evaluating power adjustment during operation A alternative concepts and conducting a series of reserve shutdown system (poison balls) trade studies beginning in early 1984 Major provides a diverse means of achieving cold features studied over this period have been shutdown MW0 20 • t1 Cor o t e0 siz20 e A vertical helical-coil steam generator is locatee steath mn di generator vessee Th l • Fuel type Prismatic bloc s v sphericak l motor-driven helium circulato s mountei r d pebbles vertically on top of the same vessel The • Vessel type Stee s prestressev l d concrete steam generator vessel is connected to the reactor vessel below the core by a coaxial • Turbine One turbine per one, two, or four cross-vessel During normal operation, the reactors Heat Transport System e primar(HTSth s i ) y s watev r • RCCrAi S mean f removino s g nuclear heat froe mth alss i primarcore t oI th e y mean f remoso v • Reactor building- Degre embedmenf eo t decag in y heat durin gnomaa l shutdown • Fission product transport- Alternative If the HTS is unavailable because of its containment maintenance needs or because it has failed, concepts the plant can be cooled down using the RCCS However, this requires several days and In August 1985, after completing trade studies subject e a vessetemperaturth s o t l e e majorit th above th n o f eyo features0 35 e th , excursio o obtaiT e desirenth n d plant MWt reactor module concept was selected for availabilit o t limie risth f d o tk an y further development This design uses investment losse MHTGth , s equippei R d prismatic fuel blocks stacked insid a estee l with a Shutdown Cooling System (SCS) The pressure vessel and a single heat transport loop heat exchangeS e circulatoSC th e d th an r f o r (Ref 1) The side-by-side vessel arrangement locatee bottoe ar e reactoth th t df a m o r vessel is show Figur1990n m nI a stud , 1 s e ywa (Fig 1) The SCS circulates helium only undertake o improvt n e estimateth e d cost withi e reactoth n r vesse t rejectI l s heaa o t t performanc e planth f teo (Re ) 2 Thif s study closed water loop whic turm h n rejects heat t MW 0 culminate45 e e choicth th f n o i ed to the atmosphere via an air-blast heat FIG 1 Reactor module reactor module describey dke hereie Th n exchanger Normally, the SCS will be the desigt additioe th s MW ni 0 n45 chang e th n ei second mean f rejectino s g nuclear decad yan of one nng of unfuelled center reflector blocks residual heat power To accomplish this, the program has e fouo th fachievee rb goalo t s i sd Nextn a , in the core defined four Goals which must be satisfied, as integrated systems engineering approacs i h Four reactor modules, each coupled to a Plant Level Requirements follows systematically applied to develop the functions, requirements, and specific design selections turbine generator set, are used in thi sMHTGE desigDO nS R tU oprogra e Th m has been • Goal 1 Maintain Safe Plant Operation e planMW t capacit0 achiev69 e yth e Each necessary to meet, m a balanced fashion, all of developed using a systems engineering • Goa . Maintail2 n Plant Protection the Top-Level Regulatory Criteria r TLRo , C reactor modul s i house ea vertica n i d l approach Rather than attemp o diriult t y cylindrical concrete building, most of which is • Goal 3 Maintain Control of Radionuclide (Ref 3), and user requirements (Ref 4) As an utiliz e largth e e bod f light-wateyo r reauor adjunc thio t t s processspecifif o t se ca , licensing embedded in the earth Each 450 MWt reactor design and regulatory experience ihe Release module, shown in Figure 1, consists of two bases has been derived (Fig 2) The licensing designers have sough consciouslo t t y analwe • Goa l4 Maintain Emergency Preparedness bases includ licensine th e g basis events (Re) 5 f vertical steel vessels - a reactor vessel and a e functionth e performeb o t s d an o thet d n steam generator vessel connecte a shor y b dt that demonstrate the compliance of the design select design features needed to perform those The approac desige MHTGe th th o f ht ns o Rha with the TLRC and the classification of the coaxial cross-vessel The reactor, reflector, core functions support structure, and core restraint devices are been applie a "top-down n i d " mannee Th r equipmen assuree tb than o responca t dt o t d installed in the reactor vessel The fuel and s witpowey A an h r plant design projecte th , process begins wit e quantificatiohth e th f o n e e eventmanneth th n i s r specifiee th n i d reflector samblocke th ee sizsshapar d ean e produco t overals i m eai lsaf d economicaan e l top-level criteria pertainin welw ho l eaco gt h TLRC II. MHTGR DECAY AND RESIDUAL HEAT REMOVAL REQUIREMENTS

MHTGe Th normallS RHT yt removeMW 0 45 s large when compare decae th o ydt heat genera UCENSmO BASIS fro reactoe mth r energe corTh e y supplieo dt tio absenc e coolingy nth an ratn I f o e th , e circulatoth s e flowini adderth s o t dga g MHTGR could absor decae bth y heat energy stream somd an , e surrounheae th loss i to t t d for many days befor e temperaturth e e tran PflMCVAL DESIGN CRITERIA ing buildin d e totaRCCSth an g lo s energ, y sients would become unacceptable delivere e steath mo t d generato s approxi r i UCENSMQ BASIS EVENTS mately 453 MWt Table 1 lists the heat transfer The HTS is capable of operating with low parameters for normal operation of the feedwater flow and reduced helium pressure (even including helium at atmospheric près EQUPMENT CLASS MHTGR sure) to remove the decay heat, along with the thermal energy stored in the reactor and ENOMEBWKl PRODUCT OTHER BASES Table 1 unavailables i s i S S HT SC vessele e th th f ,I s PLAN SIGHf To a ,er use removo dt decae eth y energ cood e yan th l MHTGR Normal Heat Transfer Parameters reactor down Figur showe3 schematisa - cre presentation of the HTS, SCS, and RCCS The Reactor Power Level t MW 0 45 SCS rejects heat to a closed water loop (the Circulator Power Level 3.55 MWt Shutdown Cooling Water System r SCWSo , ) which reject atmosphere heas th sit o t e Th e FIG 2 Top-down integrated approach Heat Los RCCo st S 1 1 MWt SCS is designed for two operating modes Heat loss to Steam Generator 009MWt During normal reactor operation, it is in a Compartment standby mode, wit hsmala l flow circulating in the SCWS by a jockey pump or natural cir Both the regulatory bodies (U 5 NRC) and the To comply with these requirements for Goals Helium Temperature, Core 288 -C culation When called upon to remove heat electric utility industry, through Gas-Cooled 1 and 2 functions, the use of inherent charac Inlet (550 T) fro circulatoS reactore mth SC e th , r induces Reactor Associates (GCRA), have established tenstics and high-quality fuel in the design forced circulatio f heliuno m ove e shutth r - top-leve requirementy l criterike e Th s aa e sar minimizes releas f radionuclideo e o thas s t Helium Temperature, Core 704 -C Outlet (1,300 T) down cooling heat exchanger e maith nd an , follows normal operational releases or any accidental pumps and fans in the SCWS are started The release primarf so y radioactivitd an w lo e yar Helium Flow Rate 2096Kg/s plane Th t• must meet 10CFR100 accident worker exposure minimizee sar d These tech- system is designed to be started at any point in dose limits thyroid(30wholm m 0Re Re e5 ,2 niques have proved to be effective m other Peak Fuel Temperature 1,420'C the transient following a loss of the HTS A bodysite eth t boundar)a conservativelr yfo y gas-cooled reactors, and have been demon (2388 T) gravity-operated, pressure assisted check evaluated events with frequencies greater circulatom ma e valvth r outlen i e t prevents strate measuriny db g release worked an s r Peak Temperature of 285'C than 10- reactor *pe r year exposures from operating plants helium from flowing backwards through the Reactor Vessel (545 T) steam generato d maian r n circulatoe Th r plane Th t • must avoi neee th ddisturo dt b An important aspect of the MHTGR is the the off-site public This mean e planth s t Feedwater Temperature 193 "G SCWS is designed for an ambient air tempera- approach taken with regard to Goals 3 and 4 (380 T) ture range from -43'C (-45T) to +43'C must meet Environmental Protection Agen- To achieve these goals with high assurance, (+110T) Table 2 indicates the SCS operating cy Protective Action Guideline r evacuafo s - the design of the plant is such that control of Steam Temperature 54 1'C conditions tion and sheltering (5 Rem thyroid, 1 Rem radionuclide releases is accomplished by (1,005'F) whole body) at the site boundary for retention of radionuclides within the fuel Feedwater Flow Rate bote ar hS 176unavailablHT 3e Kg/th d san eS SC e Ith f restrictively evaluated events with particles, with minimum reliance on active heat is removed by the RCCS The decision to frequencies greater tha 10-x reactor n5 ? pe r design features or operator actions Design Change in Secondary 2361 KJ/Kg provide the RCCS is based on the expected year and the expected frequency with features, including the low power density Coolant Enthalpy ordern I o t S SC d an S reliabilitHT e th f yo which all events could lead to doses annular core configuration, uninsulated reac meet plant design goal accidenr sfo t pre\n e requiring evacuation must be less than 7 r vesselto d passivan , e RCCS backup decay tion, accident performance availabilitd an , a v 5xlO- r planpe t year heat removal system are chosen to preclude Whe reactoe nth shus i r t down fissioe ,th n pro- third decay heat removal system was required • Worker dose mus e lesb t s than 10f %o evente th s which could challeng integre eth i ducts in the core continue to generate power by An additional functional requirement proo t , - 10CFR20 ty of the ceramic fuel coatings radioactive decarate t whiceTh a y h powes i r vide active cooling for the concrete structure generated is approximately one percent of the surroundin reactoe gth r vessel assignes ,wa o dt • The plant safety performance must be par y e RCCthif ke o t Th sa s desigSi n strategy normal power generation after about one hour the RCC Sselectes aftewa t i rorde n di meeo t r t insensitive to plant operator action or mis Its features have been developed in a loiver decae Th y heat generation curve fall verf of sy safety and availability goals action level recursion of the overall MHTGR dtrbign rapidly durin e firsgth t hour when mosf o t • The plant should rely as much as possible process Subsequent sections describe s e eth the power comes from short lived fission pro- e analyseTh s performe o determint d e th f i e n passiveo , inherent mean f achievino s g lection procesresultane th d san t desige th i no ducts d moran , e slowl s tima y e increases MHTGR meet s safetsit y goals require that acci required safety functions RCC morSm e detail because the short lived isotopes have decayed e Desigdentth m sn Basis Event frequency powee th d mostls i r an y from long liveo dis region (more frequent than 10-* per reactor • The plant must be economically competi topes which gene-rate decay energ lowea t ya r year) take credit only for components, struc live e totaleveTh l l heat capacit e corth e f yo hires, and systems which are "safety related" reactor internals, and vessel system is very Such components, structures d systeman , s -Ni mak reactoe eth r subcnhca controe Th l l rods, Configuration and Other Requirements reserve shutdown system, and a strong 00 Configuration and other requirements im negative temperature coefficient will each posed on the RCCS include the following INLIT/OUTI I-T independently stop the chain reaction How STRlXTlIRt ever decae th , y heat energy wil addee e b lth o dt •The configuration of the RCCS is dictated bv reactor, regardless of all other considerations the physical arrangemen reactoe th f o tr \ » The RCCS mus e ablb t o absort e b head an t sel and its supports The system must jllow remov t froi ee reacto mth r quickly enougo hs for the transmission of deadweight and --eis that the added decay heat does not result in mic structure loadth o st e unacceptable temperature e fuelth r , reactofo s r internals, or vessels RCCe Th S• must withstan credibly dan e près sure transient event The pressure transients The normal operating temperature, and which could load the system from inside the allowable temperature limits for various building might come from mfailura ma a f o e material MHTGe th n s i listee Rar Tabln di e3 steam e maximu linTh e m credible load from this source is less than 10 psig •The RCCS duct system must penetrate th e primary radiation shield for the MHTGR Table 3 The allowable dose limits for workers outside Material Temperature Constraints the building will result in RCCS design 540X. requirements, which wile specifie b l th o t c MAIN Normal choice of air cooling ( ONOtNSFR Maximum STEAM Operating •The designatio f "safety-relatedo n " means ÎENtRATOR OFAtRATOR Temperature Allowable (Average) Temperature thae RCCth t S mus e designeb t o resist d t external hazards, including seismic loads (0 3g Fuel 842'C 1,600'C SSE, 015g OBE). and potential tornado borne (1.548T) (2,900 T) missiles This designation is also the source Graphite 800-C 3,000'C e targeoth f t reliabilit e RCCe th Th f o yS (1 ,47010 (5,400 TO MHTGR risk assessment n e basea ar s n o d RCCS reliability requiremens sy s thai e t th t Core Barrel 288'C 482'C tern should faio morn l e than onc n 10i e* (550T) (900T) demands FIG 3 MHTGR heat removal systems Reactor Vessel •Th coole- au e d panels selecte RCCe th r Sdfo — Pressurized/ 243'C/NA 427'C/482'C can be viewed as part of the reactor building Depressunzed (470T/NA) (800T/900T) boundar allowable Th y e leakage acrose sth must traditionall designee yb qualified dan n i d Concrete Structure 43'C 176 7'C reactor building boundary (one volume per day) must includ y leakagan e e acrose th s Table 2 a rigorou demonstrato t y swa ehiga h degref eo (HOT) (350 'F) reliability and to resist internal and external RCCS panels Shutdown Cooling System Parameters hazard e designatioTh s f "safeto n y related" •The utility industry requirement for must exten l poweal o dt r suppl structurad yan l Fuel degradation begins to occur at 1,600'C functiobua s i t f temperaturno timd ean e minimal reliance on active safeguards and SCS elements which enable the performance of the avoidance of operator involvement is inter Standby Operating function In this context, the HTS and SCS preted as a requirement for passive opera Mode Mode have not been designated "safety-related", and tion, and an avoidance of any valves, fans, safety analyses have been performe demono dt - e heaTh t removal capabilit e RCCth f So y motors, or other active components Helium FloS SC w 065Kg/s 132Kg/s strate that the plant can meet its goals without mus relate e rate b t f heath eo o dt t transfe- be r Max Helium Inlet Temp 285 C 667 C credit for them The RCCS is a passive system twee materiale th n s insid reactoe eth r vessel It has been designated as "safety-related", and is and the vessel When the vessel is depres IV. RCCS DESIGN SELECTIONS Water Flow SCWS 73Kg/s 403 Kg/s assume e presenb o d functionat d an t l dunng surized the allowable vessel material tempera the full range of licensing basis events This ture rat s f higheth i heaeo e t t bu transfer r The requirements presented in Sections III Max Water Outlet Temp 1305 C 221 C and IV, viz, the heat removal and design re SCHE designation has an effect on the design require- from the core to the vessel is lower As a ments imposed on the RCCS resul e temperatureth t core more th ear n si e quirements, have been considered in arriving Water Pressure SCWS 48MPa 48 MPa desige ath t n selection e RCCth r Sfo s These severe for the depressunzed case When the design selection discussee sar e folloth n wdi Water Return Temp 1183'C " 60 C III. DESIGN REQUIREMENTE TH R SFO reacto s pressurizedi r , natural circulatiof o n ing sections SCHE REACTOR CAVITY COOLING SYSTEM e heliuth m transports heat frocore th mo t e vessee th l inducin mose gth t limiting temper Air Versus Water Peak Heat Transferred 036 MW• l t 9MW ature conditio systee th nm Therefore th e FueVessed an l l Temperature Limits most restrictive design requirement for the For passive remova f decao l d residuayan l l forceIal f d coolin e MHTGth f go losts Ri , there RCCS is the vessel temperature during près heato coolintw , g schemes have been * Natural convection mode of SCWS operation are several mechanisms whic designee har o dt surized cooldown events evaluated natural draft air cooling and forced water cooling wit a hpassiv e e modTh e The performanc f bot o e scheme th h s have RCCS Desig Componend nan t Arrangement naturae Th l draft air-cooling scheme consists been evaluated against various criteria mclud water-cooled scheme consists of two 10 f coolin0o t se g oa fpanel s mountee th n o d The RCCS design is shown in Figures 5 percent forced cooling loops which reject heat mg function d requirementsan s , operabihty, cavity wall surrounding the reactor \essel hcenseability cosd an ,t Bot schemee hth n sge through 9 It consists of three sets of compo- e atmospherth r blaso ai t a t vi ehea t two inlet/outlet (I/O) structures for inlet and nentstructuresO I/ s , ducting d coolinan , g exchangers (Fig 4) If active cooling is lost in erally mee e functionth t d requirementan s s outle f atmospherio t c coolin sea f o t gd airan , However r coolinai e th ,g schem selectes ei s da panels The I/O structures are located above both loops systee th , m revert a passiv o t s e concentric hot and cold ducting for transport grade on top of the reactor building These boiling mode wherein hea s rejectei t y b d ing air between the cooling panels and the the preferred optiofollowine th r nfo g reasons structures connect to the concentric cold/hot venting steam Adequate stored water is I/O structures This system always works m a • For the water cooling scheme, there is ducts. The ducts are routed through different provided such thapassive th t e cooling mode passive mode as the heat from the M.-.sel significant uncertainty and complexity asso- reactoe areath f so r buildin connecd ge an th o t t could laseighr fo t t day thin I s scheme th e reachin panele gth s r througdriveai e e sth th h ciated with two-phase flow in the boiling cooling panele coolinTh s g panels, which cooling water is circulated m vertical tubes system mod d wite transitioan eth h n betweee th n consist of cold downcomers and hot risers are arranged on the cavity wall surrounding the activ passivd ean e modes located inside the reactor cavity and completely reactor vessel • The air cooling scheme has fewer failure surroun reactoe dth r vessel modemors i d es an passiv e The inlet openings of the I/O structures supply • To provide the same level of reliability as the cold outsid e outleth d et an airopening, s air cooling scheme, the water cooling scheme facilitate r bacth ai e exhaus o t t kth e ho f o t needs to be provided with very complex atmosphere The concentric cold/hot ducts features As a result, it becomes undesirable transpor r betweeO ai I/ t e ho tth n cold an d structure e coolinth d t gan sho panele Th s ro««A«tATM —I/ Q due to increased complexity and cost •fêi-i-^LJh JL ——il—-

n , ~5

AID iMk.tr >-

CMCULATO*

MACTMVMML

IMUtATIO OVTUTOUCTI

COOLNM'MMU

CO FIG 4 Active water cooled RCCS with passive mode 5 Passiv G r cooleFI eai d RCCS concept poured Vertica t ducho l t sections woule db move heat fro reactoe mth r cavity proteco t , t GO ducting is surrounded by the cold ductin fres gi allo o et d an w thermar l he growtt e Th h O mal growth is accommodated by the flexibility lowered into final location afte e colth rd duct e concretth e structure from overheatingd an , which facilitates maintainin e buildinth g g provido t emeansa f decao y heat removar fo l structure temperature within the allowable of the riser tubes and flexing support plates walls are completed These flexing plates are located at three vanous off-normal events Hea s removei t d limit In the cooling panels, the cold air flows froe cavitth m y naturayb l circulatio f outno - e e bottocavitth th dowa f o vi t yo mn vertical position permio st t vertica radiad an l l V. SUMMARY AND CONCLUSION thermal motio tube th ef n o assembl y sid r througeai h cooling panels surrounding downcomers The air is then heated by the the reactor vessel vessel heat and rises through the nser panels t Plenum.t plenuho Ho rina e s mgi Th header The Reactor Cavity Cooling System (RCCSs )i due to buoyancy The hot air then exit cavite s th to f yo th thaep to t e collectath t r frose ai mth systeme th f o se comprisinon e Modulagth r The RCCS consists of two inlet/outlet atmosphere first througt ductinho d e hth gan hot nser tubes and directs it to the hot exhaust High Temperature Gas-Cooled Reactor 4x450 structures for inlet and outlet of outside air, a finally through the outlet openings of the I/O MWt Side-By-Side Steel Vessel Plant The f coolino t se g panels surroundin reactoe gth r e ductplenuplatm Th 8 s e s builm3/ i f o t vessel within the reactor cavity Tor receiving structures stiffened by transverse rigid frames It is pnncipal- functionre e o RCCt th e f o Sar s field-bolte nsee th r o tubdt e t assemblho e Th y RCCS Structural Support plenum connects to the hot exhaust ducts at the four corner reactoe th f o sr e cavityth d an , The structural support concepts for the RCCS differential movement between the maccoms i - components are briefly discussed in the modated by a flexible bellows at each comer following paragraphs Hot and Cold Ducts. The cold duct which Cold Doivrtcomers The downcomers surround integrat n duca ho s i te sth le parth f o t completely line the inside surface of the buildin mads i d unlinef eo gan d concrete Th e reactor cavit d maintaian y e concretth n e concentri d internallan c y route t ducho ds i t temperature below 655'C (150'F) by segmented with expansion joints to allow ther- circulating the inlet air against the back plate mal expansion. One support point on each seg- anchore cavite th o ydt concret /4-inc1 e Th he ment is anchored to the cold duct, with sliding thick back plate serves as a form for concrete, supports or guides at the other support point thus ensuring good heat transfer to the con In vertical runs, the fixed point is at the top of Crete surfac e remaindeTh ee down th f o r- o avoit n thdru ebucklin o deat e d du gloa d comer consist f 5/16—incso h thick cover plate compression and vertical 10-inch wide channea di l phragms spaced approximatel) m 4 (2 m c 0 y6 I/O Structure. structurO I/ e concreta Th s ei e centen diaphragme o Th r weldee e sar th o dt structur o accommodatt e e seismic, tornado, bac kcovee platth t r ebu plate boltee ar s o dt and missile loads. All concrete surfaces of the the diaphragms to permit inspection quiescent chambers whic - e subjechar im o t t pingemen t exhausprotectee ho ar y r b t ai t y db Bottom Cold Plenum. The bottom cold steel liner plate The liner is separated from plenum is a ring header at the bottom of the e concretth e surfac o permit e t circulatiof o n cavity that collect e colth s d downcomer ai r cold air The concrete surfaces in cold air and directs into the nser rubes The plenum section t lineno de sar rests on the cavity floor and consists of two assemblies Materials of Construction Installatiod ,an n Plan • Back plate with bottom plat suppord ean t AH the metallic parts of the system are con- bracket - the assembly is attached to the structed from ASTM Crade A carbon steel with cavity wall by means of embedments adequate allowanc r corrosiofo e e colTh dn platp To e • with front assemblplate th e- s yi metallic portions in contact with inlet air are welded to the hot riser tube modules galvanized to mitigate atmospheric corrosion Figur show0 e1 sschema e reactoth f eo r cavity These assemblie connectee sar eaco dt h other and RCCS installation sequenc e schemTh e e with bolts permitting dismantlin r majofo g r requires m-place installatio f assemblieno o sto repai r replacemeno r e bottoTh tm cold ple- large to ship to the site Field assembly of num supports the entire assembly of hot riser shop-fabricated items will be required to sup- tubes and the upper hot plenum porinstallatioe th t n sequence In-situ assem- t RiserHo Tubes This sectiopanee th f lno blf pre-fabncateyo d section s anticipatei s o t d consists of 2 m x 10 in x 65 ft rectangular employ bolting, since many items are tubes arranged around the reactor vessel The removable dunng major maintenance opera tubes extend below down to the floor of the lions However, to meet the leak rate require- bottom cold plenum and are welded (o the ment, most of the components would be of front plate The tubes also extend above up to welded construction Installation of horizont- the upper hot plenum and support this pie al hot duct sections is anticipated to loliow num at their top end The entire assembly of pounn strippind gan f colgo d duct floord an s the nser tubes and the hot plenum is floating walls but before the top slab is formed and FIG 6 RCCS arrangement isometnc 3TEB. PIATEB

MMATMN

^

CONCINT« HOTCOLOOuCT

coco FIG 8 RCCS panel configuration cross-section

heat and transferring it to the air, and a set of vide a simpls d reliablan e e mean f heao s t concentri cold an d t duc cho transportinr fo t g rejection which complement passive sth d ean air between the cooling panels and the intrinsic safety features of the small modular inlet/outlet structures gas-cooled reactor The system operates and adjusts its heat removal capability automa The RCC s i locateS d withi e reactoth n r tically withou y humaan t n interventiot I n building The physical envelope of the sys can neithe startee b r r stopped dno ma y db tem is defined by the inlet/outlet structures, vertent operator action The system does not steel ducting steed an , l cooling panels mclud y depenactivan n eo dcomponen t sucs a h FIG 7 Overall RCCS configuration ing thermal insulation and supports pumps, compressors, valves, etc, and relies The passive air-cooled decay and residual solel n naturayo l (unassisted) mode f heao s t OS heat removal system described above pro- transfer r0o0

SECTK5A NA- section B-B OUTLET INLET

FIG 9 RCCS air inlet/outlet structure

REFERENCES

1 "Conceptual Design Summary Report, Modular HTGR Plant", Bechtell a National t e c In , DOE-HTGR-87-092, September 1987 2. "MHTGR Cost Reduction Study Report", Bechtel National , IncDOE-HTCR-88-512al t e , , October 1990 3 'Top-Level Regulatory Criteria for the Standard HTGR", Gas-Cooled Reactor Associates and GA Technologies , HTCR-85-00c In , , Octobe2 v Re r2 1986 4 "Utility/User Requirements for the Modular FIG 10 Reactor cavity and RCCS installation sequence High Temperature Gas-Cooled Reactor Plant", Gas-Cooled Reactor Associates, GCRA-86-002 Rev 5, April 1989 5 "Licensing Basis Events for the Standard MHTGR", GA Technologies, Inc. DOE-HTGR 86- , Februar1 034v Re , y 1987 IMPACT OF INCREASING MHTGR needscope Th se include 84-columne dth MW(t0 45 , ) core, other core configuration MW(t)0 45 t d sa an , POWE PASSIVN RO E HEAT REMOVAL alternatives at higher and lower powers than 450 MW(t), corresponding to the maximum achievable for two different reactor vessel sizes The trade study was structured to provide visibility to the strengths T D DÜNN, A A SCHWARTZ, F A SILADY and weaknesses of each alternative relative to the top level requirements General Atomics, San Diego, California, United State f Americso a KEY REQUIREMENTS AND STUDY CONSTRAINTS

Abstract Top level user and regulatory requirements are given in the US-DOE MHTGR program's Overall In 1990 a cost reduction study recommended that the reference U S MHTGR module design be Plant Design Specification (OPDS) An alternative core configuration was considered viable only if it change 84-coIumnn a o dt MW(t0 45 , ) annular reactor cor attaio et n improved economics wit same hth e satisfies these mandatory requirements or musts of the plant However, since the margins by which an high leve previoue f safeto lth s ya s reference 66-column MW(t0 35 , ) MHTGR modul objective Th e e alternative meets the top level requirements vary, the requirements also served to define desirable, rather of this pape reporo t s i r t of recentla f y completed core configuration trade study that reviewe basie dth s than mandatory, objectives or wants for that recommendation with more detailed assessments The trade study examined alternate core configurations in terms of the size, shape, and power level Core configurations at 450 MW(t), an mandatore Th y requirement mustr so s defin desiree eth d limitation e resultth cord e sth an er sfo alternativ highet ea lowet a r e powerron powed an , r were considered These alternatives representee dth configuration trade study As discussed in the Cost Reduction Study the plant costs of any alternative maximum achievable power for fuel element for two different reactor vessel sizes Fuel, reactor internal plant must be competitive with coal plants Two key safety requirements for the MHTGR plant are and vessel temperatures during pressurized and depressurized conduction cooldown transients are judged essentia commerciao t l l deployment presente compared dan limito dt s neee Base th improvo d t n do e economics without sacrificine gth MHTGR's high leve f safetyo l trade th , e study confirmed thapreviousle th t y selected 84-column0 45 , plane Th t shal designee b l 1 perforo dt safets mit y functions without credi r shelterinfo t g MW(t) annular design remains the preferable configuration r evacuatioo publie th f nco beyon plant'e dth s exclusion area boundary

INTRODUCTION plane Th t shal designee b l 2 perforo dt safets mit y functions without relianc contron eo l room equipment automatee th , d plant control system r operatoo , r actions In October of 1990 the Cost Reduction Study (CRS) (Ref 1) recommended that the reference MHTGR design be changed to the 84-column, 450 MW(t) annular reactor core design That In order to meet these must requirements, acceptable response needed to be demonstrated for recommendation was based on the objective of attaining improved MHTGR economics in a reactor pressurize depressurized dan d conduction cooldowns Thes rare ear e event whicsm diverse hth e active module that offere same th d e high leve 66-columne f safeto lth s ya MW(t0 35 , ) MHTGR describen di coding systems are unavailable and heat removal is by passive means to a heat sink exterior to the the Preliminary Safety Information Document submitted to the U S Nuclear Regulatory Commission uninsulated reactor vessel Alternatives must also possess adequate shutdown margin, reactivity control 84-columne Th MW(t0 45 , ) design appeare beso dt t mee utilite th t y industry objectives Since that time, and fuel performance which ten limio d t allowabl e tth e average core power densit core Th ey heighs twa more detailed assessments have been performed on the 450 MW(t) design restricted to 10 or fewer fuel blocks to assure both seismic and nuclear axial stability, while factory offsite manufacturin shippind gan g limitations contro reactoe th l r vesse stead an l m generator maximum objective Th core th ef eo configuratio n trad recommendatioS erevieo t stud CR s e wywa th n with weights and sizes To assure that there will be no degradation in the level of safety, constraints to knowledge of the more detailed assessments that have been performed The study more thoroughly increasing power output and to more cost effective design selections imposed by the reactor core and CO Co examined alternate core configurations, in terms of size, shape, and power level, to best satisfy the utility internals, the steam generator, circulator and fuel handling and storage were taken to be the same as m S Tabl muste CR th liste 1 el th ssal use selectinr dfo 0 cor0 e gth e configuration alternative f thiso s trade TABLE 2 CORE CONFIGURATION SELECTION CRITERI THEID AAN R IMPORTANCE study

TABLE 1 Selection Criteria (Wants) Relative Importance CORE CONFIGURATION SELECTION CRITERIA MUSTS Normal Operation Normal Operation/Refueling 1 Provide cost margin relative to coal High • Provide shutdown margin of 1 % 2 Utilize development cost benefit of NPR Moderate • Provide effective reactivity control 3 Provide margin on peak fuel temperatures Low • Provide axial power stability • Meet reactor vessel/steam generator 4 Provide margin on vessel fast fluence limits Low vessel shipping weights/sizes 5 Provide margin on core pressure drop limits Low • Meet fuel performance limits • Mean busbar costs < $2000/kWe 6 Provide component design margin Low Off normal Operation 1 Provide margin on PCC temperatures Moderate Meet depressurized conduction cooldown (DCC) 2 Provide margin on DCC fuel temperatures Moderate and pressurized conduction cooldown (PCC) temperature limits 3 Provide shutdown margin during cold water Low Meet protective action guidelines limits ingress 4 Provide margin unprotecter sfo d reactivity events Low 5 Provide seismic margins Low selectioe Th n criteria that provid desirable eth e objective core r wantso th er configuratiosfo n study are given in Table 2 in order of descending importance for both normal and off-normal operation Only wants which discriminate between alternatives were included The selection criteria fall naturally into categories that improve the design in the areas of plant economics, passive safety or design margins that was used in the 350 MW(t) design (the NPR design), as well as the maximum power level [500 High priority was given to capital and operating cost reduction Furthermore, maximizing the MW(t)] that can be attained within the larger vessel used in the 450 MW(t) design Cores with lower development cost benefits of commonality with the MHTGR-New Production Reactor (NPR) was assigned power levels were not considered because they were found in the Cost Reduction Study to not be a moderate priorit t thiA sy design e stagth f e ordeo n i , meeo rt passiv e th t e safety requirements, margin economically viable Cores with higher power levels were not considered because of either conduction in componenfued an l t temperature limits during conduction cooldowns were required Other design area cooldown temperature constraint r becausso e they would require onsite vessel fabrication margins were assigned lower relative importanc assessiny B e overale gth l plant economics, passive safety and design margins, the best core configuration was selected Five alternative core designs as illustrated in Fig 1, which met the must requirements were evaluated Thesfollows a e ear s ALTERNATIVE CORE CONFIGURATIONS A The reference 84-column, 10-layer, 450 MW(t) core, which was the recommended design The purpose of this core configuration study was to examine alternative core designs which satisfy in the MHTGR Cost Reduction Study This core design utilizes 84 columns of fuel both safety and economic requirements Therefore, the alternative designs which were considered elements, with 10 fuel elements in each column Reflector elements surround the fuel encompasse maximue dth m power level [410 MW(t)] safelthan ca tattainee yb d withi smallee nth r vessel elements 90-columnA . E , 10-layer MW(t0 50 , ) core. This configuratio alss ni o simila Caso t r eC

4 COLUMN» S 7Z COLUMNS »0 COLUMNS •>0 COLUMNS 70 COLUMNS above, but achieves 500 MW with an increased power density. The core height is the same as Case C, but the alternative has an increased steam generator size, hot duct diameter, and lower plenum height to reflect the increased power level and increased coolant flow rate associated wit highee hth r power density.

COMPARISO ALTERNATIVEF NO HIGHEO ST R WEIGHTED WANTS

10 LAYER 10 LAYER IO LAYER 9 LAYER 10 LAYER üüi : alternativee Th s were ranked relativ wantse th r eac o et Fo h. wan beste th t alternative(ss wa ) Ü UU identified followed by the others as judged in qualitative categories of better, good, and ok.

Provide Cost Margin Relativ Coao et l (high weight)

MWtl0 41 ) 45O MWO 45O MWI) 50O MWCI) Comparative cost trends were calculated wit ha mode l which employ same th s e cost trend algorithms used in the CRS as supplied by program participants for their scopes of responsibility. These FIGURE 1 cost trends wer _ e_ plan fowere d rth tan e conducted using approximate method efficientlo st y compare RANGE OF REACTOR SIZE AND CONFIGURATION ALTERNATIVES EVALUATED design options. The costs are in 1989 dollars and are adjusted to a replica plant estimate.

The three most significant factors affecting cost orden i , importancef ro core ar ,e power, reactor 72-columnA . B , 10-layer, core which produce 0 MW(t)41 s . This configuratioa s i n vessel size, and core height. Figure 2 shows the total busbar cost for the five configurations of this variation of the previous 350 MW(t) commercial design, incorporating 6 additional fuel study 84-cohunne Th . MW(t0 45 , )cos % desigt 14 improvemen a n s (Casha ) eA MW(t0 t 35 ove e )rth columns and a somewhat higher power density. It is most like the reactor design being reference bese Th t .configuratio n fro mcosa t standpoinMW(t0 50 e ) th s Casi t havineE g199a 6 cost proposed for the NPR. It utilizes the same diameter reactor vessel as the previous 350 improvement. Figur showe3 capitae sth five l th cose r configurationtfo thif so s study cose Th t. changes MW(t) design and the NPR design but differs in steam generator size, hot duct diameter, from the reference are essentially the same with the largest difference being a 19% capital cost and lower plenum heigh consequenca s a t increasee th f eo d power rating. improvemen MW(t0 50 e ) th desigr fo t n (Cas. eE)

A 90-column, 10-layer MW(t0 45 , ) core. This configuratio varian a Case s ni th eA n to The 84-column, 450 MW(t) design (Case A) is in the middle of the cost range along with the two reference design which utilize extrx ssi a column f fueso l withi same nth e reactor cavity othe MW(t0 r45 ) configurations cosA . tassociates i saving % 3 f so d with improved thermal conditions to reduce the power density. of temperature and pressure and more effective design selections. The remaining 11 % is due to the economy of scale. A 90-column, 9-Iayer, 450 MW(t) core. This configuration utilizes 90 columns of fuel elements similar to Case C above, but has only 9 fuel elements within each column. The Case B, the 410 MW small reactor vessel configuration, is 3% more expensive than Case A. total core height and reactor vessel height is reduced accordingly. The 450 MW(t) power This is due to lower power with economy of scale working against it. There is some cost advantage due CD Ol ratin achieves gi increasiny db powee gth r density. to the smaller reactor vessel. The savings for this configuration are in the vessel system, reactor system, CD fuel handling system reactod an , r building These savingt enougno e overcomo ht sar decrease eth n ei O) power and this case is right at the capital cost must requirement

Case C, the 90-column, 10-layer 450 MW(t) core configuration, is slightly more expensive (0 2%) than Case A The added fuel fabrication cost due to 60 additional fuel blocks is greater than the saving reducen si d core side reflecto circulatod an r r power

laye9 e r Th core configuratio bette% 1 f Casrs o ni tha eD n numbe A Cas eA f area o r e ar s helped by this change In decreasing order of savings, they are reactor building, reactor vessel, core barrel, fuel fabrication, core reflectors, circulator power Reactod an , r Cavity Cooling System (RCCS)

corW FinallyeM configuration0 50 ,e Casth bette , % eE 6 s ri , than Cas eA Thi almoss si l al t due to the economy of scale The capital costs do not go up as much as the electrical output Savings FIGURE 2 m fuel fabrication cost contribute 02% towar totae d th highe o lt e Thidu r s powesi r density overcoming OPERATING COST COMPARISON OF THE ALTERNATIVE DESIGNS the effect of 60 additional fuel blocks

Utilize Development Cost Benefit(moderatR NP f so e weight)

All of the commercial alternatives described benefit economically from the technology transfer MW(t)0 35 fro e mth , 66-column laye0 1 , r MHTGR-NP expectes i t I RwilR dl NP provid thae th t e eth design methods, the validation and verification of computer codes used in the design, much of the required technology development, and many of the design studies and details that would be used in any of the alternatives Furthermore, the NPR could provide the prototype plant in which the performance of the fundamental safety features of the MHTGR is confirmed Therefore, this evaluation was limited evaluatino t e alternativgth e core configuration r physicafo s l identit f componentyo s with theiR NP r counterpart recognizes i t I s d that even though physical identit achieveds yi , difference desige th n i sn conditions for that component will exist as a result of differences in the duty cycle, seismic level, or other plant component design RCCS) e (sucth s musd ha addressee an , b t desige th n di n development

Of the alternatives considered the 72-column, 10-layer 410 MW(t) core provides the most commonality with the current NPR design The reactor vessel is essentially identical, having the same diameter and only a slightly increased height to accommodate a larger hot duct, lower plenum and cross FIGURE 3 vessel diamete metallie Th r c reactor internals (othe ductt r alsho e tha )e ar o n th essentiall y identical, with CAPITAL COST COMPARISON OF THE ALTERNATIVE DESIGNS only a minor difference in the height of the core lateral restraint The graphite permanent reflector will e identicae graphitb th d an el core support will differ only slightl heighn e neutroyi Th t n control have both failed to perform respective functions of providing forced cooling of the core Decay heat is assemblies, in-core flux monitorin refueline mucd th g an f ho g equipment wil identicae b l l because th f eo then removed by thermal radiation and natural convection from the reactor vessel to the natural commonalit n controi y penetrationA d locationro lNC d an s s use r refuelinfo d g Reactor service circulation air-cooled RCCS The thermal transients for these two accidents use conservative values for equipment would also be identical Core instrumentation required for the commercial plant would be the decay heat which are 12% higher than nominal The 12% is composed of two parts 2% is for the identical to some of that which may be anticipated for the NPR plant The plant protection hardware assumed uncertaint powen yi r level instrumentation account % uncertainte 10 th r d sfo an , decan yi y heat, would also be identical, although the software would be different to reflect the difference in trip material properties and calculation methods requirements The shutdown coding heat exchanger (SCHE) may also be physically identical even though they may have different safety classifications The core design is substantially different due to NPRs s par f A thio t s stud determino yt bese eth t overall core configuration, PCC DCCd an s s were inclusio f targeno t fuel element higd san h enriched fissil esteae fueTh ml generator woul differene db t evaluate l fival er cordfo e alternatives Tabl summarizee3 systee sth m parameters, temperaturey ke f so accommodato t highee eth r power level circulatoe th , r woul differene db t becaus desirabilite th f eo f yo component componend san t limiting temperatures during fiv e PCCth e r caseinitiae sfo Th s l pressure, utilizing magnetic bearings, and other plant components would be different because of different and core inlet and outlet temperature are the same for all concepts In addition, the radial and axial requirements and/or different power levels zoning s adjustewa flatteo dt n power profile reducd an s maximu e eth m core peaking factor

extene Th commonalitf o t slightls yi y othelese th sf rwit o alternative l hal largee Th sr cord ean larger reactor vessel eliminate some of the physical identity obtained with the 72-column configuration TABLE 3 However, some of the core instrumentation, protection hardware, reactor service equipment, and ALTERNATIVE CORE CONFIGURATION DESIGN COMPARISONS refueling equipmen MW(tt0 woul45 e commoe )dth b alternativesr Fo n SCHe ,th E woul commoe db n However, for the 500 MW(t) core, it may be expected that the SCHE will also be different and the extrapolation from the 350 MW(t) NPR components such as the steam generator and circulator is greater Case A Case B CaseC Case D Case E Power, MW(t) 450 410 450 450 500 The assumption that the MHTGR-NPR project precedes the commercial MHTGR provides a Fuel columns 84 72 90 90 90 significant advantage in the development and construction costs for all the alternative commercial designs Core layers 10 10 10 9 10 The cost advantage realizee sar d becaus similarite th f eo f majoyo r component learnine th d san g curve Power density, w/cc 60 63 56 62 62 by members of the contractor teams Vessel ID, ft 237 21 5 237 237 237 Pressure, psia 1025 1025 1025 1025 1025 In summary mose th , t beneficial alternativ commonalitn eo Cass yi whiceB - 72 base s hi a n do Core inlet temperatureF ° , 550 550 550 550 550 column core This alternative utilize same th s e reactor vesse MHTGR-NPe l th siz s ea othee Th Rr Core outlet temperature, "F 1300 1300 1300 1300 1300 alternatives all utilize a larger reactor vessel size to accommodate the 84-column or 90-column cores Peak Pressurized Conduction Cooldown Temneratures Limit Provide Margi Pressurizen no d Conduction Cooldown Temperatures (moderate weight) FuelC ° , 592 603 618 640 622 - 1600 Vessel side wall, °F 760 739 760 761 790 800

pressurizee Th d conduction cooldown (PCC depressunzed )an d conduction cooldown (DCCe ar ) Upper plenum shroudF ° , 1354 1420 1358 1383 1420 1500 important classes of events which challenge heat removal and therefore the control of radionuchdes The Core barrelF ° , 1325 1373 1330 1348 1390 1400 00 1478 1520 1496 1532 1578 1800 -J PCC and DCC events occur when the Heat Transfer System (HTS) and Shutdown Cooling System (SCS) Operating control rod, °f 00 The PCC event is typically initiated by loss of offsite power and turbine trip plus failure of the to start on demand The RCCS then removes decay heat from the vessel by natural convection and 00 SCS to start, the reactor trips and inserts the operating control rods making the reactor subcritical With thermal radiation The fuel temperature is the limiting component temperature for the DCC event A the HTS and SCS not available to provide forced cooling to the reactor, decay heat is then removed by compariso peae th kf no fue l temperatur functioa s ea f poweno shows i r e peaFigurn ni Th k fuee4 l thermal radiation and natural convection from the reactor vessel to the RCCS temperatur varioue th r efo s core configuration alse ar so tabulate Tabln di e4

componeny ke e Oth f t peak temperatures showe variouth Tabln r ni fo s e3 cases e fuelth , , 1700 operatin gupped controan rd plenuro l m shroud temperatures have bot e leashth t sensitivit o coryt e Cast D configuration/power change havd greatese san eth t margi componeno nt t temperatury ke o etw limite Th s da E parameters ,e mos th whic te ar sensitivh o cort e e configuration/power e closeschangear d o t tan s component temperature limits during PCCs, are vessel temperature and core barrel temperature These two parameters were examine determino dt bese eth t core configuratio standpoinC n PC fro e mth t 72 Col 10 High

1500 resultC PC se giveTh Tablm n e3 indicat e thal fival t e configurations mee l appropriatal t e 90 Col. 9 High u_

component temperature limits with acceptable margins Indeed there is only minor variation in the peak £ vessel and core barrel temperature for the different configurations 1400

standpointC FroPC e m th 72-column e th , MW(t0 41 , ) core marginallCass i , eB bese yth t design because it offers the most margin to the vessel temperature limit while still offering an acceptable margin 1300 350 400 450 500 550 core foth r e barrel temperature However 84-colume th , MW(t)0 n45 , Cas desigeA n also offers large Pnwr MWII) margins to both the vessel and core barrel temperature limit FIGURE 4

COMPARISO PEAT NA K FUEL TEMPERATURES DURING DEPRESSURIZED givea r nFo core power ther negligibls ei e difference betwee recommendee nth d g4-column core, CONDUCTION COOLDOW FUNCTIOA S NA POWEF NO R Case A and a 90-column core Case C as shown in Table 3 A 9-layer high core Case D increases the maximum core barrel temperature approximatel negligibla s y ha 20° d Fean effec maximun o t m vessel temperature as compared to a 10 layer high core Case C At 500 MW(t), Case E, is acceptable but TABLE 4 reduces margi boto nt vessee hcorth d an le barrel temperature limits when compare same th eo d t desig n COMPARISON OF ALTERNATIVES FOR DCC FUEL TEMPERATURE WANT configuration at 450 MW, Case D The margin to the vessel and core barrel temperature limits is only 10° Casr Ffo eE Peak Depressunzed Case A CaseB CaseC CaseD CaseE Limit Conduction Cooldown Temperature Provide Margin on Depressurized Conduction Cooldown Temperatures (moderate weight) Average FuelC ° , 1240 1350 1230 1300 1325 _ Maximum FuelC ° , 1550 <_ 1630 1540 < 1620 < 1620 1600 The DCC event is typically initiated by a small primary coolant leak 0 32cm2 (0 05m7) near the Operating Control 2050-2150 >.2285 2300-2400 2190-2290 2210-2310 -2)50 reactoe th f o r p vessee reactoto Th l r trips automaticall reactow lo n yro pressur side th e d reflectoean r Rod, °F operating control rods are inserted The HTS fails immediately after the initiating event and the SCS fails Side Wall Vessel, °F 735 870 740 795 845 900 Because of the 1600°C limit on the fuel temperatures, the 9-layer high core, the 500 MW 10 TRENDS IN SAFETY CRITERIA laye rsmal higcolumn2 W 7 M he l reacto0 coreth d 41 , an , r vessel design estimatee sar e righe th b t o dta t FOR FUTURE REACTOR PLANTS fuel temperature requirements. Case A and Case C have only minor differences in the core designs. As E. HICKEN one would expect, Case C, the core with the most fuel columns, (i.e., the lowest average core power Institute for Safety Research and Reactor Technology, density) slightls ha , y more margi pean no k fuel temperatureeventC DC . e Howeverth r sfo , with respect Forschungszentrum Jülich GmbH, to the peak operating control rod temperatures Case A without the six extra columns on the active core Jülich, Germany periphery resultlowese th n si t temperatures. Casrankes i eC d next followe othee th ry db three . Abstract

Design, commissioning and operation of nuclear power plants require safety criteria to achieve a specific safety level. Formally, safety criteri e relatear a o lawst d , regulation guidelinesd an s . CONCLUSIONS Becaus l majoal n ei r countrie t fixed ye tere t criterith w ,m no sne "safete aar y criteria uses i " d here in a broader sense including views from individuals. Several trends in safety criteria are discussed and justified and approaches to reduce accident dose so that evacuation of the public is not necessary Five core configurations have been examined in terms of desirable objectives related to plant reviewede ar . economics, passive safety, and component design margins. The results of the overall ranking of toe five core configuration MW(t) 0 thas si 45 te Casth , , 84-columneA laye0 1 , r core previously selectee th n di CRS remains the best configuration. While not the highest rated in every category. Case A has attractive I. Introductio somd nan e history economics and retains the passive safety margins which are higher rated wants. A close second is Case Design, commissionin operatiod gan f nucleano r power plants require safety criterio t a C, the 450 MW(t), 90-column, 10 layer configuration. However, it has slightly less cost margin and less achiev specifia r efo c safety level. flexibility in locating the outer control rods. The remaining three cases are right at the 1600°C fuel limit Increased understanding, operating experience, incident accidentd an s s wit a hreleas f eo during conduction cooldown which is too close at this point in the preliminary design. Case B which fission products to the environment established a continuous process for changes in safety criteria wise major .hTh fo rmor e changeb o et precisr o smajo- r fo e- r improvements during retains the same size vessel as the 350 MW(t) NPR has a number of advantages over and above these day s beeha s n intensifie concerno t publie e th du dy sb c abou safete th t f nucleayo r commonality, in the area of component design margin. Its chief disadvantages, however, are in the reactors. higher weighted areas of plant economics and margin on the passive safety DCC temperatures. Fro beginnine mth nucleaf go r energy productio nsafa e operatio mandatorys nwa , because the potential of danger to the public was realized from - unfortunately - the atomic bombs. For the first nuclear reactors a core melt and - due to a missing strong containment - a major fission product release with related early fatalities was predicted. Hearings in the USA in the ACKNOWLEDGEMENTS early Iff s resulte requiremene th n di copo t e with large break LWR'sf so . Emergency core cooling systems and a containment to confine the fission products were, therefore, installed - at least for reactors in the Western Countries. The first major risk study - WASH 1400 - e authorTh s would lik thano et Departmene kth managemene f Energo tth d yan f Generao t l showed that sequences requiring more systems might have a higher probability for core melt than sequences with a large break as initiating event The accident at TMI-2 demonstrated Atomics for permission to write and present this work, which was supported by the Department of this addition i , influence th o nt humaf eo n behavio sequence th n ro eventsf eo accidene .Th t at Chernobyl evidently showed the world-wide impact of a major fission product release. Tnis Energy, San Francisco Operations Office Contract DE-AC03-89SF17885. acciden mor- t e thaothey nan r even resulte- t reactioa publie toleratn o di t th t y ncb no ea major fission product releas mory ean e from nuclear energy production tha d worts i ad t .I o ht laca f informatioo ko t e du n expert publid an s c t car mostlr nucleano efo o yd r powered REFERENCE submarines and production reactors. In chapte I generaI rchapten i d I an soml II r e more detailed trend safetn i s y criterie ar a discussed e conclusionTh . e basesar written do n materia privatd an l e communications with 1. "MHTGR Cost Reduction Study Report,' DOE-HTGR-88512, October 1990 experts in several countries, especially USA, France, Netherlands and Sweden. oo

This view seems logicalrealitn i s ver i t t yi ybu , difficul o achievt t e agreement between Dose limitations outside the plant politicians about nuclear energ t leas A ysomn i t e countrie a solutios nr fa appear e th n i s future For design basis accidents usually dose limitations outside the fence of a plant exist There are no limits with regard to sequences with a major release of fission products Several trends can identifiee b d Nuclear reactors should be small, simple and safe (SSS) 1 ) e g for sequences with a probability of This view will be discussed below 10"n/plant/y core th releas e f inventorro th % limite s ei 1 0 yo dt 10 n".ï/plant/y cor e release th reth f inventoro limites ei % 1 o dt y lO'^/plant/y core releasth e ef rth inventoro limite% s e i 0 1 yo dt Nuclear reactors shoul smale db l somn i n9 e equalr proposalo 8 o st s The term "small" is used in a different sense, some believe in plants with several ten or hundred Megawatt thermal while some others define plants with 600 MWe to be small core th release f o inventorth . g % limite s ei e 1 0 proteco yo t dt grounde th t , however Therefore, always a definition should be used with the term "small" There seems to be the probabilite th specifiet no s yi d following trend o 3) ancutoff frequency for core melt scenarios is to be specified (between 10 ° - 10 large plants with 1200 -150e 0MW /plant/yr) with a specified maximum dose Sometimes the doses will be specified medium size plants with 600 MWe dependin e distancth n o ge fros e interestini planth m t I t g thae producth t f o t small reactors with severa hundreo t n te l d MWth contaminated area tunes the probability is not always decreasing with distance With It is evident that smaller reactors can be designed more easily with regard to a high safety these specification evacuation sa n outsid fence eth e necessare migtb t hno y level However, it should be kept in mind that considering the overall risk from nuclear principlen I possibilitieo ,tw limitatioe th r s fo majo a f no r fission product release exist energy productio nsmalle- r reactors mus more b t e saf f nucleaI e r energy production shall a) a design of a containment to cope with a molten core play a major role in future, one has to think about several thousand large plants or several ten b) a design with low power densities and low core temperatures avoiding a major thousand small plants Available sites, operating crews, etc have to be taken into account In fission product release practise, the existing proposals for future reactors concentrate on large and medium size reactors, using also mainly existing technology existine Th g tren favouo t s di r cas , mainleconomo et a) e ydu y

Nuclear reactors shoul simple db e Safety Goals e trenTh d towards simpler system evidens si t This would also increase robustness However, regulation d guidelinean s s sometimes prevent major simplifications g requirement e , r fo s Safety goals specify limit functioa s sa f probabilito n dosed e requiremenyan Th s majoo N t r redundancie informatiod san n abou plane th t t stattrene Th ed towards more passive systems releas f fissioo e n product r frequenciefo s s abov 6/yr/plant0 1 e a poin s i " t valu f thio e s mainly safety systems will support simplification relationship yearsWhil w o safetfe ag a ey goals were discussed intensively, nowadaye th s construco t t trenno s d i continuou a t s relationship between probabilit specif o t dosd t yan e bu y 5. Design Basis Accidents only one point as outlined above. Safety system componentd an s e reactoth f so r plan designee ar t copo dt e with design basis accidents. It is obvious that the existing design basis accidents will continue to be a ALARA basis for the design of components and safety systems. The calculations include safety margins. On the contrary, the trend is to calculate loads from core melt scenarios with s reasonable abbreriatioa Th w lo s "a e f achievableno uses "i d frequently. Thio s tw ter s mha realistic assumptions. In addition, completeness of scenarios analysed will not be components "feasibility" and "costs". A few years ago, in the USA an improvement should be required. performe coste th f sdi were below 100 reductioDollae S 0th U man-rem e r rto on f no . The trend is not to explicitely specify a relationship between a dose reduction and cost. 6. Phenomena relate Severo dt e Accidents III. Trends for safety criteria for containments of LWR It is not yet clear if scenarios with core melt become design basis accidents or if it has to In the USA as well as in Germany some trends can be identified for safety criteria for a be shown that containments can cope with loads from those scenarios with some containment. The specific interest in safety criteria for the containment originates from the certainty includo t t trene no . Th e s d i sever e accident desige th n sni basis accidents. philosophy thae containmenth t e lastth twil e barrieb l r agains a majot r fission product release. 6.1 Melt / Water-Interaction ("steam explosion")

1. Containment Failure This phenomena has to be considered in-yessel and ex-vessel. Due to the lack of firm basis, assumptions reach from exclusion (based on probabilistic assessments) containmene th A US fain e l ca tth percen wit0 n 1 I ha t probability procedur.A o t w eho to consideration cleao .N ridentifiede trenb n dca . calculate the containment failure rate could not be identified (e.g. 10 percent of all sequence percen0 1 r so t dependin expertjudgement)n go Germann I . containmene yth t should stay intact above a cut-off frequency. This seems to be the general trend. 6.2 Hydrogen Burning and Detonations EPRI require r advancefo s d system appropriatn a s e desig r mertisationo n n I . 2. Leakage Rate general a complet, e oxydatio e fuecladdind th ro limitatioa l f no d 0 gan 1 o nt percent hydrogen concentraion in dry atmosphere will be required. For AP-600 an EPRI specifies a maximum leakage rate of 0.5 weight percent per day, while the values ignitor system is under consideration while SBWR's will be inerted. %/d1 - 1 , 0. respectively r E o C d e th %/ r 5 0. .Fo - AP-60SBW2 e e fo0. th rth e d Rar 0 an trene relater Th dfo d requirement Germanclearn t si ye t . no s yi design 80 + the leakage rate is specified to be below 0.34 %/d. maie Th n philosophy behin ddose thith es si limitation . However, this specified value canno testee tb d under accident conditions. 6.3 Melt Attack on Containment Structures and the Pressure Boundary evidens i t I t tha leakaga t e rate wil specifiee b l appliee b o dt d als sequencier ofo s with core melt. The general trend is that melt impinging structures should not increase the leakage rate, very often concrete protects sensitive areas. 3. Coolin Containmene th f go t 6.4 Direct Containment Heating ("DCH") EPRI as well as AP-600 require a passive cooling system. The general requirement is that cooling must be provided also for sequences with a core melt. However, if the integrite containmene Th th f yo t migh threatenee b t mely db t injected under high capacity of the containment is large enough to store the energy released (mainly decay RPV pressure inte containmenth o t atmosphere e generaTh . l o t tren s i d heat r severafo ) l days, e coolinprovideb n ca gactivy b d e systems n additionaA . l depressurize the RPV (as persued in Germany but also recommended in the USA) requiremen capabilite th coolins e i t th f yo g system loweo st containmene th r t pressure ospecificallo rt y desig compartmene nth t belo V around wRP an e dth (e.g haldesigo e .t fth n pressure within days). 6.5 Melt/Concrete Interaction 4. Scrubbin Containmene th f go t Atmosphere In Germany a core catcher will be required; the specific design has not been A possible requirement is not yet discussed in Germany due to the lack of a contaiment decided. In the USA a design to allow a spreading of melt is recommended as well spra scrubbinA y US containmensysteme e th f th g o n I . t atmospher undes ei r discussion melte furthes watei th a t f s I .o a rp r layeto recommende n o r seleco dt aren a t o at for designs with containment sprays. meet 0.02 m^/MWth. 6.6 Pressure Increase due to Decay Heat In Germany most probably a filtered venting will be required. In the USA a trend can be identified to use existing heat removal systems. This is due to the time available.

6.7 Elevated Temperatures In German temperaturee yth containmene th f o s t atmosphere wil relatee b l o dt the containment pressure or calculated as a result from hydrogen bums. In the USA it will be required that the leakage rate shall not increase.

7. Instrumentation to survey Containment Integrity In Germany position indicator operatorA installede sUS ar e th s n shoul.I enablee db o dt survey periodicall continuouslr yo y containment integrity.

8. External Loads In guidelineGermanK RS existine A yth sUS wil e gusede b th lguideline n i ; s wile b l applied. It is not yet clear, if external loads will have the same weight as internal loads, happen ca i.e t i . n tha probabilite th t fissioa f yo n product release from internal loads si very smal t relativelbu l y high from external loads e.g militara . r planyai e crashn I . takaddition o t larg t e cleaew th no ers ho i uncertaint t ,i y bands into consideration.

9. Containment Integrity during Shut Down Periods The discussion about a related requirement has not been started. In the USA industry has been aske develoo dt p appropriate procedures.

10. Bypass Sequences Some specific pipe break o allod s e wa e containmentbypas th th f o o t s e Du . improvement of containment integrity bypass sequences have to avoided. There is a trend in several countries to require systems with primary coolant outside the containment to be designed at full pressure.

11. Fires There is a general trend to further prevent fires and to mitigate its consequences.

IV. Summary

It is generally agreed that future reactors should have a higher level of safety. This includes design as well as operation. The main trend is nowadays to concentrate on LWR's and to improve the containment function in a way that scenarios with core melt can be coped with. It is wort noto ht e tha dose th t e wil limitee b l mako dt e evacuation unnecessary. Related criteria undee ar r discussio beet t havye nt nbu agreeeno d upon; they differ sometimes substantialln yi different countries. ANALYTICAL METHODR SFO PREDICTIONS OF THERMAL RESPONSE, ACCURAC PREDICTIONF YO S (Session III)

Chairman

M. fflSHIDA Japan ANALYTICA EXPERIMENTAD LAN L INVESTIGATIONS underway in Japan Recently HTGR programmes have mainly been focused on the OF THE PASSIVE HEAT TRANSPORT IN HTRs development of modular reactor designs with some hundred MWt which mainly rely on UNDER SEVERE ACCIDENT CONDITIONS passive systems to achieve a high degree of safety The results of probabilistic safety analyses (PSA) have shown, that the risk of an W REHM H BARTHELS, JAHW , N HTR-Modul smalles ei rwit R tha hcomparativela nPW thaa f o t y scaled-up power /!/ Forschungszentrum Mich GmbH, 8 Julien, Germany Accident sequences with a probability for major release less than 10 /yr have been excluded due to supposed small risk contributions J CLEVELAND1 Oak Ridge National Laboratory, A safety goal of the HTR Module (200 MWt) for beyond design accidents (about Oak Ridge, Tennessee, 6s /yrtha0 ' <1 ) t radiation impact e withi sar e Radiatioe rangth nth y f limitb eo t nse s United State f Americso a Protection Ordinanc desigr efo n basis accidents, evacuatio publie th f no c shoule b t dno M ISHfflARA regarded as necessary to avoid health effects Furthermore, an extended safety philo Japan Atomic Energy Research Institute, sophy is m discussion Safety concepts for future reactor designs have presently been Ibaraki, Japan proposed whic excludhcan e catastrophic consequence publithe almoscsby for detea t r mirustic approach taking into account severest accident scenarios g advance e , d safety Abstract enclosures (especially containment) for PWR and FBR /2/ Burst-proof cast iron pres- Thennodynamic accident analyses have been performed with computer simulation sure vessels, for example, are being considered for the H'l'K in context of the question model o investigatt s e core heatup sequences, sensitivity analyses, power variations, How safe is safe enough9 anticipated transients without scram d coran ,e displacement considerationr fo s With this background, analyse severesf so t core cooling accidents with CORE HEATUP probabilistic safety analyses (PSA) of small gas-cooled high-temperature reactors (e g sequences and important temperature loadings of safety barriers have been performed HTR-Module) In worst case considerations where not only a loss of the active heat e HTR-Modulfoth r e design wit ) h Sensitivitieregar(A o dt f accideno s t analyses, removal system is assumed but also a loss of the vessel cooling system, the heat would be (B) Accuracies of analytical methods, and (C) Importance for the safety concept transported into the surrounding concrete structure In such a case the concrete would act as a natural long-term intermediate heat storage dissipating the heat through the 2. SENSITIVITIES OF ACCIDENT ANALYSES concrete surface The dominant safety barrier of modular HTRs are the coated particles within the Large scale and reactor safety experiments have been performed to investigate passive graphite fuel element. Important temperature loadings of the safety barners are shown heat transport mechanism whics- cooldown hca H'l'na K core during severe accident r corfo e1 coolin b inTa g accident eventsm s with total failur f activo e e systemd an s conditions - for validation basis of computer simulation codes used for accident analyses passive heat removal dissipatio y naturanb l convection, heat conduction thermad an , l In general comparisone th , experimentaf so analyticad an l l results with computer calcu- radiation For the heat transport analyses analytical methods /3,4,5,6/ have been used in lation heae th tf stransporo t goocoden i e ds ar agreemen t connection with experimental test facilities /7,8 o investigat/t e reliabilitth e d an y 1. INTRODUCTION sensitivit computef yo r simulation results ) Base(Ta2 b deterministin do c assumptionsa severest accident scenario initiate anticipatey db d transients without scram (ATWSs i ) High-temperature gas-cooled reactors (HTGRs) are being developed in Germany, the discussed in the following as worst case consideration United States, Japan, and several other countnes Especially noteworthy is the construction of the High Temperature Test Reactor (H'l'lK) which is currently well ) Failur(1 f activeo e feedwater suppl r steayfo m generatord (SGan )

1 for vessel cooling system (VCS), CO Present address International Atomic Energy Agency, Wagramerstrassx Bo O P e5 en 100, Vienna, Austria (2) Failure of the blower trip and blower shut off valve, (O Ta Cor• b1 e Cooling Accidents with Main Temperature LoadSafete th l so y The accident scenari s beeoha n analysed separatelmodulR HT a e r uniyelectrir fo fo t - o> Barriers During Heat Removal city generation alss i t oi , typica procesa r fo l s heat applicatio e probabilitTh / n/9 f yo accident sequence expectes si lese b so dt tha n 10" r reacto8pe r year Barrier Accident Design Value

Fuel Elemen) î F ( t • Loss -öl- Coolant Accident <1600-C

PftSSUJI Vessel ( PV ) • Loo - of • Forced Convtction <350'C The consequences of ATWS depend decisively on the time at which the reactor is Transient sLOF( T) tripped and what decay heat level (DHL) is reached afterwards (power input) To exa- • Anticipated Trustons wttwutScnm(XrWS) mine the mam influence on the reactor temperature response, two power transients

Concmt CtD ( CC ) • Loss - ol - vessel Cooling <1(XTC have been investigated /10 sensitivits /a y analyses with inherent reactor trip (casy b ) e2 Confinemert ( CON ) Syslem ( LOVS ) negative temperature coefficient (DHL after 1250 sec witd )an h delayed reactor trip • Pnmjiy System Ruptures (PSH) (case 1) by control rods (DHL after 20 sec) The resulting pressure transients in the primary syste showe mar Fignm . .1

The fast pressure and temperature transients would lead in ATWS case 2 to a tempera- Tab 2. Analytical and Expenmental Investigations of the Core Coolin) W Modul- M gR 0 Accident20 HT e ( e th r sfo ture-induced failurlarge th f eeo safety valv open ei n position and/o failura e o th rt f o e blowehigo t e h temperaturerdu primare th n si y circuit which increase durin oven mi r g5 Accident Analysis Expenment 600 °C. The temperature behavior of the depressunzed reactor in a LOCA is discussed later ATWS Influenc Scraed m Tin» (Hul Transfer) THERMIX /KOWEK/ KISMET /SIKADE

LOFT Sifety Mirpint LUN0 A-2 THERMIX / KONV9C / KISMET

LOCA SjJetyMjrains KTA-5 THERMIX/H&TIWS LUNA • 3D

LOVS SensSMly ol ( Concrete Belaviot ) PSR Con Arranoemeil THERMIX /RALOC

) (3 Failur controe th systemf d eo ro l s with delayed reactor trip, ) Power-,(4 pressure- temperaturd an , e transient causn sca e (a) Loss-of-coolant accident (LOCA); the reactor is depres- surized withou VCSe th t , (b) Loss-of-forced convection transients (LOFT), the reactor is pressurized without VCS, and (c) ATWS consequences, primary system ruptures cannot be

exclude long-tere th r dfo m behavior. Fif • 1 ATWj S Pressure Transient Primare th n si y Circuit e pressurTh e peak coul e reducedb ATWn di smale Sth casy l b safete1 y valve from 1200 temperaturbare 2 6 7Th . o 0t e transient primare th n si y circui showe ar t Fign n i ; afte 2 . r

the initial temperature maximum they increase slowly during the first hours nearly to 1000- powee th o t r blower e e inputh du y b tC delaye ° .A 0 50 d blower avoidefailure b n eca y db countermeasure tha o reactoe ss th t r remains under pressurized conditions. 800- e long-terTh m temperature behavio e PRESSURIZEth f o r D REACTOR durina g LOF withoud Tan t operating vessel cooling syste mshows i Fign ni . Bes.3 t estimatd ean 0600-| 5 conservative calculations of the core temperatures are far below the design value for the CO 5 400- fuel elemen shod f 125an o tw C 0grea° t safety margins delaye.A d failurpressure th f eo e o, vessel which is heated-up during a week at about 500 °C can be avoided by counter- o H Safety Mwgin measure whicr sfo h several day available sar e (Fig. 4) . 2OO- —— best tslmafe

Core maximum temperature f sensitivito s y analysee long-terth r fo s m temperature behavior of the DEPRESSURIZED REACTOR under LOCA conditions are compiled 50 1OO 150 200 250 300 Time Ihl in t significantlTabno ; the3 .e ar y y influence e vesseth y lb d cooling system. Best estimate core maximum peak temperatur s 144i e afte C 0d rise° ran o t sabouy da 1 t Firj LOF• .3 T Temperature Transient Core th esn i

700

/StetmGtnratolSO.n Astern Own» ISG)i .ou ATWS Cuti It,-20s)

T(5h).4SO-C

K Vtssel ComcUn; Veud

100 150 300 Time :tï) Time IminJ

CO Fig. 4 • LOFT Max. Temperature Transients in the Pressure Vessel Fig 2 • ATWS Temperature Transients in the Primary Components and Concrete Cell CO OD Tab LOC: 3 . A Safety Margin Sensitivitied san s Related 20OO ——————————— — — —— to the Max. Core Temperature ! Fast Depressunzation Max Core Drffe •ence Temperature p 1750- fc] CO [%]

Fig,5 • LOCA Influence of Thermal Power on the Max. Core Tem- perature

1570 °C with conservatisms for main influencing parameters which could be reduced on an updated data base About 50 °C are to be added because of ATWS initial starting conditions with the result of 1620 °C in total The temperature peak value represents a Powa Vtiutm 250 VW great safety margi a significan r nfo t fission produc e fuetth lreleas f o element t eou s Heat ink, Vess withm the core during LOCA conditions based on experimental heating tests /I I/

e sensitivitTh y influenc f POWEo e Rt modul VARIATIONMW 0 e 20 uni n e o t th r Sfo core maximum peak temperatur LOCa n i e A with vessel r coolinfo 5 shows g i Fi n ni best estimate condition thermaA s leadW M l s 0 poweaccordingl25 f o r a pea o t yk temperature of 1615 °C in some per cent of the fuel elements as a hot spot in the core, core th exposes ei f o 10 % temperature o dt oves % ove 0 r 3 125r d 140 durinan C 0 ° 0C ° g a perio somf do e days (Fi) g6

The peak temperature of 350 °C occurs in the pressure vessel after about one week and remains below failure limit se hea (Fi Th removes i t ) gradiae 7 th n di l direction mainly by thermal radiation - besides convection and conduction - from the pressure vessel surfacvessee th o lt e cooling systee concret th e e primar t th mTh a " ) o e y°C cel0 (5 l sensitivity influence of the thermal radiation by an emissivity variation of 0 6 to 0 8

reduces the vessel température respectively •6 LOC o Fi A Temperature Distributio e Fueth ln i Element s Pressure Vessel—-

— Confinement

Core-

Axial Heigh; 'm t

Fig.7 • LOCA Temperature Profile in the Reactor Pressure Vessel Concrete Cell- Cas eA Cas eB CaseC With failure of the active water-cooled VESSEL COOLING SYSTEM - as assumed in FigCor< 8 . e Displacemen Concrete th n i t C Scenari ed Celan B l . oA the ATWS scenario- the heat is transported during a LOCA into the surrounding concrete structure of the primary cell. The temperature response of the reactor com- ponents was analysed including the confinement as a time dependent boundary compact core. The corresponding core maximum temperature transients of the scenario condition. In the thermal properties of the concrete, energy absorbing processes like case (156sA 0 °C), B(1125°C) C(1690°Cd an , e showar ) Fign i n .9 (regardles f o s release of hydrated water and phase changes were taken into account /12/. The pressure loading recriticalitd san y aspects). resulting temperature transient °C/hour1 s( ) showed thaheatee th t d concrete structure The important aspect of case A is that the heat dissipation would partly be interrupted (250 0C/meter) long-tera act s a s m natural intermediate heat storage dissipatine gth insulatioe th du pressure o et th f no e vessel wit htemperature-inducea d failur thao es t heat during some months through the concrete surface in the radial direction at core cas resulteA typicaa n si l core displacemen concrete th bottoe r th casf f Fo eo t o m. eB midheight. foundation temperatures over 750 °C are calculated after several weeks in case B. The propagation of premixed corrosion products - released hydrated water/steam and 2 2 Evaluatio Heaf no t Dissipation Perturbations reactions with hot graphite surfaces - into the confinement atmosphere has not been investigated; conditions for hydrogen combustion modes cannot be excluded and are to Different scenarios with postulated rupture primare th f so y consideresystee b n mca n di be analysed. Suitable computer code e beinsar g established temperature-inducee Th . d an ATWS post accident phase. The following CORE DISPLACEMENT CASES have consequences could be decisively reduced - as a precaution - by an additional carbon been analyse examino t d e sensitivitth e f heao y t dissipatio r greafo n t perturbations brick layer between the pressure vessel and the concrete foundation structure. The heat (Fig. 8) based on the assumptions: Case A rupture of the fuel discharge tube in the dissipation in the radial direction is more effective than in the axial direction (e.g. case C CO CO pressure vessel, case B with a core dissemination, and case C comparatively with a with lower concrete temperatures). Pebble-Bed Ptmaty Circuit

Cote Vessel

Healer

LUN• 0 1 A g ExperimentaFi l Facilit Loor yfo p Circulatio f Naturano l Convection i 'C0 >30 bar0 - HTR4 T .- p s( 0 60 400 800 1000 Time Ih) ———•- general, the LUNA results can be well predicted by the THERMIX-C-2D code with some exceptions because of modelling (3-D effects) and numerical (small pressure Fig 9 • Core Maximum Temperatures for Scenario A, B. C drops) limitation show1 1 g comparisosa Fi s pretese th f no t calculation with core inter nal natural convection and Fig 12 the posttest calculation with natural loop circulation modulae foth r r desig differencee Th n s between measure calculated dan d temperatures 3. ACCURACIE ANALYTICAF SO L METHODS vere ar y smal r stablfo l e starting condition naturae th f so l convection Unstable condi- Pretest and posttest calculations have been performed to investigate passive heat tions with a flow reversal or very small mass flow rates (e g less than 0 1 kg/s) can cause transport mechanisms in HTRs to provide a validation basis of computer codes used in numerical and/or physical instabilities accident analyse ) Larg(a s e scale experiments wit LUNe hth A test facilit simulato t y e the heat transport m the primary loop under LOFT conditions, and (b) Reactor safety 2 Compariso3. n with LOCA results experiments with the AVR reactor to simulate the heat transport in the reactor components under LOCA conditions A loss-of-coolant accident is one of the most severe accidents for a nuclear power plant To demonstrat e safetth e y behavior incorporated into smalR designsHT l , LOCA simulation tests (no HTA-5) have been conducted with the AVR reactor at KFA in 3.1 Companson with LOFT results computee th Julic r Fo h r simulatio LOCe th f no A test, dynamic heat transport analyses The LUNA loop is the only test facility which has been used to simulate heat transport have been performed to determine the temperature distribution throughout the AVR by natural convection processes in HTR primary circuits in detail (Fig 10) Observing reactor during the experiment Different heat transport codes - like THERMIX C 2D geometrica thermodynamid an l c similarity conditions, different primary circuit arrange- and HEATING-3D - have been used for pretest and posttest calculations m an ments (HTR 500, THTR-300, HTR-100, and HTR-Module) have been tested /13/ In international cooperation between KFA, JAERI ORNd an , L measuree Th d core maximum temperatur wels wa l predicte C f ° 108eo 0 pretesn di t cal- culations r examplefo , , performe Interatom/Siemeny db ) Certais 13 (Fig n deviations 200- —— Measurement (LUNA) —— Calculation (THERM) 20 XC occured for the position and the time of the core maximum temperature due to preliminary data for the LOCA test 160- Our posttest calculation of the core maximum temperature (Fig 14) is m very good agreement with the temperature recorded by the monitor elements The accuracy of the

120 HEATIND 3 G analysis resul comparabls i t e wit interpretatioe hth accurace th f no f yo the experimental result within the range of +/- 8°C, the THERMIX-C-2D results show dependinC ° 0 modelline 2 - th d n go an anC gR ° accuracapproac0 AY 4 e + th f yf o h o B SO- core This also means that the used correlation of the effective thermal conductivity for the pebble-bed core is correctly modelled 40- licensine th r Fo g procedure prediction temperature th f so core th en ei component s were nt o Tu 66 wit t 3i anv 7 hb du vessel cooling made by AVR/ORNL on a conservative basis /14/ The comparison showed significant difference componene th m s t temperature regionn behavioi d an s ) withrD 15 (Fi3- g temperature effects, especiall reflectoe th yn i r nose abov pebble-bee eth d core

Flg. 11. Pretest Temperature Calculation of the Pebble-Bed , Test with Natural Convectio Core th en I In joint best estimate posttest calculation expanden sa d HEATING-3D mode uses dwa l with updated input data Improved 3-D calculations also took into account natural convection heat transfer coefficient boundars a s y condition naturao t e sdu l convection circulation horizontasm s space ga lverticad san s gap ga outeln si r reactor structures

- - MeiJwement (IUNX) 200 —— CitntatK» (THEBMX C - 2D)

1200

I ISO- 1OOO Code. Ü 800 -COMMX-3D « -THERMX-2D 3 100 - (AVRLOC/kTest) ë 600- o Q. 3 Ï o >- 5 400 50

* 200

Test no 64 with ar at 11 bar and vessel coofcng

0 5 1 1 5 2 5 20 40 60 80 100 Time h Time In)

Posttes• Fl2 o1 t Temperature Calculatio Pebble , th d l nBe eo Fig 13 • Pretest Core Max. Temperature Calculation Test with Natural Convection In the Primary Circuit (Ref Interatom) o 1200- l\î /15/. However, natural convection effects in the pebble bed core itself are not signifi- ; «70-e E»'*""*"1 ILOCA) cant under LOCA conditions as the AYR tests with open and closed primary circuit 960 shut-off-device have shown. This analytical approach showed a much better agreement wit experimentae hth l results (Fig. 16). Calculated temperature withie sar uncere nth - \vr tainty of the measured temperatures of the components. However, azimuthal tempera- ^THEBUIX-C-ZD tur e hea th effect te head flowth an st o t ssink s e analysehavb o t e d wit a hmor e sophisticated model whic underways hi . ! 480- « e modellinTh e LOCth f go A test wit e HEATÏNG-3hth D cod s meanwhili e e being 240- LUNA-3e testeth r dfo D test facility wit hdetailea d measurement. Azimuthal tempera- ture effects can be produced in an annular pebble-bed by an electrical heating of the vessel wall in one of the six segments. First test results show the separate influence of 48 72 96 120 Time I»! ——— natural convection, thermal radiation head an , t conductio segmente th n i s (e.g. signifi- can locar fo t l heat bridge primare th n si y systedecae th ms ya heat remova shows ha l n Flo-14 • Posttest Core Max. Température Calculation (Ret KFA - ISfl) THTR)e foth r .

900 TOO

800 Exponent llOCA) Code. r 560 700 -THERMIX/HEATW6 L Analysa (HEATWG- 3D) P 600 with retinal convection effects 420 o | 500- Experiment (LOCA) Q 280 400

300-

200 20 40 60 80 100 120 Tlmo Ih) 24 «8 72 96 120 Time !hl

Fi{j.15 • Pretest Temperature Calculation of the Side Reflector (middle) . Posttes F>6 D1 t Température Calculatio Side th et no Reflec - (Ret. AVR/ORNL) tor (middle) (Ref ISR/ORNL/JAERI) . 4 CONCLUSIO IMPORTANCD N AN SAFETE TH R YE FO CONCEP T REFERENCES /!/ W. Kroger, J. Wolters, J. Mertens, R. Moormann, W. Rehm: "Preliminary Results of Thermodynamic safety analyses have been performed for a modular HTR unit to inves- an Updated Probabilistic Safety Analysis for the German HTR-Module", Proc. 2nd Post SMiRT Seminar on Small and Medium-Sized Nuclear Reactors, San Diego, tigate temperature loading safetf so y barriers during severest accidents with passive heat USA, Aug. 1989 removal dissipation into the reactor structures and ultimate heat sinks. In this context, /2/ H.H. Hennies, G. Kessler, J. Eibl: "Sicherheitsumschließungen in künftigen Reak- the temperature response behavio bees rha n analyse postulater dfo d accident sequences toren", Atomwirtschaft i 1992Ma ,238-24. ,S 7 associated with: anticipated transients without scram, loss-of-forced convection, /3/ W. Jahn, W. Rehm: "Thermodynamic Benchmark Calculations for a Small HTR loss-of-coolant, loss-of-vessel cooling, and primary system ruptures, regarding to Concept with USSR Design Data", KFA-ISR-IB-14/91, Aug. 1991 sensitivitie safetsand y margins followinThe . g statements seeimportanbe mto the for t /4/ G. Meister: "A Program Module Simulating a Gas-Heated Steam Generator with safety concept: Steam Condensatio Primare th n i y Flow Channel", Jul-Spez-150,1982 151 K.W. Childs: "HEATING-7.0 User's Manual", Oak Ridge National Laboratory, Oak Ridge, USA, ORNL-K/CSDTM, March 1990 • Thermodynamic methods have been proved to be suitable for accident analyses under loss-of-forced convectio d loss-of-coolanan n t simulation conditionse Th . /6/ E. Kicken, KFA-Jülich, Schwinges, GRS-Köln: "Description of the RALOC Code Family", Personal Communication, Feb. 1992 comparison of analytical and experimental results implicates comparatively high accuracies for calculating self-limiting heat dissipation processes. . /?BarthelsH / . Jahn . W Rehm, W , : "Compariso f Theoreticao n Experimentad an l l Studies of Afterheat Removal by Natural Convection Circulation from an HTR", Proc Inth .4t . Topical Meetin Nuclean go r Reactors, Thermal Hydraulics, Karlsruhe, • Sensitivity analyses suggest greater safety margins concerning failure limitf o s FRG, Oct. 1989 safety barriers because of conservatisms, especially for the first barrier of the fuel /8/ J. Cleveland, T. lyoku, W. Jahn, W. Rehm: "Thermodynamic Investigations of Loss- of-Coolant Accident r HTGRfo s SimulatioR s baseAY n o d n Experiments", 1992 element under LOCA condition. Reduced conservatism e safetth f yo s margins ASME/AIChE Nat. Heat Transfer Conference, San Diego, USA, Aug. 9-12,1992 might be incorporated into the design of a HTR module unit in the interest of . WoltersJ / . /9 al.et , : "Probabilistic Safety Analysi Assessmend san Urban o t n Sitinf go optimizin safete gth y concept. the Modular HTR for Process Heat Application", KFA-ISR-IB-3/90,1990 /10/ W. Jahn, W. Rehm: "Untersuchungen zur Corethermodynamik und zum Primärkreis- • This approach could be achieved by maximizing the power of the modular units verhalte HTR-Modus nde i schwerebe l n Kemaufheizstörfällen", KFA-ISR-IB-15/91, 1992 together with an assembling in one closed containment to reduce economic costs /ll . NabielekH / . KrügerK , . RehmW ,. SchenkW , . VerfondernK , : "Validatiof o n (e.g. 1000 MWt power plant with 4 units of 250 MWt instead of 5 units of Predictive Method r Thermafo s l Behavio d Fissioan r n Product Retention During 200 MWt). Such a containment design would include additional safety aspects for Accidents in Small Modular HTRs", Proc. 3rd Int. Seminar on Small- and Medium- unlikely events (e.g. large air/water ingress). A carbon brick layer above the Sized Nuclear Reactors, New Delhi, India, Aug. 1991 concrete foundation could act as passive precaution for the core integrity. /12 . AltesJ / . BarthelsH , . RehmW , : "Thermodynamic Investigation f Passivo s e Decay Heat Removal from HTR Cores and Component Behaviour", Nuclear Engineering and Design 121 (1990) 211-218, North-Holland r morFo e rigorous safety criteri f futurao e nuclear power plant discusn whics- i e har - /13/ H. Barthels, W. Jahn, W. Rehm: "Stand der Arbeiten zum Naturkonvektionsexperi- sio nwit- h energy generatio procesd nan s heat applicatio industrian i l areas, improved ment LUNA-HTR-Modul", KFA-ISR-AN, Aug. 199, safety enclosures should assure an additional potential of the safety barriers in severest /14 . KrügerK / . BergerfurthH , . BurgerS , . FohlP . ,Wimmers M , , J.C. Cleveland: accident scenarios with very unlikely events almost by a deterministic approach. "Preparation, Conduct d Experimentaan , Loss-of-CoolanR lAY Resulte th f o s t Accident Simulation Test", Nuclear Science and Engineering 107, 99-113,1991 However, a comparative safety philosophy for improved containments of the next reac- tor shoulgeneratioR HT consideree db d an n R wit internationan di hLW l cooperation /IS/ T. lyoku, W. Jahn, W. Rehm: "Analytical Investigations of the AYR Loss-of-Coolant Accident Simulation Test (LOCA)", Research Centre Jülich, KFA-ISR-IB-3/1992, 8 based on PSA work to fulfill the question: How safe is safe enough? March 1992 o PRESENTATIO DECAF NO Y HEAT . The relative position of the core above the steam generators does not allow the -ft. REMOVAL COMPUTER CODES starting up of a natural circulation by thermosyphon inside the reactor between the core (hot steasourcee th md generator)an s (cold source). USED FOR GAS COOLED REACTORS . The use of prestressed concrete for the vessel prevents any decay heat removal through the primary vessel towards outside. . CARVALLOG . DOBREMELLEM , . MEJANA , E . The use of a shutdown heat exchanger, located in the upper part of the reactor, was Départemen mécanique d t t technologieee , proved indispensable in order to remove the decay heat Commissaria l'énergià t e atomique, This exchanger induces a thermosyphon opposite to the normal circulation direction, and Centre d'études nucléaire Saclaye sd , the stud f thiyo s transitioimportann a prograo D t & d n le R abou] t m [3 t hande on instabilite th ,n o - y threshold characterization, loca whica leay o ldt ma h Gif-sur-Yvette, France inversion of the flow in some channels of the reactor, - on the other hand, the détermination of the influence of volume forces (Grashofs Abstract number) on the heat transfer coefficients, for small Reynolds's numbers. For this type of reactors, it is shown that the modelling of the decay heat removal implies • the development of a thermohydraulic model representing the whole reactor, existinFoe th r g French Magnox type reactors computeo tw , r codes have been developeo dt detailea . d modellin phenomene th f go a induce presence th y db f importaneo t local analyze the transient after reactor shutdown : volume forces (local inversio flowe th f ,n o etc.) . firse tTh one - , "GITA" representativs i . shore th f te o ter m evolution (less tha daysn2 ) Two codes, subjected to a validation program, have been developed by EDF and CEA : and it includes a refined representation of all the reactor components. - The "GfTA" code for short term evolutions (less than 2 days), secone Th - d one, "LOTE* bees ha ,n develope represeno dt lone th tg term evolution "LOTEe Th - " code (with more simplified modelling lonr )fo g term evolution days)2 > s( . (from 2 days to several months) with a simplified representation of the main components of the reactor. The current studie CEt sa higAn o h temperature reactor performee sar vera ydn i different One exampl f accideneo t simulatio presentens i existinr dfo g Magnox reactor context. CEA has carried out, since two years, evaluation studies about "innovating reactors'1; Moreover, as a part of the French program on the future reactors, an analysis of the modular the MHTGR [4] has been chosen, among others, for this evaluation. high temperature has been initiated. In their modular version, the high temperature reactors have ultimate decay heat removal 2D and 3D general flow and conduction codes are used for this analysis • systems through the primary vessel, which essentially implement the heat transfer by radiation - DELFINE is a 20 conduction code including a 1D thermosyphon model. It has been and conduction, convection excluded. use r decadfo y heat removal analysis. Building dispositions have even been adopte ordedn i limi convectioo rt e tth accidentan i l flo- D w3 TRI a cod Os i e includin radiationD g3 , conductio convectiod nan n heat transfer. situatio wated avoio an n(t rr f dai ingrescodeo e GITs d LOTsa us s Aan e effects Eth d an ) It is used for detailed thermal analysis during accidental conditions. was therefore imprope represeno rt t suc situationha . As this evaluation effort could not justify new developments, the general codes, developed by CE nuclear Afo r application, have been used: - the DELFINE code, 2D code for conduction and radiation, in finite elements, 1. INTRODUCTION TRIO-Ee th - F code, genera) thermohydrauli codeD c 3 r conductio,fo radiationd nan .

The developmen f graphito t s reactorga e f 1950o s d startes en e n i Francdth y b e Eight reactors have been built; only one of them is still in operation (BUGEY 1) [1]. 2. CODES USED FOR GRAPHITE GAS REACTORS Such a type of reactor is characterized • - by a natural uranium metal fuel, magnesium-zirconium cladded, whose operating 2.1. General temperature limit respectivele sar y 650° 515°Cd Can , - by a graphite moderator giving to the reactor a large thermal inertia, Three different situation r decasfo y heat remova takee ar l n into accoun] [2 t - by a working fluid which is generally a pressurized gas (CO2 at 25 or 40 bars). evacuation by forced convection driven by the circulators towards the steam generators, Different architectures have been proposed for these reactors (loop reactors or integrated evacuatio naturay nb l circulation toward shutdowe sth n heat exchangers. reactors) thin i st presentationbu , calleo ,s onle dyth "integrated" concept wilconsiderede b l , . evacuation (in filtered open circuit) by a forced air circulation, for long shutdown regime. for instance the "ST LAURENT A" one. The reactor vessel is in prestressed concrete, it contains the whole components of the primary circuit (i.e. core, steam generators circulators)d an , . "GITAe 2.2Th . " code The circulation insid core eth s descending ei gase th , d circulatorse driveth an , y nb , passes throug exchangere hth s before goin towardp gu annulae core th sy th e b r space The GfTA code has been developed by EDF to calculate the short term evolution in shutdown transients of the reactor Concernin e decath g y heat removal, some fundamental characteristic f thio s s type reactoe Th - discretizes ri d accordin monodimensionaa go t l diagram show figurn no e1 of reactor have to be noted [2] - The reactor core is subject to a more detailed modelling; it is divided into 5 radial REACTOR SAINT-LAURENT DESCRIPTION OF THIS REACTOR FOR THE COMPUTER CODES GITLOTD AAN E

"Cuir. -rr--3 •-.:,-; ifa

UPPER PLENUM

till ! CORE | j SHUTDONt1 radia : GITA:discretizatiol n HEAT HEXCHAFI6ER LOTE o discretizatio:n n + BY-PASS axial : GITA: discretization LOTE: discretization ' i •1i i:i i:i LOWER PLENUM

SUPPORT PLATE ANNULAR SPACE

STEAM GENERATOR t

C02 flow during CEILING normal rtinn ing

1 BLOWERS - 1

o en FIG 1. Representation of Saint-Laurent A2 reactor (computer codes GITA and LOTE). zones, each radial zone being represented by a mean channel of the zone, each mean A detailed presentation of "LOTE" code, whose simplified modelling leads to strong o channel being constituted by : reduction f computatioso n costs bees ,ha n don [5]en i . 0> . the fuel. . the cladding, 2.4. Example . the graphite sleeve, . the moderator. The presented example concerns an hypothetical accident scenario on SLA reactor. discretize finitD 2 ey d b difference s (radia longitudinal)d an l . Initially e reactos nominath it , f o r operate% l power5 8 t a s ,pressurO witC ha f o e The modelling Integrate whole sth programseD result& R f sr instanco ,fo influence eth e 28.9 bars and a CO2 output temperature of 410°C for an input temperature of 240*C. of volume force hean so t transfer coefficients (figur. e2) - The sequence is initiated by a complete loss of the circulators, which stop in 150 sees. linean no f r o differentia t se e Th - l equation solves i predictiva y db ecorrectiv- e method, - The shutdown heat exchangers cannot be put in service. with the usual criteria of time step (Couranfs and Fourier's meshing numbers) used for watee Th -r supplsteae th f myo generator seconds0 6 loss i t a t . explicit method. scrae Th m- occur seconds7 s afte sequence rth e initiation. The scenario is fully computed over 10 hours with the GITA code. The results are shown 2.3. The "LOTE" code on figure 3.

It has been developed by CEA to describe the reactor evolution and decay heat removal . Durin firse gth t hour, some flow oscillation recordee s reactorar e shutdowe th th dn i s a , n for long term scenarios (post accidental situation, maintenance, final unloading of the reactor, heat exchanger lack does not allow a thermosyphon to be correctly established. etc.). thermosyphoe Th . n flow rate remain srecirculatioa lowd an , observes ni core th e n di The particular nature of gas reactors (low density of residua) power, no risk of phase peripheral channels, where both temperatures and residual power are lower. change of the coolant fluid, large thermal inertia) leads in fact to slow and monotonie . The low thermosyphon flow rate causes a high thermal stratification inside the reactor : temperature evolutions, without important discontinuities : that allows simplified modellings from 230°C in the lower part up to 650°C In the upper part. of some reactor parts. reactoe th o t r thermae Du . l inertia affordable ,th e temperatur claddin e f 650°eo th n Co g The justificatio f thino s simplified modelling relie validationn so s carrieGITy b At dcodeou . is only reached after 9:30 hours. That allows to implement the solutions in order to re- It only involves some reactor components, and the core will be represented by only one radial establish the failing functions. zone of mean power, whose components (uranium, dad, sleeve and brick) are supposed to be at same temperature t allo.no wThacorrecy a tma t representatio f locano l flow inversions. 3. CODES USED FOR MHTGR EVALUATION 3.1. General

too The modular high temperature reactor, as well as other innovative concepts, is the subject of evaluation studies by CEA. A particular attention is focussed on : 90 decae th y- heat removal (residual power evacuation). - the air or water Ingress. which are the determining points for the reactor, as its promoters have founded the complete 70 safety philosoph integrite th n yo y keepin firse th t f gbarriero . These reactors (MHTGR or Modular HTR) use a decay heat removal by conduction - radiation throug e primarhth y vessel toward externan a s l system (ttiermosypho r wateno r circulation). In the choice of means to implement for evaluation studies, three elements have been determining : . The decay heat removal system needs neither system modelling nor complex functionalities. evaluatioe Th . n work follows budget contingencies, which prohibi developmeny an t t of new codes. importann A . t effor mads CEAy wa te b , durin lase gth t decade givo t ,genera o et l codes an operatin modellind gan g versatility allowing the mcoveo t wida r e application field. Therefore wits i t hi , "DELFINE "TRIOd an " " general purpose codes thaevaluatioe th t f no modular high temperature reactor (the american MHTGR [4]) was tackled.

6 7 g 9 10 15 2 3.2. The "DELFINE" code IRt/Ra") The DELFINE code is a 2D conduction - radiation code, developed in the last 70s. It is part of the first generatio f finitno e element codes develope [6] A mais It . CE n y characteristicdb : e sar FIG . Effec2 . f buoyanco t y force clan so d heat transfer. - 2D conduction with variable properties and phase change. EVOLUTIO MAXIMUF NO M CLAD TEMPERATURE (C02) EVOLUTIO TEMPERATURF NO COOLANF EO T FLUID (CO2) 650°C

600.

500.

400.1 0. 1. 2. 3. U. 5. 6. 7. 8. 9. 10. 0. 1. 2. 3. A. 5. 6. 7. 8. 9. 10. Time (hours)

FIG. 3. GITA code results — Complete failure of the cooling systems.

- face to face or closed cavity radiation (with external computation of form factors). watee th r- reactors, fast neutron reactors, high temperature reactors. - internal coupling with one or several flows represented by an unidirectional circulation The fission environmente ,th r industryca e alse ,th o,ar application domain codee th r .sfo (forced convection circulation or natural circulation, either open or closed). Since some years, the code was the subject of a vast validation program, particularly in the natural convection, mixed convection, stratification domains. "TRIO-EF-code 3.3Th . e 3.4. Examples

thermohydrauliD 3 a s i t I thermad can l code, whose development starte aboun di t 1980 Three application example presentee sar d: integratet I . uniqu9] [7n , a : 8 , t sn i e se . the decay heat removal of MHTGR (DELFINE code), . the Navier - Stokes 30 equations in laminar and turbulent flows, with closure models evaluatio floe r ai th w e . passinth f no g throug core h th f edepressurizatioi n occuro t e sdu e typesK- .d an L oK- f a failure of the hot duct (DELFINE code), . the 3D conduction with phase change, . the flow in the upper plenum during the previous accident (TRIO code). radiatioD 3 e th n wit. h integrated computatio viee wth f nfactoro s with hidden parts. The code uses a method of finite element type. It integrates all the possibilities offered by 3.4.1 Decay heat removal (Residual power evacuation) "object" oriented structures arid takes advantag data f eao language allowin engineee gth o t r simply program the partial derivative equation system and the chosen algorithm (for example to We remind the MHTGR takes advantage of a fully passive ultimate removal of decay heat : realize magnetism flo- chemistrr wo yflo- w couplings). radiationy b - , conduction naturad ,an l convection insid vessele eth , Of course, such characteristics offer wide range of applications for this code which, in the naturay b - l circulatio r evacuatioai e th n i n circuit (RCCS). nuclear domain, is used for the whole fuel cycle and the reactors. Trie used model is a bi-dimensional axisymmetrical representation of the core zone, up to Lef s mention, for instance : the boundary conditio f naturano l convectio f radiatioo d nan n toward RCCe sth S system. - the laser enrichment (AVLIS process), Block e representear s y equivalenb d t circular ringsd theian ,r equivalent thermal o - the transport and the geochemistry of radionuclides. in the underground storages of conductivity is determined by a conduction calculation on one cell, taking into account the -•4 radioactive wastes. cooling holes, the moderator and the fuel. o spacess ga e ,th between I n blocks r betweeo , n corprimard ean y vessel heae th , t transfer EVOLUTIO MAXIMAE TH F NO L FUEL TEMPERATURE VERSUS TIME 00 is considered to be done by conduction, natural convection (according to Grashof's number) and radiation. A temporal evolution of the maximum temperature, and an isotherm cartography at 80 hours, are given in figure 4. Numerous sensitivity studies have been carried out. Some result presentee s followinar e th dn i g table: 1800

Tref T max Studied parameter« Reached after : °C °C 1600 Graphite conductivity 80 hours decrease% 0 2 y db 1663 1780

Residual power 1663 1888 80 hours _ UOO increase% 0 2 y db o

Cooling circuit 80 hours decrease% 0 2 y db 1663 1668 1200 Air (instead of helium) 1663 1709 70 hours

Initial temperatures 50 hours increase% 20 y db 1663 1746 1000

We note, particularly, the important role played by the precise knowledge of the residual power (decay heat). 800 t ductho 3.4. a ingressr failureo 2t Ai e du 0 i.0 80 120 160 200

Th t duce ho failur t a provoke f eo instantaneoun a s s depressurizatio reactore th f no , Time (hours) with an air ingress which can cause the oxidation of the core graphite. The model shown on figure 5 uses the coupling between tfiermosyphon and conduction of DELFINE code. The whole core channels are represented by an equivalent channel with adapted characteristics (for pressure drop head san t exchanges). The different time constants between natural convection and conduction lead to process : DECAY HEAT REMOVAL IN THE MHTGR ISOTHERMS AT TIME = 80 HOURS - the decay heat removal in transient regime, - the thermosyphon by successive permanent states. The results (figur highligh) e5 : t . the air flow passing through the core (in such a configuration) is limited by the important pressure core e dro oxidationth e f p;th o thin ,i s scenario, willowe b l . . the thermal power extracted by the thermosyphon is negligible compared to decay heat. Thes studieo etw s roughly confir resulte mth MHTGe s th alread n o R t behaviouygo r [4]. 3.4.3 Flow In the upper plenum

characteristice Onth f eo decae th f syo heat remova thas i l t i leadt higo st he riseth f so core temperature (600 to 800°C) and of internal structures. This characteristics results fro ceramie mth c fuel choice, working norman i , l operation, ver fror y fa technologica ms it l limits. To be exhaustive, it is necessary to examine carefully if these temperature rises are not detrimental to correct resistance of metallic structures of the reactor (primary vessel, control rods, thermal insulation). Such analysise f modellingo s e requirus e eth s more detailed tha onee nth s previously FIG . Evolutio4 . maximae th f no l fuel temperature versus timdecad ean y heat removan i l exampln useda d an ,f thie o s typ f computatioeo bees nha n carrie t wit TRIde ou h th O code. the MHTGR isotherm hourstimt 0 sa 8 e= . AIR MASS FLOW THROUGH THE CORE t I deal suppe e witth floe hth rn wi plenureactore th f mo , followin e previouslgth y processed air ingress J/U The meshing represent uppee th s boundare reactore r th th par f d o t an ,y conditions ^ (gas velocity, temperatures) come fro previoue mth s computation We note (figur tha} volume 6 th t e forces creat convectioea nuppee looth n pi r plenum \ uppee th n ro drivin shrou impnngemens n thia r ga gt o st dFo example ho e flowinr th ai f e o t ,th g 350 \ EXAMPLE OF VELOCITY RELD ~c \ IN THE UPPER PLENUM 2 350 t> \ Ta \ | 340

>fl \ oU> I \ * 330 ^ \ \ 320 > 1

i

310 0 6 0 5 0 4 0 3 0 2 0 1 0 Time (hours) EXAMPL ISOTHERMF EO S INGRESR MHTGE AI TH N RSI it 300°C

SOO-C

600°C

O CO r ingres MHTGAi FIe th G5 n si R FIG 6 through the core seems to have a minor effect on the flow patterns inside the upper plenum [7] MAGNAUD, J.P.. GRAND, D, VILLARD. M, CHEISSOUX, J.L.. HOFFMANN. A. "Recent This computation integrates Development numericae th n si l Predictio f Thermano l Hydraulics". International Meeting - the radiation between core and thermal protection, on Advances in Reactor Physics, Mathematics, and Computation. Paris, France April - the conduction in the different media, 27th - 30th, 1987 - the convection in the upper plenum. Such calculations, presented her examples ea alloy wfinema ,a r analysi f thermaso d an l [8] MAGNAUD , GOLDSTEINJP . , "ThS. . e Finite Element Versio f TRIo n O Codeh 7t " mechanical behaviour of the internal components of the reactor International Conference on Finite Element methods in Flow Problems. Huntswlle, USA April 3rd - 7th. 1989.

CONCLUSIONS [9] COULON, N , "Computation of 3D Form Factors in Complex Environments" ANS meeting on Advances in Nuclear Engineering Computation and Radiation Shielding. Santa Fe, USA Important advances have been realized during the last 25 years regarding the computation April 9th - 13th, 1989. codes and the thermohydraulic general codes, which were at their early development when the graphit reactors ega s were launched. They havreacw eno maturith a y e stag th wels r e(a fo l modellinease th f use) eo r fo , s gthaa t allow engineee analysith s y shi b r theie rfo srus works. For modular high temperature reactors, this use is all the more facilitated as the dominant mechanism f decaso y heat removal (conductio radiationd nan ease modelizear )o yt , provided thaengineee tth goo a s dr ha knowledg f inpueo usee t datth dr acodesfo .

ACKNOWLEDGEMENTS

The authors would lik thaneo t k EO r approvaFfo publiso t l h informatio GITe th An n o code , whic firss htwa develope EOF y thedd b nan , implemented jointl EO y CEAyd b Fan .

REFERENCES

[1] BASTIEN , "Twenty-ninD. , e year f Frencso h experienc operatinn ei cooles gga d reactors". Technical Committee meetin Designn go , Requirements, Operatio Maintenancd an n e Cooles oGa f d Reactors Diegon SeptembeA Sa . US , r 21s - 23rdt , 1988.

[2] LHEUREUX , AGUILERAR. . , "SafetA , y Criteri d Provision an e aEvacuatio th r fo s n of Residual Heat from Graphite Gas Cooled Reactors*, to be presented at IAEA Specialists meetin "Decan go y Heat Remova Head an lt Transfer under Norma Accidend an l t Conditions Cooles inGa d Reactors" Julien. 10th- Julh y6t , 1992.

[3] GITTNER, G., BABY, J.P.. "Evacuation de puissance résiduelle par thermosiphon dans les réacteurs à gaz". EDF-CEA conference cycles : "Natural and Mixed Convection and their Industrial Applications" Ermenonville, France. September 13th - 18th, 1971.

[4] Utility / User Requirements for an assessment of the modular high temperature gas cooled reactor. Report GCRA 91-002 (1991).

(5) BOUTARD. F. CARVALLO, G, CHEVALIER, G, "French experience in Thermohydraulics calculations of Gas Cooled Reactors" Specialists meeting on Uncertainties in Physics Calculations for Gas Cooled Reactors. Villmgen May 9th -11 th, 1990

[6] GOLDSTEIN , JOLYS. ,JUIGNET , J , , "SomN , e numerical algorithm applicationd an s s of the DELFINE computer program, numerical methods in thermal problems" Proceedings Firse oth f t international Conference Swansea, Wales. 6th- Julyd , 1972n , 9 MODELIN ANALYSID GAN HEAF SO T TRANSFER FRO MHTGE MTH R CORE THROUGHA STEEL REACTOR VESSEE TH O LT REACTOR CAVITY COOLING SYSTEM

D.A. DILLING, J.M. BERKOE, S.K. CHOSE Bechtel National, Inc., Franciscon Sa , California T.D. DUNN General Atomics, San Diego, California S.A. CASPERSSON ABB Combustion Engineering Nuclear Power, Windsor, Connecticut United State Americf so a

Abstract The commercial Modular High Temperature Gas-Cooled metr plan yca y significant decae roleth n syi heat removal Reactor (MHTGR) achieves improved reactor safety process and strongly influence the reactor vessel tempera- performance and reliability by utilizing an integrated ture profile sequence of completely passive thermal storage and heat transfer mechanisms to reject decay heat in trie event that all its active cooling systems fail to operate During such INTRODUCTION events, the initial heatup transient in the core u followed The commercial MHTG advancen a Rs i d reactos i d ran b yquasi-steada y slate cooldown process which uninf ,i - bong developed to possess improved safety, reliability, terrupted continun ,ca severar efo l day buoyancyA s - and economic performanc desige Th e n approaco ht dnven natural convection cooling system callee dth safety of the MHTGR is to provide a plant that limits the RCCS facilitate continuoue sth s heat remova circulaty lb - potentia off-normar lfo l events hav thay etma adverse ambieng in througr tai reactoe hth r cavity, whers i t ei consequences Thi achieves si limitiny db g component heate thed dan n exhauste outside th do t e environment temperatures, pressures, stresses, and chemical reactions peae Th k therma RCCe l loath n Sdo occurs approximate- thaplan experience y tth tma e during these evente Th s ly at the lime that the vessel reaches ils highest tempera- MHTGR design possesses several characteristics which ture. To confirm the adequacy of the RCCS design, enable this capability, sucs ha detailed analytical models were develope simulato dt e eth • Individually coated, heat-resistant fuel "particles" decay heat removal process and predict the maximum vessel temperature at this condition. Due to the integrated • Inert helium gas used for heat transport medium nature of the MHTGR thermal characteristics and the corw eLo powe• r density complexit RCCe th f ySo configuratio reactoe th nm r cavity three-dimensiona,a l compute0 r35 modee th f lo • Solid-graphite moderator for high thermal capacitance MWt plant design which extends fro reactoe mth r coreo t • Uninsulated reactor vessel to reject stored decay heat the outside environment was created to simulate the entire decay heat removal process • Underground installatio improver nfo d seismic performance The results of the analyses confirm that the MHTGR core temperatures remain below their prescribed design limits, Each reactor module is housed m a massive below-grade and tha RCCe tth S adequately maintains peak vessel tem- concrete confinement Figur showe1 physicae sth - ar l peratures within acceptable limits In providing increased rangemen singla f to e reactor modul reactoe Th e r vessel understandin passive th f go e heat transfer mechanisms ansteae dth m generato locate e adjaceno rar tw n di t cavi- which prevail during MHTGR accident conditions, the ties of the underground silo, and are connected by a cross Figure 1 analysis results show thacombinee th t d effect thermaf o s l duct vessel Power is generated in the prismatic graphite radiation, free convection non-uniford ,an m RCCS geo- fuel blocks of the active core These fuel blocks arc ar- Schemati Modulae th f co r HTGR ranged in an annular cylinder surrounded in the center As shown in Figure 3, the RCCS cooling panels are actu Analytical models have been develope evaluatdo t e eth riser tube arrangement), necessitates thaanalyticae th t l an sidel dal graphity sb e reflector blocks that fore mth ally comprised of individual nser tubes which surround MHTGR thermal performance during conduction cool- models use simulato dt performance eth MHTGe th f eo R ro inner bottod , outeran mp ,lo reflector s 0 Ther66 e ear the vessel alon perimetee n gth no cavit e e th f Th yro down event integratee Th s d naturoverale th f eo l heat must be highly detailed in certain areas and more simpli- fuel block reference th n si e MHTGR acuve core arranged uniform nser tube layout (i e, "gaps" between banks of transfer process requires comprehensivthay an t e analy- fied in other areas Use of this approach dictates that mul- column6 in6 axia0 1 f ls o block s that provid ecora e tubes) is necessitated by the presence of lateral vessel tical model extend from the core to the outside environ- tiple analyses be performed to obtain a complete and power density of 5 9 W/cm3 The most recent MHTGR supports cavit e locateth f yo dp Deca neato e ryth heas ti ment complexite Th . systee th f ymo geometry, which thorough simulation of the MHTGR thermal performance design incorporates a 450 MWt reactor which includes an transferre convectioy db n insid risere enaturath a o st l cir- includes many nonsymmetnc characteristics (e g, RCCS during conduction cooldown. 84-column core wit hpowea r densit OW/cm6 f yo e Th 3 culation flow of ambient air which enters at the bottom of fuel blocks and the graphite reflectors are located in a reactoe th r cavity after exitin downcomee gth r formed metallic core barrel whic turn hi encloses ni e th n di between the cavity and the reactor silo concrete wall A reactor vessel reflective pane Microthermf lo ™ insulatio locatens i d The steam generator vessel contains the steam generator along the cavity side surface of the downcomer to reduce regenerative heaung of the incoming an The decay heat maie th nd circulatoan heae th tf transporro t system earnes i d outsidawae th o yt e environment througe hth Under normal conditions heliuprovides i e w th m no y db RCCS outlet "chimney ' stac k mam circulato pumpt i s a r s cold helium fro steae mth m generator outle forced tan througt si reactoe hth r whert ei is heated before returning to the steam generator The hot helium flows downward ove helicalle rth y coiled tubesm PERFORMANCE VALIDATION the once-through uphill boiling steam generator The radionuctide retention capability of the coated sphen Natural In os maie eth n circulate stear *o m generato availt no s -ri cal fuel particles is one of the .important elements in Circulation able for service, the reactor is scrammed, and a shutdown passive MHTGR safety Testin millionf go paniclef so s of Outside Air cooling system, located at the bottom of the reactor ves has shown that rapid fission product release from the sei, is used for core decay heat removal Hoi helium particles is a function of température, time at temperature, fro bottoe mcore th th pumpes emi f shutdowe o th y db n irradiation temperature, fluence bumud silicoe an , Th p n circulator, pas watee tth r cooled shutdown hea- tex carbide coated fuel panicles will retain virtuall radiol yal - changer The cooled helium then returns to the core inkL nuclides up to approximately 1,600'C for many days of A scram-miuated event involving the failure of both the irradiation temperature, fluence, and bumup expected in ma standbd man y "active" core cooling system calles si d the MHTGR Therefore. 1.600'C has been selected as a a "conduction cooldown" If such an event is initiated by "rul f thumb*eo * peak fuel temperature designl limial r fo t losa offsitf so e power followe turbina y db e tupe ,th accidents reactor remain pressurizea n si d state. If suc evenn ha s ti Similarly, there are design temperature limits on all the initiated by a primary system leak and ensuing helium other component reactoe exampler th f sFo o r limite ,th s depletion reactoe ,th r depressunzes- of the outer control rods, core barrel and the reactor vessel Durin gconductioa n cooldown, coolin MHTGe th f go R 1,175-C.760*e ar 4ÄTCd Can . respectively is achieved via passive heat transfer mechanisms (i e, A comprehensive performance analysi RCCe th f sSs o i thermal radiation, conduction naturad ,an l convectionA ) essentia confiro lt m thaMHTGe tth R design satisfiee sth natural circulation air flow system, called the Reactor component temperature safely criteria described above Cavity Cooling System (RCCS). has been designed to Previously performed scoping calculations based on the facilitate this function. The RCCS is a continuously reference 350 MWt plant design have shown that depres- operational, safety-related system which provides ade- sunzed conduction cooldown events highesresule th n ti t quate decay heat removal without dependenc acuvn eo e core and vessel temperatures and thus represent the most component passive Th s e RCCS design incorporatesa extreme challeng RCC e evene th o Th eSt t designates da substantial performance margin, robust structurea d ,an SRDC-11 was used as a representative event to study the high degre redundancf eo achievo yt e extremely high effects of depressunzed conduction cooldown reliability and thus ensure reactor safety The RCCS is In SRDC-11, a small primary coolant leak occurs at the dependent no operaton to r performancs actioit d nan s ei reactoe th f o r p vesselo l which will depressunz reae eth c insensitiv operatoo et r error r syste to hour4 2 mm s After about hal houn fa r follow Figure 2 shows schematically the means by which heat is leake instarreactoe e th gth f th , to prètnppes ri w slo n do removed from the MHTGR during a conduction cool sure signal using the outer control rods In addition it is largs it o et dowthermae Du n l capacitance reactoe ,th r assumed that neither the heal transport system nor the core heats up very slowly following scram over a penod shutdown cooling system is available after reactor trip to of many hours As the active core temperature increases, provide forced circulation coolin reactoe e th go Th t r heat is conducted through the graphite reflector, primarily helium purification syste malss i o t assumeno e b o dt radiae inth l direction Hea transferres li d acros core sth e working so that no pumpdown of the reactor system is annulus fro outside mcore th th e f ebarrevesseo e th o lt l considered Under these condition decae sth y heat from wall via thermal radiation and natural convection. Heat is core th removes ei d predominantl conductioy yb d nan Figure 2 rejected from the uninsulated reactor vessel, primarily via radiatio reactoe th a nvi r vesseRCCe naturae th o Th t lS l thermal radiation, to an array of "cooling panels" situated RCCe th circulation i S r removeai e th heaf e no s th l from MHTGR Conduction Cooldown Heat Transfer around the perimeter of the reactor cavity the reactor cavity to the atmosphere The outer radial, lop and bottom boundary conditions 4) as heat is rejected to the atmosphere by the RCCS The were fixe constant da t temperature 27'f so C Initial reactor vessel temperature peak approximatelt sa 0 y10 temperature MHTCe th r sfo R cor othed ean r metallic hours (Fig 4) However the RCCS heat removal rale components were obtained fro msteady-stata e full peaks later at 140 hours (Fig 5), as it is dependent on the power analysis temperatures experience reactoe th y db r vesse whola s la e As shown in Figure 4, the thermal transient experienced and not on a localized peak Beyond 140 hours there is a by the active core and the reactor vessel for the 450 MWi gradual cooldown of the reactor system MHTGR dunng SRDC-1 sloa s 1i w increase over many Figur eshow6 fractioe active sth th f neo core that exceeds hour peao st k temperature followe graduaa y db l cool 1,000'. 1400" 1,400d .an * ovetransien e time th rth f eo t down This slow thermal response over lime is very A large fractio active th f eno cor abovs ei e 1,000'r Cfo REACTOR similar to the thermal response of the reference 350 MWi ovehours0 r20 , wherea acuvse onlth f yesmalo a n lpa MHTGR Other components show a similar thermal core exceed 400's1 C Figur showe4 s thamaximue tth m CAVITY response The active core maximum and active core core température reaches a peak of 1,500'C approximate- average temperatures peak at 60 and 70 hours, respective hour0 ly6 s afte slae SRDC-1f th rn o 1 This peak tern ly whil maximue eth m reactor vessel temperature peaks peratur experiences ei only db y three percen acuve th f cto hour0 agenera10 n tI s furthee th l r awa ycomponena s ti core for less than 20 hours No fuel damage is expected fro core mth e radiall latee timth yn ri experienceeu s sit becaus peae eth k temperatures (1.500'C wels )i l below COKEANNUtUS peak temperature. the 1,600"C design limit and the lime at peak temperature •• % *V* %• is idauvely short * --^ %v>5" %^'s Most of the decay heat is removed in the radial direction to the ROCS with very little heat being removed axially The core decay heat generatio RCCd nan S heat removal RCCS PERFORMANCE ANALYSIS rate showe sar Figurn i e5 Initially core ,th e heatp su slowldecae th s yya heat generatio highea t a s ni r rate than Result previoue th f so s analysis show tha vessee tth l the beat removal rate The stow increase in temperatures temperature peaks closer to its design limit than the core large th eo t hea e tdu capacits i graphite th e f yth o em or other key components The primary focus of the reactor The RCCS heat removal rale increases slowly RCCS performance analysi creato t s ethreesa wa - with (he gradual increase in the reactor vessel tempera- dimensional modeenure th f elo decay heat removal tures. Around 70 hours the active core heal removal rate proces incorporato st e non-symmetric effecte th n si exceed decae sth y heat generation rate frod m,an then no syste ordemn i moro rt e accurately predic vessee tth l tem- the average active core temperature decreases slowly (Fig perature vere Baseth y n "quasi-steadyslo, d o e wi ( -

FigureS 350 MWt MHTGR Reactor Cavity - Plan Vie t wVessea l Belt-Line

Nodal power points are defined as lying midway between TRANSIENT THERMAL ANALYSIS the bounding gnd lines of these elements These points thermae Th l transient experienced during s SRDwa 1 C1 each represent a nodal volume for which a central tempera analyzed usin ggeneraa l purpose two-dimensional heat turc is calculated transfer code called TAC2 uses i calculatin r t DI dfo g steady-stale and transient temperatures in two-dimcn The heat transfer was modeled only by conduction and sional problem finite th ey s b difference method radiation withi reactoe nth r vessel This providen sco servauve peak fuel temperature thers sa soms ei e heat A geometric modeentire th f elo reactor vesse cavitd lan y redistributio beginnine th e t transienne a th th o t f go e tdu was used to perform the thermal analysis The model natural circulatio remainine (h f no g pressurized heliumn i encompasses the acuve core, the inner outer, top, and the system Heat is transferred by thermal radiauon and 250 bottom graphite reflector graphite th ; e core support floor, conduction space acrosp ga e ssth separaun core r gth esu the core support plates core ,th e barrel insulatee ,th d upper faces and the metal support structures and shrouds Heat plenum shroud, the reactor vessel, radiation shielding ma is transferred predominantl thermay yb l radiation across lerial abov belod ereactoan e wth r concretvessele th d ,an e the gas spaces between the core barrel and the reactor ves behind the air-cooled reactor cavity cooling system pa frod reactoe an mth i se r vessereactoe th o t l r cavity cool Figure 4 nels The MHTGR is described in a cylindrical coordi ing panels Radiation is calculated one-dimensionally in aate system by orthagonal lines of constant coordinate this model A convecuve flow of air through the cooling Active Core and Vessel Temperature During SRDC-11 called gnd lines The material boundaries derme annular panels is calculated which removes most of the heat regions in which temperature nodal points are located fro panele mth s MWt0 45 Colume fo4 8 ,th r n Core MHTGR state") natur conductioe th f eo n cooldown aroune dth core powe ranalysiMWte S use l th dn s i , swa whic r hfo tune at which the peak vessel temperature occurs, a the 350 MWt MHTGR corresponds to the approximate steady-state condition was assumed for this analysis whee hea RCCe bm n th t e loaSth t heada t removal curve COMMIX-e Th computeR IA r code (herein referres a o dt "crosses" the decay heat curve (i.e , near the time of peak I - "COMMIX") develope Argonnt da c National Laboratory vessel temperature) fo depressunzera d conduction cool- was used to perform the simulation (Rcf 1) The dow lemperaambien modelee C s 3' Th nwa 3 r 4 tai s d -a COMMIX calculation scheme, based on a finite-dif- tureand 101,32 pressura 5P RCCe ih t ea S ouüete Th . I - ference (i t., finite-volume) methodology, utilizes a three- RCCS is designed for this conservative condition. dimensional mesh of computational fluid cells, either m a The cooling panels, vessel wall and core barrel were 0*caI yHM RCCS HMI femoral cartesian or cylindrical coordinate gnd system, as assumed to have surface errussiviiies of 0 8 This building blocks that compris physicae eth l system model represent sfairla y conservative valu carbor efo n steel (ie..calculanonal domain) Cell location definec sar y db exposed to ambient conditions In reality, the RCCS nser the spatial volumes create intersectine th y db planed ggn s tube geometry (see Fig. 3) creates an errussivity en- - 2 Mass and energy transfer rates are calculated at each cell hancement due 10 its ability to trap reflected energy in UK interface, enabling the internal generation of temperature, narrow gaps between the tubes. The "effective" émis pressur flod ewan velocity fields for each fluid cell sivit thir yfo s typ arrangemenf eo t would actualle yb —T- COMMIX internally computes inlet and outlet flow above 0 9 The emissivtty of the reflective downcomer 50 IOO ISO 200 250 conditions by balancing the mass conservation equations insulation (facing into the cavity) was assumed to be 0 1 Tim* (Hour) at each specified location. During RCCS operation, air flow through the nser tubes effectivelo T y model three-dimensional radiation heat fulls i y developed along tube mosth ef tlengto s i d han Figure5 transfer in the reactor cavity (e.g.. between the vessel and well intturbulene oth t regim . Reynolde i e( s number, RCCS panel! insidd vesse)an e . betweeeg th e l( n core -10,000) Heat transfer rates were calculated by barrel and vessel wall), accurate surface-to-surface view COMMIX based on a forced convection correlation of Decay Heat Generation Rat RCCd ean S factors were neede inpuo dt COMMIo tt X Accordingly, the formK*Re= u N . O 8*Pr033. wher typicalls i eK y Heat Removal During SRDC-11 for the the SSPTA (Simplified Space Payload Thermal Analy- 0023 (standard heat transfer texts) For conservatism, K zer) computer program (Re . obtainef2) d from NASA was reduced by 20% to account for potential mixed 450 MWt, 84 Column Core MHTGR Goddard Space Flight Center, was used to compute the convection effects (which could degrade heat transfer) view factors. nea tube rth e entrance The computational model was derived from the physical The calculated vessel temperature profiles for the depres- system characteristics of the 350 MWt MHTGR refer- sunzed conduction cooldown (DPCC shows )i Figurn i e ence plant design. To limit the high cost of computer 8 The RCCS cootmg panel temperature profile is shown runs due to model size and complexity, a single quadrant in valueFigure Th se9 use generato dt e these curves reactoe oth f r cavit analyzes ywa d This region corres- correspond to the peak temperature at each axial plane in pondsketce th o sht show Figurn i e3 Symmetr y condi- the model The maximum temperature is around 455'C tions (i e., zero heat and mass flow) were assumed along The RCCS outlet air température in both cases is approxi the open sides of the quadrant. mately I75"C, with a total system flow rate of about 11 kg/se inside Th ce surface-to-air heat transfer coefficient, The calculational domain was constructed using a 3-D car- which is a function of the nser air velocity and air proper lesian gnd to closely simulate the rectangular geometry of tiesapproximatels ,wa W/m2-8 y1 K the RCCS panels and cavity wall In the X-Y plane, the model extends downfroe corcentee e th m th th eo t f - ro In a DPCC, convection of low pressure helium has negli- comer The cavity wall was not included in the model gible effect on the reactor internal heat transfer Conduc- sinc remaint ei constana t sa t temperature during RCCS coree radiatiod th uo )n an n(i n (acros core sth e annulus), operation due to the continuous flow of incoming whic primarile har y directed radially, dominat heae eth t ambient air. In the Z-direclion. the model extends from transfer process. For this reason the temperature profile the cavity floor (RCCS bottom plenum) to the cavity is peaked at the location of highest power generation and ceiling (RCCS upper cold plenum) extendd an , s witha drops off significantly at elevations above and below the reduced cross-section to include the RCCS inlet/outlet active core. duct outled san t chimne axian A y lmode e vieth f ws o i l Figure 10 shows the DPCC vessel temperature along a show Figurn i e7 cross-section of the vessel wall at the highest temperature 0,0 RCCe Th S riser tubes were grouped into five "lumped" (i e , peak core power) plane The azimuthal temperature 100 150 200 250 flow channels arranged aroun perimetee dth cavue th f ryo variation fairle sar y significan t thia t s location, wite hth i nnon-unifora m manne orden ri modeo rt effecte th l f so maximum gradient across a 15* section of approximately panel geometry on cavuy heat transfer Accordingly, the 28'C Such temperature gradients are induced by the non downcomer panel was separated into sections to dis- uniformitie RCCe th cor d n si S an e annulus geometrs ya Figure 6 tinguis "gape hth " portions fro "blockede mth " portions well as the shape dissimilarity between the vessel and orden i modeo rt localizee th l d effect reflectivf so e radia- cavity wall Active Core Temperature Distribution During SRDC-11 tion in the cavity gape Th s located between sccuon risef so r tubes induce MWt0 45 Colume fo4 ,8 th r n Core MHTGR modee Th l parameter boundard san y conditions were an effect in which heat radiated from the vessel is M0 W35 e lbase MHTGth n do R design steade Th . y stale reflected off the downcomer panel, back to the vessel and 17 Upper Cold Plenum

-/n»

Riser Tube» Downcomer

•Se*

i 300 500 700 900 100 300 5ui. VESSEL MDWALL TEMP fF) RISER PANEL TEMP |'F) Figures Figure 9 Vessel Temperature Profile During Riser Panel Temperature Profile During Figure? t ReferencSRDC-1MW 0 e35 r MHTG1fo R Referenct SRDC-1MW 0 e35 r MHTG1fo R MHTGR COMMIX Model - Axial View (Ex-Vessel) (1.5 MW Steady-State Heat Load) (1.5 MSteady-StatW e Heat Load) o> R f O p To isers 60- s \ V From Vesselsurface ^ Sfl \^~~^ ^n \ X £ ; N ) •£ 40- / o s r: / to /^ > u a 30- J- >* «-"Y > <3 /^ / 20- y /- f Into Riser Paiici 10- 1 ! Bottom Of Risen Figur0 e1 OJ Vessel Temperature Distribution at Peak Power Plane 0 50 100 150 200 250 300 During SRDC-11 for 350 MWt Reference MHTGR Heat Transfer (planar avg.) W/ft/tube Steady-Stat(1.W 5M e Heat Load) Figure 11 re-radiated int cavitye oth . This effect increase locae sth l improve its overall economic potential Preliminary tran- Reactor Cavity Heat Transfer Distribution During vessel temperature, as evidenced at the peak location sient performance results for the depressunzed conduc- (upper left-hand come Figrm . 10). oppositp ga e eth tion cooldown, which are presented in this paper, indicate Referenct SRDC-1MW 0 e35 r MHTG1fo R closest to the vessel. In the core annulus, the coolant thathermae tth l MW0 respons45 e t desigth f eo nvers i y (1.5 MW Steady-State Heat Load) channel walls induce a "radiation-shield" effect, which at MW0 simila35 e t referencth o rt e plant. Analytica- lmo these locations reduces the heat transfer from the core dels arc presently be ing developed to calculate the passive banvessele th do t , thus diverting heat flo adjacenwo t t decay heat removal performance of the updated plant unimpeded locations. This effect creates the "dips'* in the design for depressunze pressurized dan d conduction cool- temperature profil shows ea Figurnn i 0 e1 down events These models build on the experience large Th e vessel-to-nser temperature differentials which obtained thus far and therefore will be more advanced and develop durin conductioe gth n cooldowns induc sigea - provide more accurate simulations of the passive heat late bot depressunzee hth pressurized dan d conduction removal process cooldown events. FIDAP is a finite-element based REFERENCES nificant natural circulation airflow in the reactor cavity 1 Blomquist, R N, et al. "COMMIX-1 AR/P- A Three- effece Th f freto e convectio hean o t transfe shows ri n i Fotransiene rth t performance analysi two-dimensionasa l program which wil especialle lb y usefu modelinr fo l e gth natural convection flows inside the reactor vessel dimensional Transient Single-Phase Computer Pro- Figure 11 The axial heat flux profile along the vessel mode beins i l g developed usin SINDA/FLU1Ne gth T gram for Thermal Hydraulic Analysis of Single and surface is much more peaked than at the cooling panel, therma fluid lan d flow analysa computer code which will (particularl plene (h an yi regions wiihid )reactoan e nth r cavity An axi-symmetnc model will be used to analyze Mulucomponent Systems - Volume 1. ANL-90/45. where it is more evenly distributed Free convection include the reactor, steam generator and crossduct vessel Argonne National Laboratory. Argonne. IL, July 1991 transports a portion of the heat rejected from the high and provide a complete system simulation of the prcssur the pressurized conduction cooldow perforo t d nan m power location vessee th t sa l mid-sectio uppee th d no rt an ized conduction cooldown event. SINDA is a network- sensitivity studies on key system parameters such as 2 'Program Manual for the Simplified Space Payload lower region cavite th f yso resistance analysis program which is especially useful for decay heat, core conductivity, and surface emissiviues Thermal Analyzer (SSPTA-3 0/VAX)." Preparer dfo modelin thermally-inducee gth d circulating flows which FIDAP also incorporate three-dimensionasa l view factor NASA. Goddard Space Flight Center, Greenbelt. MD. and radiation heat transfer capabilit three-dimenA y - by Arthu LittlerD Cambridge . c ,In . Octobe.MA r FURTHER DEVELOPMENTS can develop inside the reactor and steam generator during sional model is being developed to analyze the effects of a pressurized conduction cooldown. 1986 Based on the results of the Cost Reduction Study perform- geometry variation reactoe th n si r cavit cord yean annulas e1990n di MHTGe ,th R plant desig bees nha n reviseo dt Steady state model beine sar g developed usin FIDAe gth P peae onth k vessel temperatures during depressunzed incorporate a 450 MWi reactor thermal rating in order to computational fluid dynamics computer program 10 simu conduction cooldown events ANALYSIS OF AFTERHEAT REMOVAL FROM solutions developed for the HTGRe are the only ones of this kind in MODULAR HTGRs DURING ACCIDENTS, the reactor building and attempts are needed to apply them for other AND A CONCEPT FOR USE OF SPHERICAL reacto rw onestypene r .o s This woul t onlno dy promot e i-eactoth e r FUEL ELEMENT LWRN SI s building progres t alssbu o support developmen e HTGth R f o trendt . I.S. MOSEVTTSKIJ, A.O. GOLTSEV, P.V. MKHAILOV, V.F. TSIBULSKLF, V.D. DAVIDENKO, 1. CALCULATION F EMERGENCSO Y CONDITIONS WITE DECATH H Y V.S. POPOV, Yu.N. UDYANSKU I.V. Kurchatov Institut f Atomieo c Energy, HEAT REMOVA COOLANLE LOSTH Y TB T Moscow, Russian Federation

Abstract The calculations presented in this paper were performed for a modular reactor plant VGM, whose detail desig bees ha nn developen i d Calculation codes DIAB and KORSAR have been our countr e coolan yth e cas (1)th f o etn I .circulatio n e losth s developed and calculations of accidents with ceasing decay heat removal in the VGM is accomplished due to the core e circulationth f o d los e an f scoolano th e r fo t and reflector heat conductivity throug reactoe th h r vesseo t l modular reactor have been carried out. Some the passive cooling system (PCS) where heat is transferred to the coefficient sensitivitf so modee th lo t yparameter s final cooler (air) by natural convection of the water. The heaviest have been determined. It is shown that account for accidental condition blowes ga e s r case th aris failurth f e o n ei r eo the fuel element circulation reduces noticeably the dépressurisâtio f tho n e primary circuit s supposei t I . d thae th t maximum fuel temperature in these accidents. reacto thes ri n stopped. Possible use of the fuel elements, similar to The calculations were carrie t usino ou dtw codese , the HTGK spherical fuel elements, in the LWRs is DIAÇ and KORSAR which differ in their capabilities and the results considered s expectei t I . d that these new-type obtained e DIATh .B code calculate e two-dimensionath s l dynamic reactors would be able to withstand LOCAs, with the processe n multi-zoni s e (core, reflector, internals, vessel, PCS, fuel element d protectioan s n barriers remaining air system) (Fig.l). It also permits the helium natural convection intact. and radiative heat transfer to be considered. For the PCS the natural convectio e wate th s calculate i rf o n d taking into account INTRODUCTION possible boiling. The total number of mesh points in the DIAB-Mod5 versio =s emergence wa 340nth f .o Calculatios y hr cooldow0 10 f o nn e maith n f o problem e On s associated wite nucleath h r power conditions, with the coolant natural convection not taken into plant safety is removal of decay heat with the coolant circulation account, requires 2= four hour f machino s T 386)eA time C Th (P .e cease r undeo d r LOCA conditions e modulath r rFo . HT6R reactors, KORSAR code permits calculations of the three-dimensional uynaipic where the decay heat removal is accomplished through the reactor processe e multi-zonth n i s e system without coolant circulatiod an n vessel, this is decisive in determination of the main reactor with heat conductivit n eaci y h sone alsd radiativth an eo e heat characteristic e reactoth : s r sizd capacitan e d thereforean y s it , transfer in the gaps. The «Jislribul/ion of the residual heat can be efficiency thir .Fo s reason effort o improvst e calculational methods calculated with allowanc e spherith r cefo fuel element circulation. and codes, to perform verification experiments and bench-mark The results for the VGM were obtained primarily with the DIAB- calculations ae well as to seek for new methods and ideas on decay Mod5 e calculationTh . e blowega s e wer re th casee th f o madsr fo e heat removal intensification are BO important. The technical stop (the reactor at pressure) and the primary circuit failure (without helium). It found that even in the leaking reactor the 00 maximum temperature of the fuel elesente ie 1570 C within 39 hours 9l and doet exceeno s a dlon g permissible temperatur f o 160e . 0C However the difference between these temperatures is relatively 1 - Reactor cavern small, therefore calculation r estimatiofo s e sensitivitth f o n f o y 9 5-35-0.01i « clearancr Ai - 2 e the results obtaine o variationt d f somo s e parameters were carried --— - —————————— >i——_J Reacto- 3 r pressure vessel outparticularn I . , tabl liet1 e e these coefficient f sensitivito s y <•< ? ." ,' 4 - Helium clearance of maximum temperatures of the reactor components for variations of Thermoisolatio- 5 n two most important parameters : decay heat q and effective heat .T ,19 conductivity X for the case of primary circuit depressurization. 6 - Boracic graphite 0 X 6 •2.1 7 - Graphite Table 1 'iJ. 9 collectos Ga - 8 r 8 reflectop To - 9 r Max.T, C Time for Sensitivity coefficients 2 2Clearanc- 0 1 e Components T max Tmax (hr) 11 - Reactor core 11 12 - Bottom reflector collectos Ga - 3 1 r Core 1570 39 +5 -13 in 6.1 Graphit- 4 1 e Top reflector 545 7 + 1,6 -3,4 co 15 - Boracic graphite Side reflector 820 70 +3,5 + 1,5 Tliermoisolatio- 6 1 n Bottom reflector 871 0 0 0 17 - Supporting plate Reactor vessel 364 88 + 1,3 +7 13 - Helium clearance Heliu- 9 1 m clearance As seen from the table the T max results for the core are »ore 12 20 - Metallic construction sensitiv o variationt e e effectivth f o s e heat conductivite th f o y 21 - Radial reflector pebbled core. 22 - Cooling ducts 11 Determinatio sensitivite th f no y e connectestimateth o t s s 23 - Cooling ducts of possible errors in emergency condition calculations. 2 4- KLA K ducts Howeve o t featurere code th e thes sf du o s e e b error y ma s 16 17 18 and numerical methods used as well as to possible mistakes of the codes .o s importanTherefor e s i th o t carrt t i eou y calculations using differen e samt th codee r initiafo s l data, i.e. 6-0.090 5 to make "bench-mark" type calculations. Suc a carefuh l comparison was made for the DlAB-Hod5 and the THERHIX system code using a finer calculational mesh. This codbees eha n develope d ie dan th uee n di Julich Research Center (Germany) and the comparison iteelf was carried out in accordance with the agreement between KFA attd RRC KI. FIG. 1. Calculation scheme and main sizes of reactor. The comparison .results have reveale e caus th f do smal e l discrepancies and some changes wex-e introduced into the DIAB. y powean r reactoe intacb n rca t only with heat removed frocore th me Comparisons were «leo made by the codes DIAB and KORSAB, which by the coolant. showed a reasonable agreement of the results. Therefore the codes we Analysin f safeto e th yg proble d decaan m y heat removae w l used in the calculation of accidental conditions can be considered formulate e e problefollowinth th d n i m gt possibli «ay s i : o t e sufficiently verified. create a reactor with an average level of heat density (10-50 The feature of the HTGR with spherical fuel elements is MH/m3) wher e protectioth e n barriere fueth l f o elements s would circulation of the latters and their recirculation through the core. remain unf ailed in the total LOCA and, thus, the radiation safety It is important for the problem of decay heat rénovai because the relativ o thit e s severe acciden e ensuredtb d wha,an t woul e thib d s fuel 'elements are nixed in refueling and the decay heat fraction reacto Thi. r s e b solveproble n dca m when during this accident generated by the decay of long-lived isotopes is rather uniformly distributed ovecoree th r . With this factor taken into accounte th , Result f Fueso l Temperatur Timn i e e highest temperature fuee th l f elementso s afte depressuriaatioe rth n acciden e founar t d much lower. Fig.2 show e KOBSAth s R results obtained for the depressurisation accident. In the distribution of the decay heat over the core the fuel element circulation was taken into account. As seen the 12 fold circulation of the fuel elements in the VGH reactor leads to reduction in the maximum fuel element temperature abou t r e increasinuseb fo 200 d y Thi. 'ma C s g the nominal reacto s determinei r t i powe s a rd jus y elemenb t t temperature under depressurisation conditionse th s showA .y b n comparison of the result obtained with the sensitivity coefficient o fe nomina tablth , 1 e l e essentiall b powe y ma r y risen, with this effect considered. 1200

FEATUREE TH F O SPHERICAF E SO 2 .OS L FUEL ELEMENTS WITH COATED 1100 PARTICLES FOR SOLVING THE DECAY HEAT REMOVAL PROBLEM WATER-COOLEE ITH N D REACTORS 1000

The safety of any reactor would be ensured if the removal of yoo the decay heat could be accomplished with the fuel elements and 1,, HO their protection barriers agains e fissioth t n product release ooo remaining unf ailed. The modular HTGRs are unique in thisO respect1O 0 9 . 0 0 0 0 2 7 '.'S O O0 1 6 O -J 'SO 0 Even durin e totath g l e LOCdecath A y heat removal througe th h I -Fiifl Doii'l. MOV—.-: reactor vesse s accomplishei l y thae fuen sucwa i dth t la h elements 2-OiiM-ul.it.ion wit f-oluii " hn ""i V^louJ'v protectioe anth d n barrière remain intact. Thi a poseible i so t e edu 3-J.Ï-tli 1:1 ^oul.ïtn io high permissible fuel temperatur d lowan e. power densit f th«o y s core (=3 MW/ro ). At higher values of the power density (exceeding 2 FIG . Maxima2 . l temperatur f fueeo l elements after LOC shutdowd Aan differena r nfo t =10MH/ fuee th l ) nelement protectiod an s n barrier core f th o e n i s circulatio fuelf no . ro the fuel elements mil be t enoved from the core into a special o cooling zone where the heat lemoval «ill be accomplished at permissible temperatures This may be attained using the following reactor concept -the fuel elements aie high-temperature spheric ones similar of pebbl d HTGbe e R fuel elements e b ,provide n theca yd with antioxygenic coating, -the core consist f blocko s s accommodating these fuel elements and is cooled with the water coolant, -each block is made of a material suitable by its physical fuel element characteristic d havinsan e meltinth g g point higher thaworkine th n g block temperature but lower taken that permissible for the fuel elements (e g. a Al-based material ), -The zone for cooling the fuel elements, in the absence of a coolant s locatei , o designes d s beloi e cord dth w an e that cooling of the fuel elements inserted into are not overheated In this reacto re totath plan ln i td LOC witan Ah rising temperatures of the fuel elements and blocks the core fragmentation occurs because of melting of the block materials The spheric fuel elements coming out of the blocks move by gravity downwards and, together wite blocth h k material melt ente cooline th r g zone Fig.3 fuel elements possible th f showo e son scheme f sucso reactoha r powew witlo ha r densit 5 MW/n>31 e diagrae heat-generatin~ Th ( yth ) f o m g blpcs i k e HTGTh R 4 spheri g Fi show n ci n fuel elements wit a standarh d e usen thiI ai d s m m diamete cas e lowe0 6 th e e f ro rth par f o t cooling reactor vessel present a coolins g e decazonTh ey hea s removei t d zone from the fuel elements being in this zone through the pattive-cooled reactor vistsel wall and by the radiation from the pebble-bed surface Durin e wholth gf coro e e fragmentatio d fuean nl element movemen e coolinth o t tg aone maximuth e m fuel temperature remains _^- melt lower thae permise.Iblth n e temperature accordin e estimatesth o , g . Fig 5 shone the e&tlmatee. of typical variations of the maximum temperature of the fuel e-iemcntb fur thie leactoi and the main staget of occurring procebfcet For choosing the optimal versions of the core components FIG 3 Heat removal scheme for normal conditions (A) and after a coolant loss and a (coated particles, graphite and water coolant share) the fuel element moving to a cooling zone (B) investigation f neutronith f o s c charactei ist-ic f theso c e systems have bee ne calculatio Th carrie t ou d n lesulta show (Fi g) tha6 t negative feedbace leasonablth d an k e characteristic e fuee th b l f o s obtained if the volume share of the coolant will be not largd an e iSOOn the uranium density in the core exceed 500 kg/m"3. The above fuel compoeitions eneur e consumptioth e naturaf no l uraniue leveth lt a m currene ofth t VVERs (^ 0.2nat/MW-dayU g 4k ). The calculations of passive heat removal through the reactor vessel wall for a maximum temperature of fuel elenents of 1600°C have been made. For theee conditions Fig.7 shone the dependence of

time (hour) fuel-element FIG Tim. 5 . e dependent maximal fuel temperatur LOCr e fo shutdown d Aan .

coolant

fuel element

q

FIG. 6. Dependence of K„ on water volume; enrichment is 8%; fuel element diameter is FIG . Schem4 . reactoa f eo r block containing spherica lfuer fuepe l lelement U elements g 0 4 : .4 .d an ; U g 0 5 : 3 ; U g 0 6 : 2 ; U g 0 7 : 1 . mm 0 6 ro ro Sc.i — — $=20'ca, m1 100'

10: T

1-

0.1a i j ifljio 2o.bo Jabo ' 40.0 so!' 0' «ii»' ' »

Dependenc. FIG7 . limia f eto thermal flo coolinn wi g zon corn eo e power density.

the limiting heat flu e xnoninath throug n wale lo th hl powe r e cordensitth e f o accommodatey e standarth R n i reactoWE dd r vessel. However another geometr cooline th f o yg zone poeeiblee i t I . should be chosen from the condition of not exceeding the permissible temperature of the fuel elements. As ehown by the estimates e powe,th r densit thin i y s reacto reacy rma h 40-50 MW/r, n which corresponds to the current requirements.

REFERENCE

1. Mosevitski I.3., Grebermik V.M., Sukharev Yu.P. The USSB gas cooled reactors : statue and prospects. Paper presente t IX-ta d h e MeetinIAEth A f o gInternationa l working grou n HTGBpo . Nov. 5-9, 1390, Otik-Eidge USA. EXPERIMENTAL DATA FOR VALIDATION OF PREDICTIVE METHODS - OPERATIONAL EXPERIENCE FROM OCRs, EXPERIMENTAL DATA FROM TEST FACILITIES (Session IV)

Chairman

E.F. HICKEN Germany HEAT TRANSFE UPPEE TH RRN I PAR T studied. Radiation heat transfer was calculated analytically and net heat transfer rate by natural convectio compares nwa d wit resulte experimene hth th f so t without thermal radiation. There OF THE HTTR PRESSURE VESSEL arprioo en r paper concernin naturao gt l convectio thif no s geometry resulte presene th ,th f so t DURING LOS FORCEF SO D COOLING experiments were compared wit e studiehth s performe enclosuren di f basiso c geometries such horizontas a verticad an l l enclosures.[1]-[10] . SHIINAY . HISHIDM , A 2. Experimental apparatu procedurd san e Japan Atomic Energy Research Institute, Experimen1 2. naturan o t l convection without thermal radiation Ibaraki, Japan Figur showe1 s schematic diagra experimentaf mo l apparatuses. Figure l(a)- showap e th s paratus for heat transfer experiment. It consists of a hemispherical shell made of SUS 304 Abstract 300mm in inner diameter 8mm thick and a copper plate 20mm thick. Copper plate was heated by sheathed heater weldee plateth n .do Outer surfachemisphere th f o e - cooles wa ewa y db Heat transfer characteristics in upper part of pressure vessel of High Temperature Engi- ter flowing in copper tube soldered on the surface. Inner and outer surface temperatures of neering Test Reactor( HTTR) during loss of forced cooling were studied experimentally by the the hemisphere were measured at fourteen positions located on five concentric regions on the use of hemispheres heated from below. Natural convection heat transfer and flow characteris- hemispherical snrface(A-E shows a ) Figurn ni . Locae2 l heat fluxes acros hemisphere sth e were tics without thermal radiation were studied with several kind liquidf so s workinsa g fluidd an s obtaine headby t conduction calculations. Average heat transfer obtainerat arithewas the -dby thermo-sensitive liquid crystal powder visualizatioa s sa n tracer e rangeTh . f Rayleigso d han metical average of the local heat fluxes. Prandt 1300< r hear P 0 fo l t < number transfe6 IDx d 5 an 5. sr < studfoe a rth S y < wer * e10 Water, ethyl alcohol, 78wt% 44tufd an % aqueous glycerin silicod , an Freo 3 n 11 noil s were experiment an10d 2 x 10s < Ra < 1.9 x 109 and 6 < Pr < 13000 for flow visualization experiment. use s workinda g fluids. Liquid temperature were measure r L-typo I y edb temperature probes Heat transfer correlation flod wsan characteristics were clarifie laminan i d turbulend an r t with CA thermocouples 50pm in diameter inserted into air-release pipe welded near the top of regions. In laminar region, circulating flow existed in a hemisphere, i.e., hot fluid risen up near the hemispher 12.5mm(= r t ea r — z co-ordinate showe sar Figurn ni e l(a)) e probs Th . ewa the center of the hemisphere arrived at the top of the hemisphere and flowed downward along traversed vertically. inner surface of hemisphere. Whereas, in turbulent region of Ra > 109, the circulating flow Figure l(b) shows the apparatus for flow visualization experiment. Flow visualization exper- disappeare randod dan m upwar downward dan d flow became dominant t fluiHo .d released from imen performes hemispherewa to tw y db s 100m 300md man innermn i diameter mad acrylif eo c t surfacho e dissipates ei th d immediatel surroundinn yi g fluid. resin. Hemispheres were set in rectangular enclosures shown in Fig.l(b). Outer hemispherical Natural convection heat transfer experiment with thermal radiatio alsw ono beins ni g carried. surface was cooled by flowing water between rectangular case and the hemisphere. Thermo- Maximum temperature of a hot surface can be raised up to 600° C. Air is used as a working sensitive liquid crystal powders were used as a tracer. Since the powder changes in color at fluid. Rayleigh number range will be 1.5 x 108 < Ra < 5.3 x 108. Net heat transfer rate by distinguishee b t fluin ho ca d , difference abouC th * y 30 db t colorsn i e . Water, 78wt% aqueous natural convectio obtainee b n nsubtractinca y db calculatee gth d heat transfer rat thermay eb l glycerin and silicon oils were used as working fluids. radiation. Estimated heat transfer rat naturay eb l convectio s compareni d wit resulte hth f so Temperature of the copper plate was kept at constant value up to 45*C. Physical prop- natural convection experiment without thermal radiation. erties were estimated at the average temperature of hot and cold surfaces. Non-dimensional parameters, Rayleig Nusselgß&TRd = definee ha an ar Ä , s t3a d [K.Vnumbers Nu an d an a R , 1. Introduction A'u = Äj/AAT where g, R and q are the gravitational constant, inner radius of the hemisphere and heat flux respectively, ß, K, v, A are coefficient of expansion, thermal diffnsivity, kinematic During loss of forced cooling of High Temperature Engineering Test Reactor(HTTR), decay viscosity and thermal conductivity respectively and AT = TA - Tc where Tk and Tc are tem- heat wil removee b l d from reactor cor heay eb t conduction, natural convectio d thermanan l cold peraturean d t surfaceho f so s respectively rangee Th . f Rayleigso Prandtd han l numbers radiation. Wit increase hth temperaturn ei reactoe th f eo r core, several part pressurf so e vessel flor wfo 9 10 10x x hear 5 19 0fo 5. 1. t < transfe< a a R R < < werr 6 * experimene10 10 x 2 d an t will be heated by thermal radiation and natural convection. visualization experiment and 5.3 < Pr < 13000 for both experiments. Heat transfer characteristics in upper part of pressure vessel of HTTR during loss of forced cooling will be important from the viewpoint of safety, because hot coolant heated by the upper surface of the reactor core might rise up to the pressure vessel and the vessel might be 2.2 Experiment on natural convection with thermal radiation locally heated up. Amount of heat transported by thermal radiation increases with increase in temperature of reactor core and decreases with increase in pressure in the vessel. Therefore, the Figur showe3 s experimental apparatu naturar sfo l convection with thermal radiation. Pressure- rol f thermaeo l radiation wil more b l e important when temperatur f reactoeo r cor higs ei d han vessel-type enclosur 1000m4 e 30 consistS minnen SU i f so r diameter, uppe vessee r th hal f s o fli pressure is low. Transported heat by thermal radiation, however, could be estimated analytically a hemisphere. Lower half was filled with thermal insulation blanket. Heated surface 600mm in becaus relativelf eo y simple characteristic thermaf so l radiation. diameter made of SUS 304 was installed at the center of vessel as shown in the figure. The vessel On the other hand, heat transfer by natural convection depends on flow in pressure vessel, was cooled by surrounding rectangular cooling panel located outside the vessel. Cooling water whic s regulatehi geometrye th y db , size, configuratio heatef no cooled dan d surface temd an s - pipes are attached on the panel. Therefore, temperature of the hemispherical surface increases perature difference between the surfaces. Therefore, it is not simple to estimate the detailed with increas temperature th n ei f heateeo d surface workins usea i r s da Ai . g fluid. thermal effect naturaf o s l convectio uppee th n nrpressuro e parth f o t e vessel. Inner and outer surface temperatures of the hemisphere were measured at 23 positions. objective Th presene th f eo t stud clarifo t basis e yi yth c characteristic naturaf so l convection Total heat fluxes acros e hemisphericath s l surface were obtaine conductioy b d n calculations. heat transfer in upper part of the pressure vessel during loss of forced cooling. A hemisphere Temperature of the hot surface can be kept at an arbitrary constant value up to max. about was used to simulate the geometry. First, heat transfer and flow characteristics of natural 600°C. Heat fluxes due to thermal radiation were estimated analytically. The estimation method convection without thermal radiation were studied experimentso Tw . , heat transfe flod wan r of thermal propertie definitioe th d san f non-dimensiona o n l parameter same sar s describeea d visualization experiments, were carried out. Basic feature f naturaso l convection were obtained in the previous section. The range of Rayleigh number in the experiment will be 1.5 x 10* < CJJ froe experimentsmth . Second, natural convection heat transfer with thermal radiatios wa n I0x 3 8.5. Ra< 3 Result discussiond san s Cooling to jWoter O> 3 1 Heat transfer and flow patterns of natural convection witout thermal radiation Cooling Pone\ l Water I —4- Woter Generally the functional relationship between Nnsselt number and Rayleigh number is ex- C pressed as ~ Y 1350- ?~ Nv = CRanPrm (1) ) where C,n and m are empirical constants Average Nnsselt number N u is plotted against Cooling Rayleig propose] h[8 numbel Sek4 d a t 07 an valu , ie Droplue d ] 0 m Figurth n [1 ri s f el o A a t nee 4 Panel ^ and 0 024, respectively Figure 4, however, indicates that the effect of Prandtl number is not I r \ prominen presene th n i t t stud resulte representee b Th yn followine sca th y db g correlationships o J£ — i —— 600 01974Ä1025 (I0 8A < 10o< 9) (2) ir* 0312Ä0033 (109< JZa<55xl010) Cooling iE Electric -^ir 3! Cooling (3) Heater Water h Ceramic r Woter —c J Plate —— 1000- ^Air-Release "'Pipe Air release pipe

/////x////// / /

Figur e3 Experimenta l apparatus with thermal radiation

\Bakelite Heater Copper Plate H«l«r

(a) Heat transfer experiment (b) Flow visualization experiment Figure 1 Schematic diagram of experimental apparatus without thermal radiation 103 nj r nij i i u| i i HI wateo - r u * 44e/.glycerine Bohn et al - 78°/.glycenne Tanaka et Ethyl alcohol 10 2Freon11a - 3 ~Silico• l noi

6(deg) Eq.(3) 16 28 44 60 i i ni i i nl i i nl i i ni i i n! i in 76 8 e 10 10 109 7 1110 6 10 s 10 Ra Figure 2 Positions of thermocouples soldered concentricall hemispherican yo l surface (region) E ~ sA Figur e4 Plo averagf o t e Nusselt number against Rayleigh number The above correlationship showe sat soliy nb d figuree linepresene th n sTh i . t experimental Vertical distributions of average liquid temperature measured at several radial positions are data were compared with correlationships obtaine Tanaky db t al.[4ae horizontan ]i l enclosures show Figuren ni , 7(a8(a)s6 d )an . Average liquid temperatur obtaines ewa averaginy db g base and Bohn et al.[10] in vertical enclosures shown by dotted and broken lines, respectively. In the liquid temperature excluding the temperature fluctuations due to rising hot liquid. In the figures, cases of horizontal and vertical enclosures, average heat transfer rates on hot and cold surfaces verticae th s i h l z distanc t surfacinnee ho th fro e ro em t th surfac f hemisphereo t arbitrarea y are the same because of the same heat transfer areas. For the present case, however, the area radial positions Rayleigw lo r Fo h. number t fluiho , d rises steadil centee hemisphereth a t f ya ro , of cold surface was twice as large as that of hot surface, therefore, average heat transfer rate on therefore, liquid temperature is high near the center and becomes low with increase in radial the cold surfac just s surfaceho ewa t e hal f thath . o f f Thio t s majoe woulth e rb d reasoy nwh distance 10x fro centee 7 6m.th 5. shows = a r a FigurR n ni r fo e6 the present data lie below the results of other investigators. Nnsselt number on hot surface is show dottey nb d chai comparisone n th lin r efo . Heat transfer analysis of laminar natural convection along spherical surface based on bound- 1 ary layer approximation]]2] gaveq.(l n 0.25i value e b en th o f .)t e o Many investigations showed s almos r thaturbulenwa fo n t t 3 equa1/ o tt l natural convection. Sinc effecte eth f encloo s - sure geometr hean yo t transfer characteristics woul little db turbulenn ei t region exponene th , t -8 Ra= n = 1/3 could be applied in turbulent natural convection in a hemisphere. Therefore, equations 5-7x105 ) correspon(3 d (2an ) laminao dt turbulend an r t heat transfer respectivel flod wyan becomes fully turbulent for Äa > 10*. -6 Pr= Figur showe5 relationshie sth p between local average Nusselt numbe Rayleigd an r h number. 13000 Local average Nusselt number are obtained by averaging the Nusselt numbers on each concentric regio £ shown o frot r Rayleig Figurn ni mÀ Fo . e2 h 10< numbe sa , R Nussel f o r t number '* decreases with increas angln ei e measured from vertical existence . th Thi o t se ewouldu e b d of downward flow alon hemisphericae gth l surfac laminan ei developmeno t e r regiodu d nan f o t -2 boundary layer in the flow direction. For Ra > 10 , however, Nusselt number of large angle regions(region ) showE d s an highe sD r values than tha smalf o t l angle regions(regio. A) 9d an nB Thi randoo t s e wouldu m e db upwar downward dan d flow betweecold an d t surfacnho e caused by turbulence thin I . s region floe th ,w alon hemisphericae gth l surface woul disappearede db . 0 -2 -A -6 -8 1- The ratio of the highest Nusselt nnmber(usually at region A) to average Nnsselt number was (T-Tc)/(Th-lc) laminaaboun i 5 1. t r regio turbulenaboun d i nan 2 1. t t regiopresene th n i t study turbulenn I . t region, the top of hemisphere would not be so locally heated up by natural convection as to in laminar region. Figur e6 Vertical distributio averagf no e liquid temperature 13000 = 10x r s7 P , (Ä)5. a=

1 f I I —

Ra=8-2x106 Pr=8800 10'

10'

1 l il l III i i l il I l I II I ill 1- 5 6 7 8 9 10 10 10 10 10 10 10 10" (T-Tc)/(Tr,_Tc) Z/Z* Ra (a) Vertical distribution of average (b) Vertical distribution of maximum liquid temperature temperature deviation

Figur e5 Plo locaf o t l average Nusselt number against Rayleigh number Figure 7 Results of liquid temperature measurement for Äa = 8.2 x 106, Pr = 8800 i IV) 00 Ra=4x10K> Ra=4x1010 Pr=8 •8 r/R o oO-017 9 Pr=8 -5 nO-177 •6 • 0-419 SyteOgraptu^fly^y^-^-,^,r>4ZO«ou»* '0-579 or/R = 0-017 Q 0-177 •2 • 0-419" A 0-579 0 till 0 -2 -4 -6 6 - -8 4 I- - 2 - } -8 1- (T-TO/CJh-Tb) Z/Zh ) Vertica(a l distributio averagf nVerticao ) (b e l distributio maximuf no m liquid temperature temperature deviation

l Figure 8 Results of liquid temperature measurement for Ra = 4 x 10 °, Pr = 8 10x 4 7 , = a R ) (a PT = 290

With increas Rayleign ei h number, vertical distribution averagf o s e liqnid température show boundary-layer-type profile as shown in Figure 7 (a) for Ra = 8.2 x 10$. Flat temperature profiles could be observed at several radial positions except near both surfaces. Lower and higher temperature regions were observe outee th t rdthermaa e edg th f eo l boundary layers near hot and cold surfaces respectively. These temperature reversals which are more prominent near t surfacho e t causee upwarth ear ho y d b dcente e floth cold t wa ran d downward flow alone gth hemispherical surface. The temperature reversals could be observed until Ra = 10s. Figure 9(a) shows the flow and temperature patterns at Ra = 4 x 10 . Hot liquid released from the hot surface flows toward hemispheree th centee f o sth rise d p ran to s e ther. th Severa o et l streakf so 7 t liquiho e d th coul seee db n rising near centee hemispherth a f ro figuree th n ei . Vertical distribution maximuf so m liquid temperature deviation from average valu showe ear n in Figure - designatsT 7(b d 8(b)figured e an th )an difference e + n .eI th 7 , maximue th f so d man

minimum liquid temperature from the average value, i.e., T+ = T^^-T^ and T- = 7*,„ -T„ m where Tm„, T„, and Tram are the maximum, average and minimum liquid temperatures at z respectively. Belo10= keepa 8+ ,wT R s high l z—distancevalual r efo s cold betweean d t ho n surface centee th t a rs region. With increas radian ei l distance frocentere mth decrease+ 7 , s rapidly with vertical distance from the hot surface. For the region of r/R > 0.177, hot liquid can not reach z/Zj, > 0.3. This indicates that there exists a circulating flow in the hemisphere, and therefore hot fluid rises mainly at the center. 10> Fo9a ,R rhowever , temperature reversal disappearee ar s s showa d n Figurni e 8(a) which is the results at Ra = 4 x 1010. Vertical distribution of T+ is shown in Figure 8(b). e maximuTh m temperature deviatio almoss i T n l z—t al zer r distancesofo cold an d frot mho

surfaces except very near th+ e surfaces. These results indicate that for Ra > 109, circulating (b) Äa = 1.9 x l O9, flow disappeared and the effect of hot fluid is confined in the vicinity of the hot surface. Flow shows i 9 Figurn ]0 ni x e9 9(b)1. = patter . a ManR t na y small eddies coul observee b d d dan Pr = 6.5 dominan e observedb t te no result foln Th . wliquif ca so d temperature measuremen flod an wt visualization experiment support the results of heat transfer experiment. For high Rayleigh > Idnumbe a R , f floo r w became fully turbulen t fluiho dd releasean t t surfacd ho fro e meth 9 is easily dissipated in surrounding fluids. Therefore, hot liquids can not reach the top of the Figur e9 Visualized pattern flof temperaturd so wan e fields hemispher turbulenn ei t region. Hea2 3. t transfe naturaf o r l convection with thermal radiation Local net radiation energy dQ, from dA, on surface 2 to outward can be obtained by sub- tracting inlet radiation energy from emitted radiation energy obtaie W . n experimente Inth , hea s transportei t naturay db l convectio thermad nan l radiation. Heat transported by natural convection was obtained by subtracting the heat transported by thermal dQ dA,(ac,T?= , - e.ft. ) radiation which was calculated analytically. The radius of the hemisphere R is not equal to that t surfacho e oeth f 0.3m)Ri(R= I 0.5R — d .m Sincan e natural convection woul regulatee db d Loca t healne t flu thermay xb l radiatio surfacn o , nequag s i e2 dQ,/dA,o lt , dA t a 0, therefore by the hemisphere, the radius R was used as a hydralic diameter. Heat transfered by thermal radiation has to be estimated. There are three surfaces in the 4 1, = r, 6.- ) (13) hemisphere, which contribute to thermal radiation, i.e., hot surface, hemispherical inner sur- 1 -e, face(cold surface insulatiod )an n surfac shows ea Figurn i . Radiatee3 d heat from hemispherical surfac eaco et h surface coul estimatee db followss da : Radiation energy Bt from t generallsurfacno s i ek y constant dependen positionn o t s because We define surface numbers of hot, hemispherical and insulation surfaces as 1,2 and 3, re- temperatur t neccesaril surfacf no eo s i ek y uniform presene th n I t. experiment, howevere th , spectively. Angl smala e facto of hemispher l the are FH r aon surfaceto e l(hot surfacebe can ) non-uniformit e temperaturth f o y e effec s th small, f non-unifor o ewa d t an m temperaturn eca obtaine: das neglectede b . Therefore, average surface temperature coulcalculatioe th usee r b d dfo f BI,.no Emissivities were derived from the measured results of other invc$tigators[13](l4][15]. R l cos 6 Figur show0 e1 plosa totaf o t l heat transfer rat naturay eb l convectio thermad nan l radiation (4) against Rayleigh number surfact Ho . e temperature variees th T\150n < wa i d\ °C T 100 < "C experiment. Soli figurde linth en ei shows equation (2). e anglth wher s ei measure6 e d Irom vertical. Angle factor smal» s d FISf FHo an l aren ao Figure 11 shows relationship between net heat transfer late by natural convection and e hemispherth surfaco et e 2(hemispherical surfaceinsulatioo t d an ) n surface respectivele yar Rayleigh number. Amoun heaf o t t transporte thermay db l radiation calculate methoe th y db d obtaine similae th n di r manne: s a r described above was subtracted from total transported heat. Further experiment will be per- formed in high temperature regions where thermal radiation will play an important role in heat transfer. (5) 4. Conclusions (6) Heat transfer characteristic uppen si r par f pressuro t e vesse Higf lo h Temperature Engineer- ing Test Reactor( HTTR) during loss of forced cooling were studied experimentally by the use Radiation energy 6, from small area dA, on the hemisphere to outer space per unit time and of hemisphere heated from below. unit area is expressed as

4 6, = o-f.T, + (1 - t,)h. (7) where k, and (, disignate local inlet energy by thermal radiation from outwards and emmisivity 100 100, respectively. Radiation energy frosurfacn o m writte, s i surface2 dA : s o t na ek dA,h, =

(8)

where BI, designates radiation energy from surface k to outward per unit time and unit area. II obtainee b The n yca s da

B = 1 (9)

(10) I I I 1 I i i i i 10 10 8 9 10 10 108 109 where Ai,Ai,A3 are radiation area of surfaces 1,2,3 respectively. Ra Ra Radiation energy 6, is rewritten by substituting eq. (8) into (7) Figure 10 Plot of total heat transfer rate Figur t hea 1 Plone e1 tf o ttransfe r rate by natural convection and thermal by natural convection against Rayleigh (12) radiation against Rayleigh number number r* First, natural convection heat transfer and flow characteristics without thermal radiation 13. Makino, T. et al.,"Study on the properties of thermal radiation of metalic materials for Q were studied with several kinds of liquids as working fluids and thermo-sensitive liquid crystal high temperature", Trans. JSME., 49(1983) 1040-1047 (in Japanese) powder visualizatioa s sa n tracer rangee Th . f Rayleigso d Prandthan l number expere th r -sfo iments were 106 < Ra < 5.5 x 1010 and 6 < Pr < 13000 for heat transfer experiment and 14. Sala , "RadianA. , t propertie f metals-Tableso f radianso t value blacr sfo k bod d reayan l 1300< r flor P 0 fo w < visualizatio6 d an s 10 x n 9 experiment1. < a R < .5 10 2x metals", Physical Science Dat , Elsevier(1986a21 ) Heat transfer correlations and flow characteristics were clarified in laminar and turbulent flow regimes. In laminar regime, hot fluid released from hot surface rises at the center of the enclosure. 15. Touloukin, Y.S., DeWitt, O.P., "Thermal Radiative Properties-Metallic element- Al d san It reaches the top of hemisphere and descends along hemispherical surface. Therefore circulating loys", Thermophysical properties of Matter, the TPRC Data Series 7, IFI/Plenum Data flow exist in laminar region. Whereas, in turbulent region, the circulating flow disappeared and Corp.(1970) random upwar downward dan d flow became dominant t fluiHo .d release t surfacd ho fro e mth e is dissipated in the surrounding fluid immediately. From heat transfer and flow visualization experiments, flow induced by natural convection became fully turbulent for Ra > 109. Second, natural convection heat transfer experiment with thermal radiatio beinw no g s ncari - ried. Maximum hot surface temperature of the experiment can be raised up to about 600°C. Air useworkins a i s da g fluid. Rayleig10x heat 3 s5. Ne .t < h a numbeR < 8 r 10 rang x 5 e 1. wil e b l transfer rate by natural convection can be obtained by subtracting the calculated heat transfer rat thermay eb l radiation from total heat transfer rate. Heat transfer rat naturay eb l convection is compared with the results of natural convection experiment without thermal radiation.

References

1. Dropkin, D., Somerscales, E.F.C.,"Heat transfer by natural convection in liquids confined by two parallel plates which are inclined at various angles with respect to the horizontal", Trans. ASME. Ser.C, 87(1965) 77-84 2. Chu, T.Y., Goldstein, R.J., "Turbulent convection in a horizontal layer of water", J.Fluid Mech., 60(1973) 141-159 . Trelfall3 , D.C., "Free convectio low-temperaturn i e gaseous helium", J.Fluid Mech., 67(1975) 17-28 4. Tanaka, H., Miyata, H., "Turbulent natural convection in a horizontal water layer heated from below", Int.J.Heat Mass Transfer, 23(1980) 1273-1281 5. Fitzjarrald,D.E. experimentan "A , l stud f turbulenyo t convectio f air"no , J.Flnid Mech., 73(1976) 693-719 . MacGregor6 , R.K., Emery, A.F., "Free convection through vertical plane layers-Moderate and high Prandtl number fluids", Trans. ASME. Ser.C, 91(1969) 391-401 7. Seidexperimentat al.n e . ,"A N , l stud tref yo e convective heat transfe parallelograma n i r - mic enclosure", Trans ASME. Ser.C, 105(1983) 433-439 . t Sekial.8 e . , N "Hea, t transfe enclosen a n i r d cavity wit hrelativela y small aspect-ratio", Bull.Fac.Eng., Hokkaido University, No.87(I978) 1-10 9. Emery, A., Chu, N.C., "Heat transfer across vertical layers", Trans ASME. Ser.C., 87(1965) 110-116 10. Bohn, M.S., Kirkpatrick, A.T., "Experimental stndy of three dimensional natural convec- tion high-Rayleigh number", Trans. ASME. Ser.C, 106(1984) 339-345 . Goldstein11 , R.J. t al.e , , "High Rayleigh-nnmber Convectio a horizonta n i n l enclosure", J.Fluid Mech., 213(1990) 111-126 12. Shiina, Y., "Heat transfer from the inner surface of a sphere by free convection", Heat Trans. Jpn. Res., 18(1989) 70-86 DEVELOPMENT OF AN INACTIVE 1. INTRODUCTIO N HEAT REMOVAL SYSTER MFO Passive safety system ] hav[2 s e become mord moran ee importann i t HIGH TEMPERATURE REACTORS nuclear engineering. With this in mind and by recognizing that evolutionary approaches to these problems will not solve all of them a kin f revolutionaro d y approac neededs i h . Therefore, Siempelkam latn i p e K. KUGELER 1987 proposed a solution [3] consisting of a prestessed cast-iron Rheinisch-Westfälische Technische Hochschule, pressure vessel and a passive heat removal system, integrated in the reactor cell surrounding the vessel. This solution combines the inherent Aachen safety of a prestressed metallic pressure vessel with the advantages of a passive heat removal system and thus is a major step towards the goal . SAPPOKM . BEINB , E of substantially reducing remaining residual risks. Siempelkamp Giesserei GmbH & Co., e worTh k necessar o prov t ye feasibilit th e f suco y a systeh s wa m Krefeld divided int phases2 o . Phasperformes wa I e d from Sept 198, 1 .Febro t 7 , 1989.28 . L. WOLF The content of Phase I was the overall design of the system with special emphasi e heath t f o transpors t e inne froe lineth th mrn o r surfacf o e Battelle-Institute.V., pressure th e e coolinvesseth o t lg tower. This wors beeha kn performed Frankfurt by the companies Siempelkamp and Siemens/Interatom and accompanied by the Jlilich nuclear research center. Germany Phase II started Aug. 1, 1989 and will end Sept. 1992. The work performed was a series of tests in a large scale test setup to determine effective values for thermal conductivities, thermal resistan- Abstract ces and to verify and/or correct the calculations and engineering design Growing publi d politicaan c l interests towards incorporating passive performed in Phase I with special emphasis on the heat transport by safety features in nuclear installations, let Siempelkamp in late 1987 radiatio d convectioan n cavitye th n i n. propose a solution consisting of a prestressed cast-iron pressure vessel These tests also included a steel pressure vessel wall instead of a an a passivd e heat removal system, integrate e reactoth n i dr cell prestressed cast iron pressure vesse reasonr fo l comparisonf so . surrounding the vessel. This solution combines the inherent safety of a The work was performed by Siempelkamp and Battelle Institute in coope- prestressed metallic pressure vessel with the advantages of a passive ration with Siemens/Interato d Profan m. KugeleK . d accompaniean r y b d heat removal system and thus constitutes a major step towards the goal the Jülich nuclear research center. of further reducing potential residual risks. The design had to meet the boundary conditions for reactor core and 2. BASIS AND OBJECTIVES OF THE PROJECT reactor buildin modulae th 0 f MWto g20 r h pebbl reactod ebe Siemens/f ro - KWU. D projeco desigt e objectivR& Th s e prestressea nwa tth f o e d cast The engineering design showed that many input parameterse needeth r fo d iron pressure vessel e surroundin(PCIVth d an ) g reactor cell (Fig) 1 . finite-eleroent-analysi overale th f so l structure require verificatioa d n for the 200 MW thermal high temperature pebble bed reactor, based on the by measurement a wel n li s scaled test setup. This especiallwa s y modular reactor design of Siemens/Interatom with the aim to develop an required for the heat transfer from the liner of the prestressed inactive heat removal system. cast-iron pressure vessel to the natural convection cooling system in The syste f prestressemo d cast iron pressure vesse reactod an l r cell cele forward outee th an lth ro t d surfac celle th .f eo e requirement th ha o meet l d al t r thermafo s l behavior under operating, Th experimentx e si dat f ao s provide souna d d basi demonstratinr fo s e th g upset and emergency loads arising from the reactor core. feasibilit e proposeth f o y dr deducindesigfo d an ng important thermal The emphasis of the work was centered on establishing the concept of input parameter r codsfo e predictions. a totally passive system for cooling and decay heat removal. This system Furthermore t couli , e demonstrateb d d e transporteb thae hean th tca t d is based on natural water circulation in the wall of the reactor cell. e natura th fro core a mth vi e l convection l e outsidsysteal th r o t mfo e The concept must ensure the heat transfer from the core to the normal and emergency conditions up to the depressurization accident outside wall surface of the reactor cell as well as the feasibility of without exceeding the allowable temperatures for the various structural the passive system of natural circulation. Coupled with this concept is components. This holds also for a failure of the embedded cooling e vessee desigth th f lo n wall, sinc e complexitth e f o thiy s wall system. constrain e heath st transfe e reactoth re wal f froth o e corl r th o mt e For comparison, a reference test with a steel pressure vessel wall cell. proved thae differenceth t n maximui s m temperatures between steel wall e worTh d "passive s beeha "n used many time n regari s o safett d y and prestressed cast iron pressure vessel wall were smaller than aspect f nucleaso r component d systeman s s with many different meanings. anticipated comparioA . n pape ] provide[1 r s detail f experimentao s d an l In this paper it is used with reference to the International Atomic analytical results. Energy Agency. [2] The following boundary conditions had to be met in the design of the GrJo prestressed cast iron pressure vessel as reactor vessel and of the surrounding reactor cell with integrated heat removal system. reactoe Foth r r vessel: inside diameter 590m 0m inside night 24426 mm operating pressure 60 bar design pressure 70 bar test pressurer 1,1x7ba 7 7 0= design temperatur. °C 0 35 e The level of heat flow (Fig. 2) across the vessel wall was computed by Siemens/Interatom. Curve 1 of Fig. 2 shows the heat flow over the height of the steel vessel. Integration over the vessel surface yields a total heat flow of 350 kW under operating conditions. Curves 2 through 5 (of Fig. 2) show the heat flow for the emergency condition of a depressurization accident. Curve no. 4 governs the design. Its integration yields a total of 890 kW which has to be removed frovessele th m e emergenc.Th y condition with depressurization envelops all other emergenc ye sustaine b conditions e reactoo t th s y b dha r t I . only once during its design life.

z [m] (z=distonce from top of core)

nr ?.'•:• Fig: Vertica1 . l cross section Cpprestressed cast-iron pressure vesse ) Reacto(? l r cell ©Axial prestressin gslap Hea) tendobTo 1| t @ n exchanger (6) Hoop orestressing cable (7) Liner with q [kW/m] support structure ® Cooling tubes ® Section for test setup Fig: 2 Hea. t flow throug e pressurth h e vessel wall For the wall of the reactor cell the following design conditions r thiFo s natural circulatio o t nwork e differenceth , waten i s r were established: density between cood war an le mreacto th end f o sr cell mus e matcheb t d s outsidIt - e diameter shoul t withi fi e dreacto th n r buildins a g with the geodetic height of the cooling tower. The required geodetic designesteee th lr reactofo d r vessel. e coolinheighth f o tgn thi i tower sm case5 2 , s governee i , th y b d - The walls must incorporate the cooling system and, together with the amount of heat to be removed under operating conditions. reactor walls, provid e requireth e d radiation shielding. Between The require de hea th sizt f eexchangero e numbe,th e celf pipeo rth ln i s vessel and cell walls there must be a cavity wide enough to enable all wall and the max. flow of cooling water is governed by the emergency assembling operations during the construction phase, in particular the condition with its heat flow of 890 kW. wire winding machine that applies hoop prestress to the cast iron After extensive parameter studies with regar o numbet d d diametean r f o r vessel. the pipes in the cell wall and to the temperature gain of the cooling 2 pipes7 s e waterdecidewa us , t o evenli ,t d y distributed ovee th r circumference (Fig. 4) and embedded into the cast iron wall segments. 3. DESIGN OF THE SYSTEM The inside diameter of the pipes is 65 mm. Both groups of 36 alter- nate piped bottoan s p havm to n diamete i ee m collectorth m t a 0 r 80 f o s 3.1 Overall system ends of the reactor cell. The pipes connected to the cooling tower have diameters of 400 mm each. e overalTh l syste r whic mfo e therma th h l behavio s beeha rn investi- The cooling towers have been y coolindesignedr a gs a dtower s with gated (Fig. 3) consists of reactor pressure vessel, reactor cell, air cooling by the atmosphere. cooling tower and the piping between reactor cell and cooling tower. Under operating condition e hea th se reactot th los f o sr core flows Reacto2 3. r pressure vessel ) 7 (Fig, 4 , .1 throug e vesseth h s transferreli wald an l y convectiob d d thermaan n l radiation to the inside surface of the reactor cell. There the total of The pressure vessel consist thref so e components: wilW k lcooline 0 heath 35 p tu g water. - The cast iron body Under natural convectio e wateth n r wil le coolin th ris o t eg towers - The prestressing system iseparated2 n , redundant piping systems. After transferin e heao th gt e lineTh - r e surroundinth g atmospher e heath ty eb exchanger cooline th n si g towers, The cast iron body, which vessee form th e wal subdivides f si th lo l d e reactoe wateth th f ro e rlowe flowd th cel en ro t ls becauss it f o e d bottoan p mto headse int th ring1 1 od . an s higher density. Thi e spipin th par redundans f i gto wells a t . 10 of these rings are subdivided circumferentially into 3 segments. The higher ring in the region of the horizontal gas duct is subdivided into 11 segments. bottod an Botp m to hhead dividee ar s d int block7 o s each centee ;on r an surroundin6 d g blocks. The material use ferritis i d c nodular cast iron e e masseth Th . f o s blocks range from 30 to 70 tons. e blocTh k assembl s i yprestresse n circumferentiai d d axiaan l l directions by the respective prestressing systems. The axial prestressing system is subdivided in 33 tendons, each containin wire2 31 g s with diameter. mm 7 f o s The hoop prestressing syste y wirbuils 6 b i m1 ep u o t t windin 9 f o g layers of wire, diameter 7 mm, into 28 channels. Each layer is anchored separately. The materia e l prestressinth user fo d g wirs i Ultrafore a ( t chromium, molybdenu d nickean m l alloyed steel). This materiae b n ca l desiga user fo dn temperatur 350°Cf o e . The relaxation for 30 years of service time of the vessel and at a temperature of 350°C will be 13%. After prestress, the cast iron body of the vessel has compressive stresses and the prestressing system - which is multiply redundant - has tensile stresses. e compressivth Duo t e e stressee interfaceth t a s f adjaceno s t blocks, the load carrying behaviour is like that of a solid vessel. e linep r seal totae th s l inner surfacpressure th f o e e vessel. Fig: 3 Passiv. e cooling system To ensure tha majoo n t r buckling o compensatoccurt d an s r minofo e r CO CD Reactor pressure vessel (2) Reactor cell differences in the stress/strain relation of adjacent liner plates the 00 OJ) Redundant pipes (4) Cooling tower e anchoreb o a support line o s t d ha r t structur outsids it t a ee (Fig. .7) Fig. 5: Test setup (dimensions in mm) with cast iron wall (T) Electric heater and core vessel (?) Liner with liner Fig : .Honzonta4 l cross section support structur Pressur) (3 e e vessel wal ® lAxia l CD Core area ® Liner (3) Liner support structure prestressing tendo ) Hoo(f np prestressing cable (j) Block of cast iron body (5) Axial prestressing Cove) (o r plat hoor fo ep prestressing cabl © eCas t iron tendons (§) Hoop prestressing cable (7) Reactor cell reactoe blocth f ro k Coolin) cel(8 l g syste @ m Reinforced (Ê) Reinforced concret ) Cas(f e t iron profiles concrete © Box for sprinkler system (Q> Insulation ^9 Cooling tube ) Sectio(Q s tesr fo nt setup @ Refractory support Axial Tendon

Liner

liner Anchoring Block

Cast Iron Wall

Hoop Tendon

Fig. 6: Test setup with steel wall Fig. 7

To prove that the pressure vessel has allowable stresses under all loading conditions, extensive calculations with a Finite Element Method The 72 cooling water pipes (two redundant systems of 36 each) are computer code have been performed. inserted in drilled holes of the cast iron profiles. Sections of approximatel n lengti e weldem ar h0 1 yd togethe n openinge i r th t a s 3.3 Reactor cell (Fig. 1 and 4) interfaces of the rings. To have good thermal contact to the surrounding blocks, the pipes e reactoTh assembles ri wala celd s l an ha l thicknes m d 5 fro1, mf so are widened by overpressurization. Thus, by plastic deformation, good 72 cast iron profiles. There are 6 rings on top of each other with 12 thermal contact with the surrounding cast iron is maintained. elements each. The calculations performed showed, that the cooling system is very The profiles are filled with reinforced concrete between the flanges stable in regards to changing conditions. The following examples may in order to provide structural stability. prove this: With the cast iron profiles arranged radially through the wall of Air temperatur t coolinea g tower e reactoth r cel a coolinl d heaan g t removal system outside celth l f o e 20°C: Normal operating condition: is possible r exampl,fo a sprinkle y b e r system s ultimat,a e heat sinn i k Inlet temperature reactor cell 30,7°C eCnO the highly hypothetical case that all systems fail. Outlet temperature reactor cell 35,9°C u Depressurization accident: with superreflecting metal foil in order to simulate the thermal O) Outlet temperature reactor cell 49,8°C boundary conditions of a rotationally symmetric structure. 35°C: Normal operating condition: Therelectrin a s i e c heater e cor mounteplate th e th f o evesseln o d . Outlet temperature reactor cell 50,3CC Its are s subdividei a d int 4 independeno t heating zone n ordei s o t r Depressurization accident: guarantee an almost uniform surface temperature as in a real reactor Outlet temperature reactoC ° r 4 cel6 l vessel. Thermographie measurements made before assemblin e heateth g r Depressurization accident: into the test setup showed a maximum temperature difference of 4°C over Failure of 1 cooling system (36 pipes): the axial extension of the heating surface. Outlet temperature reactor cell 70,5°C The box for the sprinkler system at the outside of the reactor cell serves for the investigation of the amount of heat which can be removed by spraying the reactor cell with water in case both redundant passive . 4 TEST FACILIT TESD AN YT MATRIX heat removal systeme reactoth n i sr cell fail durin a hypotheticag l emergency condition n thaI . t case e radiath , l cast iroe n th rib n i s Durin e engineerinth g g design wor t i kbecam e apparent that many reactor cell migh vere b t y effectiv heas a e t conductors. input parameters used in the finite-element-analysis of the overall The water flowing through the 4 pipes near the inside of the reactor structure - especially those for the heat transfer from the liner of the a celconstans kep i t la t t inlet temperatur a specia y b e l cooling prestressed cast-iron pressure vessel to the natural convection cooling circuit. Because of the limited height of the test facility, natural e celd e systeforwarouteth an th l n i o rmt de surfaccelth - l f o e circulatio s propertnwa y simulate convectivy b d e water flow. required confirmatory verification by measurements in a well scaled test e tes Th te stee th setu lr fo reactop r pressure vessel wal s showi l n setup, because the temperature distribution in these components turned in Fig. 6. out to be very sensitive to changes of these parameters. Comparing Figs. 5 and 6 it can be seen, that the distance between the outside of the pressure vessel walls and the inside of the reactor 1 Tes4. t objectives cel s equali l . Thi s importan i e scompariso th e convectivr th fo t f o n e and radiative fraction heae th t f transporso t accros cavitye sth . e primarTh y objectiv o shot e tes wth s i t f thae decao e th t y heat The instrumentation consist 3 thermocouples12 f o s 3 glob, e thermo- which has to be removed from the reactor core during all operating and meters ,anemometers4 calorimeter,4 heaa d t an sflu x sensor. emergency condition n fac i e removee redundan b tn th ca sy b d t cooling system installe e reactoth n i d r cell using only natural circulation. 4.3 Test Matrix Furthermore e temperaturth , ee differen th field r fo ts loading conditions and test configurations are to be measured as well as the A total 6 tests have been performed; 5 tests (numbered 0 to 4) with conductive and radiative fractions in the total heat transfer across the the prestressed cast iron pressure vessel wall (test setup in Fig. 5) cavity. an tes1 d t (numbe wit) 5 re stee th h l pressure vessel wall (test setun i p e aquireTh d data shall serv o verifo correcet t r e valueo y th tr fo s Fig. 6). thermal conductivity, heat transmission coefficient and heat transfer In the following, some details are provided for the specific through convection and radiation used in the pre-calculations. boundary condition experimentad an s l procedures: After completion of the tests and implementation of the findings in Tes 0 ttesf Star o tp tu facilit y the mathematica f datlo modelat se wil e availabla b sl e thas beeha tn Starting from room temperature goin o t normag l operating verified by experiments and which will permit future engineering design conditions and then slightly increase the heat input to see how to be performed at a higher level of confidence. the test setup reacts to ramps in the heating conditions. In this context the test data are important in judging whether the Configuratio without e Figbu se e coven5 . th t r plates (Pos) 6 . calculated core temperature e choseth r n fo s setu f prestresseo p d cast shieldin circumferentiae th g l prestressing system. iron pressure vesse d reactoan l r cele differenar l t from those derived e celTh l cooling. systeon s mwa for a common steel reactor vessel. Test duration: 500 h 4.2 Test facility and instrumentation Test 1 From room-temperature to normal operating conditions, then depressurization accident. As can be seen from Figs. 1 and 4, the setup of reactor pressure Configuration: with cover plates Fig Pos, 5 . .6 vesse d surroundinan l g reactor cel rotationalls i l y symmetric. Cooling systen mo Therefore.it was decided to simulate a 20° wide sector with a height Duration: 1000 h of 2 m of the reactor vessel and cell and erect it as a test setup. Test 2 From room-temperature to normal operating conditions, then The test facility as shown in Fig. 5 is surrounded with 500 mm of depressurization accident. thermal insulation in order to minimize heat losses. The supporting Configuration: without cover plates Fig. 5, Pos. 6. structure especially the bottom part, is also designed as an insulation. Cooling systen mo e insidTh e surface e insulationth f o s , facin e cavit th ge covere ar y d Duration: 900 h Tes 3 tFro m room-temperatur normao et l operating, then depressurization Tes 5 tTes t setup with steel vessel wall (see Fig) 6 . accident, without cover plates. Cooling system on From room-temperature to normal operating conditions, then The surfaces of the hoop prestressing system (Fig. 5, Pos. 5) depressurization accident, bac o normat k l operating conditions. and the inner surface of the cast iron block of the reactor Cooling system on. cavern (Fig. 5, Pos. 7) were painted black in order to test Second depressurization accident without any cooling. changes in radiative properties. Durationh 0 70 : Duration: 1050 h e heatintese Th th t f setuo g s donpwa e electrically e inpuTh . t data were Test 4 Configuration similar to test 3 (blackened surfaces). based on a heat flux analysis of the modular reactor with a prestressed From room-temperature to normal operating conditions. Having cast iron vessel which was performed by Siemens/Interatoro. The computed reached steady-state the cooling system was turned off and the heat flus increasewa x anticipatee th y b d d heat lossee testh tn i ssetu p sprinkler e reactooutside syste bloce th th th f t o ma kf r eo cell (20%). (Fig s activate, Poswa 5 . } .10 d while startin e depressurith g - zation accident. 5. COMPARISON OF PRESSTRESSED CAST IRON PRESSURE VESSEL HALL TO STEEL After completin e depressurizatioth g n accident e testh ,t setup PRESSURE VESSEL WALL (TEST RESULTS) s broughwa t bac o normat k l operating conditions (steady-state) A comparison of tests 3 and 4 to test 5 is shown in Fig. 8. and thee depressurizatioth n n acciden s replicatewa t d again It can bee seen that the test setup with steel wall reaches steady without any cooling at all. state conditions approximatel tese th t s same a setu th ) e h t p timya 0 (7 e Duration: 1050 h—————— with prestressed cast iron pessure vessel wall.

t T[-q ^ accident conditions accident conditions ^- cooling at 38'C no cooling 500 -r

steel waj^ inj|de 400 -- - -»_,*•...„a.•-•*>-'-*;.;' ••» ^- *• outside '- -A.;--.

300 -- -» cavernwall .inside » test no 5

200 --

100

Timj e(h 000 100.00 CO Fig : 8 Tes. t results -J Prestressed Cast Iron Pressure Vessel wall compare o Steet d l Pressure Vessel wall Under normal operating conditions the liner has nearly the same The experiments showed that even in case of natural circulation CO 00 temperatur steee inside th th lr s o ea wall e PCITh .V cylinder blocs i k failure and failure to initiate the outside surface spray system, the 75°C colder thae steeth n l wall e reactoe insidth Th . f o er cell with component temperature t reac no e respectiv th o hd s e limit values which PCIV is 10°C colder than with the steel vessel. would potentially threaten the integrity of the combined structures. For the depressurization accidents, the liner is 70°C to 100°C Moreover, it was shown (Fig. 8) that the anticipated differences in hotter than the inside of the steel wall; this might be due to the low the thermal behaviour of a PCIV wall versus a common steel PV-wall were heat capacity of the liner. not as significant as previously thought. e temperaturTh e PCI e blocth V th f o walkf eo l itsel s 35°i f C lower From the beginning, all experiments were consistently accompanied by thae steeth n e ltemperature Th wa . le 1 reacto th 1 sho l f o o n ee srw pre-as wel s post-tesa l t computations e thermaTh . l calculations were significant differences for the two pressure vessel options. performed by two-dimensional, transient finite-element analysis for the instrumented section at midheight of the test facility. The 2D FE-model . FINA6 L REMARKS was developed by accounting for standard material properties, including view-factor options as well as modeling thermal contact resistances by The most important result of the test series is the demonstration the "slide line" feature of the TOPAZ-computer code. In order to obtain thae prestesseth t d cast iron pressure vesse n combinatioi l n wite th h best-estimate heat transfer coefficients at the surfaces facing the reactor cell and together with the passive, natural circulation decay cavity, pre-test computations were performed with the CFD-code BASSIM heat removal system can sustain severest, long-term accident conditons. prior to any experiments. Therefore, these experiments prove the feasibility of such a concept for Fig. 9 summarizes the evolution of the values applied for the 0 MWta20 h pebbl reactord be e . contact resistance differene th r fo s t experiments. This figure clearly

CONTACTRESISTANCE |M»2 *K/W1 CONTACT SURFACE 1 ( Cale3 . .No Exp. No. 1 (2) Exp. No. 2 (3)) (4 3 Exp . .No ) (5 4 Exp . No . pro post era. post pro. post pro post pra post

UNEFV 1.35E-2 90E-4 9.0E-4 90E-4 906-4 50E-3 50E-3 50E-3 50E-3 50E3 UNER ANCHORING BLOCK

UNEH ANCHORING BLOCK/ 296E-3 2.0E-4 2-OE-4 2.0E-4 20E-4 10E-2 10E-2 10E-2 10E-2 10E2 CYUNDEHWALL

CYLINDER WALL/ 092 092 092 092 092 092 092 092 092 092 AXIAL TENDON

CYUNDERWALiy 197E2 2.0E-4 2.06-4 20E-4 20E-4 2.5E-2 25E-2 25E-2 25E2 25E2 HOOP TENDON

HOOP TENDON/ BOE-2 40E2 COVER PLATE

AIR SPACE BETWEEN 135 10 1.35 6 1 1 to VESSE CAVERD LAN N WALL

(1 Tes) t calcula**! befor INWA-EXPERIMENT-SERIEe e(h S (2) Accompanying calculations before and alter the experiment 1 (12 4 91-24 5 91) (3) Accompanying cafciiaBons before and alter the experiment 2 (14 6 91-22 7 91) (4) Accompanying calculations befor afteexperimend e eth an r 91-241 9 0 (1 ) t3 091 (5) Accompanying calculations befor afte experimend e ean th r (71t4 1 ) 91-92 61 Fig : 9 Evolutio. f Contaco n t Resistances user 2D-Computationfo d s with TOPAZ indicates that experiment 1 sthroug 4 werh e well representen a y b d PASSIVE HEAT REMOVAL EXPERIMENTS unique set of coefficients. Changes in the coefficients were needed as par f modeo t l refinement (start-up experiment r coverin fo wels a )s a l g FO ADVANCEN RA D HTR-MODULE REACTOR changes made in the facility (with versus without cover plates PRESSURE VESSE CAVITD LAN Y DESIGN shielding). All pre-test predictions were on the conservative side with the larges start-ue tth discrepancr fo p % experiment30 f yo . . WOLFL . KNEERA , . SCHULZR , , e experimentaTh l data provide n extensiva d e data basr derivinfo e g proper input value e fractionsth alsr fo of radiativ o s d convectivan e e A. GIANNIKOS, W. HAFNER heat transport across the cavity. Baltelle Europe, More details are provided in [1], Frankfurt, Germany The generated data base can be used to validate other computer codes applie r similafo d r reactor cavity design d theisan r optimization. Abstract Acknowledgement This project was financially supported by the German Federal Ministry searce Inth farther hto r reducin residuae gth l rid possiblf ao e major reactor pressai id failure for Research and Technology (BMFT). «taring and In the aftermath of severe accidents of modular HTR's, an alternative RPV has been designed and • ample rani already fabricated by the dim aempefltamp, Krefeld, FRG. Tbk alternative RPV or more recently also with drcnmfer«BtiaBy Bat band prestressmg. REFERENCES This «pedflc Stempehamp design has been toted and quitted la * geries of experiments with the maple tot veswL TU» design WM «too used fiir the coatro! gas vessel ta tbe THTR under operational [1] Passive Heat Removal Experiments for an Advanced HTR-Module Reactor »errkt condMoni. Pressure Vessel and Cavity Design, U. Wolf et al., these proceedings la «der to demonstrate reliable decay heat removal mder meet term conditions, • 1 : 1 scale, 20' Descriptio] [2 f safeto n y Related Terms, Draft International Atomic secto»tssel/carltje th f ro , termed DrWA-fadHty (Inactive decay heat ratant) fabricatem m tested dan d Energy Agency, December 1988 at Stempelkamp. [3] Integrated design of prestessed cast-iron pressure vessel and The cavity was cooled by natural drenlatton of water flowing la tubes embedded in die cast Iron passive heat removal systems for the reactor cell of a 200 MWth structure of the cavity. reactor, B. Beine et. al-, Proceedings 10th International Conference A total of 5 experiments were performed with thta setup «»«»«hifaifl a variety of changes In Smirt Seminar Smal Mediu& l m Sized Nuclear Reactors 1989. constructive details, surface and cooUng conditions. Each experiment was performed both fbr operational condition depreanrhatiod an s n transients MW0 , 20 typica A* HTR-modaler fo l . Experimental test durations ranged np to 1000 hows* Pre- and post-test predictions with the FEM-eode TOPAZ accompanied the INWA test series. This paper describe INWA-CadHte sth experimentae th d yan l remit predictive wels th s a s la e capability of the TOPAZ-code by comparing »he data with computational remits, The INWA-resnlts qualit pre-stressee yth d cast faon vessel together wit naturae hth l drenbtion cooled cavity even for the wont of severe accident conditions. Even to the case of a total faBnre of an cooling cavite capabilitieth t ya structurr o a st vessee eth l surface temperature remains below critical vaines. . INTRODUCTIO1 N Since 1973 desige th plannin,d nan Germaf go n pebbl higd ehbe temperature, gas-cookd reactors swa accompanie alternative th y db e reactor prearare vessel design usin concepe gth prestreasef to d cast-iron (PCI) vessel pj. In 1975, developmental work started and to On following years, it was demonstrated that Pd-vtsse * 0 operateCe b 35 n o t b ca p du The required materials for cast-iron, Huer and wfacs were systematically ««"»«"^ by material testing program documentedd san . Both e controtb , s storagga l e THT Urge-Kalth e tan r th s welR a kfo s tes a R l t eHT vesse t a l StempeDcamp, Krefeld, FRG, Impressively demonstrate predictabilitye dth e th s feasibilite a th ,n we s ya UcenslblUt authoritiee th y yb s accordin standaro gt d code regulationsd san . Public acceptanc f futureo e nuclear power seemingly necessitates increased applicatio f passivno e design features without resortin activo gt e components additionn I . , emphasifartheo t t pu ro st reduc e remote residual risk contributions, such as pressure vessel failure. The prtslressed cast-iron pressure vessel combined wit passivha e heat «Ink capability Integrated Intreactoe otb r aflo structure shows CO promising passive and forgiving tolerant features satisfying a number of presently dkcassed requirements (O of nuclear system futuree th f so . In tot SiempeDtamem 1987Di e ,th p propose ] sucd[2 hcombinatioa reliabla f no e pretmre vessel construction with aw advantages of a passive decay beat removal system fully Integrated toto the reactor silo cell structure. TU» proposal WM based on the design features and overall measure« and characteristic* developed by SŒMEN&KWD (INTERATOM) tar the 200 MW» HTR-modnlar reactor design, with the substitution of the PCI-vessd for the common steel Tend and to replace the silo structure by the comporte cast- Iron/concrete ttrnctare develope Stanpelkamy db p [2,3] describes a , followinge th dn i .

2. INWA-TEST FACIUTY

giveFig1 . perspectivsa e vie DfWf wo A Test Facility which corrtsponce MW0 secto* 20 *a 20 f o rso t Slemens/KW modulaR UHT r pebbl reactod ebe r design tese t.Th fadUt contlnicte» ywa d wtt detaUl hai s of a realtte PU-vetsel and cast bon/concrète composite structure wtth circumferential tendon reinforcemen scaa t L kta li The height of the sector was chotea to be about 2 m • only a fraction of the total axial system tarai drcnlation condition m the embedded tabes had to be sbnnkted accordingly by appropriait (breed convection condition. Also, ti lsecto » rsasft f o ir shap looe eth p ] old not exert the pre •forceoe nth Pd-vcstd stricter imdes «a r normal operating conditions. However experimentae th , l stractars ewa solidly clamped ont floore oth . In orde provido rt ntthnatn ea e heat sin easkm f faUm-eo botf eo h redundant, natural dradation piping ijiHaii. a flhn cooBng device, see 10 m Figure 1, was Installed hi order to rinse down on the rccfl composite surface* The complete INWA-fadHty was embedded within 500 mm duck thermal insolation. Abo, as shown In Figure 1, the supporting base structure was carefully designed for Us isolating properties. Super reflecting metal applie s fo floorOe wa di t d,a cellin sktefacd reactoge an di f eo r cavtt ordei yh providro t e proper thermal boundary conditions. A Hdiesf o e precautionary measures were necessar keeyo t p tike beat fro INWA-facfllte mth possiblem w lo s ya . Base experiencen do s elsewhere emdaheaf e o ,th tk lro

The drMsg thermal power is provided by the electric heater ekment, see Item 1 ta «gore L This heatmg element constate Independentl4 f do y controlled healing drcnlt assuro »t snrmcea e température as uniform as possible by properly accounting for the end heat losses. Thermographie control photos Indicated mazhnmn temperature derUUom of only 4 *C located towards the beater plate sides.

. MEASUREMEN3 T SYSTEMS In order to reach the tan objective» of the INWA-ttsta, the complex -dta stncta Instrumented wtth a total of 123 embedded and sorhc* thermocouples, the majority of which wer* placed alon radiae gth l eenterltai sectore th f eo . Furthermore obtaio ,t n temperature head san t Bnzes acrass gaps Electri1 c heate cord eran vesse Cas7 lt iron reactobloce th f ko r cell and contact Interfaces between components, die emphasis hi the instrumentation was pot on the positionin saffldena f go t numbe thermocouplef ro relevane di t sa wire t di surface ei h packages d san . 2 Liner Coolin8 g system Furthermore properlo ,t y evaluat heae edi t loss numbea , thermocouplef ro « wer placeo eab d towarde sdi 3 Pressure vessel wall Reinforce9 d concrete boundary sides of dw facfflty. Special Instrumentation were positioned tasU cavtte eth ordeym determino rt e die temperature 4 Axial prestressing tendon 10 Sprinkler system distribution, velocity fields and radiative transfer. For tab purpose, 3 globe thermometer», 4 turbine 5 Hoop prestressing cable 11 Insulation anemometer* and special thermocouples were installed at movable rakes. 6 Cover plate calorimeter4 s were Installed tot embeddee oth d cooling tube«. Afte firse rth t preliminary experiment, the Importanc correcf eo t heat flnx control became apparen speciaa d tan l heat flue xdi sensot a s rwa heater plate surface facing die Huer component This allowed die redondant and diverse means to control the heat fmx. INWA-TEST-FACIUTY All measurement signab were sampled, converte documented dan PC-basea y db d measuring system. 4. OBJECTIVES AND TEST MATRIX 5. EXPERIMENTAL RESULTS Figur «bowe2 • shorizonta t througco l INWA-fadUte hsamth e th et a Orne d yan indicatee th s A typical temperature trace at positions indicated in Fig. 2 fc shown in Fig. 4. After start-up which nodallsatlon used for the thermal analysis with the FEM-Code TOPAZ [4]. Abo, this figure reveab the Includes 'overheating1 to compensate for the thermal capacities of the components. Steady state to reached variety and complexity of heat transport processes occurring la the different parts or the PCI-vessel and after Depresnrizatlo . abouh 0 t10 n event (DPI then startetimee )ny ca b showt an .t A d a Flg.n E i 4DP composite reactor cell structure. starte about da t 410 lasted han d approximately 250h. With these informations as background, the objectives of the INWA-ezperimenta were as follows: Later presented température profiles refb (1) Demonstration of reliable and predictable passive heat sink capability under operational and Fig demonstrate5 . radiae sth e lt nperatarcF ffle alon centerlme gth INWe th f eAo Test Facilitys .It accidental reactor conditions without endangering (thermal limits) individual componente th d san origioutee th s nrI surfac heat-platee th f eo . integrity of PCI-vessel structure. tese th t s facilitA modnlarls yI y constructed temperatur e stepe th th ,i sh e profite resul therman tI l (2) Determinatio f appropriatno e value thermar sto l resistances, thermal conductivities (axiad an l resistances at the different contact surfaces win* m the gaps. circumferential package cablf so e wires) head an ,t transfer coefficient Tariont sa s interfaces. Fig present6 . s measured axial temperature profile differenr sfo t circumferential angleoutee th t rsa (3) Evaluation of the importances of convective and radiative heat transport mechanisms In gaps, surfac heat-plate th f eo inne d ean r surfac Huere th f .eo Whit heat-plate eth e profite demonstrates onlya most importantly acros cavitye sth . small dependenc heighn ei circumferentiad tan l direction, tile finer show variatiosa 25*f no heighCn i t (4) Determinatio effecte ai f nchangef o so coastracttrn sI e details (addltk protectiTecorerphtta f ao e This results from natural convectiobetweep ga e heae n th line th e heatn d I th plat ran d t losseee an th o st shielding the drenmlercntial cable package, surface treatment by adding a Mack coaüng onto the bottom i surfaces facing the cavtty). Fkj. 7 to 12 present Measured temperature profiles through the Pd structure and cavity for (5) Evateaikmir impace fth tt^faitorf to natu^chxahtte th f eo ^ steady-date normal operatio depresscrhatiod nan a transien experimentr tfo Thes. 4 d se «a (0, figure3 )2, s ovtiide surfac cootta n thermae th eBu n go l iisuoasi PCI-stractnree th f so composite th d sso e show the résulta ofafl experiments without protective cover plate shielding tin hoop prestressug cabks. reactor cell structure. The comparisons between «dividual carves m the respective figures dearly demonMrate the limited A total of 5 expertneats have been performed and evaluated thus tar, testing the PCI-Teasd structure variation in the experimental results, indicating consistency and phmsfbttoy of the data despite of under différent condittoas dted «bore (Tab. I) . thousand hourf so operatiof so INWe th f nAo test Bkdttty. Under normal stady-state operation maxtaom referencA e test wit hsectioa typicaa f no l steel vesse presentls lI y underway. Upo completions nIt a , temperatures at the Bner surface range around 275 *C (Fig. 7). direct compariso demonstratiod nan versoI PC f nso steel vessel behavior wff possiblee b l . The radial temperature profiles acros cavite sessentiallth e yar y flat (Fig. #10), with steep gradients AB experiment» were performed usin deplete heae w gth tDo Flgnr dcontrollinm r fo e3 heatee gth r at the PCI vessel and the Inside cavtty surfaces, «~«--«**«t a natural drcntattng How patten along tin plate in order to reach steady-state operational conditions and to simulate accident conditions according structural surfaces wit hlargea , ffsrntiaBy stagnant core. Change différense madth r eto t experiments, to the predicted behavior during a depressarhtatk» accident of the modular HTR-reactor designed by have only a small Influence on die steady-state temperature profiles. SHMËNS/KWU. The curves shown m Flgnre 3 account for heat losses (20 %) of the INWA-fkefltty.

coding tyflea 0**«ink) IEXPNO.— I

conducdo« TIME[h] i FIG. 2 Physical processe Modélisatiod san n FIG. 3 Applied hea INWA-TEST-FACILITe t floth r wfo Y £ TABLE I RADIATIV CONVECTIVD EAN E HEAT FLUX N) ACROS CAVITE STH Y CELL

EXP. NO. TIME Radiation Convection Radiation Convection h W/m2 W/m2 % %

0 190 764 61,3 92,6 7,4 u § * 400 378 185 67,2 32,8 1 650 1170 334 78,0 22,0 330 701 61,4 92,0 8,0 2 530 1744 105 94,3 5,7 513 708 55 92,8 7,2 3 750 1786 91,6 95,1 4,9 Radial coordinate from heat plate [mm] 250 672 53 92,7 7,3 4 450 1722 54 97,0 3,0 950 1712 49 97,2 2,8 FIG. 5 Radial profile throug test-facilite hth y Experiment No.2 / Normal operation

e I s 8.

H

Height [nun] TIM) E[h

FIG. 4 Time history of temperature FIG. 6 Vertical temperature distribution at the heat plate and the liner Experiment No. 2 Experiment No. 2 / Normal operation u

I

Radial coordinate from heat plate [mm] Radial distance from cylinder rib [mm]

FIG. 7 Normal operation FIG. 9 Normal operation Cylinder block temperature profile Cell temperature profile

t-SKti 3.1-7*»

u

« I 8. «c H

Radial coordinate from heat plate (mm) Radial dislance from cylinde [mmb rri ]

£ FIG.8 Depressurization event FIG0 1 . Depressurization event Cylinder block temperature profile Cell temperature profile Maximum liner temperatures nach it value of about 470 *C (Fig. 8). For experiment 4 the radial temperature gradient decreases towards the cylindrical rib of the outside PCI vessel surface because thb experimen performes twa d withou naturae tth l circulation syste composite th mn i e reactor cell structure. The simulated failure of this system leads to an increase of the Inside cavity surface temperature and as a result to a reduction in heat transport across the gap. This can be clearly seen from Fig. 10 which shows that the gap temperature level Increases from around 150 *C for experiments 2 and 3 up to 220 *C In experimen outside Th . t4 e film coolin cavite th f gyo reduces cavity température arouny sb d *5 Cd1 an does not have a significant influence on heat removal (Fig. 12). During experimen zina t1 c coated cover plate, shieldin hooe gth p prestressing cabl mounteds ewa t I . u Is obvious that the cover plate represents a large heat transfer barrier to the cavity cell. Therefore the temperatur vesseI ePC llevele Increaseth n si considerabldo t y higher values. Fig present3 1 . s this radial temperatur vesseI e PC profile e l Cablth r sfo e temperatures ris *C e0 consideratio n froI . 38 *0 Co mt 25 n of the yielding point, a high value must be set on the selection of the prestressing cable material The result f experimeno s necessitate1 t e remova e coverplatth th e n followin di f th o l r fo e g serief o s experiments. This change in construction manifested the importance of the thermal conductivity of die hoop prestressing cables beae Th .t conductivit cablee th onl n f Interpreteye o sca y b resultana s da t thermal conductivity Including radiation effects. The following effects are demonstrated by Fig. 14. The different experiments are indicated by then- number. after experiment 1 the résultant conductivity has changed to tower vaines, Radial distance from inner cavity wall [mm] the resultant conductivit dependino yt temperaturn go e (thermal radiation effects), the resultant conductivity decrease* from experimen experimeno tt t which result smala i - sh lde stressin bundle th f go e structure. In accordance to Fig. 14 an averaged value for the resultant conductivity of 3.5 W/m/K was assumed FIG1 1 . Normal operation for pre- and postpredktkos. Cavity wall temperature profile Table I summarizes the evaluation of the special measurement tnrhniqiifa applied ht the cavity for ate determinatio dif no e respective fraction convecttvf so radiativd ean e heat transport acros cavitye sth b .Th table Indicate overwhelmine sth g importanc radiative th f eo e transport (92-9 versu) 7% convertir e sth « transport (3-8 %), with flic exception of INWA experiment No. 1 which was performed withna cover shielding plate protectin wire gth e package. Fig show8 1 . heae sth ttest-facilite looseth r sfo steadt ya y state conditio experimenr fo d nan . t0 The losses are at the vessel about 1235 W, at the cavity cell 335 W and at the cavity wan 200 W. The total losses have been evaluated to 35 % of the heat input . ACCOMPANYIN6 G NUMERICAL ANALYSES

All INWA-experiments were accompanie computationay db l analyses wit FEM-code hth e TOPAZe .Th TOPAZ-cod extendes ewa includiny db radiatioD g2 n properties usin FACET-methodologe gth o t ] y[5 derive the multi-dimensional view factors. The nodalizaflo TOPAD 2 ne grith Z f dcomputationo shows si Figurn I samwhice 2 th et h a tim e Indicates the position« MPI through MP6 for the comparisons between the data and the open post-test predictions show Figurn i . e15 Figure 15 to 17 show for INWA-experhnent No. 4, respectively, Oat after a substantial learning period [3], the optimization of the code input parameters result In excellent agreements between measured datcomputed aan d results. Even worst-case scenarios, with failur l coolmal f eo g systems testes a , d during predictee b w no experimen dn witca , h4 . confidenctNo depictes ea Figurn dI . e15 7. CONCLUSIONS Radial distance from inner cavity wall [mm] The large-scale INWA-experiments have achieve l theidai r objectives wit hhiga h leve confidencef o l . Most Importantly they demonstrate passive dth e heat removal capabilit combinatioe th f yo PCI-vessef no l and composite reactor cell structur specifiee th r efo d conditions examine i fuldh l scale. FIG2 1 . Depressurization event Detailed Insights obtained from the experimental data allowed the proper specification input Cavity wall temperature profile parameters for TOPAZ code predictions, that were not readily available previously. S IL I

Radial coordinate from beat plate [mm] TIM] Epi FIG. 13 Depressurization event FIG. 15 Comparison of measured and calculated data Cylinder block temperature profile comparison Experimen4 . tNo

i -r si* A i ^^,^— - - 415» B ^^~ r • 02 f lib E ~^*r\i Ï4> P **' -S^ 3 )lk C

27fc H

t 2**1 .

20»r 1 2 . 4 6

CONDUCTIVIT Y [W/nVK ] 17V J

MO. L

104= M

7l» W 0 100. ZOO. 300. 400. 500. 17- O TEMPERATURE pC] _,. FIG. 14 Resultant thermal conductivity of the hoop prestressing cable FIG. 16 Depressurization event cn Calculated temperature distribution Measuremen evaluatiod an t n techniques were (ally teste t higda h temperature krei durind san g -Ji. hundred hourf so eacr sfo h experiment O> The experiences gained from these experiment wely lsma supplemen desige tth n need reactof so r cavity S* A cooling systems developed elsewhere. Hie facility is available for more and possibly different experiments 47k B In the framework of an international cooperation. .71. C

44V D

419» E REFERENCES Mfc P [1] Batteiie-Instltut e.V., "Feasibility Study of a Prestressed Cast Iron Reactor Pressure Vessel, (In 3«. O German), vol I and IT, 1973, 1974 J4t H Beine. B ,[2 ]"Integrate d Desig Prestressef no d Cast-iron Pressure Vesse Passivd lan e Heat Removal nfc I System for the Reactor Cell of a 200 MWtt Reactor", Proc. 10th SM1RT Post Conf. Seminar on

l h « Smal Mediud lan m Sized Nuclear Reactors, Anahelm USA, ,CA , Aug. 1989 [3] B. Berne, 'Large Scale Test Setup for the Passive Heat Removal System and the Prestressed 30. K Cast-iron Pressure Vessel of a 200 MW* Modular High Temperature Reactor", 3rd Intl. Seminar 2Jfc L on Small and Médium Sized Nndear Reactors, New DeH, India, Aug. 16-28,1991 2l* H ] [4 A.B. Shaptro, "TOPAZ Finit:A e Element Heat Conduction Cod Analyzinr efo Solids"D g2- , 114. H University of California, Lawrence Uvermore National Laboratory, Kept UCID-20045 (1964) [5] A.B. Shapiro, "FACE RadiatioA T - n View Factor Computer Code for Axbymmetrlc PlanarD ,2 , ISfc O anGeometrieD d3 s with Shadowing", Universit f CaWbrnlayo , Lawrence Lhtrmore National Laboratory, Kept. UdD-19887 (1983) FIG. 17 Depressurization event Calculated heat temperaturflud xan e distribution

5000 W

300 W

1235 W 335W 200 W = 1770 W = 35%

FIG 18 Heat Losses Experiment 0 TEST APPARATU COOLINF SO G conduction and convection. The computational code Is applied to the test PANEL SYSTE MHTGR MFO R apparatus, and the numerical results are compared with test results to verify the validity of the code. . TAKADAS . SUZUKIK , . INAGAKIY , , This paper describes the test apparatus of cooling panel system and its Y. MIYAMOTO test plan. Departmen Higf o t h Temperature Engineering, Tokai Research Establishment, Test apparatus of cooling panel system Japan Atomic Energy Research Institute, Tokai, Ibaraki, Japan A flowsheet of the cooling panel test apparatus Is shown in Fig. 1. The Abstract test apparatus consists of a test section and systems for water supply, heliu r vacuums supplmfo ga tes d e yan tTh . sectio composes nI pressura f do e A test apparatu f coolinso g panel syste mprovides I Investigato dt e heat vessel containin n electriga c heate coolind an r g panels surroundine th g rénovai performance and temperature distribution of reactor pressure vessel. pressure vessel as shown In Fig. 2. The test apparatus consists of a test section and systems for water supply, The pressure vessel, 1000m n diametermi . 3000m heighn mi d 12man tn m i heliu supplys Bga , vacuu instrumentatiod man nt control tese tTh . sectios nI thickness, is made of stainless steel, and 19 simulated stand pipes are composed of pressure vesseK 1000mm In dianeter. 3000«» in height ) containing locate e uppestane th Th n dr do . pipeheait alse f do sar o mad f stainleseo s an electric heater( Bax.lOOkW, 600 *C) which simulates a reactor core and cooling panels surroundin pressure th g e vessel. Heliue argoar d s man nga used and gas pressure is varied fro« vacuum to l.OMPa in the pressure vessel. In the test, temperature distribution of the components In the test section, helium gas temperature, heat flux removed from the electric heater and heat flux transferre cooline th o dt g panel acquirede sar .

Introduction •Skfe Cooling ftirel In Modular High-Temperature Gas-cooled Reactors (MHTGRs), one of the © features Is a passive cooling system which is applied to remove decay heat issure Vessel from a core with thermal radiation and natural convection on a reactor pressure passive vesselth r eFo . cooling system systemo tw ,bees sha n designed surfaca : e cooler (Interatom reactoa d an ] r [1 cavit) y (General Vacuum pump Atonies othee ) th [2]n rO .hand coolina , g panel syste alss ni o installen dI Helium gas the High Temperature engineering Test Reactor (HTTR) in JAERI as a back-up of -£ÄÄ£tt sjppV -tower CoolinL g auxiliare th y reactor cooling systes heaIt td mremovaan l capacit 300kW[3Js yi . Föne! L A test apparatu s providei s o investigatdt e heat removal performance Vfater fl- and temperature distributio f pressuro n e vessea surfac r fo l e coolea ( r II- cooling panel) system. Rosvmeter A computational cod s alsi e o develope calculato dt e bot e flod hth w an e temperaturth e distributio f o componentn s with thermal radiation, Fig- 1 Flowshee cooline th f o t g panel test apparatus The electric heater, 600mm in diameter and 2000ma in height, simulates CO reactor core s maximuIt . m surface temperatur powed ean e 600*Can ar r d lOOkW, respectively. The electric heater is divided into 6 segments, 7kW heaters for top and bottom segments and 21kW heaters for middle 4 segments. Each heater segmen annulan a s i t r type block madceramlcs(90a f o e % AlaOs, 6% SIOzd fillean ) d with thermal Insulator nlchromd an , e helical coile ar s wound aroun e blockdth . The cooling panel is installed around the pressure vessel, and the cooling tubes remove the heat from the pressure vessel. The cooling panel is compose f thredo e parts uppere th : , sidlowed ean r cooling panels eacd an h, coolin2 1 d coolingan 8 tubes8 g, pane25 , s respectivelyha l . Figur showe3 s e upperth , sidd lowean e r cooling panels e coolinTh . g tubes, 31.8mn i m diameter, are made of low alloy and connected in parallel between two ring type headers. The pitch length between each cooling tubes are 60mm. Cooling wate s suppliei r d watefroo mtw r supply systems. Maximum total flow ratf eo water Is 10.0m3A In the cooling panels. In order to unify the emisslvlty, black paint Is coated on the outer surface of cooling tubes and the pressure vessel. The insulator, 2210B n diameterI B , 4000m n i heighm d 100man t n i m thickness, surrounds the- outsid coolinf eo g tube preveno st effecte th t f o s circumstantial disturbances. tese Inth t section cooline th , g panel remove heae sth t fro electrie mth c heater with thermal radiation and natural convection of helium gas (inside the pressur r (outside)eai vessele th d . an ) Temperature pressure th f o s e vessel surface heaterse th , e coolinth , g tubes and the insulator are measured by sheathed chromel-alumel thermocouples numbee Th . f thermocoupleo r s attache pressure th n do e vessel f 0.5mmf o 1.0m diameten o i d thermocouple 0 surfacm5 1 an , 14 d an s ri e e ar s Fig. 2 Schematic drawing- of the test section divided 4 series at 90' intervals around the surface as shown in the Fig. 4. Alse numbeth o f thermocoupleo r s attachee electrith n o dc heaterse th , steel and Is filled with thermal insulator. The stand pipes prevent the heat , respectively17 coolin d an , ginsulatoe 12 tubeth , .d 22 san e rar transfer with thermal radiation and natural convection from the top of the In addition, temperature f coolino s g watee measureheliud ar an rs mga d pressure vessel. Stand pipe replaceable sar investigato et effecte eth f o s by platinum resistance bulb sheathed an s d chromel-alumel thermocouples, stand pipes. The pressure vessel is supported by four legs, and four boards respectively e floth w watee d ratth an coolinn ,f i reo g tube e measurear s d attached betwee lege nth s simulat skirea t type support. Heliu pressurs mga e by magnetic flowmeter s showa s Fign ni . .1 Is varied from vacuum to l.OMPa In the pressure vessel to investigate the Electric powers of heater segments are measured by power transducers. effect heliuf o s s pressurmga e which affect naturae th s l convection. Upper coolling ponel Thermal Insulation & Cooling Panels Lower cooling panel 01000 180'

Stand Pipe

180* 7

Ring type header —— 1— —' — — fr-w Cooling tube 700 950 700 Pressure Vessel / f=Sa 5 194 0 3 20Î i 205 Side cooling panel 1883 Ring type header

60

Cooling tube 180

Fig. 4 Measuring points of temperatures in the test section

Test plan

Test e conductear s o investigatt d e heat removal performance froe mth electric heatepressure th o rt ecooline vesseth d gan l panels testse th n ,I . the following results will be acquired, - temperature distributio e componentth f no s (pressure vessels, heatert se al.) in the test section, CO Fig 3 . Schematic drawin f coolingo g panel - helium gas temperature. - heat flux removed fro segmente m th electrie th f so c heaters, widely, and to verify the validity of the computational code In the wide Öl o - heat flux transferred to the cooling panels. densitys rangga f eo . Main parameters of the test are boundary conditions on the heater surface, gas pressure In the pressure vessel and the stand pipes. The main (3) Stand pipes parameter tese detaith n I t f dlscrlbee o sar l followss da . necessars i t I investigato yt effecte stane eth th f do spipes , because eth stand pipes preven e heath t t transfer with thermal radiatio naturad nan l (1) Boundary condition heatee th n so r surface convection from the upper head of the pressure vessel. Then, the two types The electrical heate s controlleI r controo a tw s I y db le modeson e th : pressure oth f e vessel witwithoud han stane th t d pipe usede sar . surface temperature control mode which makes boundary conditioe th f o n In the test, boundary conditions on the heater surface, a kind of gas and temperature e othea poweth s I d rr an ,contro l mode which makee th s gas pressur pressure th n i e e vesse e changear l investigato dt effecte eth s boundary condition of the heat flux on the heater surface. of stand pipes. The temperature distributions of the components, especially In the test of the surface temperature control mode, temperature of the around the stand pipes, and the heat removal performance of cooling panel electric heater surface Is changed from R.T to 500 *C. In the test of the are Investigated. power control mode, power of electric heaters are changed from 0 to lOOkW. The shape of the heat flux distribution on the heater surface is changed: Conclusion uniform, cosine and exponential shapes. Both boundary conditions affects temperature distributions of the The test apparatus of cooling panel system is provided to investigate components. The numerical results are compared with the test results In heat removal performanc d temperaturan e e distributio e pressurth f o n e order to investigate the heat removal performance with natural convection vessel. and thermal radiation froe surfacee beateth mth f o rs segmente th o t s In the test, helium and argon gas are used and gas pressure is varied cooling panels. from vacuu o l.OMPme pressurt th n i a e vessel. Temperatur surface th f eo f eo electric heater e changear s e electrido 50t th fro 0T d XR. «an !c powes i r pressurs (2 Ga e pressur) th n ei e vessel changed to lOOkW. And the two types of the pressure vessel with and without e testIth n , heliu s pressurmga s varieei d from vacuum e o l.OMPt th n i a the stand pipe usede sar . pressure vessel In order to investigate the effects of the gas density which As the results of the test, temperature distribution of the components in affects natural convection in the pressure vessel. the test section, helium gas temperature, heat flux removed from the In the case of vacuum condition, the test results are compared with the segments of the electric heaters and heat flux transferred to the cooling numerical result orden si investigato t r e geometric factor emlssivitied an s s panel e acquired e numericaar s th d An . l result e comparesar d wittese th ht which affect thermal radiation between the electric heater and the pressure result orden si verifo rt validite ycomputationath e th f yo l code. vessel. The test results varying heliu pressurs mga pressure th n ei e vessee ar l At present preliminara , y tes conductes I t checo dt k characteristicf so compared witnumericae hth l result orden si confiro t r validite mth e th f yo the test apparatu d instrumentationan s . Also e computationath , l cods i e computational code which calculate the heat removal performance with developing In correspondence to the test schedule. thermal radiation, conduction and convection. Argon gas is also used and gas pressure is varied from vacuum to l.OMPa ACKNOWLEDGEMENTS In the pressure vessel, because the gas density of argon gas Is about ten tines higher than that of helium gas under the same pressure. Then, it is The authors would like to thank Dr. N. Wakayama for his helpful suggestion able to investigate the effect of the gas density to the natural convection and comments. REFERENCES SANA EXPERIMENTS RELATED TO SELF-OPERATING REMOVAF LO [1] H. Frewer et.al.,"Tlie Modular High-Temperature Reactor," Nucl. Sei. Eng., DECAY HEAT 90, 411-426(1985) [2] USNRC, "Draft Preappllcatlon Safety Evaluation Report for the Modular H.F. NIESSEN, M.G. LANGE Institute for Safety Research and Reactor Technology, High-Temperature Gas-Cooled Reactor," NUREG-1338, 5-26(1989) Forschungszentrum Jülich GmbH, ] "Presen[3 t Statu f HTGso R Researc Development"& h , 21(1988 Japanesen (i ) ) Jülich, Germany

Abstract Simulatin lose f gcoolanth so t acciden experimentn a n i t reactoe th , r conditions matchee havb o et d closs a possibles ea establiso T . principle hth f self-operatineo g heat transpor decae th r y fo t heat removal, a demonstrating experiment (SANA II) is in preparation. To prepare such an experimental , whicup t h se distinctly exceed laboratore sth y scale preliminara , y test facility SAN bees ha An1 built up. For SANA-II a horizontal section of a reactor in full scale will be setup, including a sector of the core, the reflector, the vessel and the primary cell, where the decay heat production is simulated withi coree nth .

1. Introduction

Concernin safete gth modef yo m reactor concept removae decasth e th f ye lo heaon s i t majoe th f ro aspects reliabilite Th .heae th t f yremovao ensuree lb systeo t s dmha under all accidental circumstances. Detailed design studies of modem gascooled reactors e.g. HTR-Modul show, that this demand can be fulfilled for the loss of coolant accident withou activy tan e measure withoud san t exceeding temperature limits for fission product retention (about 1600°C ) only by self-operating heat transport phenomena like heat conduction and thermal radiation. Besides the results of analytical and numerical investigations, which show the qualification of the self-operating heat removal, a demonstration experiment ( SANA H) is in preparation. For this test plant horizontaa l sectio reactoa f no fuln ri l scale wil setupe b l , includin e gsectoa th f o r core reflectore th , primare vessee th th ,d lyan cell, wher decae eth y heat productios ni simulated within the core.

. Concep2 Desigd an t n Consideration Tese th tr Facilitsfo y SANAI I

Simulatin lose coolanf gth so t acciden experimentn a n i t reactoe th , r conditions have to be matched as close as possible. Following the results of theoretical studies (fig. 2), in the first hours - the reactor power is reduced to less than 1 % - the distribution of temperatur reactoe th n ei changes ri directioe th n di heatinf no coldep gu re partth f so core. In the experiment the starting temperatures are those of the calculations after 2 hours 01 N) HTR Preliminary Experiment from accident initiation and the power of about 1 % of the specific reactor power per volume . SANAI A question of design is related to the number and the distribution of the electrical heaters (fig. 3/4). In the core the heat source is nearly homogenous, resulting in a parabolic temperature profile in a vertical reactor cut. For the electrical heating the heat source is concentrated at certain singular places. In fig. 3 the temperature profiles ° cutou90 fo a r t wit heatinh6 g element t betwee showne cu ar s a heatine n th I . g elements the temperature profile of the test facility would be in good agreement with the temperature profilcoree th f . eo Onl ysmala l temperatur ecentee dron th i n s pi ri disagreement. A smaller cut e.g. 45° results in less good agreement and the central temperature drop can only be avoided by a 360° arrangement (fig. 4).

In fig. S a vertical cut of a 90° arrangement is shown. A pebble bed of 1m in height Demonstrating Experiment an radiuda f 1,5so mheates i electricay db l resistance rods. Graphit carbod ean n stone arrangements, gas gaps, steel walls, surface cooler and concrete wall, representing the SANAH typical reactor core surroundings in a scale 1:1, follow on the radial side of the pebble bed. Thick thermal insulations at the top, the bottom and the peak side of the structure are necessary to minimize thermal losses in these directions. The insulation is built up of mineral wool at the top and the side, and self supporting systems and bricks at the bottom.

Because the loss of the coolant accident being the most severe one, the plant will be operate t normada l pressure wit inern ha t gas. Nitrogen, which represents airr ,o helium will be used.

3. Preliminary Experiments in SANA I

3.1. Description of the Test Facility SANA I

To prepare such an experimental set up, which distinctly exceeds the laboratory scale, preliminara y test facility SAN bees ha steena A I builn I l(fig p . vesseu t 6) . l witha diameter of 1,5m and a height of 1,8m a pebble bed of 1m in height and 1,5m in diamete installeds ri firse th t n steI . pcentraa l heating element wit electrican ha l input vessee s Th buil i prepare s . i linstallatioe in W tth k r 0 dfo 2 f f threo no e additional heating elements e graphitTh . e tubular heating element wit outen a h r diametef o r Fig. l : Experiments related to self-operating removal of afterheat bottoe th d minimizmo t an 32m p to powe e mthickenes i e th th rt a productio p du n ni the insulation-systems heatine Th . g tub surroundes ei secona y db d graphite tubo et protec against ti mechanicae th t l load preveno st frod pebbl e electricaman n th ta d ebe l short-circuit to the pebble bed. In the water cooled domes the connections to the electrical power supply are installed. 5 -

4 -

3 -

10 50 100 hours MODUL: Decay Power (relative to full power)

KWh Integral decay power

Heat released from reactor vessel

Heat store: in d components

fue! elements hour0 10 s 0 5 10 MODUL: Whereabouts of the decay power at loss of coolant

•c_ : i». ...— ^^— < - • — ,- — b 1500- samuH«.•—— — — .^. •™- , >X- max mum •••-/•• .../.... / 1000 - •i— - ^--. -/ / x- ave rage

500- ,

10 50 100 hours MODUL Temperature of fuel elements at loss of coolant

Fig. 2: Thermodynamic behaviour of the MODUL-reactor at loss of coolant accident

153 Structure

i.. Reference case I.. Core

\o o o

1 ..

Fig. 3: Simulation of radial temperature profiles, reference case with heat released from the pebbles, Fig : Radia4 . l Temperature profile 45°a n i s180a , d °an belo sectioo wtw n throug 90°-ha segment wit heatinx hsi g elements a 360° arrangement I pebble bed. 2. rellector. graphite. 3. carbon stone. 4. steel vessel. 5. pressure vessel wall. 6 steel cover . surfac7 . e cooler . concret8 . e wall . heliu9 . m r volume ai gap . 10 . . I I insulating bricks sellsupporlin. 12 . g insulation minera. 13 . l wool.

: Vertica5 Fig. t througcu l tese hth t stand SANI AI Fig. 6: Cut through test stand SANA I

vessee moisturth d f an flasa lo s suppl r t ga hdrivai o eou T e y eth syste m witha 3.2. Test Program throughput of up to 5000 1/h N2 is connected to the domes. Especially in the start-up phase the system can be operated in such a way, that the graphite heating element is maie Th n purpos SANe th f A eo I facilit gainins yi g operation experienc criticar efo l protected against oxyge moisturd nan e desorpt fro insulatioe mth t alsnbu o froe mth components, like heating-elements, insulation-systems, measuring devices, control- pebbles. Under expérimental conditions the throughput is reduced to the leackage of system, for the SANA II plant. According to the experience of the start-up phase in the the vessel. beginnin f thigo s year afted an ,r some minor changeoperatioe th facilit e d th san f no y show7 Fig . controe sth measuremen d an l t systetese th t powee f facilitymo th r r Fo . since the beginning of May these questions, excluding the long term reliability, can be control the set value of the power is delivered from a personal computer via a RS 232 answered positively. interfac controllee th d thyristora ean o rt , wher powee eth adjusteds ri power e th n O -. side the thyristor is connected to a transformer to reduce the grid voltage to a suitable Besides these aim tese sth t facility enables investigation hean so t transport t onlno ,y value e secondarTh . transformeye th sid f o e s directli r y connecte heatine th o t dg through a pebble bed, but also through different reactor structures under stationary or element gasanalysia n I . contente sth f oxygenso , carbon-monoxid poinw de te th d ean transient conditions, e.g. segments of graphite, blocks, rods etc. of the flash gas are detected. The datas from the gasanalysis, the actual power value, ansignale dth s fro thermo-couplee mth transferree sar hybridrecordea o dt afted an rr Becaus geometrs it f eo SANe yth A tesI t facilit especialls yi y suitabl validatioe th r efo n analo digitao gt l conversio , wher PC dat e storene e eth aar th somd sen o dan t f t eo of thermo-hydraulic computer codes e.g. THERMIX, TINTE, which are used for (Ji 01 them use r supervisiodfo tese th t f facilityno . reactor simulation, /2,3/. temperature givens si temperaturee Th . holese th r t sfo pebbl f eaceo h layer decrease o0>1 liyliridrixordcr middle th 380°d bottoe n i eth an t C ma ° 0 layer51 tope o .t th , t a C ° fro0 m60 Similar results were measured at the other steady-state cases.

/52

-61 . 2 7 _71

——[cooling water)

Fig. 7: Flow of power and data in SANA I

Result3 Tese 3. th tf sRuno s ,37 The firs ttese runth t f facilitso y Januar n weri f 199d o e an 1d ycarrie en 1992 t dou . r thiFo s purpos vessee filles eth wa ld with 9300 graphit eregula n pebbleno a rn si ,76 arrangement thren i . d Inside be level e eth s (top layer (1), bottomiddld an ) me(2 layer -68 (3)) several thermo-couples were installed (fig. 8). At these levels the wall temperature measures i d too. Complementar verticae ) som(6 on t en yi cu lpebble s were provided with thermo-couples e instrumentatioTh . s completenwa thermo-coupley db e th n si insulatio t severaa d nan l outee placeth rn so surfac vessele th f eo . Additionalle yth bottoe th coolind e mheatinth an domf p o gregisteredto e p gwatee u ar th t ra .

After drying and cleaning of the system several runs with different power of the 53 electrical heater were carried out (5 kW, 10 kW, 15kW, 20KW). The resulting temperature stationarr sfo y conditions were precalculated wit code hth e THERM1X. e expecteTh casdW k etemperature 0 compare1 e th r dfo s wite experimentath h l temperature e givear s fign . i nWhil9 . e precalculatioth e n showe a verd y small buckling of the temperature in height, caused by the different insulation, the Fig. 8: Instrumentation plan of thermo-couples in SANA I temperature experimene th n si lowee tar r tha calculatee nth d ones therd strona an ,s ei g dépendance over the height. In fig. 10 a column diagram for the experimental 10 20 60

laver

layer

o 10 20 40 50 70

z G«omttrl - r 4 -> r (cm)

Fig. 9: 10 kW power, nitrogen: Calculated profiles and experimental data, numerical simulation with heatconductio radiatiod nan n

80-50 Top 0 HEIGHT / cm 100 Bottom

600 600 600 HEIGHm c 3 T HEIGHT 50 cm HEIGHT 92 cm

400 400 400 200 n 200 200 1 A Jl . Him 0 £ 0 6 0 4 0 2 0 8 0 6 0 4 20 20 40 60 80

powerW k 0 1 , Fig: nitrogen10 . : experimental dat t steada y state

157 ai Because of several indications that the different insulation types or specific operation layep to re nearlth y hav same eth ebottoe valueth d msan laye raises highea rha o dt r 00 modes of the facility had not been the reason for the buckling effect but natural level than in the nitrogen atmosphere, though the steady state is obviously not reached. convection recalculatioa , n wit THERMIX-code hth e considering convectio dons nwa e (fig. 11). The comparison shows good agreement of the calculated results and the e comparisoTh e experimentath f no l temperature calculatee th d an s d temperatures experimental data. An additional argument that under normal pressure and at least shows that the neglection of natural convection at normal pressure for a helium under certain conditions of the test-facility natural convection has to be taken into atmospher tolerabls ei e (fig. 13). acount is given in fig. 12. There the temperatures of the inner pebbles are plotted versus time. In the first part a 10 kW steady state point in nitrogen atmosphere is achieved, wit typicae hth l temperatur bottome eth droo t p . Afterwardto fro e mth e th s Actua4 3. l Test Runs test facility was cooled down (0 kW) for a couple of hours and then after changing the nitrogen to helium the facility was heated up with 10 kW. The peak temperature Since the beginning of May the plant is in continous operation. In a first series of tests dropped compared to the nitrogen atmosphere as expected. But the middle layer and thicm c k6 insulatioa attaches innee nwa facilitth e vessero e dth t s walth d yf wa an l lo

20 30 50 «o 10 S

^ $

layer

0 10 20 50

r - z C«omtlr(« ——— > r (cm)

Fig. 11: 10 kW power, nitrogen: calculated profiles and experimental data, numerical simulation with additional convektion TEMPERATURE / °C —> r (Grau c; • lu1

TO 5 -- OJ I3 O s i I ! c i

s- 1 g TO I I

3 s TO Ê § Q.

POWEW k / R ai co o> heated withou loay an dt inside goale Th f .thes so e runs weregaio t ) n1 :knowledg e ACKNOWLEDGEMENTS o alse botto t higth oa d hm an temperature insulationp to e o lowet th e ) 2 n th ,ri s impurities cause adsorptioy db botn i h insulations tesheatino e t tth ) 3 , g devic higt ea h The authors wish to thank the Hochtemperatur-Kemkraftwerk GmbH, Hamm, for temperatures. leavin graphite gth e pebble experimentse th r sfo .

Afterwards the oven was loaded in the same way as described in 3.3, to repeat the REFERENCES primary tests with an improved instrumentation. These experiments are under way. III E. Teuchert, H.J. Rütten, K.A. Haas Tes5 3. t Run nean si r Future Numerical Representation of the HTR-Module-Reaktor Jul-2618, Research Centre Jülich (KFA) 199y 2Ma , After the present runs it is planned to install an insulation on the inner wall of the vessel and to load the vessel with pebbles. With this configuration it is possible to 111 J. Banaschek achieve higher temperatures near to those expected for the loss of coolant accident. Computation Method Analysid Dynamise an th n so c Behaviof ro Becaus e smalleth f eo r temperature gradients versus radiu e effec th sf naturao t l Power Plants with High Temperature Gascooled Reactor convection is expected to be a smaller one. In autumn this year three additional heating Jul-1841, Research Centre Jülich (KFA), April 1983 elements wil installee b l radiua n do s withi pebble nth e bed. Wit highee hth r power /3/ H. Gerwin and the distributed release of heat it will be possible to achieve the required high The Two-Dimensional Reactor Dynamic Program TINTE temperature simulato t d convesan e eth x profil f temperatureeo s ove radiuse rth . Jûl-2167, Research Centre Jülich (KFA), November 1987

. Conclusion4 s

To establish the principle of self-operating heat transport for the decay heat removal, a demonstrating experimen t preparationn i (SAN s i ) Au . First design studied an s numerical investigations have been carried out.

Ia smallen r facility (SAN ) AprimarI y test f criticao s l components have been performed successfully. Moreover this test facilit s suitablyi studo et y transpord an t transition phenomena at reactor components in the accidental range of temperature. Due to its symetrical geometry the set-up is especially suitable for the validation of thermo-hydraulic computer codes. LIS PARTICIPANTF TO S Hishida, M. Japan Atomic Energy Research Institute, 2-4 Shirakata-shirane, Aguilera, A. EDF - Centre de production nucléaire de St. Laurent-des-Eaux, Tokai-mura, Naka-gun, Ibaraki-ken 319-11 Japan Central , B.PeA , .42 F-41220 La Ferté-Saint-Cyr, France Hudina, M.B. Paul Scherrer Institute, LTH 21/422, Beine, B. Siempelkamp Giesserei GmbH & Co., CH-5232 Villigen, Switzerland Siempelkampstr. 45, G-4150 Krefeld, Germany Ide . A , Fuji Electri . LtdcCo . 12-1, Yurakucho 1-chome, Chiyoda-ku, Cleveland, J.C. Divisio f Nucleano r Power, Tokyo 100, Japan International Atomic Energy Agency, Wagramerstrasse 5, Ishihara . M , Japan Atomic Energy Research Institute, A-1400 Vienna, P.O100x Bo ., Austria 2-4 Shirakata-shirane, Tokai-mura, Naka-gun, Ibaraki-ken 319-11, Japan Billing, D. Bechtel Group, Incorporated, SO Beat Street, Kusavkov, N.G. OKBM, San Francisco, CA 94105-1895, United States of America 603603 Nizhni Novgorod 74, Burnakovsky Proezd, 15, Froehling . W , Institut Safetr efo y Researc Reactod han r Technology (ISR), Russian Federation Forschungszentrum Jülich GmbH, P.O. Box 1913, Kugeler, K. Institute for Safety Research and Reactor Technology (ISR), D-5170 Jülich, Germany Forschungszentrum Jülich GmbH, P.O1913x Bo . , Gao. Z. Institute of Nuclear Energy Technology, D-5170 Jülich, Germany Tsinghua University, P.O. Box 1021, 102201 Beijing, China Kohtz . N , Siemen, sAG Postfach 100100, Gottaut, H. Program Management for HTR Plans, D-5060 Bergisch Gladbach l, Germany Forschungszentrum Jülich GmbH, P.O. Box 1913, Kneer . A , Battelle Europe, D-5170 Jülich, Germany Am Römerhof 35, D-6000 Frankfurt am Main 90, Germany Haas, K. Institute for Safety Research and Reactor Technology (ISR), Forschungszentrum Jülich GmbH, Lange . M , Institut Safetr efo y Researc Reactod han r Technology (ISR), 1913P.Ox Bo . , Forschungszentrum Jülich GmbH, D-5170 Jülich, Germany P.O1913x Bo . , D-5170 Jülich, Germany Heck, A.V. Institut Safetr efo y Researc Reactod han r Technology (ISR), Forschungszentrum Jülich GmbH, Lohnert, G. Siemens AG, P.O. Box 1913, Postfach 100100, D-5170 Jülich, Germany D-5060 Bergisch Gladbac , Germanhl y

Hicken, E. Institute for Safety Research and Reactor Technology (ISR), McGurie Institut Safetr efo y Researc Reactod han r Technology (ISR), Forschungszentrum Jülich GmbH, Forschungszentrum Jülich GmbH, Postfach 1913, 1913P.Ox Bo . , O> D-5170 Jülich, Germany D-5170 Jülich, Germany . Y Sun, Institut Safetr efo y Researc Reactod han r Technology, o> Mejane . A , CEA - DRN/DMT/SEMT , 10 Centre d'études nucléaire Saclaye sd , Forschungszentrum Mich GmbH, F-91191 Gif-sur-Yvette, France P.O. Box 1913, D-5170 Mich, Germany Murray, J. UK Health & Safety Executive, Teuchert, E.H. Institute for Safety Research and Reactor Technology, Nuclear Safety Division Forschungszentrum Mich GmbH, St. Peters House, Baliol Road, P.O. Box 1913, Bootle L20 3LZ, United Kingdom D-5170 Mich, Germany

Mosevitskij, I.S. I.V. Kurchatov Institute of Atomic Energy, Wilson, J.T. AEA Technology, Ulits2 4 a Kurchatova, P.O 3402x .Bo , Harwell Laboratory, 123182 Moscow, Russian Federation Didcot, Oxfordshire OX11 ORA, United Kingdom Niessen, H.-F. Institute for Safety Research and Reactor Technology, Forschungszentrum Mich GmbH, 1913x Bo P.O,. Williams, P. NE-451, US Department of Energy, D-5170 Mich, Germany Washington 20585C D , , United State Americf so a

Obryk . E , Institut Nucleaf eo r Physics, Wolf, L. Institute for Safety Research and Reactor Technology (ISR), Forschungszentrum Mich GmbH, ul. Radzikovskrego 157, PL-31 342 Krakow, Poland P.O. Box 1913, D-5170 Mich, Germany Rehm, W. Institute for Safety Research and Reactor Technology, Forschungszentrum Mich GmbH, P.O. Box 1913, D-5170 Mich, Germany

Rettig, G. Siempelkamp Giesserei GmbH & Co., Siempelkampstr, 45 . D-4150 Krefeld, Germany

Siempelkamp Giesserei GmbH & Co. Siempelkamstr. 45, D-4150 Krefeld, Germany

Scherer, W. Institute for Safety Research and Reactor Technology, Forschungszentrum Mich GmbH, 1913x Bo P.O,. D-5170 Mich, Germany

Stolz, H.E. Rheinisch-Westfâllischer TÜV e.V., Steubenstrass, e53 D-4300 Essen l, Germany