e - Atomic Electricity

Nationa' Technical Infernsation ServiceU S Bepasment of Camrnerce5285 Part Royal Raaa, Springfield, Virginia 22151Price Printed Copy $3 00. Microfiche $0.95This report was prepared as an account of work sponsored by the UnitedStates Goverlment. Neither the United States nor the United States AromicEnergy Commrssion, nor any of their employees, nor any of their contractors,subcontractors, or their emp'eyees, makes any warranty. express or implied. orassumes any legal Irabdity or respansibrlity far the accuracy, completeness OFusefulness of any ivforrnatmn, apparatus. product OF process disclosed, orrepresents that its use would net infringe privately owned rights

I. ... i..

UC-80ORNL-4812-Reactor TechnologyMOLTEM-SALT REACTOR PROGUNTHE DEVELOPMENT STATUS OF MQLTEM-SALT BREEDER REACTORSXiI. W. Rosenthal, Program DirectorP. N. Haubenreich, Associate DirectorR. B. Briggs, Associate DirectorAugust: 1972OAK RIDGE KATIQNAL LABOMTOKYOak Ridge, Tennessee 37830operated byUNION CARBIDE CORPORATIONfor theE.S. ATOMIC mmeY COXMISSION

........... .=

..... .y.&....i........

CONTENTS....,.. .....&%* . .Chapter1. INTRODUCTION, SWRY, Am CONCLUSIONS . * . 0Backgroun$ ......................... 2Features of the Single-Pluid Breeder Reactor . . . 4. a0 D .Page* * hThe Processing Plant FOP the Single-Fluid Breeder a 6Nuclear and Economic Performance of the Breeder . . . e 6The Status of Development, the fijor Uncertainties,and the Alternatives . . - . . . . ae .. . a~eactor ~hysfcs ana F ~ cycles ~ I . . e a e eFuel and Cookant Chemistry . . . . a eGraphite .. -.eMaterials for Salt-Containing Vessels and Piping . ea.ee 8a a 88. 10e e * . . * . . * . e * . . 9Reactor Components and Systems . . . . eCells, Buildings, and Containment e e a a e a I e 14Instrumentation and Controls . . . . . e e e . I4Fuel Processing . . e 15Maintenance . . . . . . e e 17e .ee e 12.a.e e.e a ee e e ee 17. e a e e e 18. . - . e a a e 19e . a e 21. e a e e e e a e a a 22Design Studies and Capital Cost Estimates . . eEnvironmental Effects and Safety . a e eFuture Development Program . . e eThe Incentives for MSBR Development . . eFuel Utilization . . . e e aPowerCost e e . e e . e * a . e n 0 . e 0 0 . . e . 22Safety * 0 . a a e . . a * * e e e s a e m * e . * 0 * 25m e Likeiifaooa sf success . . . . e e a a 26Fuel Utilization . . . . . . e e a a e a 27Powerrcost * * . e e a . e 0 . s 0 0 . D 0 . . 0 D * 27Safety s . . 0 a . s 0 e . e 0 * . 0 . . e * e rn 28Overall Conclusions . . - e a e a . . e a a 38References for Chapter li . . . e . . . e e e . . . e e 312. EVOLUTION AND DEVELOPPENT OF NOLTEN-SALT REACTORS . . . e 33Origins . . a e . - C . O s . . s 0 . .Relation tu Other Fluid-Fuel Programs .r . . . * * . . * * 33..Early MSBR Concepts . . . .0 * . . . . . 0 34e e e .. aTheMSRE. * . . *..L O . . * . . . 3%o e . . . * e . s . . 35Purpose. * . . -. ea. e * . . aDescription e . . . a e .Development and Construction . eeOperation . e e 0 . . . . . . . .0 mResults...... -.Recent M~lten-Salt Reactor Concepts . aCurrent Programs . . . . - . . I e eUSAEC PloHten-Salt Reactor Programs .Industrial Studies . . .Foreign Programs . . e e s e * e ee. eReferences for Chapter 3 . . . . e..35...0 0 . .*.. . . . 0 0. . . e 0 0 3636. . a , . ... *. s o e . . . e o . 48. s . . . s * . . * 41m . . a . . . * . . 42e . . * . . . * * . 45* * . . . D * . . . 45* e . . . . a . . * 460 . . * * 0 * 1 e . 46m . * . e . . r a . 48

Chapter3. DESIGK CONCEPT OF THE SINGLE-FLUID MSBR 0omh, Reference Design * * 0 .m0 ..* 0Objectives D I e s I . 0 * e 0 e e 0Ore Csnservatisn * . . . 0 0 0 * .0 0PswerCost. a . e e e * D * e * eSafety and Environmental ConsiderationsTechnical Feasibility * D s a s D . e DGeneral e ~ ~ ~ sf ~ ethe p singie-nuia tweederFeatures of the Referenee Design e IAlternatives to the Reference Design . e . aEbasco Variations from QRXL Design e e eLow-Power-Density Core . e e e aReferences for Chapter 3 a e e e e a e e esPage........... .,.....v.-,.... ........:1-vi

-ChapterPage....).........w....'i.>.>.....*d:......,.:.. .,A+ ..2w.:

22%222222222226226

.....x.,....,&._....._a .%....*.:s..... .a,*...... >I....i._. Chapter8. (cont.)Pump Experience . . . . . e e e e e e a eShort-Shaft Bump . . . . . . . a e a e a .. . . . e e e. . . ee e a e .Scaling up Pump Capacity , . . . e e . . . e eDesi .........................Fabrication . . . . . . . .e a a ae a a I e aPotential Improvements . . . . a a eIndustrial Involvement . . . . . ..e I e a ee . eSalt Pump Development Requirements . . . .a e e a eEvaluation. . . . . . . . .. . O . . . . . O a . . O O . e . O a a a e e eDescriptisn ......................ae aStatus sf Flusroborate Coolant Technology . . .e e aUncertainties in Use of Fluoroborate a e e e a eFurther Development Work . . . . . . e e e a a e aCover Gas Addition to and Removal fromthePumpBowl.. *Corrosion Product Deposition e e . . . . e e .0 0 . m * 0 0 * * . D . . e . ee aMist Control . . . e e a a eon-Line Analysis . . . . . . . .CoolantLeah . 0 a . . . e s s D Da a e aEvaluation a . . . . . e e a a e . . . . .e ee e e e a . . . .e eRequirements and Criteria for the Brimary HeatExchangers and Steam Generator e e a eCurrent Concepts . . . . . e e . e . . . . . e e eMaterials Experience * . . . . .Status of Heat Exchanger Technology e . . . . . . .a 0 eaUncertainties and Their Sensitivity to. . . .a eFeedwater Temperature List e . . e . . . - .aFlow Instabilities in the Steam Generator . .e e aConfiguration . . . . . . . . e e e e e aPressure Fulse from Rupture of a Steam Tube . . . eaThermal Transients . . . . . . . . . . . . .e a eSteam Wastage and Leak Detection(Ste~i~ Generator) . . . . . . .aFission Product Deposition . . . e e e e e e eMaterials Compatibility . e a eFabrication Techniques . . . . .e e e e eLong-Shaft Pump . . . . eIndustrial Experience and Interest . . . . eStatus of Pump Technology . . = e .Long-Shaft Pump . . . . aEffects of Uncertainties . . . . aCoolant System e e .Experience with Coolant Salts . . . . . . aHeat Exchangers a e .Heat Transfer Experience . . e. . . -. . . e. . . . . . eNonachievement of Objectives . . . - . eaTube Sheet Protection . . e . . . aCorrelations and Physical Properties . . aix. * e e .. . . a. . a. * s .. . . ePage2292232 312 33.2 312322 322332332332332 342 342 35236,23623723324224324324324424424424524524524725Q2502512522522532532532532542 54254254254255

Chapter8. (cant.)25525525525625725725725926026826126226226226 32632642 642642652652652652662672682682682692692692702302312712732332732732 342 74276276282284286282.....c..

Chapter8. (cont.).....

Page331a 3112. rn1NTEP;rnCE a * P 0concept . s a 6 0 .Technological BackgroundEISE Preparations . 0 De e a* 0 0aExtent of MSW ExperienceConePusions from Pzsa " 0Reference Design MSBR - * 0Containment a 0 * * e eAfterheat . .0 e 0 aGraphite Replacement e aStatus, e a e e a = -Further Work . m 0 D 0 0. eEvaluation . = . .References for Chapter 1 2 033333543933934 334634 734 734935 035 l35 E35 235335435 635 635 335 835 83593.5935 936036 1362r,k.,

. :.&.....x.>>.....AS_..,.,...&.....x.9..... .!.:.:e....?*$....q*&.....;

.....,.bx

..... ,.:.:

..... u.:..2Anticipating that the SCAE statement will lead to a thorough reviewOf the mOlten-Sa%t BPeaCtOr cOnC@pfZ, We put t0gether this report toprovide up-to-date informatien for the review. The report brrbefly describesthe features of the molten-salt breeder reactor as we visualizeiwes our estimate of its performance potential, and attemptsto state the present status of the technology that is required for theco~-icept and the likcekih~od that any deve%op~~~ents needed can be accomplished.success fUuyAlthough in writing the report we have attempted to discuss allmatters that are of significance to the feasibility of the concept, wehave not by any means attempted to summarize all of the information thatexists on molten-salt reactors. However, the report does contain referencesto a number of pertinent topical and progress reports, includingmany of tire $5 topical reports that have been written ta record what waslearned from construction and operation of the Efolten-Salt ReactorExp~~i~ent. These and other reports and papers are also listed, alongwith abstracts, in reference 2, and the reader is referred to it foradditional ineraformtiono.... a,BackgroundAs noted in the authorization report, the origin of m~lten-saltreactors was in the Aircraft Nuclear Propulsion Program. In P94T9 someparticipants in that program ~sncluded that w~lten-fl~oride sdts had.useful attributes for the fuel of an aircraft propulsion reactor - nighuranium solubility, excellent chemical stability, and good physicalproperties - and work was started on a moltern-salt aircraft power plant.Easky in the aircraft development program came the recognition that themolten-salt technology offered additional advantages for civilian poweruse: avoidance of fuel element fabrication, rapid and inexpensive reprocessing,on-line refueling, good neutron econom)7 , and high temperatureQpe%atiQn at bow preSsU%e. bnS@qUentPy, in 1956 a y%OgPaElbegun at ORE% to hVes%igZite RlO%ten-§alt KeaCtGrs for Central stationgeneration of electricity. Three years later, enough progress had beenmade in defining the csncept that an &C task force on fluid-fuel reactorscould say that, while limited in breeding satis, the molten-saltapproach had the greatest chance of technical success 5f any fluid-fuelsystem [a].A result of this opinion was that in 1960 csnstructian of the Plslten-Salt Reactor &:periment was authorized. The 7.4 MWCt) MSlRE became criticalat Oak Ridge in P 65 and, after a very suceessfu% operating history, wasshut down in late 1969 SO that its budget could be used for developmentsaimed at molten-salt breeder reactorsThe MSRE experience was of major importance to the molten-salt c5ncept.Up until the PISRE begail t5 sperate well, in spite of the Task Forceeanclusion, few people besides those actively involved in the developmen%program considered molten-salt reactors to be really practical 0 Therirajor P€?ZiBObi Was that Ope%atiOll ZXld 'i@.aiTJt63X3blCe Of a systel3 COntaininga highby- radioac~ive fluid fuel that melted at over 800°F seemed extremelyb..

3...... .J .,+.;aKi,>difficult. In 1966, however, the F E E began to provide evidence to offsetthat view. When power operation began, the usual start-up pr~blemswere encountered; but sustained power operation provided a remarkabledemmstration of Operability. Starting in late 1966, an uninterruptedone-month run was made, then a three-rn~nth run, and finally a six-monthrun. Next, using a small fluoride volatility plant connected to thereactor, the OKiginalb partially enriched 235U fuel was removed from thesalt and was replaced by 233U that had been made in a production reactor.TII~ MSRE then operated a final year on the 23%, which made it tlie onlyreactor to ever have been operated on this fuel, and for a period plutoniumwas used as the makeup fuel. When shut down, the MSPEE had circulatedfuel salt at around 1280°F for a t~tal of 2-112 years.Busing the years in which the MSW was being built and brought intooperation, most of the development work on molten-salt reactors was insupport of it. As a result sf the MSE's success, however, the budgetwas increased to permit work aimed at molten-salt breeder reactors, andthe shutdown of the MSRE freed additional funds for this purpose. Themost significant product of this effort has been a new chemical processingmethod. Brought forth in 1968, this development permitted an importantchange in our concept of an MSBW. (Like many features of molten-saltreactors, this processing concept grew out of basic work on the ~henist~-gof fluoride salts that has been carried out ~ Q K a number of years as partof the AEC's Physical Research Program.)\hen molten-salt reactors were first considered for central-stationuse, it was not clear whether they wsbnld serve best as converters or asthermal breeders. A good converter could be obtained by putting uraniumand thorium in a single salt, but it appeared that they would have to bein separate salts to obtain a good breeder. This was because the chemicalprocessing methods then available were only suitable for separate uraniumand thorium salts, a fertile thorium blanket was required, and most of thefertile material wsuhd have to be kept out of the c~re to limit neutroncaptures in protactinium. The concHusion at that time was that eitherthe converter or the breeder could Bead to low-cost power, and the MSREended up having a single salt so that its engineering features resembleda converter, but the salt did not. contain thorium, which made it similarto the fuel salt of a two-fluid breeder.As emphasis in the USAEC reactor development program shifted moreand more to breeders, the design and development effort at OWL was concentratedincreasingly on the two-fluid system in spite of the greatertechnical difficulty of the reactor core. This difficulty arose chieflybecause graphite tubes were required to separate the two salts in thecore, and building a reliable graphite piping system that would withstandthe ridiation daraage of the high neutron flux appeared very difficult.%he processing advance of 1968 eliminated this problem. This advancewas the demonstration of the chemical feasibility of using Piquid bismuthto extract protactinium and rare earths from fuel sat that contains both~ K ~ ~ ~ Uand I Ithorium.IProtactinium, the intermediate in the breedingchain between thorium and 233U9 has a significant neutron capture crosssection and must be kept out Q € the core of a thermal breeder to obtaina good breeding ratis. The rare earths are important neutPsn psisonsand must also be removed rapidly for good breeding. The new process

4A conceptual design sf a 1860 "(e) KSB was prepared at ORNL andin l3Sb a report [4] was issued describing t concept Ebasco Services,Inca, and a number of industrial firms and utilities associated withthem in the privately-funded Molten-Salt roup, have reviewed the moltensalttechnology [SI and the OQX design [ 3 , and Ebasco and its industrialpagtnelPS have a%SQ $egla'Ll a SC?pZi%Eite COIICepsUEd design Study Qf MSBRqSmder an OLtT subcontract [SI 0 These efforts have produced useful suggestions,and some have been incorporated into our concept of an MSER.We visualize the basic concept as shown in Pig. 1.1. Tine core isformed from an array of bare graphite bars, SQ designed that they canbe replaced from above, and having open channels that provide for thepassage of salt. Die volume fraction left for salt is different indifferent regions sf the core, and in an annular space it has been made13igh enough thag: this volume is undermoderatea and XES like a blanketwhere most of the escaping neutrons are absorbeeh in the thorium.n e fuel salt is a mixture of lithium-7, beryllium, thorium, anduranium fEaaorides that me%ts at 93B"F, At reactor operit has a viscosity about like that of kerosene and a revapor pressure. As shorn Fa the fF re9 this salt flu2the $Ore W'rtet-e it iS heated %O l300"F, and i% then iS PUDIped thKoUgh aheat exchanger where the heat is transferred to a sdim fluoroborateintermediate coolant, The coolant transports the heat to a steam systemwhere supercritical steam at l088"F is generated, leading to an overalltherm1 efficiency 0% 44%. All of the salt-containing equipment is madeout ~asteiioy N, a nickel-base developed especiany for withmoPten flusride salts in the aircraft propulsion program.All of the salt-containing equipment is located in steel-lined concretecells that can be heated to raise the equipment above the meltingpoint of the salt. a drain tank located below the reactor has a naturalconvectioncooling system that is always in use and that serves as anextreme1~7 re~i~ieileat-a-ej ection system %issisn-proauct decay t ~ e ~ tFuel is drained into this tank during shutdowns, and the cells are designedso that in the event of a salt spill, it would also reach thistank and be cooled.....u.2i...-I'ss.. .

.....

e- .*~ubbles of helium are injected into a bypass stream Qf fuel saltand swept back. obit in a CyclQne Seg%aPaton^ to purge XelIon-)i3%, the majorneutron poisons and oqhet- noble gases out of the reastor. another andmuch smaller side stream of fuel salt is passed through the chemical processingplant ts remove the protaetini~rn and the salt-soluble fissionpPBdU6 ts eEnclosing the reactsr and the chemical plant cells is a conventionallin&-c~nc~ete containment building, which backs up the cells theto provide an additional barrier to the escape of ~adioa~ti~ity. mearrangement of the cells and the building layout provide access fromabove through removable shielding to all the radioactive parts sf theplant that might require maintenance.The Prscessiang Plant for the Single-Fluid Breedert eaters the processing plant at 0.9 gpm, a rate whichcontents of the reactor system through the plant everyten days. It goes first ts a fluori~~ati~ncolumn where the uranium isremoved 88 volatile UF6. Newt it flows to 8n eXt%a6%8% where it istacted with liquid bismuth containing some diasskwed Eitliiutn, and herein a reductive-extraction process, the lithium enters the fuel salt inexchange for protactinium which enters the bismuth e An additisnal steptransfers the protastinim into a separate salt where it is held untilit decays to uranium and is returned to the reactor.The fuel salt, now free of uranium and protactinium, goes to the1.'metal-transfer prsceSsF: where it is contacted with a captive vslmesf bismuth into which the rare earths and some other fission productspass. From this bismuth the fission products are in turn transferredinto lithium chloride. In the metal-transfer process9 the bismuth actssomewhat as a selective membrane which permits the passage of the fissionproducts between the fuel salt and the lithium chloride without the passageOf thQriUm.Finally, the UFE removed in the initial step is contacted with thepurified fuel salt and reduced bask to UF4 for return t~ the reactor.Waste prsduets in the plant are atcuaulated, given a final fbusrinatisntreatment to recover any uranium that might have passed that far, andput in storage in the reactor building where they are held as long asdesired before transfer to a central waste-disposal facility -Special ~ateltiabs will be required in the processing plamt. Thewall of the flu~rinator ill be protected f r ~ m sarrssicsn by a frozenlayer of salt, The transfer li_raes will p~obabl~gr be made out of molybdenumtubings and some of the large vessels may be built out ofb.k...-pww

..... .:

8.....Tne Status of Development, the Major Uncertainties,and the Alternativesjor objectives of this report are to assess the status ofMSBW technols tu identify any needed developments whose successfula ~ ~ ~ is uncertain, ~ ~ ~ and i in those ~ ~ cases, e to ~ see what t alternativesare available if the approach does not succeed. Q U ~csncPusions about these subjects, broken down in the same way theyare treated in the report, are as f~llows:.......-.z.,.....x.:.y,Fuel and Coulant C he~stryThere is no doubt about the choice of the fuel salt - theE~P-E~F~-%~F~-UFL+ system meets EBR requirements far better than anyother mixture. Its most serisras shsrtcsming is the low solubility ofoxide (30 ppm s'->, which will require that the ingress 0% air or m0i.sturebe carefully ~ ~nts~lled. Mere the Xds experience provides confidence,since its oxide content changed little during plant life.Fissi~n-p~~d~ct shem5stry in fluoride salts is well understood,with substantial input from the :+SE experience me physical behaviorof the noble-metal fission products, which exist in elemental form,....

3cannot be predicted with as much confidence as we would like, however.The reactor designer, therefore, must use conservative assumptions asto their deposition on metal and graphite and transfer into the offgasuntil data from operation of another reactor have reduced these uncertaintieseA fluoroborate mixture (the 92-8 m d e % MaBF4-MaP @utectic) appearsto be the best choice for the loop that is needed between the fuel andsteam systems. Effects of steam inleakage on corrosion and consequencesof mixing flusroborate and fuel nust be explored further. Perhaps themost important need is a better understanding of the behavior of hydroxideion and of mechanisms by which tritium, diffusing from the fuel system,can be trapped in the Slu~~obo~ate.Fuel and coolant salts in the HSRE were analysed by removing samplesfrom the reactor and taking them to an analytical laboratory, However,considerable progress has been made in the last few years towards developingon-line methods of analysis for salt. Most involve electroanalyticaltechniques but visible-light and infra-red spe~troscopy also of ferpromise. One corrosion loop is now operating with a controlled voltammetryinstrument that reports the U3+/U4+ ratio in the salt, which isextremely important in reactor operation, Methods suitable for hydrogen,chromium and other corrosion products, salt impurities, and certain fissionproducts are beginning to emerge from the development effort and,wish continued progress, may become usable on a reactor.GraphiteGraphite in molten-salt breeder reactors must meet three particularrequirements: %t must stand up to neutron irradiation; it must have poressmall enough that capillary forces exclude fuel salt, which does not wetgraphite; and it must have a %ow enough permeability to gases to keepdown the absorption of xenon. The graphite manufactured for the MSE hadto exclude salt, and a special small-pore material was developed by themanufacturers, but the total radiasion dose was too ISW to make radiationdamage a problem, and exclusion of xenon was not a specification. Thus,although a graphite stringer removed from the MSZ showed no effect oftwo and one-half years in contact with fuel sdt, it would not have metthe radiation damage anel gas permeability requirements of an EgSBR.Radiation damage in graphite is caused by high energy neutrons andin most graphites results in shrinkage followed by expansion. Thesechanges in conventional graphites result in the volume starting to increaserapidly at neutron fluences that are too low to be of interestfor MSBR'ss. However, in the last several years9 special grades ofgraphite that appear to be made by an uncalcfned-coke process showlittle contraction and a longer period before rapid expansion begins eOne that is said by the manufacturer to be commercially available hasbeen tested in the HFIR at MSBB temperatures and bund to be .able tomeet the 4-yr life assumption of: the reference design. Consequently,a material that has adequate radiation resistance seems to be available,but longer graphite life is desirable, and there is hope that our grawingunderstanding of radiation behavior will lead to longer irradiationlife 0

Progress with seaZing graphite to exclude xenon has not gone as far.Two techniques that invslve use of pyrolytic carbon - one that depositsit in the surface pores and the Other that puts On a Ohin CQatiIlg - Canseal the material adequately, but the permeability 0% most ~f the s m d lsm1ples tested has increased excessively under neutron irradiation,Some understanding of why the permeability of the coated samples increasesha been gained recently from renarkably sharp photographs obtained usinga neb7 technique With the SCanniRg @leC%rOn IIIicrOscOpe. me failUPes aFeI ~ W thought to result from defects seen in the unirradiated materiaE, anda new procedure for depositing the coating has pr~duced fl~~q-freethat are now being irradiated. However, the sealing method has not yetbeen proven to Wabk, and scale-~p of the process to \&ere it be usedfor large pieces is still in front of us even if the method turns out tsbe a success.Sealing graphit@. by iIIIp~egIIatiEIg the SuZ'faCe pCJres With fUs%--freesalt is a. possibility if pyrocarbans cannot be used, but if no methodwill work, the breeding ratio of the MSBW will decrease ssmewhat becausesf increased neutron capture in '35Xee %e additional loss in breedingratio will depend on the rpate of stripping by the noble-gas spargingsystem but wBlZ prob By lie between 0.005 and 0.01.S X I I ~ ~ ~ S.=.....,.... _....... ..&..Bastellsy N was developed for use with molten salts at the hightemperatures ne ed in aircraft power plants, and since it has goodstrength ad good eo a%ibi%f-ty W i t h fLbkQ?Zide SdIsS, it Was Used forthe cQnstruetisn of e KSa. \dhFle the MSE being built, ewerimencsrevealed that the creep ductility of Mastellay N is reduced byneutron irradiation. This embrittlewent is caused by heliiam producedby therwnal-w@utron captures in the alloy, in contrast with the embrittle-Icent due Void foKEEition bq" fast neu%rsl?S that ha% been Of COnCePbl.for fast reactors. Analyses showed that stresses in the %fSlow enough for the reactor ts be operated safely in spite sf decreaseddu~ti~ity2 but this would not he true of future reactors I ana 8 developmentprogram was begun to find a cure for the problem.The approach followed was that of adding carhide-fsrnahg elementsWhich have been US@d t0 alIl62liarate the efa$rittleKEnt Sf Stainless Stgelby fast neutrons, a d 0.5% titanium was found to sustain the ductilityof Masteaiaoy M at the ssm temperature of l200"F. However, at k300"F,the outlet temperature in the 6KiJL-design NSBW, changes in the structure0% the carbides in the alloy caused the remedy to be lost., but this wasove.pcBme by raising the titanium content to &out 2%. some further gainwas made by ad ng niobium with titanium, and hafnium in csnjunctianwith niobium w found to be very effective, but problem with weldabilityand cos have caused us to limit our effort on hafnium-containingalloys Transmissisn electrons micrographs that disclose the carbidestructure have been of great value in revealing the factors inand per&% the properties of a s le to be judged rather re1ibefore it. is irradiated,k.L

11....,:w:

12Based on these observations, there appear to be alloys that are notaffected by tellkarium, and ELIXXIg thew are €Klodi%iCationS Qf Hastelloy%f an increase in the chrsniuw content of HastelEoy P; is required, or anIneonel or a stainless steel must be used, the ~ O~KOS~CXI rate will behigher than that with standard Hastell~y E9 brat the increase will probablybe tolerable, Corrosion experiments will be needed to find out. Exceptfor the stainless steels, where the work has already been done, a changeto one of these materials is likely to mean that modifications to conferradiation r'2SiStanCe W i l l have LO be fOUnd. If thboUgh gQod fQZtlX?@ XIaddition of niobium to Hastellsy K shodel suffice, acceptable corrosionbehavior is fairly well assured, tind the effect a% niobium on irradiationresistance has already been investigated to some extent. In any ease,in-pile capsule tests will be needed to show that the same effects areseen in-pile as with telBurium addit ions out-of-pile, and preparationsfor them are underway. A clear demonstration that a satisfactsry materialhas been found will be necessary before another malten-salt reactor can bebuilt D..... . .~;Although aany of the components and systems an PlSBR power plantare similar to those needed for solid-fuel reactors, the design requirementson others are different, and a number are unique to the ms%ten-sa2t~yisten. N?I~~Y 06 the different 0% unique aspests were investigated 111 thedevelopment programs far the aircraft reactor and the blSRE, but not allhave been used or teated, and increases in size or performance are requiredin most cases.Starting first with pumps, vertical-shaft centrifugal pumps withoverhung impellers were developed for mo%tea~-sdt service and used satisfactorilyon the Aircraft Reactor Experiment and the PfS 9 as well 85 usedand tested in a Elumber of salt loops. (A small oil-leak from the MSREp5.mary pump caused problems with the off-gas system, but the pump itselfwas used witho~t trotable for %Re reactor life, and the leak was easilycoa~rectecl in a spare penEp.) ~ t h s ~ gsteps h tim aad up to a 10- to i5-maincrease in capacity will be needed in a progression from 'die MSaE to faallsizeXSBPS, the same b ~ i design c as that on the MSE is specifiedth@ refer@nCE? MSBR design, and the Scale-up Shod$ be ?ZehtiV&ly§tKEiight-forward. We believe, consequently, that although several years will berequired to develop and test larger pumps, the problems are well understoodand satisfactory pumps can be obtained. Byron-Jackson, an associate sfEbasca in the Molten-Salt Group3, has expressed similar confidence,The MSRE intermediate heat exchanger and air-cooled radiator operatedwithout difficulty I and analyses showed decrease ist pert-ormnce th~~ughoutthe plant life. Heat-transfer experiments, as well as the operatisn ofthe E E Z units, indicate that salts act as ordinary fluids and their heattransferbehavior can be predicted reliably as lung as accurate physicaldata a%e 8VaiPable. The ZkspeCtS Sf the MSBR that differ fE"QIhlaside frona size, have to $0 with the need for high performanceQfL the MSBR to limit the fuel-salt inventoryp and the requirement thateither failed tubes can be located and plugged in place, or that the tube

13bundle or entire unit can be replaced. Both of these create design problems.To obtain eoqactness, either smaller-than-usual tubes or tubesdeformed to enhance heat transfer have been shm in our heat-exchangerconcepts, and use of either will require a testing program. Some increasein fuel-salt inventory will result if the compactness shown inour concept is not achieved, but since only 17% of the fuel salt is inthe heat exchangers, a moderate increase in their volume will have alimited overall effect. Providing for heat-exchanger repair is a partof the overall MSR maintenance problem, but new techniques for pluggingtubes being developed %or other uses should be helpful.There were no steam generators on the ARE and MSE, and as far aswe ~ Q W , there has been no experience with generation of steam with highmeltingsalts. The major problem is that in conventional steam cyclesthe feed-water enters the steam generator at a temperature below themelting point of the MSBR co~lant sdt. As a result, unless othermeasures are taken, some salt would freeze on the tubes. Allowing alayer sf salt to form might be acceptable, but to get around the questionin our reference concept, we altered the stearn cycle to increase thetemperature of the steam. entering the steam generator. A supercriticalsteam cycle was adopted but modified to mix some exit steam with feedwaterto raise its temperature to cl~se to the salt melting temperature. Thepenalty is some additional equipment and a small loss in efficiency, butthe net effect does not appear to be very great. Other ways of overcomingthe salt freezing problem also appear feasible, such as the reentrant tubeapproach that appears in some sodium-heated stearn-g~~~~ator concepts.%e Poster Wheeler Company is now exploring molten-salt steam generatorconcepts under an OREu'6, contract, and they will consider our concept andothers before recommending a design. Whether Hastelloy 1'9 has adequatecorrosisn resistance for use in a high-temperature stearn system is beinginvestigated at present.The sodium fluoride-sodium fluoroborate coolant proposed for theMSBR melts at 725°F- This gives it a 125°F melting point advantage averthe LiP-BeP2 used in the MSRE and it is much cheaper, but until a fewyears ago we had .had no experience at all with fluoroborates. Duringthe past several years, howeverT, an issthermd MSRE-scale loop has beenoperated with fluorobsrate, as we11 as two small forced-convection loopswith heaters and coolers, and a number of natural-circulation loops. Afairly extensive chemistry and analytical chemistry program has also beencarried out. The major difficulty with fluorobsrate is that it has agreater tendency to pick up moisture than the other salts we have used,which makes it more corrosive, but the corrosion rate with clean salt ismodest. The BF3 vapor pressure over the salt requires special provisionsin the cover-gas system, but these have been worked out satisfactorilyin the loops that have been operated.The likelihood of steam generator leaks that introduce moisture intothe coolant will require that a cleanup systembe provided. One sf theways to prevent tritium getting into the steam may be to trap it in thecoolant and extract it from there, and the processing system to accomplishthis probably serve both purposes.

14The noble gases are imsoluble in fuel salt, ad, csnsequent%y, the6 is sion-proauet pa is stain in an rnBR can be greatly reduced by spargingon from the saLt. This was demnstrated to be very effective inE, where over 88% of the 135xe was removed. A more effective andbetter controlled system is proposed for the XSBR, however, that involvesinjesting helium into a bypass stream Of salt and removing it and thenoble gases with a centrifugd. separator Experiments using water haveprovided designs for the equipnent and indications of the performnce tobe expected, but testing with salt is needed and is planned for an tamscaleloop now being built.The containment philosophy and the containment building desim forthe MSBR differ little from those for solid-fuel reactors, although the1% require more extensive filtration and cleanup provisions. The prosalto use the reactor and coolant sells as ovens in which to heat theaE% systems, h eve%, iS not on%y UKliqUe to the MSBR but also iS differentfrom the FE where components and pipes were enclosed in insulationand individually heated. The uncertainties that exfst mainly have to dowith she best way of insulating the cell and how the equipment will besupported and restrained to resist an earthquake. We foresee no limitingprsb%ems here, bus if some arose, an alternative would be to return tothe concept used successfully in the FIsa.The method of providing access for maintenance by removing sections0% shielding froa, the tog of the cell. was used without difficulty on theI%=, and the major diffePenCf2 OR the fu%l-SCa%e MSBR is the larger Sizeof conponents to which access will be required.Instrumentation and ControlsMSBR's have somi features that are favorable with regard to controlsf the reactor and a few features that add diffi~lties. chief among thelatter are the high freezing temperatures of the salts, which require thatspecial provisions be mde in the control and grotestisn systems to avoidfreezing during transient conditions and part-load operation. In addition,rates of temperature change probably have to be c~~ltrolled during loadchanges to prevent excessive transient stresses in the system. Wt1iPecontrulb methods have not been worked suc in detail, several alternativeschemes appear ts be possible, and at least one shodd be satisfactory.The small amourt of excess reactivity required in an MSBR and thedynanic characteristics of the reactor greatly simplify the reactivitycontrol requirements * Maintaining a long-term reactivity balance willbe difficult on an MSBR because of the csntinmous fuel processing, andnew techniques my have to be developed.Significant experience with instrumentation systems has been obtained,with high te erature facilities of various kinds, and in particular withthe MSRE where the reliability of the thermocouples was particularlyw.

satisfactory. The pressure and flow rate of the fuel salt were nbtmeasured directly in the PISRE, however, and having the entire ELSE381~eactor cell heated adds the complication that all of the instrumentsin the cell must be able to operate at high temperature. Hence somenew instruments a d measurement techniques will be needed for the MSBR.Fuel Proces si*g.....=,A........i', ,.&Achievement of a significant breeding gain in a thermal spectrumreactor is dependent on rapid removal of the fission products, and inthe case of the single-fluid &fSBR, is also dependent on separation ofthe protactinium from the fuel salt. The chemical. steps involved inthe processes presently proposed for accomplishing these separationshave been thoroughly investigated and appear to be well establishedHowever, engineering development and the demonstration of satisfactorycontainer materials have not progressed nearly so far.Fluorination to recover u~aiiiurn from radioactive fuel has beenused several times, most recently in the fuel rep~oeessing at the >ERE.However the scheme proposed for the MSBR involves continuous fluorination,which has only been demonstrated in sm11 equipment and which requiresbetter corro~ion protection for the fluorination vessel than hasbeen achieved before Although many aspects of contin~o~s fluorinatorshave been investigated, including the formathan of frozen salt layersthat are expected to protect the fluorination vessel, a significantdevelopment effort on the fluorinator lies before us. some of thenecessary experiments are underway and are progressing satisfactorily.Direct reduction of UFg for return to the reactor by its absorptioninto fuel salt has been demonstrated in small laboratory experiments andthe behavior found to be satisfactory e However, no engineering experimentshave yet been operated, although they are planned for the near future.One favorable result of the recodaination experiments was the discoverythat gold is not attacked by the process fluids, and gold plating mayprovide adequate corrosion protection for the recombinerReductive extraction sf realistic concentrations of uranium and zirconium,whose behavior is similar to that of protaetiniu~n, has been demonstratedwith fuel sdt and molten bismuth in a packed-column contactor.Flooding velocities and mass-transfer rates were measured and found tobe as predicted from data on mercury-water and aqueous-organic system.A demonstration of reductive extraction using representative concentrationsof protactinium is needed, and 38 to 50 gram sf 231Pa is beingobtained to make that possible.All of the steps in the metal transfer process have been ~~IIIQI-Istratedin a small single-stage integrated experiment, and preliminaryoperation of a larger, but still single-stage, experiment has begun.A three-stage experiment that will be 5 to 10% of MSBR scale is beingdesigned.Losses of fissile material from the processing plant must be keptlow. Although the process fluids circulate repeatedly through the plant,actual %~sses can occur only in the wastes. These are therefore collectedtogether, held to await protactinium decay9 and then batch fluorinated to7-r......:.=.....cz,&

‘P~GQV~P any traces of uranium before discard. Experience with fluorinationshows that the uranim curltent can be reduced to a very low levelby this technique,Fuel salt and lithium chloride are compatible with some commonconstruction materials but nick~cel. dissolves in bismuth , and the solubilityof iron is great enough for mass transfer to occur rapidly ina system having significant temperature differences. Consequentlymaterids such as molybdenum, graphite, and tantalum will be requiredfor the processing plant. Fabrication of molybdenun has always appearedvery difficult, but we have developed a variety of forming and joiningtechniques during the past ma yearsp and a fairly complicated processingfacility is presently being built completely from molybdenum.Graphite shodd be less expensive to me tha molybdenum, and the P~OCessingvessel for the three-stage metal transfer experiment will. bebuilt out of graphite. However, we presently have insufficient dataon the compatibility of graphite with bismuth csntaining large concentrationsof lithium or trace quantities of other materials. A smallI_-;f’natural-circulation loop has been built out of a tantalum alloy and*e -----‘ operated with bismuth; the corrosion rates of tantalum and of the alloyseem acceptab%y small, but there Ss some inconsistent evidence of embrittlementof the alloy,Carry-over of significant quantities cf bismuth t~ he reactorwhere it could attack Hastel%oy N nust be avoided. One Hastellsy Nnatural-circulation loop containing fuel salt has been run with anopen capsule of me%ten bismuth in contact with the salt. No effectof the bismuth has been seen, but more needs to be learned about howtolerant the reactor would be of small quantities of bismuth in thesalt. Little is yet known about the tendency of bismuth to be entrainedin salt, and its sol~bility in salt has not been measuredaccurately although bassic them~dynami~ considerations indicate thatthe so%ubKLi%y must be very low. Our approach to preventing bismuthcarq7-over is to attempt to develop salt-bismuth contactors with stirredirnterfkzces in which the bismuth and salt are not dispersed, and a preliminarydemonstration of this technique has been made in a metal transferprocess experiment. Even if such techniques are used, however, carefulanalysis for bismuth, plus a final de~~tup step, such as passing the saltthroplgh a bed Of nickel KCOQL, WiLl likely be Wed.Failure to develop sya tern for rapidly removing protactinium andthe rare earths would prevent the attainment of a significant breedinggain in a single-fluid PISSR. Fortunately alternative approaches appearto exist for mst, but net all, parts of the process. Graphite may beusable for the plant material if molybdenum will not serve, and tantalumand its alloys represent mother possible alternative for parts of theplant If some parts of the presently proposed fluorination-reductiveextraction process cannot 3, made to wurk, an oxide precipitation processthat has been investigated in a Limited way appears to offer an attractivealternative; pret.actinium can be seEectively precipitated as Pa205 bytreating fuel salt with a mixture of steam and hydrogen fluoride, anduranium can subsequently be remved by a similar process. Indeed, furtherwork my show that the oxide process has advantages over theflus rinat ion- reduc ti ve ext ract i on me thodL.,r,w.,w.2

17.....:

looking at some features of the reactor in reater depth than has beendone by OR&&. The mjsr aspects of the des gn that they have identifiedas requiring f ~rther study have to do with the transient thernaE stressesin the primary system f ~ll~ing rapid changes in reactor operating csnditions,ad the methods sf supporting the reactor components and providingKes$raht to PeSPSt Sh&illg by earthqUak€?s . E ~ ~ S CfaV0rS Q rep%acinggraphite an element at a time rather than replacing the entire core asa unit, as we had proposed.In the studies we have made to size components and evaluate dternatives9 only simplified e%astic-stress analyses have been made Beforeonents can be built for reactor use, however, additionalmechanical pr i e ~ measurements will have to be made (the extent dependingan w terid is used), ad extensive stress analyses willhave to be perfarmed. Design rules analysis methods atid stress limitsthat reflect the time-dependenee of materid properties ana stmcturaibehavior will have to be used because the strength sf likely materialsat reactor te eratuxes wiei be limited by creep effects. ~ ~ ~ aretilsas i g nts C Q V ~ these ~ requirements are currently being developed in the LWBWprogrm and wFll be available for use in MSBB design.Our capital cost conparison between a fully developed XSBW and apresent-day PI43 indicates that the costs are rougtsPy the same. Althoughthe ascuracy of an estimte such as this is not dependable, mainly becauseof the limited depth of the design, we believe that the tecZbnPqueQf using actual PWR cost breakdowns for the basic cast data helps tolinit the uncertainty. Vnen its design study is completed, Ebaiseo willmake a cost estimate for the %BR that shodd provide additional insightbecause Of their eXperit3ICe 85 aPCkiteCt-engheer QlI SnSny %ight-Waterreactorpower plants 0Environmental Effects and SafetyThe major uncertainty with regard to environmental effects is howto $ea% with tritiun. Tritium is a special problem because of its highrate sf production in the fuel salt and because it readily diffusesthrough metals at XSBR temperatures. '%he distributisn of t~itium inthe X3.E was determined and found to agree reasonably well with ananalytical model developed for predicting its behavior e When thismodel is used to estimate 'now tritium wodd behave in the YfBR withno special measures taken to block its passage, an excessive ammt(790 ~ilstay) is feud to reach the steam system. Several modificationsand operation offer ways for drastically reducing escape bythis route. The objective of limiting tritium release to within presentAEC guidelines for light-water-cooked reactors appears attainable, butthe best measures are yet. to be chissew a d demonstrated.The situation with regard to kinetics and nuelear safety is uniquebecause of the circulating fuel. The kinetic behavior af molten-saltreactors is ~elP understood, however, and predictable by methods provedin the Msm.problem, as demonstrated by operation of theeffective delayed ne~t~o9tThe smlP delayed neutron fract n causes no difficdtRE on a 3 3 ~ fuel with anfraction of only 0.8619. Thus there is ampleu,s.....

....,.Z&,;!:e.........,w;a....&S3s. ..:.y, u..... ,wbasis for being confident that damaging nuclear excursions are highlyimprobable. Of the potential sources of reactivity inc~ea~es, the onethat will require the most study is the hideout of fissile material-Conditions that codd lead to such hideout are known (oxide precipitation),but it appears that these conditions can be safely avoided.The afterheat situation is also unique. The major source is mchless intense than in solfd-fuel cores because in an MSRR the bulk O€ thefission products are incorporated in a large mass of fuel salt. Furthermre,this heat source can be gotten into a rePi ly cooled situation(the drain tank) under any accident condition. A ssmewhat separate problemis the smaller radionuclide heat sources in the processing plant, inthe reactor off-gas system, and deposited om surfaces in the fuel system,which will also require cooling. The MSE provided usefd information onfission product behavior but uncertainties in nob le-metal behavior dictateconservatism in design for cooling the fuel loop and off-gas system.On the whole, however, afterheat promises to be less of a problem in GdSBR'sthan in other reactors. In particular, the dilute heat source makes the"China syndrome" less of a concern.The design-basis accident in an M§BR is a rupture in the fuel systemthat quickly spills the entire fuel imventory, and the containment of theradioactivity in this event is the chief safety consideration €or an MSBR.The containment must be tight, but the behavior of the spilled salt and ~its fission products is predictable and there appears to be no need for "innovative development of containment technology to take care of thFs .event.It appears from basic considerations that site requirements for anMSBR plant should eventuzlly be no different from those for other reactorsof like power and its safety provisions should be no more expensive.Because of the unusual nature of an MSBR, however, it will be necessaryto begin with fundmental principles and develop criteria appropriate tothis kind sf reactor, then to perform a safety analysis conparable indepth to those for reactors now going into operation.Future Development Program-.....%.A, . .......,.::Ea.....\....L,. As in any reactor development program, achievement of economic.mo%ten-sdt breeder reactors will require that the basic technologybe well established in research and development programs and be demonstratedand expanded by the construction and operation of several increasinglylarger reactors and their integral processing plants. Thetechnology program is in progress now, and we favor the construction ofa 150- to 208-MWCt) Molten-Salt Breeder Experiment (MSBE) as the nextreactor in the sequence to an MSBR. The MSBE would have the powerdensity and dl the features and system of a full-scale breeder ~eactor.Other steps are possible, and one favo~ed by the Molten-Salt Breeder WeactorAssociates, an organization consisting of Black 6% Veatch ConsultingEngineers and a group of utili-ties 181, is the construction of a largerbut lower performance converter reactor that wodcl evolve into a breeder.We believe, however, that the more direct route of the breeder experimentis preferable

In the technology program several advances must be made before wecan be confident that the next reactor can be built and operated successfully.%le most important problem to which this applies is the surfacecracking sf Hastelloy KO Sone other developments, such as the testingof 68DIe Of the CQElPOXI€ZI%s QT the batter Stages Of the pl33Cesshg p%antdevelopment, could actually be completed while a reactor is being deedand built. The mjor developments that we believe should bepursued during the next several years are the following:A modified Haste%koy Ns or an alternative material that is imune toby teiiarrium, must be selected ana its eompatibiiity wit11 fuelsalt demonstrated with out-0%-pile forced convection loops and in-pilecapsule experiments; means for giving it equate resistance to radiagemust be faun if needed, and comerci proauction Of thehave to be demonstrated; the mechanical opesties dataneeded for eode qualification must be acquired if they do not alreadyexist 0i”. -2.-.-. .. 3.4.5.of intercepting and isolating tritium to prevent its passageinto the stea~zs system mst be demanstrated at realistic ~oslditi~zns anda large enough scale to show thkt it is feasible for a reactor.The various steps in the processing system must first be demonstratedin separate experiments; these steps must then be combined in an integrateddemonstration of the complete process, including the materialsconstruction; ana ~ n a ~ after ~ y , the KSBE plant is con~~pt~aliydeSipC?d, a WCk-up COb%tailling COIRpORentS that as Close as possiblein design to thase which will be used in the actual process must bebuilt and its OperatiOn and minteellance procedures demons trated *The various components and systems to go on the reactor must be deveSspedand dem0asCrated bander conditions and at sizes that a%How confidentextrapolation to the XSBE itself. These include the xenon strippingsys~ern for the fuel salt, ~ff-gas and cleanup systems for thecoolant salt (facilities in which these could be done are alreadyunder construction) tests of steam generator modules and startupsystems, and tests of prototypes of pumps that would actually go inthe reactor. The construction sf an engineering mock-up sf the majorcomponents and systeans of the reacts% wouBd be desirable, but whetheror not that is done would depend on how far the development programhad proceeded 2n testing various cowonents and system individuallyGraphite elements that are suitable for the MSBE shsubd be purchasedin sizes and quantities that assure that a co pro due t ion capsexist, and t,he radiation behaviorky-produced mteriai shsula be confsealing graphite to exclude xenon shodd continue to be explored.

21Other research and development will be required in a number of areas,but those listed are the major and most costly undertakings. They representa desirable program for advancing and testing molten-salt breedertechnologj7 in the absence of a cornitmeat to build a reactor, and mostbecome necessary if a ~eactor is to be built.%he Incentives for WBR Development'phe foregoing discussion indicates that csnsiderab le progress hasbeen made towards the development of molten-salt breeder reactors butalso reveals that a substantial development program will be requiredbefore commercial MSBh become an actuality. We turn now to the questionof whether there are incentives for pursuing such a program.A full statement of the rationale for the development sf moltensaltbreeder reactors codd logically make three points. First, breedersare needed. Second, in addition to the LME3RR, whish has already beenestablished as a national goal, one or more different concepts that arealso capkble of meeting breeding needs should be pursued at appropriatelevels of effort. Finally, the MSBR has a unique and significant roleas an alternate breeder.Tke argument for the first point has been adequately made in anumber of places - mast notably in the 1962 report to the President [la]-and does not weed to be repeated. The second point involves considerationsthat go beyond the seope of this rep~rt and thus the arguments to supportit will. also be omitted, although we are convinced of its validity. Weare thus brought directly to the third point and must consider how wellthe MSBR meets the requirements for an alternate breeder.For the development of any breeder to be worth pursuing, the system,including the reactors and the associated fuel industry, must potentiallybe able to meet three basic requirements -.... ,a......iii. ..m.........,.:

222s discussed in the following paragraphs, MSBR's appear potentiallycapable of satisfying all three criteria.Fuel Utilization. - The r pid removal of fission products and protactiniumfrom a molten-sale breeder reactor, coupled with the absence Sfstrongly ~~U~KQI-I absorbing materials in the core structure, makes it feaaiblet~ achieve a significant breeding gain in a thermal reactor, Whencombined with its low specific inventory, this results in good utilizationof uranium ore resources. Tiiis point is illustrated by Fig. 1.2,~hich shows that 11either the breeding gain nor the doubling time ita themselvesare adequate measures sf the ability of a breeder reactur to limitthe amount of uranium ore that must be wined to fuel a growing nuclearpower economy. The fissile inventory is also important, and it is their~ Q W specific inventory that makes it possible for molten-salt breederPeactors to serve a8 well as fast breeders in limiting the resource requirementso Tu demonst~ate f~rthe~importance of this point thepeak UPaniUPn Ore Kequirefnent§ Obtailaed. from CUPTTeS like those Show72 inFig. 1.2 have been cross-plotted as a function of the specific inventoryand doubling time in Fig. 1.3.Cansequenthy, our conclusion is that molten-salt breeder reactors,ill Spite Sf their Small breeding gELiEk, Can Serve W a l l &S fast lreaCtQPSin esnserviwg uranium. In addition, molten-salt converter reactors fedwith plutanium can have high conversion ratios and very favorable fuelcyclecosts. Thus, even if m~lten-~a%t breeders are not S U ~ ~ ~ S S 8% ~ U %not needed, the molten-salt teehnology can be used in a ~onverter thatserves as a co~~paniola to a fast breeder to provide low power COS% and abalanced fuel ECO~OIII~ [9, p. 6-521~hrotng~=~~t this section tnave discussed the conservation of uraniumore but have n ~ mentioned t thorim, even though the MSBR is a thorimcyclereactor. The reason is that it is fissile fuel that is in shortsupply ami not fertile material. 111e W.S. reserves of bot11 2 3 8 ~ andtikssium are adequate to supply the need for fertile materials for taun-$reds sf years, and the cost of power is relatively insensitive to theircost. As already noted, doubling the cost sf thorium from $5/_lb to $lO/lbwould add ondy OoQ5 rnihI/hh to the fuel-cycIe cost if no changes wereto ecenonnize on t~mrium use.Power Cast. - Avoidance of fuel fabrication, rapid removal ~f fissionproduct p~isons and pratactfnium, and a low fissile inventory resultin low fuel-cycle costs for 'a.ISBWPs in spite of inclusion sf a substantialcapital cost for the on-site processing plant, Capital casts for thereactor are less certain than fuel-cycle csosts, brat a detailed comparisonindicates that when fully developed, the construction casts of MSBR'ssh~uld be about the s e as tbmse 06 light-water reactors. The higha. efficiency of the MSBR, the ]bow primary-system pressure, andthe large temperature differences avaibable for heat transfer are thekey factors w~xich hoia NSBB capita^ costs dot%m9 wllereas remote maintenancerequirements on parts of the plant entail added c~sts.*... -

23CUMULATIVE URANIUM REQUIREMENTSOWNL- BWG 72- (01 3AW282+x&MSBR 21 f.5 1.06.... w...., ,~ .MSBW 90 8.9 1.07FBR 5 3.0 f.63NOTE : (A) ONLY LWRs BUILTTHROUGH ENTIRE PERIOD..... .m.... 2,s1970 i x KIQO ~ ~ moa 2010 2020 m oYERWFig. 1.2. Effect of breeders on the ore requirements of the U.S.nuclear power CXORO~~ if breeders are introduced in the mid-1980's andonly breeders are built after 1998.x.&$........ .-. c....

24OWRlL- BW6 92- 5897a59 2 3 4 5 6 7SPECIFIC INVENTORY (kg/MW (elw.Fig. 1.3. The effect of the fuel utilization characteristics ofbreeders on the cumulative uranium ore requirenents for a growing U.S.nuclear power economy,

I....25...... >......:=.>Some penalty must be paid for the cost of replacing graphite, but ifthis must be done as frequently as every four yearsis our estimate is thatthe cost will fall in the range 0% 0.1 to 0.2 mill/kwh. The need to performremote maintenance on parts of the plant may add additional downtimerequirements but this should be offset by the high availability resdtingfrom on-stream refueling, which obviates the need for annual refuelingshutdowns 6In sum, we believe that power costs of MSBRPs should be competitivewith those of light-water reactors, and the attainment of low ~ Q W ~ Tcost-does not await development of a large fuel-cycle industry......./......?..A>...-__ .;.......,.:.x4,.& .....Safety. - Molten-salt reactors have certain inherent features thatassist the designer in providing a safe plant. The salt systems sperateat low pressure with little stored energy; the salts do not react rapidlywith air cx water; some fission products are removed from the primarysystem continuously; and iodine and strontium form stable compounds inthe salt. Continuous fuel processing eliminates the need for excessreactivity, and a prompt negative temperature coefficient is associatedwith the heating of the salt.A safety disadvantage is the accumulation of fission products in theprimary system, the off-gas system, the fuel storage tanks, and the processingplant, which requires provisions to insure that the fission productswill be contained and their decay heat will be removed under all conceivablecircumstances. Partially offsetting this is the ability to drain thefuel into a tank that has an always-ready, redundant cooling system. Inthe Y§BR reference design, this tank is used also as a hold-up tank forthe strongly heat generating off-gas, which means that its cooling systemis always in use and need not come into operation just in an emergency.An added advantage of the design is the use of a natural circulationcooling system that does not need power to operate and can pick up increasedload without action by the control system or the operator.Two other factors may provide Some safety margins over SQlbid-fUebreactors. One is the comparatively %ow power density of fission productsin the fuel salt, which should permit catch basins or crucibles at thebottom of the containment: building to be coolbed well enough that "Chinasyndrome" penetration of the containment still would not occur if allelse failed. The other is the on-site processing, which eliminates theneed to transport fission products at a time when their heat generationrate is still significant.In addition to being able to satisfy these general criteria for abreeder, the MSBR is particularly suitable for development as an alternativeto the Liquid-Metal Fast Breeder Reactor. The reason is that it iscompletely different from the L R, and if the LMFBW were to encounterdiff icubt technical , safety, or economic problems, there is a good chancethat those problems would not be shared by the MSBR. These differencesinclude: fluid fuel versus solid fuel, slow-neutron spectrum versus fast.;.;..... ..B

26BpeCtPUbB, 233u-nl fkkd CyCke V@YSUG h-23 8u Cyde, Il'~0lteIl-Sd.t CQOlalatversus sodim c~olant, processing fuel on-site versus shipping to a centralfaeiliey, graphite core st~ucture versus stainless-steel claddingStrUCtUre, and Slow %eac%or kinetic8 VeTSUS fast killetics. A furtheradvantage arises from the dissimilarity: if both reactors are developedsuccessfully, the differences will provide an opportunity for full economkcompetition in the supply of breeders to utilities, extending back to themining 0% thorium and uranium ore.Thus we conclude that the MSBR can satisfactorily meet the requirementsfor a breeder if its potential is realized, and the differences benit and an DfFlR provide advantages to it as an alternative to theR. In the next section we examine the prospects for success in aehievingthe potential Of the concept.The Likelifisod of SuccessThe development program for a power reactor might be considered asuccess FP it brings all sf tile technology needed for the reactor to thenufactusers can use it to prsduce power plants that can besold to utilities. After having eo~~pleted this review, we conclude thatthere is a ~eas~rtable expectation that an MSBR deveHopment program canacconpiish this T~ a. so, it find s~~.~ti~ns for thetechnical problems we have identified and muse develop dl of thematerials c5nponerits and systems that are needed a Our CORC~US~O~Sabout the major problems are as fo%lows:1.2.3.Recent experiments indicate that there probably are solutions to tineinter gKanU%aX= CraCkFng prQbl@T€l %hat has hung IlleKMCingl~ OVe% theprog~arn during the past year. 8x1 important question at the momentis Whether 'Che reIX2dg Call be a SIIlEda Change ill the CoDIpOSitisn sfHastel$oy N, 0% instead we must substitute an alloy of significantlydifferent csmpositisn that will require an extensive program toqualify it for reaCtClP Use. miS fnatter is being intensiVe%y investigatedat present, and. a preliminary answer sh~uld be availablewithin a few months,Our research program gives hope that means will be forthcoming forreducing the escape of tritium to eke steam systemwitho~t a significantchange in our concept of the reactore We should be able toselect the mast promising meth and demonstrate it in a system ofreasonable size within one 01: two years.:+iueh work remains to be done in the development and demsnstratimOf the processing system for the single-fluid breeder reactor, but:progress so far has been very satisfactory, and this remains one ofthe bright spots in the >mBR development program. The prospects forsuccessfully developing the fluorinstion-reduceive extraction-metaltransfer system are god, and there are same alternative appr~oachesin ease parts of the presently preferred system encounter ~ K I S U ~ Q U ~ ~ -able difficulties. We should n5te that a csmplete demonstration ofthe p?ZCXesSFElg plant Call Only be l€Eide QL% a %$aCtO%, Where repreSenti3-tive concentrations of protacrinim and short-cooled fission productsagfe availablek.3..... C&

274 After these major technical questions have been favorably resolved,a host of tasks remain to be accomplished in taking the engineeringPEE scale to MSBR scale. Those that we recugnize are describedin this report; some are difficult, but all seem achievable. Theremay be others; if so, they can best be identified by doing the detaileddesign of reactor plants and can be brought into the sharpestfocus by development of equipment and systems for those plants.....iu.Solving the technical problems and developing an operable reactor,however, are not done a .guarantee of success for a breeder. The breedermust not only be operable, it must meet the perfornnance criteria we listedearlbier. Therefore we turn to the question of how likely it is that thecharacteristics claimed for the MSBR reference design will be. achieved a.....,.:.a....s.3Fuel Utilization. - Our experience with the MSRE and with the HighTemperature Lattice Test Reactor at Hanford shows that there is littleuncertainty in our ability to calculate the breeding ratis and fuel concentrationof a m~lten-sdt reactor that has a specified care composition.More uncertain is the behavior sf fission products in the MSBR; here theMSE data do not provide a complete basis for prediction, but the u~certaintydoes not appear to be great enuugh for there to be a major errorin the estimated breeding ratio.The fissile inventory depends on the volume of the reactor primab-ysystem and the amount of uranium hold up in the processing plant, as wellas on the concentration in the fuel salt; and these could be somewhathigher than estimated in the reference design. The uncertainties arewhether the heat exchanger can be as compact as postulated in the design,whether the plena in the reactor vessel are large enough for adequate flowdistribution, and whether the hold-up time in the processing plant can beas short as pustulated, While we could have been overly optimistic aboutsome of these, none appears likely to change enough to have a major effecton the fuel utilization....,!:.=Power Cost. - The probability of achieving our power cost criterionis more dffficult to evaluate because it not ~ ~ l involves l y uncertaintiesin MSBR costs but also uncertainties about what the cost of the competingsystems will be. The majo6- cost item in the fuel cycle is the capitalcost sf the processing plant, and this is probably the most uncertain ofthe estimates. We can only say that we think we have mde a reasonablyconservative estimate, including, for example, an allowance of $200 apound for the cost of fabricating molybdenum; and there is additionalconservatism in the processing costs being based on using the processingplant for only 1000 MW(e) of reactor capacity, whereas the unit costs ofprocessing plants come down very rapidly if the throughput is increased.Because the fissile inventory is fairly low and the credit for saleof bred fuel is modest, the fuel cycle ec~nornics of M§BR's are not verysensitive to these factors nox to the cost of enriched uranium. IncreasingUraEltum ore cost from $8 to $16 a pound Without reQptidZatiQi3 Qf thereactor would only increase the fuel cycle costs by about 0.1 milP/kwh.

Other factors have a small effect on the fuel cycle cost; dsubbingthe cost: of ?Lis for example, would add 0.06 miYLs/kwii9 and doubling thecost Of thorium Would add 8,05 mial/kWh. h re%atfVe8y high fiX@d-Chal-gerate Qn the fuel inVetltoPy has been Used in she eS6iPnateaThe graphite replacement cost was estimated assudng that the graphitelifetime will. be only as great as that 0% graphite that has alreadytested, a d exgectea impro~efilents in lifetime w i u reduce tile fiequencyand thus the cost of replacement, The cost of the graphite hasbeen estimated to be about $kO/lb, which is much higher thof most s'sspecialtyP' graphites but could be %ow for a seabemeeting the unique requirements sf the I4SaSBR. However, allowing an additional$5/lb to cover a possible underestimate of the sealing cost wsuPdadd odgr abaut 0-1 mi%l/kwh to the power ssst.We KLUst aCkIIOW%ed@? that eStiE3tiCXt Of Capita%. costs Qf plant§ %O bebuilt far in the future wit11 yet undeveloped techwoi~gy is m-1uncertainties. Because of the way EWR cost data were used, ttaese uncertainties,as we judge It, h ve more to do with the design of the plantthan W i t h OUP ability to DldCe COSt cCiE@3risCX%3 far Et given design. Nevertheless,there is limited room for error in the comparison with an LWRbecause the cost sf 8~react0r equipmenttF (which includes 9the reactor itself, the salt pumps, the heat exchangers aators, the salt storage tanks, and the off-gas system) is Only one-thirdof the total 666% af the power plant.actor that can affect the power cost is plant availability.t reactors do no% have to be shut down for refueling, andtile frequensy Of graphite replacement is %OW aask be sc'kseadea to COincidewith major turbine maintenance, MSI3R's start off with an availabilityadvantage over L\-R'ss. The KSBR plant must, however, be made reliableand must be specially designed so that the initial advantage is satoffset by the increased difficulty of maintenance.Our- estimate has the cost of power from an KSBR being about 0.5 mil%/kwh less than that of a light-water react^^ at present uranium ore prices.If uranium ore costs increase by $$/ab by the time breeders are introduced,the cost advantage sf an MSBR would increase by Q -3 mill/kwh e Thus thereiS 2, faiK marP@n f8H $%Tor in CoIllpariSC?n With present day uJRgSs. MOWever,LWR costs are certain to change same in two or so decades, and KTGRPsrather than %WRSS could be the cQlnverter with which to compete at the time.Thus strong conclusions about the MSBR meeting our cost criterion are notpossibbe, but the chances seen reasonably good with low uranium prices,of course, increase as the cost of uranium ore goes up...... u.s,~*....1.L

29En conttast to the probability of a big accident, the wider digtributionof radioactivity in a molten-salt reactor and its processingplant increases the chances of small releases. The ability to restrictthe release of radioactivity during norm1 operation or maintenanceperiods to a desired level appears to be a matter of cost, and the provisionof adequate containment and clean-up system should reduce therelease rate to 85 ISW a8 desired.These considerations convince us that the likelihood of being ablet~ develop an ICBR that can meet the requirements for a successful breederis good. We must note, however, that skeptics will sometimes acknowledgeuments that we have cited but then raise questions about otherfactors that they think d&t be iKLIpOKtant. One question has to do withthe practicality of operating and maintaining a reactor in which fissionproducts spread throughout several parts of the plant. We believe thatthis is a serious matter to be treated thoroughly in the design, but weare convinced that it can be handled economically. Our views are basedon experience with the operation and maintenance of four fluid-fuel reactars,and most importantly, by our favorable experience with the MSW.Larger plants will have larger COTD~Q~~II~S to be handled and higher levelsof radioactivity, but we believe that the ISRE mintemance approach, usedwith careful design of the plant so that maintenance requirements areanticipated, will permit repair or replacement to be done in reasonabletimes aA second question is whether the chemical nature of a molten-saltreactor and the requirements for operation of a processing plant in conjunctionwith it will make utilities unwilling to purchase such systems.Our response is that utilities have been willing to take on other advancedtechnologies, such as nuclear energy and the use of supercritical stearn,and they seem willing to undertake operation of system c~ntaini~lg moltensodium. None of the utilities that have evinced an interest in moltensaltreactors has indicated to us that the requirement for chemical reprocessingwould cause them to be unwilling to sperate an MSBR,The final question is sometimes stated this way: "If molten-saltreactors are as g~od as you in Oak Ridge say, why aren't industrial firmpresently working on them?" The answer to this has several facets. Firstof all, the commitments of the present utility suppliers to other systemswere a11 made before the MSW had operated and provided a demonstrationthat the technology was more practical than some had expected. Second,the manufacturer of an MSBR cannot expect the significant fuel supply andrepr~cessing business - the "razor blade" business - that can accompanythe sale of a solid-fuel reactor, and thus the potential for future profitsseems less with molten-salt ~eactors. Finally, the budget of the AtomicEnergy Comission for molten-salt reactor development has been relativelysmall, and the AEC has mde no commitment that it will assist in the developmentof molten-salt breeder reactors. Hence a manufacturer who consideredundertaking molten-salt reactor development could not be sure ofreceiving the kind of development support that has been forthcoming forother reactor concepts e

II30The objective of developing breeder reactors is to obtain a sourcesf low-cost energy for ourselves and for future generations Molten-saltbreeder reactors have attributes sf fuel utilization, ecsnsdcs, and safetythat make them well suited to serve that purpose. '%he highly successfuloperating experience sf the Molten-Salt Reactor Experiment and the developmentsin chemical processes that have allowed an important simplificationin the bPeeder ConCept provide SkkPpO%t f0BT the CQnteHnCiCNI that P&3CtOBfGhaving these characteristics can be successfully developed Because theydiffer in lD.2ifa)p respects f%Qm LiqUid-Metal FSt Breeder &38CtoKS, ?d%BFP%serve particularly well as insurance for the nation's energy supply inW encounters insurmuntable obstacles a Moreover, if bothsystem are suc eSsf%illy deVelOped, the 2hility sf the HQlten-Sdt reaC%C!I?to startea IL as a breeder or operated economically as a converter onpEutonium, 245Y or 2 3 3 ~ makes it: a suitable companion for an UEBW toaneed fuel economy. In this case, the differences betweenthe two concepts, starting back at the mining of ore, also provide increasedsppsrteantity for a competitive breeder industry *Thus far the work on nolten-sdt reactor technology has establishedwhat we believe to be a fim foundation for success. There are stillsome basic proble notably surface cracking of Mastelkoy N and tritiumment 9 whose solutions h e not been fully demonstrated, We alsoize some major engineeri tasks, such as demnstration of thedependable, econo~eal maintenance of an MSB and scale-up of reactorequipneat to MSBR size, Nevertheless, the p tential of molten-saltreactors is proesing ensugh to ju ify a continued effort.Favorable resolution of the b ie problems must come first, thenmore extensive development of cs%ap ents and system. If these proceedsatisfactorily, the construction a% a reactor should corn next, but thescale-up need not be large since a 156 - 200 N"(t> reactor codd demonstratedl of the techlnOlogy that is esSenfLial %or a EK3Eten-sa%t breeder,This reactor should provide sufficient Pnformakion for concluding whetherfull-scale EBB'S will be technicalPy and econo~cally attractive enouto justify completing their development 0 We believe that a strengly~ o t i ~ and ~ t ad : ~ UEitely ~ funded pPJgrZLrrt that fOlHOWS this route W i l llead to molten-s t bPeedeP Ke%rCtO%S that 6aR play a 'iilZi.jOr role in providingthe natio s future enerh..E..

~FebruaryReferences for Chapter 11 e “AuthorEzing Appropriations for the Atomic Energy Condssion forFiscal Year 1973,” House of Representatives, 92d Congress ReportNo. 92-1066..&.......... (1111.~....Lm...“.y .~....&:.a .........s&2. D. Fb. Cardwe11 and B. N. Naubenreich, Indexed Abstracts of SeZectedRefeyences OM Molten-Salt Reactor TecianoZogg, OWL-m-3595, December1991.S. USMC Task Force, Repore on the Fluid PwZ Reactors Task Fmce,TID-8507 ~ 19594. ConceptwZ Berip Study of a Single-FZuid MoZten-Salt Breederi?eactor, ORNL-4541, June 6371.5. H’sitelz Salt Reactor YechnsZogy, Technical Report of the Molten-Salt G~PQUP, Part I, Ebasco Services, Inc., December 1971.6 e EVahatiOl., of a IOOO-MW(e) MoZter,-SaZt Breeder Reactor, TechnicalReport of the Molten-Sa%t Group, Part 11, Ebasco Services, Pnc.,October 19717 2800 ~Wle) bfoZten-SaZt &ee&r Zeaetor Conceptual Desigrz Study,Final Report - Task 1, Ebasco Services, Pnc., February 1972.8. Molten-Salt Breeder Reactor Associates Fina’c RepYdt, Phse IStudg - Project for Investigation 07 J!o Zten-SaZt Breeder R~acto~~Black and Veatch Consulting Engineers , Kansas City Mo. (1970).9. Ps-keiztiaZ JPzcZear Fouer G~outh Patterns, prepared by the SystemAnalyses Task Force under the direction of the Division of ReactorDevelopment and Technology of the USAEC, WASH 1098, December 197010. Cost-Beneflt AnaZysis of the U.S. Breeder Propam, Division ofReactor Development and Technology, USAEC, ILGH 1126, April 1969.11.CivLZicnz FPzcZearr Power - A Report to the Fpesident - 2962, U.S.Atomic Energy Cormissfon, U.S. Government Printing Office,Washington, D. C

2. EVOLUTION AND DEVELOPMENT OF MOLTEN-SALT REACTORSP. N. HaubenreichOrigins....&:a.....**. . :. ..,. . . . &... ".y.L2. ..h%en the idea of the breeder was first suggested in 1943, the rapidand efficient recycle of the partially spent core was regarded as themain problem el]. This problem, which is still crucial in breeder economics,was attacked in two ways - by striving for very Long burnup andby seeking to simplify the entire recycle operation. The. latter pursui%inevitably led to consideration of fluid-fueled reactors as the ultimatein fuel cycle simplificatian.Neutron-multiplying systems consisting of aqueous solutions and slurrieswere investigated soon after the discovery of nuclear fission, andthe first effort toward a fluid-fueled breeder was based on these systems,in which the fluid is both the fuel and the moderator [2, pp. 1-91 TheHom~geneous Reactor Program, organized at ORNL in 6949, had as its objectivea reactor with a uranyl suPfate-DZ0 solution core and a thoriumoxide-B20 slurry blanket, separated by a Zircaloy cure tank, This concept,with its superior neutran econumy, offered good 233U breeding performance*About 1950 the idea of a very different fluid-fuel reactor for powergeneration emerged at Brookhaven National Laburatory fmm studies on Powmeltingalloys and slurries of uranium and thorium in liquid metal. Thiswas the liquid-metal-fuel reactor concept [a, pp. 699-9252]. A version ofthe LQ'R using graphite moderator, U-Bi ss%ution core, and ThOZ-Bi slurryblanket appeared capable of breeding.Meanwhile yet another fluid-fueled reactor had been conceived f ~ an raltogether different purpose - aircraft prapu%sion. Several differentc~ncepts of compact reactors were being considered for generating heat t~be used in a jet engine. The Oak Ridge idea was to use a high-temperatureliquid fuel that could be circulated to remove heat from the core and bedrained for refueling. Experiments to investigate molten-salt fuels werebegun in 1947, and 3 years later molten fluorides were chosen for the maineffort sf the Aircraft Nuclear Propulsion (ANI?) program at ORB% [3]. Thefluorides were particularly well suited because they offer low vapar pressureat jet-engPne temperatures reasonably good heat transfer properties sand immunity to radiation damage, and they do not react violently with airor water. (See Chapter 5.) A small reactor, the Aircraft Reactor Experiment,was built that used a fuel mixture of NaF, zKP4, and UFb eireplatingin Pnconel tubing through a moderator assembly of Be8 blocks [2, pp. 673-881.In 1954 the ARE was operated successfully €or 9 days at outlet temperaturesranging to above l680'F and powers to 2.5 MGJ(t) in investigations of thenuclear dynamics of the circulating fuel system.It was recogni~ed from the outset that molten-salt reactors might beattractive for civilian power applications, and in E956 a group was formedat QRNL to study the characteristics, performance, and economics of moltensaltreactors for central station power generation [4]. A wide variety sf33

34Relation to Other Fluid-Fuel ProgramsEarly in I959 a task force &sembled by the M C made a comparativeevaluation of the three fluid-fuel reactor concepts then being pursuedThe ccsnclusion was that the rioTten-salk reactor, although limLted inpotential breedin gain, had "the highest probability of achieving technicalfeasibility" [5] * Soon thereafter work on the ~ C ~ U ~ QhomogeneousU Sand Liquid-metal-fuel reactors was discontinued, Leaving the ni~bten-sabtreactor as the Tone fluid-fuel breeder concept still being supported bythe usme.*Although the Nolten-Salt Reactor Program, as such, was relativelyyoung, there was an extensive technological base from the ANP program,where $60 million ksaa beern invested in molten-salt reactor techno~ogy.Some of this had gone for developments specific to the c~mpact: aircraftraration, but a large fraction of the technology was equally appPitothe civilian power reactors that were being envisioned. Thephysical chemistry of interesting fluoride salt mixtures had been explored,and a container alloy had been developed that was especially cornpatiblewith fluoride salt mixtures and which had significantly higherstrength than Hnconel at the 1500-l6OQQF temperatures required in anaircraft reactor, Originally called IKOW-8, this alloy is now generallyknown as Hastelloy No Techniques for producing, purifying, and analyzingfluoride mixtures had been worked out, and considerable experience wasg2iRed in handPing the fuel. The fl~~ride volatility process wasdeveloped and was successfully used to recover the uranium from the AEfuel in 1957-58,En addition to the generally applicable &'P work, there was somespinoff %S the mobten-s2lt teChnSlogy frclRl the aq%keQUS kOlni3geneoUS Peactorana liquid-metal-fuel reactor program e he H C X W ~ ~ I I ~eactor~ Q ~ ~Program had built and aperated two reactors using circulating aqueousfuel solutions at 250-3OQ"C. considerable maintenance w2s required onradioactive parts of these reactors, 2nd one significant contributionto reactor techno1ogy was this experience with maintenance of highlyradioactive systems [7]. A chemical pr~cessing scheme explored for theliquid metal fuel reactor involved molten salts and molten bismuth. Theexperience of this effort, and the general background of experience withmolten bismuth, proved valuable when extraction system involving moltenbismuth became the heart of the processing concept for MSBW'ss.a,:.>i,........ .=. .. c.sdiWork has continued at KEKA2 Arnhem, Netherlands, on a z-~~c~QL-concept using an aqueous suspension fuel [SI.

35Early MSBR Concepts.y.*...,.*>4.'.....'"il :.:........,.XdIn the early days of the Molten-Salt Reactor Program, serious CORsiderationwas given to homogeneous reactors in which the core containednothing but salt. These ideas were abandoned after calculations showedthat the limited muderation by likely fluoride salt constituents alonewould result in a thermal reactor with inferior breeding performance.Breeding appeared possible in intermediate-spectrum reactors, but theirgains were not high enough to compensate for their higher fissile inventories.Studies Sf f%St-SpeCtrum mC9lten-Sdt reactors (Using chloridesalts) indicated good breeding ratios, but fissile inventories were excessiveunless unconventional heat transfer systems were used to minimizeholdup outside of the core.After experiments showed that bare graphite could probably be used inthe core of a molten-salt reactor, MSB efforts concentrated on graphitemoderatedreactors having well-thermalized neutron spectra and How fissileinventories e Two general types were considered - single-fluid reactors inwhich thorium and uranium were combined in one salt, and two-fluid reactorsin which UF4-bearing fuel salt was separated from fertile salt containingThP4. En any ease the dilraent fluoride mixture would be 'LiF-BeF2 insteadof the NaF-ZrFb mixture used in the aircraft reactors; the 7LiF-BeF2 absorbedfewer neutrons and dissolved more ThFk without excessive liquidustemperatures e The single-fluid reactor was relatively simple and promisedlow power costs, but breeding appeared to be impractical because of R~Utronleakage atnd losses to protactinium and fission products [8]. (Atthat time it was not clear that Pa and fission products could be separatedeconomically on a very short cycle.) The two-fluid reactor could be designedwith a fertile blanket to reduce leakage, and Pa losses would bereduced because the fertile salt would be at a lower average flux. Theonly processing required for the fertile salt was fluorination to recaverthe bred uranium. The fuel salt could be processed by a combination offluorination and an aqueous process. The two-fluid reactor was more complexin that it used tws salts that had to be kept separate, but it didoffer attractive breeding performance.The MSWE.....=.... ,:paPurposeBy l96Q a fairly clear picture of a family of molten-salt reactorshad emerged. The technical feasibility appeared to be on a ~.oundfooting - a compatible combination of salt, graphite, and containermaterial - but a reactor was needed to really prove the technology.'khat was the purpose of the Molten-Salt Reactor Experiment: to demonstratethat some of the key features of the proposed molten-salt powerreactors could be embodied in a practical reactor that could be operatedsafely and reliably and be maintained without excessive difficulty. Forsimplicity it was to be a fairly small, one-fluid reactor operating at16 W(t) or less, with heat rejection to the air via a secondary salt.

36The x3R.E flowsheet is shown as Fig. 2 s 1. Figure 2.2 shows some detail0% the 5-ft-diameter reactal- vessel. The fuel was LiF-BeF2-%rFq-UP4(Q4-%Q-5-l mole x>, the secowdary sa%t was LiF-BeF2 &M-34 mole X), themoderator was rade CGB gYL3phite, and all. other parts contacting saltwere of Hastellsy N. The bQW1 of the fueB &M ~2s the surge space forthe circulating loop, and here absut 50 gpm 0 fuel was splrayed into thegas space $6 a%Bcw xencm and krtyptsn to escape from the saEt. Also inthe pump bowl was a port through which saLt samp%es could be taken orcapsules 0% concentrated %uel emishing salt (UF4-EiF or I?uF3) csuld beintrsduced. The fuel system was located in sealed ce%%s, laid out formaintenance with long-handled tosls tkmu apernings in the top shieldimg~A tank uf LIF-BeFp ~~31% was used ts flush he fuel circulating systembefore and after maintenance. In a cell adjacent to the reas%sr was asimple facility ~CJP bubbling as th~~klgh the fuel or flush salt: B2-I-IFts remove oxide, F2 to remove uraflium as UFg o References 9, 10, and %kprovide more detailed descriptions of the reactor and processing plant.Beve%opment and ConstructionMost sf %he 'PEW effort from 196Q t&rough 6964 was devoted to design,developmen%, and cQnskPuc%isn of the %%sEE, Pr~ductisn 2md further testingof graphite and Bastelloy N, both in-pile and out, were major developmentace ivisies * Qthers included wosrk cm reactor ekemis try9 developmentof fabrication techniques for Hastelloy N, development of reactor csmponentsp and remote-maintenance gPannin apld preparations. (A csnvenienksummary of developments through the end of major construction is givenin reference 12.1Befsre the MS development began3 tests had shown that salt wouldnst permeate raphite irk which %he pores were very sma%%, Graphike wiU-kti-ie desired p -re structure was available mby im small, experimentaklyprepared pieces, however, and when a manufacturer set out. to produce anew grade (6GB) to meet the MSWE requirements, difficuPties were encsuntered[la, ppe %-I%-%89]. A series of pitch impregnations and heattreatments praduced the desired high density and small pare structure,but im the final steps occasional cracks appeared in many of the 2-1/4-h.square baafs o Apparently the cracks resulted because the structure was sotight that gases from the pyrolysis of the impregn t csubd ~0% escaperapidly el40ugh D Tests shswed, howeveap) that the cracks did not propagate,ekes when filled with salt and subjected to repeated freeze-tAfter analysis showed that heating in salt-filked cracks woulcessive, the graphite was accepted and used in the MSRE.The choice of Hastelloy N for the MSRE was on the bases 13f the prsmisingresults -0% tests at k&P csnditfsss and the availability sf much ofthe required metaliurgical data.* Develegment for the SE generated theki!xAny attempt ks develop a less-expensive aHHoy, possibly even bettersuited to KS CsRditioms s was precluded by time and s~st considerationsgs2, p* %%k].

?,*.... ,;.a:.... . .:.:.;.):....:*a....,*.:2........*. .,!

38GRAPHEVE SAMPLE ACCESS PORTFLEXIBLE CONDUIT TOCONTROL ROD DRIVESCOOLING 818 LINESd166ESS PORT CQOblNG JACKETSREACTOR ACCESS PORTCONVRQL ROD THIMBLESOUTLET STRAlNEWCENTERING GRIFUEL BMLETRE&CTOR CORE CAREACTOR VESSELART[-SWIRL VANESMOD E RATOWSUPPORT GWBOk..

39further data required for ASME code approval. It also included preparationof standards for Hastelloy N procurement and for component fabrication.Material for the MSBE, amounting to almost 200,000 lb in a varietyof shapes, was produced commercially. After weld-cracking in experimentalheats was overcome by minor composition changess these was no difficultyin obtainhg acceptable material Requests for bids on component fabricationwent to several companies in the nuclear fabrication industry, butall declined to submit lump-sum bids because of lack of experience withthe new alloy. Consequently all major components were fabricated in AECownedshops at Oak Ridge and Paducah [12, pp. 63-82]. After appropriateprocedures were worked out Hastelloy N fabrication presented no unusualproblems *At the tfme that design stresses were set for the MSE, the fewdata that were available indicated that the strength and creep rate ofHastelloy N were hardly affected by irradiation. An arbitrary allowancewas made for possible effects, however, by extablishing design stresses28% below Code values for unirradiated Hastelloy N. After the constructionwas well along, the stress-rupture life and fracture strain werefound to be drastically reduced by thermal-neutron irradiation. The MSREstresses were reanalyzed, and it was concluded that the reactor wouldhave adequate life to reach its goals. At the same time a program waslaunched to improve the resistance of HastelEoy N to the embrittbement.(See Chapter 7 and reference 13.)An extensive out-of-pile sorrosi~n test program was carried out forHastelloy PJ [12, pp. 334-3431 which indicated extremely low corrosionrates at MSE conditions - Capsules exposed in the Materials TestingReactor showed that salt fission power densities of more than 200 W/cm3had no adverse effects on compatibility of fuel salt, Hastelloy N, andgraphite. Fluorine gas was found to be produced by radiolysis of frozensalts, but only at temperatures below about 100°C 61.2, pp. 252-2871.The results of this program are described in some detail in Chapter 5.Components that were developed especially for the MSRE includedflanges for 5-inch lines carrying molten salt, freeze valves (an aircooledsection where salt could be frozen and thawed), flexible controlrods to sperate in thimbles at 1200°P, a d the fuel sampler-enricher[12, pp. 167-1901. Centrifugal pumps were developed similar to those usedsuceess€ub$y in the aircraft reactor program, but with provisions for remotemaintenance, and including a spray system for xenon removal. Remotemaintenance considerat ions pervaded the MSRE design and developmentsincluded devices for remotely cutting and brazing together 1-1/2-inchpipe , removable heater-insulation units, and equipment for removing specimensof metal and graphite from the core.The MSW development program did not include reactor physics experimentsor heat transfer measurements. There was enough latitude in the MSREthat deviations from predictions would not compromise safety QP accomplishmentof the objectives of the MSKE*Construction of the primary system components and alterations of theold ARE building (which had been partly remodeled for a proposed 6O-W(t)aircraft reactor9 were started in 1962. Installation of the salt systemwas completed in mid-1964. OWL was responsible for quality assurance,

nagement of construction 6141. The primary system wereL forces; subcontractsrs modified the building and installedancillary sys tens *.....

.41....c.:.:

42_-_P Pthe reactor. Limitation of oxygen access to the salt proved quite effective,and the tendency sf fission products to be dispersed from contaminatedequipment during maintenance was less chan we had anticipatedOperation of the XSRE provided some insights into the unusual problemof tritium in a molten-salt reactor. It: was observed that about 6-10s/,of the calculated 54 Ci/day production diffused out of t.he fuel systeminto the containment cell atmosphere and another 6-10% reached the airthrough the heat removal system [IS] Tie fact that these fractionswere not higher indicated that so~ething (probably oxide coatings) partiallynegated the easy transfer of tritium through hot metals.The one quite unexpected finding of great importance was the sha%%swintergranular cracking observed in a11 metal. surfaces exposed to the fuelsalt. This was first noted in the specimens that were removed from thecore at intervals during the reactor operation. Post-operation examinationa% pieces of a control-rod thimble, heat-exchanger tubes, and pumpbawl parts revealed the ubiquity of the cracking and emphasized its i m p ~ -tawce to the MSR concept - Further investigations and possible consequencesaipe discussea in chapters 9 ana 14 of this repo~t.Since the MSE, the Xolten-Salt Reactor Program has been a technologyprogram, not focused on building a particular reactor, but seeking toidentify and accomplish the develwprnents that are needed before moltensaltbreeder reactors can became a reality 1181. In the furtherance of thisprograrr,, efforts on conceptual design have been essential in defining theneeds for development, while experimental findings , in turn; shape theconcept. This section describes the reactor concepts that have beenconsidered in the course of this intertwined precess.As described previ~usly, at: the time that she MSRE was conceived,the two-fluid reactor, despite its relative complexity, seemed to heldthe most promise as a breeder. During the early years of the MSRE,relatively little effort was devoted to refinement 0% conceptual designs.Basic chemistry studies ~sntinued, however, and led in 1964 to an importantdevelopment that simplified the processing in the two-fluid breederplant [l2, p. SQS]. This was the separation of rare-earth fluorides frontLiF and BeP2 by distillation at 1QOO"C. (The practicality was Eaterdemonstrated with a portion of the MSWE: fuel [I91.> Thus it was, whenthe MSWE settled into operation, that desiefforts focused on the two-k...,....*ii;-Ifluid concept eStudies at first indicated outstanding resource utilization, mainlybecause sf an extremely low specific inventory of about 0.8 leg fissiEe/m$(e)[%O]. Then in 1969, when irradiation of graphite to very high neutronflueraces revealed more rapid dint~nsi~~al changes than had been proj e~ted,the two-fluid concept was dealt a severe blow. Accommodation of the dtfferentialgrowth of the graphite EIde the COre design and asSePil$ly 80complex that it seemed necessa y to replace the entire reactor vesseland its contents whenever the raphite core tubing became unserviceable eme reference aesi -NM(~> included four 556-m~t)

43..

~ .~"-f~-uorination~ auvd),44One alternate uffers a way around the necessity of replacing theat bk-year iRtE2rValS. This difficult job is avoided by shply makingthe core ~f the breeder Barge enough and the damage flux low enough. thatthe core graphite will. last the %Q-greah- life ~f the plant. Such a "perm~e~~t-~~re"K~ELC~QE- can have a breeding ratio as high as in the referencedesign, but the large core means a greater inventory and longerdoubling tiate pq.hother possible simplification is the elimination of most of thechemical processing. Hf it should happen in a breeder plant that pratactiniumand fission-product removal were stoppedS the reactor couldcontknue t~ sperate for months, or even years9 as a near-breeder wFthonly the 5f u.k-aniuw. ~aternatively the reactor souidbe built as a co~lverter with no processi~g (other than perhaps oxide re-_-Some studies indicate that an economical made of operatian would> be to run for 6 equivalent full-power years, recover the uranium by batch(as demonstrated in the KSRE), discard the salt with the-4fission proaucts, ana ~I~SUI~I~ with fresh carrier salt. ith her 235u cxplutonium could be used for startup and feed of such a molten-saltconverter reactor.A limited amount of c~~~~eptual desip was done on a 356-MW(e) converterthat positively o~ercome~ the problems of graphite replacementana triti~rn co~~tainm~~~t with a mini~~mof additional development. mecore is large enough SO that presently available graphite would last30 yearss and tritim containment is ensured by using an intermediatesalt loop containing Hitec [24]. The Mitec, a "iOg-69aNO~-Nao~ mixture,reacts with tritium to form water, which could be stripped continuousLyfrom the salt. Because the nitrate-nitrite mixture wouLd precipitateuranium and might react vi~lentry with graphite if it leaked into thefuel system, a compact hop containing LiF-BeF2 (which is quite compatiblewith the fuel) is interposed between the fuel and Hitec system,The disadvantages of the extra bosp and temperature limitations Hitecare weighed against simplification uf the steam system because of therelatively &ow melting point of Eites (288"F>. In any evenc this studyis an offshoot from the main line of 8KXL effort, which is directed atthe high-performance breeder.In considering the most expeditious route to the ultimate moltensaltbreeder, we have devoted some attention to the conceptual design~f what we call the K~kten-SaIt Breeder Experiment (MSBE). This reactorwould be designed to operate under conditions at least as severe as thosee plant. Design studies show that a graphite damage flux twicethat in the reference XSBR and power density, salt composition, and prstactiniumand fission-product concentrations like those in the referencedesign could be attained in a ass-m(t> reactor with a 4-ft core in a7.5-ft vessel [25]. The NSBE would produce steam and would include acomplete processing facility. Although its small dimensions limit itsbreeding ratio to about 0.96, the MSBE provide a definitive testof a16 the basic equipment and psacesses required for a breeder.r,L........ . ....,

4%Current Programs"',.....:..%...._..^.....:.:.y... .y.:.4....924....;,.:.:.,......Z.@. >-.............s;

46The emphasis in fuel processing is on ~eductive extraction - thechemis try of fluoride-bismuth and chloride-b ismuth sys tern and engineeringequipment to exploit this shemis try - Fluorination and fuel reconstitutionare being developed, and alternate processes such as oxideprecipitation are being studied.. ..liii.".-here have been ~ V Q privately fmaed conceptual design studies andevaluations of MSW'ss. The first vas by the Molten-Salt Breeder ReactorAssociates (NSB >, headed by the engineering firm of Black & Veatch anddwestern utilities. The MSBM study identifaed problemareas but c~n~l~ded that the econsmics of molten-salt reactors were at tractiverelative to light.-water reactors, and favored a program leading t~early comercia% application of a moIten-salt c0nverter [as]. Since theconclusion of their study in 1490 the XSBhas bee^ rekat ively inact ive eThe second privately funded organization is the Molten-Salt Group(MSG) Whose %Ormatfon Was afanQUnC@d ill 1969 [as, Po l-l] a me MSG isheaded by Ebasco Services and includes 5 other industrial firms* and15 companies easi. pn the fan ai the MSG conpietea anevaluation of the state of b%% technologgi 6281 and a critique of theOR", lOQO-MFi(e) breeder design 1291 They con~%uded in the first reportthat the existing tC?Cht%OlOgy Was Sklfficient to justify CQnStKUCtion Ofa d€XmnStPatiQn plant that W Q U ~ breedp ~ provided the processing worksas intended, but that its maintainability, reliability, CQS~S., and plantlife could not be predicted reliably from the existing te~h~ologgr [a$,p. 61. En the second report, they concluded that the OmL referencedesign ~HlbPaCed SQBlE! teChRQbogical diffiCU1ties but Was a SUitahledeparture point for exploring MSBR tech~~lbogy [29* p. 41. The MSGmembers have agreed to support studies of a demonstration plant andalternate molten-salt reactor concepts and an updating of their MSBRtechnology evaluation. These studies are proceeding concurrently withthe Me-funded MSBR design study that is being done by part of the GPQUP'SWQrking fo%Ce.There have been other indications of interest in molten-salt reactorsby industry and utilities. Xany have sent representatives to MSRP annualinformation meetings, and several have made private studies of theirpossible role in XSR development. A few have assigned staff members towork in the MSRP (far up to 2 years), and materials producers have contributedby providing experimental graphite and alloys for evaluation.....t.C.3&.,The Indian Department of Atomic Energy is interested in molten-saltreactors in connection with their vast thorium %~SQUE^C~G and the anticipatedavailability sf pHutonium in India for MSR startup. A smll program* Babcock Ed tJilcox, Byron Jackson9 Cabot Corp. Continental Oil,and Union Carbide.

47of research on the MSBR concept has been underway at the Bhabha AtomicResearch Centre since 11969. The program is mostly concerned with investigationsof the chemistry of melts containing PuF3. The USAEC and the DAEhave been exchanging molten-salt information, and an agreement providingfor somewhat broader cooperation is being considered.Euratom has supported research OR various topics pertinent to moltensaltreactor technology. Prom 1964 to 1966 there was an exchange ofmolten-salt information through a formal agreement with the USAEC, butthe exchange is now relatively inastive. Development work involving constructionand operation of a molten-salt steam generator that was begunwith Euratom support is continuing at Delft University, however. OtherEuropean molten-salt reactor work was conducted at KernforscherngsanlageJPiliek in 1963 to 1967, mostly in connection with the epithemal reactorconcept, MOSEL [3Q]. This has been discontinued.The U W A has supported a sm11 effort on molten-salt reactors forseveral years. Studies and a limited amount sf experimental work arecontinuing, with emphasEs on chloride-fueled, lead-cooled fast reactors.

References for Chapter 21.2.3.J. A. Lane, H. G. :~lasHherssn, and F. Maslan, Fluid Fuel Reactors,Addison-Wesley, Reading, Mass., 1958.....U.X.'... ~. -. .....4.5.6.. ..u.:s7.8.4.16 011 P12 013.15 D

4916, B. E. Prince, S. J. Ball, J. €2. Engel, P. 1'9. Baubenreich, and T. W.Kerlin, Zero-Power Ph2sics Experiments WL the lGShZ3 OWNL-423.3(February 1968).17 e R. Be Briggs, "Tritium in Molten-Salt Reactors, I' Reactor Technokqd,- 14, 335-42 (Winter 1976-72)18. PI. G. MacPherson, "Development of Materials and System for theMolten-Salt Reactor Concept, I' €?@actor Technol. - 15 136-155 (Summer1972) -14. 9. R. Hightower, L. E. McNeese, B. A. Hannaford, and H. D. Cochran,LozJ-PTesawe DistiZZation of a Portion of the Fuel Cmier Salt f~om-the MSZE, ORNIL-4577 (August: 1971) e20. I?. R. Kasten, E. S. Bettis, and R, C. Robertson, Desip Studies sflBOO-MTY'(e) Molten-SaZt Breeder Reactors, ORNL-3996 (1966) I21. R. C. Robertson, 8. B. Briggs, 0. %. Smith, and E. S. Bettis, 2%~fluidMoZ.ten-SaZt Bpeeder Reactor Design Stud$ (Status as of Jan. 1,2368), OmL-4528 (1970) -22 Concept-ml Design Study of a Single-FZuid Molten-Salt Bme&r !%actor,OBWL-454a. (1971).24. E. S. Bettis, L. G. Alexander, a ~ PI. d L. Watts, Desip Sttrdies of QIdolten-Salt Reactor Derzonstration Phe, OWL-TM-3832 (June 1972) .25. J. R. MtWherter Molten-Salt B~eader Eqwfiment Design Bases,QML-TM-3177 (November 1970) e26. Molten-Salt Breeder Reactor Associates Staff, Final Report, Phase1 Study - Project f ~ r Investigation of Molten-Salt Breeder Reactor,Black & Veatch Consulting Engineers, Kansas City, Mo. (19709.27. b000-NbJ(e) Molten-Salt Breeder Reactor Conceptual Design Study,Final Report -Task I, Ebasco Services, Inc., New Ysrk, February1972.28. MoZten-Salt Weactop TeehnoZogy, Technical Report of the Molten-SaltGroup, Part I, Ebasco Services, Ins., December 6991.29 Evaluation of a 1000-1We MoZt-en-SaZt Breedem? Reactor, 'FechmicalReport of the Molten-Salt Groups Part PI, Ebasco Services, HIPC.,October 1971 030 e P e €3. KasteR, "The MOSEL Reactor Concept '' a"vzird Ye Us Intern. Conf.Peaceful Uses of At. B~argj, Genawa, A/CBEu'F 28/H/528 (1964)

3. DESIGN CONCEPT OF THE SINGLE-FLUID MSBRM. W. Kosenthal.......SW9"ne preceding chapter discussed the development of molten-salt reactortechnology and the evolution of design concepts over the years. Inthis chapter is described the single-fluid thema1 breeder reactor thatwas selected in 1968 as the focus of QmL's development pr~gr~i:rra. At thattime we began the conceptual design of a lOOQ-PIW(e) MSBR that exploitedthe new developments in chemical processing and the new approach to coredesign that were mentioned eaPlier, and in 1971 we issued ORNL-454l [I],which describes the concept as we visualized it in 1970. In its generalfeatures,, that concept continues to be the basis for the QRNL program,although our ideas about specific features of some of the system andcomponents have changed. The OWL-4541 concept with some modificationsis generally referred to as the reference design. Most sf the ~ ~ ~ ~ h i n gchapters of this report treat specific features of the plant, and sametimesnewer ideas are presented than those given in ORNL-4541. When thisis done, hmever, the changes from the earlier design and the reasons forthem are generally discussed.In addition to OWE'S appr~ach sometimes having changed in the lasttwo years, the views of Ebasco Services and its associates on specificfeatures of MSBRs sometimes differ from those of BmL. The major pointsof difference are listed later in this chapter, and more information aboutthem appears in the later chapters.Finally, at the end of this chapter we call attention to ways inwhich the design of single-fluid MSBWs might differ if the criteria werechanged from those that we followed in selecting the reference concept.Om% Reference Design..., .....:.:,&,In selecting the reference design, we generally attempted to attainobjectives in four areas: Q P ~ eonsematisn, power costs, safety and environmentalconsiderations, and technical feasibility * The goals werenot stated quantitatively, however, and our approach has been more to seehow well the colacept can be made to do and to decide whether that is 86-ceptable, than to meet specific criteria. Co~~~p~omises among the factorswere a necessary and continuous part of the design effort, and OULK viewsof the require~~~ts have shifted some as attitudes in the nation and inthe atomic energy programs have shifted. Nevertheless, we have held tothe following guiding principles......;.y x..4..... . i/,d

52breeder that by itself can satisfy both the wear-tern and long-term needsfor a breeder reactor 0% (2) a reactor having lower perfomance QK differentcharacteristics that, as a campanion to another type of breeder,could improve the overan1 nuclear eeonomy. our goal was a concept thatsatisfy the first requirement, although we reeo ized that the techneededfor the second wou% in m3s-C aspects be demanding thanf the hi~h-~~rfo~~ncereactor and, thus, would generally becomeavailablbe also from the work on the other concept.This general objective w translated into wore specific requirementsby computing the cumulative uranium ore requirements for the U.S. ueinthe AEC's projection of nuclear power growth betmeen now and 2600 andpostulating dates of introduction sf breeders havin various fuel utilizationcharacteristics - These studies showed that $ since the breedingratio sf a thermal breeder cmnot reatly exceed 1.0, the fuel inventorymust be kept low to achieve 8 tiSft3etOr)p pePfQPaanCe, Further Studiesthen identified the MSBR desi conditions that gave what appeared to beacceptable uranium ore requirements. Consideration sf the availabilityand anticipated demands for other materials needed to support a large MSBRindustry indicated that MSBW development would not be hindered by shortagesand rising costs 621 *.... cs.0 - As with other types ofreaetbeen fie pPOtectidaIl Qf thepublic from ~ T O S U to ~ ~ radioactivity, both as the ~~nsequenee of 8 ~ -from lower level release durim nomal uperaticans or maintenance.NSBRs differ from solid-fuel reae ors, not only in that the fuelis in liquid form, but also in that the processing plant is at the reactorsite and an inte pal part of the plrequires Consideration from those Of Solid-fUel SyS tethe safety question has been to examine the basic cansieratfons that relate to safety an to attempt in the design to provide

53means for treating each of these adequately. En considering how thesefactors affect the concept, we have tried to make the design such thatit could be located on any site acceptable for uther kinds of reactors.The guidelines on permissible release of fission products hairechanged in recent times. Thts has required that we re-examine some ofthe approaches towards low-level releases that appeared acceptable earlier,and, in particular, we have had to devote more attention to means forkeeping the tritium release low- Safety and environmental considerationsare treated together in Chapter 14,..........-

54/...E..

55......;y.2&,.....:

.... ,... iii~.-. ..u..' YREfiCTORVESSEL- --._.___-- _-CONTROL rnDS - - - --mGRAPHITERE~LECTOR- __-.IGRkW#TEREFLECTOR-,a,.....

57.....:.w,....,r. >. .?.... s.:;v....

.....59...../.A i......... ......:+x,..:.:.:.:,.....ORNb- BWG 74- 9004:...... (..x;:*....::gp,....>a..... :.:

gable 3.1. Characteristics of a 1000-hI&V(e) MoItern-SaIt Breeder ReactorUseful heat generation. MW(thjNet electrical output of phnt, MWCc)Overall plant thermal efficiency. %Fuel salt inlet and outkt temperatures, ' FCoolant salt inkt md outlet temperatures,Throttle steam conditionsReactor vessel inside diameter and height, ftCore height. ftCore diameter. ftRadial blanket thickness, ftGraphite Eflector thicknCS§: ftNumber of core elementsSize of core eIernents, ftSait volume fraction in core. 'TSait volume fraction in reflector, 5Tutal weight ofgrxphite in reactor, kgMaximum salt velucity in core. fpsPressure drop through reactor due to flow, psiAverage core power density. \v/crn3~axirnum therms1 neutron flux. neutrons cmm2 sec-'Graptritc damage flux (>io ire^) at point ofmaximum damage, neutrons cm -' sec-'Estimated graphite life at maxirnuni darnagepoint, yeas"~ ~ tsalt c volume i ~ in primary system, Et3Thorium inventor). kgFissile fuel inventory of rrrictor system andprocessing piant, kgBreeding ratioFissile fuel yield, Z;iyearb~uei douhiing time iexponential). years*22501000441050.1300850, 11503500 psia, 1000°F22,2013141.52.514120.33 x 0.33 x 14.813 and 37

I .,Alternatives to the Reference DesignEbasco Variations from OWL Design.... .fila......:*4Ebasco Services and their associates in Reference 3 present theirconceptual design of a single-fluid molten-salt breeder reactor thatis expected to satisfy the same general criteria that were the basisfor the ORNL design. This study represents only the first phase ofa series of tasks that fiasco is to undertake for QmL. Its objectivewas to identify the gene~al features of the concept, whi& is beingexamined in more detail in s~bseqpae~t studies.For this initial examination, Ebasco was restricted in some aspectsto using conditions specified by OWL; for example, the composition ofthe salt and the salt volume fraction in the core were specified, andthe performance characteristics of the processing plant were definedfor Ebasco and were not to be studied. However, in their examinationof other aspects of the plant, Ebascs in some cases selected differentways of designing features or components of the reactor from those inthe ORNL reference design. Notable were the following:.....%!&'1.In order to avoid the difficulties of sealing the interior surfaces,Ebasso uses a slab geometry for the graphite core elements ratherthan the square element with a central hole used by QRNL.....ii,.,....,.....*:>; . ..;,> .....a1

62In the earlier discussion of the principles follmed in selecting6, refeh-ence desi we noted that if a thermal breeder alone isto satisfy the requirements of a growin nuelear econof&4p, the fuel inventorymu~sl: be kept r@assnabl.y IQW. If this requirement is altered becausethe molten-salt breeder is not the only breeder in use ar becausesf either a lower rate of nuclear power growth 5r a greater availabilityQf UrXL~UITl Ore, en the r@qUir wts fer low inventory will be eased.e csllld lead to a des with a %m

.,.,.,. ..A, .63.....ir

.....19-.

..Y....;..4ei24A. M. Perry.....c.z,......x.x.....r:......; .*.,.....i .i...&....i,..... :.x,*,.....:=&..........y .%A......=*Reactor physics considerations in the design of molten-salt reactorsare, for the most part, stmilar to those for other thermal reactors. Inparticular, the graphite-moderated MSR has much in cornon with the High-Temperature Gas-Cooled Reactor (HTGR): graphite moderator at an averagetemperature of 600-700"C Th-2 33U fuel cycle, and similar fuel-moderatorratios. Thus, much of the design technology of thermal reactors in generaland of HTGR's in particular is directly applicable to the FER. Thisincludes nuclear data and methods for calculating neutron fluxes, powerdistributions, effects of heterogeneities in COR structure, control rodworthis, temperature coefficients of reactivity, and so forth.There are, however, a few important differences. The need to establishaccurately the reactivity lifetime of the fuel, SO important to theeconomics of s~lid-fueh reactors, has no analog in the NSR, since fuelcan readily be added or rem~ved, as required, in order to maintain criticality.The problem of accurate calculation of the reactivity life ofthe fuel is thus replaced by the problem of designing, building, ad operatingdevices to mintain the desired composition of the fluid fuel.This is not the same as adjusting the feed rate for solid-fuel reactorswith COII~~~UQUS on-stream refueling, since in the latter case the dischargeexp0supt.e of the fuel remains a crucial economic parameter.Similarly, the problem of calculating power-density distributions inKSR cores takes a somewhat different form than in solid-fuel reactors.In particular, since the fuel is constantly mixed and its cornpositionremains essentially uniform thr~ughout the reactor, the problem of timedependentpower distributions is much less important in fluid-fuel thanin solid-fuel reactors.With respect to their dynamic behavior (i.e. the short-term timedependences€ operating variables such as power and neutron flux levels,fuel temperatures, ete.) the circulating fuel reactors - and especiallythose operating with a33U fuel - have unique characteristics requiringspecial study, and these will be discussed in a subsequent section of thischapter. We simply note for the moment that an EBR typically has a muchsmaller delayed neutron fraction than do other types of reactors and thetransport of the delayed neutron emitters in the moving fuel alters theresponse of the reactor to changes in reactivity. On the other hand, mostof the power generation is in the circulating fuel itself, which is alsothe primary reactor coolant; thus, no heat transfer lags between fuel andcoolant are present. As we shall see, these two factors together producesafe, stable operating characteristics over a wide range of core designparameters.%he central question in the physics of the Molten-Salt Breeder Reactorwould thus appear to be the breeding ratio itself. This is so becauseof the very small breeding gain (breeding ratio minus one) that is65

characteristic sf thema1 breeder reactors. mereas fast breeders typisa!'by achieve their disirable breedin perfomance With largains, in combination with relatively large fuel inventories, the thema%breeder, with a much smaller breeding gain, must succeed by maintaininga relatively Haw fuel inventory." Because the breeding gaim is lows theexpected perfo~~~~nce of a themal breeder reactor is especially sensitiveto uncertainties in the calculated breeding ratio, and it becomes wesessaryto establish the ne~tron ba8a11~e in the reactor, on which the breedingratio depends, with greater precision than would be required for ailli"fast breeder reactor 0Breeding in Molten-Salt ReactorsThe quality of perfomanee of a breeder reactor, in a nuclear powereeowomy postulated to follow a particular pattern of rQWfI1, is Oftenevaluated in terms of the cumulative amount of bllrani ore that wouldhave to be mined up to the time when the breeders themselves could satisfyall requirements for fissile materials. It is ene rd re CQ gnh e dthat this quantity depends both on the breeding rati of the reactor andon its specific fuel inventory.~n ~ ig. 4.1 (aiso staSta as pig. a,2>, we shops the anticipated orereqUireUlentS for a U.S. nUCl@aP power industry based On VaKi5US 2>oS%U%atedcombinations 5f present-day converters and possible future breeders, havingdifferent breeding miss ana specific fuel inventories (~cg fissiie/m~e)).me total installed nuclear-electric capacity is assumed to reach 140 W(e)in 1980 and 930 &'(e> in 2000, and %a increase thereafter at a rate ofLOO aJ(e) per year. It is further assumed that only light-water reactorsare bUiPt prior to 1982, that the first breeders come on line in 1982 andthat only breeders are added to the system after 1998, Also ahom onFig. 4.1 are the price ranges within which the usmc currently estimatesthat incremental amounts of uranium my be available from domestic U.S.ores including estslfgsted prob&,ke ~ ~ S O U ~ as C well ~ S 3s reasonably-assuredk.3reserves rA cross-plot of data taken from curves like those of Pig. 4.1 producesa plot of resouree requirements vs specific fuel inventory, fordifferent values of the doubling time, as shorn in Fig. 4.2 (also shornas Fig. 1.3) E% may be noted frr m these curves that a breeder reactorwith a (compound) doubling time 0 20 Yearsp ana a specific inventoryof 1.5 kg/MW(e), would be expected to require very little mre ore thana breeder with a doubling the of 10 years and a specific inventory of4 kg/Md(e). (These might be, respectively, a thermal breeder with aratio of 1.07 and a fast breeder with a breeding satis of 1.35.)kThe combined importance sf these two factors, breedingfuel inventory, w ill be explained in the next section.

&.&.-.......y,!*...-CUMULATIVE URANIUM REQUIREMENTS(millions of short tons bi30,)......x

4ORML- DWG 72- 589940 0 2 3 4 5 6 -7s P EC I F I c f MV EN TO WY k g / M w. (e)Pig. 4.2. Maximum uranium ore requirements. Basis: U.S. nuclearpower capability reaches 140 W(e) in 1980, 930 GW(e) in 2000, thereafterincreases 100 GW(e) per year; first breeders on line in 1982, onlybreeders built after 1998,

69Thus, a thermal breeder reactor with a low specific fuel inventorymay have breeding performance comparable to a fast breeder with a largerfuel inventory, even if the doubling time of the fast breeder is muchshorter than that of the thermal breeder* However, it is clear that thelow breeding gain of the thermal breeder makes its perfomance wore sensitiveto uncertainties in breeding ratio, and it bec~~~es necessary toexamine carefully the SQ~~C@S and probable magnitudes of such uncertainties.Uncertainties... ..... _.-.. .,........,'.:.La....,;- . i = 0.08.We regard this, however, as an extremely improbable combination of circastan~e~,a view that is strengthened by the satisfactory results ofcriticality calculat~ORs dESCl?ib'Zd bdOW.)

AbsorptionsFissions2XTh2”pa233u234~235u‘Mu237YPLi3-i%eI9FGrsphiteFission productsLeakage.rjeBreeding ratio0.99688.00450.92420.08190.07580.00740.00440.00320.01 410.00700.02030.05100.01500.0221c2.23171.07080.00300.81450.00040.06180.00448

.... ";,ỵ L71.....i-i:.=+........ . .:.w...?,c......2'.&A..;...e-....i.x:>>233u43SuQO 2.293-QTIQO -1Y 2.50QF 0.18-QoQ,iVO -1V 2.43% 0.50234u -ua236u -+0.01020.5%+O.OIsO.Ol020.5%+0.01+a02+IO%215%0.0080.0090.0030.0060.0010.001

72.’-Assigned0.0030.0020.0010.0010.0020.0030.0040.0030.0050.0030.0020.0020.0030.0040.0060.005

73,.;. .7....x.*~.;.;&..... !.d....$3cs......:.;

74u (meaured) 0.223 * 0.0039 0.2509 a 0.00380: (calculated) 0.1226 0.2500CKm/Q, €.OM x 0.032 1.054 f 0.015qStl = u(L Jr a)-'2.226 * 0.QW2.224 z 0.022''I$ 2.227 i.9441.943 * 0.006b1.943 i- O.0lOd

.&........ :am.:*a .......... :. .3..&....,:.:.:*:.-75carrier salt). Instead, the design has been optimized on the basis ofthe ratio of the breeding gain to the square of the (specific) fuel inventory,which is an approximate measure of the cumulative. uranium orerequirement for establishing a self-sufficient power econ~my based onbreeders aThe inventory is nonetheless subject to some uncertainty, to theextent that the volume of fuel salt in the system depends on details ofthe mechanical and thermal-hydraulic design of the system, including heatexchangers, piping, plena, etc. It way be noted that an increase of 200ft3 in fuel-salt vokume, without any reoptimization, wou~d produce a fractionalchange in the ratio G/12 of about the same size as the (fractional)uncertainty in G(= BR - 1) that arises from nuclear data and reactorphysics uncertainties.There are additional potential sources of error in the breeding ratiothat lie outside the area of reactor physics. These include the chemicalbehavior of certain fission pr~ducts, notably the %oble" metds, a, NO,Ru, Te; a part of these fission products is known to be deposited andheld on primary-circuit surfaces, including core graphite. The estimatedbreeding ratio for OUT reference MSBR design includes an allowance [6]for dep~sition of these nuclides on graphite to an extent Qi.e., 10% ofthe mounts produced) that appears reasonable in view of data obtainedfrom samples from the MSRE and from an in-pile loop.* The effect ofthese nuclides on the breeding ratio is shown in Pig. 4.3 as a functionof the in-core life of the graphite. For a 4-year graphite life, theaverage poisoning effect is about 0.004. Since the behavior of thesefission products is not fully ~nderst~od (see Chap. 5) we cannot excludethe possibility that a larger fraction of these fis~ion-produ~t poisonsmight be retained by the core graphite.Other factors that could have an adverse affect on breeding perfornancewould be failure to achieve adequate sealing uf graphite surfacesagainst adsorption sf I35~e or adequate stripping of xenon from the circulatingsalt, failure of the chemical plant to remove fission productsfrom the salt, or non-recoverable losses of fissile material due to upsetsin reactor or chemical plant operation. The first of these WQUICI at mostreduce the breeding ratio by 0.815, while the last two would cause temporaryreductions in breeding ratio pending restoration of normal plantoperations.According to the curves of Fig. 4.2, the cumulative resource requirementsare less sensitive to errors OK uncertainties in doubling time forlow-speci%is-inventory breeders than for hfgk-specigic-inventory breeders.This is partly due to the irreducible ore requirements far converter reactQrS,assumed t~ be built during-the early years of the postulated5;Chapter 5, this report..-,..... i . ,

0,0940.0720.0100.0.00820 2 4e Change in breeding ratio due to noble-metal fissionproducts in MSBR.

77growth pattern, as well as to the smaller absolute mounts of imventoryrequired for the low-inventory breeders. For example, a system predicatedon molten-salt reactors with a specific inventory of 1.5 kg/ml(e) and adoubling time of 35 years would require perhaps 2.5 x lo6 short tons ofU308, while a system using breeders with the same inventory but a ZQ-yeardoubling tine would require about 2 x 106 short tons of u ~ Q ~ * ~hus, * areduction of breeding ratio from 1.09 to 1.04, should it occur for anyof the reasons outlined ab~ve, would not markedly impair the ability ofthe molten-salt reactor to limit the m~unt of ore necessary to achievea self-sufficient nuclear power economy.Reactor StaticsMultiplication Factor".;* _.2-......... . .....&

7Concentmtron (gjliterP' 35 U loading' * U loadin$Predicted concentration 32.8 15.1 r &leObsened concentratieid 32.8 i 0.3 15.4 t 0.1Qbsewed/predict& 1.00 i 0.01 1.02 f 0.01"Rractor isothernial at -E bOO" F, fuel not circulating.b~issiie uranium, grams per liter of salt.rsce Ref. 9.%e, Ref. io.'Uncertainties in adjustments for residual plutonium andfission products from L K U and ~ for dimensional changes ingraphite cure structure due to fast-neutron irradiation,fLJncertainty due mainly to uncertainties in salt density andsdt volume.g(M/k) (akialrf) = 0.36; I% &l.E- 0.4% irk.....

,.... .....~79..... ,&,

k,20°C 300°C 627°C 1000' cAk, (200" -+ l0OO"C~k, (measured) 1.0291 f 0.0012 1.0124 f 0.0018 1.0045 f 0.0010 1.0037 * 0.0022 -0.0090 k 0.0016k, (calculated) 1.0300 1.0138 1.0057 B.0030 -0.0108measured/

8%Reactivity Control...,.. >.&Requirements for reactivfty control in molten-salt reactors are somewhatdifferent than in most solid-fuel reactors. hng-tem reactivitychanges are to be compensated by adjustments in fuel salt composition.Shutdown rods worth a few percent: in Clk are likely to be required, butthese wodd nOrmally be full. withdP8Wn for the re8Ctole. SmLlhl ZlMoUlltSof reactivity adjustment for normal operational maneuvering should beavailable, and if possible these control devices should RQ% have an adverseeffect on the breeding ratio. We therefore visualize an adequatecomp%ement of neutron-absorbing safety rods which WQU%~ be fully withdrawnduring normal operations. Maneuvering, as now planned, will be aecomplishedby graphite rods which displace fuel salt from special passagesin the core.Techniques for calculating the effectiveness sf the absorber rods arewell established, and generaPPg. reliable to within less than 10% of therod worth, which is quite sufficient. Predicted and measured rod ~~prthsin the b%sRE agreed to witkin 5% of the rod worth, as may be seen fromTable 4.7.The graphite displacement rods, to be used for maneuvering, are unconventional.Their reactivity worth is expected to be small (e.g.,6k/k PI, O.OQ% per rod) and they depend on somewhat different physicaleffects than d~ normal "black" absorber rods Calculations show thatdisplacement sf fuel salt by the graphite rods increases reactivity,rather than decreasing it as one might expect, and it appears that changesin resonance ne~tr~n capture in neighboring fuel passages are at leastpartly responsible.* Some of the reactivity specimens in the HTLTR experiments161 gave qualitative confi~~~ati~n of the positive reactivityassociated with displacing fuel by graphite but calculations preciselycorresponding to the experimental situ~iti~ns have not yet been made.The reactivity requirement for maneuvering is not a rigid ones but itwill be necessary to establish more firmly the effectiveness of the graphiterods before they could be adopted finally as the basis for MSBR design.Reactivity CoefficientsA large single-fluid MSBR has temperature coefficients of reactivity(see Chap. 14, Table 14.1) which are nut especially large for a fluid-fuelreactor. Both the overall, isothermal temperature coefficient, and thecoefficient of the salt alone represent small (algebraic) sums of muchlarger individual effects, e.g. thorium Doppler coefficient (negative)and moderator coefficient (positive). Direct experimental confimatisn* A similar effect, with a similar explanation, is noted in naturaluraniumgraphite reactors, where removal of all fuel from a single channelincreases reactivity. In this case, good agreement is found betweenexperiments and calculations e

2Uranium concentration23SUbConfigurationReactivity (% 6k/k)Observed=RedictedInitial crrtrcal concentration 1 Pod 2.26 2.283 rods 5.59 5.891.1 Y initial concentration I rod 2.08 2.09a33Ur1 rod 2.58 2.753 rods 6.9 4.01

....&aa........of the calculated magnitudes of all these effects has not been obtainedfor a molten-salt reactor, though all the important effects will be inferredfrom the-HTLTR-MSBR lattice experiments. The criticality calculationscited in Table 4.6 show satisfactory agreement between measuredand calculated changes in k, over the temperature range between 28" and1000"~. ~itho~gh effects of salt expm~ion are not directly reproducedin these experiments, the major components of the overall temperaturecoefficient (Doppler and thermal-base coefficients) are included. Fueldensity effects were tested in one of the reactivity samples, but analysisof these measurements is not yet complete.Calculated and measured temperature coefficients for the b%sRE [9]showed agreement to within about lo%, brat these depended mainly on newtron leakage effects; although the agreement is encouraging, the resultsare not directly applicable to a large ER.The reactivity effects of fuel additions were measured in PEE forboth 245Y and 233U fuel loadings, with results shown in Table 4.8.Radiation CalculationsDesign prsb%ems related to radiation transport, such as shielding,radiation damage, and gamma or neutron heating, present a somewhat differentaspect in molten-salt reactors than in many other reactor types, butthis is because of the mobility of the fuel and other radiation sources(e.g., fission gases), and not because of any essential differencesin the transport problem par seeIn gas-cooled reactors such as the natural-uranium, graphite reactors,radiation streaming from fuel channels was peculiarly important.In EPFBR's, neutron penetration through thick composite regions of ironand sodium has proved to be an important problem requiring new experimentaldata and improved methods of analysis. In the case of the moltensalt reactor, while it may be premature to conclude that no unique problemsof radiation transport will appears none are now evident. Instead,it seems apparent that the principal problems will relate to determiningthe distributhns of the sources of radiation, and especially of fissionproducts, throughout the reactor plant. These aspects of the problemare treated elsewhere in this report.Keactor DynamicsQuestfons of control and safety of molten-salt reactors are discussedelsewhere in this report (Chapters 10 and l4)* Et is our purpose hereonly to clarify the status of our understanding of the relevant neutroniccharacteristics of the reactor.Cireulat2ng-fuel reactors especially if operated on the TII-~ 3 3 ~fuel cycle, have unusually small delayed-neutron fractions. Calculatedvalues for molten-salt breeder reactors fall in the range 0.00lQ-0.60l5,depending on details of reactor design. At the same time, the graphitemoderatedER, in common with most other graphite-moderated reactors, hasa very long prompt-neutron generation time, e.g., 0.4 msec.

4Table 4.8. Messwed and calculated reactivity coefficientsof fuel concentration forEMd0.2230.3690.2440.389...

I. .,85With respect to temperature coefficients of reactivity the situationis qualitatively similar to that of the natural-uranium, graphitemoderated,gas-cooled reactors, i.e, a prompt negative temperaturecoefficient for the fuel is accompanied by a delayed positive coefficientfor the moderator. In the case of the reference MSBR, these coefficientsnearly offset each other in magnitude. This gives rise to a small S~QWoverall coefficient, in contrast to the much larger negative prompt coefficient.We have not found it very useful to try to formulate simple generalizedstatements about the dynamic behavior of EX'S in terns of theseunderlying characteristics. Instead, we have tried to develop reliablecomputational models for predicting their behavior, and have gained confidencein the use of these models by comparisen with observed behaviorin the case of the MSRE....._.i......;.;.:2 ...-....Effects of Fuel, CirculationModels for the reactivity effect of fuel circul~~tio~~ must take intoaccount the transport of delayed neutron pre~ursors inn the moving fueland the weighted contributions Of delayed neutrons emitted outside thecore, as well as the skewed distribution of delayed neutrons emittedwithin the core. The usual reactor kinetics equations must be modifiedto take these effects explicitly into account, since their importance isnot independent of reactor period. Confirmation of the models developedfor the MSRE was obtained during the control rod calibration experiments.The reactivity effect of fuel circulation (with 235U fuel) was measuredto be 0.212 C 0.604% 6k/k, and the calculated value was 0.222% 191.(Neglect of delayed neutrons emitted in the upper and lower plena yieldsa calculated value of 0.30% &/k.) In addition, good agreement was obtainedbetweeA rod calibration CUKV~S inferred from period measurementsusing the conventional inhour equation with the fuel stationary, and themodified equations with the fuel circulating. The reactivity effect sffuel circulation is thus believed to be well understood.Frequency Response and Reactor Stability.... ....,.d

Fuel-Cycle Economics

.Y...,&&i87....i.,...;.v3..,1035ORN h- BWG 7 8- 52 19........,:-.....:.:.:410-3 2.......A1.,9060..... .:.a.........% !.e!:a--e,- 830wmoQ:Ia- 38-60-90f6-2 2Fig. 4.4.5 lo-' 2 5 100 2 5FREQUENCY (rodrans/seelMSaE f requermey respsnse-PrnS signal.IO'

88Table 4.9. MSBR fuel eyeb costKilIs!kWhr( e)Fixed charges on carrier salt at 13.2%year7LLFBeF*ThF4Salt makeup (15 calendar yea cycle)7~iFBeF2ThF4Fixed charges on fissile inventory at I3.2%/yearFixed charges OR processirrng plant at 13.4%/yrarRoccssir.g pIant operating costsGross fuel cycle costFuel productions creditNet fuel cycle cost,0.0580.0080.033-0.0790.0190.0040.0160.039Q.36ff0.49 - 0.690.051.02-1.22-0.090.93-1.13

89....*:::b......

Be r y P I iumUraniumFuel cycle costs for the reference MS w istfiuencea by the priceof fissile uranium primarily through the inventory charges, which are0.36 dbla/kkhCe), as compared with a fuel production credit of 0.09rnills/krn(e) (for 8 breeding ratio of 1.07). Doubling the price 0% fiasilebaramaim, without compensating changes in the fuel cycle, would thusadd 0.27 ~LPls/k~(e). (Doubling the cost $8 to $16/Pb Uq08would add 0,09 wills/kbh(e).) Some adjustments in fuel salt c~mp~siti~n,as well as in details ~f core design, would of course be possible. Etis not to $e apected, baowever, that appreciable reauctions in fissileinventory wouhd be possible without a mjor reduction in breeding gain.TdiliLe the trade-~f%s between specific inventory and breeding gain dependon many details of reactor design, a sough approximation is that a 10%reductisn in inventory fwitfisut: altering the design Of the primnary circuitatePnaP t0 the reactor CQ%e Id cost about 0.015 in breedingratio. Applied to the reference this indicates that a saving of0,636 in inventory charges would be partly offset by a reduetiom of8,021 in the fuel production credit. Actually, the relation betweenratio and specific inventory is hi laPy non-1inear: a majordecrease in inventory (e.g. ~ Q reduetion) X~ ~ ube i a pOssit~e, at theexpense of reducing the eonversisw ratio nearly t~ zero; on the otherEaand, a large percentage increase in inventory could gain at most a veryfew percentage points in breedin ratio. Thus, the possible range foradjustwents is rather limited whether breeciimg performance or fuel costalone is the guide,k.,k.. .,Graphit eWep%aeement of the graphite core structure, as a consequence of radiationdamage, is estimated to contribute 0.17 milPs/kWh(e) t~ the powerCost for the reference NSBW. A partial breakdown of this cost is shownin Table 4.10, (Inclusion sf this cost component with the fuel cycle costis of C Q U ~ S arbitrary, ~but conventional, in keeping with the inclusionof Ee or B2Q mkeup costs for gas-cooled and heavqT-~ater-moderated reactors- )Graphite alone contributes &QU~ 6.12 miIs/kk%(e) to the cost ofCore P@p%aCelE%lt (at foulf-year intervaks). bkKh Of the Ullc@Ktahty ingraphite cost is associated with coating or impregnating the graphite toreduce its pemeability to xenon. In our cost estimates9 this representsabout $5/Eb sf an average $l1/Hb cost of the graphite that is replaced,The uncertainty in graphite eoatim costs is substantial - probably a

....,x.w91.....!.%...i..........i.,,. .....:*jTable 4.10. MSBR core replacement cast4 calendar year replacement schedule; replacement assumed totake place during planned shutdown for maintenance ofturbines and other equipment [ 151Miilions of dollarslilills/kWhr(e)Hastelloy N 1.09 0.035Graphite 3.75 0.119- ~Special labor cast 0.50 0.0165.34 0.17.:.:.*.....

92,...~. ....

930.200.1 8+ORN b-QWG 92- 8579. ~ .0.1 60.1 48.1 20.1 00.08,;.;.;q..>8.06$ Il/lb..' ...;,.,.....;.;.;, . 90.040.02T--I-...,y.*;00 4 8 12 16 20 24 28GRAPHITE LIFE [years).....:.:.:.xFig. 4.5. Graphite replacement cost as a fun~tion of graphiteprice and replacement life.....

94References for Chapter 42. Y&R PPogrm Semiann. PPogE?. Wept. Feb. 28, 1971, ow-4676, p a 44.5. A. M. Perry, H. P. Bauman, €luoZ. AppZ. Tech. - 8, 214 (1970).. .. ...e&. .,.6, A. E%. Perry, E. I?. Bauman, PJtscZ. App2. Tech. s8, 215 (1978).11. T. W. Kerlin, S. J. Ball, R. C. Steffy, 'Theoretical Dynamics AnalysisOf the EISRE,"'NZh@Z. Tech. - 10, 118 (n971).12. T. W. KerIin9 S. by. Ball, R. C. Steffy, PI. R. Buckner, "Experfencesnamis Testing Methods at the MSrn,8' NikcZ. Tech. - 10, 103 (1971) 015. Conceptual Design Stud3 ef a SingZe-PZaid Molten-Salt Breedela Reactor,8mL-4541, p. 188 (1971) 0

....4i.....Y....i.~.. ....... ..,i iW. R. Grimes, E. G. Bohlwann, A. S. Neyer, and J. M. Dale...,........ i .,....,X.

96high concentrations of thorim at temperatures comfortably below the ternperatrareof the primary heat exchanger. The mixture must be thermallystable, and its vapor pressure needs to be low over a temperature range(600-75Q"C) sufficient8y high to permit generation of high-quality steam.The fuel mixture must possess heat transfer and hydrodynanic propertiesadequate for its service as a heat-exchange fluid. It must be ~~naggressivetoward some material of construction and toward some suitable moderatormteria9.. The fuel must be stable toward reactor radiation, must beable to survive fission of the uranium (or plut~niumt), and must toleratefission gr~duct accumulation without serious deterioration of its usefulproperties. An additional demand is that the fuel be menable to effectiverecovery of bred fissile material and to removal of fission productpoisons9 as discussed in Chapter 11.The reqUaKeIIIent that the mBB fuel CoRsiS%, except for it8 fissileand fertile material, entirely of nuclides of very low neutron captureCTB88 Section restricts the Choice Of mteria1.S to CORIpOundS Of beryl1iuID9bismuth, boron-ll, carbon, deuterium, fluorine, lithium-7, nitrogen-15,oxygen, and the fissile and fertile materials. Other nuclides can be toleratedonly as minor ~~nstituents. Host of the compounds of the potential81rnajor constituents" are elbinated by the other fuel requirements. Nodeuterium-bearing csmpounds are practical in such melts. Carbon, nitrogen,boron, and oxygen form high-melting, and quite unsuitable, binary compoundswith the fissile and fertile metals. The oxy gnated anions either lack therequired thermal stability (e.g., NQg- NQZ-1 or fail as solve~~ts forhigh COIlCeRtPatiOn§ of th0riLml ColRpOUndS (e,g. 6032'). It quickly de-~ e l ~ theref~re, p ~ , that fluorides are the only suitable salts.UrilniUlll tetPaflUOride @P4> PS the Qnly flUOPide Sf UraniUH1 thatappearsas a constituent of mo~ten moriae fuels; m4 is reiativelystable, nonvolatile, and nearly nonkygr~~~opic. In the pure stateuranium txifluoride &UP 3) disproportisnates at temperatures above about1BBQ"C by the reastion....&.

........ .._.,.:.:.:.:....:,5;2...._,.y,.. .' heZrFb was added, as dissus%edl below, to preclude inadvertent precipitationSf uo2.Fuel selecti~n is no great problem for the MSBR since, as will bedetailed below, the phase behavior of LIF and BeF2 with UF4 and ThFhis entirelg s%tkfaCtO9p. The requiPE%l,ent Of Very IOW CTOSS Sectionrequpires, in principle, that the fuel mixture be free from extraneoushigh-cross-section ions. Purity requirements dictated by oxide-fluorideequilibria and of compatibility - both of which are described in somedetail below - are rather more stringent than those posed by cross-sectionconsiderations.Experience with Molten Salt Fuels....x.:.:,......:.:

98ny chemical studies which immediately preceded operation ofwere directed at LiP-BeFz-based fuels, and much was learned thatars directly releQant to the El3R system and that is described insome detail in subsequent sections of this chapter.Operation of the HSRE, with its fuel mixture of %iF-BeP2-%rP2-UP4,provided much chemical information that was reassu~ing 191- Salt samplesremoved routinely from the fuel and c~olant circuits (one to three perweek froen the fuel system) were analyzes for uranium, mjor fuel constituents,possible corrosion prsducts, and (less frequently) for oxide ioncont minat ion 0khaagrSeS far UlPaELFW by cou~OmetPiC titration 1191 Showed good re--producibility and high precision (0.5I) but on-line reactivity balancecalculations were about 10-fold more sensitive than this in establishingchanges in uranium ~oncentrations within the circuit. A11 the data suggeststrongly that the fuel was completely stable and that losses ofuranium, if any, were extremely sma3.P. Betemination of uranium OW orin a graphite moderator bar led to the ~oncl~sion that the entire stackcontained less than 10 grams of uranium, a quite negligible mount [20].Oxide concentration in the radioactive MSKE salt was determined bycareful evaluation of B20 produced upon treatment of the salt sampleswith an~~ydrous PEP. AH samples examined showed less than IBB ppm of G-;no perceptible increase with time was apparent [21]. This is moderatelyreassuring insofar as the practicality of maintaining oxfde contaminationat very low levels in fut~re reactor systems is concerned, though bettermethods of analysis for oxides are clearly needed.KS%%E maintenance operations involved flushing the interior of thefuel circuit with a S~i~-~e~2(66.0-34.0 mole X> mixture. ~~-~a~grsis ofthis salt before and after each use showed that an amount of uranium wasadded to the flush salt in each flushing operation equivalent to 23 kgof fuel-salt residue (about 0.5% of the charge) from the reactor circuit.The magnitude and the reproducibility sf this figure seem to confirm thenarkwetting characteristics of the clean fuel salt towardmetal ane fission process is niiaiy oxidizing toward dissolved ~ 3 f in thefuel. In the MSRE a convenient means for restorin the b'3+ concentrationwas to suspend beryllium rods in a perforated capsule of nickel in thesalt in the pmp bowl. This active metal r~i~iicted with UF4....

99x.d......:s*

Although operation of MSRE generally verified the behavior predictedfor the fuel salt, not all the news was good. Fission product behavior(to be described subsequently) was even more complex than anticipated.Our methods for sampling the M%RE salt were restrictive, and some ofOUK methods for analysis needed ked improvement. Behavior of tritium(see Chapter 14) portended a problem in large molten-salt reactors.Finally, in spite of the excellent picture on generalized COKIPOS~O~,the grain bsrandaq attack resultin in superficial cracking of the HastelloyN exposed to the fuel was a major and disappointing observation(see Chapter 7).Present Status of Fuel ChemistryPhase Behavior among Fluorides. - Phase equilibria among the pertinentBEXiR fluorides have been studied in detail, and the equilibrium diagram,though relatively complex, are well understood.The binary system LiF-BeF2 has melting points belm 560°C over theconcentration range from 31 to 68 mole % BeF2 [24, 25, 261. The phasediagram, presented in Fig. 5.1, is characterized by a single congruently~elting compound, 2LiPeBeP2, and a single eutectic between BeP2 and2LiFOBaFz.The BeFp-UFb 624,251 and B~E'~-%IRI?~ [a71 systems are very similarin phase behavior. Both systems show simple single eutectics containingvery small Concentrations of the heavy metal fluoride. ThP4 and UFk areisostructural; their binary phase diagram shows a conti~ous series ofsolid solutions with neither maximu111 nor minimum.The binary diagr for LiF-UFk [B] amd LIF-ThFb [%9] are relativelySimilar. The LiF-UF4 system ShWS three collapoUfldS (aton@ Elre Congruentlymelting) and a single eutectic, at 27 mole X Up4, melting at 490°C. TheLiF-ThP4 system contains four binary eomp~und~., one of which (3LiFeThF~)melts congruently, with mo eutectics, at 570'C and 22 mole 2 ThF4 and at560°C and 29 mole X T~I%PL+.The ternary system LiF-ThF4-UP4 [30], shown in Fig. 5.2, shows noternary compounds and a single eutectic freezing at 488°C with 1.5 mole 2TkP4 and 26.5 mole % UF4. Liquidus temperatures decrease generally tothe LiF-UF4 edge of the diagrm.Because the NSBR fuel needs a concentration of ThF4 much higher thanthat of UF4, its phase behavior is dictated by that ~f the LPF-BeF2-ThF4system. Figure 5.3 gives the ternary system LiF-BeF2-ThP4; this systemshows a single ternary eutectic at 47 mole X LIF and 1.5 mole X ThF4,melting at 360"~ 624, 291. The system is complicated by the fact thatthe compound ~ L I F ~ T can ~ P incorporate ~~e2* ions gn both interstitial andsubstitutional sites to form solid solutions whose compssitional extremesare represented by the shaded triangular region near that compound. Inspectionof the diagram reveals that a con~iderable range Sf compssitionswith more than 10 mole 2 ~ h will ~ 4 be completely molten at or below 500°C.The maximum ThF4 concentration available at this liquidus tekperatureis just above 14 mole 2. As expected from the-general similarity of ThF4and - and especially from the substitutional behavior shown by the.....ww..

........:

16 2Pig. 5.2.The system LiF-fiF4-UP4.

..... .:.:

10 4LiF-UF4-ThF4 system (Fig. 5.2) - substitution of a small quantity of UFc,for ThF4 scarcely changes the phase behavior. Accordingly, and to a verygood approximation, Pig. 5.3 represents the behavior of LiF-BeP2-ThPt+-UFqmixtures in which the mole fraction of ThP4 is much greater than that ofUP4 *Effect of Qxide. - Phase behavior of the pure fluoride system LIP-BeF~-ThP4=-UP4, as indicated above, is such that a wide choice of adequatefuel mi~tu~-es assured. The behavior of system such as this, however,is markedly affected by appreciable concentrations of oxide ion,which might be produced by inadvertent contamination of the fuel system.wken a melt containing only LS, BeF2, and UF4 is treated with areactive oxide (such as B20) 9 precipitation of UB2 ,-Jo results gs, 311 0If the melt contains, in addition, considerably more ZrF4 than UP4sinadvertent oxide contamination yields monoclinic Zr02 containing about250 ppm of UQ;1 9321. Precipitation of cubic U02 (containing a smallconcentration of Zr02) begins only after presipitation of Zr02 has droppedthe ZrP4 concentration to near that of the UF4.T R ~ effect of added oxide on the NSRR fuel mixture, with its containedThF4, UF4, PaF4, and perhaps PuF3, in addition to LiF and BeF2,has been carefully examined in a series of recent studies [%3,34,3%,36,37,%8,39,40]. The findings of these studies are s ~~~~na~ized in thefollowing :The solubilities of the actinide dioxides in EBBR fuel salt are low,and they decrease in the order ThQ2, PaQ2, WO2, PuO~. The solubilityproductsQ-02 = A&+ xi2-....C.%....a,:.are presently estimated as follows:,... Y.(kO.8)Since all these oxides have the same fluorite structure and nearly thesame lattice ~~x-EII~I~~~Ks, they can fom solid solutions with owe another..... t,&....

.....:.:

0 WRL -BWG 72- 8 3265007005r"' -.,20058.....y,.&.20I*.,104 .o 4.4 4.2Fig. 5.4.Oxide tolerance sf MSBR fuel.

~ .,.=,....107...i.:.*for which we estimate the equilibrium quotient....."C.?.. i.....>io.."+...A .....

Table 5.1. Cornpsition and properties of MSREand WBR fuelsMSWE fuelMSBW fuelComposition, mole '% LLF 65 EIF 71.4BeF* 29.1 BeFa 16ZrF4 5 ThF4 12UE4 0.9 GF4 0.3LKpidul."C 434 500"F 813 932ROFItESdt 6QO"C (1112°F)Densty, 'g/;'cm3 2.24 3.35Heat capacity, c&(g-'''C~or Beu/(Ib-" F) 0.44 0.324Viscosity cen ti poises 9 12Vapor plZSSUIe, toFES 4 . 1

.....,:.:.a109....

,......... ,.%>Table 5.2. Standard free energies sf f ~ m t b ~ ~for species in molten 2LiF.BeP.p773- 1000" KMaterialQ=The standard state for tiF and BeFa is the molten2LiFeBeFz liquid. That fer MloFs(& is the gas at oneatmosphere. That for all species labeled (d) is that hypotheticalsolution with the solute at unit mole fraction and with eheactivity coefficient it would have at hfiite dilution.,>)11_..ii.

BPIIt is also clear from Table 5.2 that, of the structural metal fluoridesshown, CrP2 is the most stable. Accordingly, Cr should be seBectivelyattacked in alloys such as Htstelloy N by any extraneous oxidantsin the system. Impurities in the pelt should reactas sho :Ed oxidized films on the metalfollowed by reaction of the NIP2 with Cr. Reactions such as these willproceed essentially to completion at reactor temperatures; they canlead to rapid initial corrosion but not to a sustained attack.If the fuel is pure and the metal clean, UP4 is the strongest oxidantin the MSBR fuel system. The reactionhas, from the data of Table 5.2, at 900°K an equilfbrium constant~f the MSBR fuel with nUFLI = 3 x 10-3 and with no cr2+ or U F ~ presentinitially were permitted to equilibrate at 900°K with a Hastelley N surfacewith aCr = 6.05, the equilibrium solution would contain slightly morethan 10 ppm of ~ r + ~ 1vUF3 = about 2.5 x IO-', ~n principle, therefore,a mixture in which the UP3 mole fraction could be maintained at about3 X IOm5 should corrode the metal very little. Indeed, it seem likelythat corrosion could be kept Go quite tolerable limits in this way evenif alloys (such as ~nsone~) with considerably ~.ai&er chromium concentrations(and correspondingly higher values for a ) were used.Cr

112The corro~ion by UFq cannot, in principle, be completely eliminated,since the UP4-Cr reaction has a small temperature coefficient. Consequently,circulation of the salt through a temperature gradient tends toremove Cr from the hottest surface and to enrich the coldest alloy inthis element. The rate of such a reaction is controlled by the rate atwhich Cr can diffuse from the bulk alloy to the surface or (more likely)the rate at which the Cr can diffuse from the surface into the alloy inthe coldeb- region. Experience indicates that no real difficdty is to beexpected from this reaction.Modified Hastellsy N will contain titanium, while that in MSRE didnot. Estimates Ill] of their free energies of formation suggest that thefluorides of titanium are slightly mare stable than CrP2. Titaniumshould, therefore, be expected to react appreciably with the UP4. SinceTi diffuses less readily than does CK, however, it would not appear thatsuck ccsrroaion would be particularly deleterious.Graphite does not react with, and is not wetted by, molten fluoridemixtures of the type to be used in the EBR. Available thermodynamic dataIll] suggest that the most likely reaction:b........should Come to @gEuilibri~ at CFL, pressures below atm. This Consideration,taken with the wealth of favorable experience, suggests that noprUbleUX5 are likely fKoHl this SOLXrCe.It should be noted that at least one source e481 lists chromimcarbide (Cf~362) as st&l@ at MSBR temperatures. If 80, it should bepossible to transfer C'hPO1I1iUltl, at the Kate it Could diffuse from thebulk alloy to react with the salt, to the graphite. NO evidence ofsuch behavior has been observed with Hastellop M in MSRE or other experimentalassemblies. Although it may be possible with alloys of higherchromim content, it should not prove greatly deleterious, sirfce itsrate would be controlled by the rates at. which chromium could diffuseto the alloy surface and should be limited by a film of Cr3C2 farmed ORthe graphite eIt: must be emphasized that n~ne of the above throws any light uponthe special grain boundary attack upoat Hastelloy N in the MSIU fuel circuit.It is conceivable that some heretofore untested combination ofoxidizing regime, radiation, and fuel interaction was responsible, but itseems much more probable that some fission product (likely tellurirare) wasresponsible. This matter is discussed in detail in Chapter 7.. ..e.:;Interaction of EBW Fuel with Extraneous Materials. - n e complexmixture comprising the NBR fuel reacts readily, though not violently 87~even energetically, with water vapor to produce oxides of the actinideelements and HF vapor. (In fact, water vapor is a possible reagent in aselective precipitation scheme for fuel processing described in Chap. 11.1Rapid addition would certainly prod~ce a mixture of oxide products whoseequilibration would be relatively slaw. This admixture of water and fuel'salt would lead to appreciable corrosion due to the HF so produced, butwould hardly prove catastrophic.

....,.:.=113None sf the MSBR fuel constituents can release P2 upon reaction withoxygen (or nitrogen). Reactions suck as....I.:.:.:.;have been postulated [49], but such reactions seem most unlikely in dilutes~Pution in MSBR fuel. No reaction of Lip, BeP2, or Tub With 02 is OSsibleaAn inleakage of air into the MSBR fuel circuit cannot, therefore,cause energetic or violent reactions. However, since the WBR metal surfaceswill be oxide-free (because of the fluxing action of the fluorides),rapid reaction of the air with the metal circuit will occur. Such reactionsasfor example, will imediately result inand rapid corrosion. It will, aecordinghy, be necessary to minimize ingressof air both during operation and during maintenance. This is ofespecial importance because, as indicated earlier, the oxide toleranceof the KSBR fuel is low (not much above 30 ppm of 02-1 eThe consequences of mixing MSBR fuel with the secondary coolant, aswould occur as a result of a leak in the primary heat exchanger, are discussedunder Coolant Chemistry in a subsequent section of this chapter........>. .

in large (ca. 300 lb) batches, while the enriched uranium fluoride waspurified separately (as the LiF-UPk eutectic) e It seems likely that agenerally similar proqedure would be desirable got- BfSBR..The most difficult specification for MSBR will probably be that forvergr oxide concentration. ~xperienee with the process for high T ~ F ~concentrations is limited. However, all the preparations to date andmny kattoraeory-scale researches have success fully droppea the concentrationbe lo^ that required to precipitate WQ2-Th82, SktouPd thisprove very difficult on a large scale, it would be possible ts processthe EiF-BeF2-ThF4 melt with fl~orine; it seems possible that the rep~~eessingfluorination (see Chap. 11) might be used If necessary as thefinal polishing step in this treatment.Radiation Stability. - An early concern was the possibility thatradiation (includin recoiling fission fra ents) from the fission processmight lead to adiolytic instability the fluoride. a@-combination in mobten salt sf '*dissociated" species Ci.e. $ a fluorineatom and an deetr~n) should be very rapid. Nevertheless, it seemed(and Still SeeElS> likely that thehe exists a pOWeK level sufficient todamage a molten fl~~~ide by dissociation into metal and fluorine.irradiation tests W@XT COndUCt&?d prior to 1959 W i t h NaF-ZrF4-'LIF4 ~i~t~rres in %nconeab at temperatures at or above I5BQ"F [6,7]and quite high fission power densities, from 80 to lOQO watts/cm3 offuel. Eo instability of the fuel system was apparent, and the corrosiorndid exceed the COnsidePELbabe 2SiWuHat expected froIl3 baboratOry-SCdetests 0The ARE tests had not included graphite, and several irradiationtests were performed in the early days sf the NfsbFR program [53,54,55,561 primarily to test wetting of graphite under irradiation, These testsused mixtures of LiF, BePzP and ZrF4 with ll mole 2 ThP4 and 1.5 mole XUF4 in sealed Bastelboy N capsrx es, irradiated at power levels above 200watts/cm3 fuel to burnups as hi as 8% of the 235U. Examination ofthese capsules after storage at ient temperatures for many weeks revealedappreciable quantities of 6%"4 and, in most cases, considerablequantities of fluorine in the cover gas [57,58]. Careful examinationstrongly suggested that the P2 generation had not ~cc~rred at the hightemperature, but by radiolysis of the ~~~xture in the solid state.This suggestion was confirmed by irradiation in EflR of two arrays ofMastelloy bJ capsules, all containing graphite and EiF-BeF2-ZrP4-UF4 mixtures.Two 0% the capsules in each array had gas imlet and exft linesto PeTXlit SaDIpling Sf the COVB% as as desired. as samples dram fromthe test capsules at operating temperatures and at various p ~~er levelsup to 80 wattslcm3 showed no ~2 (t~aough an occasional sample from thefirst array showed detectable traces sf CF4). During reactor shutdowns,however, with the capsules at about 35'C, pressure rises were observed(usually after an industion period sf a few hours), and F2 was evolved.In the second array the capsules were kept hot during reactor shutdown aswell as during operation; no evidence of F2 or GF4 was observed. Such F2generation at ambient temperatures was subsequently follo~ed for severalL hot cells. The generation diminished with time in a, manglerk .p .... ->

115,:.y&.corresponding closely with decay of fission product activity; Fz evolutiomat 35°C corresponded to about 0.02 molecule per 100 ev absorbed, could becompletely stopped by heating to 100°C or above, and could be markedlyreduced by chilling to -70°C. The F2 evolution resumed, usually after afew hours, when teMperatUre was returned to 35-50'~.These and subsequent experiences, including operation of WW,strongly indicate that radiolysis of the molten fuel at reasonable powerdensities is not a problem. Radiolytic fluorine must be dealt with, k~wever,if irradiated fuel mixture is chilled below about 100°C.X>..........i.......:.a.: >. i.....;'it . 2..., . ..:.=.......>.....,.P. .,....+x.?.'A;

cases. A plausible mechanism has existed for years to explain the minorcorrosion observed, but it is still possible that extraneous oxidants arepartly responsible for the observations. Sme especially careful testsof these points would be welcome -and newer on-line analytical methodsseem to make these tests possible. Such testing will become especiallyvaluable if albloys Other than khtelloy N must be Considered for MSBR,While the equilibrium bekavi~r of MSER fuel with contaminants suchas stem, air, EF9 and is well under~tood, little is ~ H Q about I ~ therates of some sf these reactions. Studies of the really pertinent onesshould be initiated eIt is still necessary to demonstrate that large-scale initial purificationof the MSBW fare1 solvent can drop the 02- concentration to thedesired value, well below 30 ppmn. mile other facets sf the purificationscheme also lack demonstration, no other specification would appear to betroublesome aRadiation stability of the PSBR fuel at elevated temperatures wouldwst appear'to pose problem. The 72-16-12 mole % composition of LiF,B ~ F and ~ , TP~F~ has scllid compounds quite different from the mixes tested,and there is reason to suspect that P2 generation may be less of a problem(perhaps negli ib1e) at ambient temperature. This point, to whicha favorable answer might simplify some storage and handling problems,shoukd be checked.It should be emphasized that not all MSB chemistry is well. understood.Special problem areas that are chemical in nature, though coveredelsewhere in this docmeat, Are concerned with (1) retention and controlof tritium, (2) with the special grain boundary corrosion and crackingin Mastelaboy N, and (3) with behavior of fission products. Considerablechemical d@QElkOpment W i l l grObEibly be ZeqUi%ed to define andl to Solvethese problems.. .. >I.. d?,Y .Fission Product BehaviorGeneralCirculating liquid fuel simplifies fuel recycle by making possiblein B ~$M processing, as required to increase breeding gain. However, thischaracteristic also means that fission products are spread through a131parts of the fuel circulating system and peripheral system suck as offgas,drain tanks, etc. 'khis substantially affects the reactor operationand performance as re ards breeding, materials behavior, afterheat, andmaintenance. Thus a horougk of the fates of the fissionprQdUCtS in MSBR's is illlportalat $8 their devdopwent.As operation of the MSE and concurrent investigations emphasizedthe inpsrtance of the fission products, investigations of their fates inwere pursued in a variety of ways. Arrays of graphite and metalspecimens were examined after exposure in the core for periods of severalthousand hours. Samples of fuel salt were dipped from the agitated poolin the pump bowl or were pulled from beneath the surface of the same poolinto evacuated capsules with fusible seals. The gas above the salt in....c.*..

117....&.

11%PiSSklllPercent yieldproduct 4 3 3 ~ 23SU 239h6.160.760.341.984.000.223.476.000.020.131.340.740.566.496.1 f5.345.154.414.802.9 11.802.606.411.130.442.365.730.283.985.730.030.171.711.130.476.404.276.095.786.306.065 .oo3 .oo2 .QO7.171.320.802.074.560.623.134.560.230.421.73B .32I.@15.095.035.655.897.106.105.915.672.60....u.x;. ..g..;.:,

..*=...,....~_&x . s2 .....>-Table 5.4. Free emergy of formation at 650°C (LLGF, kcal)Lit, Be2+, and E’- are at unity activity; all others,activities in mole fraction units.......V. .SolidDissolved in2LiFmBcF2._ .....IJ.:.=....:iii..a;.,.,. .......a . ‘:&3....::x.*..........:s5+-363.36-364.64-341.80-310.92-389.79-316.93(- I5 0.7)-138.18- I21 58- 39-126.49-354.49-356.19-332.14-216.16-300.88-392.52-308.10-392.92(-296.35)-152.06-134.59-113.40(-186.3) -62-50.29-74.58-449.89-366.49-306.65-159.13-232.26-200.59-98.36-42.15-189.57-66.112-173.72....Source: C. F. Baes, Jr., “The Chemistry and Thermodynamicsof Molten Salt Reactor Fuels,” Symposium on Reprocessing ofNuclear Fuels, ed. by P. Chistti, Nuclear Metallurgy, vcl. 15, p.617, USAECCONF-690801 (1969)......c.:.:.:,....w

40" 4 o4 IO2 f Q3 4 8" 4 05 4 8" 9 8'%a /XUF3Fig. 5.5. Variation of equilibrium concentration of st~tr~turalmetal fluorides and the distribution of isdine as a function of theUFLJ1FF3 ratio in an BEu.:;.....

0-bFig. 5.6.Variation of partial pressure of vslatile fluoridesas a function of UF~/UFQ ratio in an MSR fuel 1353.

3. A nist shield encLosinmentthe sampling point provided a spesiak environ-4, Lubricating oil from the pump bearings entered the pump bowl at a rateof l to 3 cs/day.5. There was c~ntin~ously varying flow and blowback of fuel salt betweenthe pmp bow% and an overflow tank.In spite of these problem, useful infomation concerning fissionproduct fates in HSKE was gained..... c;cStable Salt- Soluble Fluoridesa - Stable fluorides showed little tendency to deposit onMastellay N or graphite 6403. Examinations of surveillance specimensexposed iPa the C01pe Qf the %w ShSkT@d Qaly 8.1 to 0.2% Of the iSQtopeSwithout noble gas precursors on graphite and Ha~telE~g~ N. The bulk of"the mount present s emed from fission recoils, based on an estimateby compere 6611 ana eneral consistency with the flux pattern 1601.The penetration of several fission products into the relativelygraphite is shown in Fig. 5.7. Note the flat profile for376S (ph-C?CUsSOK 37 xe, ~ " ; 2 = 3.9 min) in contrast to the 95~rand144Ce9 which do not have noble gas precurs~r~. Similar data were obtainedfor 89 r (precursor: 89Kr, TI/ = 3.2 fain) and 14a~a E ~ Q I (precursor:14% 3 ra/2 = 16 see), althou h in those cases the concentrationfalls off more with depth in ttas graphite. s he dip in "'7~s concentrationat the free-flowing salt surface (left side of pig. 5.7) was estab-Pished as real in this and other examinations [629, and similar dipswere noted in eXaK~ifaiY3g the graphite CQre Stringer removed fKOEn the mwmCompere and Kirslis If531 attributed these dips to significant diffusionof cesium atom in the graphite; such diffusion would also explain thevery flat 14'7cs profile shown in the figure. The mobility 0% 141csk... .%>

Table 5.5. Stable fluoride fission product activity as fraction of calculated inventoryin salt samples from 233 U operationNuclideWithout significant noble-gas precursor-.--With noble%as precursor‘5Zr 14”Ce 144Ce 147Nd 8gSr 131CS 91Y 14%aWeighted yield, %a 6.01 6.43 4.60 1.99 5.65 6.57 5.43 5.43Half-life, days 6.5 33 284 11.1 52 30 yr. 58.8 12.8Noble-gas precursor 89Kr ‘“‘XC2 91Kr “‘OXt?Recursor half-life 3.2 min 3.9 min 9.8 set 16 setActivity in salPRuns 15-17 0.88-1.09 0.87-1.04 1.14-1.25 0.99-1.23 0.67-0.97 0.82-0.93 0.83-1.46 0.82-1.23Run 18 1.05-1.09 0.950.99 1.86-1.36 0.82-1.30 0.84-0.89 0.86-0.99 1.16-l As 1.10-1.20Runs 19-20 0.95-1.02 0.89-1.04 1.17-1.28 1.10-1.34 0.76)---0.95 0.81-0.98 1.13-1.42 1.02-1.20aAllocated fission yields: 93.2% as3U, 2.3% 235’ti, 4.5% 23gPu.bAs fraction of calculated inventory. Range shown is 25-75 percentile of sample; thus half the sample values fall within thisrange.

c . . * -

l 25TI/^ = 25 sec; precursor:deeper graphite penetration by l4lCe than Ib4Ce shown in Pig. 5.7.Some observations indicate similar mobility for '9~r141Xe, TI/^ = 2 sec) may account for theGas Samples. - Gas samples obtained f r ~ m the gas space in the pumpbow% mist shield were consistent with the above results for the saltseekingIsotopes with and without noble gas precursors. Table 5.6 [64]shows the percentages of these isotopes whish were estimated to be in thepump bowl stripping gas, based on the amounts found in gas samples.Agreement with expected mounts where there were strippable noble gaspresurs~rs is satisfactory in consideration of the mist shield, contaminationprsbbems, and other experimental difficulties. Gama spectrometerexamination of the off-gas line showed little activity due to salt-seekingisotopes without noble gas precursors 1681. Exa~~inati~ns of sections ofthe uff-gas line also showed only small amounts of these isotopes present[691*Xoble MetalsThe so-called noble metals showed a tantalizingly ubiquitous behaviorin the MSRE, appearing as salt-borne, gas-borne, and metal- and graphitepenetratingspecies. Studies of these species included isotopes of %,Xo, Tc, Ru, Ag, Sb, and Te,..... ._.,...A,,=.:.........>C

Table 5.6. Fission products in pump bowl gas samples, percenta of MSRE productionratePrecursorISOtOpe Caldated percentISOtOpeTl/zstrippxl into gasPercent found in gas samplesbGaussCoarectedIsatopes with gaseous precursorsR9Sr “Kr 3.2 tin 14 6.5 * 1 5.7 f 1.21 “‘cs ‘37Xe 3.9 min IX 33 tb 25 + 6“IY 9’Kr 9.8 set 0.07 0.4 1 0.2 0.004 f 0.01“%a ‘““Xe 16 aec 0.16 a.1 f a.02 0.06 ?r 0.01Salt-seeking isotopes9SZr 0.06 f 0.01 0.01 * 0.004‘““Ce 0.03 + 0.01 -0.003 f 0.003““Ce 0.3 f 0.w Q.05 f Q.03147Nd 0.02 * 0.007 0.002 ?I 0.002“As noble gas precursor.‘Mean value.T!orrectd for estimated mist content [66,67].

u.3rnu5..iii, ..&...z 5......;.:.x.. ..:.M....‘.:.:.:.32 5wu ew“ 2......&.,.a52....,.:&&+&10-22 4 12 (9 44 45 46 9 24 36 42 44 45 55 57 58 59 76 ii 99SAMPLE NUMBERFig. 5.8. Noble metal isotopes in salt samples (as percentage ofcalculated inventory) [633...... +...., .%!A.. .?....., 153 .y..... :.&

4913031740.9 f 0.21.5 2 Q.50.7 f 0.32.3 9 (9.711 f 3-1 2 0.8-1 f 0.8

129.....r:x,......?.!.l

a. 3Q........- e .>. - %W CQwSidePkIg the resultsassumed, for lack sf otherfission products per unitarea fokllldk OK2 the speCim@IKS Were repreSC2ntatiVe 0% all the graphiteHastelkoy N surfaces. It was, of course, recognized that this assumptionwas tenusus at best, and the exarnibtati~n of post-operation specimens fromthe KSl3.E showed that it was not a very close approximation. Table 5.8shows fission pr~du~t distributions caaeulatea in this way for severalSets Of SUrVeiPIance spl2CiKllen.S and for parts Of the COKIpo€l@RtS PelllOVedin the poSt-operatiOn eXallIinatiOn. '%he CaEcUlate distributions varywidely for different areas both in the cfrculatin system and on thesane component.The f surveil%ance specimen array, exposed for the last fourmonths of operatisn, had graphite and metal specimens natched as tocsnfiguration in varied flow conditions [76]. The relative depositionintensities (1.8 if the entire inventory was spread evenly over all surfaces)Were a.$ ShUWn in Table 5.9.The examination sf some se ents excised from particular reactorcomponents, including core neta and graphite, pump bowl, and heat exersurfaces, one year after shutdown also ~evealed appreciable accmulatisnof these substances. The relative deposition intensities atthese locations are also shown Zn Table 5.9.~t is evident that net deposition generally was intense OWmetal than on graphite, and far metal, was more intense under more turbulentflowe Surface ~~ughess had no apparent effect.Extension to all the metal and graphite areas of the system wouldrequire howledge of the effects of flow conditi~ns in each regfs~, andthe f~actioni of total area represented by the region. (Overall, metalarea was 26% of the total, and graphite 74%.)A theoretical approach [76a] treating the noble metal transport asan ordinary convective turbulent diffusion p~ocess including gas interfacesas imperfect sinks shows some qualitative agreement with the sbserveddeposition behavior.Flow effects have not been studied experimentally; theoretical approachesbased on atom mass transfer through salt boundary layers, thougha useful frame sf reference, do not in their usual form take into accountthe deposition and release sf fine particulate material such as was indi-cated to have been present in the fuel system. Thus, much more must belearned about the fates of noble metals in molten salt reactors beforetheir effects OR various operations can be estimated reliably...w3k.,

Table 5.8. Percent noble metal distribution in MSRE based on examination of various specimens removedPercentages assume amount found on a specimen is representative of all graphite or Mastelloy N surfaces.Graphite- _._.. - ...-Hastelloy N“Nb 9gbfo g’Tc lQ3wu lQeRU lllAg lZSSb 129mTe 132~~ 9sNb 99$fo 99Tc 103~~ 106~~ lllAg lZsSb 129111~~ 13”TeCore surveillance specimensGroup 2 (32,700 MWh) [16] 36 11 7 10 34 41 15 70Group 3 (65,600 MWh) [8] 41 9 4 4 6 5 12 19 3 3 12 10chxlp4 (18,SOOMWh) [60] 7 4 2 3 5 3 3 5 3 26 4 8 16 12 18 16Group 5 (11,800 MWh) [62] 15 16 5 4 3 10 53 4 4 56Core stringer, topCore stringer, middleCore stringer. bottomControA rod thimble, bottomControl rod thimble, middleMist Shidd. QUtSide iFI SaltHeat exchanger, shell.Meat exchanger, tube17 5 12 47 33"79 77 138 47'39 11 56 44aPost-operation specimens [75] E36 32 40 13 85 43a27 19 15 11 35 14=7 19 7 10 74 23""9 27 21 5 68 35’”7 31 28 14 113 67a

Table 5.9. Relative deposition intensities far amble metalsSurveillancespecinr~ewsGxaphiteIaninnrTurbulent0.20.2a.2 0.06Q.Q40.16 0.15(9.10 0.07MetalLaminarTurbulent0.30.30.5 0.11.3 (4.10.30.3a.92.QReactorcoI-qmlentsGraphiteCore bar channelBottomMiddleTQlJTurbulent0.541.090.230.070.25 8.65 0.4ha!B.06 1.90 0.w0.29 0.18 0.6PMetalPump bowlHeat exchanger shellHeat exchanger tubeTurbulentTurbulentTusbulent0.260.330.210.73 0.271.Q 0.101.2 0.110.38 2.85 0.89”0.19 2.62 1.390.54 4.35 2.eCoreRod thimbleBottomMiddleTurbulentTwbwlent1.421.001.23 1.54 a.50 3.21 1.690.73 0.58 0.42 1.35 0.54”

133Although the noble metals are appreciably deposited on graphite,they do not penetrate any more than the salt-seeking fluorides withoutnoble gas precursors, as shown by the io3Ru data in Fig. 5.7. Deleteriouspenetration of Hastelloy N grain boundaries is discussed in Chapter7,.....>-Iodine....>.;&.....=, -.....

Table 5.10. Indicated distribution of fission products in mollden salt reactorsFission product groupExample isotopesDistribution (7%)In salt To metal To graphite To off-gas QtherStable salt seekers ‘“ZP, 144Ck, 147Nd -99 Negligible

L 35Future Work...............>..=.;. ....>.

136The c~~bant should be inexpensive, it must possess good heat transferproperties, and it must melt at temperatures suitable for stem cyclestartup. An ideal coolant would consist of compounds which are tolerablein the fuel or which are easy to separate from the valuable fuel mixtureshouPd the fluids mix as a consequence of a Peak.rejected its heat to an air-cooled radiator at a minimbun temperatureof about 1015°F 1781. The coolant mixture chosen [8,781 forthat application was BeF2 with 66 mole % of %iF (see Fig. 5.1). Thismaterial, which melts at 851sF, was shown to be completely satisfactoryf o that ~ w e [8,9]. An MSBR coolant, however, must transport heat tosupercritical stem at minimum temperatures only modestly above 700°F[IO]. Use of the high-melting %m coolant in EBR would pose realprobIems in stem generator design. The eutectic mixture of LiF andBeF2 (Fig. 5.11 melts at near 70Q"P, but its viscosity is high. Moreover,both these L ~F-B~F~ mixtures a ~ quite e expensive if ~ LIF is used.If normal LiF is used the cost is much less, though it is mot trivial,but a leak in the primary heat exchanger greatly damages the expensivefuel. Such ~~ixfures have, accordingly, been disc~unted as MSBR coolants.The alkali metals, excellent co~Iants with real promise in othersystem, are quite undesirable hers since they react vigorously with boththe MBR fuel ad With steEU3l. More Elofole metals Such as Bi Or Pb UndelrgCJmo violent reactions with eithe~ of the other working fluids, but theyare only fair coolants, and they are not compatible with the nickel-basedalloys intended for use in mBb.miten lnixturea of N ~ N O ~ mo3, , md mo2 have been used industrially(Hitec is a cornon example) as heat transfer agents, but not at temperaturesquite s~ high as are desirable in the MSBR. These materials m ybe stable toward irradiation in the primary heat exchanger, but extensiveinvestigation of this point would be required [79]. They are clearlynot stable toward the fuel, from which they would precipitate actinideoxides rapidly but not violently, and they would react dangerously withthe moderator graphite if a leak occurred in the primary heat exchanger.Thus they are not being considered for use in the MSBR secondafgr system.It should be noted, howeveb-, that these nit~ate-nitrite mixtures haveexcellent melting points, quite compatible with water and steam,and should certainly oxidize to T2Q any tritium diffusing from the reactorfuel system. They mayp if the tritium problem (see Chapter 14) cannototherwise be satisfactorily managed, prove valuable in a third coolantHOOP, as mentioned in Chapter 2.Several binary chloride systems are known to have eutectics meltingwell below 700°F [80]. Many of these systems are unattractive since theycontain high concentrations of chlorides which are easily reduced and,therefore, corrosive or which are very volatile. The only low-meltingbinary systems S% stable, ~~omn-v~latile chlorides are those containingEiC1; LiCl-CsCl (330°C at 45 mfe Z CsCl), LiC1-KC1 (355°C at 42 mole% KCB), Li61-R.bcl (312°C at 45 mle % wbC1). Such system Would be rehtivelyaq~~tasive if mde from 7 ~ ~ and 1 % they c ~uld iesa to seriouscontamination of the fuel if normal LiCl were used.Very few fluorides or mixtures of fluorides are known to melt attemperatures below 700°F (370eC). Stannous fluoride (SnF2) melts at212°C. This material is certainly not stable during long-term service.....-.:

.....i.... .1.37

138ORNL-DWG 67- 9423AWPig. 5.9.The system NaF-NaBP4.

139Behavior with Hydroxide Ion. - Sodim hydrowyfluoroborate (NaBF30H)is a common minor contaminant in comercial NaBF4 that is prepared byaqueous processes. Pure NaBF30H can be decomposed completely at temperaturesnear 100°C 1831 through the bimolecular reaction:This material cannot, accordingly, be used as a major constituentfor high-temperature coalant. However, when the NaBP38H is present invery dilute soPution in NaBPk it is "matrix isolated," the bimoleculardecomposition reaction is inhibited, and the material is considerablymore stable [84]. Such hydrsxyfluorobarates are appreciably more reactivethan fluoroborate to structural metals, as is discussed in moredetail in a subsequent section; however, there is considerable evidencethat as much as 30 ppm of H can be present as OH- in the fluaroboratemelt without undue corrosive effects upon Hastelloy N E8.51.Relative stability of such quantities of OW" (and, by analogy sf ST>may be an important aid to control of tritium (see chapter 14). If, forexample, the coolant can be mde to contain several ppm of H (18 ppm ismore than 4 kg of I% in the nearly %$s of coolant) ana if the exchangereac tisn....'.=a.....,:.=.,.:.;.; .-.9could be made to proceed, the coolant mixture would afford a means forholdup of the tritium diffusing from the primary heat e~hanger.Some initial studies have been perfomed in which B2 has beenbubbled into, or alternatively allowed to diffuse through nickel tubinginto, NaF-NaBFh melts 6861. These melts contained hydroxide ion equivalentto some 25 gpm of I% and a considerable quantity (perhaps 3000 ppm)of oxide.Direst observation by infra-red spectr~s~opy showed that QD-grew into the melts. No corresponding decrease of OH- occurred. Thesestudies strongly suggest that the exchange reaction indicated above isnot occurring but that the added deuterium is reacting with oxide impurities,perhaps byDq 9 1B2FgQ2- + NiQ 2BF30D' + N i ......mFurther study of such reactions is clearly necessary to determine theprecise mechanism and to establish whether the reaction can be madesufficiently rapid and quantitative to fix the diffusing tritium......

140

141.;.a....Compatibility with Hastelloy N. - Acceptable compatibility of theNaF-BF4 coolant with Hastelloy N under realistic conditions seem assuredby numerous csrrosion tests [42,87,88], though the number of such testsis much simaller than that for LiF-BePz mixtures.Thermodynamic data for possible reactions of the NaF-NaBF4 mixtureis in a somewhat less satisfactory state than is that for the fuel mixtures.The free energy change for the chemical reaction....:.:.>y....i.. Q YrFpIf the fluoroborate fs sufficiently impure the sorrssion product CrF3exceeds its solubility limit and precipitates as Na$rFG** While someof the initial COK~QS~~I-I tests with quite impure flu~r~borate mixturesproduced such precipitation (and severe corrosion) [42], none sf themore recemt studies have encountered this difficulty.......x.:q.... ,:.a+* Similar corrosion to produce CrF3 and ultimate precipitation ofNaM2CrP6 was observed long ago in studies of UF4 disso%ved in the ternaryeutectic of NaF-KF-LiF [!In].........A. ~,.,).....i .&.A .....523

142Interactions with Steam. - Flusroborates react readily with stem.The reaction, which is not violent or particularly energetic, isIt: is well known that the HF so prod~~ed is quite corrssive; it is capablesf reaction with Ni and Pe as well as with Cr and can cause generalizedattack. The FeF2 and NiF2 so generated are then capable sf reaction withere in the alloy [ 71. Cantor and Waller e851 have recently shown thatthe reaction sf OH- in dilute solution (2% ppm and below) with Hastelloyh;* CXCMP~S S%SW~Y; this ~eacti~n can apparently be represented asress of steam, as from a leak in the BR steam generators, torsborate ssokant c2rcuit must, accord g%y, be expected %Q %esultin increased corrosion. such an increase has, indeed, been observedin deliberate additions sf steam to an operat ng test loop containingremovable specimens in hot and cold regions 1 9,931 Relatively rapidcorrosio~~ foilowed the ste addition; chrsmim concentration in thecirculating salt rose from 82 ppm to about 320 ppm in about 11900 hoursand remained essentially constant 6933. Examination of specimens removedat intervals from. the loop also indicated that the attack decreased withtime after the initial exposure to stem [8%,93]. These data strong1ysuggest that the Hastellay N system can tolerate inleakage of steam and,once the leak is repaired, could continue to operate without extensivedamage even if the salt were not repurified [93]. Et is clear, however,that csntinuous inleakage sf stem in appreciable quantity would requirea pu~ification system for the fluomborate salt.b.,Interactions with Fuel Salt. - The E BR design [lo] will assure thatthe pressure on the coolant is slightly higher than that ow the fuel salt,so that a minor leak in the primary heat exchanger will result in coolantsalt entering the fuel. Such a leak, even though small, will be recognizedat once because of the marked reactivity loss resulting from admissionof boron into the fuel.*It is worthy of note that the NaF-NiBP4 mixture for that study,which was shown to contain at least 25-38 ppm of H along with about350 ppm CP~+ and 3900 ppm 02-, was drawn from an operating test loopwhose perfor nce with respect to cor~osion was quite sati~fact~~y.. .

14 3....,&........w...&Mixing a small quantity S€ NaF-NaBF4 salt with the fuel effectivelydiss~ciates the MaBFb into NaF and BF3. The NaF will siissol~e in theLiF-BeF2-ThP4-UP4 salt mixture; the BF3 will distribute between the fuel.salt and the available vapor space and, if operation of the reactor werecontinued, would be swept into the off-gas system. Additions of largequantities of coolant, as might occur in the unlikely event of a grassfailure in the primary heat exchangerp produce more complex effects andcan result in appreciable pressures of BF3.The solubility of EFQ in the FEBR fuel solvent has been carefullymeasured [94]. The solubility decreases with increasing temperature andincreases linearly with BF3 pressure. Prom these data one can calculatethe equilibrium pressure of BP3 from small leaks of NaF-NaBF4 into 'PISBRfuel in a closed system. This equilibrium pressure clearly depends uponthe amount ~f NaP-MaBP4 admitted and the available vapor space. If weassume that the only vapor space available is the small bubble fraction(assumed to be 1%) in the fuel,* then one can show that inleakage at1250°F of 1 cubic foot (116 Ibs) of NaF-MaBP4 coolant results in a BF3pressure of 0.17 atmosphere. It seems clear that leaks of this order donot result in a dangerous situation.Addition of substantial quantities ~f NaBF4 t~ several fluoridemixtures generally similar t~ the PBBR fuel resulted in liquid-liquidimmiscibility [9%]. Mixing equal weights, for example, sf 2LiEeBeF2and b?aBP4 at 600°C produced ma liquid phases. The less dense phasewas rich in BF3 and contained little BeF2; the dense phase was rich inBeF2 with little BP3. LIP and NaF partitioned between these phases, withthe light phase rich in NaF and the heavy phase rich in LiF. Other experiments[95] showed that UP4 (presumably 'PhF4 wou%d behave similarly) wasalmost completely retained in the heavy phase. No such equilibrations todefine the miscibility limits or the equilibrium distributions of theseveral ions in the two phases have been performed with the MSBR fuelcomposition. Consequently, the extent to which this liquid-liquidimmiscibility limits the 13F3 pressure upan mixing Parge quantities ofNaF-MaBFb, with MSBR fuel cannot be accurately assessed. However, thefact that considerable pressures of BF3 (in reasonable accord with estimatesbased on the measured sslubibity 1941 of BF3) are developed uponmixing MSBW fuel with one-third ita volume sf NaF-NaBF3 coolant has beendemonstrated experimentally [%I.It appears certain that small Peaks of coolant into the fuel systempose no real problems. It is clear% however, that additional study ofmixing S% these fluids in realistic ge~wefiries and in flowing system isneeded before one can be certain that $10 potentially damaging situationcan arise as a consequence of a sudden major failure of the heat exchanger.* This is clearly a worst case. It completely ignores the largevolume in the off-gas system.

144Purification ~f Flusroborate Mixtures. - Initial purifieati~n of%Sd?-b?aB??~, Elf-xtUreS for LIS€! ila COTXCJS~O~ and Other engh2eriag tests hasnot posed serious proble~i~s, since the purity specifications are lesssevere in several ways than are those for the fuels, and quite pure MaBF4preparations are available commercially at reasonable C O ~ ~ S Recent .preparation E971 methods have consisted simply in (1) mixing the MaBP4with the required quantity of NaF, (2) heatin the mixed powder at acontrosled rate to 30O6C under redused pressu e, (3) heating the materialto 500°C under a static atmosphere of inert gas9 (4) mixing the moltenmixture by briefment with flowinwith BF3$ (5) purg ng the BF3 by a brief treatand(4) transferrin the salt mixture to itsstorage containe cak analyses uf mt rial prepared in this simpleway show oxide concentrations of Out 325 ppm and hydlrogebl (prOtofl) Concentrationsof about 15 ppm. Ana ses for ~e2+ S~OW values near 150 ppm;this eoaahd almost certainly be decreased by a K L Q stringent ~ ~ purifieatiQnpPOc@dUreSimple prosadurea such as that described above my suffice for theoriginal coolant loadi of an %BR, but it seems certain that additional,on-line treatmnent W i l l e required during reactor operation. The inevitsbilityiof small inleakages of steam, for example, will ~ equf~e somerepurifi~ation reprocessing. &reover, if the fluoroborate mixture isto help with the tritium problem (by tritium exchange with NaBF30H UTother reactions yielding NaBF3QT), some processing method for removFng thecontained T on a reasonably short time eye will be necessary.The afeactian Of fPUQrObOPatE2 With Stebl.. -.is a reversible one, and treatment with P%F (appropriate pressures ofin31 condition. It appears, however, that this reaetioin m y be~ O Q inefficient for we in a practical system. Such purification c~ubclcertainly be aseomplished by use of fluorine; this reagent (mixed withBFg) should certainly remove the 6pb COT> as well as the contained oxides.andIt is possibhlsle that direct fluorination (presumably using a frozen-wa$%fluorinator) may be necessary, but. studies are needed tu determinewhether less active rea ents can serve the pu~pose.

.,. ....*rn . 145I....i.:.=Radiation Stability. - The fluomborate coolant salt will be exposedto intense gamma radiation and an appreciable delayed neutron flux in theprimary heat exchanger. Consequently investigatiuns of such radiationeffects were undertaken [98].The gama irradiation was carried out in spent High Flux IsotopeReactor (HFIR) fuel elements. A 2.0 cm ID x 9 CHI Hastelloy M capsulecontaining 32 g NaBF4-NaF (92-8 mole %> was irradiated for 1468 hr at600°C in three successive elements discharged from the HF]CR. The totaldose was 9.7 x IO" R (0.1s w/g average, Q -5 ~ / maximurn) g ; for comparison,the anticipated gama dose in a PBOO-m(e) mBR is "4.25 W/ga Thepressure in the capsule was monitored during the irradiation and showedo~ly results consistent with the expected BP3 equilibrium pressure.Analysis of the gas at the conclusion of the experiment: (ambient "C>showed I%Q HF, BP39 OP fluopine.Examination of the salt on opening the capsule showed no observableeffects uf the irradiation A 29-crn2 ~astelloy N co~~osion specimenexposed in the capsule showed negligible weight loss (0.8607 g). Theresults in the irradiated capsule were in all respects comparabEe tothose observed in an unfrradiated control. experiment.A literature search and analysis indicated the direct n,a reaction....s . 2....a.3wo~ld be the mjor source of fluorine due to the delayed neutron flux.The y radiation and other considerations indicate that secondary effectsfrom the Li and 0: recoils are of little consequence, apparently becauseof rapid recombination of fragments from collisions with BP4-. Thus,if N~BF~-N~F (92-8 mole X > is exposed to an estimated 15 x 1016 delayedw/sec in an MSBR primary heat exchangers the fluorine prod~~ed wouldmount to onPgr 8 moles of fluorine per year.Evaluation and Sumary of Needed Work....

Of dilute xaBP36K SOBUtions With netds 32eed definition. In additiotl,the mechanism by whish tritium, diffusing from the fuel system, can beBF3OT- needs additional study before its value can reallybe assessed.As imdicatecl above, several of the physical property values haveted. These esthates are almost certainly adequate for thepresent, but the pra ram needs to provide for measurement of thesequantities.Compatibility of the UaF-Ha F4 With HastdlOy N UndeP IIQlX€laE Ope%-sting conditions see assured. Additional study in realistie flowingsystems, of the corrosive effects of stem imleaka is nesessaq. Thisstudy, claseHy allied F.Jith the equilibria and kine ics among the hydroxidesana oxides described above, needs uitimate~y culminate in a demonstrationloop capable of simulating steam inleak e and coollant repurificatisn.xiwg of coslant and fuel clearly requires additional study. Theon which results from equilibration of these fluids is reasonablywell understood, and, even where large leakages 0% CCKII~TI~ into the €ere%are assmed, the ustimate sv@quiiib.k-imrF seem to pose %to real danger.MOW~Q~P, the real situation may well not approximate an equilibriumsystearas are Backin and necessary ePurification roceciures €Q% the coolant mrrixture are adequate for thep~esent and can be used to provide mteriai far the msty necessary tests.These are not adequate fer ultimate an-line processi~lg of the coolantmixture during operation. Flu~ri~ation of the coolant, on a reasonablecycle time, would almost certainly suffice thou h it has not been demonstrated.A process using a less aggressive reagent is clearly desirable.The fluomborate mixture has shorn completely adequate radlationstability in the single ethsugh realistically severe) test run. Adaitionalradiasit3n testing Ofthis Wtepial in a flOWiIlg syStEil wouldseem desirable and should ultimatdy be done, but this study would motat present have a high priority.Qf the several alternative secondary salts mentioned in the text thebest candidates lithim. mere seems littie doubt that S~V-@ral of these mte%ids (LiGl-KCl mixtures, for example) CQdd be shownto be suitable coolants, although a considerabie program of chemicaldevelopment and co~rosion testing would be necessary. The same wouldbe true for the nitrate-nitrite mixtures before they esdd be used athigh temperature in an MSR tertiary l ~ ~ p .Analytical ChemistryReauirementsIn order to exploit fulPy the unique features of the MSR concept andensure safe and efficient reactor operation, it will be necessary to mintainadequate sur~eillance of the composition sf various reactor streams.Ideally, all such analyses would be performed autonatisally with transducerslocated in the salt stre , since analysis of discrete samples inhot cells is subject to unavoidable delays and is expemsive.

...,.; and UP6in gaseous streams from the reprocessing system.It should be noted that in addition to economics of time and expense,the in-line techniques will provide information not attainable by discretesampling methods. A notable example is the i33*/u4+ ratio in the fuel.'This ratio is prohibitively sensitive to atmospheric contamination duringsampling and sample transfer in hot sells, and is rather neaningless onfrozen samples because the ratio undergoes changes during cooling as aPeslldt Of I2qUilibriU shifts. hother example is the detE?I-RIinatiQn oftrace csnstituents in gases, which is notorio~~ly difficult if not impossibleto do by withdrawing samples.It is evident that the ultimate need is an analytical S ~ S ~ I f~ o M an ~MSBR that includes the most complete in-line analytical measurements possible,backed up by adequate hot. cell and analytical laboratories. Inthe interim period it is necessary to develop capabilities and to provideanalytical support for the technologgr programs.ExperienceThe mjor development of analytical methods for discrete sampleswas associated with the operation of the MSRE. The analytical methodsfor this reactor were developed to support the objectives of the chemicalsurveillance program [S, p. 13. With the exception of in-line analysesof the off-gas and remote gama spectrometry (described later in thissection), all analyses were performed on batch samples either in East cells8% bg. bench top Hlgthoas.Prior to the MSRE program, we had substantial experience in the laan-$ling and analysis of nonradioactive fluoride salts in the program aIonic or instrumental methods had been developed for most metallic constituents,and methods were available for P- (gyrohydrolysis) Kg91 andd. * me ~3+/~4+ ratio is a measure sf the redox potential of the fuelwhich infbuences the rate of corrosion and the distribution of certainfission products and tritium in the reactor system (see 5-21 to 26).........

sulfur [lO0]. For RE application it was necessary to adapt thesenethods to ho -cell OpeKatiOnS. ki teChniqUe invOkVhg the @VOlU%iOHa ofelemental oxy en by reaction with BF3 has since resulted in the versa-PPL) method [loll for oxide in inorganic samples. A nonselective~ e ~ s ~ rof e "reducing ~ e ~ t po.~e~" of adequate sensitivity had been developed(hydrogen evoabution method) [l6B2]. A general expertise 6183% in the radiochemicalseparation and measurement of fission roducts was availablefrom earlier reactor program at 8experience with in-lineas analysis, particularly processy [%04], was availablefKOm Other proation of the and in the subsequent technoa_ogyPro ~ ~ ~ of e EE?%kOdS ~ f Q fSCrt2te ~ ~ Sampl@S @ Was ~ cQntiI2Ued, ~ andthe Laboratory has acquired instrumentation for newer analytical techniques.Instrumental methods whisk have or are expected to contributeinclude: x-ray absorption, diffraceion, and fh~resceneespark source mss spectro~~~et~y, ESCW and Auger spee-%on ~ croprsbe measurements> scanning electron microscopy,try, Fourier transform spectrometry, neutron activationanalysis, delayed neutron nethods, photon activation analysis, and scanyparticles, e.g. protons. The detailed descriptionsf all analytical methods availabbe to t%se program is beyond the scopeof this report. A tabulation of our analytical capabilities is givenelsewhere [lO5]. Certain deveEopmewts merit additional mention and aredescribed below.0 - The preparation of homogenized smples of MSelemental analyses presented problems beeabase of the radioactivity andthe hygroscopic nature of the salt [9, p. 273. Salt samples were takenin small copper ladles which were sealed under helium in a transportcontainer in the sampler-enricher [IO61 for delivery to the hut cell.There the ladle was unloaded and sectioned. The salt was removed fromthe truncated ladle and homogenized by a vigor~~s shaking in a pulverizervessel. Salt transfer was then mde with minimal atmospheric exposureto a polyethylene vial threaded into the bottom of the pulverizing~ssei. his p-oceaure a free-f~owing powdered S E Z T I within ~ ~ ~two hours of receipt of the ladle. Atmospheric exposure was sufficietktto compromise the determination of oxide and U3+ but did not affectsther measurements.oxide. - ~ecause of the sensitivity the pulverized sale: to u11-avoidable atmospheric contamination, we adopted transpiration techniquesin which the entire samples could be analyzed. For the more criticaldeterminations, the most successful application sf transpiration techniqueswas the determination of oxide by hydrofl~rination [lO7]. Themethod is based on the evoEution sf water which occurs when melts aresparged with mixtures of anhydrous MF in hydrogen. By removing surfacenoisture with a premelting hydrofluorination step and by measuring thewater evolved from 50-g samples as an inte rated signal from an electro-Bytic moisture monitor, we measured oxide concentrations of about 50 ppmwith precision better than +lo ppm.w.:.,....

149Uranium. - Analyses for uranium by eou%ometric titration [Is] showedgood reproducibility and high precision (0 -5%) but on-line reactivitybalance calculations were about 10-fold more sensitive than this in establishingchanges in uranium concentrations within the circuit. We havedemonstrated that it is possible quantitatively to collect the decontaminatedUPg from the fluorination of 50-g samples of molten fuel [lQ8].The technique was used primarily to separate uranium for precise isotopicanalysis, but sufficient work was done to establish its potential for amore accurate uranium determination by measurement of the separated uraniumoutside the hot cell.,.:.:.=....;.&...,:*s .....- U3*e - We tested a hydrogenation transpiration method for the determinationof ~ 3 + in fuel [IOS]. he rate of produstion of MP from thespar ing of fuel with hydrogen is a function of the instantaneous ratioof U5+/U4+* Because corrosion products also contribute HP the integratedyield from a batch hydrofluorination is related to fuel composition by anequation [log] that cannot be explicitly solved for uranium ratios. Byuse of computer techniques we devised a 4-step, %-temperature reductionprocedure to produce HP yields sensitive to ratio change. During 2 3 5 ~operation (with 0.9 mole z UP^) we obtained ~ 3+/~4+ ratios in reasonableagreement with "book values" obtained from reactor charging and operatingdata (Fig. 5.10). Nowever, the method proved inadequate for the lmerconcentrations of uranium in the 2 3 3 ~ fuel.we also applied a voltametric method to the measurement u3+/u4'ratios in remelted fuel samples [1aC1] (see section on Elestr~~he~~~icalResearch for a description of voltametry). We performed these measurementswith electrodes inserted in samples remelted in their ladles. Moreatmospheric exposure was incurred than in the oxide determinations, becauseit was necessary ts cut off the top of the ladles to acsornmodatethe eHectrodes. AccordingPy, we obtained ratios below those expected.We were, however, able to observe normally shaped VO~~~XIETIQ~I-~IILS for Cr2+and u4+waves and to follow the reduction of the fuel by hydrogenspargirtg. We also observed changes in ratios with temperaturethat were consistent with thermodynamic predictions sf equilibrium skiftsbetween the uranium couple and corrosion product ions. This indicatedthat the radiation level of the samples had negligible effect on themethod and supports the potential of voltametry for application to reactorsalt streams.Spectrophotometry of Radioactive Samples. - We constructed a facility[IlO, pp. 202-284] which permitted the measurement of highly radioactivesamples- within a hot cell- by using the components of a spectrophotometerLocated outside the cell (Fig. 5.1%). A system sf extended optics directedthe chopped reference and sample beams through the cell walls,focused the sample hem at the center Sf an optical i=Ur:ba~e, and ret~l-nedthe two beams through the wall to the m~nochrsmatoH-detestsb- section ofthe instrument. The system design included devices for remelting largesalt samples under inert atmosphere and dispensing portions to spectrophotometriccelPs, but because of the imminent shutdown of the MSRE we

I .et.69.4a ANALYTICAL RESULTS( H2- HF TRANSPIRATION METHOD)L1.2s-”Y- 4 .63b-45-IiQ0.83Y0.66.46.26full power hours3+ 44-Fig. 5.10. u /v in the MSE fuel. salt runs 5-14.

,... .......Pig. 5.1%. Spectrophotometer installation....C.$j< -.. . .:.:.x....

devised a sample system to fill windowless cells by direct immersion inthe fuel ~espite precautions to prevent atmospheric exposure the ~ 3 +in these samples was completely oxidized before measurements ~ould bemde. we were ame, ~~Q~eveg-~ to the spectra sf u4+ ana (i=~ilowareduction with urani~ metal) that of u3t C~KIEJ~I-~SO~ sf: the spectrasf these samples with those of nonradioactive preparations indicatedno adverse effects from the activity of the fuel and demonstrated thefeasibility sf the technique. he facility has since been used to measurethe spectra of transuranium elements and protactinium in molten salts.&. .uipment was installed at the mm to perform limitedcell as a transducer. By means of an oxidation and absorption train Ella8,p. 1961 we were able to measure both total impurities and hydrocarbons inthe off-gas. The sampling station also included a system for the cryogeniccollection of xenon and krypton on molecular sieves to provide concentrateds2mples for the precise determination of the isot~pic ratios sfkrypton and xenon by mass spectrometry. During the last two runs Of MSW,we set up equipment [Ilk, p. 183% at the reactor ts convert the tritiumin various gas stre S tCi Water for EiC2asuk~tllent by sCin'Ci%latiOn COUflting.. ..K.*... ".a,,... . - .,* - By means os a precise collimationsystem mounted OB a maintenance shield, rad ation from aeposftea fissionproducts on compsnents was directed to a hi h-resolutfon hithiUlll-dri%tedgemaniuna diode [aQa, p. 361. Prom the g a spectra obtained we wereable to identify specific isotopes such as noble metal fiseiaw productsand to reap their distribution by moving the ssllimating system [ll2].the latter runs of the reactor such me U%emC?ntS Were KWde dUKingpower operations [ll2].Bismuth. - The investigation of the metal transfer rep~o~essing systemdescribed i~ chap. 1.1 required the development sf more sensitive methodsfor the determination of bismuth in fuelsa We found ehe inverse pokarographictechnique to be most wefbll, with detection aimits of about 50 ppb[ll6, p. 2081. The bismuth is deposited in a pendent mercury drop eEectrodefrom HCE solutions in which copper is masked with thiocyanate, andmeasured duriw an anodic scan. 19e also developed a speetrsgraphic methodof at least equivalent sensitivity; however, it incorporates a preesncewtrationby extraction with dithieone and requires Parge salt samples.

153.:3

154Fig. 5.12.current-voltage wave and derivative %or reduction of9 experimental curve (-1 0Y++ to u3+ theoretical points g .>

@......i..*.. +:.&.......... 2.s:pa:.:.= ................... ..>g.'.:.:.: .;.;.,.y.+-0.2 - 0.4 0 0.2 0.4 0. 0.8volts PLATING TIEWEVSFig. 5.13.Chromium stripping peaks for various plating times.....,.:. w, . ......,!.X,

Vi

w-in.i.....Fig. 5.14. Construction of the lanthanum triflusride membranereference electrode.

15 9.....A:

The latest innovation in cell design is an optical probe which lendsitself to a sealable insertion into a molten salt stream [H14, p. 711.The probe makes use of multiple internal reflections with a slot of appropriatewidth cut through some portion of the internally reflectedlight beam (Fig. 5.l6). During measurements the slot would be below thesurface of the molten salt and would provide a known path length for absorbancemeasurements. It is proposed that the probe could be made of.. ;.I’.....,A.. .,.&: PU~+, ~r’-%~ ~d3+? ~m3+,~r and H O ~ + ~ Semiquantitative characterizations including absorptionpeak positions, approximate peak intensities, and possible assignment ofspectra, have also been made for ~i~f, v2+$ v3+, E U ~ + sm2+9 ~ cm3+, and 02-*Me &re d S Q investigating protonated species in the proposed coolantsalt by spectrophotometric methods. Evidence for the existence ofhydrogen-containing impurities in XaBP4 was first obtained from nearinfraredspectra of the molten salt and in mid=-infrared spectra of pressedpellets of the crystalline material [116, pp. 94-96]. Deuterium exchangestudies are being perf~mecl to characterize the protonated species in themolten fluoroborate melts [114, pp. $31. Two very sensPtive absorptionpeaks have been identified and are attributed to species that contain-OH and -OD. The abs~rption spectra of several other species have beenobserved in flusroborate melts [U6, p. 1361. Our spectrephotometricprogram is also providing data for the identification and determinationof solute species in the various melts of interest for the fuel salt reprocessingsystem [l15’je......:.:.:.:

LIGHTBEAMPig. 5.16.%%stted optical probe for spectral analysis.

163Commercial gas chromatographic components for the high-sensitivitymeasurement of permanent gas contaminants are not expected to be acceptableat the radiation levels of the EBHP off-gas. Valves contain elastomerswhich are subject to radiation damage and w%iose radiolysis productswould contaminate the carrier gas. The more sensitive detectors generallydepend on ionization by weak radiation sources and would obviously be a%-fected by sample activity. A pr~t~type of an all-metal sampling valve61231 has been constructed to effect 6-way, double-throw switching of gasstreams with closure effected by a pressure-actuated metal diaphragm. Wedeveloped a helium breakdown detector capable of measuring sub-ppm concentrationsof permanent gas impurities in helium. Use of this detectorin a simple chromatograph on the purge gas of an in-pile capsule teat(&nR-49-61 demonstrated that it was not affected by radioactivity [l24].Ironically, subsequent tests with a more highly purified carrier gasrevealed sporaaiciuy noisy operation caused bgr unstable discharges ~1231.Tests to circumvent this by controlled impurity additions were suspendedbecause of more exigent problems.The analysts of the coolant cover gas involves less radioactivitybut more complex chemical problems. Currently we are investigatingmethods for the determination of condensable material tentatively identifiedas BP3 hydrates and hydrolysis products [llO, p. 2Q%] and forother €oms of protons and tritium. We believe that "dew-point" anddiffusion methods offer promise for such measurements ElI-4, p. '921. Wehave developed an improved Karl Fischer co~lo~~ietric titrator to providecalibration measurements of "H20'' in both simulated and actual C Q V ~ ~gas smples.In-Line Applications. - We recently demonstrated the first successfulchemical analysis for a flowing molten fluoride salt stream [ll4,pp. 69-70] by measuring U3+/U4* ratios in a loop being operated to determinethe effect sf salt QR Hastellsy N under both oxidizing and reducingconditions. The test facility is a Hastellay N thermaP--convectisnloop (NCL-21) in which LiF-BeF2-ZrFq-UP4 circulates at about five linearfeet per minute. The analytical transducers are platinum and iridiumelect~~cl~ that are installed in a surge tank where the temperature iscuntro.~eci at 650"~. ~e have monitored the u~*/LJ~+ ratio over a periodof several months on a completely automated basis. We desfgned a newcyclic voltameter, which provides several new capabilities for elect~ochemicalstudies on molten-sale system, for use with this system. Thevoltammeter can be directly operated by the PDP-8 family of computers[U6, p. 1381. A PDP-8/1 computer is used to control the analysis system,analyze the experimental output, make the necessary calculations, andprint out the results. As equilibrium c~nditi~n~i were being established,increases ~n the ~ 3 + concentration were followed as chromium SIQWIY disss~vec~from the ~aste~loy N, causing ~ 4 f to be reduced. Precipitousdrops in the U3* concentrations were also observed due to the intr~ducti~nof oxidizing Contaminants when metal specimens were inserted into themelt (Pig. 5.17). Work is continuing toward the application of our electroanalyticalresearch experience with the CO~~SS~QI-I product ions fortheir determination in this system.

16 4OWNL-DWG. 78 - (43791 EIIIII1IIb82 0LOOP OPERATION hourse5.17. In-line U3+ detednation in simulated NSRE fuel..... r..;;.~ .

165Provisions have been mde in the design of the two MSRP engineeringIoops, the Coolant Salt Technology Facility (CSTP) and the Gas SystemTechnology Facility (GSTF), for the installation sf our analytical transducers.This will provide us with experience on systems where the operatingconditions will more nearly represent those which will be encounteredwith an operating reactor.Future Work.......? -.,

16 6In-line methods for reprocessing streams must ultimately be developed,but the analytical requirements are not yet established.Hethods for the in-line analysis of the fuel will be readily adaptableto fluoride reprocessing streams, and measurements in the less corrosivechloride stream should be much simpler. Moreover, this less formidablesolvent has encouraged others to work in the medim and the literaturecan be exploited when requirements are established.The low tolerance of the $EBB fuel to oxide contamination wecessitatesthe development of in-line techniques for oxide measurement.Presently, we feel that a counter-current hydrofluorination technique(transpiration) offers the most promising approach to the in-linemeasurement ~f oxide, with ekectrocherdcal measurement of OH- afterBF equilibration or the spectrophotometry of sxyanions [I16 pa 136 ]as alternatives. The in-line application of transpiration methods willrequire the development of precise metering system for loa salt flows.We have an immediate need for transducers (under development) formeasurement of hydrolysis product in the cover gas of the GSTP. Later,methods of improved sensitivity will be required to analyze the heliumcover gas of the GSTP. We have acquired an ultrasonic detector 61291which we will couple with an existing chromatograph to provide aninstrment for the sensitive determination of contaminants Pn GSTP

16%References for Chapter 51. Clark Goodman et al., ihclear Problems of Wsn-Apeou Pluid Fwled!?eactors, Massachusetts Institute of Technology ~ USAEC Report11HT-5000 (Oct 0 15, 1952) *2. George Scatchard et al., Chemlcal Problem of JJon-Aqueous FluidPue Zed Reactors, Massachusetts Institute of Technology , USAECReport MIT-5BOB (Oct. 15, 1952).4. W. R. Grimes and %a. G. ~ 1 1 High , Temperature FtleZ Systems.Litaature Smvey. Y-659 (July 28, 1950).a5. W. R. Grimes et aZ., "Chemical Aspects of Molten Salt Reactorfuel^,'^ in Fluid Fuel Reactors, J. A. Lane, E. G. EfasPherson,and Prank Mashan, eds e , Addison-Ideskey P~bliski~g Coo InsCambridge, Mass. (1958)6. W. 8. Grimes, "Phterials Prsblems in Molten Salt Reactors," inMaterials md Fuek for High l'emperutum JJucZear Enepgy Appli-@&ions, ed, by M. T. Simnad and E. R. Zumwalt, the M.B.T.Press Mass (1964)8. hTs R. Grimes, NucZ. AppZ. Tech. 137 (2.970).9. R. E. ~homa, Chemical Aspects of !.ERE Operat-lon, Om%-4658(December 1971) L10. ConceptzlaZ Design Study of a Single-Fluid Molten-Salt Breeder Reactor,OWadE-454% (1971).11. L. Brewer et aZ., PIDBC-155% (1945) and L. Brewer, The CkmLstq ma'MetaZZurgy of IdiscelZaneous MatemGzk; Thermodynamics L. L. Quill,ed., McGraw-Hill, Mew York, pp. 76-192 (1950).13. PI. Flood and T. Forland, Acta Chim. Scand. 1, 592 (1947), andH. Flood, 91. Forland, and K. Grjotheim, Z. znorg.. AZZgem. C"nem.276, 289 (1954) e-14. R. C. Briant, A. M. Weinberg, E. S. Bettis, and W. K. Ergen, e t ale,JJuc. sci. EEg. 2, 79% (1959).

.....riii. %.!.%17. W. R. Grimes, No V. Smith, and @. M. Watson, J. Phys. Chem. - 62, 862(1958) e21. fGR Program Se~ann. Prsgr. Rept. Feb. 28, 2968, om-4254, p. 88. I....-24.25 *R. E. Thoma, ed., Phase Diagpms of Nuclear Materials, om-2548(NQV. 6$ a959>.L. V. Jones et al., ""Base Equilibria in the Ternary Fused-SaltSystem LiP-BeF2-UP4," J. Am. C~PCTRI. SSC. - 45, 79-83 (1962)26 s27 &'R. E. Thoma et ale, "Phase Equilibria in the System BeF2-ThF4 andLiF-BeFy-ThF4,'s €7. Phys. aem. _p_ 66b9 865 (1960)*28 e29 030 e21 a"Phase Equilibria in the Fused Salt SystemLiP-rnFi+ and NaF-TkF4,B' ei. PkIS. nem. __g 63, 1266 (1959) sR. E. Tksma et d.,e. F. w@aver e$ GZ., '2?hase Equilibria in the systems UP4-ThF4 and%~F-UPL,-T%FL+," 6. Am. Ce~m. So@. - 43, 213 (1960)&E€? Program Semiam. Progr. Rept. JuZy 32, 9964, Om-3708, p. 214.z" .-....

16932. K. A. Romfserger et aZ., "Phase Equilibrium Studies in the Uranium(1V)Oxtde-Zirconim Oxide System," presented at 151st National ~ eetingof American Chemical Society Pittsburgh, Pa a %Iarch 21-31 1966 a34. C. E. Bmberger, R. G. Ross, and C. F. Baes, Jr., J. ITn02ng- nsilcz.eiZem. - 33, 767 (1971)35. C. F. Baes, Jr., NucZem li&taZZw?gy, Vo1. 15, Symposium on Reprocessing0% Nuc%ea~- Fuels, ed. by P. Chfotti, USAEC-COXF 690801, 63.7(1969) e46. MSR Program bS@mimne P~ogr. Rept. Feb. 28, 2970, ORNL-4548, p. 152.41. S. Cantor et aZ., Physical Prspcwties of MoZ-t:e~ SaZt ,?eactor Fuel,OmL-TM-2316 (Augue t 1868) 043. W. D. bfanly e-t: ak., Ppogress in Nuclear Pnergg Series 117, Vsl. 2,p. 164, Pergamon Press, Inc., Lond~st (1960).44. id. D. Manly et aZ., FZuid FueZ Reactors, p. 595, ed. by 9. A. Lane,E. G. MacPherssn, and Frank Maslan, Addison-Wesley Publishing eo ,Cambridge, Mass. (1958) a.......... . .:.:

51 D52 *53 *54 a55 *56 057 658 D59 0 C.F. Baes, JP. "The Chemistry and Thermodynamics of Halten SaltReactor Fuels 9'' Nuclear MetaZZUP6Jl.J g, 617-44, USLIE6 Coni=. 690803.(August 1969)...... c.962.63 e64 m65 *6667.68 e69 078 *

71. MSR Frogram Sevriann. Progr. Re@. Peb. 28, 1969, OWLYL-4396, p. 139.74. P. F. Blankenship and S. S. Kirslis, Reactm Chemistry D;V. Ann.PPO~P. Rept. Des. 34, 9969, Owp9L-4229, p. 15.75. MSR Program Semiann. Progr. Rspt. Azlg. 31, 1971, 0 E-4928, pp. 54-5.76, IdSR Program Semiann. P’rogr. Rept. Peb. 29, 1972, OkNL-4782, in press.77. 8. P. Wichner, Side Stream Processing for Iodine and Xenon Removalfron the MSBR, Om-CF-72-6-12, in press.78. P. N. Maubenreich, J. R. Engel, AhcZ. Appl. Tech. s8, 118 (1970).78. E. G. Bohlmann, Heat Wansfer Salt $02” Pifgh TempeFature Stern Generation,OWL=-Tbf-3744 (in press) e.:&.

90 es%aFJm(Joint Amy-Navy-Air Force) InterimThermal Research Laboratory Dsw Chemicalgan *Thermochemical Tables,Company, Wdland, Michi-9192 e93 094.95 a96 esa.98 0E. L. Covere, H. 6. Savage, and J. PI. Baker, J. Nuc2. Mater. 34,99 (1970).99 a100 D181182 D183 0104.105 s106 e107 e

173109. MSR Paoogram Serniann. Progr. Rept. Feb. 28, 1967, QRPJL-4119, p. 158.110. NSR Prsgrm Semiann. Pragr. Rept. Peb. 28, 2969, OWL-4396, p. 200.112. A. Hsutzeeb and F. F. Dyer, G m a Spec-trrometric Stm&'%es of Fiss?:cnProdwts in the M%E3 OmL-3151 (August 1972).113. B. L. Manning, e. Maoaantov, "Rapid Sean Voltmetry and ~%ranopotentiometricStudies of Iron in Molten Fluorides ,Ii J. EZectroanaZ,Chem. - 7, 102-108 (1964)114. MSR Progrm Semiann. Progr. Regt. Aug. 31, 2971, OWL-4728, p. 75.l119. 3. P. Young and J. C. White, "A High-Temperature CeHb Assembly for~pectrsphotometric studies in Mo1tem Fluorides " Anal Chem. - 3% ,1892 (1959).120. J. P. Young, "Windowle~~ Spectrephotometric Cell for Use with CorrosiveLiquids," Anal. Chem. - 36, 390 (1964).121. !dSE Program Senriaiizn. Prop. Rept. Auy. 31, 2965, QRNL-3872, p. 145.124. 1GR Program Semiann. P~ogr. Rept. &Zy 32, 1964, QKXL-3708, p. 328.125. Bulletin MQ. 7012, Avcs Everett Research Laboratory, Everett, Mass.(197%).126. D. %. bknning, "Voltametry of Iron in Holten Lithium Fluoride-Potassium Fluoride-Sodium Fluoride ," J , BZectroanaZ. Chem. - 6 302(1944).127. D. L. Manning, "Voltammetry of Iron in Ms1ten Lithium FluoridesdimFh~dde-htassi~m(1964) eFitloride, '' J. BZectroanaZ. Chem. 7 , 3021128. G. Namantov and D. L. Planni-ng, "VoLtametry and Related Studies ofUranium in NoPten Lithium Fhuoride-Beryllium Fluoride-ZirconiumFluoride,'8 AnaZ. Chem. - 38, 1494 (1966).

......-.>

w. P. Eatherly....?.Y!Relationship Between Graphite and Core Design.....A d........., ...... ,:.:

xenon exclusion implies effective gas permeabilities of cm2/sechelium STB or less, roughly a pore entrance diameter requirement of0.01 p or less. Such pore textures are not attainable in the ordinaryfabrication of bulk graphite. The requirement can be easily met by pyrolyticdeposition of carbon onto a bulk graphite, although questions ofradiation stability again arise.The existence of a finite graphite lifetime forces the reactor designto %ow power densities or to periodic graphite removal, as discussedin Chapter 4. We shall concern ourselves here with the problem of graphitefabrication and its effect on design. The requiremmt for salt exelusion,and thereby a fine-grained graphite, also determines that thegeometric cr~ss-section sf the graphite prisms be kept small. This isnecessitated by the inability to co~tr01 microstructure to the desireddegree in forming and heat-treating large cross sections. Small sectionsalso have the advantage of minimizing thermal gradients in the graphitereactor operation and thus reducing the rate of radiation damage.Pat the several design studies for MSR's, the problems of graphite removaland prism geometry have been solved in several different ways.In the reference design 633 the care consists entirely of squareprisms approximately .four inches on a side. In the central core zone (Zonee> the reqUir@lllent 0% 13.2 VSlUrn@ percent salt leads to SeCtiOnS With SmEllEholes or interprism slots (see Fig. 6*l)* Salt volume is traded betweencentral h~le and slots to simplify orificing of the salt flow. For %heouter core region (Zone 11) the required salt volume of 37% is obtainedby opening up the central axial hole (Fig. 6.2). The entire core is supportedby Mastelloy tie rods and grid plate a d is periodically removedas a unit. Although the square cross-section permits easy fabricationof the base stock graphite, the possible need to pyrolytically treat theinterior surface of the hole presents a formidable fabrication problem.~tais pro~em was circumvented in the ~basco study ~41, which substituteda slab design for the square prisms (see Figs. 6.3 and 6-4)- Tkeslabs are assembled into hexagonal elements, each of which can be removedas a unit ts permit partial core replacement.A similar slab design was utilized in the study of a demonstrationreactor 151, except that RCJ prsvisi~n was made for graphite replacementdue to the lOW power sbeHlsity in the Core.Mechanical analysis [6f has indicated there are no significantthermal or radiation-induced stresses in any of the designs.c.s....General BackgroundGraphite has been employed as a nuclear material as lofag as therehave been reactors. It formed the moderator in the Stagg Field experimentand the prototype Oak ~idge I%X3phite Reactor. The first extensive ex-

I in. 1,A,.,3 698 in\ORkk-DWG 69-575OA0680 in /B --in-SECTION A-AZONE 1-4I -43021n 3698 in -30.302 I11/SECTION 6-6ZONE f-A141,......=T-. 11.340 in....*Fig. 6.1.refeKf2IIC@ df2Sigfi.$14 inSECTION 6-BZONE 1-8Graphite moderator element for Zone P sf the MSBR

c.3"71

.+.&.....:.a...,. .x.:s;.&$... .. ...m...... .,.,...d.4,*a..... ,&>,Fig. 6.3. Top detail of the gb-aphite moderator elementas proposed in the Ebasco Design Study.:;sa........l:.= .,..

186--.,.... c.. .QRNL-DWG 7s -131'35Fig. 6.4. Cross-section of the Zone 1 moderator elementas proposed in the Ebaseo Design Study.

181perience with graphite was acquired in the Hanford production reactorsand somewhat later in the British ga~-coolied reactors, albeit at lowertemperatures and fluences than apply to the MSR's.basre recently the Dragon Reactor initiated the use of graphite attemperatures in the lOQO"6 range, followed shortly by the Peach BottomReactor in the U.S. and the AVR Reactor in Germany. Currently, the gascooledreactors are beimg designed or operated to take graphite temperaturesin the 1200-l300"6 temperature range.Pauch of the experience and data obtained in the gas-cooled reactorprogram is directly applicable to MSR's. In particular, data taken atHaford in the mid-1968's spanning the temperature range 300-%100"6 firstindicated the finite lifetime of graphite subject to neutron-induceddamage, i.e., its eventual dimensional e~pansion and loss of mechanicalintegrity.In late 1968 a program was initiated at BRNL to evaluate graphites%OK molten-salt reactor application, and more specifically, to determinewhat limitations graphite might impose on reactor design. A program planwas proposed [7] to demonstrate feasibility of improving graphite by 1895,and to bring such improvements to commercial application by 1988. Theproposal was ambitious and has not been fully implemented due to fundinglimitations: Considerable progress has, nevertheless, been achievedtsward demonstrating the capability of existing materials to meet MSRrequirements and to delineate areas for future development.Current StatusThe MSR graphite program has evolved into a four-pronged attack -to survey existing commercially available graphites for their applicabilityto an MSBR; to gain sufficient insight into the damage mechanismto be able to estimate the degree of improvement to be expected in futuregraphites; to develop an in-house capability to fabricate graphites inorder to relate damage behavior to structure and fabrication technique;and to develop methods of sealing the graphite against xenon-135 diffusiom.Included in these areas are the necessity to develop design data andcost estimates .Irradiation Damage StudiesThe basic irradiation damage phenomena in graphite are determinedby the extreme anisotropy OS the crystal. The carbon atoms are arrayedin tightly-bound hexagons in planar array. The planes are well-separateda d weakly coupled. Pnterstitials produced by weiutron bOmbaPdment wanderfreely between planes and reintegrate as new planes. Vacancies Left behindare collapsed. We are thus left with the picture of a single crystalexpanding indefinitely in one direction and contracting in the other twowith little change in net volume. an a polycrystalline material, eachcrystallite is thus expanding and contracting in varying directions, andit is hardly surprising the material eventually deteriorates. What is

182remarkable is its ability to withstand these changes Partially orientedpyrslytics irradiated in BPIR to fluences of 3 x 102’ neutrons/cm% at MSRtemperatures expanded ~BOX in the preferred e-axis direction without losingneckanical integrity!During the period 1963-1971, over eighty different experimental andcomercially available graphites were irradiated in HFIR t~ establishtheir dimensional behavior 683. A picture of the mi.crsstructura% propificantto radiation damage has gradually emerged. Our coneas follows.08’C temperature range of interest to molten-salt reactors,bulk graphites can be classified into three behavior modes dependingon their fabrication historys namely c~nventi~nal materials,black-baaed materials, and monolithic materials 181. By conventionalmaterials we inc$ude a11 normal commercial graphites formed from calcinedcoke 01- graphite fillers bindered with thermosetting or thermoplasticmaterials and subsequently heat trea,ted. These materials way be isotropicor aIIiSQt.rQpiC, but show an immediate volume contraction under damagefollowed by rapid and catastrophic expansion. Their lifetimes are in therange fHQlrrr 1 2.5 X BO2’ n@UtrOnS/cI?l2 (a 566 kev) (see Pig. 6.5). meexpansion is characteristically parabolic with fluence.The second class black-based graphites, employs carbon blacks asfillers, the individual blacks having a rou ly spherical crystalliteorientation capable of wiehs tanding high ta entia% strain. Dependingon heat treatment temperature, they may contract rapidly at first anelexpand linearly with further irradiation (see Fig. 6.6) a The differingexpansion behavior f HOW the more conventional graphites is tentativelyexplained by the ability of the black particles to withstand strain.The third class, and the one of interest to us, is the monolithicmaterials which appear to be binderless, or for which the filler materialis chemically active and reacts with the binder. The result is an extremelyhomogeneous structure usually unmarked by microstructures. Underirradiation these mateska1s undergo a prolonged stable induction periodhefore breaking into parabolic expansion (see Pig. 6.7). The lifetimesof those tested to date lie in the 2 to 3 x 102’ ~e~tron~/c~~? range(E > 50 key) The best terials of this type are invariably isotropic,and their induction period is attributed to their microstrength andability to flow plastically to relieve strain. The parabolic expansionof both these materials and the conventional graphites can be related tovoid generati~n as the structure finally fractures at the crystallitelevel E91Ira sumary, we feel the general nature of damage in polycrystallinegraphite is understood, and its relationship to microstructure at leastqualitatively demonstrated...... .*;eBased on the RPIR irradiation data, a program was initiated in 1990to explore the fabrication of monolithic graphites specifically aimed atradiation damage resistance. The program is rather modest in Both objectiveand scope. Small samples up to three inches in diameter only are

.:;u. .I -....*.:.L,...... .. . !....i.:.S'588 ~..... ,:is.>SM-I.........,. .-.... . '.:

I I I I IPig. 6.6. Length changes for various black-based graphites andPOCO-AXF irradiated at 7115°C.

. ..-.----Fig. 6.7. Volume changes for monolithic graphites irradiated at715°C.

,.........Xenon ControlThe problem of effectively sealing the graphite against xenon cantake three forms: direct P~~p~-egnati~n by hydrocarbons followed by heattreatment to leave a carbon relic in the pore; impregnation With a liquidOIP Solid Salt $0 fill the pores; Or SbaKfaCe treatment to Seal Qff thepores at or R~BP the surface.Sufficient experience has been enerated in the graphite industryto indicate the limitations of direc carbonaceous impregnation. ?Tlaedecomposing hydrocarbon generates gaseous products which muse escape tothe surface or rupture the bulk iece. A practical limitation in persueabilityis of the order of 10- cm2/see, a fact~r 100 larger thanrequired for excluding xenon. This approach has therefore not been exploredin our program.The use of salts to fill the pores has been looked at cu~sorikgr [8]and is still being investigated. 'Pke limitation anticipated is the diffusionof uranium into the salt. Nevertheless, it has been demonstratedthat suck a technique can reduce gas permeabilities bo the desired range.The use of pyrolytic decomposition of hydrocarbons has been extensivelystudied for reactor applications primarily under the ga9-cooledreactor program a The background developed there on pr~cess parametersproperties and irradiation behavior formed the basis for our program.The first approach Ell] was to utilize a gas impregmation process tofill the pores near the surface, this being preferred QVW a coatingprocess because sf its greater resistance to handling damage, The processsists of alternately pulsing hydrocarbon gas and vacuum, thus effectadecomposition of the gas deep within the pores of the graphite, andhas proven to be easily controllable a d effective. Apparent permeabil-%ties* in the 10-8 cm2/see rage and below are readily attalneci. MO~Olithic-typesubstrates were impregxated and irradiated in BFIB, but the=;. .y". . .,.!&* The permeabilities as used here are derived f~om gas flow measureboththe sealed region and the remaining unaffected substrategraphite. The actual permeabilities at the sealed s~~face layerare probably a factor of 180 lower.. .... u.s

187.....

188.....tu.s,Fig. 6.8. Scannin electron microscope pictures of coatedgraphite. Left: Thin coating insufficient to cover surface.Right: Soot inclusion with radiating crack structure.

189OWML-DWG 653-5538500~-..... ...., _:::I900.... ,:.:.x..... z.m.;.;.y,+=......v)Mw erW2a=l v)200.;&....,. . 1.51100.....:.w0-60O 1 2 3 4TIME AT 80% PLANT FACTOR (years)

Manufacturing Capability and CostsThe extensive survey program of various commercial graphites describedabove netted one graphite that is acceptable to the MSBR referencedesign, a second acceptable but limited in available sizes, and two othergrades which are p~tentially acceptable. These span, happily, four inde-pendent vendors eGreat Lakes Carbon Corporation grade H-364 is available in the propergeometries and possesses a lifetime before significant expansion occurs ofthe O P ~ ~ of P 2.5 x 10%' neutrsns/cts2 (E 9 50 ke'$i) at 7 ~ 5 ' ~ mis . is 19%less than specified in the reference desi but is close enough $0 becornpensatable by allowing more expansion reducing the maximum powerdensity, or replacing the graphite somewhat sooner. POCO grade M F isthe best comercial material we have encountered with a lifetime of theOrden^ Of 3 to 3.5 X lo"%, but is Currently available only in Shortlengths. Material submitted by Airco Speer and Pure Carbon Companiesalso may fall into the class of these materials but to date have onlybeen irradiated to 1.5 x 18"'. Their behavior appears to be similar tothe best of the monolithic grades.We have been unable to obtain fi~m price estimates on these gradesfor QU~C application, but they appear to fall into the range of $5-BB.00per pound of finished graphite, evem on a first-order basis. Prices of$5.06 per pound or lower ~e~tainly appear to be probable if the marketbecewes suffi~iently large to per~t- the graphite to be handled as a stockitern.tie coating 01- sealing, there is an existingsuch coatings prinarily for aerospace applications.However, the most irradiatian-resistant type of coating (%'%I%, orlow temperature isotropic) employs process parameters quite different.from the industrial-developed processes a Cost-es timting is difficultsince a process has not been developed, but we guess that $25-30 per %bof finished graphite in the slab geometry should certainly cover anyreasonable process during its early development. EventualPy costs shoulddr~p to $5-7.00 per kb for reasonable production quantities..... .*x.>$.A*........ I.....uncertainties and Further WorkA number of uncertainties remain, none of which particularly affectthe viability of the molten-salt breeder concept- These uncertaintiesmay effect further compromises in desi and perhaps economics, but inno case are they vital to the technical OX economic use of graphite inthe reactor. Since these uncertainties are readily resolvable by furtherwork on graphite, we discuss the two topics together.

191Irradiation Damage and Graphite Fabrication.:.=. .....*:.;.x.......... YllclThe potential for improved graphites that can be employed in MSBR'sbeyond a fluenee of 3.5 x lo22 is, we believe, good. %he question remainsspeculative as to the degree of improvement to be anticipated. Our confidencein anticipating at least incremental improvements has increasedsignificantly in the past year as our mderstanding of the relationshipsbetween microstructure and radiation damage has improved. The availabilityof HFIR (or fast reactors in the future) to obtain full fluencesin less than a year enables a graphite development program to proceed ata reasonable rate.The fabrication prscess whish is c~rrentHy being studied at ORNE hasonly yet been briefly explored. Areas awaiting examination where frartkgrprogress can be anticipated are the use of blacks in ~ K O C ~ SanalogousS ~ ~to the green-coke route, and the use of high-pressure processing for bothraw materials and carbonization. To date very little effort has beenexpended by industry and other government laboratories to increase thelife or stability of graphite, and the development of new processes anddiagnostic techniques suggest much more rapid progress can be made. Webelieve these alternate fabrication techniques and their relevance todamage resistance can be at least indicated within two to three years,The fact that pyrolytic materials have survived to fluences greater than1.5 x neutrons/~~~? at 1258'6: implies lifetimes of the order of5 x 1022 neutrons/cm2 at ~QO"C should be attainable.Xenon ControlThe ability to exclude xenon from she graphite by means of pyrolyticlempregnation or coating has not been demonstrated. Alternative techniquesexist but currently remain unexplored.Both coatings and substrates have separately been shown to surviveto >3 x Hence, this is strong ~:~BSQII to believe a m~110layer coatingcan be made to work in the MSBR. However, in the gas-cooled program wherecoatings have been shown to survive equivalent fluences, it has been foundnecessary to decouple the substrate and coating. Both low density p y ~ ~ -lytic and silicon carbide intercoatings have been employed, and analogoustechniques can be utilized he~e albeit at an ecsnomlc penalty. If suchtechniques are required, some two to three years' effort will be neededto develop them. In any event further work to upscale and prepare fortransition to comercial suppliers will be required.Two further techniques are the filling of porosity in the graphitewith either liquid or solid salt. Solid salt, and bismuth as astand-in, have been shown to at least yield the right order of gaseousdiffusion rates. The uncertainty remains as to whether the liquid orsolid diffusion rates are sufficiently low. The potential of this app~oachcan be determined in about one year.Underlying all of this are questions as to the efficacy of thehelium bubbles and the impedance of xenon diffusion across thesalt interface to limit di€fusion into graphite quite independently ofthe permeability, a subject wh'ich is discussed in detail in Chapter 8.

192me theml conductivity degradation with damage remilas to be established.This leaves uncertain the maximum cross-sectional area of thegraphite prisms permitted. Although representing only a question ofdesign, it must be answered. We estimate two to three years will be requiredto obtain the necessary irradiation history and data.,., _.\Stresses and CreepPresent 'benowledge on creep in graphite leads to am analysis indicatingsnly trivial stresses are developed im the graphite. Recent unpublishedresults obtained at Hanford m y indicate the ability ofto creep deteriorates at high fluences. If so, th@ problem can again beavoided by design, but the situation is uncertain. We estimate a minimumof three to five years to obtain the necessary information......Y.WG. ... :.EvaluationIn general9 g~aph2ite presents no serious problem to the molten-saltreactor. At least one vendor has available a satisfactory material inthe required sizes, and there is every reason to believe the lifetime offuture graphites can be incrementally exten ed to perhaps twice that ofthe best existing materials. Cost estimates utilizing present materialcapabilities yield 8.17 mill/kWnr, for replacement sf the entire core ona four-year cycle in the refe~etnce design, or a somewhat lower cost ifonly the most hi hly irradiated m.ate.IPk3l fs KC?plZLC@d E%ZiCh time. Inc?ZeaSingthe lifetime to eight years reduces the fuel egpeabe cost by about0 l mIll/kWhr but beyond this the cost savin s are small. We can thusstate that existin base graphites are accept ble, and future graphicescam probably be developed to reduce replacement costs. 'The present methods of pyrolytically coating the graphite appear tobe satisfactory, although ~adiatisn testing will require at least anotheryear to demonstrate this. Alternative routes exist, but the question iseventually one of ecsaamics - the value of narginal increases in breedinggain versus the cost of the coated material.We have not yet acquired sufficient data on the thermal and mechanicalproperties of graphites of the type used in the mBR. Enough ishorn to be certain these do not affect the exact shapes of graphite perwittedand the mems by which the graphite is supported, but additionaldata must yet be secured for design purposes.Ill SUlll, there are no PB;a%OlaS to expect graphite to limit the OVe~Edlfeasibility of molten-sdt breeder reactors, and acceptable materials areavailable today r....

193References for chapter 6P. J. M. W. Simmons, Bgdiation Damage in G~qhite, Pergamon Press (1965).2. 6. B. Engle and W. P Eatherly, ''4 Review of High-Temperature GraphiteIrradiation Behavior" in High Temperature-'&& Pressures (to bepublished).3. Conceptual Uesi_p Sku& of a Single Fluid Molten-Salt Eeactor, OWNE-4528 (l970).4. 's1060 m(e) Molten-Salt Breeder Reactor Conceptual Design Study,"Final Report - Task I., Ebascrs Services, Inc. (1971).5. E. S. Bettis, E. G. Alexander, and H. L. Watts, Des

I... u.2

....i......, "-. ..6L.a_._ .....?7. MATERIALS FOR SALT-CONTAINING VESSELS AND PIPINGH. E. McCoyMaterial Requirements.s . ..,.?....;il%The metal used in fabricating a molten salt reactor will be exposedto several environments. The inside of the primary circuit will be exposedto %~P-B~F~-T~FL+-UF~~ the coolant circuit to NaP-NaBF4, and the steamcircuit to supercritical steam. Thus, the tubes in the intermediate heatexchanger will be exposed to bot11 salts and those in the steam generatorto both coolant salt and steam. The bulk fuel salt temperature will rangefrom 1650" to 1306°F and the coolant from 850 to El50"P. The steam inthe salt-heated stem generator enters at 780°F in the reference loopsand is heated to 1000'F. The outsides of the metal components will beexposed tu containment cell environments composed primarily of nitrogen,with enough in-leakage of air to make it oxidizing.The most basic requfrement of the structural materials is that theybe chemically compatible with these various environments. The chemicalprsperties of the salts were discussed in Chapter 5, where it was pointedout that the 5elective leaching of chrsmium would be the primary mechanismfor co~~osiota of iron- and nickel-base alloys by molten fluorides. Thusthe concentration of chromium is an important consideration in selectingan alloy to be used in molten-salt circuits. Iron is more easily oxidizedby the salts than nickel, so the preference (although not necessarilya requirement) of a nickel-base alloy over an iron-base alloy is immediatelyobvious. Good resistance to oxidation in N2-02 enviruramewt:~ isfavored by high chromium concentrations.The material. requirements for steam generators are discussed inChapter 8. Satisfying the need for compatibility with the coolant saltand resistance to stre~s-c~rr~si~n cracking in the steam be difficdtwith a single alloy suggesting the use of duplex tubes.The subject: of design stresses will be discussed more fully i~Chapter 13, but it is obvious that the material must be capable of withstandingwithout %ailure the stresses that will be imposed during service.The relatively hFgh temperatures involved will require that suitableelastic-plastic analyses be made of all structures. An MSBR will sperateat: relatively low pressuress so a high-strength material does not seemnecessary. However, thermal stresses will likely control the designand it is quite likely that a material with moderate strength will beused *The ~KIIIE~~Y circuit, pa~ti~ula~ly the reactor vessel, will be exposedto neutron irradiation, but no metallic structural members will be in thehighest neutron flux regions. At the vessel wall, the peak thermal andfast (>6.8 MeV) fluxes will be 6.5 x and 1.2 x 18" neutr~ns-cm-~~sec-~,and over .a 30-year lifetime with an 86% load factor, the peak thermal andfast fluelaces will be 5 x .and 1 x lO2O weutr9ns/cm2, respectively.These relatively low fluence levels are due to the effects sf the graphite.:=195

e~~ector ma souid be redtlced even further if necessary. me fatfkuence is not high ensugh for void formti~n to be a problem, and theirradiation damage is primarily the high-temperature embrittlement dueto helium generation 111.obvious requirement of the structural material is that it mustbe fabricable into the form needed to build m en ineerlng sys tern.Basic shapes required include plate, piping, tubin 3 md forgings- POKassembly, the ~~~terial be ~eldable both der ~~ll-c~ntrolled shopconditions and in the field. Many iron- and nickel-base alloys satisfythese requirements, dthotagla the technology is more advanced foralloys than for others. e basic ability to work with a material ismore important, however, an having a currently viable technologyin several fabrication shops. The Patter factor would simply nake thefirst unit cheaper.ry9 one mst keep sight of the basic requirements thatthe material be CB atible with its environments, have acceptablemechanical properties, both ~nir~adiated and after exposure to themaximum expected neutron fluence, and be capable of being fabricatedwith reasonable ease.Backgssmd

Table 7.1. Clmeqical composition of Hastelloy NElementContent (% by weight)'"Standard Favored modifiedalloyalloyNickel Base BaseMolybdenum 15- 18 11-13Chromium 6-8 6-8Iron 5 O.€bManganese 1 0.15 __ 0.25Silicon 1 0.1Pho sphurus 0.015 0.01Sulfur 0.020 0.01Boron 0.01 0.001Titanium and hafnium 2Niobium 0-2....

19is important because it forms carbides that restrict grain gr~wth duringhigh-temperature treatments and improve the strength. Elements such assulfur, phc~sphorus, and boron, and many others not included in Table 9.1,are tramp or impurity elements that serve no known useful purpose in thealloy D These elements enerally have little effect on the alloy behavioras long as they are kep at reasonable concentrations.Silicon is introduced by the refractories used in the air meltingpractice and is an important element. kste%loy %a c~ntaining 0.5 to 1%silicon contains stringers of coarse carbides and will form some finecarbides during annealing at 1200 to 1606°F [s] e These carbides are 0%the MgC type, with M having the composition of 27.9% Ni, 3.3X Si, 0.6% Fe,56.1% No, and 4BX 6r. They are not easily diss~lved during annealing,so the alloy has stable properties over a broad range of operating temperatures,Several meltin practices are currently in use that result inlow silicon concentrations. e carbides in these alloys are of the N2Ctypeo Where %% is 80 to 98% mokYbden%yqp with the reminder ChrQmim. ‘%heyare more easily dissolved than the MsG type that contain silicon.%US, &Std%ogP M is basiCa%ly an alloy strengthened With DlolybdenUmand containing enough C~KOUI~UB for m~derate oxidation resistance. Thecarbide type is controlled by the silicon ~~n~entrati~n.COITQS~U~ Resistance of HastePloy NSeveral hundred thousand hours of corrosion experience with Bastel-Isy N and fluoride salts have been obtained in thermal convection [2,3]and pumped systems 681. As discussed in Chapter 5, these experimentssh~wed that the predominant corrosion mechanism in clean fluoride saltscontaining uranium was the selective leashing of chromium. Only 7% ofthe alloy is chromium and this must diffuse to the surface before itcan be re~~~ved by the salt. DeVan measured the rate of chromium diffusionin ~asteHloy N 691, and the measured diffusion coefficients were usedtu estimate the chrolaim profile after 30-year service at 650 and 7’00°Cin salt oxidizing enough to csmpPete%y deplete the surface of Cr. Aseven in this extreme situation, the depth of removalThe early work with kstell~y K and other alloys revealed the importameof c~~-~troPlin impurities in the salt. rarity flwrides such asFeF2, HoP2, and NFP will ~eact with Cr to for rP2, a more stable fluoride.Water will react with the fluoride mixtures to form MF that willform fluorides with all the structural metals. Such impurities led torelatively high corrCl%iQn rat@% Of even HCistePlCIy N the early experiments,However, we learned how to prepare pure salt mixtures, md verylow corrosiom rates were obtained. The ultimate proof of this abilitywas the operation of the NSRE where the overall COZ-I-YAS~~~ was held lowfor almost four years at temperature.me preceding diseussion ap ied to the removal of material, butdeposition is also of coneern. the salt circulates from hotter tocooler regions, the solubilities of the corrosion products in the saltdecrease, and if concentrations are high enou In, material way be deposited.This pr~~ess is complex, depending upon chemical driving forces and factors.... Vl,$j

........i.._,, .199....+:.sU.d1.0ORNL-BWG 72-85260.80.6G30.40.200 2 4 6 8 f0 12DISTANCE FROM SURFACE ( mi Is)Fig. 7.1. Calculated contentration profiles for Cr removal andTe enrichment based on measured diffusion coefficients. The timesused in the calculatisns were 3 yrs for the MSRE and 30 yrs for theMSBR.

such as geometry and flow conditions. Portunatdy, the salts being consideredf ~ an r MSBR tend strongly to deposit mterial rather uniformlythroughout the c~ld region and have a minimal tendency to plug heatgertubes and salt passages in cooler parts of the system.The physical and mechanical properties of Hastelloy K were evaluatedrather extensiuely before the MSRE was constructed. These properties hauebeen sL.mlmarized previously [a 1. ?%e Strength sf this allQgP iS quite goodlbecause of the 16% mslybdenum. The property changes with time are sw;lP%since the alloy does not form intermetallic compounds but only smallZinQUlItS Of fine carbides.Fabrfcation. - Although the power level of the MStaE was small, thesystem was complex and required the ability to carry out all of the basicfabrication steps [ll]. Eany thousands of psunds of basic product formswere procured from three vendors. Some of the components were built bycsmercial vendors, but most of the fabrication was done in the AEClnlonCarbide shops at Oak Ridge and Paducah. Welding, brazing, and inspectionprocedures fer the reacts% developed. one ofthe final steps was to make use of the heaters an the vessel to postweldanneal the final vessel closure weld.Operation. - As discussed in Chapter 2, the react~roperated verysatisfactorily. The primary system was ihb~ve 5OO'C for 30,807 hoursand filled with fuel salt for 21,046 hours. The only failure invoEvingHastelloy M was through-wall cracking of a freeze value c~irtcident withfinal shutdown of the system 6121 e This failure was due to fatigue fromdifferential. thermal expansion in a part that was constructed too riC ~ X T C S ~ Q - ~ . C ~rrosi~n clurirtg the operation of the PISRE was followedboth by analyzing the salts and examining surveillance specimens removedfrom the COBTe. %e pt-iWrgr COlPt-oSioR prsduct, crF2, KeIlItiined below itssolubility limits in the salt; so its ~ ~~~entration could be used as ameasure of the amount of chromium being removed from the metal. Theresults of such analyses have been described in detail, (reference E3and Chap. 5) A simple summary is that the chrsrnim removal was very11: the total amount accumulated in the fuel salt was equivalentts that which would be rem~ved from all metal surfaces to a depth ~fQ,4 mil; the E1ElQra6nt appearing in the cOO~ZXl% Sal% Was practically nil.Surveillance samples located in the cure of the MSRE were periapicallyremoved for examination and testing. Samples of both standardand modified HastePEsy N were always in excellent physical conditionwith only a slight amount 5f discoloration [1,14,l%,16], and visual

281....:.:.y,j....;!w.....,:

as that shown in Fig. 9-2 that had been strained in the hot cell. Cracksf0Wd after StKEtiKh material that had been exposed in the core were nomore pervasive or deeper than those in the heat-exchanger tubes, whichhad heen exposed %Q insignifi@ant laeUtr5n flux. By CQntleolled dissolutionof several samples, a n of fission p~od~ncts were found within thematerial to a depth sf several mils. '$$lis suggested that the crackinparticularly tellga~iurn, which was found at the ighes t concentrat ion -and set off the investigation which is describ= .....Current State Of Materials DevelopmentHastelEoy N (both standard and modified c~~~~positions) has been shownother in-pile tests, and large number of out-of-pile loopsto have excellent corrosion resistance in salts containinmF4, illEd UF4 [2,3,8,31]. This extensive experience confirm the behaviorthat would be predicted from calculations such as those plotted in Fig,7.1.Corrssion studies with the proposed coolant salt, sodium fluorsborate,have been more limited [32,33,45]. However, we have operatedfour therm1 convection loops and two pumped system in the materialsprogrm for tQtal test: time Of about l%B,QBO h.% (See chap. 8) e Thisexperience reveals that: the flusroborate salt absorbs moisture quitereadily, W i t h atteIldaIlt generakized COrrQsiQn. &I OCCZSiOIX When leaksd~el~ped, the corrssion rate has imcreased and then decreased as theimpurities were exhausted. During these periods of high corrosion, allCOmpon@-6k$s of the allogr Were remOVed UnifO9d.gi frQm the hot leg a d depositedin the soEd leg. Crystals of Ka3CrF6 a d NagFeFG deposited inthe cold regions as their solubilities were exceeded. Neverthelesspunped loops in which the salt is heat d ana coolea between 1270 a d 7 9 5 ~have been operated for several thousan s of hours with corrosion ratesof 4l.l mpy, so we believe that satisf CtOry pPQCE?dUPes for Using thesalt can be developeme ssrrosion b avior of several other nickel-base alloys was investigatedin screening tests in the aircraft. propulsion p ~ ~ g ~ [2,3] a m -The proposed service temperature was l508'F and most of these alloys werenot considered further because in tests at that temperature large amountsQf ChrorniUEL Were remaVed, With fOl3ltl.tion Sf Voids in the hot regions Of&!§e: loops afld d@pOSitiQn of ChKUIIIiUIll Crystals in the Cold IfegionS. b-csnel $80 received the most study of arn~ alloy besides Hastebloy N, andthe evaluati~n p~~9grtl.111 on it involved several thermal ~~~~vecti~nloopsand 4 forced convection loops that operated for a total of 79,368 hr [8].Although the CCXX-CX~QTI resistance of PTI~OR~%600 was not as good as thatof Bastelloy N, at temperatures in the range of MSBWfs the rates weres~metirne~ 1o-w enough to be of interest. For example, one Inesnek 600lusp operated at a peak temperature of 125Q'P for 8801 Ear with intergranularpenetrations of 1.5 mi This penetration is high by our currentstandards, but was only sli %ly high@K than that ObSerVed for ksteHl5gT N. ..&%

..., .....:.:,:.:.:iFig. 7.2. Hastelloy N samples strained to failure at room temperature.The MSRE rod thimble was at high temperature for 31,000 hr, withthe Power surface exposed to N2-Q2 gas. This surface was oxidized, andthe cracks only pernetrate the oxide. The upper surface of the thimblesample was exposed to fuel salt for 21,000 hr, during which time the Teconcentration built up. The Power sample was vapor plated with enoughtellurium to produce a concentration of 8.1% in the oute~ 5 mils and annealed1000 hK at 650°C.

204tested in salt of comparab%e pu~ity. Thus, it is likely that m alloycontaining 15% ch~~~ium(e,g,, Inconel 600) would have acceptable corrosionresistance at 12(30°F 8% less.The compatibility of iron-base alloys with fluoride salts has received~elatively little attention because the thermodynamic data indicate thatnickel-base alloys with a minimum chromium content will be mre corrosionresistant than iron-base alloys. %e initial screening tests on types 300and 400 stainless steels indicated that these alloys were unsatisfactolPy[2,3]. Eowever, one type 384L stainless steel thermal convection loopcontaining a fuel salt has been in operation for over 9 years without pluging[%Sa; the co~=rosion rate at the peak te eratulpe of 1278°F is about.% Dpy. It iS quite likely that the CQFrQS n rate could be ~ed~ced toan acceptab%e value by decreasing the temperature to 1280"F, but tests athigher velocities would be required to p~ovide more conclusive informtionA nost k€lpOHti%Ilt COnsfdePation in %he Suitabilfty Qf iTXXI-baSe alloySis the possibility that, throu some misoperation, the salt could becomeoxidizing enough to corrode this process would not. be diffusionCQntrOlk?d, and thus large EUWUIl erial c~uld be transferred quicklyfrom the hotter re ions to the cooler regions of the system. While thisis also true for n ckel, it can occur with iron at less severe conditions.l~mepTer~ we believe that the oxidation state Can be eontroiiea closelyu..atibility of sodium fluorebarate with ir~n-base alloys isunknown. Corrosion in this salt is ont trolled primarily by impurities Ebut iron seems to be attacked as readily as chromium. If the salt couldbe maintained very pure, iron-base EaPloys might be acceptable.Irradiation Ernbr it tlementThe peak fast fluenee at the inside surface of the reactor vesselwill be of the order ofE~~K~IIS/CII?, which is too low to causedetectable swellinpeak themad fluen e of 5 x 1021 neutrens/era2 is reat enough to producesignificant axmaants of helFm, about 5 pprn from r alduaa 1% and possiblyansther 100 to 200 pm by transmutations involvin nickel that have onlyrecently been reco ized ks occur [SS] 0 In stand rd Hastelloy N thehelium would reduc rain boundary cohesion md increase the tendencyrain boundary fracture, with the result that the fracture stains attea temperatures become quite HOW.Our approach to combating &is embrittle nt problem is to add dementssuck as titanium, hafnium, zirc8niuI11, and n obiurn that promote theformation of finely dispersed MC type carbides [3 1. These carbides producenUIX63rOUB intE2KfaCes that $rap the heliLEFi %at er than allowing it to8 the grain boundaries. Typical compositions sf modifiediven in Table 7.1.all of the carbide-fomin elements are beneficial in improving the%KaCtUre StPaiIl, but there Zi%e several practical reasons why titanium andniobium are more desirable. Zirconium has been f o ~ to d cause weld metalin concentrations as s.65W E391 and for this reason,a very undesirable alloy ition. Hafnfm causes weld metalcracking at conce~~t~ation~ sf about O.7%, but the greatest problem withc.*....

285......:.=.:,......:.:.a....''i.2 .?. ,...... i......'.:

206As already noted, ‘intergranular cracking was observed in the surveillancesamples and several components from the MSRE [l7]. me mostsignif icant characteris tics of the cracks are :1.Cracks were formed on dl surfaces exposed to fuel salt.2.3.4.Some cracks were visible in polished sections from some components(particularly the heat exchanger) whem they were removed fromservice, but deformation at ient %eVeKatU%e W a reqUf.h-ed tQmke moat show up.The material removed from the PISRE had been heated and exposed tofission products %IX times ranging from 2500 to 25,080 hr. Althoughthe fKequWlCy Of Cracks ita@lpeaS@d W i t h time, the ximum depth didnot increase detectably eWe have not been able to produce similar intergranular cracks byCO~IXXS~QR. To determine if cob-rosion c~uld be the cause, the salt inone fuei-sdt loop was made quite oxiafzing adding F ~ P ~ selective .ELI-IUI~~ attack occurred, but the attack was very shallow and theea grain bo~~~dariesdid not open further during straining. ‘$hed most convinein evidence that co~ros ion (chrsmfm depletion)e directly from examination of HSRE samples.caepietion coeb~d not he detected in sawpies from theand in a section of the control rod thimble that was under? these samples t~ere cracked as severely as those (e.g.the bare control rod thimble) in which chromium depletion was detectable.unlikely that chr~mium depletion alone can account for the.... *:f”ible mechanism to be considered is that OW or severalelements diffused into the material preferentially along thebo~adaries and degraded them in some way- The prucess responthe cracking could be (1) the formation sf a compound that is verybrittle, (2) fsraation of low-melting phases along the grain boundariesthat become liquid at operating temperature, 0% (3) a change in compositiomalong the grain boundaries so that they are still solFd but veryWeak. ‘61a@ firs% and third FWchank3mS Wodd require SQEE defOKHlathn toform the cracks, but the second mechanism wodd ~lst require strain, andsamples could have cracks present before pos tsperatian defoClearly, it is extremely important that the elements responsiblefor the cracking be identified and that the HngchaniSn? be determined. halyticaldata from materials from the MSRE show that all sf the fissi~nproduets with sufficient half-lives to be detectable after two years were

.....'.:.:

ElementTnb,le 7.2. Fbssible effects of several elements on the cracking of kiastelloy PFree energyCracking of Effect on Effect on Effect on Effect onExpectedMelting Concentrated of fmxlatianpoint in cracks invapar and tensile CrCep tensile creeplocationof fluorideelectroplated properties propertics of properties of properties of0E(“Cl MSRE sarnpresat 1000°Kspccimensb of nickeF nickel alloyd Ihstelloy Np’ Hastelloy h”b &(..I mole-l yY -1 )e elem%tOVkZdlratingSulfur 119Selenium 217Tellurium 450Ai%fliC 811Antimony 630Tin 232%inc 420Cadmium 321Ruthenium 2500Technetium 2130Niobium 2468Zirconium 1852M0lybd~Wl-f 2410strontium76XCesium 29cerium 804Rhodium 1966-+t-t . ..-+--tInsolubleInsolublet++.-f-Insolublet-+tt-tt*ft34 Deposited-27 Deposited-39 Deposited-62 Deposited-55 Deposited-60 Deposited-68 J4-64 g-51 Deposited-46 Deposited--70 g-99 Salt-57 Deposited125 Salt-106 Salt-120 Salt-42 Depositedtt+-tttt+ttttttizaThe symbols used in this table should be interpreted in the following way: A plus refers to nondetrimental behavior, and a minus indicates detrimental effects. Twominuses indicate particularly bad effects.hWesults of current rrJtiruch.cC. 6. Bieber and R. F. Decker, “‘The Melting of Malleable Nickel and Nickel Alloys,” Trans. AI&E 221, 629 (1961).%. R. Wood and W. M. Cook, “Effects of Trace Contents of Impurity Elements cm the Creep-Rupture Properties of Nickel-Base Elements,“” AWetullurgia (BOO, 109 (1963).ePrivate communication, W. R. Grimes, QRNL.hay appear as H2S if HF concentration of melt is appreciable.Way appeu in salt if salt mixture is sufficiently oxidizing.

289. :.d. .,....,. . .,.... .;i.;,....,:.:.y!..,&X*.....,A,. ............&A %The diffusion rate sf tellurium into Hastelloy N was measured.Samples with 127~e deposited on the surface were annealed for 3000 hrat 650 and 760°C. At 650°C the penetration was so shallow that thelapping technique used did not give very reliable values, but accurateresults were obtained at ?60aC and the diffusion coefficient in the bulkmaterial was 1.01 x cm2/sec, about equivalent to that of chromiumat 650°C. me penetration profiles also were used to obtain the productof the grain boundary width and the grain boundary diffusion coefficient,and the measured quantities were then used with the Fisher mode1 [45] tocompute the grain boundary penetration. Ar; shown in Fig. 7.1, the M~Ximumpenetration of tellurium in an blSBR at 760°C would be 8 mils in 30years. The less accurate experimental values obtained at 650°C were usedto estimate that tellurium should have penetrated the grain boundariesin the MSRE to a depth of 2 to 3 mils, and the penetration of an MSBRoperating at 650°C for 30 yr should be about 4 mils. The relatively lowsensitivity of the penetration depth to the time is due to the variationwith time to the one-fourth power for grain boundary diffusion comparedwith the one-half power for bulk diffusion.These computed depths of penetration are quite acceptable, but severalfactors can move the curves. &e factor that could reduce the penetrationis that the supply of tellurium to the Hastelloy N would controlthe rate of penetration rather than diffusion through the metal. This isquite possible, since the concentration of tellurium in the salt wouldbe very low, and other fission products such as mlybdenum would depositand possibly interfere with tellurium actually reaching the Hastelloy N.At least one factor could increase the penetration. Cracks SOU^^ formand the diffusion front move inward, such as probably occurred in thespecimen that was exposed to tellurium while being stressed in a creepmachine. Another factor could move the profiles either way; compoundssuch as nickel-tellurides may form along the grain boundaries, and thetellurium may diffuse at higher or lower rates through these compoundsthan through Hastelloy N. Thus, although the diffusion measurementsprovide an explanation of the limited penetration in the MSRE and offersome encouragement that the depth of penetration of tellurium would notbe very great in an MSBR, they cannot be taken quantitatively.Little 2s known about the chemistry of tellurium, but most likelyit is similar to that of sulfur. The basic problem with nickel alloyscontaining sulfur is due to a low melting nickel-sulfur eutectic thatforms when sulfur segregates in the grain boundaries and causes theseregions to be weak compared with the matrix. Alloying additions such aschromium raise the melting point of the eutectic and reduce the magnitudeof the problem. Some proprietary work on superalloys shows that about16X chromium is required to make a superalloy resist embrittkement bysulfur. Assuming parallel behavior of telbkurium, its deleterious effectson Hastellsy N might be offset by the addition of chromium.Additions of tellurium, selenium, and sulfur are often made to steelsto obtain improved machinability but they cause embrittlement at hightemperatures. Small cerium additions have been effective in reducing theembrittlement. Thus cerium additions to Hastelloy N may also be effectivein making the tellurium innocuous.

210Fortunately, tellurium probably behaves in non-fissioning melts muchas it does in a fissioning salt, so that laboratory experiments can beused to answer many questions. Assuming this is so, over sixty alloyswere electroplated with telluri~m and annealed for long periods of time toate the effects Sf COSIlpQSition, ibtC]lkllding higher ChlrOmiklnm COnceIltrationOR the cracking phenomenon. Included were several nickel-basealloys, representative alloys of types 200, 300, 400, and 500 stainlesssteel, nickel, copper9 iron, Moanel, two cobalt-base alloys, and severalheats of modified and standard Hastelloy N. After being annealed, thesamples were strained at room temperature and sectioned for metallographicexamination.No crack formed in iron, csppe~, M~~bel, the stainless steels, orthe n8Ckel-baS@ a%lQyPs COntaifliIlg more than 15% ChrsmiW. However,cracks did form in stellsy B (1X eh~o~~~iurn maximum), Hastelloy W (5%C~~QEI~UEI) and in most heats of Wastelloy N (7% chromium). Some of theheats of modified Hastelloy N had better resistance to cracking thanstandard Mastelloy N. These alloys contained several additions, but theonly addition CQITIITI~~ to the improved heats was 2% niobium, and the twoalloys that contained 2% niobium were completely free of cracks. Typicalpkotsmiersgraphs of several alloys after exposure to tellurium md deforanatlsnat room temperature are shown in Pig. 7.3.A similar type of experiment was run in which test samples wereexposed tCl Sknall ~ O U t l t S sf tellblriUID Vapor. meSC? @Xgerilllents were ‘KUHlin quartz, which is nonreactive with the materials, and included vacuumoutgasing and bakestat. Thus, the oxygen levels were low and the conditionsshould rep~esedk those that would be expected in a reducing salt.Seve~al materials have been exposed under these conditions and strainedto failure. fie results obtained thus far generally agree with thoseobtained in the experiments where the tellurium was electroplated thetest sarmple. HasteEIoy N formed intergranular cracks, but the intensityof cracking varied with composition. Nickel formed some intergranularcrach. Type 384 StaiIl%ess Steel did not Crack. InCOngl 660 had DOCeraeked in the plating expe~iments but did form shallow intergranularcracks im the vapor experiments. Assuming that 14% chromium is requiredfor protection against tellurium embrittlemen as it is against sulfurP,the 15% chromium in Ineonel 600 should be mar inal in preventing embrittlement.Thus, the different behavior in the two types of experimentsnay be a result sf the experimental techniques or something more nebuloussuch a8 small. chemical variations in the two heats of material involved.These limited observations indicate that many materials are moreresistant than Hastelboy N to incergranular cracking by tellurium, withmost showing no deleterious effects. Among those unaffected in thetests, a noted, were nkkel-base alloys containing 28% or more chrodu~~,stain%ess steels, coppers and Monel. me results on Ineonel 600 (15%CklfQfllibura) Were inC0IbclWSiVe. me tests Of the modified heats of HaStel%OyN offered some encouragement that this alloy can be made resistant byc~mpositi~nal changes. These results suggest that there may be severallIlaterkals Whose Use FJOUld avoid the CraCkhg prQbhXl.....ii..

.....-. ! .,211..... ..,\r.....:.:.:j.....s;:;a....::c

212me psrecedb cussion indicates that the metal to be used forfabrication of an IRUSt Satisfy three IWib r@guiPeIilentS: (1) COKIpatibi%i.tywith the working fluids (fuel salt, coolant salt steam)(2) adequate plasticity after neutron irradiation, and (3) resistance tointergranular cracking by fission products. A single material. need notsatisfy all sf these, since all parts of the system do not have the samerequirements. However, the use of dissiIl-Lk3.r materials introduces cornplexity0% desi so the possibilities for matting the system of onematerial need to be considered.Severall of the materials that we presently view as being reasonablechoices are listed in Table 7.3. The first shown, a mdffied MastelloyN, is highly preferred but its use, of course, depends upon being able toalter the compos it ion to stop the intergranular cracking Whether Has telloyN can be used in the steam generator will depend upon its compatibilitywith steam, which, as discussed more fully in Chapter 8, is stillsubject to question. The peak temperature with the materials used inselection h wsdd IikePy be 1300°F.OUK second choice tsouEd be to use type 384 stainless steel in thery circuit because it appears to have excellent resistance to damageby tellurium. The addition of about 0.2X Ti to this material has alreadybeen sho~~xi to be an adequate solution to the problem of irradiation embrittlelIEnt."%he lE3h qUeStk3Il KegaKdin type 304 stainless steel concernsits corrosion in fuel salt, and the outlet fuel temperature in thereactor might have to be r@dUced to IoWelP the COlfrClSiQn Kate.StZiiH-rlesSsteel will likely not have adequate corrosion resistance in the soslanttransition to WastePEoy M would be made in the intermediateera This would require a duplex tube of type 384 stainlessinside and HasteElsy N on the outside.Bur third choice at this time would be a system made entirely of anickel-high chromium alloy. The available COK~OS~QII data on Iheonel 608(1%&romfum) sug est that such a system would have an acceptable corrosionrate with fuel salt at about 1280°F. However, these data wereobtained a number sf years ago on ~~~ativeiy i~lpure salts, t ~ aaaitisnai~ dCQrPOSiOn testing Will be r@qui?Zed, hCoEX3l 688 has no%: been testedunder control%ed additions in sodium flrasrsbaara es but OM% presmt derstandingof the chemical behavior of this salt ives reasonable hope Ofacceptab~e compatibility. nis alloy has been in many steam generatorsand the experience has been favorable. The main problem withInconel 600 is embrittHment by neutron irradiation. We have been ableto improve the resistance of Hastelloy N and types 304 and 316 stainlesssteel by controlling variables such as grain size, heat treatment, andc~~~positiow, but this ability muat be demonstrated for a nickel-higl?chromiuln alloy.The fourth materials selection would involve the use of the partiallydeveloped 22 Ti-modified Hastellsy N in the entire system with the surfacesexposed to fuel salt being cov ed (weld overlay, duplex tubing, cladding)by stainless steel or Monel. is would require that the details of theprocesses be developed to include joining to insrare integritydUKhg SeKViCee.:. .. G......,$.&

.,...... y.213;.:.=.:.........i...........!SelectionPrimahy circuitCoolantcircuitSteamgeneratoraUncertainties.... .:.&........ .:.:;i.:?............... .y(..A,

214several of these selections involve duplex tubing and: clad structures.Methods for wakin duplex tubing techniques for wela overlayinggenerally exist fo the materials involved. %e greatest complicationcomes about in joinin where, alth~~ghthe basic ability to makesuch joints is available, jo t desips a d welding procedures must ~sedeveloped for each material.h e further method that has been considered for handling the crackingproblen is to getter the fission product tellurium frorn the salt withsome reactive material. The ammt of tellurium produced is quite small,and a very efficient filter would be required. To be effective, thisfilter would have to be placed near the reactor outlet, and the pressuredrup, heat generation rate, and salt holdup associated with it would likelybe Very high. 'FhUS, We have deV0-d Only lianited attention to this approachaOur preferences in Table 7.3 clearly favor staying with HastePloy Nas a structural material. Its resistance to corrosion by fluoride saltshas been well demonstrated, and irradiation embrittlemat appears to betaken care of adequately by the addition uf titanium. Consequently, achange to another material should be made Q~PY if it becomes clear thatHastelloy N cann t be further msdified to improve its resistance to interranularcracki In the event that a change of materials is necessaryrFckel-base all appear preferable t~ iron-base alloys because the saltcan be &uoWed to be more QXidiZill with nickel. However, th@ studiesthus far indicate that the iron-ba e alloys offer more resistance to intergranularcracki~g. Thus, these two factors must be balanced againsteach other in shoosin a material, and this will be possible only whenmore data are available on the resistance of different alloys to crackinga d their corrosion resistance in salt.Further Mark6%OSt pressing problelll W i t h &iStE?l;bQy N is its Su§Ceptibi%ity tograin boundary cracking when exposed eo fissisn products in fuel salts,Work in the immediate future will concentrate on determining whether acceptablechemical modifications to Hastelloy N will adequately fmprsveits resistance to cracking. Experinents already run indicate that theaddition of 2% niobium may be effective, and Hastellays containing morethan 162 chromium are not attacked. Nodifications of Hastelloy N con--taining various concentrations of chrumium, iron $ man mese, silicon,titanium, and cerium will be annealed in the presemce of tellurium, andevaluated for crack susceptibility. If an alloy near the Kaste11oy Ncomposition can be shown to be imune to tellurium, it must then be irradiated60 determine whether it has adequate ductility in the irradiatedcondition, and tests in salt must be run to determine the operating ternperatureEi~~Ltati~ns imposed by carrssion. Further work to develop theirradiation resistant microstructure in the modified alloy will be reqtliredif the modificati~n has altered its resistance to radiatio18.In addition to working on Hastellsy N, we will evaluate a nickelhigh&KOIE~.UITI alk~y. Inconel 600 (with 15% chromium) seem to be borderlinein its tendency to fo~m inter ranular cracks, auad an alloy with about

.....-:.:.:>.....;.

Following the path that has been outlined, we should be able to makea conclusive choice of materials within about two years. However, manytion tests will have to operate beyond this time. Additionaleffort will be needed if a duplex system is required, since methodsmust bi developed for mking clissiaiaar joints, duplex tubing, and weldoverlays 0h.,~...YEvaluationThe basic requirements of a structural material are that it be compatiblewith its environments , have acceptable mechanical properties bothWirKadiated and after exposure to the maximum expected neutron fluence,and be capable of being fabricated with reasonable ease. 'Two cornpatibibityproblems exist, one bein the selective removal of chrsfnim andthe other being intergranular c~~~lbinp due to the infusion of fissionproducts (likely tellurium) eOUK experle~ce with Hastelloy %a has been very favorable so far ascorrosion is concerned. Chemical modifications have made the irradiationembrittle~nt tolerable, and there is reasonable evidence that f~rtherchemical modifications can be made to control the intergranular cracking.Tfie development sf a suitable modification of Hastelloy N that can besafely used with fissi~~itlg fuel salt should be the central thrust of thematerialbs development program.Nickel-high chromim alloys appear to resist i ergranular crackingby tellurium, but the extent of their irradiation e KittleIXnt n%Ust beevaluated by experiments and their CO~X-CIS~Q~ behav r aust be studiedin nore detail. Previous tests with relatively inpure salts and ~~nsiderationof chromium diffusion rates in the netal indicate that the peakerature for Inconel 600 (15% ch~~mium) would be about 120Q'F.essary first to determine the minimuan chromium concentrationrequired to prevent intergranular cracking, and then the questi~nirradiation e~~b~ittle~nt md corrosion must be evaluated.Type 304 stainless steel offers excellent resistance to intergranularcracking by tellurium and has acceptable resistance to embrittlewentby neutron irradtation, but its c ~rr~sion resistance must be evaluatedmore ~ o m ~ l and e ~ very e ~ ~ likely it will net be usable above l260'P.Other materials such as Hone1 aRd copper resist cracking by telluriUEaibFbd Can PQssibIy b@ Used as Coatin s in the psl-imary circuit. However,COnSiderZib%e deV@lOpElent Work WQU d be required to follow thisroute *Our work thus shows that there may be several materials that willsatisfy the basic requirements for MSBB piping and vessels. However,some further inv~i~~tigatio~ will be necessary before choosing the mostdesibable QptiCXl.Tellurium does not form very stable fluoride in MSBW fuel salt,80 it deposits on metab surfaces in the reactor leading to the intergranularcracking that we have observed. Since this pr~cess involves theinteraction Of the t1K1 meta$s, ft Should net differ in ii nbad.eaP eXlVirOIllnentand a laboratory e eriment. This means that most ef the e. .,......... -.,;. .... ā-h.

......:.:.:

218.......... ._References for Chapter 71.2.3.4.5.6,'.e....7.J. H. DeVan, MS aaesis (University of Tennessee, 1940).12 614 e15 0

~21916. H. E. McCoy, JK~, An EzPahat&m of the Molten-Salt Reactor ExperimentHmbelZoy 19 SmeiZlance Specimens - Third Groxp, OR%9E-TM-2647(1978)17. H. E. McCoy, Jr., The HSRE md Its OpePation, report to be published....,:.:.:a......:.z,;&.....;.rr)....;.y*.....'..L*......:a

29 *W. %a. Martin a d J. R. Weir, Jr., "Postirradiation Creep and StressRupture of Kastelloy N," NwZ. Apple 2, p. 167 ($967).30.31 032 e33.34 *35 D36 E. E. Bloom, "Nue%eation a d Growth of Voids in Stainless SteelDuring Fast-Neutron Irradiation, Rdiation-Indufled voids in MetaZs,U.S. AEC Office of Information service^, p. l (1972)3% e38c.u ....39 m40 *41 .C. E. Sessions and T. %. Lmdy, "Diffusion of Titanium in ModifiedHastelloy Nsf8 s. IimZ. Mater. - 31, 316 (1969) 043.44 045 eC. G. Bieber and E%. P. Decker, "The Melting of Malleable Nickel andNickel Alloys s. Adm - 221, 629 (1361).D. W. Wood and R. M. Cook, ""Effects of Trace Contents of ImpurityElements om the ~reep-~ugture B~operties sf Miekel-Base Alloys "Me%dlwgia LQ9 (1963) eJ. 6. Fisher, J. AppZ. Phys. 22: 74-77 C195%>.

8. REACTOR CCMPONENTS WESB SYSTEEMSw. a. HuntleyBunlap Scotta. G. Grindell H. A. McLainA. I. Krakoviak E. S. BettisJ. L. Crowley 6. E. Bettis..... .,.,...A%+The purpose of this chapter is to describe the present status of thetechnology of componertts and systems for molten-salt reactors, to indicatethe importance sf the uncertainties remaining, to identify the additionalwork needed, and to evaluate the probability of success in obtaining re-Liable components and systems. Except for the ~~ntr01 and safety rods,the reactor vessel and internals are not covered in this chapter; theseare discussed in Chapters 3 and 6. The problems related to the chemistryof the salts and the materials of construction are discussed briefly onlywhere the me~hazficab design OF operation of the plant is affected. Otherwisethe reader should refer to the respective chapters (5, 4, and 7) formore detailsIn preparing this status report, we used the reference design forthe MSBR [l] to determine the requirements fur the components and system,exa~~ined the various conceptual designs of the components proposed in thereference design md in the studies prepared by Ebasco Services [2-4],Foster bh@ele% [SI, bfCb?herteK [6], and $@tti§ ffz. [7] to detednethe possible difficulties. We them reexamined the prior experience toestablish the status of the various te~hn~l~gies and to determine thework needed to resolve the uncertainties. as outlined in chapter 2, thisprior experience began during the Aircraft Nuclear Propulsion (I$NP) Programpand progressed through the development work associated with design,construction, and successful operation of the Aircraft Reactor Experiment,and subsequent development for a larger aircraft reactor. This experienceserved as a basis for and carried over into the development studies conductedin support s% the design, construction, and successful ~perati~nof the MSW and subsequently in support of the conceptual design studiesof the MSBR. Operation of the MSRE for more than 13,000 equivalent fullpowerhours provided most of the experience related to nuclear operation.Although not all the uncertainties related to fission product distributionuncovered during the MSRE operation have been resolved, the componentsand systems operated about as expected and provide confidence in theseareas of nolten-salt technology.Presently we are constructing two MSm-scale facilities for the studyof the problems associated with the handling and circulation of moltensalts and the operation of auxiliary components in large loops. TheCoolamt Salt Technology Facility (CSTF) is nearing completion and shouldbegin operation circulating NaBF4-NaP (92-8 mole %> in late sumer 1992.The Gas System Technology Facility (GSTF) should begin operation in thes~~ll~fie~ of 1973. The GSTP will. circulate molten LiF-kF2-nF4-rn4 (71.7-16-12-0.3 mole %> and will provide a means for testing the operation ofthe gas handling system and other components with a typical NSBR fuelsalt. We expect that the operation of these facilities over the next twoyears should resolve many of the remaining uncertainties in the gas handlingand the coolant salt technology....:g.u221

222The discussions which follow provide an evaluation of the te&no%ogywithout much detail, but the references cited contain the material neededto substantiate the eastence and status of the te&nology.Salt PumpsRequirements and CriteriaThe pumps for molten-salt breeder reactors must circulate fl~o~idesalts in primary (fuel) and secondau (cosPant) salt system reliably attemperatures approaching P3QQ’F and meet the general hg~d~aulic requirementsp~esented in Table 8.1. The table presents the design temperatureand the hydraulic characteristics of the salt pumps for the primary andsecondary salt systems for a lQ00-rn(e) MSBR, a 288-W(t) Molten-SaltBreeder Experiment (MSBE) and the MS . other criteria fa% %I[SBR Saltpumps are presented in condensed form in Table 8.2. Inert gas has been~i~c~lated by salt pumps to aid in preheatin and cooling down mm andtest facility salt systems, and the need for this pump capability maypersist for PISBRVs also.k.15,h.,.

223.... .;.:.sTable 8.1. General pump requirements for MBR salt systems. MSBE,

224Table 8.2. Condensed lit of criteria for an M%BR primary salt pumpItemPump hydraubic requirementsPump design pressure and temperatureOperating life requirementsPump structurePump dmerBearingsShaft sealsCodes and standardsMateriais of iwnstmctionVibrationShaft forcesAmbient temperatureNormal operating conditionsStartupMeatup and cooldown cycIcs(to and from room temperature)Zero to full power cyclesFull to zero power cyclesNuclear heat depositionPump tank vollrmeVariabk frequency supply €or drive motorLubricant-coolant packagePump instrumentationPump speedTemperatureLiquid leverMaintenanceDirect maintenanceRemote maintenanceIncipient failure diagnosisCriteriaSee Table 8.1. Pump ma~iufiictuier to study relationship between pump speed. efficiency,NPSH required, salt volume within the pump tank to recommend ahydraulic design suitable to both purchaser and manufacturer.See Table 8. I.30 years.30 years, conveniently replaceable.TQO,OB0 tis, conveniently replaceable.59,006 hr, conveniently replawable.Pump to be designed, fabricated, inspected, and tested as though subject to therequirements of ASME BPV Code, Section 111.$0 meet the requirements of ASME BPV Code, Section IIE, as supplemented forHastelloy N.Avoid resonances that are harmful to the pump and other salt system mmponents.Reduce to level consistent with pump bearing life of 100,000 hr.As specified, which wil1 reflect choice between oven heating of salt system andheating individual system components.Pump should safely withsta~d across-the-line motor startups that produce maxirnnmacceleration.58 cycles specified in pump specification.360 cycles specified in pump specification.240 cycles specified in pump specification.Provide cooling as required to protect pump components from overheating. Heatdeposition rate to Ire supplied in pump specification.Provide for maximum anticipated thermal expansion of the primary (fuel) salt.To meet requirements of pump specification.Furnish lubricant-coolant system to meet lubrication and cooIing requirements ofpump rotary element.Shaft seals shall be used lo prevent leakage of lubricant-coolant into primary saitor ambient atmosphere. Provide split flows of purge gas (helium) in shaft annulus.Upward flow to prevent diffusion of vapors of seai oil leakage into primary salt.Downward flow to prevent diffusion of salt vapors into shaft seal region.Provide three independent pump shaft speed sensing systems consisting of electromagneticpulse generator, sensor, and readout. Provide capability of sensingdirection of shaft rotation.Provide ungrounded sheathed thermocoupbs to measure pump temperatures asdescribed in punip specification.Fabricate and instal1 liquid level sensor in pump tank to the design provided by theCompany.Provide static shutdown shaft seal and purge gas provision to accommodate removaland replacement of shaft bearings and seals subassembly.Provide dl features of alignnrent, quick disconnects, bolting, and devices needed toaccomplish the remote removal and replacement of the drive motor and pumprotary element.Make provisions to detect malfunction of principal pump components to anticipatetheir failure.,....w.&.,h.

~225Table 8.2 (continued).... .&.........:.:&,

226Short-Shaft Pump. - A conceptud drawing of a short shaft centrifugalpump is shorn in Fig. 8.1. It is the pump for the primary salt system inthe reference design for a single fluid WBR [9f 0 Except for its largecapacity, this pump is very similar to those used in 1954 in the ARE,those proposed for the Aircraft Reactor Test (ART) in the late 1950'sin the 1960Ps[lo], and those operated for over 20,006 hours in the PIS[la]. It consists of three principal parts: the rotaryr element, thepump tank and the driver. The rotary element contains the conventional,oil-l~b~icated bearings that support the short shaft from which theimpeller is overhm and also the Shaft seds that hUPd the beaKiTngand seal lubricant- oolant in the bearing housing. "he pump tank incorporatesthe pump casin (volute), the necessary nozzles for the inlet anddischarge of pumped salt ad inert cover gas, and the mounting bra&ets;the tank would be welded permanently into the salt system in an MSBR.me pWkp driver is preSefatay COIlSide~ed to be a thPee-phaSe iHndUctiOn(squirrel cage) electric motor installed in a wate ooled vessel. mer0taI-y element, pump tank, and drive motor are ass led with gasketedjoints to form a gas-ti t unit; the gasketed joints are connected to aleak-detector system for nuclear reactor application."!A-- A comeeptual drawing 0% a long-shaft centrifugalPump . 8.2. It was considered for th@ reference two-fluidMSBR [la] and is similar in overall features to that preferred by BytonJackson [13, 141. The principal configurational feature that dtstinguishesthis cOllceI>t fPoHl 8 Short-shaft pump is the long shaft SUppQrted at itsabmer end by a m~lten-sabt lubricated bearing. Because this pmp configuratFowwas considerably outside our experience, we had the rotordpa~cs sf the ~htlft-bea~ing-housilrag system and the sharacteristics ofmlten-salt lubricated bearings examined in some detail by MechanicalTechnology Incorporated. The results of the study were reported in [-%SIand summarized in 6161. The pump shaft length and design speed for thepump shoulbd be selected to operate safely below the %irst shaft criticalspeed if at all practicable. If this cannot be done, then a practicalmeans of performing precise dynamic balancin of long pump shafts must bedeveloped. ons side ration should also epe given to constructing a rotord~adcsi~i~lation facility of the shaft-bearing system to e~aluate theeffectiveness of the dynamic balancing pro~edures, the bearing performance,system stability.wp is suitable for reactor salt system that requirelocating the impeller at relatfvely low elevations in the system layout.It is suitable for ci~culating gas only at temperatures above the saltwelting point, and then only if special pravisiom are made to supply-salt lubricant to the 1o~e.g: bearing when the salt system isut the time 0% the completion sf the m~ study, the focus ofattention shifted from the two-fluid to the s%ngle-fluid systeuLs, and noXlKlKe Work Wa8 pE?rfOKmed OH1 the long-shaft pUlDp.%&. .h.,

227OWL-DWG 6Y-6802.....,.x.:.y.d.........,sx!.....l,..- .,13: in.....x.u.....,:.y,.>......'.i' . dFig. 8.1. A conceptual salt pump for Molten-Salt Breede P Reactor.An example of t :he short shaft pump configuration a

228

..... ~.~ ....l229.z.+$. .......A.-.......... ...&..... :s*.....s.:g....l...i A,&% .,.;.;.;.....:.zc>r.%;......=....":.s.....:.;.;*.-.Short-Shaft Bump. - Numerous short-shaft cewt~ffugal pumps for mo%tensaltand liquid-metal apglfcations have been designed, constructed, andE. Table 8.3 is a r6sum6 of these pumps giving design valuescapacity (flow) and speed, the nuder of umits built foreach model, and the total operating time accmulated with each model ofA very wide range is represented: capacities have ranged fromm to I500 gpm; heads, from 56 to 400 ft; and design temperatures, betweenI200 and 1500°F. May pmps have been operated at 150Q"F, Sinunits that were operated continuously and satisfactorily include severalEFB pumps for periods of 15,008 to 20,000 hr and one MF pump for morethan 25,000 h ~ * More than 58,006 hr satisfactory operation was accumulatedwith the MSW salt pumps in the fuel- and coolant-salt systems inthe reactor. An MSRE prototype pump was operated for approximately $000hr with molten-salt. mis was followed by 14,000 hP of operation of theMa&-2 fuel salt pmp which had a deeper pump tank and a slightly longershaft OVerhang for the iIIIpe%leK than the Heactor pump. I&US it ketal Ofmore thah 86,080-hr operation was accmulated with four %RE salt pumps.The models LFB through Pa pmps were designed to the requirementsof the 1950 versions of the ASm Boiler and Pressure Vessel Code (ASHEBPV Code), Seetion VIPI, supplemented with metallurgical data taken duringthe ANP P~o~~~III. The MSRE: and ALPHA models of molten-salt PUIR~S we%@designed to the requi~ements of the 1960 versions of the ASKE BBV Code,Section VIPI, as s~pple~~nted with low cycle fatigue data obtained duringthe early polstion of the MSU and with thermal strain-fatfgue analyses of'the pump tanks [I71Several concPusions important to pump design were made and reaffirmedas pump operating experience was accumulated. The short-shaft pump configurationcan be designed and built to operate satisfactorily in moltensaltreactors amd satisfactorily scaled up in size from 5 to 1580 gpm.Pump performance characteristics obtained during water tests [I$, pp. I1and 131 of a cewt~ifugal pump can be related reliably to the molten-saltoperation of that pump. Experience has pointed out the importance ofquality assurance in all the prime function^ of design, materials prscurementp fabrication, assembly, test and inspection, installation, andoperation to the production of sattsfaetory salt pumps. Occasionallyduring the scaling up of pumps to larger capacities, a principle previouslyapplied may be ove~l~~ked with annoying consequence, such as thelocation of the parting plane in the catch basin in the MSRE fuel- andcoolant-salt pumps. The eat& basin is provided to catch the lub~icatingoil that leaks past the lower shaft seal and thus prevents it from runningdown into the moaten salt in the pump tank. mere are vario~ designapproaches to providing the basin cavity, and prior to designing the MSREsalt pumps, all cavities had a parting line which requjired a static sealat an elevation considerably above the floor of the basPn. In the PISREsalt pumps this parting line was placed at the floor of the basin, andduring pump ope~atfon the two components joined at this line underwentsome ~elative displacement that opened the joint and permitted seal oilleakage to K U at ~ a slow rate down into the molten salt in the pump tank..Iln .,.,. _u.t.g2

2364LFBDANADACIn-Prie LoopMFPKAPKPMSRE fuel silt pumpMSRE coalant salt pumpMSRE Ahk-2 fuel silt pumpALPHANa, NsK. andrnuntert sdtNa, NaK, andmolten saltMolten saltk'fdten SdtNaK and molten saltNaK and molten saltN& and molten §dtblolte-n saltHeliumMclteri saltHeliumMolten saltMolten salt923005010504038050785030051506011900375150012QQ800120030400037501450300030003550350011751775117565001100-14061000 - 15001000- I400B I Q0-I5001900-1 50070Q-15001800-1225100-12001000 1225100 12001000- 130 0850- 14004610383242e21I466JKM3b57,0004,000f4.UOOC4 1,000"21,50045.0053 1,6006,00024,6004 QQQ14,0006,000.....*&.S.. .. c.3..k.Total83734,700

241The permanent fix, which required seal welding the edges of the jointswas adopted only after nuclear operation of the MSRE and did not forestallthe C Q I X ~ and ~ ~ activity needed to handle the Consequences of the oil thatleaked into the MSRE fuel-salt system.Long-Skaft Pump. - The principal resuHt~ of the study 0% the rotordynamiccharacteris tics and molten-salt lubricated bearings for the longShaft pbatnp are presented above in the section entitled Salt $UlTIfl Concepts.We had some experience with operating test bearings in molten salt in theearly 1968's, which is reported in [19]. Hydrodynamic lubrication withmolten salt in journal bearings was demonstrated over the tempe~aturerange %200-1580°F. A BKA model pump was modified [XI, pp. 56-57] tosuppc~rt the l ~ e end r of the shaft with a molten-salt lubricated beariand was operated for approximate.ly 12,500 hr at H225'P. Although alljouma%s and sleeves in these test bearings were constructed of HastelaboyN and were operated with some degree of satisfaction, it was decided onmetallurgical grounds that long tern reliability required the applicationof hard materials such as refractory metals and cemented carbides. A fewmaterial specimens, hard faced by the plasm spray method,Subsequent concentration on the pump requirements for thedeemphasis of this effort, and the test program for the specimens was notcompleted.hdustrid Experience a d IwterPest. - Efforts Were made to elicitindustrial interest in the production of salt pumps for the ARE, ART,MSRE, and associated molten-salt test facilities. Hasever, the quantityof pumps to be produced at any one time in the two decades during whichthese projects we~e viable was apparently too small to obtain and maintainserious industrial interest. In 1969 a specification E211 was writtenfor a short-shaft centrifugal pump to provide the requirements of theprimary and secondary salt systems in a 100- to 20o-NW(t> MSBE, (see Table8.1). Westinghouse EHectro Kechanical Division, Bingham Pump Company,and Byron Jackson Pump Company expressed interest in cementing OR thespecification and receiving a request for proposal to produce the pumps.Westinghouse produced a very good response to the request, the only onereceived, but the deemphasis of the MSBE in 1970 obliged us to drop theirproposal. We have had no contact with the pump industry to produce saltpmps sf the long-shaft configuration.Status of Pump TechnolornThe precedtng picture of the status of the technology of salt pumpssupports the belief that satisfactory salt pumps of the short-shaft configurationcan be pr~duced for MSBR's. What follows in this section tendsto emphasize problem areas, none of which should be insuperable for eitherthe short shaft or the long shaft pumps.

I~- .........232ai

.....233,Z,M.....r...... -2,.:.;.x.,....,saFabrication. - A continuing effort has been expended in developingin-house fabrication expertise with and introducing segments of industryto newer construction materials such as Pnconel in the early 1950ps andHastelloy N in the early P960's. In the pr~dueti~n of satisfactory HastellOgiN castings for irripeal@Ks and casings %OK the MSm Salt pWIlpS it Wasnecessary to take the best sand castings produced by the casting industryand upg~ade them at OWL by identifying flaws with radiography and thenrepairing the castings by grinding out: the flaws and depositing weld metalin the cavities until 'a satisfactory metal structure was achieved. However,for MSBR salt pumps, the prospect of making pump impellers and casings byweld joining machined pieces appears to be an attractive alternative teacastings 0.....::sfi.....;.I" .y I........w ?ry . ......iii.,.:.x,...$

2 34beneffeial to salt pump technology also. An iterative approach dth selectedvendors should lead ts simpler and, hopefully, less expensive pumpdesigns. Et should also lead to lmer pump construction costs as expertisein Bdentifying and resolving the design problems peculiar to nuclearpumps is obtained and fabrication a d assembly talents are sha~pened.Mu& of the industrial e erieace derived from the development of pumpsfor R's can be expected to apply to pumps for EfSBRs.=.- . ..Effects of uncertainties. - BR salt pmps have the usual turbomachineryelewnts upon which are UperilIlpOSed the reqUirETWnts fOK Sperationat temperatures in the range lO0Q to 13QO"P and in a nuelearradiation en~i~~~~~ient. These latter requirements generate some problemof heat removal and radiation damage that will require directed effort.seals having reater rubbing veHmever, neit er of these problwill bar the desiities than our eqerience has covered.presents a genuine uncertainty thatOK short shaft pumps, but if long shaft pumpsshould be needed,previously mentioned should yield to adequatedevelopment effort a Confidence in their long tern re12abikLty willreqUiPe endUrac@ test opE?.ratiQn in mOBten salt.If satisfactory operation sf the long shaft p t supercriticalspeeds is proven to be difficult to achieve, the d sf the pump shaftmay be modified to increase its stiffness and rais first flexuralresonance frequency of the shaft safely above the pump desiAlternatively, the hydraulic design may be modified to deerspeed satisfactorily below the sritieali freqsaenq of the shaft eThe principal features of &e development requirements for (a shsrtshaftsalt pmp include proof and endurance testing of the shaft seal,it WateK perf13 ce test sf the pump hydraulic desi , and endurancetesting of a stype pup with molten salt. In 8 ition to these item,-shaft salt p will require a bearing materials developmentoumal mounting method program, a rotordynamicogrm, and probably verification of the capability tofabricate IpUlIIp Shafts to pr@Cise cOIlcent%iCity, stKai&tne%S anddynamic balancing requirements. Each production pmp, whether short 0%long shaft., should be subjected to a shakedown md m~lten-salt proof testprior to installation into an NSBR salt system,The diameter and rubbing velocity of the shaft seal for an HSBR saltare laKger than ally We have e 62Ktestced. HnnanUfactUE?K Shouldselect the shaft seal design from among the bellows mo~~lt, fluid bearing,Visco seal, etc., and provide performance and endurance tests until aSatisfaa~pgr aesi is evolved. The proof test of shaft seals installedin promtype and pr~duction salt pmps is performed as part of the shakedm and molten-salt proof testing.

235A test stand should be designed and constructed to perfom a seriesof tests with molten salt on the prototype pmp that will verify itsperformance and give confidence in its ]bong tern endurability. The prekidnarydesign of such a test stand suitable for MSBE salt pumps was completed[22, 233 in December 1969; it was estimated to cost $860,000 atthat time. A partial listing of important tests planned fur fie facilityy,WhiCh are pP2Sented in references 22 and 24, hClUdeS:I.Obtain the pump hydraulic perforwamce and cavitation inception characteristicsover a wide range of pmp speeds a d capacities and temperatures2.3.4-5.6.7,Determine the characteristics of the purge gas flow in the shaftannulus 0Determine the distribution of salt aerosols produced by pump ope~ationand obtain the performance characteristics of a e ~ ~ removal o l devices,if necesetary.Bemomstrate the operabiPity of the incipient failure detection (HFD)devices sObtain long tern endurance operation with a prototypal pump.Make molten-salt proof tests of advanced instrumentation for moltensaltsystems as it becomes available.Evaluatf~nwe know hm to make reliable short-shaft centrifugal pumps formolten-salt reactors kavfmrg built and operated many with capacitiesto 1500 gpm. Although it m y take several years to produce the largerpumps fog demonstration of fdf-scale MSBR's, the problem are well understood,and there fs little question that satisfactory pumps can be obtainedon a schedule compatible with obtaining the other principal reactorcoHLpO3X~RtS.We have had very little e eriencle with the long-shaft pump configurationfor mlta-salt system. However, it has been used in sodium system,and plans are being processed for its use $n the liquid-metal fast-b~eederreactor ~E-S~KEUII (LMFBIQ. The incentive to use long-shaft salt pmps isto reduce the total deve~opment requirements for MSBR'S anel LmBI[%'s. Indeed,the MSBR short-shaft pmps can benefit from the LB-EBR sodium pumptechnology program because the two pump configurations share many eomonrequirements. The advanced analytical design methods that are being developedfor the sodium pumps should have much direct application to thedesign of molten-salt pumps, We expet that the results of the shaftseal development PKS~Z-EUII for sodium pmps will also have dire~t applbicationto salt pumps. TIM fabrication technolog developed large sodium

236pumps should have application to the fabrication of components for MSBRpumps after making some d1mmces for a prob He difference in constructionmaterials e The themal shock capability f the Sodium Pump TestFacility (SPTF) could be utilized to shock a prototype BR salt pump,However, the requirements for sodium and salt pmps are also sufficientlydifferent that in other important respects they are unique. '&ledensity of salt and sodium il~e quite far apart and for apaefty and head, the pmer ~e~pired to drive a salt pumpdeveloped wodd be several tines the pwer a d pressure for a sodPum pump.Hwev@rs besame of the hi her volumetric heat capacities of the salt,the total pump power requi are less for salt than sodium. Thetemperatures for the 'S are 180°F Or lllOBTC2 those fordiU?& reactor SySt therefore the salt pumps dlbl have to beea for saightiy 1 able stresses eObtaining rtaaterials for molten-salt P~b%icated bearinbe SatisfactOFj for Steadap KUIIniwg cofsditiOIlS 2s EEQt likely to be diffk-Cult. MCWeVeT, StaIptkIg and Stopping Co~ditions b'SCOW3 mor@ SeVeHe asthe temperature is raised above Ib20Q'F. Also, as pointed out above,suitable devices would have to be developed (I) t~ attach salt-bearingmaterials to retainer pieces having a greatly differing coeffic%ent ofthermal expansl%onintain alignment bemeen journal andbearing surfaces with other shaft support bearing at 1PSO"F.Economy in design, production, operation, and maintenance favors theEitteIiQt to m&e the hydKad%C and ?JECh ical designs sf pumps for boththe primary- and seeondantgr-salt system very sidlar, if not identical,Coolant sys tern.....

237,%>....__ iiih......iii.*....223other than the eutectic NaBF4-IRPaF ($2-8 mle %> which has a melting pointof 725'~ and costs approximately $68/ft3 ($0.5/lb). ~t appeared that itmight be possible to contain a small concentration of hydrsxl ion incomplex in this coolant which couId be continuously removed and replacedto function as a trap f o the ~ tritium. bong its disadvantages are (I)the seed for BF3 Fw the cover gas because of the evolution of BF3 fromNaBF4 at MSBR operating temperatures and (2) an affinity for water, whichcauses increased corr~si~ene~s in bo%h the salt and cover gas systems.As will be described below, these problems are not so severe as to affectour Choice Of the flUQrObOr%t@ IIliXtUTX? as the coo%Eknt in future MSR's,A lower-melting-point fluoride mixture of LiF, NaP, and BeP2 with aPiquidus temperature of 640"F, a mixture of Lf61 and KCB with a liquidustemperature of 680'F, and a mixture of 'KF and ZrP4 with a Piquidus temperatureof 752°F appear to be alternatives. As mentioned in Chapter 2, someconsideration also has been given to a special design using both LiF-BeF,and a nitrate-n%trite mixture bemeen the fuel and the stem. Addltisnalinformation on coolant salts is given in Chapter 5.The simplified schematic diagram in Ffg. 8.3 shows one of four secondarysalt circuits required to soslb the reference design l00Q W(e>molten-salt breeder reactor. In each circuit sodium fluoroborate sal%flows from the outlet of a primmy heat e ~&~inge~ at ll5Q"F to the suctionof a sump-type centrifugal pmp, Salt discharged by the pump flows inparallel through several steam generator-super~eater and steam-reheaterunits and returns to the primary heat exchanger at 858'P. The flowthrough each secondary circuit is 20,000 gpna and each circuit contains2888 ft3 of sodium fluoroborate. The total ~im~unt required is ~458 tows.Valves are provided fm the salt lines to c~ntr~l the distribution sf flowto the StGXiI-KaiShg units and to bypass coolant flow around the heatexchangerunder partial load conditions. (These valves need only tothrottle, not to shut fxag,) A pressure-re~ief system is proviaeafor each seeondaq salt circuit whereby high pressure produced by ruptureof one more tubes in the steam-raising units is relieved by the burstingsf rupture disks and rapid blowdown of salt, noneondensable gases,and steam. A COmOfp GOVeP-giLS System pPOVided for the four Circuitswhich supplies a mixture of BF3 and helium and possibly other gases tothe tank of each salt circulation pump to provide a protective atmosphereover the free surface of the sodium flusroborate. A purification systemis provided to maintain the purity of the salt in the secondaq cireuits.As a mfnimum the purification system fs expected to remove some cokrosisnprctducts by cold trapping a d to convert oxides and hydroxides to flusrides.The processing system may also remove HF from the salt by contactingit with gas or by other means.&,&....i-. .,Experience with Coolant SaltsA LiF-BeF2 mixture was circulated at 860 gpm in the secondary systemof the MSRE for approximately 26,008 hr over a temperature rage of l000to 128@"F. DUKing the l8fetiFiW of the mRE (%4-1/2 gTr) the &romiUm Contentof the salt remiwed constant at absut 32 ppm, indicating remarkablylittle eorrosfon. Metallographic examination of the piping after shratdswn....,,&..,a...

238BART OFORML-OW- 69- (0494ARFREEZE VALVE FLOW RATE6ARE TOTAL FOR 1000 Mar(@) RANTSECONDARY SALTDRAJN TANKe 8.3. Secondary salt and steam system molten salt breederreactor flow diagram-EOQO M%(e) Unit-Mark %I.

239e....w..........:.:so...Z&....d.74.......a..Status of Flueroborate @solant TeehnohgyThe compatibility of sodium flusroborate with the proposed MSBRcontainer material (Haatelloy N) is satisfactory (csrrosion rate of 0.1to 0.2 mils per year) provided that moisture is excluded from the system.Data from eorrssion experiments show that addition of a small amount ofwater to sodium fluoroborate salt in a Rastelloy N container at about1100°F results in the metal eo~roding at a rate that is initially manymils per month and that decreases gradually until it becomes less than8.3 mil/yr [28]. Until recently chemkaP analyses indicated that. c~r1-0-sion rates this low were obtained with salt containing mare than 1608 ppmof water and of oxide. This ked to the suggestion that sodium fluorsboratecould contain water in two foms: one that is highly corrosive andOne that iS Only Slij$lt%y cOK333siVe to H%stelfOy %a.study has revealed that the analytical results were in ~ T K O ~ . Salts thatWCK~ ~epo~ted to csrntain more than 1008 ppm of water are now found to containsmall amounts Of a hydroxyl Compound (presumably sodim hydroxgrflusr~borate)and to contain more than 1000 gpm of oxide.%eCent iRteHlsiVe,aa....

......:

242h sodim flussoborate Undergoes a solid phase transition at%470"P, no unusbpaa. problems associated with this phenomenon were encounteredwith the. operation of the freeze valves used in the forced convectionor XP kmps. During the operation of the PKP loop, the freezevalve was maintained bemeen 268 and 488"~ and was sthjected to a totalof 16 fill-drain cycles.Evidence of the corrosion product, M~$PF~~ has been found in virtuallyall loops cireulatin sodium fluoroborate. The deposit is found inthe CBId@St part sf the c rcuit and is expected to deposit on the tubesof the steam-raising equipment if not removed by some means. %heref~r@~the solubility of NagCrP6 in sodim fl~o~~borate w detedned and coldtrapping techniques were used with limited suc@ess 8 isohte this materialin natural- and forced-circulation loops. Excessive drain times wereencountered that were attributed to the accumulation of the corrosionproduct, Na3CrFg, in drain lines. As a result of this experience, thedrain line temperatures of sodium fluoroborate loops should be maintainedtly higher than, the main loop temperature. Also, the freezed be located near the drain tank rather than near the mainloop, This confi uratisn will lengthen the diffusion path between themain loop a d the coia spot in ~e drain kine. ~n MSP~ the concentrationof Na3CrF6 will be kept low to protect the steam generators heat transfere PKP pump shaft annulus splits into.9> flows dmn the shaft and into thepump bawl vapor space; the remainder flows up the shaft, past the rotatingshaft seal, and through the oil catch tank. The shaft purge inhibits$a& diffusion of aoxisus gases from the pump bowl to the sea% regionmd oil catch tank. The BFg content sf the lower seal purge stre= wasITBiIItairsked b e h W 1008 ppm, and no de%@FXKiOUS OperEltiIlg exTperieflCY2 WiiSencountered.uncertainties in Use O f Fluorsboratew.5

243A system of injecting hydrogen into the salt circulating system, removingthe bubbles by means of a centrifugal separator, and extracting the HFfrom the off-gas could result in the accsmodation of a water leak of l...;....ICover Gas Addition to and Removal from the Pump Bowl. - Certain impUriti@S,VagfQUsby identified 88 Water, hydrffw f%UoP3bOric acid, BP3-I-fzOreaction prQdUctS, etc., have @aUS@d high CoPlfosiCKi rates iB9 cibCUlE&tingsystems and flow restrict$an in the off-gas systems of sodium fkuoroborateloops. The identity, SQUTC~, method sf IXIWV~~, and the maximum impurityconcentrations that can be tolerated in the cover gas and salt need to beestablished. The CSTF will utilize the helium purification unit from theMSRE to purify the inc~ming helium and thus eliminate one possible sourceof moisture and oxygen.The coolant pump for the reference design MSBR will operate at ll%QQPwith a shaft purge rate af not less than 5 liters/min (STP) and a pumppressure not less than 10 psig. If no recovery provisions are macle, thefour pumps wou~release one cylinder of B F ~ (ZOO ft3 STP) into the atmosphereper day. The CSTP is equipped with a recove7 unit that is expectedto reduce the BF3 release rate by a factor of 100. Disposal of the residualBP3 in the off-gas stream from the BF3 recovery unit (

244Coolant Leaks e - As explained in Chapter 5, vis%ent e~othe%dc~eactioiis occur when f%usrsborates are mixe with steam OF with fluoridefuel Salts. h fact, flUQrQbOrat@S are ilTlIl scible with mullten mixturesof Ifthium and beryllium fluorides over a significant range ~f cuaditisns.Uranibam and Other tpi- and fetKaVa%enfZ elem@wtS a%e not extracted intoflusroborates, and the only high-melting CQIR~XIII~ that might be fopandis sodium fluoride. Thke~e is some migration of LiF to the fluomboratephase, and replacement of NaF by LIF in the MaBF4 csmplex results in analmost immediate release of some BF3 gas, The detailed consequences sf0% the fuel salt with the sodi~m fluomborate depend on (a) thef traixhg of the two B~H~ZULS, (b) the solubility of the BP3 gas inthe resultant phase, (c) the relative temperatures of the two fluids,(d) the kinetics of reactieans bemeen the distributed components 8 and(e) which fluid constitutes the continuous phase 0 Present evaSbuation ofa%P of these factors indicates that there is no mechanism ~ O K the concentrationof uranium to prsduce a critical configurati~n and IL~Q compoundswill be formed which cannot be redissolved through the addition Q€ app~opriatechemieal agents 6321. For pils R use, BIBKeoveP, the accidental introductionof fluorsborates into the circulating fuel would cause a largereactivity decrease because of %he borona, and thus even a small leak

245would be quickly detected. The boron could be removed from the fuel saltby treatment with ?a'. some studies of mixing in a loop system are neededin order to better evaluate the effects.......... ,:.:.y,EvaluationAlthough sodium flusroborate is somewhat of a nuisance salt comparedwith LiP-BeF2, Its ope~ational probPem are not insurmountable, and its%ewer liquidus temperature and cost make it our choice for the MSBR. Themajor 3?2qufPE%leftt SefaSitiVe and early detection Of bnroP.5 tUr@ iEll€%kageto minimize corrosion of the containe~ material and corrosion-productdeposition in the steam-raiahg equipment. Detection of the resultant HFin the off-gas Pine and detection of chromium by the on-line salt ZX-G~.~Z~Kare two promising methods for detecting moisture inbeakage. Processingof the coolant salt to make it serve as a t ~ i t i k l a a slnk ~ is discussed inChapters 5 and 14.Beat Exchangers....?$dPrimary heat exchangers (salt-to-salt] and steam generato~s (saltto-water)comprise the coupling between the circulating fuel reactorsystem, the coolant system, and the steam system. Othe~ heat exchangersless i m p ~ ~ tfrom a ~ ~ the t standpoint of development required, are the steamKE!heZIterS (Salt-tQ-stea$ and %he 9eP1@at-st@a~-prehE-neatePs (SteaIll-tO-ste~).SOllE? Of the ilRpOrtant teTDpeKatures and pKesSUPeS Within Which these heatexchangers must operate are given in Table 8.5, More specific informationwill be found in the references shm.The steam reheater and the reheat steam preheater are regarded asconventional heat elgchawge~~ and no fundamental problems of heat transferor devel~pment are foreseen. Even though the 656°F reheat steam entersthe reheater below the 7BQF liquidus temperature of the coolant salt,the Pow steam-side heat-transfer coefficient leads us to conclude thatthere will be RO sig~~ificaatt PPoblem with freezing of the salt. The remainingdiseussion is therefore directed toward the primary heat exchangerand the steam generator.Requ%rements and Criteria for the Primary Heat Exchangerand Steam Generator,The general requ5rements and criteria cornon to the p~i~ta~y heatexchanger and the steam eHaeleat0P OUP Kef@9@nce design BPe i68 follOW§.Both are shell and tube eat exchangers fabricated sf HasteEloy M and bothmust maintain their structural integrity during a 30-yr desiineludes thermal transients caused by nomal operations, various plantupsets, and emer @facies DLffeKent$a% e~~lXSi0T.I bef3TeeKI tubes a d S3hd.lmust be accsmodated without the use sf bellows, and thema% stresses atcrftbcal locations, such as tube sheets and nozzles, should be minimizedwithout the use of a gas space if possib.B)e......,:.:.:.>CIC.?G2

246Neat exchangerSteam generatorSteam reheaterReheat steam preheater Steam (shell) 550 650 595Supercritical steam (tube) I000 869 360044(.. I

247The steam generator tubes will have supercritical steam on one sideand sodium fPuoroborate the other. ?The peak t@mpekatUr@ Of the steamside will be about POOO'P and that on the salt side will. be l150'F. Bothof these fluids are c~rrosive under certain c~nditi~ns. As discussed inChapters 5 and 7, sodium fl~o~oborate is aggressive when water is presentin the salt. Steam is oxidizing to metals and can produce st~ess C~ITUsioncracking when it contains small concentrations of chlorides. Thusthe material used for steam eneratolp tubes in a molten-salt system mustresist corrosion by both sodium flusroborate a d steam, or duplex tubesmust be used.With the use sf thermal baffles for the pPOteetiOn of the tube sheet,the 788'~' feedwater entering the steam generator will cool some of thestatic salt below its 725'F liquidus and freeze it on the colder surfaces.The desip must accommodate the freezing and thawing of this small amountof salt without structural damage to the steam generator.We believe that direct maintenance of the steam generator will bepossible even though. there is some induced sodium activity in the sodiumff~or~borate. Estimates place the 24Na activity low enough (about 11 ucuries per gram of salt) for direct maintenance if drained of salt [333.Although the extent of the problem is uncertain at this time it is possiblethat the maintenance plan must be able to accommodate two other sourcesof radioactive contamination. Trace elements (such a8 cobalt) in theHastelloy N will be activated and could be subsequently dispersed by thecorrosion process. Failure of a primary heat exchanger tube couPd allowfission products to enter a d be dispersed thr~ugh~ut the secondary circuit.The design of the steam generator must aceomdate the necessarymaintenance and inspection to meet the requirements for in-service inspection,for minimizing down time, and for detection, location, and pluggingof leaking or damaged steam tubes.Two special requirements -minimum fuel salt inventory and maintainabilityby remote means - are imposed on the primaq heat exchanger. Thelow fuel salt inventory is desi~able to minimize the doubling time andinventory costs, The use of enhanced heat transfer surfaces on the heatexchanger tubes will. serve to reduce the total surface area, and the useof small tubes will further reduce the fuel salt volume. 'Fhe rapid replacementof the tube bundle (or the entire mit) by remote means is necessaryto minimize down time in the event repair is necessary in thehighly radioactive primary system.Current Conce~ts..::aa....A,+.,.....AConeeptual designs of molten salt primary heat exchangers and steamgenerators have been or are being prepared by OWN%, Ebasco Services I~COKprated,Foster Wheeler Corporation, and by Black and Veatch. These arebriefly described and referenced in Tables 8.6 and 8.7.One steam generator configuration which will receive attention infuture design studies and has promise of alleviating the feehater temperaturerequirement is the bayonet QK re-entrant tube. Although investigationsthus far indicate the bayonet tube configuration is not practicalat supercritical pressures because of the thick tube wall required, it........,:.x

248Table 8.6. Molten &t steamConceptConfigurationThpmal Tube Tube size,rating (hlpvv) Number OB (in.)Tube sheet totube sheetlength (R)ReferenceOW". Reference Horizontal U-shell 121 393 % I6 [401and tubeEbasco (E3 & W) Vertical helical 483 815 Pcoil uFoster FVheeler Vertical L 483 1025Molten Salt Vertical huckey stick 483Breeder WeastorAssociates (B & V)3450 Y112~=.= .......... .. i. ... i..1

249........., .a,..........

has definite possibilities for a subcritical pressure system. Ebaeohas Hooked at this eonfiguration [38] and the Dutch propose this conceptfor a rYMalten-Salt-heated SteaHa eflerator [39].Heat Transfer ExperienceThe operation of the MRl3 represents the most recent large scaleerience with salt-to-salt heat exchangers [MI. 7315s operation proeda consfder&le momt of confidence in the design techniques.The heat transfer correlations used for the MSrn primary heat exchange~were based on the previous development tests which showed thatfluorPde salts behave as noma1 fluids. When the HSRE operation revealedthe ~~eral% heat transfer coefficient to be less than predicted, reevaluationof the physical properties disclosed that the actual thermalconductivities of the fuel and coolant salts were below those used inC~lCUlationS d acCoWted for the difference [49]. The Q~er-~ansfer coefficient of the MSRE heat exchanger did not. changeduring some 22,000 hr sf salt circulation and 13,OtBQ equivalent fullpowerhours of ope~atfon thus indicating no buildup of scale and noevidence of gas filming.'Fke sdt CCKilp3s~t~~nS Used iH6 the Msm iage not &e Same as preSentlyproposed f o the ~ MSBR. Hmever, some preliminary heat transfer infomationfrom the operation of a smkl ssrrosioat loop w5th sodim fluoroborateindicates general agreement ~ 5th the Sieder-Tate correlation [50]. Thusthe conclmion is that tihe use of accurate physical property data withcorreaatiows for noma1 fluids is adequate for heat trasfer design withfluoride salts. There is, ~ O W ~ V ~ no K , molten-salt heat transfer experiencewith the knurEed tubes proposed for the primary heat exhmger in the 0reference aesiCorrelations for hea! transfer and pre UPe drop Were chosen fromSpETl litePatU%e md adapted for Use iH1 desi ing the mBR reference primaryheat ex&an er and steam enerator. D cussions of the we of thesecorrelations certain correction factors applied because of baffle spacing,bypass flow factors, etc., are fomd in Ref. 51 for the primaq heat exerand 52 for the steam generator.Although the operating experience with salts is quite extensive, theavailable experie~aee in the generation of steam irt high-temperature ~noltensalt-heatedsteam enerators is nil. mere is a considerable amorant ofe~pe~ien~e, both iltk the USA atad Europe, with the use sf Iow-melting saltcalled Hitee. for the generation of low-pressure steam and for a heatt%fa?ilSfer m@d%UKD [53]. This Salt WoUEd be 83[1 effective barrier for tritium,as described in Chapter %, and has therefore been sOwS%dered for US@ in amolten-salt reactor designed to be built with a minimum of further developmente541 *b:,*%.. ....Flaterfals ExperienceAs discussed in Chapter 5 and in the discussion of the coolant saltearlier in this chapter the ope~atio~n of the Poops has shown that sodiumfluorobor~~te is compatible with Rastellsy N in the absence of impurities I

:.:.:.>;....251.xs*:;:;::v........:-3...::WJmainly moisture. In the absence of moisture, metal removal rates of a0.lmillpr have been obtained, which is ~ Q W enough for use in a plant designedfor a 3O-yr life. However, since the corrosion rate accelerates whenmoisture is introduced and since the plant must have the capability of~eeovering frsm a steam leak, a smll purification system must be pro~idedfor removing water and corrosion products from the salt. A method forremoving water has been demonstrated on a small scale and the use of coldtrapping to remove cor~osf~n produets has also been demonstrated.The compatibility of Hastelloy N with stem is being investigated ina test facility in TVA's Bull Run Steam Plant. Unstressed tabs exposedat 1000°F have very acceptable metal loss rates of acb.25 mil/yr. TheBull Run facility has been modified to accomodate stressed specimensamd a few have failed to date. The failure times do not seem outsidethe scatterband for failure times obtained in inert gas. This work isbeing conducted to evaluate the possibilities sf stress corrosion ofHastellboy N in the presence ~f steam as reported by Spalaris et aZ. e551based on their rather limited work.Some tests are in progress on duplex tubing manufactured by theInternational Nickel Ckxnpany with Incsloy 800 on the steam side (inside)and nickel on the salt side (outside) This combinatisn contacts fluidswith the alloys that have excellent compatibility aStatus of Heat Exchanger Technology.... ...,.3:*x.:.:.:+ ..iii,....'.!.2+.!...,:.:.'A. .,,:.:.:.:......:.::eWith the backgr~~~~dof MSRE experienee and the subsequent de~elopmentoperation, we are confident of being able to predict salt-side heattransfer coefficients with accurate physical properties available. Muchmore testing and operating experience is needed to verify present infsrmationand the revised salt compositions concerning physical properties,heat tranSfek, and pressure d1POp cOK~@lations, and the avoidance Qf f%OWinstabllity and stratification problems eme experience with Hastelloy N in steam is mostly good, but somequestions must be resolved before a firm conclusion can be reached. "heduplex tubing of nickel-Incoloy 800 looks p~~~i~~fsing: the compatibilityis excellent, and produstion methods for the tubing have been developedso that a high quality product can be obtained at less cost than HastelloyE tubing. On the other hand, steam generat~r fabrication would be complicatedby use of the duplex tubes.a considerable amount of information is available on the use of supes-CHitiC23.1 st@alIl: Over 140 Supercritical preSSblKe Steam geI3ekatOlP UnftS %$@now operating, under construetion, or on order in the U.S., Europe, andJapan [56]. Design of molten-salt steam generators can draw on thisexperienceSome work has been done with analog simulations on the control ofthe MSBR system. The results indfeated that in order to maintain boththe primary salt and the csohnt salt cold leg temperatures above theirrespective freezing temperatures, either of two csntrol schemes would besatisfactory; both require a variable coolant salt flow. One controlSCheme WQUld peQbBire atfZelllperatiOn Of Outlet Steam telllperature b@fOKereaching the turbine and the second would require a controlled bypass..........x.:.....5%

252

*$253...........

254fidence in their use.e - The mcertainties in heate probably less than those ofnce it has been established that saltsere was no evidence of gas film9 uncertainties remain concernratorsurfaces and in the use ofheat exchanger. Verificationphysical properties and resoludedto prodde incresed con-Themal Transients - Transients in operating conditions cause thewd heat exchangers by causing the metre rapidly than the interior. Blthounot be very sensitive to the accurate definitionis a problem which must be more ully explored50 that necessary ~ cco~o~a~i~ns can be made in the desierg and the control system. some work has been done in thethe MSBB reference system on a hybrid eomputer system 1641 ee- The wastageesence of sodium0 - me existin uncertainties in the quantity seation of fission produet depositio on components md pipin(see Chapters 5 and 14) affect not on the heating problem but also themaintenance planning. In particular, e distribution sf fission p~sductsd the heat loads imposed must be evaluated i order to provide for emerofthe empty prfmq? heat exchan er in the event the saltscould not remain In the system.L..

255Fabrication Techniques - Specific joint designs fabrication procedures,and inspection procedures have not been chosen at this timae. Hmever,these present no unusual problems and can be expected to ev~lvewFth the heat exchanger desip. Specific problem will be resolved asthey arise with the partfcipation of industrial firms. The tubing dthenhanced heat transfer characteristics proposed for the primary heat exchanger~epre~ent~ an uncertainty which must be explored. Using unenhancedtubing would increase the heat exchanger fuel salt VQ~UITE about40 to 58% which means a 4 to 7X increase in total fuel salt inventory.Development Program for Heat Exchangerse - n e most important uncertainty concerningthe primary heat exchanger to be resolved by development is that sf materials.The materials problems are discussed in Chapter 9.The next m~st important uncertainty for the primary heat exchange^is whether the enhanced tubing will have the assumed structural and heattransfer characteristics necessary for a low salt inventoq. Heat transfertests with salt would be necessary to verify the structural and heat transfercharacteristics @ The testing should also fnclude the provisions forremote maintenance that are incorp~rated in the heat exchanger design.Steam Generator Materials.:.x.:,.....i' .:.: .$We plan to continue operation of the stem corrosion facility atTVA's Bull Run Steam Plant. This facility will accommodate 688 unstressedcoupons and 13 stressed specimens. The evaluation of Hastelloy N, particularlyin the stressed state, will continue until a conclusion can bemade about its compatibility with steam under stress e Enough infomationalready exists ~neoiogr ~ O Q and ~nconei $00 in steam that evaluationp~ograms on these materials will not be necessary.In order to evaluate the use of Inconel 606 in contact with coolantsalt and steam, tests would be needed to dietemhe the compatibility ofInconel 488 with sodium fluoroborate. The initial tests should be thermalconvection l~~jps, with tests in pumped system initiated if the resultsfrom the thermal convection loops look favorable.'Eke work most needed on the duplex Enc~loy 80hickeI tubing is anevaluation of the mechanical integrity of the product. Each side of thetube is compatible wtth its respective environment namely Incoloy 888with steam and nickel with fluorobsrate. The tests of mechanical integritywill be primarily long-tam tube burst tests in inert gas enviton~~nt.The main questions to be amwe~ed concern the integrity of the bimetalinterface and the resistance of the nickel to intergranular separation.If these tests are favorable, the question of joi~ing techniques must beevaluated. Welding heads currently exist that permit welds with fillermetal to be made inside %/sf in. diameter tubes. The evaluation ofjoint desfgns that utilize this equipment will be a part of our futurepro gram a........>:+:A....y,.:.>>r.. I

256ex&anger and the steam generator.coolant system would have to beplan for molten-salt heated stemdescribed in [67].Task 1. Prepare a conceptual desi of a stem generator for the EOOQ-W(e)MSBR reference plant, This steam generator is to produce steamat l($QOQF and 3500 psia from feedwater at 908°F.....u.:2In addition to the desi study some small-scale tests will be performedand a ste eTaelPator dev~~~p~ent r@pOKt Will be prepared if fUfldfllcase, the plan is to proceed with the prelidnaq deenerator,to conduct tests and evaluations9 and to demonstrate in small tests of properfomabout as expected. Perfomnce testsof models that contain a few full size tubes would be conducted under fullheat flux conditions in a 3-rn facility as described in [69] *The third phase will consist of entests on prototype stemgenerators, The testa would be perfst of the operation of theunits in the steam system of a reactor experiment or demonstration plant.h e or more ind~~trial companies will prepare the detailed desimanufacture the steam ewerator prototypes With the experiencduring this phase, the manufacturer would have demonstrated the capbilbitiesneeded for 8 %e Steam generators for futureneeds of the MsBW PKorequirements ?Kill depend on wBEetlaerase three specifies a modular steam gen-L referencegenerator (stash as prwheeler and

..... .:.=:,....< .A.I.-.,..... ... ii.............. t........,.%....... .:.:.:

.....258ORML-DWO 70-11007.....m.:id 5a HgT.....6 sI

259..... .:.=.......?.!d...?%$condensers operating at about l-1/2 in. Hg abs. Full-flow demineralizersare used to obtain the high purity necessary ~ O K once-through steam generatoroperation a%ter which the feedwater is raised in temperature andpressure to the final 700°F and 3800 psia steam generator inlet condition.me of the two uncsnventional features in 0rnLSs adaptatisn of theTBA steam cycle is the reheat of extraction steam in two stages by use 0%a prime-steam-heated preheater and a salt-heated reheater. The otherunconventional feature is the heating of feedwater to 700°F by directmixing of the 866°F steam exiting the above preheater with the high pressure556°F feedwater leaving the top extraction heater. Two barge motordrivencanned-rotor centrifugal pumps in parallel then ea& boost about19.,680 gpm of feedwater from about 3508 psia to 33800 pia. Eight stepsof feedwater heating are used in addition to the direct mixing ts obtainthe 700°F feedwater temperature. The use of the supe~c~-itical pressurea%Hsws the direct mixing without serious problemEbasco examined alternatives but selected essentially the same stemsystem as that of the QRME reference system 6721. They concluded that theuse of the direct mix8ng for feedwater heating and the special boosterpumps are feasible and within foreseeable technological development eEbasco proposes a weans of improving the cycle thermal efficiency by reclaimingreactor decay heat and chemical processing heat. This heat isintroduced into the cycle as an additional S O ~ K C of ~ low pressure feedwaterheating steam 9731.......*

260Dtcanned-rotor booster p needed for the MSBR stem cycle are about 502K than the EabgeS WIX pmp built to date [77]. Consequently, dementof larger eapaeity9 multistage pumps would be needed. The mixinre also larger than those in use, but this does not appeardevelopment problem since a ch,&er similar to one spec-R system md about four-ffftks %IS large is in US& at thePlant 1781. As an alternative, a high-pressure heater could be used to obtain the 780°F feehater. The exit keatinsteam could be thela heated to 1000°F in a salt-h~~ted ex&anger and reintroducedinto the cycle thereby eliraainating the pressure booster pups eEvduatisnefficiency can be obtained with the supereritfeal pressuresteam system g44,5x per Refe 79) even though the molten-salt steam generatOKITMY TZ@qUilPe %OO"F feedWZi%X%r. T improvement in the efficiency andreduction of capital costs which c realized by lowering the feehatertemperature requirement provides a g incentive for the developmenteffort in this directfon. The low of the steam pressure and 05 feeswatertenparatute could actually be mutually compatible, For example,as discussed in the seetion on heat exchangers, the bayonet tube configurationFapaich was msatfsfactoq at hfgh pressure may be satisfactory withsubcritfeal pressure steam a d lower feedwater temperature Although theoverall plant: efffcierracy would be less wieh the suberitisal pressurecycle, tne simplification aata reduced capital costs be offsettinThe mIten-salt reactor concept is not strongly dependent on the detailsof the steam system. Haasever, since thhe steam-electric equipwent representsmre than half of the cost Qf the plant, the optidzatisn of thissystem merits much development and analysis effort.mere are several options 6aiIZible %O% the st@iXIl system Qf EL KIOltellsaltreactor plant, none of whi involve fundamental uncertainties. Alluncertainties are red to be resolvable with the application sf anadiequateiy t-maed pro of analysis and development, Since the detailsof the steam syste o ae~ela~~~t an the requirements of the stemenerator, it is i e that these requirements be definitely establishedand verified..;.;.;.:. .-

,:......261.... ,:.:.;#;sw,.. .,....A. ,...... ..... . .,...... 51......:.

2626is not restricted and the mcertaiwty of maintaining multiple gas pocketsis also avoided.As far as we hm, no rupture disks have been operated with moltenSalt. bilthOUgh KUpture disks have been Wed 1~6th liqUid-EEtd SystemR program, in most c es the disks have been located in a gasspace. me ma r uwcertaianty is one of ~ e ~ @ sf ~ a ~ ort.abinatfsm ~ ~ e ~ of tmaterials and aCtUatiOa% dlevfee fllZlt Will operateperiods fn contact with molten sa%$, Although the develaysten rupture disk appears feas5b e9 failure to achieve this goal wsuldrequire reversion to the use of a as pocket fob the p.k-oteCtiQn Of thedisk assembly eLittle mcertainty 5s associated with the retention system necessaryto handle the effluent from the relief device. However, 8 developmentmust be established for the csmpsnents and operating proceduresneeessa?ry to provide for separation, handling, and disposal of the gaSeOus,liquid, and sohid effluents resulting from a rupture, The developmentinfOrmEt.t%on needed also hc%UdeS the plPoceduKes d equipment necessaryfor &&2 c%E%3SlUp Of the KelaaaiIIder Of the SeeOnda salt system to allmresmptisn of operation in a dni. ..

263.%.k> ...,...........ustvalve consists simply of a section of salt-filled piping which can be ~cooled when desired to estabP8eh a short section of frozen salt whicheffectively blocks flow. Both the freeze valve and mechanical valve havespecific advantages. The mechanical valve can prodde faster opening andclosing and can be used for flow throttling. The freeze valve prsvideshigh reliability and zero salt leakage across the valve and thus is partfcularlywell suited to applications when tight shutoff is needed forlong periods of time.Freeze valves have operated successfully in the MSRE and in manyout-Of-pile tests at 8 L in sizes up to 1-%82-in. IPS pipe. Kore than125 have been used, and they have accumulated more than 650,008 hr Of ope~atingtime. These valves operated with salts as diverse. as the eutecticmixture of NaBF4-NaF (92-8 mole W), which has a 725°F melting point,and LiF, which has a 1560°F melting point.The operation of the 12 freeze valves in the MSRE demonstrated reliabilityin an actual reactor system [8%]. These valves were of twosizes, lb2 in. and %-1/2 in. IPS, and ascumu%ated 208,000 hr of operationduring the reastorfs lifetime. me of 1-112 in. valves failed and leakeda few grams of salt du~ing the final shutdown of the reactor. The failurewas due io the ma^. stresses caused by a field modification of a shroud onthe valve to aid attachment of a larger cooling line. Suitable designmodifications should eliminate this prob abem in future valves e..iii. .a,.y.x.*.....i...A ,....,:.:

264I~elbsws-seaEed valves have been used suscessfully in many sodium applications.For up to a few inches in pipe size, it appears that similarvalves would be satisfactory for rase in molten salt if fabricated ofmterials. Hmever, there is veryIIth bellows valves larger &and ~ e ~ e Of ~ o ~ ~ e ~ ~dfscussed by SeFm [8lten-sdt salves. Eost sohowever, have had sodim freeze seals on the vais not available for moltela-sa3it valves, Aabthosealed valves for molbten-salt reactors conseq ntly have very little sodiumfKOm, VdVe desi fabrication, and operationZB the future prsdde muchuseful infomation to aid molten-salt valve development.ve techrasIogy in sizes up to 1-1/2-im. IPS pFpe sizer than 1-1/2 iesign to prsvik..,reasonable tiexperience of thesis of a freeze valvepredfetab le temperat~~eIt seem likely that prototypes@actors will have to be exffreeze-thaw cycles eattendant seat-pevelopmerat prsb aeveloped, even ile for use where needed. The1 valves, partie-for molten salts.69s comercia1 technolo y eXiSts fOg fEib3dCatiOn Of BaS%elloy N valvesand bellows at the present ime. HBweVer, there are no Lnmm meta%%urgieaIwhich should prevent fabrication of large valves when reIEffect of uncertainties of the MehmicaB ValvesThe two major uncertainties Fnvol selection of the pabmafLeKiak3 %Or ShUt-Off Valves a d desi and development ofThe seat-seal of the mechanical shut--0 valve requires thatseal force be applied at the seal surfaces and be maintained b7hile thevalve is elused. The valve must also open oft demmd without dmaging theseal SUK~EE~. 1% such a valve proves to be musually difficult ts develop,it would be reasonable for most appIications to use a freeze valve in‘L.X..

265:$&%......:.:.:......:.:.:......:.:.:$Aseries as a backup to this mechanical valve. In this case the mechanicalvalve would be closed to choke-off the salt flow while the freeze plugis established and then reopened to prevent self welding of the seat andplug. This combination has been used in the drain line of some test loops.The valve plug can be mechanically guided into the valve seat OKthe throttling sleeve either in the salt or on the gas side of the penetrationseal. The latter scheme requires stronger aneders but can be doneif the valves are to be located in an area outside of the cell furnace orif the operator-guide can be in a small separate cell where it could becooled such that the bearing materials for the guide are not operated athigh teTllpeKatUr@ eThe mechanical throttle valves will require operators to cont~01 thevalve trim positions. a%hOUgh there are SOme diffePenceS in the requirementsfor these operators when used with a shut-off valve and with athrottle valve these differences are swab1 If a rekkb le throttlingvalve cannot be developed for use in the coolant salt system then theoptimum control scheme for the reactor and the steam p w e plant ~ will beaffected. A description of the use of the throttle valve is given inChapter 10.Development Program Remaining for Valves.:.;.;.;

266Ii

.....:*..... .:.:,a......:.:.:.:,........., ..A,...". 2 .&that separated them from the salt and therefore required auxiliary gascooling. Because the rods and drives were separated from the salt, therewas IIQ problem of gas seals for fission product control, The safety rodsfor the ARE were suspended from magnets which were moved with lead-screwtype drives. This drive was also used f5r the regulating rod. The KSREcontrol rods were suspended from a continuo~s chain drive which was connectedto a motor through magnetic clutches to pernit fast insertion,These rods and drives operated without significant difficulty during thelives of the reactor experiments,The pSiSQn YfkateKial for thhe safety rod Was 134c and stainlesssteel served as poison for the control rod [%I. The poison element forthe MSW cont~ol rod 1971 was a mixture of gadolinium and aluminum oxides,formed and sintered into a number 0% short cylindrical tubes which werein turn canned in Inconel. Vfsual observation of some of the canned elementsat the end of the seeactor operation revealed no significant chanin dimensions OK external appearance.There has been no experience with co~ltrol rods operating directlyin the salt; however, the physical arPangement and proVisionS for coolingmight be similar to that in the liquid-metal-cooled reactors. In oneproposal for a control rod f ~ use r with a liquid-metal-cooled reactor,the pOiSSll Section iS guided Within a duct Which is pOsitiOIIed Withinthe core. The MSBB reference design uses graphite ducts for guiding thesafety rods and the graphite control rods.......A%!......;.A.i....,..... .:.:.a.....A

should be adaptable to XSR use althsar some of them require more headroom than we would like to provide ab e reactor. En any event manyof the techniques will be useful in d ng a drive system for XSBR's..... y...zEffect of uncertaintiesThe use of a graphite rod to provide the reactivity control andsh2ming necessary is uncertain both because of its very Pow reactivityWorth ad the problefla 0% gknidhg if2 Within the COX'@. If the graphiteOd k3 deterwine to be unacceptable, then a poison rod of somewhatreater worth CB ld be used with an acceptable loss in breeding ratisof 0.005 for 8.2% Ak/k pea] eThe use of B4C 89 the poison, with the required heat renaovaP and thee Of the helium produced, might m&e such an absorberaacceptable If so, che BL+C could be replaced wfth rare earths such aseuropLum or adsliniura sxfde which release energy as gama rays insteadof alpha particles, Hs helium is generated and the gama rays should notdeposit as mu energy Hocally as would the alpha particle. me MSBRreference des uses the B4C poison in the safety or shutdown rods andexpects no trouble because they can be held out of the reactor COK~ untilneeded and theref~~e do not have a helium generation OH heating problem.If further study sf the shutdm requirements determines that rapid skutdownis neededs then the rads might have to be inserted near enough tothe core to a a h require consideration of the use of other poisons.. ..i.xG..... _.. .Further Workh y further Work dOtae in the deVE?%OpHlellt Of CXXItKO% and Safety rodsd rod drives shod be directed toward the exact needs sf a specificHnoEten-salt reactor. Req~i~ements such as the exact amount OS reactivityts be controlled, the change rates of reactivity, and the response timesfor the contro~ ana safety system should be established before the meckaanFcaldesi of the system is started. Prototypes of the parts of therelated to the salt- as interface, the gas seals, the guides or8 for the rods Withi the core, and the transmission of motion tos would be fabrisated and operated first in water and air to assistthe design, and then in salt and inert gas to proof testctancy and overall performance. The seals or activity control,l%e bearin materials and arran ement, and the methsd of eoolinthe Control eb2EX2nt Over its entire range of trave%. Would KeCefVe theearliest attentian. After the initial. component testing full scale rodsdrives WOUPCI be tested 'UIX+-X sf~~~~liated reactor ~~nditi~ns...........I1cs.'EvaluationNo fundam@n%al difficulties are foreseen in designing control andsafety rods and drives for use in MSRBs, but a thorough development andprototype testin program will be required to a5sure adequate performancein a high-temperature environment before installation in a reac%oP. mere

269a%-e reasonable alternatives for most of the uncertainties. Perhaps aproblem requiring much attention is the bearing of the moving rod on thegraphite duct and obtaining an acceptable lifetime from such an arrangement.HQW~V~P, because the fit can be loose and the contact pressure low,We believe that SatIsfaCtQ-~?j guides CiLn be deVehp@d.Fuel Storage and Afterheat Remuval~epositories are needed for both kinds of salts based in the MSBR toprovide safe and convenient storage any time they are drained. me fueldrain tank, because of the intense ~adi~a~tivity sf the salt, must providereliably far the removal of afterheat as well as the sure containment ofvolatile radioactive products. In some ways, as diseussed in Chapter 14,it is the major safety feature of the reactor.The fuel drain tank must have sufficient capacity to hold all thepr5mary salt plus the amount of coolant salt that is in the heat-exchangershell; in a double-ended heat-exchanger tube failure, this much coolantCould drain iRtQ the primary system. PUek salt CXIInot enter the Secondarysystem rapidly because 0% the pressure differential between the fcwo systems,and monitoring for fissiun products in the secondary coolant willgive immediate evidence of outleakage, so that the primary system can bedrained before much fuel salt Is transferred. Consequently the cooPantsaltdrain tanks are simple tanks, with no special cooling provisions,located in heated sells.In addftion $0 the primary drain tank, there is a fuel storage tankinto which the fuel can be pumped in case a repair of some palet of theprimary drain tank system is required. This tank has a heat remuval capabilityOf Silly 1 SEnC@ the Salt Can be heEd in the drain tank Unti%the radioactive decay has reduced the heat load to this level.Draining the ReaetorIn normal circumstances, the primary system can be drained into thedrain tank t:kkOU@3 either Of plug Valves pElralhi?l. This EiXTaXIgementwith two valves is, we think, sufficient assurance of an ability todrain the system. The valve plug does not depend on a valve seat forsealing, but is p~~vided with a heating and cooling system for the seatso that the final sed can be made by freezing the salt under the seat.The valves, described in detail on page 51 of ~ ekrenee 99, are locatedin the drain line leading from the Pow point of the primary system to thetop of the drain tank. For maintenance purpusess they are located in thedrain tank cell with an access plug in the shielding directly over them.Leakage of Fuel SaltIf there is a leak in any part of the fuel system, the leaked saltmust be conveyed to the drain tank and its afterheat removal system. Theleaked salt is es$%ected in a catch pan under the reactor whish in turn

2 76drahS ht0 the? dKZh tank pit. &re the Salt enters 8 fUIlIE?lb-bik@ dKaiXIpan on tap sf the drain tank which dfrests the salt to a frangible-diskvalve which can be punctured on demnd to drain the spFlled salt into thedrain tank. No pipe comes belm the catch pan so that unless the d~ainIeaks, all primary salt will. positively end up in the drain tank.rovide for the contingency of a leak in the drain tank, it is enclosedin 8 seeowdaq tank. h y crack or break in the drain tank would only le&salt Lnts the annular space between the drain tank wall and the SUKI~OU~~~FItank eAfterheat Removal.........ys.:;..........~~II

......:.:.y.271....,.i...).../.......:.*,......L.xi,'.:.=....~.....

272x. .....~ ......QIFig. 8.5.EjlSDR - 380 &f?d(e> primary salt drain tank sectional elevation.

..... .y,2.7 3..... .:.-.....',E

2741.12.k3.4.5.as system must be able to acco te various for 0 f csntaminasuchas noble metals, hydro68particulates that may be carriedecomposition pby the purge gas,ducts, or anyather,. ..Gaseous effluent is to be discharge from the reactorplant E7.Positive containment of all radioactive mat@rial must be proriand accident co~~~~~o53s0- an the reference desi 13R [1QQ p xenon-l35 andDare stripped from the f alt by injectinHn sf fuel salt t en from the discharge of the fuel saltpump as shown in Fi 8.6. Just upstream of this pins, bubbles eontafnnoblegases are removed %rsm the salt by the use of bubble separators.as along with my salt entrained in it passes throuto the drain tank [,%SI% an the drain tank, the eis separated from the gas and returned to the main circulation loop byjet pumps. The line running from the bubble sepa tor to the drain tankis provided with a siphon break to prevent d~afna Of the main loop incase sf a pump failure.. This siphon break is in e form Of a loop thatrises above the liquid level n the pump bowl and a venturi that communicateswith the pmp bowl as space thkrsugh a pipe. During noma1s passing through the p bowl is drawn into the venturiinto the drain tank alaS-Salt miXtUKe flSWleseparator etank, thFs gas has a residence time of the order sf Ito 2 hr to permit the separation Of most Of the particulate matter anentrahed salt and to greatly reduce the decay energy [l02]. The gasthen recycled through a cleanup sy tern back to the bubble generator andthe pump shaft. seals. Before the as enters the cleanup system, a particletrap will be prodded if the drain tank is not capable Of separatinla. the particulate tter and entrained fuel salt from the offasthen flms thrs &arc081 beds having a 4s-hr xenon holdupr,........... I&.._si;.,.&>..... ..u.2to the bubble generator, but part of it is cleanshows a gas compressor in the line Of the gas fa.

2 75..........*a:....".yi*>......,:.=.......m.... :;;.9,.L.X, :......_Fig. 8.6.Schematic flow diagram MSBR off-gas system......*2

276enerator, but present efforts are being direeted to developing aenerator that does not require this cogas being recycled to the pump seal pes through a charcoalbed dth a 98-day xenon holdup time and an off-gas cleanup system.The additional charcoal bed is sufficient to 1low decay of practicallyall the noble gases except krypton--85. This aseous isotope, along withany tritium present in the off-gas, is remve and stored in the offcleanupsystem. Although not shown Pn Fig. 8.6, provisions will be madeto remove any oxygen or mofsture that may be present in this recycled gasstream. These provisions are in the fom of molecular sieve beds to drythe gas and titangum sponge beds $8 remve the oxygen. Maketap cover gaswo~abd enter the MSBR fuel salt gas system just upstream of these beds........... ->.l...Noble Gas Stripping. - me emphasis on the use of gas bubbles forremoving the nable ases from the MSBR fuel salt is because of the demwstrata-capability6f bubbles to strip out these gases in the experimentalaqueous and molten-salt reactors that have been operated [183,l04,%Q5,186,107,108] Also, it appears that such a system would result in the smallestfuel salt inventory when compared to other noble gas removal systems eIHa EIO%ten-Sdt reactOrs dth graphite HItgPde%BtO%, the nObk gases tendto diffbase h%o the graphite ;End X'eRlEdXt fZheP@ Until they &?Cay 8% reaetWith neutrons. To limit this, methods are being developed to put on a pyrolyticcoating on the raphite (see Chapter 6) 0 The gas-stripping systemas migration into the raphfte effectively to rectionto an acceptable level. Small bubbles circulatingwith the salt appear competitive because of their relativelylarge surface for core void fractions less tha 0.01 6106 1 e>...% u

271A nets calculational model that considers the effects of the gas solubilityand variations in the temperatures and pressures throughout thefuel salt loop has been developed. The pu~pose of this new model is toCheck the assUlIlptions Used in the pKeViOUS EIlodel and to m&e 73eCeSsary Imodifications. Results using the new model are preliminary, however, and {it is too early to assess the differences in the results produced by the'h?O IIlodels.Results from the well-stirred tank model [l(B6] were eo~related interne of a time constant defined asvcStripping rate = __TIIIWhere v is the W3in fuel Salt lOSp VOlUm63, c %S the fSOtOpE? concentrationin the fuel salt, and T is the time constant. Figure 8.7 shms the eontributionsf xen~n-135 in the graphite and fuel salt to the poison fractionfor both uncoated and pyrolytic coated graphite. The contribution of theXellOla-135 tIae gas bUbbleS t0 the poiSOn fPtlCtion Was about 0.0803 formost cases that were studied since it was assumed that the purge gas enteringthe fuel salt system contained low quantities of this isotope. Thediffusivity and the porosity of the base graphite were assumed to beftz//kar and 0.1, respectively, Increasing this aiffusivity to 10-3 ft2/hrhad little effect on the results even when the graphite was uncsated,since for uncoated graphite the migration rate of the noble gases to thegraphite is eontrolled by the adjacent salt boundalpy layer. A strippingsystem having a time constant of about 30 sec (quite rapid) would berequired to reduce the xenon-135 poison fraction to 0.005 in a reactorhaving uncoated graphite. Putting a pyrolytic coating on the graphitepermits us to use a stripping system having much longer time constants eFor example, Fig. 8.7 shows that a coating with a ft'/hr diffusivitywsblllbd require a stripping system having a time constant of about 300 secto limit the xenon-135 poison fraction of 0.005. The effect of the csathg pOrQSitgf On this Wa.8 fOUIld to be SITd.1. IT this @88e, the Tiate OfXRBgHatiOTl Salt th€? bulk Of the g??aph%te iS COIlt??oll@d bythe Pate Of the d5ffusiCKl thPQU@i %he coating mat@l?ial.The thickness of the pyrolytic coating for the cases shmn sfn Ffg.8.7 was assumed to be 0.011 in. As shmn in Pig. 8.8, varying the coatingthiCknesS CaIl have Et gPeat influence the results. Here fop %L strippingsystem having a time constant of 316 see, decreasing the diffusivity ofthe coating would allow a much thi~~te~ coating. The ealcu%ations alsoshowed that not coating the graphite surfaces in the reflector and plenumregions had a negPigib1e effect on the core poison fraction.For the bubbles to strip the noble gases, the migration rate to thebubbles must equal the stripping rate. If the gas phase resistance tomass transfer is negligible,Migration rate of gas to bubbles =

2 78tde 8.7. Effects sf stripping rate and pyrolytic coating on35Xe poison f faction DBIf

2795....s2>. :. :ss-4zPese........:.:.:.i' ./r...2332(bp390

.........&%aWhere \ 2s the BkaSS tranSfeK COefffCient, is the bubble surface area,c is the Bsotope eoneentratisn in the salt, Cb is the isstope concentra-tion b the gas bubble, W is the Henry's law co%bgtant, R is the gas eonstant,and T is the temperature. Bssuning a bubble diameter of 0.020 in.,a mass transfer coefficient of 2.0 ft/btr, the time constants wereculated for uarisus void f~action~i and fraction of bubbles removed perlOQp cycle. %t was CSSUreed that the gas bdhtg E-etuTXled from Off-gaSsystem was essen ially free of xenon-135. The results plotted in Fig.8.9 show that a .2% circulaefng void volume and a 0.1 f~aetiganal bubbleremoval rate giv 8 8 tine! constant of about 258 SBC. Comparing this withPig. 8.5 yields a tar et xe~lon-135 poi~~nfracti~n of 6.005 if the graphitehas a pyrolytic ssat g O.OPI in. thick andl diffusivity. me kneesin the curves of Fig. 8.3 occur where the partial pressure of the xenon-135ifisant resistance to the transfer of this. ..,x.s.h%en this resistance is not siapp%OXhlated by the PelatiOIIbiwhere db is the bubble diameter and rl, is the void fraction. This impliesthat reducing the bubble diameter and/or increasing the mass transfer coefficientand the void fraction will reduce the time constant. I% willbe seen later that the bubble mass-transfer e erinents indicate that farthe conditisns in those experiments the mass ansfer coeffieient is propartisnaPto the bubble diameter. In this case, the bubble system timeconstant varies inversely with the void fraction, However, as the bubblesmall, it 19 questionable that the bubble ma88 tKaIlSft3r coefubblediameter, Ifthis proves to be true, the bubble system time co ant would decreasewith decreasing bubble diameter below some liaiti size and could possiblybesome very small.me above cah2ulat~om indicate at the fraction of the bubblesrefkacpVed per loop Cycle ahtd/OK the Off as recycle time could be smallerthan the values selected for the refe rice design P'iSBR since the bubblesare far from bein saturated with xenon-135. Because of the uncertaintiesin the model, the parameters above were selected, but as further data andexperi6nce are obtained, it may be possible to reduce these values. ROW-ever, it should be rem ered that there is some small increase in theCore psison fraction asAlso the shorter recycle times in the off-gas system might not greatlydecrease the sum sf the volumes of all of the charcoal beds since even-tually almsst all of the noble gases and their daughters w ill have to beremoved %row the purge as stream before it is recycled to the pump.w......*Z.A

. .......A. I .d2815ORM k- DIN6 72- 8% f 82NO RECYCLE OF OFF-GASMASS TRANSFER COEFFsCIENTTO BUBBLE = 2.0 f8/hBUBBLE DlAMETER = 0.020 in......:&.._,.;

282IIodine Removal. - Since most of the xenon-135 produced in a reactor isproduced indfrectly by the decay of 6.7 kr half-life iodine-135, use ofsidestream iodine strippers has been sug ested as a method for minimizingin the MSBR [EE4,115]1 D %h "effective" solubility of iodine,e ratis of the dissolkved iodide concentration to the sum ofthe partLal pressure of the iodine and hydrogen iodide in the gas phase,has been found to be a function of the salt chemistry as well as theteTTlperatU%e. 'Ghfs "'effectfve" SQlUbilitY quit@ hf.&ls 90 Strippingthe iodine with an inert as stream would require very larof gas. ~o~ver, adding yd9-Qgen fluoride to the Stkippinwouid oxidize the ioaide and reduce its FieffeetidB S O ~ ~ Hcourse the stripped salt must then be reduced before it is returned tothe main fuel circulation Esop.The minimum rate that this fuel salt must be processed for iodineand xenon to achieve various xenon-135 po son fractions in the reactorwith uncoated graphite are shown in Fig. .IO. It can be seen that strippingonly the iodine from a side stream of the fuel salt is not sufficientto reach the tar et psisow fraction in the reactor. However, removing theiodine with the enon is very effective in reducin the poison fraction.nlerefOre, if the graphite Cannot be Sealed, all iodine stripping §y%teElmight have to be used in conjmetisn with some type of a xenon removalsystem in ordeb to achieve a xenon poisort fraction of 0.005.design studies were made of stripping units capable in one63% of the xenon and 6% of the iodine from an 8% fueland in another case of removing 60% of the iodine froma 8.8% fuel salt side stream [115]. In both cases e atmo%pherie pressuregas stream contains 1% hydrogen fluoride to ox ize the salt, andas flow rate through the stripping unit about 80 ft3/seceme heat produetisrn rate af the tripped gases and their daughters alonwith any removed noble metals w estimated to be about 6 m(t). Thiswould imply a fairly sizabPe ga system and the drain tank could not beused as a delay tank because uf the presence of hydrogen fluoride in thisgas. me hydro en fluoride could come in esntact with the fuel salt thatConsideration lIlbllSt be given $0 K@I€IoVE%lb sf the afterheat fn the event Ofa failure in the gas circulation system.Spray towers venturi csntactors , ramp flow units and packed coEumshave been considered for the stripper. AliE appeared to be feasible altkohlghELUch deVelOplllellt WOUld be required FOE. any one sf them. They Wereall fairly Har e mits (roughly 10 ft hi h amd 18 ft in diameter), butthe liquid holdup in most of tkemwas not excessive.Lm.:Ibe - In the bubble stripping system, the purge gase separators and the pump bowls and flows directly to thedrain tank [l%6]. Since this gas contains many short-lived fission products,nonvolatile fission products including the noble metals, entrainedSalt, and deCo€TlpOSed hydfOCaafk20llS Or other ford material, muchattention must be given to the layout of the line discharging from thebubble separators and the pump bowls. This is to assure that this linewill not become plugged with particulate matter and that it will not..... *.&

283I I I I I I INO IODINE REM I I I IiFig. 8.10. Xenon-13% poison fractions using combined xenon andiodine stripper with uncoated graphite.

284IIoverheat even mder reactor accident conditions. Conceptual %ayout studiesindicate that this can be done.To separate the particulates and the entraimed salt and to ~ emo~elarge amounts of ener y generated by these materials and the radioactive88 during the first decayP period, a large VeSSeP having large VS%UEEand heat sink is desired gaa7g. me drain tank appears to me@t theserequirements and the sff- as is routed first to it. Mere, as described@aKl%er h this chapter, ab0ELt 1 mgt) of heat and mos of the nonvolatilematerials are KeIDDVed from e gas stream. After hour holdup,generation rate in the off-gas stream drops f by about twotude due to the decay of the short-lived fission products,le attention must be iven to the development of a drainthat these ai are accomplished and thhgs such as shortcircuiting of the flow of the gas in the drain tank do not occur. h yparticulate material that is not rem~ved from the off-gas in the draintank should be removed by a fil e%. before the off-gas enters the charcoaleds so that thethe bests win not be restricted by the@cmuHation of fot nueh effort has been directed tothe design sf a filter for this se in the MSBR. Probably sinteredmetal would be useerial and much attention would haveto be given to desbe cleaned or replaced and cooledadequately at all times.BE,Noble Gas Absorption on Charcoal, - IBynadc absorption of the moblegases in activated charcoal reduces the volume actually required to allowthe desired decay of these gases. me factor by which this voiume can bereduced from that of an empty vessel or pipe varies from 500 at 8B"F, to58 at 250"F, and to a at 886°F. me capability of the charcoal beds forremoving the noble gases has been calculated from the computer programsdeveloped for the Hrn-2 a d the F%sm [118,113,128,%21] 0 Both of thesepFpOg'gaE18 COnsidep %he effects of the gas and isotope flow Pates, pipegeometry, and te~~peratures on the amount of noble gases removed. A revisedcomputer program is now bei developed, incorpsrating the informationfrom the previous programs, at will opti~ee the design Of thecharcoal beds on the basis of selected parameters.Beat generated in the charcoal beds was calculated by a computerprogram that considers the movement of the noble gases throughout theoff-gas system [3117,122'6. This code assumes that the nonvolatil@ daughtersof the noble gases remain at the location of their birth. Typicall resultsof this program are shown in PiThis heat is remove from the charcoal beds by having the pipes contaiI2ingthe ChaPCOd ilDIlErSed $W dosed tanks of Water. WateP fS EilhWecfto boil in the 47-ha~ xenon h~%d~p bed tanks with the heat: beinfrom the steam in a reflux condenser. However, the 90-day xenon holdupbed must be kept cooler and the water in these tanks is circulated througha cooling unit, Potential problem associated with removing heat thisway that have not been investigated thoroughly are the esnsequenees of aWater leak ifl$O the ChaKCoal beds and What htapgens ff the COOliytg Water

285$3.:.=....RESIDENCE TIME (hrl 0.ot 0.01 0.005 0.01 O.oB5 ( 6 7 4 9 5 2 4DECAY HEAT IMul GAS + OBUGHTERS 0.002 a.oa2 e ooi 0003 0.0014 0.42 0.56 0.59 0.37 0.(7 0.+1DECAY HEAT (Mw) GAS ONLY 0.OOf 0.001 0.0005 0.001 6.0006 0.24 0.34 0.33 0.25 032 0.07.:39.....L:.:.;g“29600 230.137200.004B eo4RECYCLE......A+. pig. 8.11.Typical distribution of decay-heat in the MSBR off-gas system........&,&j..,......m . _............

286Ie - After the purge leaves the 90-day xenon holdupe radionuclides have decayed to neglfgtble valuesexcept for the krypton-85 and my tritium present in the gas. It hasbeen p~opo~ed [I221 that this gas flow through a cleanup system consFstinof a ~ oppe~ oxide bed to convert the tritium to %2Q and low-temperature(0°F) charcoal absorbers where the 3H20 and the remaining kggrptsn a dxenon are removed. A compressor returns the slemed gas back to the reactorthrough the pump seals* etc. These Pow-temperature charcoal bedsare regenerated periodically by passin t of the clean recycled heliumthrough them at 500°F. This helium st then passes through a liquidnitrogen trap where the 3~20, k~ypton enon are remved a d stored.after all of the KadiOnUClFdES have bean ZfeIW2Ved from the purge gasstreamD, there might be some moisture or oxygen remining in this gas stream.Potential points of entry for these eontaminants into the reactor system@odd be illlpbarfties in the Ill.&eUp gas SUpplgi OK a Small leak one sf thecharcoal beds. &ikcular sieves at room temperatu~e (~86'P) have beenused successfully to remove moisture ~KO, the fuel salt purge gas stream[l23]1. These beds Faere HegeRerated periodi@dPgP by pubging them Withdry helium at 500°F. It also has been demonstrated that oxygen can beremoved from the fuel salt pur e gas strean by passing it t h ~ o ~sponge beas at 1200~~ elas] aMSRE Gas Systems..I?X.., .. .

....287.....Y, .,x.,.....,:w......:.%I.... &.&.....&Ya;&,;

28BQBWere Of UXIffOXTB Size Xkdl that their er was constat. Solution of therexations derived fo this model indicated differences in the void fractiondifferent Kt2 ions of the reactor, brat these differences were nota StKOHng fUnctiota 0% the bubble size.d its precursorsiodine-135 and xenon-l35m in these four re iom using the void fraet'dfstrtbutisn calculated by the ~ elatfo~~ d scribed above. A large H%of calculations usin internakPy consist t parameters were made to describethe noble gasThe reasonably good fitthe calculated poisonred values shsm in Pig.was obtained by use sf one set of parmeters.Consideration of th@ solubility of the cove gas was not sufficientby itself to obtain the a .l2. In addition, themass transfer coeffic%ent for the xenon diffusin $8 the graphite had tobe reduced by about a factor of sFx, the bubble tripping efficiency hadto vary with the munt 8% gas ingeste the pump, and an additionalle EE?ChaXkfSm for XePaQW %Pansport rap~lite was required to producewima in the curves.. (Direct inte ion of bubbles withwas postulated in order to provide a process menable to mathematicalese same parameters to describe the transient xenon poisoninUdden pWeK Chandid not accurately reproduceehaVioK. Thus he improved calculatisnal procedure was nota COlTlplete 8UCCe88. b%OKe WOhk i% neCeSSc3Ky to elucidate the XenOHP behaviorin the m and to accurately predict the %a avior in an MSBR.se of the S ~ o ~ t C and ~ ~ the ~ n more ~ s comp sated nature of thisodel for the MSRE, the well-stirred pot insoabubated that they would not contribute as much toas they did in the 0 It is expectedeing developed for ribiwg the noblebehavior in the MSBR will shed some light on this. HBW@TP@F, wore exp@rimentaldata probably will be required before this situation can be cornpletelyresolved *ated@neKdlY opernuisancetype problem were encountered9punp bob71 produced 8 lXkt having CQI1-centrations similar to any ordinary mist. Some of the salt mist driftedinto the off-gas line at the rate sf a few gram per month, and some ofthe b"@SUabtant Sal$ depClSitS had to be Kf@mOTPed at iIlteETak3 Of Six KlOnthSto a year. Foreign materials, partieuParPy those resulting from the fuelpump lubricating ail leaking into the fuel salt famd their way to thecontrol valves 9 flow restristors 9 and filters a These item graelualbly becameplugged by these polymerized OK~ZXXLCS and they had ts be cleanedout or replaced periodically a This problem was corrected by installingnew filters of a modified design [130, 13Hq. Considerable attention was

Ia289....&.&....&$ORNL-DWG 74-483......a....MCURVES CALCULATED.POINTS MEASUREDoHEb1UM COVER GASeARG8N COVER GAS 1........>&00 0.4 0.2 0.3 0.4 0.5 0.6 0.7CORE VOID FRACTION (Yo)Pig. 8 e 12 e MSRB s teady-state poisoning by 35Xe during f~ll-p~we~operation. Calculated cu~vee are based on the "lumped, sshble gas"model, with adjustment of the parameters to achieve the agreement shorn....&....a;....

290where Sh is the Shemood modulus based on the pipe diameter, Re is thes madulus based 011 the pipe diameter, Se is the Schmidt mddus,Gs is the Sauter mean bubble diameter, and D is the pipe diameter. Forthe conditions studied so far, it can be seen that the mass transfer eoefficientis directly proportional to the Sauter mean bubble diameter,dvss Pipe diameter, D, was included as a nondifnensianalizing parameteralthsugh no actual variations in pipe dimeter were made. In vertical.

................sB103 2 5 10" 2 5 2PIPE REYNOLDS NO., Re d G q VFig. 8.13. Typical mass transfer coefficients sbtabed for strippingoxygen from aqueous glycerine SO~U~~ORS.

cP292

293;.= .....;"" .,Ya.......m.....(:&.....%,a....,2%,s...U.i%....,%,. . . ..... . .L.&channels where buoyant forces become significant, it can be seen in Fig.8.13 that the mass transfer coefficients pass through a minimum and tendto approach the mass transfer coefficients for bubbles rising freely ina liquid.These results are a little surprising in that they are a little loverIthan anticipated and show a direct dependency on the bubble diameter.IThe reasons for this are not clear and more data are necessary. Additionaldata are now being taken using another size test section. ~xtrapolation 1of these results to bubbles in the fuel salt may introduce uncertaintiessince the salt physical properties amd chemistry are different from thosefor the glycerine-water mixtures 0When first considering bubble generator concepts for the MSBR, btwas fomd that very little infomation was available in the literatureabout th5s subject for systems similar to the MSBR [l(B6]. This unit shouldgenerate a uniform dispersion of bubbles over the pipe area to give thedesired void fraction and bubble size distribution. Also we decided thatwe would attempt to develop a unit that would use the available gas supplypressure, since we preferred not to design and build a gas compressor foruse in this environment e Both mehanicai and fiuid-pmerea generatorsWere COnSider@d. We decl&ded that the fluid-pOWer@d type gC?neKator is thebest approach sfnce no moving parts are involved. This led to the selectionof a venturi type device in which gas is ingeetea in the venturithroat and bubbles are generated by flufd turbulence in the diffuser section[ 135 $106 ] * "Teardrop" and "multivane" type bubble geRC323tOrs werealso tried and appeared to be capable of giving satisfactory performance,but we believe that the venturi concept offers the greatest simplicity.Tests on MSBE size venturi type generators using demineralized water withand without the addition of surfactants aqueous glycerine solutions, andaqueous calcium chloride solutions indicated that the size of bubbles generatedby this; unit was not very dependent on the nature of the fluid.HOW~WX, a theoretical analysis indicates the diameter should vary as the0.6 pmer of the ratio of the surface tension to density, (c/p)Q*Ge Additionaltests are in progress to verify this relationship.A most important observation on these bubble generator tests in aloop is that the bubble size in the loop is affested strong%y by the kydKaUliCSSf the Circulating pWp and Of the circulating system [1%,106]."he circulating pump in rhe best loop sheared tke bubbles to a very smallsize (about 0.081 to 0.802 in. diameter). These small bubbles coalescedquickly in demheralized water, but not so quickly in the aqueous glycerineand calcium shloride solutions These observations indicate thatthe effect of the pump on bubble size and the coalescence of bubbles mustbe studied in fuel salt before the final design can be established forthe bubb Be generat or1The in-lfme centrifugal gas separator developed for the aqueous homogeneousreactor program showed great promise of meeting the MSBR reguirements[106,l28,137,138]. These requirements are that the gas removalIefficiency must be high over a wide range of gas inlet flms, the pressuredrop must be compatible with the pressure drop available in the primarysystem, and the mount of entrained liquid in the g= removal streammust not be excessive. C~~~~iderable effort has been directed to developingit MSBE size separator which has a &in. inside dimeter and a salt

294flow rate of 408 to 550 pm Kl06,%39,14Q,l41,1421. Figure 8.14 shows thetest unit that has been eveloped that appears to give a suitgas removal effieiency. This unit has a 44-211. separation letapered casing, and gas removal at both the swirl and recovery hubs.The early test units pe~fomed well when using demineralized water,'$Ut had poor gas SepaKa$iokn efficienC!ies When Using aqueOUS glycerine OKcalCiUl3l Chloride SCllUtiofaS With khematic ViScOsitieS Shdb3K to that ofthe MSBR fuel salt. we postulated that the higher kinematic ViSCOSitY?reSU%ted in a redlaced radial bubble Vehcfty to the CeIItKal Void Of theseparator and an inc~ea~ethe effect of the fluid turbulence. The pumpproduced much smaller bubbles which did not coalesce as ~eadilgr in theseaqueous solutions as they did in the demTneraPized water and the smaller....L.A.aratos unit ave better separation but resulted in a vortex instabilitythe mount of gas that couPd be injected and removed fromthe test fluid. By tapering the casing and reasssvin gas from inlet andoutlet hubs the separator was made capable of hmdl ng a%% of the anticiuidand gas flow rates.ain some mderstmding sf ubble formation and coalescence in theds and in molten salts, a hakr test Kig WEkS bUFlt [l42]. Transparentcaps Ul es con t ain is these liquids were shaken to given the desiredbubble size and void fras ton. The shaking was stopped and photographswere taken at frequent Fntemals to study the bubble behavior in theliquid. Tests SO far indicate that small bubbles coalesce very quicklyin demineralized water, but very slowly in aqueous glycerine and calcium&loride solutions. A very significant void fraction of small bubblesremained in these sslutions 28 see after the sh&in had stopped. In66-34 mole x LiF-BeF2 salt, m e was not as rapideralized water, but more rapie glycerine and calciumchloride ao%utio s. The void fraction of saksel3b bubbles in the salt 20 secafter the shaki, Rad stopped was very small, indicating that there wouldaration from the mw fuel salts than from the calcium8. This indicates that the probability of th& presenting the desired performance in the MSR fuel salt systemEffects sf Unkaoms and UneertaintiesiII

295l~ops with aqueous solutions that have ‘kine~tic viscosities similar tothat of the fuel salt. The shearing of these bubbles into finer bubblesby the pumps in the loops suggests that the bubbles in the MSBR fuel saltwill be smalle~ than we have assumed in our calculations, This observationand the indications that bubbles cf~culating in the MSRE fuel were verysmall led Ebasco to suggest that elaborate bubble generator is not requiredfor the MSBR [14%,144,%45]. Also these smaller bubbles would resultin a larger bubble surface area for a given void fraetFon and Ebasco suggestedthat the desired PQW xenon poison fraction might be achieved withoutSealing the graphite mOd@Ka%oP.We believe that we can generate bubbles in the fuel salt satisfactorily,but this still needs to be demonstrated. The shaker experimentsso far suggest that bubbles in the fuel salt will ~cpalesee more readilythan those in the aqueous ‘test solutions, tending to make the bubbles inthe fuel salt larger than those observed in the test solutions in the Iosps.Smaller bubbles will provide a greater surface area for a given vodrnrneof bubbles to compete for the noble gases in the reactor. .Although thebubble mass transfer experiments show that the bubble mass transfer coefficientis proportional to Sautes mean bubble diameter over a very limitedsize rage, there are indications that the relationship does not extendfar outside the measured range. Consequently, it may well be that use inthe fuel salt of bubbles smaller than those in the experiments would result%n a higher sate sf noble gas transfer. However, our present. knowledgedoes not provide proof for such a conclusion.The bubble separator, like the bubble generator, has been developedto operate satisfactorily in loops using the aqueous test solutions but isyet to be tried in a molten-salt system. While we are quite confident thatit will give the desired separation in fuel-salt system even with smallerbubbles, it still has to be prsven.If for some reason any part of the bubble system does not work, somesort of side-stream gas stripping unit would have to be installed in thefuel salt system, The pemalty for this probably would be a larger fuelsalt inventory, much larger mounts sf gas would have to be handled, anda gas co~~pre~s~r would have to be deve%oped for use in a high temperatureand highly radioactive environment.As discussed in Chapter 6, considerable success has been achieved inputting pyrolytic coating on graphite, but a coating that reliably retainsits low permeability when irradiated has Rot been demonstrated. Althoughwe are optimistic that a satisfactory coating can be developed, we mustconsider the consequences of failure to accomplish this goal. Also thereduction in the cost of graphite would be substantial if it did not haveto be coated. As stated above, smaller bubbles wight be able to strip thenoble gases from the fuel salt to the desired level even without theIgraphite being sealed, but this must be investigated experimentally. In- Icreasing the amount of bubbles in the salt to a core void fraction of 0.02 Ito 8.03 has been suggested as a means of limiting the poison fraction 11439Ithat might compensate for lack of graphite sealing. The 0.01 void fraction 1selected by ORHL is not a firm limit, but it is believed to be reasonablebased on the MSBR design requirements and the MSWE operating experience[l46]. Also with 0.02 to 0.03 void fraction in the core the void fractionin the salt entering the circulating pump would be around 0.18 which isundesirable.I*Iit

296IIt was shown earlier that isdine strippin PIUS XefaOF1 strfppiwg Qfa fuel salt side stream has proaaise of achievi g the desired poison fracevenwith uncoated graphite. This method has the several dipreadvansthat were discussed, but it probably could be made to work. Thisprocess probably would be mre e enaive than the bubble stripping systemand would require continuous oxi tion and reduetion of the sidest~eam andcareful €mWitOrin of the redox eofsditiom of the fuel salt.The fraction of the noble-metal fission produets that will find theirway into the MSBR off-gas system %a not ertain; it has been estimated tobe frofn 5 tQ 35x OH1 the basis Of %lesm e erience %1471. In any ease, theoff-gas system must be desi ned to prevent this material from accumulatingin depeasi%s that would restas flaw OF overheat parts of thee sf and other particulate II%att@Kthat could be carried into the off-gas system is ratneeptain, but use of ilfilter OPI the dkah-tank Will pKOb&%blby be deSLrk3ble to thatthe operation of the rest of the off-gas system including the ckareoal bedsiS Sa$b%%C%Qv eI" A-bFuture WorkVork is continuing on extending our knowledge of mass transfer toe8 fkswing in Hfqui so The bubble mass transfer experiments are nowkUIl UShg other Si e test sections in the experimental loop. Futureefforts are beimg planned to investigate mass transfer to bubbles in areasthe highly turbulent ~egi.011~ in the reactor such as the elbows,ansions, and the pump volute. Attempts w5Pl be made to investitransferto smaller bubbles e @cause the present method forthe bubble size is not capable of identifying very small bubbles,a method nust be improved 0% another nust be aevised. MEGStransfer rates from the fuel salt to the bubbles will be determined in aSSP OW being built and discussed belm. we may have to fse satisfies withjust the PafOdUCt Qf the EELS% tPEXisfe% Coeffic%f2nt a d the bubblesurface area in this Poop if methods are not devised to measure the bubblesize in flowing salt streams.Work also is continuing on studying the behavior of bubble generatorsand separators under various flow condit ons e Pressure drop, vePocity,and flow stability measurements are bein made. New units will be buflt,tested in the fuel salt test loop now be ng constructed, and modified ifnecessary to obtain the required perfsmance sThis fuel salt loop called the Gas System Technology Facility (GSTF)[nlc$,lsss] w ill be used for tests of the MSR fuel salt and off-gas system.In addition to the tests mentioned above we plan to investigate nobleIEX38 tPXE3fer to graphite, the @f%eCt Of the eoIIQositiOl8 Of &he Salt QTIthe Vah-htlS test KE?SU%t%, pe~fo?ZRaIlCe sf the? Off-gaS System COIllpO~eHnts,and the behavior of the noble gases and noble metals in this system.Various components of the off-gas system that must be tested beforethey are built and used in a reactor system include the salt mist sepa~at~~,filters, and units to remove and store tritium and krypton-85. Also testswill have to be made to assure ourselves that we can sat2sfactoriIy predictwhat will happen to the noble metals and any other particulate matter inthe off-gas system.

299CWe~all EvaluationPresemt evidence leads us to believe that removing the noble gasesby use of bubbles in the fuel salt is the best approach to achieve thedesired poison fraction in molten-salt reactors. Experimental data indicatethat we have a workable bubble generator and separator and thatthe mass transfer rate of che noble gases to the bubbles is acceptable.It may be shown in the future that turbulence in the pump impeller orother parts of the fuel salt system could reduce the bubbles to smallsizes, but it is premature to assume that now. Use of smaller bubblesmay eliminate the need for sealing the graphite to achieve the desiredpoison fraction, but this cannot presently be assured so work should continueon sealing methods.Use of a side stream stripper to remove the iodine and xenon fromthe fuel. salt appears to be a feasible alternative for controllinnoble gas poisoning, but is not as attractive as using bubbles unless itis required for tritium removal.Data are available to show that all the components in the off-gascleanup system are workable. Care must be taken in the design of thissystem to assure that particulate matter, ~adioactive and nonradisactive,does not accumulate anywhere in this system such that it would retard thegas flow or cause overheating.Although much development and testing of various parts of the moltensaltreactor gas-handling system have been and will be done, the systemwill have to be proved in an operating reactor. Many subtleties andinteractions affect the performance of this system, and a ~eactor is thesmly place where we can demonstrate that they all have been conside~ed.The major uncertainties of a fundamental nature in the componentsand systems for MSBRs are related to the provisions that must be made for~~co~t~~dating the widespread radioactivity anel to the exact distributionof fission products and associated heat production centers around thereactor system. These uncertainties affect the mechanical design sf manyof the ~~mpo~~ents and system such as the off-gas. Although the MSBE or ,a first demonstration reactor would have to be overdesigned in many respec~s,it sho~ld be possible to build and operate it safely and reliablywhile obtaining basic infomatisfa needed to optimize later plants. Theoperation of such 8 ~eactor is an essential part of a program to providethe technology for EISBRs.A summary of the evaluation of the status of each of the importantcomponents and system is given below.Operation of the pumps in the ARE, the =WE:, and in the large-scalepump test loops demonstrated the reliability of the pump designs a d thepresent state of the technology for salt pumps. We know how to makereliable short-shaft pumps having capacities up to 1500 gpm for m e inmolten-salt reactors. Although it may take several years to produce thelarger pumps for intermediate- or fdl-scale MSBR'S, the problem a%e welliII

29 8understood, and there is little question that satisfactsry pwps can beobtained on a sehe ule conpatible with obtaining the sther principal reactorcomponents. The pump pX'ogr?Xil Will benefit froan %he LmBks Sodiumpump technology program because pmps for liquid metals and for moltensalts have many COKZMXI requirements.Our experience with over 11,000 hr of operation of an I9.Slng sodium fluorsborate and with the operation of the- and farced-convection loops for material compatibilityom that we can safely handle the salt and that it is agood candidate for the coolant salt for malten-salt reactors. Furtherwork is needed EnethodS fQr earlby detectiCKl Sf EtlC?istUaf@ inleakage, Corrosionpr~duct handling, and the migration of tritium within the coolantsystem, At this time webasic problems which would cause elirriinationof sodium fluoroborate as the coolant salt; however, there arealternative salts that have only slightly less favorable characteristics.The experience gained from the heat exchanger studies in the &a?Program and from the design, fabrication, and operation of the heat exchangersand air-cooled radiator in the &ERE has show that the moltensalts behave as conventional heat transfer fluids amd that, except forthe enhanced heat transfer tubes, none of the uncertainties associatedwith heat exchangers and steam gen rat~rs 8r& fundamental. All UncertaPntiesare related to engineerin design optimizatisn and to maintenanceahpa are considered to be resolvabl in a reas~nable fashion with an adequatelyfunded pPogPiiw s% analysis and development. The review made bythe Holten Salt: Group headed by Ebascs Services reached the same cornclusion.an5the-n: study devoted to stem enerators now unde~ way by theFoster Wheeler Corporation is expected to result in conceptual designsof steam generators for use with molten salt and to descriptions of thedevelopment pregrams needed to produce them.The studies sf steam systems f ~ use r with molten salts conductedby ne~~be~s of QRXL and Ebaseo Services have shorn that the conventisnaPstem system of an existing modern steam power plant can be adapted %ofmolten-salt use by the additisn of only three new components. These area reheat steam preheater in the intePrtledia%e-pressure-~~rb~~e circuit,chamber f o using ~ prime steam for the final stage sf feedwaterand booster pumps for delivering this feedwater to the steamgenerator, We found also that changes were needed in the startup systemto provide for the high liquidus temperature of the salt. Several sptionsare potentPalLy available for the steam system for a molten-salt reactorplant, none of which involve fundamental uncertaii-fties. All engineeringuncertainties are considered resolvable with an adequately funded p1-8gra111of analysis and development.The experience with the design, fabrication, and operation of controlrads md drives in the ARE and the PISRE provide information which &rill beusef~bi in the design of rods ana drives for the ~ B R . me IIB~OP differenceis in the need to operate the rsds in the salt in the MSBR to providecooling and reduce the parasitic EQSS of neutrons. Control rods anddrives for L?.ZFBRss have this design feature and much of the informationproduced in developing thew should be applicable to the development ofrods and drives for MSBb. Although a thorough development and p~~totypeprogram will be required to assure adequate performance in a hiu..k..,

.....i .299.....:.:E3.........cw,l%..:.=,....:;*z...,.... .:.:.d.......A i..... ., %.. i..........-.>.,......., >,.;

.I'.... *,.,References for Chapter 8....I.... e2L2. MoZtePz-Salt Reactor techno lo^, Technical Report of the Molten-SaltGroup, Part I, Ebasco Services Inc., December 1971.3. EvaZudion of a 1000 ,%de ivoZtere-SaZt Bmeder Reactoa?, Technical Reportof the Molten-Salt Group, Part PI, Ebascs Services, Inc., October1971.5. Design Studies of Steam Generators for Molten-Salt Reactors, MonthlyProgress Report /k3 Feb. 14-MarcHz 31, 1993, Foster-Wheeler Corp~r;atiQnDOCUTRefnt NO. NDl72622..7. E. s. ~ettis, L. G. Aiexmder, and H. L. Watts, Desip Studfes of aMoZten-Salt Reactor Demonstration Plant, 0 E-TM-3832 gsme 19S%> -9. Conceptual Desip St~dy of a Single-Fluid Molten-Salt Bpeeder ReactoT,OmL-4541 (1971) s14. EuaZv.atim of a 1000 [me M~lten-SaZt Breeder dieacto~~ Technical Reportof the Nolten-Salt Group, Part 11, Section 6.gP Ebaseo ServicesIns. October 1971.

30116. Summary of Study of Feasibility of 80tor-Be~t5~1g System for a 1250 kpMolten-Salt Fuel Bump Conducted by PTTI on Subcontract 2942, Intra-Laboratory Correspondence, A. G. Grindell to I%. B. Briggs (June 27,1968) *19. P. G. Smith, High-Temperature Molten-Salt Lubricated HydrodynamicJouma1 Bearings, ASHE T~msactio~~s - Lt9 263-274 (1961).20. PBR Prsgparn Semiann. Prog~~ Eept. Feb. 28,2962, OWL-3282.21. Oak Ridge National Laboratory, Reactor Division Job Specification61I-%0544-RB-O01-X-l (July 21, E969).22. A. 6. Grindell and 6. K. McGlsthlan, Conceptuai System Desig~ Deecriptionof the Salt Pump %st Stmd for the MoZten-SaZit; BreedeyExperiment, OKTE-TM-2643 (August 1969).....*y......!.&. 2%. E. V. Wilson and A. 6. Grindell, P~eZiminaq Systems Design. Desc~ption(Title d Besicpi of the Salt Pwnp Test Stand for the IdoIten-SaZt Breeder Eqeriment, ORNE-TM-2980 (December 1969).24. A. G. Grindell and R. B. Briggs, Frogrm PZm for the Pmcu.rementWLd Testing of I!!SBE Salt Pwnps, MSR 69-94 (Oct. E, 1969).25. J. P. Sanders, A Review of Pos~L5Ze Choices for Secondaq C~oZant~fop Molten-SaZt Reac%op?s, internal memorandum, Au.....:.:,.............:.:.>>29. W. R. Huntley, B R Program Semimn. Pmgr. Rqt. Aug. 32, 1974,QRNk-4728, p. 152.

30238. R. B. Briggs, An Assessment of the Effects of Leakage of Water fromthe Stecun-fi'aising &pipent m Corrosion in the Secondmg System ofan MSBR, Iwtra-Laboratory C~~resp~ndence QRNL-%aSR-71-27, March 191971.33. Conceptual Besipt Study of a Single-Fluid Molten-Salt Breedem? Reactor,0mL-4541 (1971), p. 54.34. Ibid.. p. 65.35. Ibid., p. 69-36, Ibid., p. 81.42. Gesign Studies of Stem Generator5 for Nolten Salt Weactom, M~ntklyProgress Report fa, Feba 14-March 31, 11972, Foster Wheeler CorporationBQCUIRent NO. NB/72/22.43. Project fop Iszznesttgation of MoZten SaZt Breede~ Reaeto~, Final Reportof Phase 1 Study by Black an? Veateh Consulting Engineers,September 1970, p. 70.II44. Conceptual Desi Study of Q Single-PZuid Mo lten-Salt Bm?eeder Reactor,QrnTL-4541 (1971) > p. 54.45. 5. R. HeWherter, Molten Salt Breeder Eqeriment Design Bases, ORNE-TM-3177, November 1970, pa k8.44. E. S. Bettis et aZ., Design Stzldies of a MsZten-Salt Reactor BemonstrationPlmt, OWL-TK-3832, June 1992, p. 26.

......s;?303......:.:.ya-_..... ,;.:.:.> ...47. 1000 Mie) Molten Salt BPee&r Reactor CmceptuaZ Design Study, FinalReport -Task I, Ebases Services Inc., February 1972, Pig. 4.23....WJ49. e. H. Gabbard, Reactola $over Measurement and Heat Transfer Pe~fsrmaneeip2 tk Molten Salt Reactor EqeKrnen-t, OW%9E--'1%.f-3002, Kay 1970.,$

30466. %bid., p. 156.67. Plan for the DevsZopment of Stem Gene~atsrs $089 Molten Salt Reactors,enclosure with letter dated Bee. 11, 1978, from %). B. Traugerto HiBtsn Shaw.$0. Molten-Salt 8easztor ?'eehmlogz~, Technical Report of the &ltew-SaltGroup, Part I, Ebases Services Ins., December 1971, p. 53.71. i%meptuaZ. Design Sttidg of a Single-Fluid Molten-Salt B~eeder We-L-4541 (1971), p. 74.73. Ibid., p. 5-28.45. Purchase Order Subcontract No. 91X-886786 between Union Carbide HuclearDivision and F~~ter-Wh@eEe~ Corpo~atfon, Task I1 sf DesignStudies for Molten-Salt Reactors.C...-.74. EvaZtlation of a 1500 &Ne hfoLten-Salt B~eeder Reactor, Technical Reportsf the Molten-Salt Group, Part 11, Ebasso Services, Inc., Ostsber1971, p. 118.80. Pbid., p. 123.L.

....305......:.:.

30699. E. S. Bettis, L. G. KLexawder, amd H. L. Watts, Design Studies of a:MoZten-Sal75 Reactor DemonstPation Plant, om%-TP1-3832 (June 19 72) 0ea, P. N. Haul3in the 16m, 6, and A. Houtzeel, SpPay, Mist, Bubbles3627 (June a9so),105. J. R. Engel and Steffy, XePzon Behmisr i-n the MoZten-Salt Re-L-234-3464 eoet. 1971) e109. Co?aceptmI Design Stud2 of a single-Fluid Molten-Salt Bpeedefl R%-L-4541, p. 61 (1971)..C.&>.........

.....:.:.:.$.... I.... . .,. . . :.A,. %. ..,y..*.. ~..,..........'.:

134. %. S. Kress, Mass Transfer Bemeen Srnal1 Bubbles and Liquids inCocurrent Turbulent Pipeline Plm, QREJ%-Tb%-3718, April 1972.135. mR FpIopm Sepsaimn. Prop* Rept. Peb. 28, 1971, OWL-4676, p. 49.137. J. A, Hafford, Development of the Pipe Line Gas Separator, OWL-1602, Pes. 18, 1954.138. F. N. Peebles, The Xoti~n of Gas Bubbles in the "2 Reactor Core,ORESL-1E71, Jan. LCs 1952.49-51 0Semitenn. PP0p0 aqt. Feb. 28, 1971, OWL-4676, pp.146. J. a. En el, persoqal communication, July 5, 1972.I

....;*>A?.+2 .....'.A y, .%. .9. CELLS, BUILDINGS, AND CONTAIWNTE. S, Bettis......+. .y *Requirements.....'is.' . !.A.... ..,. .*.. The molten-salt reactsr plant differs from others in that bargemounts of fission products are dispersed throughout the reactor primarysystem and ~eveb-al auxiliary systems. Much of the equipment and pipingmust be preheated and held at l800"F or higher to keep the salts molten.Maintenance of the radioactive equipment must be accomplished throughshielding by use of remotely operated tools. Facilities MUS^ be providedfor the handling and storage of fission products discharged from the fuelprocessing plant and for materials removed from the reactor systems. Becauseof the differences, considerable attention was given to the containmentrequirements in the NSBR reference design and to a reactor buildingand equipment cells that would satisfy those requirements.The major criterion is that double containment be provided at alltimes for equipment that contains the bulk of the radioactive liquids andgases. The inner containment must remain sealed at all times when theequipment therein is operating. The outer containment must also be sealed,but controlled acsess through airlocks is permissible at any time when theradiation levels are below the levels specified for human occupancy.When maintenance of the doubly contained equipment is necessarys thebulk of the radioactive fluids must be drained or purged from that equipmentamd secured in tanks in other sealed cells. The inner containmentcan then be unsealed and openings made as necessary to accomplish the maintenance.A degree of inner containment must be maintained by limiting thesize of opening and providing a flow sf air inward th~ough the opening.Equipment that is removed must be withdram into casks for transfer torepairp disposal, or storage facilities within the outer cantainment.During all these operations the outer ~~nt~iinme~~t must be kept sealed.All ventilation streams must be filtered, passed through absorbers, andrecycled where feasible. Gases that are discharged to the atmosphere myCOntaiHl SRPY trivia-? ElIEloUIlts Of fissioll pKldUCt radioactivity.DescriptionThese general requirements were applied in the MSRE, although lessstringently, because much less radioactivity was involved. The MSRE fuelcirculating system was installed in a reactor cell that was a steel tankimbedded in consrete. The fuel drain tank system was similarly enclosedin a steel-lined concrete cell. Salt-containing equipment in the cellswas enclosed in insulation that contained electric heaters to maintain thesalt above the melting point while the cell atmosphere was held to 150°Ft

y space CQQP~~S. The top closure of each cell was f~med by a steel~drane sandwiched between two layers of concrete shield blocks andwelded to the tank walls. The reactor and drain tank cells were insidethe reactor building. This building was a steel frame structure thatwas covered with corrugated sidin and lined with steel sheet to makeit moderately tight 0When the reactor was operating or maintenance was in progress, andat most other times, the reactor building was closed and the interiorwas kept at a slightly ne ative pressure by drawing air from the buildingthrough filters and discharging it up a stack. During operation the atmospherein the reactor and drain tank cells was maintained below 5 percentin oxygen by admitting nitrogen and at subatmospheric pressure byexhausting a small stream past a radiation monitor and tap the stack.When a cell was opened for ~~~intenance, the size of the opening wasninimized and a flow of air into the cells was maintained by means of aline from the cells to the building exhaust system. Materials removedfrom the cells were withdrawn into casks or bags for storage in othercells in the building or for removal from the plant. The containmentsystem worked well and provided satisfactory protection for the publicand for plant persQp%nel.....

MSBR Cells..... ....,....,.a............83.V.....:*,>>.. ..:.yAll cells that contain radioactivity have thick concrete walls forshielding and are lined with metal to provide a tight containment, Thesecells also provide a controlled atmosphere environment for the equipment.The reactor cell and fuel-salt drain tank cell require both heatingand cooling. They must be heated to about 1080°F before the reactor canbe filled with salt. Being able to cool them is useful for removing radioactivedecay heat from the primary system after the salt has beendrained and for lowering the temperature before doing maintenance. Forthis reason, these cells are lined with thermal insulation and have aforced circulation closed gas system in which the gas can be heated orcooled. The off-gas cell and the chemical processing cells have heatersor coolers on the components themselves, and the ambient temperature ismaintained at about 100°F by space coolers.Building the reactor and drain tank sells as ovens introduces severalproblems. The insulation must be able t~ expand and contract withthe heating and cooling of the cell and be effective for at least 30years with little repair. Some equipment supports and restraints mustoperate harmally at 1000°F but be able to accornmsdate occasional coolingto 200°F.mosphere.Reliable blowers must be provided to circulate the cell at-Industrial blowers are available for circulating gases inovens at temperatures up to 200B"F, but they probably will have to beupgraded for reactor service. Special attention will have to be givento penetrations through the cell walls and to disconnects for instrumentand SerViCe IfneS. POX' 56m6 SePviCe the inner ends Of the peRetratiOnSand the disconnects my be at high temperature. POP others the penetrationswill terminate behind the insulation at the inner ad.% of the celland be at ]bow temperature. Some specially cooled thimbles may be requiredto bring nuclear instrumentation close to the reactor vessel. Coolingmust be provided in the cell walls to remove the heat that passes throughthe insulation and that is generated in the walls by nuclear radiationsin srder to keep the concrete at low temperature.The benefits.derived from installing the reactor equipment in avensas opposed to putting the insulation and heaters on the equipment andpipi~g are substantial. Insulating the walls is potentially much lesscomplicated and less expensive than prcsviding many specially fitted andremotely installable pieces around the vessels and piping. In the oventhe exterior surfaces of the reactor equipment are fa^ mre accessible forremote maintenance and inspection and for installation of instrumentation.The multitude of electrical cables and the~~~ocouples and associated disconnectsand penetrations that are required for individual pipeline andvessel heaters are eliminated. Although the heating systems on the MSRE >; ....

the cell, passed through abs~lpbers and filters, and reeyeled t~ the reactorbuilding or discharged up a stack. Radiation instruments monitorthis effluent gas to prevent release sf activity QUtSide the buiikdhlg.%he bottom Of the Heactor cell ~OKITLS a Catch pala Which CQnneCtsthrough a line to the drain tank. Any break in the prinaa~y salt circuitwill result in the leaked salt being conducted to the drain tank. Thisfeature is discussed further in Chapter 8.The stem calls are located outside the main building containmentmembrane. This is to avoid any possibility of a stem leak gutting pressureon the containment structure. Because the S~COE-L~~K~- coolant is radioactiveand toxic, these cells are als~ sealed when the equipment isStatus and Uncertaintiesfeneral design of the reacto~ containment building for an MSBRthe design that has come to be rather standard for nuclearand m unusud construction techniques seem to be involved inhe cylindrical shell OH the various cells. Heating the reactorthe Salt, kiCWeVer, iS UlSniqUe.The method employed for heating requires no inventions and shouldhave minimal troublbes. The uncertainties winlly relate to the method ofin~ulatirtg the cells and mkiwg the penetrations, and, as discussed inChapter 13, the way to support the reactor components in the hot celland to restrain them a ainst seismic forces. No limiting problems areforeseen, but the desi in to be €dly worked out. A c~nsicteamount of developwent w needed in proving the design. Theapproach Of Separately insUlEitiHl and heating the components can beadopted if the diffiCUB'kk?S Qf U ihng the Sells as BVefbS prove to be toogreat. Scaling up the MSW method for an MSBW would require much clevelopxentand testing also.Top shielding plugs that are removable for maintenance access wereused successfully in the KRE-2, and MSRE, and a sealing membrane waswed in the same general way at the ERE-2 and %RE. Plow of air into thecell through openings was also used to prevent the dispersal of radioastivityduring maintenance. Ear er cell spenin s will be required ow anmBR, the amounts sf activity in the system will be greater, and the restrictionson discharge of activity will probably be tighter. Methodsand equipment will have to be developed for sealing openings ar~~atd workshields and tools to provide better shielding against radiation and toIp@StriCt the required air flow %Q Pates that W i l l not require eXceSSiVe%yParge absorber and filter banks.W,&....I

.:E.:.>Evaluation.....,z

.... YII

....,.:.:.:;.;.;s....:.A,Normal operation of an BR includes all phases of startup fromcold hot standby conditions, production of electric power at demandedloads between 20 and 108% of design capability, and seheduled shutdown.The control systems must reeognize the different requirements forthe various operating modes and establish and maintain safe and appropriateoperating conditions. The systems must coordinate the operationof the reactor, the primary- and secondary-salt l~ops, the stemtogs, and system auxiliaries. In general, the load demand is the primarysignal to which the control subsystems must respond. Mowever, while~l%tchLng the power generation with the load, the control system mustmaintain system temperatures and their rates of change within acceptablelimits. Specific areas of concern are the temperature sf the steam atthe turbtne throttle, the rate sf change of temperature in the salt loops,and the salt temperatures, which must be maintained well above the freezingpoint throughout the circulating system (with the possible exceptionof surne areas in the steam generator).ne pPe§ent concept is to control the nUCleELr power genePatiol3 bygraphite rods, which are used in an automatic control loop to maintainreactor temperature at a set point programed according to the needs ofthe steam system. This arrangement is a variation sf a scheme successfullydemonstrated on the IGRE9 where the temperature set point wascontrolled by the operator. Such a eontrok system makes the reactorpower slave to the load, with a temperature base li~e independentlydetermined to provide steam at the desired temperature.Maneuverhg from one power level to another requires C Q I I ~ ~ of O ~steam temperature during the transient. The current concept involvesautomatic control of sec~ndaq salt flow rate through the steam generatorto take advantage of the thermal capacity of the salt while thereactor power level is being readjusted to the new requirements.In the multilbsop plant, such as the reference mBR, the controlsystem must adequately respond to loop interactions. The most satisfactoryapproach appears to be one in which each IQS~ is controlled asa unit to produce a specified amount of stem under well-defined comditions,wFth the bala~ci9b-g of the loops under the control of a masterpr~gra~r~~~er - perhaps a digital computer - that is responsive to theneeds of the power grid. This programer would adjust set points onappropriate closed-loop ~ o R ~ ~ Q B I associated ~ I x with the coolant ISO~S.The control system requirements of an MSBR are basically the sameas those sf sther power reactors. However, they do differ in SQIWdetail. These differences will be discussed later.315

316IIn addition to providin womal control functions the instrumentationand control system must provide protection against a variety of mormalousor accident conditions Although the entire control system should contributeto safe and orderly ~perations, there is always a system dedicatedto protection of personnel and to the p ~~ention of major equipment damage.This system, the plant protection system (PPS) includes monitoringinstKUllleIltatiOn t0 detest Off-9aO~~l COnditbns logic sUbSystelE3 &Smake decisions and initiate corrective actions and actuators to effectprocess control actions 0The plant prstection system must be capable of shutting down theplant when necessary an out other protective functions 9 suchas insu~hg that system are in order for containing the radioactivityin the event ~f a major accident and the removal of afte~heat folllwingan emergency shutdown where normal cooling is impailred.Whereas the control rods will be graphite partially inserted intothe core so that positive and negative reactivity changes can be made,the safety rods will be neutron-absorbing poison KO~S of considerablymore reactivity ~o~th. Because of their effect on neutron economy andbreeding, the safety rods probably will be normally withdrawn out of theactive core region. However, in some nol-~~il circumtances, having thesafety rods partially inserted for a short time my be desirable. Therefore,continusus adjustment of theirp position must be possible.It is not dear at this the that a fast "scram" capability will berequired. Tke p%aIIlpt negative fuel temperatu~e CQeffiCiat plus theconsiderable therm1 capacity of the salt and graphite EIK~ factors whichmake the plant less sensitive to reactivity excursions (see Chapter 14).However, only detailed analyses of a particular plant design will establishthe precise requirements of the PPS.'. ..... .-.ex:. ........,vi. -,.a .IThe ~perationd cont~ol system and the plant protection system willrequire extensive instrumentation to provide input into the automaticdecision-=kin process of control Measurements of neut~on flux levelsas well as of the nonnuclear variables such as flow rates, p~ess~res,telBpe?ZatUre§, etc. W i l l be Vital ti3 effective control of the pla'~It.Some instrument B ~ ~ S O Kand S signal transmission lines, and possibly someContainTlEnt penetrations W i l l b@ Keq red to operate reliably in hostileenvironments of high temperature, hi radiation levels, or both. Thehigh residual radioactivity in the reactor cell will make direct maintenanceimpossible in many locations, so accessibility of instrument cowpoenentsfor ntem~te disconnect and replacement will be necessary.~w general, conventional electronic or pneumatic signal conditioningequipment can meet the needs of the NSBR. However, the size and complexityof the plant will make it highly desirable to use digital computer rechniquesfor multiplexing, data storage and retrieval, calculation, andother functions 0 Optimization of the plant output will require developmentof sophisticated ccp~~t~ol schemes. A high degree of automation ande.. .I

,.... .....

318lXSSlaK@ loIlg-tel2Il ShUtdOWn wargin lander UFlUsUal CirCUElst EZlCeS 80 fastactingdrain valves are not required.The reference design NSBR has the entire primary salt system in anoven where the temperature is maintained at about 1000'F. %is arrangementmakes it unnecessary to F~sulate 'the pipes and vessels and removesfrom the reactor cell a mu%titucle 0% electrical cables thermocouplesdisconnects and containment penetrations that would $e required forheaters di~e~tly on the pipes and vessels. Some c~nnect~rs and signaltransmission lines must either be designed to sperate properly in suchan environment or must be located in penet~ati~ns that are cooled rebtab ly.Experience with the MRE and Other FacilitiesA large number of out-of-pile a d in-pPle Poops and sther facilitieshave been operated with molten salt at ORNL over the past 28 years, andall Sf these have had instrU€R@ntatiQn andl cofltrol System Of Varyingdegrees sf eomplbexity. In addition, we and others have operated highturesystem containing molten metals or gases$ and a number ofs of these types have been aperated in the U.S. and abroad.e provided some experience in instrumentation that isirect and useful infomation has come fromISuccessful operatfen of hundreds of thermocouples attached to thesalt system walls in the MSRE gives confidence that reliable temperaturemeasurements can be made a$ the elevated temperatures of molten-saltreactor systems e Although there was considerable scatter in the readingsof the csupkes lander different heaters when the salt was actually isothermal,techniques of biasing the out uts Were used to provide thorouacceptable measurements [3, ppe 22-24] The impurtance of careful seltim and calibration, details of fabrication and installation, and stricty control was evident, as only 62 of the 330 thermocouples failefive years of operation.ressure and differential pressure measurements in the coolant saltsystems were made using NaK-fiIled transmitters. No direct measurementof salt pressure was made in the primary system, where gas pressure measurementswere used to infer salt-sys tern pressure 0 Direct meaaure=ntsare desirable ire an MSBR, md additional development may be requiresuck applications mThe measurement sf salt flow rate in the secondary system was madeby meam of a vef~turi and a: NaK-filled differential pressure transmitter.No direct II1eL3SU%k3I[leWC Was HLade Of flow ill the PriIELKy systefs. bEtsFJtnChas the reactor W ~ B operated with constant primary-salt flow rate, no particularproblem existed in normal operation. ks part 0% the plant proteetionsystem instrumentation, pump motor current was measured. Thoughnot a precise indication of flow, this measurement plus pump speed gaveadequate ass~rance of flaw. That is, the pump mtsr current would beless than normal even thou the speed was normal if for some reason the

319..... k.iii3I;.;$...:.,......x,*......:.:.w,:.:.w........,.,.. ......A2....., ,.:.:

Control Analysespreliminary studies StartbBp, standby and shutaown p ~ ~ ~ e d ~ e shave been carried out on the reference design MSBR, dthoafgk only tothe point of deter~nain feasibility. Ita naakimg these analyses seve~albasic restraints on ape ation of the plant ~ e reco ~ e ized. The freezingtemperatures of the primary and secondary salts are ch that the saltsystems must be filled and circulating issther ally at l808"P beforepower withdrawal can be initiated by desreasin zhe coolant-salt temperature.To avoid freezing of the saltent excessive temperaturegradi@tat%, the EknifnarTB feedwater or steam teTRperatUre to the steam generareferencedesign system must vary between 1000°F at zeroin the 8 ts 168% power range. In addition, the afterheatx.2 . ..the feedwater and heat rejection system remain in operation followinshutdam of the main steam system*MSS~ of the special syste equipment needed to handle the startupand shutdown conditions in W station are the~efok-e associatedwith the scea-power system. me qeairementa impose some departure fromtern used in conventional fossil-fired supercriticaltsand will require further study. Different steamfor example reentrant or concentric-tubeot as 768"F, and wold1. An auxiliary boilsary far startup from the cold sonditisn.me pr-oposed general arrmgement of the steam system of the KSBRreference desi is described in the desi PepoKt 61, Sect. 51 and in&ZLpteK 8 Sf this %epOrt.-........

3211computer Models.....;e.*,,*.

322these limitations imposed, we found that the secondary-salt temperatureat the steam generator outlet could not be held above the salt freezingpoint at part loads beLow about 50%. A need is indicated for effectiV&deCQtaplil2g Of load effects On Sdt teIXperatUKes. Some possibilitiesare: (1) allowing the steam te~~~perature to increase above its 1000°Fdesign pint as the Isad decreases, with subsequent attemperation sf thesteam with injected feedwater; (2) inereasin the fee dwa t er temp er at ur eabOVe its 780°F design point as the load deCKeaSe%; (3) reducing thenumber ~f steam generators in use as Isad decreases; and (4) using a saltth~ottling valve to bypass some of the secondary-salt flow around theprimary heat exchanger to reduce the temperature of the salt enteringthe steam generator. Steam attemperation and secondary salt bypass werethe subject Of additional steaay-state analysis of the piant concept eEither schene or a combination of the ~ W Q appears to permit the establishmentof acceptable part-load operating conditions.=

322.:.>aThere are certainly other possible control schemes %or achievingsatisfactory plant perfor~~n~e. Analog simulations have shown that anintegrated control scheme is essential to good performnce. I€ excessivethermal stresses are produced by the transients that accompany loadchanges, they can be reduced by varying the primary salt flow rate as afunction of load.ake small isotherml temperature coefficient 0% reactivity (seeChapter 4) implies that only modest amounts of control reactivity areneeded to accompEish plant load mneuvering. FOP the reference design,a typical maneuver from 58% to 100% power at a rate of 5%/min required0.05% &k/k and a Kate Sf Q.OOOB%/seC 6k/k. m@ maXhUTU SYStCZEE t@mperatramrate of change for this transient was about O.3"F/sec at the reactoroutlet .Accident Analyses.... i,;; ..d.:.:,x,....*.:*'The most likely abnorml power excursions WQUE~ result from suddenchanges in load demand OK rapid changes in salt flow rate in either theprimary or secondary system, as a result of pump or power failure. A fewlimited cases of thPs type have been examined. on the hybrid simulation.Some less-likely reactivity anomalies were also briefly examined on thehybrid model. As discussed in Chapter 4, conceivable reactivity changesmay result from primary flow variation, fuel addition accidents, coregeometry changes, or failure of one or more control rods.A reactor shutdown QP protection system must be coordinated with thesalt ci~~ulation loops and the steam plant. If the load is suddenly lost,the reactor power generation must be reduced to avoid ov@rh@athg.Similarily, if the reactor is shut down the steam load must be quicklyreduced to avoid subcooling or freezing of the salt. Similar situationsarise due to salt circulation pump failure. Fur example, if primary flowis lost, followed by an appropriate reduction in load demand and reactorpower, the primary salt could still freeze in the heat exchangers becauseof the ilICrei3sed dwell time Unless SecQndaafy %POW iS also lfeduced.The plant emergency procedures clearly will be somewhat complex becauseof the high salt freezing temperatures, and a careful analysis willbe needed to derive satisfactory solutions eReactivitv Control......:%cLong-term reactivity adjustments are expected to be accomplished byvarying the fuel concentration. Normal regulating and shimming functionsfor load following and shutdown are within the calculated capabilitiesof a few graphite rods, as proposed for the reference design e%%, p. 641.Some events may be anticipated that would require additional negativereactivity or reactivity rates beyond the capability of the graphite rodsalone. Poison rods could be used to provide this additional reactivity~ontrol and to provide substantial shutdown margin, Having poison rodsin the core during normal ~pg~fation is undesirable because of theiradverse effect on breeding. The probable role of the poison rods, then,is to be held outside of the core, poised for rapid shutd~wn if needed.

324"hey may also be used for additional skimming during core loading or forother operattons or abnormalities but the exposure at power will be shortand che effect on breeding and r d life w ill be smll. We do not presentlyanticipate that extremely fast insertion, or "scram,t' of the rods will benecessary, although reliable insertion must be assured e....YIns trumentation,........- _IIn the reference MSBR, the entire reactor cell will be at about 1000°F.Thus any nuclear detectors which are located in this space must be capableat Such high temp@ratUI'e QT else the)' mUSt be Cooled. Noe presently available for operation above 6O0-9QOoF, and evenfor these terngePatUres Cmlgr SpeCiaP deV€2%8plllental tnsdeh eXiS t. Sidlarproblem exist in the liquid meta% breeder reactor prog~am, and somedevelopment work is being done, but no significant progress has been announced[%8]. A coordinated detector aeveiopment program is needed forthe %%oPten-SaPt Reactor Program.It may be possible 5s locate ionization chambers in specially cookedwells OK thimbles located outside the reactor vessel. For use in thereactor protection system, such chamber welabs WQUld have to be designedwith performance reliability as required of the protection system, sincefailure sf the cooling systems would bring about failure a$ the detectors.Neutron fluctuati~n analysis proved to be a valuable tool for monitO%ingaIl0Ed.QU.S bC?haVior in the MSm [g, 18]. UnfoPtUlately, Qne Ofthe requirements for obtainin good results with this techniq~e is a highdet@ctiQn efficiency f8P CQPg coupiec~ ~leutrons. mile no detailed ealcbnlationshave been made of i-~e~tr~n fluxes outside the reflector and vessel,we estinate that in the NSBR these fluxes will be too low to provide %heal-to-noise ratio for some types of fluctuati~n analysisthe proposed nuclear detector locations are within the hightemperaturepri ry containment oven and have in ~o~frm~n the problem ofcontainment penetration seals. Generally, the specifications for penetrationsfor nuclear detectors are demanding than those fer processsensors, because ~f the typically very smll signal currents delivered.The location md detailed requirements of these penetrations have not yetbeen determined, but the need f5r some development work in this area isanticipated. The signal transmitting lines will require special designand SQHle deVelopwent because Of the high-t@mperature enViPonXllent.Process Iststrbaraentrat$~~~

325;.=,>.....applicable to the MSBK. Similarly, experience being gained by the utilitiesindustry with instrumentation of supercritical pressure steam systemswill be applicable to the PTSBR.MSBR process instrumentation located outside the biologicallyshielded areas and not an integral part of the containment system canbe conventional equipment. Some standard components, however, may requireupgrading, and a strict quality control program will be requiredto ensure a level of reliability and performance commensurate with MSBRrequirements.All process instrumentation components located within the containmentcells or as an integral part sf the containment system must probablybe considered developmental. These components are predominantly primarysensing elements for measurement of flow rates, pressuress levels,weights, and temperatures in the salt-containing pipes and vessels, inthe ass~~iated purge and off-gas systems, and in the salt chemical processingfacilities * Other such components are final control elements(such as off-gas C Q R ~ ~ valves), O ~ Peaa-tJire and piping connections tothe sensing and final control elements, remotely operated disconnects,and containment penetration seals.The electrical conductivity of the MSBR salts will be a factor inselecting the type S% primary sensing elements that can be used. Theconductivities of MSBK salts are estimated to be about 1 mho/cm - aboutthe same as MSRE salts. This means, for example, that magnetic flowmetersprobably cannot be used, and most of the devices will be similarto those used o~ the MSE. Some new techniques are being investigatedwhich show some promise for use with Pow-conductivity salts.SOM~ development will be required to adapt MSRE control c~mpone~ltsto the higher pressures and temperatures that will exist in portions ofthe MSBK. Development sf other equipment and techniques, such as electricalpenetrations into salt-containing pipes and vessels, would undoubtedlylead to improved instrumentation.Rod Drives....,...-1-The detailed rod requirements will be a function of the particularreactor design. "he rod worths will vary with type, size, and location,SO that drive speeds cannot be specified until the physical and nucleardesigns are we18 advanced. The mechanical desPgn of the drive mechanismis discussed in Chapter 8........&*..........,A ,........i........JSalt Throttling ValvesOne of the more promising plant control schemes depends upon thrst-%ling in a secondary salt bypass around the primary heat exchanger.(Tight shut-off is not required af the valves in this bypass.) Themaximum and minimum flow rates and the rate of change of flow rate necessaryto achieve satisfactory plant control will weed to be factoredinto the development of a valve far this service. Valve design is discussedin mapter 8.

Digital Computer Application for Control and Data HandlingChemical Plant Instrumentation and ControlThe instrumentation requirements of a full-scale chemical processingplant have, so far, received only niraimal attention from instrument desig~ters.%e p~ocesses involved have been instrumented a d controlled onlaboratory or pilot plant scales, but not with the volumes and radiationlevels expected in the full-scale plant. Early evaluation of potentialproblems is needed to enable initiation of required development on atimely basis-Uneert ain t ies and Alt ernat ivesThere are several uncertainties regarding the overall control of: anMSBW that will require additional analyses and csnsideration of alternatives.En particular, the high freezing temperatures of the salts createthe need for special considerattons in load and flow control. Urnusualconditions or failures, such as a failure of a pump in a salt loop, willrequire specific responses to avoid salt freezing. The behavior of multipleloops of a lar e plant will need to be carefully prog~amed toavoid freezing prolaleas.As mentioned earlier in this chapter, maintaining desirable or acceptablesteam a d csoiant-sait temperatures in the system under papt-ioadnditians is a fundamental problem. Seve~al approaches have been investitedfar possible resolution of this difficulty. One is the use of flowproportioningvalves to bypass coolant around the primary heat exchanger $a d this is a very attractive scheme from strictly the esntrol vfeqoint.Relatively complete separation of load effects from the salt operatingcsnditisns is possible using one 0% two valves, and salt temperatures andsteam conditions can be optimized more or less independently. proportioningvalves fop molten salt will have to be developed if this scheme is tobe used.The second agsp~oach, which has been briefly examined, is to allowthe temperature sf the steam leavin the generator to rise at reducedhade and to attemperate it with feedwater addition to regain proper eonditionsfor the turbine. Attemperation is used in modern supercrit%ca%fossil-fired steam plants but is usually foll~wgd by a reheatel- ~r super-$&..,. . . -..... -

.....327.....I.:.:*

32 8EvaluationIassessme t of the cont Dllabilfty Of aR HSBa COmp Ked With Otherpower plant co-ncepts reveals a number of favorable features and a fewfeatures which add difficulty. bong the favorhble -features are then~clear charasteristics ID which be described as aocik. Comfortablylong promp%-neutrOn %ifetheS a d a large pKODIgSt n@gatiVe teI!lperEktU%ecoefficient of react ivity yield very desirable control characteris tics[3, p. 621. Thus the capability for fast transients as a result ofconceivable reactivity anomalies is minimal [l, p. 1173. A positivemoderator coefficient contributes to make the isotherm1 temperaturecoefficient of reactivity fairly small. This results in ve~y modestcontrol reactivity requirements for mneuvering and perFaits very smallloaded excess reacti~ity. The large heat capacity of the molten saltserves as a buffer to absorb the effects of reactivity transients andLessen their inf%wnee on the plant, a d melting of fuel is mot alimitation since Ft is already molten. Gaseous fission products arecontinuously strippea from the salt, greatly reducing their usualreactivity effects, ZEKd the entire %eaetOr fuel system opeKat€S froma low base pressure.The high freezing temperatures of the salts are negative featuresthat complicate both control a d protection to some extent in that specialprovisions must be made to avoid f~eezing of the salt during power OKe Analysis indicates that special control measures are neeintainproper salt temperatures under part-load conditions.1 situations, such as loss of salt circulati~n, requirehe load md reaucin flow in adjacent loops as well as re-e sirculatin fuel system alters the effective delayed neutronfraction since some of the delayed neut~on~ are edttea outside the reactorcore area. Variation of the fuel salt flow rate, therefo~e, willaffect: reactivitgr SQIWZ extell%, but this FS not expected to present iaSignificant CCRltrpOl prob%@m. gfth On-line f U d prOeeSsing, EkCCUKateFIIV~D~O~~. ~ontrsl a d l~ng-&er~~ rea~ti~it~ aceomting will be difficultthan in past experience, since a large number of variable factorsaffect the reactivity balance to some degree. The net effect of dl thesefactors, however, is not e pected to present any short-te~m control or-term bookkeeping uncertainties.BVe%l CQnCep% fob^ eating sf the salt system will make maintenanceoperations easier than if the piping and vessels were individuallyheated and insulated, but more of the instrument systems will be exposedto the high-temperature environment. Because of this and some otherrequirements adequate instruments for all applicati~ns are not availableoff the shelf. Development has been S~QW, however, because the requirementsare such a strong function of the partic~la~ application and~ ~ ~ i2 i ~ ~ r n ~ ~ ~In sumary, same of she control and instrumentation problem willbe challenging, but B Q W ~ is expected to be beyond existing technokscapabilities. Instrumentation and csnt~ol. development problems shonot have a detrimental effect on the overall development schedule furOs...... ~.~..... \x.xk..

329References for Chapter 181. Conceptual Besip Study of a Single-Fluid Molten-Salt B~eede~Reactor, ORNL-4541 (1971) 0$.Fa-2. a. L. Moore, Fwther Discussion of Inst-ntgtisn md ContmlsBeuehpment &xded for the Molten Salt Bmedero Reactor, ORNE-TN-330% (Aug. 5, 1971).5. MSR Pmgmm Semimn. PPO~P. Wept. Aug. 34, 1968, 88EaE-4344.7. 6. D. Martin, Jr., hstmmenta%ion and Controls &k. Ann. Pmg.Report, Sept. 1, 1970, ORNE-4620.8. R. C. Steffy, Jr., P~equency Eesponse Testing of the IdoZ$en-Sa%tReactor Experimerzt, OIRE3%-T?+-2823 (March 1970).10. J. e. Robinson, B. N. Fry, Det-e~nrincrfiion of the Void P~action inthe MSRE Using Smll Induced Presswe Pertwhation, OFW%-a%l%-23b$(Peb. 6 , 1969) 012. W. PI. Sides, Jr., ControZ Studies of a 2080-Mu(e} MSBR, OmL-W-2927 (May 18, 1970).13. w. H. Sides, Jr., MSBR Control Studies: Analog Simulation Progrc&m,OR.NL-TPI-31Q2 (May 1971)14. 0. Id. Burke, Hybrid Conputer SimZation of the IGBR, 0RML-'6p1-3767(May 5, 1992).15. C. K. Samathan and A. A. Sandberg, University of Illinois, Chicago,IPPinois, and P. H. Clark, 0. W. Burke, and 8. S. Stone, ORNL,"Transient Analysis and Design Evaluation of a Once-Tfnrough SteamGenerator with the Aid of a Hybrid Computer," submitted for publicationin NucZear EiagineeAng md Besip.

330

11. FUEL PROCESSINGL. E. McNeeseOperation of a molten-salt reactor as a high-pe~%~~-~~~nce breeder ismade possible by the continuous processing of the fuel salt in a facilitythat is located at the reactor site. The most i~~portant operations consistin HC2IilOVing fission product9 (pPhcip8PBy the Hare earths) and %SOlating233Pa from the region of high neutron flux during its decay to23% in order to hold neutron absorption in these mterials to an acceptablylow level. It is also necessary that excess uranium produced in thesystem be removed %OK sale, that the fuel salt be maintained at the properredox potential, and that the oxide and corrosion product concentrationsin the salt be maintained at tolerable levels.me rates at which the fuel salt must be processed for 233~a removaland raPe-eartk removal are mutually dependent. It will be convenient todefine the tern o'processing cycle as the time required for processinga volume of fuel salt equal to that contained in the re~ictor system.me 'erewovaP timest for a given material is then an effective cyePe timethat is equal to the p~ocessing cycle time divided by the fraction of %hematerial that is removed in a pass through the processing system. Asshown in Fig. 11.1, for a particular single-fluid MSBR having a breedingratio of l,O7, the required rare-earth removal time can range from 58days for a protactinium removal time of 2 days to about 18 days for aprotactinium removal ti= of 20 days. The optimum choice of protactiniumand rare-earth removal times is largely dependent on the characteris ticsof the processes employed, For example, the present ~a~e-ea~th removdprocess requires that protactinium be removed from the salt prior to theremoval of earths. Hence, with this process, the rare-earth removaltime will always be as long as or longer than the protactinium removaltime. As will be discussed later, a protactinium removal time of 10 daysand 8 ra~e-eartla removal time of about 27 days are used with the referemeeprocessing sys tern.....&Processes involving the selective chemical reduction of materialsfrom the fuel salt into liquid bismuth appear to be the most promisingprocessing methods currently available, and the development of theseprocesses has been the subject of most 0% the recent WOK^ om fuel processing,We have noted previously [I, p. 1701 that the iscslatisn of protactiniumis straightforward since its extraction behavior is significant~ydifferent from that of uranium, thorium, and lithium. Hmever, untilrecently, the removal of rare earths was difficult since the rare earthsand thorium extract in almost the same manner from molten fluoride mix-tures. In 1969, Smith and Ferris [2, p. 2851 noted thatdistribute seleet%ve.ly into molten lithium chloride fromcontaining thorium; this observation allowed McNeese [3,the rare earthsbismuth s slut ionspp. 2-15] to331

33%ORff b DWG 72 - 78990 0Fig. 11.1. Rare earth and protactinium removal time combinationsthat resinkt in a breading ratis sf 1.07.

,:.=...........:.=*=.= ....33%devise a significantly improved rare-earth removal sys tem known as themetal transfer process aDistribution of Metals Between Noaten Salts and BismuthBismuth is a low-melting (271°C) metal that is essentfaPPy* idsciblewith molten halide mixtures consisting of fluorides, chlorides, and bremides.Tke vapor pressure of bismuth in the temperature range of interest(500 to 700'6) is negligible, and the solubilities of Iftkium, thorium,uranium, protactinium, and most sf the fission products are adequate forprocessing applfcatisns.Under the cOElditiOnS Of iflterest, PedUctiV@ extraction PeaCtionS betweenmaterials in salt a d metal phases can be represented by the followingreaction:in which the metal halide mn in the salt reacts with lithiurn from thebismuth phase to produce M in the bismuth phase and the respective lithiumhalide in the salt phase. The valence of PI in the salt is +fi9 and x representsfluorine, chlorine, and bromine. It has been found [4] that ata constant temperature the distribution coefficient D for metal M dependson the lithium concentration in the metal phase (mole fraction), XLi9 asfollows:log B = n log x 4- log ,y,* 0The quantity K * is dependent only on temperature, and the distributioncoefficient ismdefined by the relation:LfD-mole fraction of K in metal phasemole fraction of MX in salt phase enI:.....: -.- .,.....,The ease with which one component can be separated from another is indicatedby the ~ a t of i ~ the respective distribution coefficients, that is,the separation fastor, As the separation factor approaches unity, separationof the components becomes increasingly difficult. On the otherhand, the greater the deviation from unity, the easier the separation.Distrib~tion data obtained [4] for a number of mate~ials betweenfuel salt (72-16-12 mole 2 LiF-BeP2-ThF4) and blsmuth at 640°C are summarizedin Fig. 11.2. The lines for the various elements have slopes thatcorrespond to the indicated oxidation states. Under the expected processconditions, the Pa-?& separation factor is about 1200, which indicatesthat protactfnia~lrn as well as UP~~~UEII and zirconium can be easily extractedfrom a salt stream containing ThF4. However, the rare-earththorium separationfactors are cldae to unity (1.2 to 3.5), indicating that removalof the rare earths from a salt containing thorium fluoride will be difficult.The p~evious ~a~e-earth removal system, which was based on these.....*.&. i .,....

3 34MOLE FRACflOM Li IN BISMUTHFig. 11 2 eDistribution data between fuel salt and bismuth e.. ~.~ ....

335.&&.Em separation factors, required a large number of stages, a high mtalto-saltflow ratio, and a large electrolytic cell for providing thoriumand rare earth reflux at the ends of the extraction cascade [I, pp. 190-75, pp. 52-77].F%e have found, however, that with LiC1 or LiBr, much more favorablethoriUU€-Tar@-@aKth SeparatioIl factors are obtained [2, p. 2851. DiStributiondata for L i ~ [6, l p. 3.71; 71 at 640°C are shmn in Fig. 11.3. Thedata fall roughly into three groups. The divalent rare-earth and alkalineearthelements distrfbute most readily to the LPC1, with thorium-rareearthseparation factors uf about POs. The trivalent rare earths formthe seco~d group and the thori~~t-~are-earth separation factors are aboutlo4 e Tetravalent materials , such 8s thorium and protactinium, distributeonly slightly to the LiC1. Studies on the temperature dependence of thediStribUtiOn data ShCW @SSentiallgP no effect fOk the divalent e ~ ~ ~ e ~ t § ,a WhOK effect for the trivalent elements, and a SOllEWhat greater effectfor the tetravalent elements. The distribution coefficient for thoriumis decreased sharply by- the addition of fluoride to the LiC1, althoughthe distribution coefficients for the rare earths are affected by only aminor amount. nus, contamination of the LiCl with several mole percentfluoride will not affect the removal of the rare earths but will cause asharp increase in the thorim discard rate. Data with LiBr [7] are similarto those with LiC31, and the distribution behavior with LiCl-LiBrmixtures would likely not differ appreciably from the data with the purematerials.The potential held by LiCl for selective extraction of the rareearths from MSBR fuel salt is best illustrated by considering the equilibriumconcentrations of rare earths, thsrium, and lithium in fuel salt,bismuth containing ~ed~ctant, and LiCl as shown in Table 11.1. The eoncentrationsof the rare earths and alkaline earths in the fluoride saltcorrespond to a 25-day removal time for these materials in the referenceMSBBa me thSkiUTR COnCf3RtPEitiOn in the bislmuth 1s 90% Of the ~fl%lSlpi~sslubility at 64O'C. As can be seen, the ~a~e-earth and afialine-earthelements are present in the LiCl at Pow concentrations and al-e associatedwith a negligible amount of thorium.A rare-earth removal system based on this effect will be practicalonly if a suitable means is available for remving the l-al-e-earth andalkaline-earth ele~~nts from the Lie1 e The distribution coefficients forthese elements are strongly affected by the concentration of lithium inthe bismuth phase, and the best method for removing these materials fromthe EiCl appears to be extraction into bismuth containing lithium at aconcentration of 8.85 to 0.50 mole fraction. Sufficient data have beenobtained with 1ithium concentrations in the bismuth as high as 8.38 molefraction to show that no deviation oecurs from the relations establishedinitially with much lower lithium concentrations.Pl-ota~tinim Removal System

336QWIUL-DWG-7Q-12482I I I I

,.a.....a.....w.....eC w iTable 11.1, Equilibrium concentrations in fuel carrier salt, bismuth, andlithium chloride at 640°C...ElementMote fractionIn fuel carrier salta In bismuth En lithium chloride

338....

339UP4 and approximately 0.0035 mole z BaF4 is withdrawn from the reactor.About 99% of the urawiUR is rem0Ved from the salt by fluorination in Orderto avoid the use sf large quantities sf reductant in the subsequent protactiniumremoval seep. The salt stfeam is fed countercurrent to a bismuthstream containing Pithim and thorium, whe~e the remining uranium andthe protactinium transfer to the metal strem. These materials are transferredfr'om the bismuth to a captive secondary salt by hydrofluorinatingthe bismuth stream EeavPng the extraction colurntaa in the presence of thesecondary salt. The secondasry salt which flows through the hydroflusrinatoralso circulates through a fluorinator, where about 3OX of theuranium is removed, and through a tank that contains most of the prstactiniurn.Lithim is added to the bismuth leaving the hydrofluorinator,and the resulting stream is returned to the top of the extraction colbum,The salt leaving the extraction column is essentially free of uranium andprotactinium but csntafns the rare earths at essentially the reactor concentration.This stream is fed to the rare-earth removal system.Rare-Earth Removal ProcessA simplified flowsheet for the ~are-earth emo oval system [9 pp.1-15] is shown in Pig. 11.5. Fuel salt, which is f~ee of uranium andprotactinium but contains the rare earths, is countercurrently csn%actedwith bismuth containing reductant in order to extract a significant fractionof the rare earths into the bismuth. The bismuth stream, whichcontains the rare earths and thorium, is then countercu~~e~tly contactedwith lithium chloride e Because of highly favorable distribution c~effi-CientS, Significant fractionS Of the earths transfer to the LicPalong with a negligible amount of thorim. The final steps of the processconsist fn extracting the rare earths from the %i@k by contact with bismuthhaving lithium concentrations of 5 and 50 at. 2.This process has a number of very desirable characteristics. Ofprimary i~~portance is the fact that there is net consumption of reductantin the two upper csntactors. The process is not sensitive to minorvariations in operating conditions. Essentially no materials other thanthe rare-earth and alkaline-earth elements are rem~ved from or added tothe fuel salt; the major change consists in repdaeing the extracted rareearths with an equivalent amount of lithim as LiF. The amourat of L%Fadded to the fuel salt in this manner during 30 years of operation wsuldbe less than 10% Sf the EiF iEWentOr)P in the reactor.ConceDtual Processinn FlowsheetThe reference processing flowsheet [a, pp. 3-21] is shmn in Pig.11.6. Fuel salt is WithdrENll frofa the reactor (an a 10-day Cyck; for al000-m(e) reactor, this represents a flow rate of 0.88 gpm. The fluorinator,where 99% of the uranium is removed, has an active diameter of 8in. and a height sf 15 ft. The protactinium extraction co%um is 3 in.in diameter and is packed with 3/8-in, Ras&ig rings, The column isequivalent t~ five equilibrium stages and has a height of 15 fta The

340---IIFUEL SALT(No U or Pd--- Bi - Li(0.5 MOLE FWAC. Li 1OR.....LL..!bi-Li- - m + DIWALEN?I--Bi -Li-g + TRIVALENTRARE EARTHS..... . Fig. 11.5. Metal transfer process for remvd of rate earths fromsingle-fluid SBR fuel salt..... as..

I341....c.:..... ..:?m,

34%

343contains the equivalent of 5 mole % Lip. It appears that the fluorideconcentration in the LiCl can economically be as high as 2 mole %$ whichcorresponds to a thorium discard rate of 7,7 mles/day. Discard of thoriumat this rate would add only about 0.0013 miHl/kWhr t~ the power cost.he effect of fluoride in the L~CI on the removal of rare earths is negligibleIn fact I the rare-earth removal efficiency increases slightlyas the fluoride concentration in the LiCl increases. In addition, contactOf EiCl COntZi.tPigng fluoride with has been fOUnd to rC2Sd-t in foETlationof volatile BPq 610, p. 1061, and thus fluoride can be removed fromLiCl easily by this means.The reliable removal of decay heat from the processing plant fs animportant consideration because of the relatively short decay time beforethe salt enters the processing plant. A total of about 6 ba9 of heat wouldbe produced in the processing plant for a 1000--b/ICa(e) MSBR. Since moltenbismuth, fuel salt, and LiC1 are not subject to radiolytic degradation,there is not the usual concern encountered with processing of shortdecayedfuel.Waste Stream Produced by Processing Plant,=.......... ,:.M....:*..... aAll high-level waste streams produced by fh protactinium and rareearthremoval systems can be codined [$, pp. 3-21] for uranium ~ecovelpyprfor to disposalb, as shown in Fig. 11.7. In this operation, waste saltfrom the protactinium decay tank would be combined with the discard streamof fuel carrier salt e The lithium-bismuth stream from the trivalent-rareearthStripper would be hydrofluorfnated i93 the presen@e sf the resultil'lgsalt, and the combined stream would be held for protactinium decay. Theprotactinium concentration in the combined stream would be only 500 ppminitially, and the specific heat generation rate would be acceptably low.The salt in the waste holdup tank would be fluorinated before discard torecover uranium in order that the loss of fissile material can be madeacceptably lowe The composition of the di-carded salt would be 74.7-13.5-9.5-0.8 mole % EiP-ThP4-BeP2-ZrP4, 1.2 mole % trivalent-rar@-ear$h flusrides,and 0.3 mole X divalent-rare-earth fluorides. The salt tempe~aturewould have t~ be maintained at about 600°C so that the trfvalent-rareearthfluorides would not precipitate, This processing scheme would requirethat salt be discarded at the rate of 60 ft3 every 220 days. Weanticipate that the waste wfll remain in this form unless the requirementsof the federal waste repository make further processing necessary.Thorim is discarded from the system at the rate of about 50 moles/day. Alth~~gh the cost of replacing this thorim is low (S.0084 mill/kWhr), the resulting thorium utilization is only about 20%. Flowsheetmodifications have been developed, h~ever, that will not require discardof thorium and which will result in almost complete utilization of thoriumif desired.An additional high-level solid waste stream, which contains most ofthe iodine and bromine removed from the reactor, is produced by the H2-HPpurification and recycle system (shown in Fig. 11.81. The B2-HF streamsleaving the fuel reconstitution step, the hydrogen-reduction C0lUrnSBpurge coPums and hydrofluorinators are combined, compressed to about

344

3456sORNL DWO 72-1803..... .=....&$HF-ti2CONTAININGHI,HBr2,2.........,aFig. 11.8. Hydrogen - HF purification and recycle system......m........,a,.... &.&....ss3

2 atm pressure9 and shilled to -40°C in order to condense HP from thestream fsr proLarge fractionen and fluorine for r pcle by electrolysisected to be dissolvedin the condensate. These c ounds are morevslatile thanand can be separated by low-temperaturedistillationtop of the distillation coPum, whichs codiwed with the gas stream leaving chen a small quantity of HP, and the resultingas stream is ied in regenerative silica gel sorbers andAbout 5% of e hydrogen is fed through beds of activatedalumina and charcoal for removal of SeFg, TeFg, and noble gases, whichare not removed by the KOE.The halides are accumulated in the KOH scrubber soEution for a periodof 34 days, after the solution is heid for a 45-day decay period.The solution is then evaporated in 24-ine-dian, IO-ft-long waste cowtainerse Two waste cogntaine~s are filled annualLy *Alternate Processing Methodswe presently know sf no other rare-earth removal method as attractiveas the metal transfer process; hmaver, protactiniu~~ and uranium appearts be removable from fuel salt by alternate methods based on the selectiveprecipitation of the oxides of these materials. Protactinium ispresent in the NSBW as PaF4. Baes, B r, and Boss [IObQ, p. $21 haveshown that Pa4+ dissolved in a molten Fz-ThPb, solution can be oxitothe 5' state by hydrofluorination according to the reaction:

Y-349&.&.....&...,.:.:.y,....5::s.,_. .......Me have aade prel5mina~y evaluations of several conceptual fl~wsheetsbased on oxide precipitation [14, pp. 237-403; the most prodsing of theseis shown in Fig. 11.9. Fuel salt would be withdrawn from the reactora 3-day cycle, and about 60% of the protactinium wou%d be removed as Pa285in ordemr to obtain a protactinium removal time of 5 days. The Pa205 precipitateWould be hydrQflU0Kinated in the presenCe Of a captive fluoridesalt phase which would be circulated through the protactiniu~~~ decay tankand through a flu or in at^^ in order to maintain an acceptably low uraniuminventory in the decay tank. Part of the salt in %he decay tank wouldbe returned to %he react~r p~~f~di~ally to conapensate for salt that istrEla9Sferred tQ the hydafOflUQrinatQ1- With the Pa205. Ten percent Of thesalt leaving the Pa205 precipitator would be pro~es~ed for rare-earthremoval by the metal transfer pr~cess;; this would result in a 38-day processingcycle and a rare-earth removal time of about 50 days. Most of theuranium must be removed from the salt prior to the removal of the rareearths. This codd be accomplished either by fl~~rination or by oxideprecipitation The separated uranium would then be recombined with theprocessed salt leaving the metal trans%er system and wsuld be returnedto the reactor.evel lop went work on a rider of aspects of the reference ana alternateprocessing plant fl~wsheets either has been completed recently oris in progress.Metal Trilwsfe~ Process Development.:.:.: -h engineering experiment completed recently demonstrated all aspectsof the metal transfer process El51 for the removal sf rare earths. Theequipment consisted of a &in.-diartr compartmented vessel in which werep~ese~t about 1 Biter each of MSBR fuel carrier salt, bismuth saturatedwith th~rium, and EiCl. The fluoride salt initially contained 147NdF3at the tracer level and LaF3 at a concentration sf 0.04 mole fraction.During the experiment, the rare earths were selectively extracted intothe LiCl along with a negligible mount of th~ri~m. Provision was madefor circulating the LiCl through a chamber containing bismuth having alithium concentration of 38 at. %? where the rare earths and thorium wereremoved. The distribution ratios for the rare earths remained constantduring the empe~i~~~ent at about the expected values. About 58% of theneodymium and about 78% of the lanthanum were c~llectred in the Li-Bi SQlution.The final thorium concentration in the Li-Bi solution was below5 ppm, making the ratio sf rare earths to thori~a in the Li-BI greaterthan 105 times the initial concentration ratio in the fuel salt and thusdemonstrating the selective rem~val of rare earths from a fluoride saltcontaining thorium.A larger metal transfer experiment [k4, pp. 254-55; 12, pp. 289-12;131 has been put into operation that uses salt and bismuth flow rates that....,S&.... .

348

349are about 1% of the values requi~est for processing a 1860-MJ(e) %BR, andthe p~elhminary design has been carried out for an experiment that willuse a three-stage salt-metal eontactor and fbotJ rates that are 5 to 10%of those required for a 1008-&fW(e) MSBR [%PI,.... .X&...,3a...

350

351supply 1s musk simpler than that required for induction heating. A facilityis planned in which a continuous fluorinator can be operated thathas a molten zone diameter of 5 in. and a molten salt depth of 5 ft@Fuel ReeonstitutionStudies of the absorption of kTF6 by MSBR fuel carrier salt containingW4 are being carried out [L%]. Absorption of UFs in fuel carrfer saltcontaining UF4 has been sham to result in the formation of soluble nomvolatileUP5 according to the following reaction:.A....C.Xl....,.,.......,:.if4

352values &odd be obtained by fmprovin the contact of tihe gas with thesalt eom 50 to 90% of ~ 6 UJeaIliW 2 WBS pkecipitated EL$ oxide iHa most 0%erfments. Samples of the oxide contained about 98% U82 even thoulower uranium consentrations in the salt the solid in equi%ibrfumwith the salt would contain 58X 6702 or less. we believe that under nonequilibriumprecipitation conditions, such as would be present in a proc-@SShg phIlt:r uo2-Th02 Solid SolUtiCKlS fOmed askniCh are in eqUflibKiPamwith the salt at the moment of precipitation but which once fomed do%tot Papidly reeqUF%ibrate. "Ells, Solid SolUbiOHls that BPe folXEd eElPayin the precipitation process and that contain 90 to 95% U02 are stillpresent durin the final sta s Of precipitation when solid solutions arebeing formed at contain mu less u02. mi5 effect appears to allowpreeipitatisw Of 99% of the urani as a solid containisingle-stage bat& precfpitator. Pea contrast, earlier[lo, pp. 202-31 9 based On the assmptisn that the oxide and salt wouldrim throughout the precipitation processs had indicatede bat& countercurrent precipitation system would be1 of this fraction of the uranihom without theremoval of more tthe thorim in the salt as Th0-2.The oxide precipitate was served to settle rapidly, and more than~ Q X sf salt be sepa%a d from the oxide by simple deeantatisn.sts that the removal af uranium from KSBR fuel salt from whichctinim has been removed can be a ~ ~ o ~ easily p ~ ~ by s h ~ ~....&!%xon by contacting the salt with as eous B20-Ar KLxtures ERemoval of Bismuth from Fuel Saltplant, the fuel salt will be contacted with bismuthin 0lede.e: to P@EIOVe prOtactinilXil and the rare earths.It will be necessary that ent ained or dissalbved bismuth be removethe salt before ito the reactorp since nickel is -quiuble fw bismuth Q1 at the reactor operating temperature. Effortsto measure the s0Pubility 0% bismuth F salt have ithe so%ubility is 1me than about 1 ppm, i6fkbismuth fn the salt unsmuth can onlyant concentrations in the salt as entrained metallicCteHfZe biSt8tath CCJZlc@II$PatiOII lik@ly to beter it is contacted with bismuth, we have beeq eriments invo 1vinindicate that the bcentration in the salt in most e es from $0 to 100 ppm aftercsmtercurrent contact of the salt and bismuth in a packed-eo3bw contaetor;however, concentrations below 1 ppm are observed in salt lieavina st2rred-interface 8 It-metal contactor in which the salt andwst dispersede present aiffieuatfes is fiat ofC O ~ t a of ~ ~ ~ ~ les ~ with ~ o swllP ~ quantities of bismuth&emaafcsa_ maHysea e

35 3A subsequent phase OB: the experimental pro ram will consist in testingthe effece%Veness Sf VariQUs dE?VEceS and mate?XblS f8P 1PeRloViag entrainedbismuth from salt. It is expected that contact of the salt withnickel wool will be effective in removing entrained or dissolved bismuth,since a large nfekel surface area can be produced in this manner.A natural eirculatisn loop constructed ~f Hastellboy M and filledwith fuel salt has been operated by the Hetals and ceramics Division forabout two years; a molybdenum cup containing bismuth was placed near theb~tt~m of the loop, To date, the reported concentrattons of bismuth insalt from %he loop (

354me seleetisn of molybdeam as a prseessi plant material was basedon corrosion investi59; 16, pp. 189-951 andelsewhere [ 2% 23 whslution and chemicalattack in molten biswere esnducted in smallKIRZd eOa%VectiOn laops WhiGh provided 63 temperature radient of 100 to"c in the bisnutk circuit. Tests were conducted on IOW-C~X% BI mohyb -denurn and the %loy TZM in pure bismuth and bismuth containing up to 0.01ible in the temperaturehr. Tests carried SUt instatic bismuth also have spawn no effect of stress on the corrosivfty 0%mol yb denurn eb ~ o l ~ b has ~ eexeellent n ~ resistance to corrosion9 there areother dffficdties With fts use, MoIybden~m is a particuhfly Structuresensitivematerial; that is, its meektanieal properties are known to varywidely, depending upon how it has been metallurgically processed. meduct i %e-b rit t %e fCKan8 i tiOI2 temp eHat UrPe E moabybdenm varies from belowroom temperature ts 26U-300"6, dependin both upon strainm%CKoStrUc%Ure 0% the metal. MaXhUlD d ctility is provideworked, fine-tion. Recent advances fw vacuum-me%ting pracuctionof material with impraved and moreprope~ties The arc-melted Em-carbon, law-Qf I€iOlybdebnr%m, aVaik3bh CX3mer@ial1gP9 affords KehtabVPeIYof grain size and interstitial impurity leveP. Nevertheless,the use sf molybdenum as a structural material requires highly unproceduresand imposes strin ent Girdtations on systemstandpoint of geometry and rigces in the fabrication technology sf molybdenum havein building the mlybdenum system in which bismuth ande somtercurrently contacted in a I-in. -ID, 5-ft-highurn that has 3.5-ln.-ID upper and lmer dis aging sections eismuth will be circulated in the system by -lift pmps thatte the streams to 3.5-in.-I%d head pots for ling, gas separation,and flow ~~iea~ut-ement. A salt-bismuth interface eteetor af the typedescribed earlier will be provided in the lower dfsen gemenat section fordeteWwing the pressure drop thro, h the cslum and e holdup of bismuthin the collmm.T@chwiques have een developed %$, pp. 253-54; IO, pp. 184-85; 14,pp. 219-20; 12, pp. a. 7-69] for the prod~etion of cl~sed-e~~d ms%ybdenumvessels by back extrusiow, which involves %he flow af metal into a dieand the backward flow sf metal over an advancing plmger. 'This processhas the advantages that the dimet r sf the part produced is as large aser than that of the startin metal blank; the configuration of theed by relativelyes in the die and mandreland sufficient defo tion can be accomplished that a wrought or

355......:.y4,&$.........,:.:.:a.....sz

The results Of BUT work to date on molybdenum fabrication techniqueshave been quite encQIu ing, and we believe that the material caw be usedPants if proper attention is given to its fabri-CatPQll &a?XCtesgiS tics EGraphite, which has excellent compatibility with the fuel salt, alsoshows promise for the containment of bismuth. Relevant infomation onraphite is p~esented in Chap. 6, which reviews the development sf graphiteas a moderator for molten-salt reactors. Of course, in a chemicalp~ocessffag applfsation, the absence of a neutron flux allows greater flexibilityhl ~ E SeleCtion 2Of graphite rade and fabrication taistorq. thanfor a 3?&3ctOK COPe.

... s.9357......*.....;.:;:;:g....&: were conducted for 508 hr at '?6BBC using both high-puritybis~l~th and Bi-3 wt. % (48 at. X) Et. Although penetration by pure bismuthwas negligible, the addition of lithim to bismuth appeared to increasethe depth of permeation and, presumblgr, the wetting characteristics ofthe bismuth.There are several approaches that have potential for sealing a porousgraphite against penetration by the bismuth and bismuth-lithium alloysTwo well-es tab lished ones a ~e (1) multiple Liquid hydrocarbon iwpregnatfonsthat are carbonized andlor graphitfzed and (2) pyrocarbon coatings EAnother possible approach is the use of carbide-forsdng sealants. Eachof these sealing approilches is being evaluated in bismuth loop experiments.We are also studying the wetting characteristics of raphite as a functionof surface pretreatments such as dedustfng, alcohol wash and oven. dry,and vacuum degassing at 700 to 1000°C.Fabrication of a processing plant from graphite would necessitategraphLte-graphite and graphite-metal joints. We have conducted developmentstudies [25,26] on both types of joints wing high-teqerature brazesand also metals which bond by fsming ca~bidle~. Several of these experimentaljoints show promise for the chemical. processing application. Otherworkers [27,28] have pioneered mechanical joints which may be satisfactoryfOP the pHOpOSed applicatfsn......:;:.:.:.......... .,.....:.=.:....;.!i

458(20 different materials) were Ioeated in &e fluorinators. Several specimenshad lower rates of maximum corrosive attack than "L" nickel. Thespecimen showing the least attack, Hpu 80, had a maximum bulk loss rateof 11 mi%s/mowth based on total time in molten salt. Other corrosioncoupon tests at 600°C [3S] shwed that IMBR-1 is dso more resistant tocorrosion than spLsp nickelsThese ~peratisns have afforded useful guideligles and background infOl3Eatiofnfor the SelecfZ.iOfa of COIlStrUctiOn lWLt€%bkd.S for the pKCSposedprocess applfeations They shw the importance, however, of inerting themetal surfaces in a fl~~rinat~rwith a passive frozen-salt Payer.Effect of Uncertainties on ProcessingThe successful operation of a p~oce~sing plant based on the referenceflowsheet is contingent on the develspment pro ram meeting severalobjectives. These include the following:-1 developing continuous fluorinators having an acceptably low corrosionrate and an adequate uranium removal efficiency,%. identifying materials of construction that are compatible with moltensalts and bismuth containin4, developin on-line instrumentation necessary for plant operation..... a*;The eonsequenees of these objectives not being met are discussed in theremainder of this section.

.........ir. !,359..... .:.:

esntinususly of the concentration of uranium, protactinium, bismuth, andchromium as well as the redox potential of the salt leaving the processingplant. If on-line instrumentation could not be relied upon, it wouldprobably be necessary that processed salt be held long enough before beiwreturned to the reactor fop the desired analyses to be made. The fuelsalefnventory would increase about EO%, wfth an accompanyinof about 0.04 d%lb/kh%r in the fuel cycle cost, if a one-day holidup wererequired 0k...,

361Overall Evaluation of P K O C ~ SCapabilityS ~ ~ ~A

362ferences for Chapter 111. 6%. E, rnatley, L e E. KcNeese, w. L. carter, L e M. Ferris, andE. L, Nieho3isan, WueZ. Appl. Tech. - 8, (1970).I.. .Y7. Lo M. Ferris, F. J, Smith, J. c. Nailen, anJ. S’norg. MucZ. ChSrlI. - 34, 313-20 (1972).,.;.;.>:. ...--L.X......CY.

363.....

....i..;;...-

.... .;.y>.is.Concept.....:w......:.:.:.:. .._,The maintenance of reactors requires the performance of variousmechanical operations on equipment which, because of radioactive contaminationand activation, is not. directly accessible to maintenancepersonnel. Depending upon the level of activity, the size of equipment,and the design provisions for mafntenance, anything from simplelocal shielding ts fully remote manipulation may be required. Thetime required to do maintenance and the cost of the maintenance provisionsincrease with the de ree of remoteness required.The cir~ulating-f~el reactor has fission prod~~tsand intense~adiati~n tu contend with not only in the reactor vessel but also inall of the primary circuit through which the fuel salt circu1ates andin the off-gas system. If the fuel processing paant is integral oron-site as it will be for an MSBR, the maintenance of that plant isessentially part of reactor maintenance. Thus the circ~bating-fuelreactor requires radioactive maintenance sf a mater scope than doesa fixed-fuel reactor. On the other hand, the refueling operation issimpler, the radioactivity is retained on-site within one containment,and the necessity of a separate maintenance organization and equipmentfor a fuel reprocessing plant at another site is avoided.Although maintenance design efforts cannot affect the size andactivity level sf the components in a reactor, much can be done in thedesign stages of a plant to influence strongly the degree of accessfbilityand the complexity of the maintenance operation. The maintenanceconcept f ~ an r MSBR is characterized by the general principles:1. Each system is composed sf manageable units joined by suitabledisconnects.2. Each unit is aceessibEe and replaceable from directly above throughremovab He shielding e3. Failed units are removed amd replaced.The concept is described succinctly in the following quote [I] asapplied to the mm.isRedYced to fundamentals, the Msm is a collection ofcomponent parts which a ~ e capable sf being disconnected andreconnected remotely. Access to these units is providedthrough removable shielding sections that make up the roofssf the various cells. A portable maintenance shield is installedover the component, the roof section is removed, and

IIThe ability to completely disconnect a particularcomponent is basic to this system. The disconnects mustbe remotely operable by the Isng-kandled tools. They mustbe reliable both fop the service conditions and for thehigh radiation d in some cases must satisfy nuclearsafety considerat ions of containment leak tightness andleak detectability. A number sf different disconnectsare used at the 3fSW for the various applications. AP~nostall the piping in such auxiliary system as the offgas,lubricating oil, air, and couling water systems havestandard ring joint flanges, with minor TIlodificatiQns,special designs Were tlsea for leak detector tubing,

._ .....,.,367Technological Background..... ,:.:.>

368radiation. Lifting devices permitted the remotely controlled buildingcrane to remove shield blocks or major equipment items, Each removableitem had a bail at the center of gravity or other provisions to simplifylifting. Guides were provided where necessary to steer replacement partsinto place. To ensure proper fit, jigs were built for all the major replaceablecomponents. %he 5-inch salt lines were provided with flangesthat used frozen salt as a barrier to keep molten salt away from the ringgasket. Machines were developed (and fixtures installed in the cell forthem) that could cut the l-P/2-fnch lines to the drain tank, prepare theends, bring old and new pieces together, and join them by brazing. Theonly tools required for most operations were simple, long-handled hookswrenches md clamps characterized by their reliability rather than bytheir versatility -During the installation of the reactor equipment and the prepowertesting, many of the maintenance provisions were tested. All of the primaryloop was assembled on a large jig before going into the cell andoptical tooling was used to locate precisely reference points in the cell.Maintenance items that were tested included the crane, lifting and viewingdevices, and all kinds of disconnects, Freeze flanges were opened andc~osed; a cell space, cooler, the c~nt~o1 rods, KUCI drives and a coresample array were removed and replaced; %he primary heat exchanger andfuel pump bowl were installed using in part the remote maintenance provisions.&f*ter the nuclear startup expe~imgnts the fuel pump rotary elementwas removed, inspected and reinstalled. During this time personnel weretrained and procedures were perfected.

.....X.!/i36 9.....%. ...Table 12.1. Semi-remote work in MSWE reactor and drain cells after beginning of niaclmr operation",. ;.;.....?&.....:

the reactor: long-handled tools throu the maintenance shield. Some ofthe tools were quite different, however, to handle the special task.The control rods and drives were removed, the 10-inch core accessplug with the rod thimbles was taken out, and a section of rod thimblewas cut off with a grinding wheel. A grinding wheel, mounted on a specialtool, was also used to cut through the pump tank around the sampler cageso that it could be removed. An Il-inch section ~f the heat exchangershell was cut out with a plasma torch, then secti~ns of 6 tubes were cutwith an ab~asive cutoff tool and removed for inspection. Examinationshowed that, as suspected, a small amount of salt ($2 in.3) had leakednear a freeze valve in a drain line. me freeze valve ana adjacentpiping was cut out and renraved. The holes in the pump tank and the heatexchanger shell were patched, the latter by welding, ana plugs were installedin the severed ends of the drain lines.Although the major components of the fuel system, whose replacementwould have been more complicated than any job that was done, required nomaintenance, the MSRE experience was of sufficient extent to thoroughlyy and many of the specific design features.e was reliable: no job arose that could not bedone. h important factor in this was the flexibility of the maintenancesystem which allowed obstacles t~ be circumvented and mfo~e~een jobs(such as those in the off-gas system) to be accomplished.The experience with the MSRE emphasized that the payoff for preparati~nis tremendous in the case sf radioactive maintenance. Jobs such asreplacing the core spe~i~ns weant quickly compared to others thatwere basically less difficult, but for which no special provisions hadbeen made.Especially valuable information on fission-product csntamipationcame from the MS experience e Noble gases behaved predictably and couldbe purged before ystem were opened. Salt-seeking fission products wereno problem - the salt drained cleanly and any that was trapped froze andretained the fission products. There was no c~rr~sion film QK scale toflake off and form dust. At leasc part uf the noble metals that depositedon surfaces in the off-gas system were fairly easily transferrable butparticulate eontaminat ion was generally c~nfi~~ed to the tools which wereswabbed agld ba ged as they were purled from the shield. Iodine that wasproduced by the decay of tellurium on surfaces soon appeared in the gas.Ventilation air frokg the reactor building and ~~QEII the c~ntainment cellswas passed through particulate filters and up a stack. The greatestamo~nt of activity disel-aar ed in any week was less than 0 -2 CP (mostlyiodine) and occ~rred durin work an the off-gas system.Although radiation levels in the reactor cell were typically on theorder of several thousand Whr while maintenance work was going on, the.general background to t&iich workers were exposed while mantpulating toolswas only about 0.01 a/hr. ~rocedures were planned to minimize personnelexposure in locally higher radiation fields and the work was never seriouslyinconvenienced by havin to rotate WOKkerS. NQ IlEiintenanCe workerever received more than the ~QKEBB limit sf 3 rem in any quarter.I.... -

....i*,;+,371......:.:.y,Reference Design MSBR('),. I.:, 9In designing and p anning for the maintenance of t..e MSBW we haveevaluated a d adapted the @X.l>@KienC@ with the Msm Etrad other radioactivesystems, The maintenance system, as applied to the reference MSBR, isdiscussed in some detail in ORXL-4541 661.NSBR maintenance requi~ements fit into the following four generalclasses eClass I - Permanent Equipment. - This category contains all thoseitems which can reasonably be expected to require no maintenance duringthe design lifetime of the plant. Examples are the reactor vessel, thepump vessels primary heat exchanger shells , the fuel-salt drain tank,thermal shielding, thermal insulation, the connecting proce~s piping,etc. No special provisions are included for maintenance af these items.Emergency maintenance to some extent is possible, however, because ofthe access that is provided primarily for in-service inspection.Class PI -Equipment Allowing Direst Maintenance. - This group includesthe item which can normd$y be approached for direct maintenancewithin a reasonable period of time (typically after the secondary salthas been drained and flushed and the remaining activities allowed todecay for about 10 days) The steam generators, reheaters, csolant-saltpumps, and the equipment in the heat rejection cell fall into this class.In the unlikely event that one of these components did become highly contaminatedwith fission products, its removal would be treated as a ClassPIE or IV item, discussed below, Qnce the sources of activity were removedfrom the cell, cleanup and component replacement could proceed inthe norm1 fashion using direct maintenance.Class 111 -Equipment Requiring Semidirect Maintenance. - Muck ofthe equipment in the sffgas and chemical processing cells, such as pumps,blowers, valves processing vessels, filters ets., will become radioactive.In general, the sizes s% these items are comparable to the MSNequipment. WBR radiation levels may be a factor of l6 higher than inthe MSE:, however. The maintenance tools for this class of equipmentcould be similar to those for the MSRE, but the shielding and containmentprovisions would have to be more effective because of the more intenseSQUPC~S of radiation. - ThisClass IQ -group includes items which are clearly beyond present experience becausesf a combination of size, radiation level, afterheat removal, and disposalconsiderations. Ex~~~ples are the pump rotary element, the primaryheat exchanger tube bundle, and the core graphite.The reactor primary system, because of the large size of the highlycontaminated equipment , presents the greatest problems in containing theradioactivity and dealing with afterheat and is therefore used as thebasis for the discussion which follows.

co nt a irlmerttfithougk the fUe% Salt and highly Hadioactive gases Will be reEiQVedfrom arny system before it is opened for maintenance, the reactor primarysystem will still contain Barge amounts of radioactivity, some of whichwill be transferable. The MSBR building and cells and the maintenanceequipment and procedures must, therefore, be deai ed so as to limit thespread QE radioactive material within the reactor building and to preventmore than trivial. amounts from being released ~~tside the building duringmaintenance eme heating and the amounts of short-lived activity that must bedealt with decrease rapidly during the first few days after cessation ofpower operation. For this reason it is unlikely that the primary systembe opened sooner than ten days after flk~i-poWer operation is stoppea.Tea days after shutd~m from long aperation of 2256 W(t), the noblemetabfissi~nproducts ~n surfaces may total about 2.2 x 108 curies, orroughly 3 x IQs Ci per $t2 of metal surface that had bee^ exposed to thefuel salt. (This figure is based on the assumption that 75% of the noblemetals deposit on metal surfaces in the Hoop.) The deposited telluriumwill be generating I~dli~e, some ~f which will go into the gas or air contastingthe surface. at see (11.6 clays) the calculated total rate atwhich 2.341 1321 is generated from 78-11 ~e on surfaces is 2 x 106 Cijhror ab~ut 30 ~i./hr per ft’ sf surface in the fuel circulating system.Eight-day l3II will be generated on surfaces at a tQtd rate 0% abaut5 Ci/hr or 7 x 10-5 Gi/hr per ft’.*%he noble gases in the graphite after PO days CQU%~ amourtt to asmuch as 1.2 x IO6 Ci (almost all 13%e) I assuming none diffused out duringthe cooling period. %&e radioactive daughters of noble gases Ln thegraphite would likely range up to I x $07 6%.The in~entorie~ of the fissisn products that dominate at 10 daysafter shutdown may be as much as 170 times those in the MSBE. (Thisfactor may be considerably Iawer if, as expected, the MSBR gas-strippingsystem removes much of the noble metals.) The amounts per unit areashould be less than 5 times as great as in the MSM, however.* Thus theobserved behavior of the dep~sited fission products in the MSRE shouldbe rather similar to that to be expected in the MSBR.Based on the MSRE experience, we expect that the noble metals onsurfaces will be more or less adherent, depending 5n whether they are onsurfaces in the salt loop Q‘P in the off-gas system, but that care mustbe used to avoid knocking 011 scraping them off. Although there was uncertaintyin the fraction of the iodine generated on surfaces that cameoff into the gas in the MSRE, it is clear that in the MSBR measures mustbe provided that are adequate to deal with all Q€ the iodine so generated.,... ....ii..a,: ....* For the shortlived noble metals that are dominant a few days aftershutdown, the intensity on fuel loop surfaces is nearly pr~po~tfonal tothe ratio of power to loop surface area. F Q the ~ reference MSBR this is2.25 x 106//%.2 x 104 = 43 k~/ft’; for the MSRE the ratio was 7.4 x 1~3/8.5x 102 = 9 kW/& 0

37%..........-,a....$a......x:.:., .=2.....,:.:.:*....*&..... it.:',The noble-gas daughters in the graphite will not be readily transferable,and should cause no sontarnination problem.The MSBR building and equipment layout described in Chapters 3 and 9are intended to permit safe containment of the radioactivity during maintenance.Before maintenance is started, the fuel salt will be securedin the drain tank. (Normally it will be circulated for several days tohelp remove afterheat before it is drained.) The system requiring maintenancewill then be purged sf radioactive gases and, if necessary, cooleddown. The containment cell will then be unsealed and a maintenance shieldset up. The cell will be maintained at a slight negative pressure by anair exhaust system. Tools and other items penetrating the maintenanceshield will be sealed eo the shield so as to minimize air leakage intothe cell. (Boots or gas-buffered seals might be used.) Primary systemwill not be left open to the cell longer than necessary; if an equipmentitem cannot be replaced inmediately a temporary closure will be applied.This will be required to minimize both the radioactive contamination ofthe cell and the ingress of oxygen a d moisture into the systema. Whencontaminated items are removed through the shield, they will be withdrawninto casks through openings equipped with valves or flanges that willclose the reactor cell and the cask except during the transfer. Generallythe gas streams passing through the cell will be filtered, passedthrough absorbers and recirculated. Any excess gas will be stripped ofradioactivity before it is discharged to the atm~~phere. Item removedfrom the primary system will be repaired or prepared for disposal in ahot cell where similar precautions must be observed. All operations willbe carried out inside the reactor building to assure complete containment.The assurance of public protection during maintenance is comparableto that during operation, although the lines of defense are different.During operation, the fuel salt and highly radioactive gases are circu-Hating within the vessels and piping, which are in turn doubly contained,being inside sealed cells inside the reactor building. During maintenance,the fuel salt with most of the fission products will be sealed inthe drain tank, and the precautions that ensure that it will not bebrought out during the maintenance constitute primary containment for it.Any system to be opened will be drained and purged so that when it isfinally opened the amount of radisacttvity that could conceivably escapeinto the cell will be far less than that circulating (and liable to beingspilled through a pipe rupture in the design-basis accident) during operation.The first line of defense ("primary containment") for this limitedamount of radioactivity is the ventilation sys tern that maintains the cellsat a negative pressure and K~II~CJVES any radioactive contaminants from theexhaust stream before it is discharged into the reactor building. Secondarycontainment during maintenance is the same as during operation -the sealed reactor building.Af terheatThe fission products that will remain in the graphite and on surfacesin the fuel system will produce significant heating in the largeequipment items while the maintenance operations are going on. 'This must

374be taken into account in the provisions for the m a j ~ r operations: replacementof the core graphite, the tube bundle of a primary heat ex-Changer, and perhaps the rotary element of a priIEiry pUHnp.we estimate that 106 seconds (11.4 days) after shutdown from 2250MTJ(t) the total heat generation rate in the primary system would be 790 kW,consisting of 210 ISW in the graphite of the core, 125 kW in each S€ thefour primary heat exchangers, and 66 kW distributed over the other surfaces.What the temperature in the pri~ib-y system nust be before it isopened has not been decided. With the cell cooling system that we haveenvisioned, the fuel system temperature 10 days after a shutdown wouldlikely be in the range of 588 to 1QOB"F. If further study of the maintenanceprocedures indicates that this would be unacceptable additionalheat removal must be provided or additional time allowed to reach a satisfactorytemperature. Some c~~ling must be continued while the item isbeing removed, but the ra%e of temperature rise in the absence of coolingwould be ISW - 3B"F/hr for a heat exchanger bundle and 2.4"FIhr for thegraphite core. The temperatures of the piping and rotary element of thepump would rise even more slowly.Graphite ReplacementbS f?Xpbained elsf3Wfaere iR this HepQrt, beCaU%e Of IleUtron irrad2.3-cion damage, it will be necessary to replace the core graphite severaltimes during the life of the MSBW plant-* Considerati~n of the effectson breeding and possibly on power diseribtuion Leads to removal of thegraphite while it is still structurally sound. Thus, although the removalprocedure must be capable of dealing with broken graphite elements, thestrength of the graphite should R Q hamper ~ its handling.In the OWL reference MSBR 963, the upper head of the reactor vesseland the entire core (176 tons of raphite and 97 tons of metal) are replacedas a unit. This constitutes by far the largest maintenance taskin this conceptual design, and explains some important features of theb~ikding layout and equipment described in Chapters 3 and 9. The majoritem of special maintenance equipment required for the core replacementis a 28-ft-cliam x 40-ft-high shielded transport sask. The carbon steelwalls of the cask are about 2 in. thick, whish is sufficient to reducethe radiation level on contact with the outside of the cask to about 1080R/hr and at the outside wall of the reactor containment vessel to lessthan 6.1 R/hr after a PO-day decay period for the core. Conservative estimatesindicate that the 210 kW of heat being generated in the core canbe safely dissipated through the cask wall so that no cooling system forthe cask will be required. The reactor core assembly is prepared forremoval in a semidirect fashion through a work shield. The lifting sfthe core assembly into tne transport sask and transport to the spent core&.,*~n exception would be the aow-power-a@nsitp MSBR discussed in Chapters2 and 3, in which the cure is made barge enough and the power densitylow enough that the graphite will last for the life of the plant.

storage cell, the installation of the new core assembly, a d the replacementof the shielding are accomplished from the remote maintenance controlroom in the same fashion as the removal and replacement of large item ofMSW equipment.Ebasco Services and their associates in the industrial design studyques cioned the desirability and practicality of replacing the entire soreat once. They elected instead to make the core of individually replaceableelements, and decided that 15-in. hexagonal elements were the optimum[SI. They would replace some of the graphite elements at 4-year intervalss during waj or turbine-generator overhauls. Adoption of this schemewould obviously greatly alter the requirements for handling tools andcasks for the exposed graphite.Status:.pa'....As stated earlier, the MSBR maintenance concept depends upon accessfrom above, a system S% replaceable units, approprtate disconnects andtools to operate them. Thus it is clearly essential that maintenancedesign be concurrent with plant design. This has been the case in theconceptual studies to date. The maintenance techniques for fluid-fuelreactors have evolved as the size, complexity, and radiation levels ofthe reactors have increased. Design studies have not indicated any insurmountableproblem in maintaining a 1000-FgW(e) MSBW, and no seriousconflicts have arisen in imposing the maintenance requirements on thereactor sys tern.Most of the techniques and many of the tools have been developed.Several. flexible maintenance shields have been built and used. Opticalviewing equipment - window inserts, periscopes, adequate lighting - allare available. The use of a shielded maintenance control room with windows,remotely-operable TV, and remotely-controlled crane and tooling hasbeen successfully demonstrated. Remotely-operable disconnects for electricalpower, instrumentation, and service piping are at a satisfactorystage of development. The remote fabrication of brazed joints in smallsystem piping has been demonstrated in connection with the MSRE [2].Two important techniques that are requisites for large power reactorsare not: available, however. They are remote welding md in-service inspectionand repair.It is highly desirable from the standpoint of reliability that theMSBR circulating fuel system be of all-welded construction. Thus remotecutting and rewelding of the system piping will be required in the replacementof major components. The status of remote welding as of 1969and the required development program were presented in reference 8. Aportion of that program has been accomplished and the present generationof automatic welding machines are reliable and capable of making highqualitywelds. These machines are not now capable of fully remote welding,but appear to be adaptable to this purpose.The provisions in the MSBR for access to equipment for maintenanceoperation are equally applicable to in-service inspection. The state-ofthe-artof remote inspection of welded joints is reviewed in reference 9.

Dependable application e% common methods for nondestructive inspectionof welds is difficult or impossible in high-te erature, high-radiationfields, Some methods promise to be suc~essf~l, however, and current AECand industrial p r ~ g r a are ~ ~ developing equipment, nanipulators, and interpretivemethods for acoustic efnissi~n and ultrasonic holography m~nitoringand inspeetion. These programs are expected to culminate in remote in-SpeCtiohn UEthQds for reactor welds which should be adapt&k? to l?fSBR Conditionsand needs. The equipment and techniques for remote repair do notexist; however, much of that development as well as the inspection devel-~pment is interchangeable with remote welding developmene.Further WorkThe program for maintenance development for the MSBR [l,6] consistsOf three major elements:1. The develapwent of remote cutting and welding capability for replacementof major items of equipment in the circulating fuel system.3. ‘Ehe concurrent design of a maintenance system that is integral withthe design of the reactor system and the development and demonstrationef tools and techniques required for that specific reactor.%e development of remote cutting and welding requires the adaptationor devel~pmgnt of manipulators for remote placement, pipe alignment ~and control of the automatic welders and she plasma torches that are alreadyavailable. Such a program could be started at. any time.The in-service inspection and repair operations will use essentiallythe same tools and techniques that are required for the cutting and welding,Techniques for nondestructive examination should be pursued and themost promising [9] should be adapted for the particular allays and configurationsexpected in the HSBR.The development of equipment for maintaining a particular MSBR cannotprecede the design of the plane and the development sf plant components,but it should keep pace with it. The general scheme of sewiremote~laintenan~e that was proved in the MSRE can be used for some parts ef theand off-gas systems, with only miner changes in shielding tocompensate for the higher intensity of radiation. For physically largesystems, QII the other hand, it wodd not be practical simply to scale upthe tools used at the MSRE because they would tend to become too unwieldyfor use by hand. nus some aeveiQpments in handling equipment and changesin the tool and component design will be required. The basic philosophyof designing components to be maintained with simple, reliable tools manipulatedintelligently and flexibly should be retained $ however #......... ~.

....377The conceptual design of system for limiting the spread of radisactivecontamination during maintenance can proceed and should be carriedwithout delay far enough to define requirements on filters, seals, etc........ 1....,.&Eva 1 u a t ion..... . .......w,......Y.. >> ,:.:.=. ...........L.*.....ps:*.....*pa....l,.:.:.i' .AAn abssbute necessity for practical maintenance of an MSBB is theearly recognition of the eventual maintenance requirements. Maintainabilityconsiderations must imbue the entire design effort, and waintenamepreparations must be tho~o~ghly developed and tested in advance.We are aware of these needs and they are reflected in our conceptual designs*The size of the equipment and the intensity of the radioactivity inan MSBR are greater than we have dealt with before. Nevertheless, thegeneral phihsophgr that we have developed, many sf our techniques, andsome of our t~ols are either directly applicable or readily adaptable.Other tools to handle large, heavy equipment must be developed as theneeds are defined. In addition, the techniques and equipment %or weldingand inspection that are now being developed in connection with otherkinds of reactors will have to be adapted and developed further for useon an MSBR. It does not appear, however, that the maintenance of an MSBRW i l l impose UIIP@asQnable Kequirem€!ntS for invention Or developftnellt.In summary, consideration of the state of the art and the foreseeabledevelopment of the technology in relation to the needs of molten-saltreactors leads us to conclude that by adequate planning and preparation,the maintenance of an MSBR can be wade reliable and safe. 'Ehe costs ofthe special provisions that must be made for maintenance of an MSBR havenot been estimated in detail, but they appear mlikePy to be a decisiveeconomic factor.....$92....

378References for Chapter 126. CmceptwzZ Desip Study of Q S~ng'ke-PZwid MoZten-Salt BP@@&P Reactor 9OWL-4541 (1971).

13. DESIGN STUDIES AND CAPITAL CQST ESTIMATESK. I. Lundin e. w. colainsIntroductionThe Molten-Salt Reactor Program at the Oak Ridge National Laboratoryhas, since 1968, been focussed on a 1088-W(e), one-fluid MSBR. A studythat was completed by ORNL in 1970 pr~stuced the conceptual design that isdescribed briefly in Chapter 3 and in detail. in reference 1. The Nolten-Salt Group, as part of their privately funded assessment of WBR technology,reviewed this ORNL design and issued a critique of it in 11971 [2]. Meanwhile,OWL issued a request for proposals for an independent, AEC-funded,indust~ial design study of MSBR's. A proposal from Ebasco Services wasaccepted and in 1971 Ebaseo and its industrial. associates began work ~ndera subcontract with OWL [3]. Task I sf the study included devePapirag anMSBR plant concept from specified criteria. This task was completed anda final 4iego~t issued in February 1972 [4].This chapter deals with design work the OWL reference concept andthe alternative approaches proposed by the Ebasco group. It identifiesthe information on materials and the developments in high-temperaturedesign methods that will be needed. Ffnally it discusses the capitalcost estimates that have been made and their sensitivity to uncertainties.Primary Systems Layout and Structural DesignBackground and StatusThe OWL reference concept el] was based on a "top" s~pp~rted primarysystem utilizhg structu~al members in the cell roof for supporting thesuspended major equipment. The reactor vessel was anchored while theheat exchanger and pumps were free to move except for the restraints imposedby the hangers. The piping flexibility analysis showed all stressesto be reasonable. However, only cursory evaluation was made sf the effectsof seismic forces, inertial stresses, or stresses due t~ thermalshock. The basic assumptions wade were that dashpots or other seismicrestraints could be added and that the effects of extreme transientscould be prevented or mitigated by proper system c~nt~ols.Ebasco Services, im their Task I conceptual design [4], came to theconclusion that when the cell roof penetrations for maintenance accesswere considered, there would be insufficient space in the structure forequipment supports. In addition, when consideration was given to responsefrequencies and amplification factor due to the 50-ft height sf the equipmentlayout, the requirements for seismic restraints made it desirable toinvestigate alternate support systems. The concept that Ebasco chose toinvestigate used "bottomt* reactor vessel and heat exchanger supports

380with a three-level system of horizontal trusses (lateral supports) forcontrol of earthquake motions mey successfully ran flexibility andstress analyses, and alth~ugh an arrangement with acceptable stresseswas found, it turned out to be very sensitive to relatively minor changesin layout and temperature assmptiuns. The design of equipment supportsand restraints will, therefore, require further detailed analyses.Ebasco also took an alternative approach to ORNL on the question ofttaemal transienlcs They investigated the requirements for directlyacsomodating a shock without usin special contrul features or devices eFor example, a system scram on loss of secondary coolant pumps or on asteam line break could result in the primary system outlet line changingfrom 1300 to lQ50BF in 14 sec, or the inlet line changing from 1850 to1308'~ in 30 sec. such transients would result in quite wacceptableess levels. As a remedy for this, Ebaseo selected and analyzed a deusinga thermal sleeve whisk isolated the flowing fluid from thepiping by an almost "stagnant" fluid layer between the piping and sleeve.BY taking advantage 06 the effect of low Reynolds on heat transfercoefficient in the annulus, they were able to reduce the~~~il-~h~~k*cran-,..,.,...sient stresses to negligible levels eSensitivity tu Uncertaintiesme differences in approach between Ebasco and OWL-4541 concerningreactor system layout and structural support center mainly on methods ofsupporting primary systems components and controkling temperature gradientsin components due ts transients.Neither the primary system supporting arrangement shown in the 6mLreference design nor the one proposed by Ebasca has been completelystudied to the extent that all problems are horn and solutions are inhand. However, both organizations csnclude that it is feasible to designhigh temperature system having flexibility for expansion and yet capableof resisting seismic loads The OK" concept requires yet-to-be-developedhigh temperature snubbers for seismic restraint , while the Ebasc~ conceptmust S ~ O W the effectiveness sf a high temperature support system tu sustainedand shock loads. Final design of the primary system as well asits support system wik% require detailed analysis to achieve an optimumlayout Having the required flexibility and meeting the stress limitationsfor all operating conditions. The stresses due to expansion have beenshown tu be controllable to an acceptable level by adjusting the lengthsof piping while still maintaining a compact Payout. Lengthening the linesinvolves some increase in fuel salt inventory; however, most of the invemtoryis in the reactor vessel, pumps, and heat exchangers.The Ebasco analyses indicate that unless special measures are taken,transient thermal stresses at several paints in the primary system canbe excessive. Ebasco's s~lutio~~, the use of them1 sleeves or linersto eliminate direct contact between the c~~lant flow and the pressureboundary, appears to provide protection, and in surne cases may be theonly way stresses can be held down to acceptable levels. Thermal sleevesprobably involve additional cost, although in some places they may permitless material in component fabrication, thereby compensating for at leastpart of the cost of the liner.

381Future WorkFuture work ow methods for limiting thermal stresses is dependentupon selection QE a material for the salt systems components. Additionalwork can then follow to determine temperature profiles throughout thesystem for all operating conditions, to perform detailed analysis ofstresses at critical locations in components, and to evaluate temperaturecontrol systems, thermal liners, components supports, etc., that havepotential for enhancing feasibility, safety, and economics of the system.Other areas which would receive design attention include use ofexpansion joints at sell penetrations, methods of vessel fabrication(e.g., shop vs on-site) and provisions for in-service inspection of components-EvaluationAlthough much work remains to be done before attempting to selectopthiurn configurations, methods for controlling transient thermal st~esses,and support systems ~ O K the primary system, we believe the conceptualwork of both ORNL and Ebasco have defined the major problems and suggestedapproaches that will lead to satisfactory so4butions.Deslign Methods - Codes and Standardsi...... 2..........:>ys+ .The mechanical design of MSBR vessels and piping must deal with avariety of problems. Because of the high operating temperature, Pargethermal expansions must be accomodated; but considerations of salt inventoryand pressure drop dictate that the piping be no longer than the minimumnecessary to provide flexibility and acceptable stresses. The largetemperature differences around the system lead to potentially high stressesdue to temperature gradients, particularly near mozzles, tube sheets, orother structmra% discontinuities. Materials must operate in the creeprange, requiring design methods, appropriate for this situation, that areonly now being developed. The material contemplated for the salt systemsis modified Hastelloy N whose allowable design stresses, while expected tobe superior to those determined for standard Hastellsy M, have not beenestablished. Further, there is a possibility that the surface crackingproblem (Chapter 7) or some other consideration may require a differentmaterial such as Inconel QK stainless steel.Status;.=).....Due to the limited effort that has been devoted to conceptual design,only simplified elastic stress analyses were employed, primarily to sizecomponents and to evaluate design alternatives. Before actual csmponents.....,.,. ..A*. .,....

382could be built to satisfy the stringent requirements of the Nuclear PowerPlant Component Code, the K6etalEurgical development work must be completedto establish a thUleoUgh hQWled e of the properties of the chosen alloyand an exhaustive stress analysis must be made of all components over theentire range of design conditionsStandard Hastelley N is approved for use UndeP Sections 111 Ellad VIE%of the MME BoiIe~ and Pressure Vessel Code through code case approval.cas€? 1315-3 ZippPoves Use Of hSt€!llOy M for pKesSUre vessels Constructedin accordance with provisions of Section VEIP, Division 1. Ml~w~lblestresses are given for te~~~peratu~es to l300"F. Case 1345-1 approves useOf HastePlOy N for Class l73UcleaK VeSSt21S construC%ed in aCcC?rdanc€? Withprovisions of Section PEE of the Code. Design stress intensity valuesare provided only to 8OO"F, in cornon with other materials approved foruse under Section 111.Case 1331 provides rules fOP COnStrUctiC~n of Class 1 stUC%ear Vesselsthat are to operate at temperatures above those provided for in SectionTIT. H O W ~ V ~ the ~ , recent revisi~n, Case 1331-5, includes only 304 ana316 stainless steels and requires a thorough howledge of the materialproperties to establish the design stress intensities. Extensive stressanalyses are also required using techniques just now coming into use and,in fact, unproven for many materials, ineluding Hastellay N. Thus, beforeHastellay N can be used at elevated temperatures, a materials testingprogram adequate for eode approval must be co~pleted and Case 1331 mustbe revised to include the new material.Strength of standard Mastellay N above 956"P is limited by creepeffects, making it necessary to employ design rules, analysis methods,and stress limits which reflect the time-dependence of material propertiesand ~tr~otu~ai behavior. of failure costsidered by the designrules of code case ~331-5 include: ductile rupture from short-term loadings,creep rupture from I~ng-te~m loadings, creep-fatigue failure, andgross distortion due eo incremental collapse and ratchetting. Brief out-%ines of design ~ t h a& ~ also d ~ provided for the foihwing modes offailure: loss of function due to excessive distortion, buckling due toshort-term loadings, and creep buckling promoted by changes in geometrydue to creep deformation associated with long-%em loadings.Design methods to cover these requirements are currently being developedand applied to the LmBW program. They can be used to designBR components and assure confidence in their reliability and safety.Should current work result in the selection sf stainless steelinstead of Wastelloy N, much sf the needed materials information wouldbe available from the R program. There are indications that hnconelbm y be used in some L components in which ease materials informationfor it also would be forthcoming......sens itivity %a uncertaintiesSince no detailed analyses have been made of the thermal stressesdue to rapid temperature changes, stresses at nozzle to shell intersecthX3,Otl other SecOtndEtPy and peak StKeSSeS, We do not hc3W quantitativelywhat the values may be. The preliminary analyses do indicate that

383.....:

384If stainless steel is used for the ?EBR reactor material, inelasticmethods of analysis developed in the %k%FBR program will be directly applicable.If Inconel or Bastelloy N is used, most of the theoretical knowPedgefrom the LKFBR program will be applicable, but additional materialsinformation will be necessary to establish the hardening model for thedif f ePent s %KEIh cOmpOn@n&S aEvaluationExcellent progress has been made recently in developing codes andstandards through the ASME and the M3Css RDT Standards Program which willfacilitate design and construction of PEBR components The LEGBR High-Temperature Design Methods Program is expected to develop analyticalmethods for designing components with assurance of satisfactory perfomanceover their planned lifetimes. Even though a material other thanstainless steel will likely be used for NSBRs, the theoretical bases forthe design methods will have been established and only a reasonableamount of tesging of the reactor material will be necessary to developsufficient information to apply the design methods.Capital CostsBackground and StatusOmE estimated the capital costs for building the reference POOO-MJ(e)WBB after completion of appropriate research and development programs andfound them to be comparable to current costs for light-water reactors ofsimilar size [I]. The Nolten Salt Breeder Reactor Associates (MSBStA),headed by mack and veatch ana funded by midwestern utilities, arrived ata similar conclusion. En a study completed in 1970, they estimated thatthe capital cast of a 100O-MM(e) MSBR WoLLd be abaut BO percent less thanthe cost of a pressurized water reactor sf the same capacity [5,6]. TheReactor Assessment Panel of the Edison Electric Institute, in their 1970evaluation [TI, used capital costs for EBR’s equal those for light-waterreactors.Ebaseo has not yet made a detailed cost estimate, butwill do so as part of their current study.The accuracy of absolute cost estimates for the MSRR reactor plantequipment is limited by several factors, including the preliminary natureof the designs to date, .and especially by the uncertainties disc~ssedbelow. Direct comparisons with light-water-cooled reactors are hamperedto some extent by the major differences in design and materials for thetwo reactors. On the other hand, only about one-thi~d of the total costsf a nuclear power plant is for reactor equipment, the refor the heat-power system, general facilities, and indirect costs, whichare expenses that can be accul-atelgi compared. It is the comparativeapproach to cost estimation that has led OaMiE to conclude that the costsof a fully developed MSBR will be roughly the same as for a PWR./....Yu...,

385... .&$:.@.:&.:BWe estimate the cost of reactor plant equipment (excluding contingencies)to be about $3 mil1ion more for a %OOQ-MG$(e) MSBR than for aU306-MW(e) PWB. (The MSBR involves more expensive materials, specialprovisions for maintenance, and other unusual design features that areonly partially offset by savings due to its lower pressure and Powerthermal rating.) Our estimates allow $6 million more for contfngencieson reactor plant equipment in the MSBR than in the PhX. Special materialsadd $1 million more LQ the cost of the MSBR. On the other hand,the cost of the turbine plant equipment are $12 million less for the hightemperatureMSBR than for the PWR, according to our estimates. Afteradding in the indirect costs we arrive at a difference of $2 millionbetween the two reactors [I, p. 1511, which is insignificant comparedwith the uncertainties....Sensitivity to Uncertainties.....&....... GV.9...;......High-temperature stress considerations could conceivab ly requirelowering the reactor outlet temperature from 1300 to 1200'F. Loweringthe fuel temperature does not necessarily mean that the steam temperaturein the heat-power system must also be reduced. The stearn csubd be heldto the reference design conditions of 3500 psia 18O0"F/10QO"F, by increasingthe primary heat exchanger area by abuut 60% and the pumpingcapacity by the same amount to compensate for the reduced temperaturedifference. This would probably be done by adding cooling loopst resuktingin 40% additional cost for the heat exchangersp pumps, and piping,plus additional building cost to provide the needed Hayout area. Theoverall capital cost would increase about 6% in this case. hother approachwould be to lower the steam design condition to 3500 psia98BSF/900"F, thereby reducing the thermal efficiency of the cycle fromabout 44 to 42X. For a thermal efficiency of 42%, the thermal capacityof the reactor plant would have to be about 2400 MW(t) rather than the2258 m(t) used in the conceptual study. If it is assumed that capitalcost is directly proportional to the thermal capacity, the estimated capitalcost will increase about 6.6% due to the lowered efficiency. Thusit appears that either approach would result in about the same captialcost increase. (Fuel costs would be different.) If the same materialis used for the salt systemss lowering the temperature does allow higherdesign stress limits, whi.ch could permit thinner sections and less mate-rial in csmpsnent fabrication. Alternatively, less expensive materialmight be used. No credit has been cPaimed for these possibilities inthe foregoing estimates eIn our estimates for the reference MSBR, the cost of equipment madeof Hastelloy M amounted to 292 of the total plant capital cost [I]. Onlyabout one-third was the cost of Hastelloy N, the remainder of the cost ofthis equipment being mostly shop labor. Thus, although we recognize thatthere is considerable uncertainty in projecting the cost of Hastelloy 69to that which will prevail when FlSBR's are being built fn quantity, theeffects on the plant cost estimates are not ]likely to be dominant. If it:were possfble to substitute stainless steel or Inconel for Hastelloy N

386without any penalty from having to operate at lower temperature, significantsavings might be achieved in the 29% of the plant cost that isassociated with Bastelluy N equipment eThe cost of the HSBR graphite is particularly uncertain, partlybecause it is not yet clear whether sealing, which may account for halfof the cost, is economically desirable. (See Chapter 7.) The graphitecost is not a large fraction of the plant capital cost, however, amountingto 6% in the estimate for our reference aesis 111.Of equal importm~e with the capital cost is the plant availability.Here molten-salt breeder reactors tend to have an advantage since theydo not require periodic shutdown for refueling as do solid-fuel reactors.In the ORNL design, ~hartgi~g out the graphite core is scheduled everyfour years to coincide with the major turbine maintenance, so no additionaltime is lost for this. (E~SCCI's design involves replacing portions ofthe COR at similar intervals .> No credit €or the additional availabilityfor not having to refuel was taken in our cost estimates, however, inorder to prsvide an additional mar in in case the maintenance of radioactiveequipment requires mere time than is needed for maintenance ofsolia-fuei reactors.MSBW capital cost estimates involve considerable uncertainties dueto uncertainties in design. %tae cost of Hastelloy N and graphite affectsabout 35% of the capital investment. eo arisons indicate that the capitalcosts 0% MSBR's will be roughly the same as for light-water-cooledreactors aby-$ .eo Rc b us i ortsUncertainties still exist as to what IXS~~CX material will be Einallyselected for MSBW's ana in some of the methods for designing hightemperaturec~~ponents. Once the choice 0% materials is made, work underway for theplus a reasonable amount of material propertiestesting, should provide satisfactory high-te~~pe~ature design methods €orMSBR s .Several capital cost advantages af MSBR's over light-water reactorsappear eo offset additional costs associated with the dispersal of radioactivityiR %he liquid-fuel System. k?EPC@, Our estimate that the Capitalcosts of fully developed MSBR's will be about the same as those of presentLWR's seem plausibk, and an examination of the uncertainties in the costcomparison indicates that they do not represent a large percentage of thetotal cost......V.2&

(flis?387x4w.......xu,,.,......B . .:&1. Conceptual Design Study of a SingZe-Fluid kfoZten-Sa 2 t Breeder Beactop,OlUL-4541 (June 1971) *2 a Eva Zuation of a 2800-MVe #Q Z ten-Sal t Breeder Reactor, TechnicaLReport of the MoLten-Salt G~oup, Part IT3 Ebasco Services Inc.October 1973. -3. "Oak Ridge Picks Ebasco for Molten-Salt Breeder Design Study",Nuel. Tad.. 18(5) : 29 (May 697%) a4. ZQQO-Mi(e) E BR Conceptual Design Study, Final Repori: !Task I,Ebasco Services N.P. (February 1972).6. Project for Investigation of Molten Salt Breeder Reactor, FinalBeprt -%me 1 Study, Molten-Salt Breeder Reactor ~sssciates,Kansas City (Sept. 1970).7. Repmt of the EL5bb Reactor Assessment Panel, Edfson ElectricInstitute, Publication No. 70-36 (April 1970).

R. B. Briggs, 9. R. EngeP, P. N. MaubenreichGeneral Considerations....

390of radioactivity d~ not escape under the worst credible condition. Thetwo areas are, of course, interdependent, since the absolute c~iterionsf nuclear safety is that conditions which could ~vemhKh the radiologi-Cal safety pPQVhiOnS ElUSt B@\Te% Occur.In an MSBR, fission products are always distributed throughout thereactor and the p-troCeSSing SyStelras, in Con$raSt With CQnVentiOnal reZiC%O%Swhere the fission products are normally contained within fuel elementsin the core. Thus, for an MSBR to have equivalent overall containment,greater requirements must be placed on the c~ntainwent barriers fromthe fuel salt outward. On the other hand, the fuel-coolant barrier ina soiid-fuei rea~tor, interpssecl as it is between the heat source andthe cooling fluid, is the barrier most vulnerable to damage in a nuclearexcursion so that its pr~teetion and the consequences of its failure tendto impose mare restrictive nuclear safety requirements on a solid-fuelreactor. These are the obvious differences; others are brought sut inthe discussion of the reference-design MSBR which follows.This section deals primarily with the large KSBR s6atiun, includinga reactor and fuel. p.k-~~essing plant, that is described in Chapter 3.The purposep however, is to delineate the important factors that must. beconsidered fur any mobten-salt reactor. Envirsnmental effects of normaloperations a ~ e considered first, then the various topics that relate tothe ckreat sf a large accident.k...,Environmental Effects of Normal OperationWaste heat must be dealt with as in any thermal power plant, butthe high-temperature MSBR can use the most efficient steam cycle thatis available so that heat reje~tion is minimal. In the reference design,with a plant thermal efficiency of 44%, 1225 W(t) of heat goes intoCQndeETSeP Water When the net eabectrical Output Sf the plant is 1800 b%?d(Et>.Another 43 N(t> is rejected from the drain-tank cooling system and plantheat losses.The expected rate of discharge of radionuclides in gaseous effluentsduring normal operation of m MSBR station is extremely small. Heliumthat contacts the fuel salt in the reactor is recycled after passingthrough a cleanup system whose only output is tritium and noble gasesin sealed cuntainers [I, p. PQ8]. Gases in the processing plant are alsorecycled except for a 0.5-scfnr stream of hydrogen that is vented afterits fission products, are trapped on ckarcoaP [2, p. 251. Alternatively,this hydrogen might be sent to the reactor part of the plant for use inremoval of tritium. The coolant salt c~ver-gas system includes provisionsfor re~novhg tritium that reaches the coolant system. The tritium thatdiffuses out sf the high-temperature salt system into the containmentcells is removed from the cell atmosphere recirculation system as PITO.Much of the tritium in the steam system blowdown (see below) probably=.:.:Ib..k.,k.,

391would be discharged into the air if- wet cooling towers were used. 0therwisevery little of the steam system tritium would go into the air.Tritium presents an unusua1 problem in molten-salt reactors becauseabout 2400 Ci/day is produced in the fuel salt of a 1600 paW(e> reactorand it readily diffuses through most metals at the high temperatures ofMSBRs * DCalculations indicate that in the reference MSBR about 790 Ci/dayof tritium might reach the steam system [3]. Virtually all of this wouldbe released by normal system blswdswn i n t ~ the condenser cooling water.Discharge of 790 Ci/day in a 560,006 gal/mrin stream produces a concentrationof 260 x kQe3 p@i/liter, This is 52 times the 5 x loma pCE/literconcentration used as a design objective for liquid effluents for lightwater-coolednuclear power reactors [4] e Presumably molten-salt reactorswill be required to attain similar low concentrations, so the unhindereddischarge of tritium in liquid effluents from the reference-design MSBRis unacceptable. Several modifications in the design or operation ofthe reference plant have the potential for drastically reducing theamount of tritium that reaches the steam system. These will be discussedlater in this chapter.Sampling such highly radioactive fluids as those in the MSBR withoutreleasing radionuclides to the environment is IXI simple matter. Experiencehas shown, howeverB that zero release is quite practicable when samplingequipment and procedures are designed for total contaiiiment To accomplishthis, MSBR sampler lines have multiple closures and terminate in chamberswith a controlled atmosphere that is recirculated through a purificationsys tern.Another occasion that requires special. precautisns to prevent releaseaf radioactivity to the environment is the opening of a system f ~ mainte- rnance. As described in Chapters 9 and 12, the MSBR containment and maintenanceprevisions are designed to confine any radioactive contamination torestricted areasNuclear SafetyThe general principles of nuclear safety are the same for all reactors.Small fluctuations in reactivity should produce only htghly- damped poweroscillations. Large, rapid increases in reactivity should be difficult toproduce and be easily controlled before the resulting power excursions producedamaging temperature or pressure excursions. The characteristics ofthe MSBR plant are such that these principles are satisfactorily met.The continuous removal of fission products and the adjustment of thefissile inventory in the fuel salt. during operation of the MSBR minimizethe amount of excess reactivity that must be compensated by controlb rodsand hence limit the potential for rapid increases in reactivity associated*Nearly ab1 is produced by neutron absorptions in Iitkium. By comparison,tritium production i~ 1600 MM(e> pHants amounts to 40-50 Ci/dayfor light-water, high-temperature gas-cooled , and fast-breeder reactorsand 3500-6000 Ci/day for heavy-water reactors

392with this excess. In the reference-des%gn MSISR, the maximum amount ofexcess reactivity that must be compensated by rods under n~r~lcal conditionsis expected to be less than lX 6lc/k.The fissile material in the on-line processing systems amounts toless than 1% of the reactor inventory. If all this CQUP~ be returned tothe reactor, the excess reactivity wodd be increased only 0.4% or less.Furthemore, conceivable rates of introduction are quite inconsequential,*and any unwanted reactivity increase from this source can easily be stopped.Decay of precursors in the fuel circulating outside of the core reducesthe effective delayed-neutron fraction from 0.38% to 0.12% in an operatingMSBR. Thus one result sf a cessation of flow is a 0.18% 6k/k reactivityincrease in a time on the order of the ha%%-lives of delayed-neutron pre-CUTSO~S. Somewhat larger reactivity effects of stopping and startingfuel circulation could result from the perturbations of temperatures andthe effects of changing pressure on gas bubbles in the core.%%le maximum effect of temperature changes is shown by the followingargument to be quite manageable. The reactivity coefficient for changesibl telltpe%atu%e Sf the entin^&? core (-8.9 X '6-l) i§ much smallerthan the coefficient for the fuel alone (-3.3 x 10-5 'c-1 for a uniformchange of fuel temperature over the entire core) e Thus the upper boundon reactivity effects due to a temperature change in the core correspondsto c~oling all the core fuel (and none of the graphite) from the maximumoperating temperature (705°C) to the fuel liquidus temperature. (5QQ"C).This is only about 0.7% 6k/k. affects actually attainable are smallerand, because they can be produced only by inflow of couPer salt, occurwith time constants ~f at least several seconds. The safety rod systemis quite capable of preventing power excursions due to such effects.Because of the strong absorptions in thorium in the fuel salt, displacementof a small fraction by voids has a positive effect: ~n reactivity.There are, in principle, two ways this could occur: 1) by increasesin %he volume fraction sf-circulating, nsncondensable gas and 2) by boilingof the fuel salt. Heither of these processes appears to be capable ofproducing changes of sufficient magnitude to represent a safety problem.Under normal operating c~~iditi~ns the fuel salt in an MSBR contains0.2 to 1.0 vol 2 of helium bubbles. This gas is introduced arid removedcontinuous~y to strip 135~e~mus, changes in the rate of addition orremoval or chan es in system pressure will change the core void fraction.At 1 vol 2, the voids in the reactor core represent abaut 0.039X in reactivity.A complete depressurization of the fuel system, which would alEQWthese bubbles to expand by a factor of 2 to 3, would cause a reactivityincrease of only about 0.lX 6k/k. In addition to the bubbles, the saltcontains some dissolved helium and the pores of the graphite containsubstantially more. Although the total amount of helium in the graphite&.dkThe rate sf reactivity increase would be only 5 x 6k/k perSeCOd if the 3 ?..iter/miam salt stream from the processing plant to the49,OQQ-liter reactor file1 systen contained twice the noma1 uraniumconeentratisn 0

39%is Large, the rate at which it can diffuse aut is limited so that forreasonably attainable rates of pressure Loss in an PaSlBR (% -2 psi/sec)the combined reactivity effect due to bubble expansion and graphite outgassingis only 0.005 (% b;k/k)/sec or less.Voiding of a few fuel channels by local boiling (as might resultif flow blockages occurred in individual fuel passages) is not a severeevent. me positive reactivity effect associated wfth 100 empty fuelcells* at the average nuclear importance far the central core region isless than 0.5% 6k/k. Because the boiling temperature (at ab atm) is morethan 700°C above the normal operating temperature of the salt, the energyinput required to heat the salt to the boiling temperature over MU& ofthe core in a short time would require a nuclear excursion larger thanany produced by credible reactivity inputs. Thus, boiling in the corewill not come into play as a positive reactivity feedback in any nuclearexcursion originating with the reactor at or near normal temperature.Displacement ~f small amounts Q€ fuel salt by graphite produces apositive reactivity effect. At the center of the core (the most sensftivespot) the effect amounts to 2.9 x 6k/k per cm3. Conceivableevents, including sudden redistribution of clearances in response toflow changes or accumulated stresses, produce no reactivity increase ofmuch CQ'ilsequenCe eA unique consideration in fluid-fuel reactors is the possibility ofinhomogeneity of the fissile material in the circulating fuel. Specificallyof C O R C ~ is ~ gradual ~ segregation of fissile material ~~tside thecore, f~llowed by rapid introduction with the incoming stream. %he MSBRfuel salt, as described in Chapter 5, is quite stable over a range ofconditions much wider than the anticipated deviations Seuranium could conceivably be produced by introductdon of reducing agentsor oxygen into the salt, but adequate protection against this fs providedin the MSBR.The response of the nuclear power to reactivity increases is g~vernedby the temperature coefficients of reactivity and the action of the controlrods and safety rods. Because the delayed neutron fraction is unusuallysmall, the MSBR power responds rapidly to reactivity increasesThe reactivity coefficients for uniform changes in fuel and graphitetemperatures are listed in Table 14.1. In response to reactivity transients,core temperature changes will not be uniform, however. In partfcular,the graphite will change temperature much more slowly than will thefuel salt. (In the central core region, graphite comprises about 90%of the heat capacity but only about 8% of the nuclear heat source is inthe graphite.) &~isequently, heating of the fuel salt results in a prompt,negative response of reactivity to a power excursion. This respo~se isgreat enough to limit effectively the initial power surge caused by anycredibly rapid increase in reactivity. Thus safety- rods are not requiredto sperate unusuaPly fast. The total core temperature coefficient (fuel

39 4ComponentFuel saltDoppler effectQThermal basebDensityTotal fuel SBt-3.28Gmphinite:Thermal hsebc2.47Bemsity -0.12~Total graphite 92.35COX -0.87“Primarily due to thotiurn.b~pw;rr~ shifts in thermal spect~~in increase reactivity becausefissile cross section decrease less rapidly than the thoriumCKQSS section does.

395plus graphite) is quite small, however, and might be positive. ('%he calculatedvalue is negative, but the margin of uncertainty is such that the.coefficient could turn out to be positive.) As a result, safety-rod actionis required to prevent wide variations in temperature that would otherwiseresult from any reactivity change that persists more than a few seconds.In summarys the nuclear safety characteristics of the reference KSBRare such that a reactivity excursion leading to a breach in the fuel systemcontainment is highly unlikely eProm the standpoint of reactor safety, the chief importance of radionuclidedecay heat lies in whatever threat it my pose to the integrityof the p%bElry System.In an MSBR, the fission products and transmutation pr~ducts and theheat produced by their decay are distributed through the reactor systemand the processing facility in a manner that depends on the physicalfeatures of the plant, the chemistry of the radionuclides, and the extentOf repPQcessing Of the fuel salt. In the ref€lreIlCe design HSBR afterseveral months of operation at its design power of 2256 m(t), the totaldecay heat rate is about $52 MW, distributed as shown in Tabbe 64.2.Most of the heat is generated in the fuel salt by the decay of all classesof radioactive products with half lives of a few minutes or less and oflonger lived products that are soluble in the salt.Krypton and xenon diffuse into the pores of the core graphite orinto the helium bubbles that circulate in the fuel salt and are rem~vedto the drain tank. About 80% sf the energy of decay of the gases thatreach the drain tank is released in the tank. The remainder is releasedin the carbon beds. For the purposes of conservative design, gach daughteratom of a noble gas atom is assumed to deposit on the surface nearest thepoint of decay of the parent atom and there to release its heat, Thiscertainly is the situation in the graphite and in the carbon beds. Dependingon the design, daughter atoms born in the off-gas lines or drain tankcould either be dissolved in fuel salt and returned to the primary systemor be ca~ried on with the gas stream.The fission products from element 419 niobium, through element 52,tellurium, are not stable.= metal ions under the normal redox conditionsin the fuel salt and are rapidly reduced to the elemental state,* Thesemetals are highly insoluble in the salt and are not wet by it, so theytend to deposit on metal and graphite surfaces and cohlect at gas-liquidinterfaces The distribution of the afterheat produced by these materialsthen depends OR their distribution among the various surfaces. Table 14.2shows two calculated distributisns for the afterheat from metallic fissionproducts reflecting uneert ainties in their phye ical $is tribution a Thefirst is based on an assumed sticking coefficient - that is, the probabilitythat a fission-product atom remains on the surface to which it migrates -* Because of this tendency to exist in the elemental state, thesefission products are frequently referred to as "noble metals" in MSW'ss.

Meat generation rate (MW)LOCatiOKl Of herat SOUPC& Bubble sticking Bubble stickingcoefficient = 0.1 coefficient = 1.0Fuel salt - all classes of radioactive products 102 102Gmphite in reactar vesselNoble gases and daughtersNoble metal depositsMetal surfaces in primary system - noble metd diepaGts 18 12Drain-tank systemNoble gases and c8;aughtersNoble metals and daughters1.44.8L.43.68.9 9.91 .% 8.3Off-gas system - noble g m s and daughters 2.4 2.4Fuel reprocessing plantFission groBuctsProtactiniumTotal6.6 6.65 .o 5.0- -152 152

397of 1.0 for metal and graphite surfaces and 0.1 for bubble surfaces. Forthe second distribution, the sticking coefficient for gas bubbles wasincreased ts 1.8 and no reentrainment was allowed in the salt in thedrain tank. The daughters of noble metals are also noble metals whichtend to remain where the parent was deposited except for iodine, thedaughter of tellurium. In the presence of flowing salt, iodine returnsto the liquid phase; otherwise it remains on or near the surface at whichit was formed.The fuel processing system is the final location in which a substantialamount of radioactive decay heat is released. The source ofheat there is roughly equally divided between 233Pa and fission productsof intermediate to long life. Decay heat generation is a most importantfactor in the design sf this systeml. The operation of the processingsystem scarcely affects the decay heating in the reactor during operation,but does reduce the heating rates at longer times after shutdown (anddecay of the shorter-lived fission products). Prom the foregoing, itis evident that three conditions of operation must be considered withregard to release and removal of decay heat: normal operation at variouspower levels, reacto~ shut down but fuel salt remaining in the primarysystem, and reactor shut down with fuel discharged to the drain tank.During power operation, the decay heat is only a small fraction ofthe fission heat and is of no consequence in the primary fuel-saltcirculating system. In other parts of the fuel system and in the off-gassystem, the design must accomodate the heating from this source. (Stagnantlines and pockets where radioactive liquid or gas can reside mustbe avoided, or cooling must be provided as necessary.)The drain-tank cooling system normally operates at only one-thirdof capacity, so here again the primary concern is to assure that theheat production is reasonably unifsrm thrsughout the tank and that linesleading from the primary system to the drain tank are cooled properly.In the off-gas system and in the reprocessing plant the heat loads duringnor-mal operation are the design heat 4i~aels. There are no abnormal. conditionsthat could cause these heat loads to be exceeded by significantamounts, so the primary concern in the design is to assure that thecooling is distributed properly Redundant capacity must be provided inthe cooling systems for the reactor drain tank, for the carbon beds inthe reactor Qff gas system, and for the fuel reprocessing system to assurethat cooling will be available at all times.When the reactor is shut down, the radioactive decay heat decreaseswlth time as shown in Pig. 14.1, In a normal reactor shutdsm, the fuelsalt is retained in the primary system for many ~QUX-B, the primary pumpscontinue to operate at full flow on the normal electric supply or at 10%flow on pony motors driven by an emergency power supply. The secondarypumps also csntinue to operate at flows in the range of 10 to IQOX, dependingon the power supply. Heat is transferred from the primary saltto the secondary salt, steam is produced at much reduced rates in thesteam generators and is discharged to the turbine condenser or to othercoolers. With PO% of their rated normal flow, two primary loops andassociated secondary losps will hold the temperature at or below thenormal level until 5 minutes after shutdown when one hop combinatisnis sufficient. In the absence of cooling, the te~~p~ature of the primarysystem would rise to 1408 and %5OO"F in 70 and 120 minutes,

3989ow M b- e, WMG 72 - 8 5 89A-FROM FlSSIOM PRODUCTS IN THE PRIMARY SALTB-FROM THE NOBLE METAL FlSSlON PRODUCTS,(Nb, Mo,Tc, Ru, Rk?, Pet, Ag, Sn, Sb, AND Te 1C-FROM THE NOBLE GASES, Kr AND Xe, AND THEIRDAUGHTERSB-FROM =%Pa AND LONGER LIVED FlSStON PRODUCTSIN THE FUEL WEQROCESSENG PLANT0.56.020.04Pig. 14.1. Time dependence Of decay heat. soklrcesin 8 1008 m(e) MSBR plant.

.... :.x.*399.....YS,,.... =..a 7 ......x.>.,.:j....-.,........I.. -respectively. The heat load in the drain tank begins to decreaseimmediately after such a shutdown. Heat production in the off-gassystem decays more slowly and the heat production in the processingplant is little affected for several hours,Under accident or other unusual circumstances the fuel salt 2sdischarged to the drain tank for cooling. Discharge sf salt from theprimary system into the drain tank could begin at shutdown and be completedin about 7' minutes. At this time the heat production rate in thedrain tank would be about 40 EN; the temperature of the salt would riseto a maximum of 1400'F in a few hours and then fall to about lO0O'F ina few days, where it W Q U be ~ ~ maintained by control of the cooling.Conditions in the off-gas system and in the reprocessing plantwould not be affected significantly by draining the reactore, but theconditions in the reactor primary system would be markedly different.Draining of the fuel salt would remove the fluid that transports thedecay heat from the graphite to the primary heat exchangers. If thesecondary salt were drained from the heat exchangers at the same time,the preferred means fur removing the heat from the decay of noble metaldeposits on the heat exchanger tubes wouPd also be removed.Calculations have shown that the cornp~~~nt~ and piping in the primarysystem could be designed to be cooled adequately by pl~~viding asystem that would maintain the cells walls at HQBO'F. Heat would betransferred by sadiat ion and conduction within the components andwould be radiated to the cell walls. The temperature at the centerof the graphite core in the reactor vessel of the reference designwas estimated to reach a maximum of l90Q"P after 14 hours, but thevessel walls would not exceed 14OQ'P. With some modifications S€ thecurrent reference design, the center tubes of the primary heat exchangerswould not exceed 20BQ"F, and the outer shell would not exceed P400"F.These temperatures are believed to be acceptable for the Sew times thata drain at shutdown would be expected to occur in an MSBR. Belayingthe drain by 24 hours a d cooling the plant to 1058°F during that timewould reduce the decay heat rates by a factor of about 10 and wouldsubstantially reduce the temperature riseInteraction of MaterialsAccording to the preceding secti~ns, the integrity of the salt containmentis not seriously threatened by either nuclear power excursionsor afterheat. In this section, we consider threats from (a> normalsystem corrosion and (b) possible pressurization and enhanced attackresulting from a small leak between coolant and fuel salt or betweensteam a d coolant salt......'.!.51.1C~rrosiort. - Hastelloy N is corroded by MSBR fuel salt under normalconditions by reactions of the type....i . I. 2.L....x.p;....#@2

The ratio UF~/UFL+ in the fuel salt is maintained at a value such thatequilibrium is reached with concentrations of EJI~+~ pe2fg and MQ~+ inthe salt that are much less than 1 ppw and a chromium concentration lessthan QIE hundred ppm. The predominant corrosion reaction then is reactionQf UFL, With Cb-n~Qmitlm in the rsgtal to form e9-F~ Which dissQ1VeS fnthe salt. Chrsmi~m becomes depleted near the surface and after the firstfew thousand hours corrosion is limited by the rate of diffusion of chromiurnto the surface. At NSBR temperatures this limit is on the order of0.1 mil/year of chrorniua depletion.Introduction ~f moisture md air into the fuel system produces BFand metal oxides, which dissolve in the salt and make it WOE oxidizingand more COI-KOS~V~to all ~onstituents of Hastellsy N. During normaloperation the CmtaHlinantS can be kept Pow by CQntrQbling the composit~srmof the CW~K gas e Xaintenance operations will almost inevitably introducesome moisture, but with reasonable precautions to minimize air inleakage,COPI-Qsion from thiS CilUSe W i l l haVe negligible effect On CQn%ai~?JEIIt.(eorrOSiQn in the fuel system which was opened two cx three timesa year, averaged on1 0.1 millyear [5, pp. 71-79].)A phenomenon that presumably would affect Hastelloy N in the fuelsystem of the reference MSBR is the intergranular attack and crackingobserved om surfaces exposed to the MSRE fuel salt. As discussed atlength in Chapter 7, the cause appears to be fission-product tellurium,but the existing information does not permit a reliable prediction ofbehavior over many years and numerous stress cycles. Some indication ofthe seriousness is the observed effect on the MSRE heat exchanger tubesafter 24,580 hours at hi temperature following the beginning of poweroperation (and fission-p duct deposition) Surfaces eXpQsed to fuelhad crack at almost every grain boundary, to an average depth of 5 milsafter being strained. If the depth of the intergranular attack continuedto increase with the one-fourth power of time at temperature (asgests is reasonaFke) the average crack depth after 30 yearshours) would be only 9 mils. This depth of cracking in thereference MSBR would have only insignificant effect on khe strength ofthe r@aCtOK vessel and piping. On the other hand, if the attack proceedsmuch more rapidly, as is conceivable, it could seriously affect the reliabilityof the fuel containment.* It is clear that safety considerationsrequire that these ques tions be favorably and conclusively resolved.~sssibilities are discussea later in t'tl~s Chapter.. ..u.5Coolant Salt Interactions. - The eutectic of NaBF4 - NaF (92-8 mole X>was chosen as the secondary salt in the reference MSBR because it hasreasonably good coolant properties ~ it is relatively inexpensive, and itsmelting point (925°F) is ICW compared to that of other suitable fluoridedXtUreS. FrOlll the Safety StEXldpoknt it is iElpQ$t:ilhlt that HliXillg Of fuel

......-401.....,:.:,A.xw. . :.

482and disposing of the contaminated coolant salt promise to be unpleasantoperations The chemical toxicity of the boron trifluoride precludesits indiscriminate release from the plant, but the presence of sodi~mfluorobsrate does net otherwise affect the safety of those operationsSodium fkuoraborate, if spilled into the reactor cell or into asecondary cell, must be contained. The salt contains radioactive sodiumin a concentration of 0,6 Ci/ft3 and some tritium, and the sodim flue-Pdde Xid boron trifluoride %re both toxic CkaeMiCdS * %e COnfainEEIltcreates no special prsblem however, because the cells operate at:temperatures below 1100'F at which temperature the BF3 pressure overthe salt is only about 0.3 atm,Water and steam react with sodium fluoroborate to produce primarilyhydrogen fluoride and sodi~m hydroxyfluoroborate The reactions arenot destructively exothermic, but the hydrogen fluoride is corrosive tothe metals of the react~r secondary system and the tubes that separatethe fuel salt from the coolant salt. Although the corrosion rates arenot catastrophic under any foreseeable circumstance, the Beakage rate ofwater from the steam system into the secondary system and the hgrdr~genfluoride concet~trati~n in the secondary salt must be kept low in orderto maintain a pow corrosion rate sf piping ma equipment.In the event of a rupture of one or more tubes in a steam generatoror superheater, the sapid pressurization of the secondary system and thepossibility of transmitting that pressure to the primary system is themjor concern. Isolation valves must be provided to stop the flow offeedwater and steam to che faulty steam generating equipment and pressurerelief devices must be provided on the secondary system to keep thepressure below the system design pressure. The steam and salt that aredischarged through these devices must be contained. me affected secondarysystem must be purged of hydrogen fluoride and moisture and the CORtaminatedsalt must be purified ur replaced while repairs are made on thesteam generator before operation of the plant cam be resumed.The use of a chennieallby reactive coolant in the secondary system ofthe EIISBR intraduces same problems in designin the plant for Upset CQRditions.The interactions of the coolant with the materials, with fluidsin contiguous reactor systems and with the cell atmospheres, however, donot appear to be 80 vigorous or the reaction products so aggressive asto create major safety concernsThe basic function of the engineered safety features in a moltensaltreactor plant is the same as in my nuclear plant - to prevent anyu~cogltro~ied Pelease of radiQaCtiVity maer acciaent conditions. medetailed requirements are unus~al, however, because sf the nature of thefuel - liquid, but practically nonvolatile and not highly reactive withair or water.Previous disc~ssion has indicated that abnormal conditions withinthe primary system 06 a molten-salt reactor -nuclear excursions trnduncontrolled fission product heating - do not pose major threats to itsintegrity. Of the CQnditiOnS considered, only the prevention of pressure

483.....L.x,.......-.e,:.a.........y,.: . Y .;excursions that could be initiated by large leaks between the steam andsecondary salt sys tern require the implementation of specialized safetydevices to protect the primary system boundary. These devices must insurethat the secondary salt system is reliably and effectively vented in theevent of a steam generator failure.Despite the low probability of a breach of the primary systemboundary, the consequences of such a failure must be considered. Becausethe fuel is in liquid form, my primary-system rupture releases largequantities of radioactive material into the imediate surroundings Toprevent the dispersion of that activity throughout the reactor building,the components of the fuel system are enclosed within a primary csntafnmentsystem of sealed cells from which water is excluded. The systematicexclusion of water guards against the generation of Parge volumes ofsteam from the sensible heat of the fuel salt and thus Pimiss theincreases in primary containment pressure to small values even for majorsalt spills. A secondary containment system that encloses the equipmentcells provides additional protecti~n against the release of radioactivityto the environment -The most unusual of the engineered safety features in the MSBR isthe provision for dealing with afterheat under accident conditions - theheat source is Ped to the c~oHing instead of vice W P S ~ (as in the ECCSfor a light-water-cooled reactor) As described elsewhere, the bulk ofthe fission products stay in the fuel salt, making shutdown cooling forthe fuel salt essential for prevention of excessive temperatures - Theultimate cooHi~ng system is in the drain tank, so the reactor and containmentare designed so that the fuel will get to thac tank under any credible accident conditions o The heat removal sys tern is simple ruggedalways operating (being used to remove heat from off-gas sources), andcan continue to operate without electric power and unattended to coolthe fuel as long as necessary in the design basis accident.In connection with the afterheat removal, it is worth noting thatproblems associated with it are much less i~-~tense in an MSBR because themajor source is inseparably associated with a very large mass of salt.(me ratio of heat source at shutdown to heat capacity in the MSBR fuelsalt is only about one-tenth of the ratio in the dry core of an LWR.)~ecause the heat source is so dilute, the "~hina syndrome" does notappear to be a serious problem in an HSBR.Siting Considerations.:.s

404regulations my have been developed for "power parks" that include reactorsand a reprocessing plant. For now, the only question that can bewell defined is whether or not an MSBR plant, similar to the referencedesign, can meet the guidelines and restrictions on siting that nowapply to commercial power reactors. We shall address this question inthe course of the following discussion.Factors affecting the siting of a reactor include:1 transportation requirements during construction,2. transportation of fuel, etc., to site during operation,3. tPaTLSp09t.%%ion Of fuel and Wastes from Site duaflskg OperatiO€2,4. discharge of materials and heat to the environment dusimgnormal operat ion I5. consequences of credible accidents (inventories of fissionproducts a d actinides, fractions likely to be released,etc.), and6 disposal 0% materials after decomissioning.We shall consider each of these in turn.Transportation During Con~tru~ti~n. - The Parges t components of anHSBR are the reactor vessel (2%-feet diameter x 33 feet high, 155T), thesomewhat smaller drain tanks, and the primary heat eXCfnaRgePs (6-ftdiameter x 24 fe: long, 53 TI. These are similar in size to item ina IOQQ-PaWfe) light-water reactor and pose similar transportation problems. - %he flow of materials i~to amdout of a 108O-rnJ BR plant are as shown in Fig. 14.2. (Graphiteshipments are expressed as average rates, equivalent to replacement 0%the 176 '6 of graphite in the core at &year intervals.) Plant inputspose ka0 p-POb%eSIlS of sr2lnSpOr%ZltiOne BeCauS@ 0% the Qn-Site processingana decay of high-level was s, the awemt of intensely radioactivematerial shipped out of an BR plant each year is far less than thatLeaving any other reactor station sf cowparabble electric capacity. Insteadof short-cooled fuel elements, there are separated fission productsthat have decayed for years: high-level wastes from processing are accumulatedin tanks for 4-5 years, then stored on site f o nine ~ yearsbefore shipment. The volume and radioactivity of these wastes are aboutthe same as. those of high-level waste and cladding hulls that are ultimabt@EJTShipped from Zl PepPoCeSShg plant Serving 8 1066 m(e) m R [6].Krypton-85 and tritium will be stored md shipped in l.5-ft3 cylindersat 1800 psi - seven per year. Thg reference-design MSBR provides foraccuneulation within the reactor building of all the graphite removedfrom the reactor over the life of the plant. Thus, it can be packagedand disposed of on a convenient schedule, possibly- as part of the decommissioningprogram.

NOBLE GAS AND TRITIUM REMQVAL7 CYLINDERS /YEARHIGH LEVEL SOLID WASTECHEMICALS20 TQNS/ YEARPRQCESSINGICI6 lb/YEARHIGH LEVEL SOLID WASTE 136 d/YEAR( 7 CQNTAlNERS / YEAR, 0.25 #w/CQNTAINER 1- LOW LEVEL SOLID WASTE (4000 ft”/ YEAR)p UNCONTAMINATED LAUNDRY AND SHQWEW WASTE ( 300,000 gal/ YEAR 1

486Effluents Practically the only radioactive effluent eitherga%eQUS Or liquid, from &f%BR phnt 5s the tl-itiUlTl that: reaches thesteam system. In the reference design, where 790 Ci/day is released ina %GCE,QOQ gal/min stream of cooling water, the effluent concentration(0.26 x ~ C i / d > is a factor of 12 below the current lQ CFR 20 limiton releases to unrestricted areas [7]. Ora the other hand, it is a factorof 52 greater than the AEC's numerical guidelines for effluents fromlight-water-cooled reactors [43.As described Eater, there appear to be ways of limiting effluenttritium to about the same as the LhJR guidelines. In this case, tritiumwill pose no unusual siting requirements on the MSBR.Site requirements connected with the discharge Of waste heat are thesame for an MSBR as for a modern fossil-fueled plant of equal electricaloutput and less demanding than for current power reactors.Radionuclide Inventory. - As a result of the on-site processingfacility and the attendant storage OE separated fission products, theinventory of radioactive isotopes expected to be present at an NSERsite is considerably reater than that present in other nuclear powerE--.be present not in the reactor but in isolated, protected: waste storagetanks in the form of relatively stable fluoride salts. The inventoryof radioisotopes in the reactor will be considerably lower than in otherreactor types as a result of the c~ntirauous processing of the salt. Themaximum in~ent~~iesof several raaionuciiags expected to be present inthe fuel salt and precessing plant of a lQOQ-~(e) MSBB are listed ina(ZLblE? %IC a 3 a hV€!ntorieS %OW in a PIR and MFBR of comparable size areshown for comparison. The MS B inventories are iven just prior to shipmentof high level waste to a federal ~pository while the PWR and LHFBRin-~e~~tories are given just prior to refueling.Besign-Basis Accident. - In the MSBR, the design-basis accident isa rupture of one of the main fuel circulating lines that ~ccurs while thereact~r is at full power a d quickly spills the entire charge of ~n~ltenfuei. me primary containl~l~nt is desipea to prevent any release sfradioactivity into the reactor building 0% environs in this event. Considerationsf the fuel salt chemistry, the intensity of afterheat sources,and the dependability of the drain-tank ~~oling system support the conclusionthat this objective is attainable. Thus the design-basis accidentshould not affect the health and safety sf the public.Deconwnissisning. - Presumably at the end of plant life the radioactiveequipment and materials in the reactor and processing systemmust be removed to some ultimate disposal facility.here will be 1700 ft3 of highly radioactive fuel salt, 8400 ft3 ofcoolant salt, 175 ft3 sf Pa decay salt, 60 ft3 of bismuth, and 20 ft3of lithium chloride, all radioactive. There will also be several hundred

...... i. rc,40%....;.:,

408md about 3000 ft3(STP) of helium, with little or no radio-The fuel salt, bismuth, and LiCl are sufficiently valuablethat they will likely be recovered for reuse. The NaK would also besssibly as much as 1200 tons of graphite may be in the reactorbuilding at the time of deeo issioning (the fixed graphite plus 6 replaceablecore assemblies). This can all be broken down into piecesthat can be conveniently shipped in shielded containersThere will be many equipment item that are highly contaminatedwith fission pr~ducts. The largest, such as the reactor vessel, thedrain tank, and the primary heat exchangerss must be cut into pieces fortr~~-~sportati~n and disposal. This can presumably be done within the containmentalready provided within the MSBR building for use during maintenance.The pieces then should be as manageable as the most radioactiveportions of other kinds of reactors upon desomissioning.Thus it appears that disposal of an MSBR, while clearly a majorundertaking, will be manageable within the teChnQlsgy needed far maintenanceof the MSBR and dispgtsal of other types of reactors.c,Sumaary. - On the basis of the foregoing comparison it appears thatthere shgt~ld be no major differences in the siting requirements for a~UPE~- aeveioQedi MSBW a d for other types of advanced power react~~s.Reactor ~rechn~loa in general and the years of AMP and %RP work inparticular provide the information needed to answer nearly all of thequestions that are i ~~p~ta~t to molten-salt reactor safety This sectioncontains a brief review of the pertinent information that is in hand,with some discussion of its adequacy. The few important gaps that remainto be filled are noted in passing. The significance of uncertainties andthe needs f~a: further work are discussed in the next major section ofthis chapter.The ultimate reliance for protection of the public from an MSBRaccident rests on the csntaiment system that is in effect during operation.This system does not involve any untried construction techniquesbut it dues have a major untried feature. The inner walls Q% the reactorcell and the fuel salt drain tank cell must be insulated and the cellsmust be heated and operated at temperatures above 1080°F. This featureis discussed in Chapter 9. As pointed out earlier in this chapter, thecharacteristics sf MSBR fluids innpose no severe problem with regard topressurization 0% danger of chemical reactions Therefore, PSBR cellsand b~lildi~~gs can be mostly designed and built with the ~ontain~~ent technologythat has been thoroughly aeveiopea for other reactors.

409....

418Unavoidable uncertainties in the breeding ratio alone will be equivalentto perhaps one percent of the reactor fissile inventory per year, anduncertainties ifa long-term fission product poisoning fulrther obscure reactivityevidence sf fissile hideout. The reactivity balance should revealsignificant segregatisn that occurs within a few days, and there mightbe other clues that would permit detection of gradual hideout, but thenost dependable, safest course is to preclude trhe possibility. This means(a) using a salt mixture whose behavior is thoroughly known, and (b) operatingthe plant so as to steer well clear sf any condition that could resultin segregation of uranium (ur plutonium).The phase relations in pure EiF- eF2-'%IaFl+-UPL+-UP3 mixtures arequite accurately known. Conceivable variations of the fluoride ratiosin MSBR fuel from the nominal composition do not approach any region offuel segregation. Oxide behavior is also well known, but here there isless latitude. Lntrod~tion of oxy en into MSBR fuel w ~ l result d inprecipitation of a uranium-rich mixed oxide (TTI~~-UQ~) when the oxideion concentration reaches somewhere between 36 and 156 ppm (see Fig. 5-4).The concentration in an operating MSBR must be kept below about 30 ppm,and there is good reason to believe that it can be. The ingress of oxygencan be limited, as shown by MSRE experience, to rates and amounts thatcould easily be removed by the EriSBR processiiag system. In any event itwill be necessary to verify that the oxide concentration in the fuel issafely low by frequent accurate measurements e Techniques for oxide malysesthat are currently available are mot adequate for the HSBR needs.u .,Fiss ion-Product BehaviorThe general behavior of fission products in molten-salt systemshas been largely established by a variety of independent studies andby analyses of the MSM perfol-man~e, both during and after its operation.Tie details of that behavior were described in Chapter 5, bus there aresome aspects that are of particular interest with regard to nuclearsafety. expeetea, the nobie-gas fission products (xenon ana kqpton)were readily removed from the circulating fluid and transported ks theoff-gas system. Although there was significant transport of thesematerials to the unseaEed, graphite used in the MSE, a factorof 6 reduction in xenon poisoning was achieved with a simple gas-strippingsystem. once reHtoved, these gases be effectively and predictablyhandled in the off-gas treatment facility. Post-shutdown release ofgases previously held up on the graphite my, however, be a radi~logicalsafety consideration in primary system sf an MSBW.In the XSRE, at least some 0% the noble-gas daughters formed in theoff-gas system were carried along by the gas stream, and large quantitieswere accumulated in the particle traps upstream of the charcoal beds.Since the 0ff-ga.S had nCI further eXpQSUri2 to the fuel Salt, I20 i~%C?~~t~Onwas obtained with respect to redissolution in the salt, which c~uPd beiElpOrtsLnt ill the MSBa drain tank.The fission products that from thermodynamic cmsiderat ions wereexpected to remain dissolved in the fuel salt were shown in the MSRE tobehave as expected. hareg the species of particular interest from aw.,

411. &i:U!radiological safety standpoint, both iodine and strontium showed nutendency to escape from the salt. Some iodine did appear in the gasin the reactor loop after salt drains, due to decay of precursors thathad been deposited on system surfaces. Again, special steps will berequired during some stages of post-shutdown operations of am MSBR toprevent the release of iodine formed in this way.Many of the noble-metal fission products were fumd on surfaces inthe MSRE. Ef the apparent sticking coefficient of noble metal atoms tometal surfaces is taken as l.Q, then data %row the MSRE indicate that theapparent sticking coefficient to graphite was 0.5-1.8 amd to gas bubbleswas

must be made and appropriate criteria must be developed by which MSBWsafety can be measured before molten-saEt reactors are accorded thedegree of confidence now enjoyed by Pight-~ater-co~led reactors,More information is required an the behavior sf fuel salt boilingin contact with graphite at temperatures to 3000°F. Theoretical consideratiansand the few data that have been obtained indicate thatthere will be no significant interaction but this must be confirmedill tes ts Of longer duratiOKl. CQIIsiderable infOrmation 2s aVailElb leon the volatility sf fission products in fuel salt at high temperaturebrat additional data are needed that are more directly related to theconditions of a design basis accident. At high temperature, mst ofthe iodine released by the decay 0% teal~riuw deposits on surfacesand would be expected to react rapidly with the metals in the reactorsystem md be retained there. Experience with the MSRE gives no duesconcerning the distribution of the iodine between the metal and a gasover the metal. Data are needed for a variety of conditions so thata good analysis can be made of the behavior to be expected of thisiodine under accident conditions and durin maintenance of the reactor.area in which moaifieations in the reference design are certainlyneeded and essential informtion is lacking is tritium containment .The remainder of this section is therefore devoted to this topic.The amount of tritium that could reach the steam system sf thereference-design MSBR has been estimated to be about 11.3 of the2420 Ci/day production rate in the fuel. baodifieation of the steamsystem and its operation to retain that amount of tritium would beimpractical. Ss would be attempts to separate the tritium from thenormal hydrogen in the steam. Clearly means must be provided to limitthe tritium that reaches the steam to an amount that can be dischargedsafely to the plant environs with the cooling water from the turbinecondenser. At least BO Gilday a€ tritium could be released to a river(or to the atmosphere) in the condenser cooling water from a 1000-rn~(e)plant ana still be within the guidelines 0% the propose~l ~ppendix I to10 CFR Part 50 "as Paw as practicable" criterian for light water reactors.A tentative design objective for the MSBR is to limit the tritiumrelease to about 2 Cidday or 8.1% af the production. (This is the rateshown in Fig. 14-2*)Several modifications in the design or operation of the referenceplant, separately or in combination, have the potential for drasticallyredwing the amount of tritium that escapes into the steam system andin some instances into the reactor and coolant-system cells. Thesemodifications involve adding hydrogen to the fuel salt I reducing thepermeability of the metal walls, substituting side-stream contactingof salt and gas for injection of gas bubbles into the primary and SETQ~darysystems, exchanging tritium for hydrogen in hydrogenous compoundsor reacting it with oxide in the coolant salt, and using other fluidsto couple the primary system to the steam generators. Several possibilitiesare discussed in the following paragraphs.We have inferred from some of the salcukations that the effectivepermeability of the metal in contact with air in the MSRE mfght havebeen Qnby l/rsos of the pe9lIEability Sf UnOXidiZed metal. Oxide fikElSOk9 met&. SUI-faceS have been fQUId at 0 L and by other investigatorsh...-.&.

....

414With more drastic changes in plant design, helium containinunts of oxygen and water vapor could be used as the coolant idized to water and prevented from passing into the steam. Objection tothe use of helium in the secondary system is in the high pressure,the larger primary heat-exchanger surface, and the larger fuel-saltinventory that would be required. These objections might be sireumventedby employing the helium in the annuli sf dual-wall tubes in thesteam generators at the expense of larger and complicated steamgenerators eUse sf the nitrate-nitrite salt mixtur@, generally known as HTS orBites, in the secondary system would also keep tritium out of the steam.Tritium entering this salt would be oxidized to water, and the waterwould be vaporized into the purge gas at high temperature. Thermalinstability &QVe 1 l ~ ~ ' md P reEtCthOn§ W i t h graphite if it Were %Oleak into the primary system are objectionable features Q% this salt.These difficulties c~uld be circumvented by use of the salt in a circu-Hating system between the reactor secondary system and the steam generators*In view of all the possibilities, it seems certain that the escapeof tritium into she steam can be limited to acceptable amounts. Thetritium problem is not completely solved, however, until methods arespecified for conftning the tritium that is removed from the reactorthe sell atmosphere ana this has not yet received KUS~ atten-Most sf the critium is likeby to be extracted as water or tritiumfluoride. The water could be stored in tanks; and the tritium fluoridecould be sorbed on sodium fluoride beds, or those compo~ds might bedecomposed and the tritium converted to a solid hydride for storage. Inany event, excessive dilution by hydrog@~ must be prevented. A productionrate of 2420 Ci/day is equivalent to a trivial 0.8 ml/day of T20, mevolume of water resulting from a dilution of with hydrogen WOUI~ beof little consequence, but, if the dilution were 106, ~ OOO m3 of tritiatedwater would be produced during the life of the piant. safe StOKage forsuch a large volume would be expensive, and mans probably would haveto be provided for ConCentPating the tKitiklm.Although there is experience with tritium in the MSRE, analysis ofits behavior in large molten-salt reactors requires more detailed infermation.Some extrapolations of data from the literature can be mde, and apr~g~am is in progress to obtain confirmatory data. The program includesYneaSU%€XEntS Qf (E) the %olUbi%ity Of hydrQ en in salts; (2) the permeabilityof metals and oxide coat%ngs at low partial pressures of hydrogen;(3) the capticities sf graphite and 0% potentid coolants to retain hydrogena d tritim under simulated reactor condieions; md (4) reaction ratesen fluoride in %ow concentrations in salts with metals. Investigationof methods for separating tritium cowpomds from process streamand frola cell atmospheres and for storing the tritium safely and econom5-calPy while it decays will be included later.u..... c.3k...

.,. ....;.m. 415Eva1 uat ion.... s.22...g&.:&Waste heat from a high-temperature NSBR power plant is as %ow asfrom the most modern steam plants The plant can be desithere is practically no radioactivity other than tritium in the planteffluents. NQ shipments of short-co~led fuel leave the plant; instead,fission P ~ Q ~ U C ~ S shippea as concentrated higa-le~e~ waste afterseverd years? decay, while other wastes (such as core graphite andcharcoal) are ascumulated an-site to be disposed of at any convenienttime 6Tritium is a special problem because of its high rate of producti~i~in the fuel salt and because it readily diffuses through metals at MSBRtemperatures. In the refe~ence MSBR, with no special measures for blockingtritium diffusion, about $98 Cilday (33% sf production) would reachthe StC%Uil Syst@m. SeVeKal atQdifiCatictnS in d@Sfgn and Operation havethe potential for drastically reducing tritium escape by chis route.The objective of Bi~iting tritium release to within present Me guidelinesE Q ~ light-water-cooled reactors appears attainable , but the bestmeasures are yet to be chosen and demonstrated.The situation with regard to nuclear safety and afterheat is unique.The very limited excess reactivity and potential for reactivity increasesin an WSBR, c~etpled with favorable dynamic characteristics make damagingnuclear excursions highly unlikely- Afterheat prob kms are not intensebecause the bulk of the fission products are incorporated in a large massof fuel salt. Furthermore, thls heat source can be gotten into a reliablyco~led situation (the drain tank) under any accident condition. Radionuclideheat sources in the pro~essing plant, in the reactor off-gassystem, and deposited on surfaces in the fuel system require cooling,but simple, reliable measures appear to suffice.Although a breach of the fete1 system is highly unlikely, the designbasisaccident is taken to be a major rupture of a fuel line that quicklyspills the entire fuel inventory. Containment of the radioactivity inthis event is the chief safety consideration in an NSBR, This task issimplified because the acti~ides and the bulk of the fission produsts stayin the salt, the salt has an extremely low vapor pressures and it is nothighly reactive with m~istbtre OF air.It appears from basic considerations that site requirements for anMSBR paant should @Qentually be different from those for other reactorsof like power. Because of the unusual nature of an HSBFI, however,it will be necessary to begin with fundamental principles and developcriteria appropriate to this kind of reactor, then to perform a safetyanalysis comparable in depth to those for reactors ROW going intoOpe.k-atiOn.

416References for chapter 14I