4’04 ORNL/TM-6413 Molten-Salt Reactors for Efficient Nuclear Fuel Utilization Without Plutonium Separation J. R. Engel W. R. Grimes W. A. Rhoades J. F. Dearing Printed in the United States of America. Available from National Technical Information Service U.S. Department of Commerce 5285 Port Royal Road, Springfield, Virginia 22161 Price: Printed Copy $5.25; Microfiche $3.00 This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, contractors, subcontractors, or their employees, makes any warranty, express or implied, nor assumes any legal liability or responsibility for any third party's use or the results of such use of any information, apparatus, product or process disclosed in this report, nor represents that its use by such third party would not-infringe privately owned rights. - -) ORNL/TM-6413 Dist. Category UC-76 Contract No. W-7405-eng-26 Engineering Technology Division MOLTEN-SALT REACTORS FOR EFFICIENT NUCLEAR FUEL UTILIZATION WITHOUT PLUTONIUM SEPARATION J. R. Engel W. A. Rhoades W. R. Grimes J. F. Dearing Oy » Date Published ~ August 1978 NOTICE: This document contains information of a preliminary nature and was prepared primarily for internal use at the Oak Ridge National Laboratory. This report was m;:'c:m account of work . i S Prepared by the zmmmxfigixnmfimfikfififlfiafi OAK RIDGE NATIONAL LABORATORY | ey, expem o it o s iy b Oak Ridge, Tennessee 37830 | o it ot any et eeoon o operated by oty vy at s e would o ) UNION CARBIDE CORPORATION - for the DEPARTMENT OF ENERGY DISTRIBUTION OF THIS DOCUMENT I8 UN. ur - ) xn # iii CONTENTS SUMARY ¢ 0 ¢ & 0 o 2N E s 00t eSS ABSTRACT INTRODUCTION ...... BACKGROUND HIGH-ENRICHMENT MSRs ... S 0 o @ & 8 b o H s & e e s eSS e S s e s BB BOREEEIBEEEEBE B * ¢ ® O F O PSS 8PS e s ORNL Reference Design MSBR ...¢c00... esssesssessnasessens Reference Design Variations . ..... Plutonium Transmuter for 233U Production ...eeeecesocecssesss . DENATUREDMSR .......l‘......‘I...'...._’.. S ® & 9 S S ST S S O 0D S eSO GeneralCharacteristics ® 0 2 0 QP00 E PP SR OO ESOeNS e e N Reactor Characteristics .vceceesses Core Thermal HydraulicsS ...cececececsss: a0 0" " O e P e Chemical Processing '.......'....I.j......'.........-I..'. Balance Of Plant @ % 9 % &0 00000 MSR TECHNOLOGY STATUS .cceceeecee REFERENCES ® 9 * & 0 08 " 8 0 e e 0 0 Page i W= = 11 13 14 14 16 31 33 39 39 40 -t ot SUMMARY Research and development studies of molten-salt reactors (MSRs) for - gpecial purposes have been under way since 1947 and for possible applica- tion as possible -commercial nuclear electric power generators since 1956. For the latter, the previous .emphasis has been on breeding performance and low fissile inventory ‘to help limit the demand on nonrenewable natural re- ‘sources (uranium) in an ‘expanding nuclear -economy; little or no thought ‘has been given to alternative uses of nuclear fuels such as proliferation of nuclear explosives. As a consequence, the conceptual designs that evolved (e.g., the ORNL reference deéign'MSBR)Vall favored enriched 233U as fuel with an on-site chemical processing facility from which portions of that fuel could be diverted fairly easily. With the current interest in limiting the: proliferation potential of nuclear electric power systems, a redirected -study of MSRs was undertaken in an effort to identify concep- tual systems that would be :attractive in this situation. It appears that practical proliferation-resistant MSRs could be designed and built, and this report describes -a particularly attractive break-even breeder that includes an on-site chemical reprocessing facility within the reactor pri- mary containment. - The point of departufe~for-this~9tudy (as for other recent MSR studies) was the ORNL reference design MSBR, which in many respects, reflects the state of MSR technology at the end of the reactor:development " program in fiscal Year*1976;1:Thisfreactor*Was”chafacterized%by-a'moderate breeding rati01(®1i07);'aélo# specific inventory of fissile ‘fuel [v1.5 kg/ “MW(e)], a reasonable-fuélfddubiing time (Vv20-years), and -almost no plu- toniUm’from*thé~fué1*cy¢1é;*5ThisipérfOrmance3waé‘t6*befachiewedithrOUgh ‘the use of ‘fuel highly ‘enriched in 233y and -23%y (v72%) in-d high-power- ‘density ‘core and afieon;sité>f1sgion;pfoduét-c1eanufi,systemvwifih»a 10-day fuel'brocessing*éjéle&?*TWdiimportant*Stéps5in-inéstoéésSing’CYCIé'fiere (1) the isolation’of the enriched uranium ‘from, and its subsequeat return to, the fuel salt and (2) the isolation of 2?%Pa for decay to *®®U outside “:“the reactor neutron flux to prevent counterproductive neutron captures in vi the protactinium at the high flux levels* in the reactor. Both of these steps, aldng*with the ready availability of'excess bred fuel, were perceived “to contribute to the proliferation_sensitivitj of the reference concept. ‘A preliminary studylwas undertaken‘1ate‘invéalendar-yearA1976:to see ~1if the reference MSBR concept could be modified to significantly enhance " its proliferation resistance. Among the modifications considered were ~* ‘elimination of the breeding gain, a reduction in power density (and spe- cific power) so that protactinium isolation could be avoided without ex- i cessive penalties, and ‘several conceptual variations -in the fuel processing cycle. Reduction of ‘the fissile .uranium enrichment (i.e., denaturing) was not considered at that time because'offperceived problems with the- attendant plutonium production. ‘The net conclusion -of this study was - that, while some enhanced proliferation resistance could be achieved, : . the reference MSBR concept probably could not be made sufficiently re-. - .gistant to allow its deployment outside areas that would be "secure" against diversion of fissile material or proliferation. In a minor extension of the above study it was shown that, if MSRs were confined to "secure" areas, they could also be used to produce power from fission of plutonium (generated by other reactors) and to convert- thorium to 233U for subsequent denaturing and use at dispersed sites. - Since the confinement of MSRs exclusively to "secure' sites did not ap- pear to be desirable, no further consideration was given to concepts without denatured uranium. / The current study of proliferation-resistant systems is based on the premise that MSRs would be attractive for dispersed deployment if they could operate with denatured uranium fuel, have good resource utilization ,characteristics, and require no fuel reprocessing outsidélthe reactor - . .primary containment envelope. A number of molten-salt concepts may meet these requirements, but the one that currently appears most attractive, is a system with_denatured7fuel and a net effective lifetime breedingnz ~ratio of 1.00. This implies that, once such a reactor-were-supplied with o v}w»*Npt related to proliferation, but a potential technical problem, was the fact that portions of the moderator graphite in the MSBR core would have to be replaced every four years because of neutron radiation damage at the projected high flux levels. L vii a fissile fuel charge, it and succeeding generations of hardware could operate indefinitely with no further addition of fissile material. Addi- tions and removals of fertile material — both *3%U and ?°?Th — and other salt constituents would, however, be required to maintain a stable chemical composition. : .- Break-even breeding in a denatured MSR is achieved by making several changes in the reference design MSBR concept. First, changes were made in the reactor core size and salt-graphite configuration to lower the core power density and to enhance neutron resonance self-shielding in the 238y in the fuel. These changes increased the fuel specific inventory somewhat (to about 2.4 kg fissile uranium plus 0.16 kg fissile plutonium per electric megawatt), but they also reduced the neutron losses to fission products and 2%3pa and captures in 238y to help compensate for the reduced breeding performance imposed by the presence of the 238y denaturant. In addition, the lower neutron flux associated with these changes would extend the life expectancy of the moderator graphite in the core to approximately that of the reactor plant, thereby obviating the need for periodic graphite re- placement. It would also substantially ease the graphite design constraints and allow for simpler geometric shapes. Although the neutronic calculations indicate that this reactor could operate indefinitely with the assumed chemical processing system, there is relatively little margin for error. However, a substantial margin could be provided by allowing the addition of small amounts of 235U (well within the denaturing limit) with the fertile 238U, and some additional margin probab1y7c0u1d be obtained by adjusting the nominal core design and/or the fuel processing cycle. Aside from the core nuclear cohcept, the other substantial change from the reference design MSBR ié in the area of chemical processing. The requirement for break-even breeding would imposera need for continuous chemical processing, but the cycle time apparently could be increased to 20 days (fromrlo days'for the MSBR). However, a more significant change would_be the elimination_of'the steps. to isolate 233p, in order to avoid the loss to waste of’plutonifim; Since plutonium, the transplutonium actinides, and fission product zirconium all follow the protactinium, this change not only would preserve the plufonium reduired for neutronic sur- vival, but also avoid chemical isolation and accessibility of proliferation- viii attractive materials. (An additional step would then have to be:provided in the process to:rémove zirconium on some reasonable time schedule.) The change'actua11y ATION | » - } e’ 371 @-moles Aoy ; . S : 1 . : 1 pr—— PATCH DISCARD FINAL { § : : - ouR - - . R WaTion | pe stonage sauy [7F =220 dde o, ! , RIS LEGEND ‘ : e ' ascmdrsec < f2 | : FOR ULTIMATE ' ———— FUEL OR FUEL SOLVENT 1 DISPOSAL . ——— - BISMUTH - P 1 | WASTE SMT : —e-e- UFg ) e e —————— e e e e e B e i o e e e e e e it ——————— - wemvaen LICE ) : S : L_ : g - "_' cpt— WASTE SALT ) . - e e e — ——— ——— e o o ——— . kY S = o o i S o 2 2 T o o — e e e ————— . Fig.2 Flowsheet for fuel proéessing plant in 'reférence. desi’gn MSBR. -} properly adjusted concentrations of litbium. Beryllium is not extracted; Zr, U, Pu, Pa, the rare earths, and Th are extractable in that order.®?® Such reductive extraction processes from fluoride fuel can effectively separate uranium from protactinium (but not from zirconium) and protactin- ium from the rare earths"and-thorium. Rare—earth fission products are par- tially extracted from molten fluoride mixtures by bismuth containing moderate concentrations of lithium, but they areraccompanied by an appre- ciable quantity of thorium. Separation of thorium from rare earths (and from Y, Ba, Sr, Cs, and Rb, which behave similarly) must be accomplished by transferring all these elements (except thorium) to molten LiCl from the bismuth-1ithium alloy 5,10,11 Uranium can be separated and recovered by reductive extraction, but fluorinatiOn.to UFs isymbre effective and convenient; The UFg and F; are absorbed in a sufficientzquantity of purified fuel solvent containing UFr.s’6 Uranium in thisrsolution is reduced to UF, with H;, and the re- constituted fuel salt is returned to the reactor after final cleanup and adjustment of the average uranium valence’to about 3.99; Bry, I» (and probably SeF¢ and TeFg), which are volatilized with the UF¢, pass through the sorber and must be removed from the off-gas stream. A small processing plant is sufficient, The reactor fuel passes through the plant every ten days with a processingfrate of 55 em®/s (0.87 gpm). Table 1 summariaes the removal methods and cycle times anticipated for such a plant.® The several separations required are well demonstrated in small—scale experiments, but engineering—scale demonstrations are still ,largely lacking, and materials to contain ‘both molten fluorides and bismuth ;alloys seem certain to pose some problems. fNonproliferation attributes Once placed in operation, the reference design MSBR would require no sthipments of fiesile material to the reactor and only occasional shipments " of bred excess 233U to fuel other reactors. Accordingly, 1t would present a very low, ‘and perhaps acceptable, profile toward diversion by subnational or terrorist groups. However, as far as weapons proliferation — a national decision to exploit the machine to produce nuclear weapons — such a reactor has pronounced and obvious weaknesses. The uranium within the fuel is Iable 1. Methods and cycle times for removal of fission products and " salt constituents in an MSBR processing plant Adapted from Ref. 5. - Group . Component thgzal Primary removal operation . Noble gases Kr, Xe 50 sec Sparging with inert gas in reactor fuel T _ circuit : _ Seminoble and noble Zn, Ga, Ge, As, Se, Nb, 2.4 hr Plating out on surfaces in reactor vessel_ metals Mo, Tec, Ru, Rh, Pd, Ag, and heat exchangers - ‘ t cd, In, Sn, Sb, Te : Uranium 233y, 234y, ZQSU; 236y, Volatilization in primary fluorinator' ’ 237y - returned to carrier salt and recycled ‘ to reactor Halogens Br, I. 10 days Volatilization in primary fluorinator ' - followed by accumulation in KOH solution _ in gas recycle system sy L Zirconium and 2r, 23%pa 10 days ,Reductive extraction into Bi-Li alloy f protactinium followed by hydrofluorination into o : ‘ , Pa decay salt ‘ Corrosion products Ni, Fe, Cr 10 days- Reductive extraction into Bi—Li alloym S ' ' - followed by hydrofluorination,tnto Pa , _ decay salt o - : _ ',Trivalent rare earthsb Y, La, Ce, Pr, Nd, Pm, 25 dayse _ Reductive extraction into Bi-Li alloy‘ ' ‘ Gd, .Tb, Dy, Ho, Er followed by metal transfer via LiCl , 7 : into Bi-5 at. % Li solution . ' Divalent rare earths Sm, Eu, Sr, Ba 25 daya® Reductive extraction into Bi-Li alloy and alkaline earths ’ followed by metal transfer via LiCl - ) into Bi—5 at. % Li solution o . Alkali metals Rb, Cs 10 days Reductive extraction into Bi-Li alloy L o o followed by accumulation in LiCl Carrier saltfu f Li, Be, Th 15 years Salt discarda ‘ ‘ Y is not a rare earth but behaves as the trivalent rare earths. . Effective removal time —-varies for the different elements. ot - 11 clearly usable material for weapons, and its removal in relatively pure form by fluorination could be accomplished with little difficultyffiy use of the available processing system. Of course, snch-an ac;ion would be an overt andlqbvioue_treaty violation (the reactor could no longer furnish power), but given suitable other preparations the fiwarning time" could be quite short. Less obvious (and probably more‘insidious) routes for _prpliferafion are,'in,principie, avaiiable. The reference MSBR produces more *3°U than it requires; this 233U_is_genereted;in‘quite;pure form in __.the'prptac;iniumraccumnlatipn system and is aveilable_via flnorination with the installed processing gear. ‘Attemprs,tp remove it secretly should be obvious“foen inspector, but succesefniirenovai;would beiundetectable for a:moderately}long_periodir:Itfiis probably eaey_to underestimate the dif- ficulties in such scenarios, ,The preeence_panppreciable;quanitities of ?E?U and of more than traces of fission preincts4will e&d to the difficul- .ties, but a well-planned. end determined effortrcould obviously surmount | them. As a consequence,_theureference'MSBR nould ‘seem more Vvulnerable than most reactor types to rapid results from an overt proliferation action and would offer significant opportunities for covert action. Reference Design Variations Because of the perceived proliferation sensitivity of the reference design MSBR, a briefastudy%g.waS‘undertaken’infithe fall of 1976 to deter- mine whether the basic concept could be modified to make it sufficiently - proliferation resistantafor;widefdeploymentras a power: producer. The re- vquirementwfor,a:positiveebreeding1gain?was,eliminated,;but;the high- : enrichment:fuel;cdmposition;was-retained:to completely avoid the need to deal with plutonium, Theflonlyuother_changefconsidered:in the reactor was a -lower power density (higher;fiSsile;epecific-inventory);to,reduce the -zzsignifieanCe_bfaneutron?absorptions:in;%3?Paf(if;Pa;isqletion_were;aban— 'é;doned);and;;afieliminategthe=needkfpr?periodic:replacement{of;moderator n,&graphite*in'the~reaetor*corer;=Five variants of the basic system, including the fission—product-cleanup concept, were considered.x.-f»r -:The first-wvariation. modified the ‘reactor performance capability and eliminated the breeding of excess fissile material. Such a system would have all the proliferation resistance (or sensitivity) of the reference 12 concept but would lack the potential for" continuous removal of fissile fuel while maintaining reactor operation. J o ' The second variation eliminated all fluorination steps'—-the most 7proliferation-senSitive’procedure in the entire fuelécleanup“process.‘“ This WOnld'preVent the isolation of 2%%Pa and would require more isotopi- caiiy'separated37Li; since uranium removal prior to fission-product clean- :up'would:belaccompiished by rednctionflwitfi‘lithium.‘ It”appeared’that‘fuel 'self-sufficiency could be maintained in such a system with a reduced ' reactor power density (to limit Pa losses and reduce the relative poison- ing effect of other fission products) and a significantly longet fuel ;'processing cycle. The longer processing cycle would also reduce the re- 'quirement for "Li. The elimination ‘of the fluorination steps was felt to "represent a significant increase in proliferation resistance. - The third variation involved a major change from the nominal fission— product-cleanup concept; it was proposed to substitute a CeF3 ion exchange system for all the chemical fission—product-cleanup operations. (Gas R 'stripping to remove xenon and other volatile fission products would be retained.) Such a system would remove only the rare éartfié*{by'sabst1; tuting Ce, which has a lower neutron cross. section) and leave a variety of other fission products in the salt. Some degradation in breeding per- formance would be experienced, but it appeared that self-sustaining opera- tion could be achieved at the lower: core power density. Since this process completely avoided separation of the fissile material, it appeared to be significantly more proliferation resistant :than the-reference“concept. However, the ‘technical feasibility of this approach has not been demon-- strated, and substantial research, development, and demonstration -(RD&D) would .be required to reduce it to practice'if.it is feasible.. . The use of some form‘of'vacuum:distillation for fuel cleanup was proposed as a possible fourth approach to enhance the proliferation’resis— tance of the reference reactor concept. Although such an approach would eliminate many of the proliferation-sensitive stepe, it was not clear 'that it would be workable with a salt containing ‘thorium. The technological -uncertainty of this approach tended to rate it relatively low among the - possible alternatives. ' ' ' ’ fiyiinventory of product. Moreover, ‘the MSR permits recovery of ‘the 13 .The final alternative considered was the elimination of all on-site cleanup processes other than physical removal of noble gases. The poten- tial feasibility of this approach was based on some earlier studies of high-performance converter MSRs in which the unprocessed fuel charge was - simply. replaced every few years. It appeared that, if reactivity varia- ~tions-could be managed, such a system might require replacement of: the fuel charge only two (or possibly three) times during the life of a reactor plant. Although such changes would require the application of additional safeguard measures, the infrequency of the changes was judged to make this approach reasonably acceptable. Although some of the proceSSing*modifications to the high-enrichment concept appeared to be clearly technically feasible and all provided some enhancement of the proliferation resistance of the reactor, it did not appear that the antiproliferation gains were of sufficient magnitude to ‘justify an extensive effort to develop the reactor and the associated fuel-cleanup system. Consequently, nondenatured MSRs for power generation at dispersed sites were not considered further. Plutonium Transmuter for >3y Production At may,be,that‘any high-enrichment MSR would have to be located at a site where special safeguards would be in effect and thus special-purpose MSRs might also be acceptable. Of particular interest in this regard would be MSR systems that consume plutonium and higher actinides (produced by other reactors) and producegga3Ugforgdenaturing and subsequent utilization at dispersed sites. o | e Thermal or near-thermal reactors (Whlch include MSRs) are inherently h"filess efficient” burners of plutonium than are fast reactors and are at ‘some ‘idisadvantage in'"fuel—factory applications. However, MSRs ‘have minimal para31tic absorbers in their cores need neither head-end reprocessing steps nor fuel element refabrication, and have a much smaller in—process 233U B product .as8’ soon as it is produced' hence, very little of ‘the" product-— whose greatest value is as an export commodity -is consumed by fission fbetween replacements of solid fuel elements as in ‘the fast reactor system. 14 “Thus, ‘any advantages MSRs might have as' "fuel-factories” would be.related “to-their fluid fueli® o oo e © "The net production capability would be a major,Vbut~notrthe only, “eriterion for7evaluating‘"fuel factory" options. ' Other ‘significant cri- “teria would include the technological feasibilitj ofrthe’concept,?indus— tria15acééptability,VcommercializatiOn*potential,?safety and reliability, licensability, time ‘to commercialization, and the probable net cost of the produCtQ*‘Molten—salt reactors have not:been seriously considered Vheretofore”as?safeguarded-producersflof\233U;>perhaps’they should be. 'DENATURED MSR . . MSR systems containing substantialmamountstof‘?agU have:not been considered,in_most prior studies because of the perceived difficulties ingdealingawith the plutonium that would be produced.f_lnnaddition, such . 1 gystems would not be compatible with the high breeding performance and low inventories that have been among the traditional system goals.. How- ever, with the current emphasis on proliferation resistance and ultimate ‘resource utilization in fission energy systems MSRs fueled with denatured uranium may have significant overall technical advantages. The denatured MSR" (DMSR) described in the following subsections is based on-& preliminary JCOnceptual“study of this';ystem.' It is anticipated that a more precise and detailed description will be evolved as the study is continued.’ General Characteristics - The principal characteristics desired in a DMSR are (1) that it meet "to the maximum extent practicable the currently perceived requirements for -‘resistance to proliferation of nuclear explosives and (2) that it provide for a very ‘high level of resource utilization._ ) . . | At equilibrium,+ the principal fissile material in the denatured system is uranium with 2330 and 235U in a ratio of about 10 1. Sufficient Their efficiency. as net energy producers would be an. advantage in comparison to accelerator—driven fuel producers. ‘ +Isotopic equilibrium for fissile uranium is effectively reached in -3 few years and is independent of whether startup was on 235y oy 233y, 15 238y 4g present in the mixture to dilute the 233U by 6:1 and the 2%°U by 4:1. - Additional denaturing is provided by the 23U and 23U in the steady- state mixture to achieve the-preferred'dilutions for nonproliferation. Although substantial plutonium is produced from the 238y, the high neutron cross sections of the plutonium isotopes and the fact that all plutonium -.1s retained in the fuel salt keep the total plutonium inventory relatively low; about- 10% of the fissile material is plutonium (239 and 241 isotopes). The long effective exposure time of the piutonium results in the buildup of substantial amounts of 2*%Pu and 2%2pu. Although these isotopes have significant fission cross sections (particularly at high neutron energies), - they also undergo spontaneous figsion (i.e., produce neutrons), which tends to detract from their value as explosives materials. In addition, there is no provision for the isolation of plutonium from a number of other radioactive and otherwise undesirable nuclides. One other potentially attractive material is ?aaPa,-which:is present to about 84 kg in the fuel salt at steady state. If this material could be isolated from the rest of the fuel, it would eventually produce high-purity 2°3U, which would be proliferation'sensitive; "HoueVer,xprotactinium'isolation is not part of the concebtuaIISystem and'modification-of the‘system to provide for such ”’”isolation would be difficult, time consuming, expenéive, and readily de- tectable. - " ' Utilization of all natural resources in the denatured system appears “to be”quite:favorable;”'Significant"amounts'of“7LiF‘(and'hence beryllium ~and’ thorium fludrides)* must be continuously removed from the fuel salt " as' 'Ll 1s’ added in the fission—product-cleanup system; however, these materiale could be’ recovered by a variety of aqueous processes if it were -economically attractive to do s0. ‘The effective breeding ratio can be maintained at 1.0, so that, after the initial fissile loading; no fissile “‘materfial need be added or removed for the life of the plant, howeVer, thor- “"jum and 23%U must be-added’ continuously to maintain ‘the concentrations of thése nuclides. At the end of plant life,'only a small’ amount of addi- " tional’ uranium would have ‘to be added to that recoverable from the old ' These materials must all be included as potentially limited natural - resources. 16 “plant (to substitute for plutonium that is not recovered) to start up ‘a new plant. Alternatively, the entire salt charge from a retired plant (including in-salt fission products,'plutonium, and higher actinides) icould be transferred to a new plant with no new fissile addition and no plutonium left over for storage or disposal., The basic reactor flowsheet for the DMSR is essentially the same as that for the reference design MSBR. The only differences are in the core :.configuration, details of the fuel-salt composition, and the fission- '-product-cleanup*(chemiCal.proceSSing)~system, Thus the primary-system temperatures, pressures, and major flow rates, as well aS=the;entire ~secondary system and balance of plant, would be the same as for the refer- -ence plant. The remainder of this section is devoted to those portions of the denatured concept that have not been described previously. Reactor Characteristics The principalicriterion for an attractive DMSR is survival in a neutronic sense. It is axiomatic that adding 238y to a thermal spectrum V‘MSR degrades its overall breeding performance because the plutonium that is produced has a lower effective fission neutron yield than 233U in such a system, In addition, it was recognized that protactinium isolation would not be acceptable and that neutron and hred 233y josses due to neutron captures in 233pa would have to be accommodated. Thus, the nuclear design problem became one of halancing a low core power density (to limit protactinium losses and graphite heating) against a higher fissile inven- ‘tory in a‘oore of reasonable size and balancing a more heterogeneous _ (lumped) core‘(to limit neutron absorptions in ZSBU) against potential N cooling problems in large moderator elements, | One of the first requirements established for the DMSR was the need _for break—even breeding. This requirement probably applies to all denatured fluid-fuel reactors and to any other systems in which the entire fuel charge has one homogeneous composition.* The actual "critical point" for o In such systems it is not possible to upgrade the average core en- richment by removing below-average (depleted) fuel and adding near-average (but still denatured) material. 17 operating feasibility occurs when the fractional rate of production of fissile isotopes equals that of consumption of fissile isotopes with ap- t propriate consideration of the rate of burnup of fertile material. At this point it becomes possible to sustain reactor operation indefinitely by additions of denatured fuel. For denatured feeds containing 137 233y in 238U and ZOArzasU in 238U the minimum acceptable MSR conversion ratios are 0 98 and 0. 97 respectively. However,‘such systems would be signifi- *:cantly less attractive than a true break-even reactor in that they would . require transport of substantial denatured fissile material to the site. A DMSR must have an effective fission-product-removal system and must use the plutonium produced from 23°U efficiently to achieve break—even fbreeding over its lifetime. The plutonium and protactinium, as well as uranium, must be removed from the fuel before rare-earth and other fission products ‘can be removed. Accumulation of 233Pa for deCay outside the | reactor (as was planned for the reference MSBR) could not be permitted for the DMSR since it would make high—quality 233U available with moderate ease. It is convenient to remove plutonium and protactinium together from the fuel and to reintroduce them immediately to purified fuel solvent for return to the reactor. Such retention of 233Pa in the reactor tends to " lower the tolerable neutron flux (and the power density) to limit losses of 233Pa by neutron capture.' This decreased power density increases the fissile specifie inventory for the system but also has some favorable effects. | 1. If the neutron flux must be reduced, it is reasonable to reduce - it to values that limit irradiation damage in the core graphite such that the graphite lifetime is equal to that of the reactor, thus eliminating ‘the need for scheduled moderator replacement. JV . ) A the lower neutron flux, the xenon poison fraction for a given ‘jxenon concentration s reduced thereby possibly eliminating the need to iI;imPregnate the graphite surfeces to reduce their permeability t° gases. 3, The attendant lower graphite power densities lead to lower tem- ="perature rises in the graphite, thereby substantially easing the design constraints on moderator elements. 4, The poison fraction associated with the shorter-lived fission products is somewhat reduced, providing slightly more margin for operation. 18 Core configuration | Consideration of the above factors led to the selection of a nominal . reference reactor concept with the characteristics described below.:__ y 1. A cylindrical reactor about 10 m in diameter by 10 m high in- - 'cluding the reflector. The core size is determined primarily by the “n)_neutron damage flux to the graphite With little influence (at these 1arge B sizes) from criticality or conversion ratio._ Hence, effective flux flat- tening in the core might allow selection of a smaller reactor size or a ; | longer graphite life with minimal reactivity penalty. | - 2. A nominal fuel fraction in the core zone of about 134 subject to optimization and minor spatial variations (axial and radial) for flux flattening. - -3, Absence of a high—fuel—fraction "blanket"'zone, comparable to ‘the 37A salt zone surrounding the core in the reference ~design MSBR. This 'zone was used to help limit neutron leakage in the original breeder con- cept. 4, Simple cylindrical design (25 cm OD) for the graphite moderator elements with relatively large—diameter (v5-cm) central fuel passages. Refinement of the design might lead to modification of these properties. This basic reactor design appears to meet the neutronic and thermal- hydraulic requirements of the system while providing latitude in several areas (core size, fuel fraction, and moderator-element size and shape) for adjustment of the system performance to cover uncertainties. | In addition to the above features, the reactor would include salt " inlet and outlet plenums (between the core and reflector) at the bottom and top of the core that would be characterized by high fuel—salt volume 'fractions. A similar, though smaller, salt zone would be present between the core and reflector in the radial direction to accommodate the differ- ential thermal expansion between the metal reactor vessel and the graphite moderator. (The reflector is attached to the vessel 80 that it moves outward as the vessel expands on heatup ) The effects of these zones are included in the conceptual design. 19 Neutronic properties Nuclear composition and the basic fuel cycle. The reference graphite and fuel characteristics and compositions are shown in Tables 2 and 3. The isoéopic:composit{on &f the actinide component of the fuel at equilib- rium depends on the refueling policy, the removal process, and the flux- averaged cross sections. The fuel circulation is rapid, so that fuel everywhere in the core can be assumed to have one composition. After startup, the basic refueling policy is to add thorium con- tinually in the amount required to hold the concentration constant and to add 23%U as required to satisfy the'"denaturing‘inequality;"'Nzgau-3 6N233U + 4N235U » where N refers to nuclear number density. The actual amounts fed at equilibrium, assuming a Q.75 capacity factor for a 1000- MW(e) plant, are 601 kg of thorium and 116 kg of uranium per year. Thorium Table 2. Reférence characteriétics'of fuel salt and moderator for a denatured MSR Characteristic : : Value Graphite moderator density, Mg/m? - 1.84 Fuel-salt density, Mg/m3 3.33 Salt volume in reactor vessel, m® -~ 80 Salt volume outside reactor vessel,‘ma ‘23 Core salt-volume fraction -~ - ~ - . 0,129 Table 3: Nominal chemical ~+ .composition of:fuel salt: . Material = ' . Molar percentage BeFp ... 160 Fission products o Traece T aX'refefs éd all aétinides. represents 84% of the total feed on either a molar or a weight basis, and - elither depleted or natural uranium could be used with only insignificant differences. (Pure 238U was assumed in these studies.)‘: | , oA fission—product-cleanup Pprocess much .like that. described for the reference design MSBR.(see,also_Iable_l)risfpresumed to operate contin-. uou81Y=t9_remove,mate?ials‘f??@-thesfuelpéaip- A 20-day processing cycle was assumed for the denaturedvsystem_(vs%lo.days;for*thehreference'MSBR), so that effective removal times from-Iable'l;are approninatelyfdoubled for those elements*_whose_remova1:is a;function_of theiprooessingfcyclegV_Other differences‘from‘the,reference;cyclerthatfarise}from,changes in the nominal e + reprocessing concept are: . 1. The ?%%Pa remains with-the fuel:salt. indefinitely rather than being isolated on the nominal 20-day processing cycle. - ! 2. The transplutonium actinides are recycled into the fuel salt. 3. Selenium and tellurium are’ removed with the halogens on the nominal 20-day processing cycle rather than plating out on metal surfaces on a very short cycle. L 4. Fission-product zirconium, because it requires a special separa- tion operation, is removed on a rather long (%300-day) time cycle. 5. The fuel carrier—salt replacement cycle 18 about 7.5 years. The breeding and burning of fissile fuel proceed approximately as shown in the nuclide charte (Figs. 3 to 5), which illustrate the Th-U U-Pu, and transplutonium chains in the EM?R,_respectivelx.__Although the actual branch fractions depend'on“tfie“flUX'level“as well as the energy distribution of flux, these simplified chains indicate the potential for a mixed-fuel breeder. The data shown on the figures indicate a total of - 2.36 neutrons absorbed and 2.51 neutrons produced for each thorium atom consumed in the 2%2Th chain, while the 2°%U chain has a "cost" of 3.20 neutrons and a yield of 2.88. From this, we .can see that a combined neutron yield gives a small'surplus‘to”accounthfor’nbnactinide losses. Halogens, corrosion products, trivalent and divalent rare earths, alkaline earths, and alkali metals. TThe modified reprocessing concept is described in more detail in a later section. : - ORNL-DWG 78—10031 238p,, | g 2.12d g 237N nyY o 238 g~ |6.75d nl 7 l~-t[ 914 nl 'Y 235 nr 7 236 nv 7 2 > > » —_—237 10.6% ' V=% Y U S . o R 27.0d " H eV 98.3% - v =2.42 - 7 16.66h C233p, MY o34 - “Pa 17% Pa g~ |22.2m. sy, 99.7% ;. Fig. 3. Simplified thorium—~uranium nuclide schematic. 1¢ 22 " ORNL-DWG 78-10033 = x_. ———=——= FROM #Z¢m = i . _ 28py e 00 29y, MY 20p, AN 37.7% N 62.3% n, 238Np u = 288 : f~ 12354 239~p p g~ | 23.5m v=275 Fig., 4. Simplified uranium?fiifitonium nuclide schematic. Since the feed material is 84% thorium,;the net neutron yieldeis 2.88 Y = 0.84 =% 4 (1 —-0 84) 3—53-= 1. 04 . 2 3 The branch fractionsafld*the-Th/chhein ratio are both sensitive to the neutron energy spectrum; ae discfiSsed later. The above equation shows that the effect of the 238U chain is an important loss of reactivity and that efficient use of the resultant smaller neutron yield 1is required. The overall effect of the higher transplutonium actinides is of par-~ ticular interest. .The DMSR .is unusual in that these nuclides are recycled indefinitely as an alternative to includiqg them with the waste stream. This reduces the long~term waste problem, but it can have a significant e "““"‘3%'9?"" TN U S S N f.n,Kaa,%' 3 e t | Q s | N O 3 » 5 ""'msd 65% Cm_'l.'_'.‘.. 2“5cm——>2“sc ORNL-DWG 78-10934 vz 33 nl f\ss% L DY oare, Y e DY 2e0n L . g 1oan 5 cere e g SATE VY ‘. | / ade Sy P ey N %% . . Dok v=294 ° v=45 -\ SO S & 311d 52% - 64m . 490 16% Zfléakm_zsoak 48% g~ 3.22h LY _asp Ny . 2851 262 > Cf , wa Ct a] 2.65y TO 2480"1("__- —.—. —_ —J Fig 3. Simplified transplutonium nuclide schematic. £c 4 g e T AT 24 effect on the neutron yield of the system. Data taken from a 200-year operation study show that each atom of 240py produced from 23%py is joined by 0.11 atom produced by o decay of 2““Cm. 1If the additional 2I+°Pu, 241py, and 2"2pu reactions are. ‘taken as-a- part of the transplutonium effect, we. can characterize: the total effect as follows: For eachlabsorption in Zh2py calculated,without the transplutonium chafn, 4.0'additional absorp- tions, 1.0 additional fissions, and 3523additiodal fission neutrons are born. The net result is a loss of O.Sgneutron fier "nornalfabsorption in 262p, : Lo [ ] - The fiasions in 2"50m ?“lpd? and 2“7Cm, in descendiné order, are the largest neutron contributors associated with the higher actinides. At the low power density of this system, the a decay of 24%cn leads to an impaired neutron yield compared to that at higher power densities._ Also, the 8~ decay of 2*!Pu becomes .a nontrivial loss of fissile material. Neutron absorption in %33Pa represents a significant loss of reac- tivity in this concept, since each atom would otherwiseidecay to a fissile 233y aton yielding 2.2 neutrons directly for each‘absorption. Each ab- sorption -in protactinium leads to another in 23"U before - ‘a. fissile material is finally produced. Higher power density would make this situation worse. The'nonfissioning capture in 235U is similarly unprofitable. A total of three additional-captures are*required to produce a fissile nuclide, 239py, Some of these chains would take many years to develop fully, for example, 238y would saturate with a time constant of approximately 30 years. Even so, the full equilibrium value would eventually be reached. Consideration of all these factors leads to the equilibrium fissile inventory of the reactor. The total inventory of 233y + 235U is thus 2.4 kg/MW(e), while the fissile plutonium* (229py + 2"lPu) inventory is 0.16 kg/MW(e), | Neutronics results. The concentration, absorption,“and fission data corresponding to the fully developed breeding chains in the DMSR are shown in Table 4. More than 98% of all fissions take place in 233U 235y, 239py, and 2*1Pu. The U/Pu fission split is 5 to 1, but the plutonium neutron The total plutonium inventory is about 0.37 kg/MW(e), S0 only about 437% is highly fissile. 25 Table 4. Nuclide concentrations and reaction rates in the DMSR after long full—power operation ' Nuelid Concentration® -~ Neutron - Fission clide -(x1029)‘ , absorption fraction 232y 3,211.0 1 0.32775 0.00248 233p, 2.12 0.00396 0.00001 233y | 54.7 0.32284 0.75133 234y C 24,0 - 0.03420 0.00043 2355 6.07 0.03403 0.07272 236y - 10.0 0.00610 0.00008 237p ) 2.01 0.00607 0.00005 238y5. 348.0 ' 0.06769 0.00119 239p, . 2.69 . 0.06723. 0.10896 240py, 1.63 0.02538 0.00006 21»1?1‘l : ‘ 1.26 _ 0.02435 0.04687 2“2Pu‘ o | 3.43 0.00635 0.00006 Transplutonium o 0.02605 - 0.01577 Total actinides 0.9520 1.00000 Fluorine 47,800 0.008 Lithium , . 22,400 - 0.007 Beryllium _ 5,010 0.001 Total fuel salt | | 0.968 Graphite = | | 92,270 0.020 Fission products 0.004 - Total , . - o : 0.992 Nuclei per cubic meter of salt or moderator. bAbSOrption per neutron born; leakage is 0.008. Includes some 2“9%Pu, - 2%1py, and 2*2Pu produced from o . decay of | l”’Cm.-_ Lk ' ' yield per fission is significantly higher.i Neutron leakage is only small loss in this system, and captures in nonactinide nuclides are also f_loy The neutron utilization can be summarized as follows “:Absorbér*tjge*f> L “-*Absorgtioni(zzgfir' Actinides v Lo 005020 Nonactinide salt nuclides 1.6 Fission products \ 0.4 Graphite =~ o 2.0 Leakage 0.8 26 The depression of thermal flux in the fuel is of some interest because it governs the allowable size of the moderator 1ogs from a neutronics standpoint. . If the flux depression is large, graphite and resonance cap- ture will be enhanced._ Table 5 shows that flux depression would not be excessive in the reference core design. _ TableiSQ Fuel disadvantage factors “Neutron energy . Inner fuel Moderator Outer fuel - group o zone - zone pl-(fast) 118 . 0.98 1.11 . 2 (resonance). = ~ 1.00 . - 1.00 : 1.00 The spatial peaking factors for both power density and fast-neutron flux (E > 50 keV) have gignificant effects on moderator graphite lifetime in MSRs, particularly in the low-power—density concepts ‘where a moderator lifetime equal to that of the reactor system is desirable.’ The power- density distribution primarily affects the graphite temperature, which in turn affects the amount of graphite damage for a given neutron fluence; the neutron flux directly affects the carbon-atom displacement rate as well as the temperature. The peak-to-average values for:both power density and neutron flux are the same in the nominal core design, both in the radial and axial directions; the values are 1.69 and 1.35 for radial and axial directions, respectively. The core average neutron damage flux is 3.1 x 10?3 neutrons/cm -sec (E > 50 keV). If a fast fluence of 3 x 1022 neutrons/cm2 is agsumed as the 1imit of useful msderator life, this value | would be reached in the highest-flux region in 13 equivalent full—power years (17.3 years at 75% capacity factor) Less conservatism in defining the upper limit for damage fluence and flux flattening may allow an exten-— sion of the useful graphite life to the desired 30 years at 75% capacity factor. ‘ ' — Startup'and control. The startup of;the‘denatured;system can be accomplished with either 233U or 235U at the appropriate‘denaturing level. 27 The effect of the denaturing is such thatreithgr;fuel_will give approxi- -mately the same perfo?fiénée. .The initial.?eéc?i?ityflis’very senéitive to ~the initial fissile loading, as shown by Fig. 6. The calculated f;ssile ~loading required to achieve initial criticality_and_qve;come_equilib:ium fission product loading ie 2371 and 3115 kg for %0 and ?%°U, respectively. Figure_6,also‘shows,that_a_2%rerror_in the griticality calculations could _;be,cqmpensatedfby.a_SZ;éhange.in fissile loading. - After startup,. an increase in .reactivity on the order of‘Z.SZ will occur due to the greater effect of buildup of new fissile material over .. -that of fission products. Short-term reduction of fgactivity could be . acqqmplished}by-withholdifig uranium from_the_inpufi étream, A short-term ~ increase could;be,accomplighed‘by redficing;hefithorium.cqntent,although the long-term effect of this action might bé legs_fissile production. Thus, reactivity increases would more likely bg:providéd by smali fissile additions. B | | ORNL - DWG 78-10932 0.100 [ W 0075k 0050 .. & REACTIVITY CHANGE . ' o ECEE | 20 - 30 40 50 TLwt T FRACTIONAL CHANGE IN'LOADING OR -+ 7+ o ... . MAKEUP ENRICHMENT (%) ' - Fig. 6. Effect of initial loading and enriched makeup on equilibrium reactivity. | - | 28 " Long-term breeding and nuclear design flexibility. ' The meutronic :jcalculations indicate that the DMSR would start and run for ‘the life of the moderator on fuel which 1t manufactured’ internally. However, more is ‘ex- 5pected of it. In this scenario, it ig intended that the fuel be- recycled ‘°”Findefinitely in a succession of new reactors as the useful life of the old ;fones ends. This would eventually lead ‘to a buildup of ‘the "trash" nuclides (236U 237Np, 238py 2%2py and the various americium and ‘curium nuclides). - of these, the data used for 2%%Pu and the various ‘americium and curium nuclides must be described a8 estimates and are perhaps subject to errors of 30/ or more. If these chains develop as predicted ‘the ‘ultimate ’effect'would be a slow'approach“after‘msny‘yearS“to an'absorption fraction. of 0. 0633 (including plutonium ‘&nd’ transplutonium ‘effects) due to ‘absorp- tion in 23®pu, with a yield of 0.0377 for a net loss of 0.026. Present "Ecalculations indicate ‘that a system using “this fuel ‘would be no more than barely critical if the calculations were accurate. S | What can be done if these predictions are true? What if reactivity is even lower than predicted? Potential alternatives for increasing the overall system reactivity include (1) altering the spectrum to improve neutron production, (2) enriching the 23°U added, (3) altering the fuel salt processing concept, or (4) adjusting the denaturing limit to reduce the 23%¢ additions somewhat. The poténtial for improvement by spectrum modification seems attractive. Certainlyfithe fission neutron yield is sensitive to_the energy spectrum. To illustrate this point, Table 6 shows a three—énergj?group structure used in soue:of the analyses. Absorptions in groups 1 and 2 show a net loss of neutrons, while there is a gain in group 3. Thus the neutron yield is sensitive to the ratio of group 2 (resonance) absorption to group 3 (thermal) absorption and thus to the fuel/moderator ratio. The relative importance of group 1 (£ast) absorp- tions is small because both absorptions and productions are much smaller than for the other two energy: groups. Some additional information on the spectrum effect may be - obtained - by- intercomparing the group-average neutron ‘absorption cross sections and the effective neutron yields for some of the heavy-metal nuclides in this three—group structure (Table 7). For example, it is clear from a comparison of the Th/zsavecross-section ratios in the resonance and thermal groups that the ratio of Th/%3%U neutron:ahsorptions “th 29 - Table 6. Three-group-neutron structure and reaction rates for the postulated DMSR 'Ener"; ' Flux Relative Fission Net -Group rangy - volzme' . neutron- neutron fission o g_ , absorption production source 1 14.9-1.00 MeV 21 0.008 0.005 0.69 2 - .1.000.55 eV 223 - 0.378 0.197 0.31 _3: 0.55—0.005 eV 145 0,606 | 0.798 0.00 Total 389 0.992 1.000 1.00 Table 7. Selected cross-section data for fissile/fertile nuclides Group 232y 233U | 235y 2aafi 239p, 24lp, 233p, 1 (fast)? - T Ua - 0.19 2-2 ’ 105 0.51 o 2.1 2-0 0-81 nb 1.2 2.6 2.6 2.4 3.2 3.1 2.1 2 (resonance) | | | ; Oa C 1.6 51 25 5.8 28 39 53 n 0.00 = 2.1 1.6 0.00 1.7 2.4 0.00 3 (thermal) PP - . ' Og 3.0 250 274 1.2 1400 1000 16 0.00 1.8 2.2 0.00 n 0.00 ‘2,3 2.0 See also Table 6, ; | Defined here as vc /g .flf"‘* },would be increased by reducing the resonance flux in relation to the ther- ‘mal flux.; The _same would. be true of the ratio of ?33U/Th absorptions.‘ In this system, almost -every neutron absorption in thorium also results in a neutron absorption in 233U, thus, ‘an . increase in the absorption effective- nessigfm%??U with.reducedxresonanceafluxlleads‘to_a\lower allowable inj ventorybof§%§3UAIelative;toLtfiOrium.;fiSince the required;%saU'inVentOrY is governedfprincipally byfthe'emount-of 2331‘I_present, this also leads to a lower 23-aU_loaLd:I.ng._ Both of these effects work to increase the relative 30 - importance of the more . productive Th-233y chain (as measured by the higher weighted—average value of n for 233U in this spectrum). All these factors tend to make the neutron yield larger when resonance flux is reduced by increased moderation. Acting contrary to this trend is the tendency of ‘the resonances in 238U to capture more as the concentration is reduced. This effect does not dominate, however. A thermal spectrum results in more- absorption in nonactinide salt nuclides and graphite. Also, it is necessary to increase the moderator volume fraction to make the resonance flux lower. These effects result 1n more parasitic absorp- tions, which tend to offset the beneficial effects of the more-thermal spectrum, In the reference MSBR, a "blanket" with a relatively high salt frac- tion and a harder spectrum was used ‘around the optimum-spectrum inner core. This“tended to increase reactivity,” with the fissile material being pro- duced—in-a;hard spectrum.with.lowmparasitic capture and consumed‘inva softer central spectrum. Although the resulting core would‘be more com- _plicated (and difficult to manufacture), the alternative might be accep- table if one were required to provide the added reactivity. o Figure 6 shows the effect of enriching-the 238U makeup. lhehamount of ?3°U added would remain’as before, but some 233y would be added. If the material were enriched to the nominal denaturing limit, V1% of reac- tivity could be gained. A 50% enrichment:vould yield 7% of reactivity. This would require special protection of the material added but the amount would be only 155 kg of fissile material per year. Enriched 235U could also be used with somewhat inferior results. Since fission products constitute only a very small reactivity loss in this concept (cf. Table 4), the reactivity gain that could be realized o by modifying the fission—product-cleanup process probably is insignificant. | However, in the equilibrium fuel mixture, there is significant" poisoning associated with neptunium, plutonium, and the transplutonium ‘actinides. o Thus removal ‘of some of these materials,' possibly between movements of . the salt'from one*reactOr plant to another, could effect a significant . extension in the useful life of the fuel charge. (In the 1limit, the " "'entire fuel charge could be consigned to storage or ‘disposal at the end 'of life of a given reactor.) It seems apparent, however, that this 31 approach would have an unfavorable effect on the antiproliferation at- tributes of the concept. . _u‘ d ~The final option — reducing the denaturing ratio — may be inferior to the other three from an antiproliferation viewpoint, although it would not add to the fuel cycle cost as would enriching the feed material. Allowing the 233y denaturing factor to drop to 4 as for ??SU would produce a O.7Z increase in reactivity. . Further reductions in the ZS?U loading Would also improve the reactivity but would decrease the proliferation resistance of the system. . 4 L In summary, it appears that an attractive, proliferation—resistant DMSR with break-even breeding is neutronically feasible and that suffi- ~cient latitude and alternatives exist to ensure its technological success in this area. Core Thermal Hydraulics The reactor core thermal-hydraulic features, particularly with respect to graphite temperature and xenon transport to the graphite, were major considerations in the reference design MSBR. Although the design con- straints are considerably relaxed in this area for the DMSR they remain '31gnificant from the standpoint of overall technological feasibility of the concept. Because of the relatively low power density of this reactor concept, simple core configurations which were not possihle in the MSBR reference design may. be considered Three simple designs were considered: (1) a core made up of Spaced graphite slabs, (2) a core made up of stacked hex- agonal graphite blocks- with circular coolant channels, and (3) a core con- sisting of a hexagonal array of graphite cylinders With central coolant channels. ", o .“,. Constraints which must be considered in selecting a core design in- clude maximum graphite element temperature, local ‘salt volume fraction, and the 238U self—shielding effect, which imposes a minimum limitation on the coolant channel dimensions. The temperature rise between the coolant channel and the hot spot im- the graphite moderator element is especially important because of the strong dependence of graphite dimensional change 32 with temperature: The salt volume fraction and the 23°U self-shielding effect strongly couple the thermal-hydraulic and thefnéutronic'core designs. ' These combined constraints appear to rule out the possibility of a " graphite slab core configuratiom. Mechanical problems, especially the " loss of coolant chamnnel geometry due to shifting'of'the}stacked hexagonal blocks, rule out the second option. The third design seems to fill all the ”Erequiremente'an& is also very appealing because of its structural simpli- " éity — which ia:importantjin'a core expected to last the 1life of the plant. The outer diameter of the cylindrical graphite elements would be V25 cm and the diameter of the inner coolant chanmel %5 cm. This yields a salt volume fraction of 13% and equal coré salt temperature increases of 140°C in the central and outer coolant channels, Figure 7 shows the basic core geometry and the two types of salt flow channels.(the central and the outer channels) which are formed between the moderator elements. The 30° annular section of moderator element used in the thermal analysis is also shown in /Fig._7. If the heat transfer from the surfaces of this element were uni- form and characterized by a Dittus—Boelter correlation film heat transfer coefficient the maximum temperature rise in the moderator at the center of the reference core would be %60 C. The heat transfer, however, is B obviously not uniform to the outer channel because (1) the salt (which wets " ORNL-DWG 78-10035 FUEL SALT . MODEL SECTION FOR THERMAL ANALYSIS GRAPHITE Figt.7. Reference corewconfiguration for denatured;MSR.. -t - 33 graphite poorly) will not penetrate all the way to the point of contact of the moderatorgelements, (2) the salt velocity near the point of contact will be greatly.diminished, and (3)_regions of{low salt velocity.willrhave temperatures greater than the channel average because >90% of the power;is generated in the flowing salt. In addition, the Dittus-Boelter correlation may not apply, because a thin film of helium may exist on the graphite surfaces. N , 7 4 - In the absence of information on salt heat transfer coefficients, penetration depths, and turbulent velocity profiles, an estimate (probably conservative) of the moderator temperature structure was obtained assuming a salt film heat transfer coefficient of 0 within 15° of the point of moderator contact and a salt film heat transfer coefficient equal to 80% of the value obtained using the Dittus-Boelter correlation elsewhere. ‘With these boundary conditions, the heat:conduction.equation in cylindrical finite-difference form was solved in the 30° graphite section using the method of successive over-relaxation. ‘Constant heat generation and thermal conductivity within the graphite were assumed. . This analysis yielded a maximum graphite temperature 80°C above the salt temperature at the core center and a maximum graphite temperature in the core of 740°C at an axial location 2.1 mfldownstream,of the core midplane. The hydraulic diameters of the central and outer channels are 5 and 2.6 cm, respectively, which means the central,channels will need to be orificed to more nearly equalize the salt velocities and hence the salt ntemperature rises in two channels. This could possibly be achieved by Vmachining small channels in the graphite near the inlet and outlet ends. The possibility of spacing the moderator elements to eliminate the .prob- T_lems caused by low heat transfer and low salt velocity near. the contact - points has been investigated, but at present it ‘appears this would entail ;?a salt volume fraction significantly greater than 137% to be effective. -;Chemicaerrocessing" '_,_ Unit processes and operations generally similar to those in the flow- 'sheet for the reference MSBR can be used to process fuel from the DMSR. Processing for the latter reactor has not yet_been analyzed in detail, 34 but it is clear that the flowsheets must differ in some important aspects. The fuel volume in the DMSR must be considerably larger, and, although the cycle time can probably ‘be appreciably greater than 10 days, the processing ‘plant will be somewhat larger than that of the MSBR. 'The DMSR will contain “a considerable quantity of plutonium which must be retained within the reactor circult. The MSBR system in which ?2%Pa was accumulated outside the reactor core and allowed to decay must obviously be abandoned ‘since such a system would furnish weapons—usable 233y 4 upon treatment with Fa. Since’ protactinium and plutonium, along with uranium, must bé removed from " the fuel solvent before;yttrium'and the rare-earth fission products can be removed the DMSR must contain a system which provides for removal of plutonium and protactinium and minimizes proliferation opportunities by immediately reintroducing them to purified fuel ‘solvent’ for return to the reactor. Such a protactinium-plutonium reintroduction circuit has the " considerable disadvantage cbmpered‘with the MSR plutonium accumulation ° system that it also reintroduces fission product zirconium to the puri- fied fuel solvent. However, the protactinium-plutonium reintroduction cir- ‘cuit has the advantage — insofar as waste management is concerned — that ' . it also reintroduces americium, curium, californium, and plutonium to the reactor fuel and permits only very small losses of any transuranium elements to the waste streams.* It seems apparent that the DMSR'can_manage the noble-gas and the semi- noble and noble-metal fission products in the manner and with the same removal times described earlier (see Table 1) for the MSBR. Operation of the DMSR with 5 to 10% of the uranium present as UFj, as‘seems-feasible,_ would apparently result in essentially immediate reduction of fission =2 and their complete reten- product selenium and tellurium to Se~2 and Te™ “tion (with little or no interaction with the Hastelloy N) by the fuel. Any other seminoble and noble-metal fission products:that appear appreci- ably in the fuel stream to the processing plant could befeffectively re- moved by a simple wash with bismuth containing no reducing agent. W course, it is not known whether solid LiF-BeF,-ThFy containing fission products can be considered a suitable form for disposable waste. * It does seem certain that very low levels of transuranium nuclides will offer some advantages whatever the waste form. o - 35 The DMSR processing flowsheet, shown as a simplified block diagram in Fig. 8, would recover about 997 of the uranium by fluorination to UFg and would reintroduce it to purified'fuel solvent as proposed for the MSBR. The quantity of UF¢ to be produced and absorbed per unit time would be several—fold larger than'that fordtheEMSBR. Also, if the DMSR were operated -with 10% of the uranium as UFj3, the quantities of Ser and TeFe¢ to be recovered by the off-gas treatment system would beimarkedly increased. Fission product zirconium is produced in high yield, and its removal from the fuel is highly:- desirable. Although the zirconium isotopes are ‘not important neutron sbsorbers, any contained zirconium must be reduced with expensive L1 and reoxidized each time the fuel is processed. It should be possible to remove zirconium (on a cycle time of about 200 days) by partial extraction —-along with a portion of the uranium, plu- tonium, protactinium; and transuraniumgelements —ain bismuth containing a small concentration of‘lithiumffollowedfby selective and essentially complete reoxidation of plutonium; protaetinium, and the transuranium elements into purified:fuel solvent in a muitistage operation.* The pregnant solvent from this operation serves as the absorber solution for the UFs. Sinee the zirconiunrbearing bismuth solution cannot be completely freed from the 238p.233%y mixture by selective oxidation, the zirconium and uranium must be transferred by hydrofluorination to a ' waste fluoride salt and the uraniumtrecovered‘as UFs by fluorination before discard of the waste salt at a rate:corresponding to about 4 moles of zirconium per day. A simple method for zirconium removal on a much shorter cycle time'would be very desirable and may be possible.+ 1,;7‘; This reoxidation of plutonium and protactinium must be essentially quantitative since any of these elements (and the other transuranium ele- ments) that remain. with the zirconium are’ consigned with the zirconium to waste, - . L o i TZirconium is known to form a very stable intermetallic compound (Zr- | 'Pts) with platinum.13 This ‘¢ompound should form when fuel containing 107 of the uranium as UF; 1is, exposed to platinum, and the ZrPt; can be decom- posed to dissolved ZrFs. and solid platinum upon hydrofluorination in the - presence of molten fluorides. It appears that. neither wuranium nor thorium would be removed with zirconium from the fuel mixture by platinum,'® but there is no information about protactinium, ‘plutonium, or other trans- uranium elements. —tm- YFE ; enppomgepe LICH - e WASTE SALT Figl Z 8. denatured MSR. Preliminary flowsheet for fuel reprocessing plant in a ¥ o A _ o ORNL DWG 78-576 : C P, e i e Ce AND R |, _ iy - ‘ e—=- - == WF 70 ACCUMULATOR [~+~*3 200 L1 o T T » - . a . —— 1 - - REVE Je——— g ——upy : R ' N B 1 b . _ , - , unuu'non ‘ 1 N vALENCE] | ' - ace ki 1 . . " UsST- W' UFfs 1 : 1 t . . ADJ . REOUCTION AB!ORPTION , ! ——t L ! ! i MENT Hp | - ] P : _l | i o T T T ) ADD Li | . e e e e e ' REMOVAL| Lma i }‘ ‘ | RESE . ' = ' - . i ucu C|STRIPPER L., ! | ; : W . *- . - —————y : T . > R ; t " 1o . mmm e e Xy = T d 4 ; R | 1 -F s 1l T | ;- ‘ =X - 2 T mmsrtn ! Lo 1 PARTITION EARTH | e 4 ' b , EXTRACTION ' i | | O Jeta¥asomeiogin s v i _ L—-_.—'——-f_ ‘ L |- *_._-—_... + 1 : - ! : T &- : — - : T S ». o ) : % 1 Pa 1 | ¥ + e ' 1 g | XTREE EXTRACTION] A : | SR | wees 0T vorme-l 11 1 i J i EXTRACTION _xffliflon { - i SOLVENT 4 . > STRIPPER : | FLAl!r?on:‘N- L . ) - : W e e 3 —d =T ! T e e T 1 L. FROM - 4. . ———————e - mrmem— e T ‘ REACTOR | ' L~ I I [ pmmany | .y o iK--z | e i |Fwommaton] F2 | Secomanv e+ pvoro— & | PFPOQNDEWQSEE AT e | b C FLUORIN= | ety |Taron L1 ULTIMATE DISPOSAL ‘ fa,| ATOR | HFMp | | ———t . ! . - - -2 | e e - R o F 1 N " v : - f’ M Y. -‘—.— - - o o v e o e e San Sk Sl D SN EPe WAL L e SN SR MW S R PR RN e e ke g LEGEND ? ; oo _ > -‘-.—rufl.onmnsowenr —h * - R = —m———— ———— SMUTH : A i ot v e 1 o e e i —— i - -_— - - e e, ———— 9€ 37 If the partial reductive extraction of zirconium were used, the fuel salt would then pass‘topa_multistage extractor;where»the balance of the zirconium, uranium, plutonium, protactinium, and transuranium elements would be recovered by extraction into a bismuth~lithium alloy at somewhat higher lithium concentration. By use of about six countercurrent stages with lithium in bismuth maintained at about_Z,lSAX-lO’? atom fractionm, the protactinium losses can be kept to_completely,negligible values and the -plutonium losses can be made very satisfactorily low.* The pregnant bismuth.(containing U, Zxr, Pu, Pa, etc.) would be sent to the UFs5 re- duction and final valence adjustment stages, wheré the values would be. recovered in the fuel for return to the reactor. The fuel solvent (LiF-BeF2~ThF, containing a very large fraction of yttrium, the rare- earth, alkaline-earth, and alkali-metal fission products) from this extractor passee to the rare-earth extraction column.: - ‘The process for removing yttrium and rare-earth, alkaline-earth, and alkali-metal fission products from the fuel solvent is the same as that proposed. and described above for the MSBR. dThe,effective removal rates of the several fission products depend upon the. element removed and upon the size, flow rates, and number of effective stages in the rare-earth extraction, - transfer, and stripping systems. However, it appears that by processing 5% of reactor inventory per day (a number that may prove . .uneconomically-large), the rare earths and barium could be removed on a cycle time well below 100 days. Such removal would require discarding about 100 moles of lithium per day through hydrofluorination of the rare earths into waste salt. Cesium could be removed with a cycle time of 100 days by discarding about 100 moles of LiCl per day. ) | ~Since: the quantities of uranium, plutonium, zirconium, and trans- *uranium elements that must be reduced and reoxidized are much larger than in the MSBR, ‘the use of lithium by the DMSR will be relatively large. On ‘La 20-day processing cycle,rabout 2000 moles would be required as reductant ';; each day (with most of this entering the fuel) This corresponds to- about - *It appears that protactinium losses could be kept to less than 25 ,g/year and plutonium losses to about 100 g/year in the combined zirconium- removal system and the main extractor. . . 38 0.05'm® (1.8 £ft?) of purified fuel salt that must be removed each day.* About 280 moles of ThFy, 7 moles of 23°UF,, and 430 moles of BeF, must be added each day.' These removals and additions constitute replacement of Jthe ‘fuel solvent (LiF-Ber-Tth) once per 7. 5 full-power years of opera— tion: o o | L R " Removal of radidactiVe'speciesifrbmithe several exit gas streams could presumably be accomplished ‘in ‘the manner proposed — though not yet developed in detail — for the feference\design'MSBR.--Krypton'and xenon isotopes, along with small: quantities of salt, radioactive particulates, .and€traces of radioiodines, must be removed and_:ecovered,as'wastessfrom the reactorisbsrgingacireuit‘:'Tritium”must'be recovered from the second- ary coolant. Insofar as practicable, the several streams containing HF- .and Hz would be combined for recovery of the HF for recycle through the. system for generation of F2. It is clear that essentially-complete 'recovery of radioiodine and radioselen1Um and tellurium from the gases passing the UF¢ absorption system will-prove"to-be'difficult.ff* - fAil in all it is certain”that; even if all the gystems indicated = above prove feasible, a great deal of development isirequired=5efdre the I fuel processing plant could be designed in detail, ~Indeed as indicated in a subsequent section of this document,: design of the processing plant ’ will be further complicated by the paucity of materials of construction | that are adequately stable toward both molten fluorides and molten bismuth alloys. *This salt contains essentially the proper quantity of LiF, BeF,, and ThFy along with some rare-earth, alkaline-earth, and alkali—metal fission products and virtually no fissionable or transuranium isotopes. It con- - siderably exceeds the quantity needed for the hydrofluorination of waste - materials. It may be that an appreciable fraction of this could be stored “and used for startup of additional DMSRs. ‘Alternatively, and especially 1f solidifed' fluoride cannot be considered an adequate disposable waste, it may prove economical to recover at least the ’Li from the salt during its conversion to suitable waste. +The very short cooling time for this fuel will, of course, intensify ' the iodine retention problem though the abséence of complications from. or- ganic solvent—iodine interactions should be of some benefit. - ' .. to'the structural alloy question 39 Balance of Plant ' As indicated earlier, the purpose of this study is to examine the features of a DMSR that would differ significantly from those of the reference'design MSBR. Since the fuel salt for the denatured system would have essentially the same thermal-hydraulic properties as the MSBR fuel salt, there is little if any basis for considering changes to the reference system other than those described above for the reactor itself and the the fission-product-cleanup system. Hence, the remainder of the primary- coolant (fuel) circuit the entire secondary circuit including the second- ary salt the steam system, and the plant auxiliaries would be essentially as described for the MSBR. One possible exception to this is the shutdown cooling system and relatedrequipment, which might be simpler for the de- natured reactor because of the lower fuel power density. Other differences might appear if a detailed design were developed ‘but the reasons for such . changes would involve engineering judgment, safety analysis, and/or eco- nomic choices rather than: basic differences in the reactor concepts. As a consequence, most of the design study work that was directed toward the MSBR balance of plant could be applied to a denatured aystem. ' MSR TECHNOLOGY STATUS A comprehensive review 7 of the status of molten-salt—reactor tech~ nology was published by ORNL in August 1972 This document was comple— ‘mented by an AEC review of ‘the statusls which also identified a number of technical issues needing solutions before an MSR could be successfully “built and 1icensed., When the technology development effort was resumed in _ 1974, work was directed toward several of these issues, including the pri- mary-system structural alloy,;chemical processing technology, and tritium 'management.; Significant Progress. was made in these areas with 1aboratory demonstration of the requirements for an apparently satisfactory solution 17518 and an engineering—scale demonstration of tritium containment ‘4n the seCondary salt. Design and construction of engineering—scale tests of several parts of the chemical processing con- cept were under way when the program was discontinued in 1976. The nature 40 of the technical progress théffiwés fiéde;}in:éonjunction with the less- , stringent.requirements of the low-power-density denatured system, suggest that such a system could eventually be sucqeéefuliy.defieloped._chqver, ‘_Vsubstantial time .and effort would be required to develop the MSR into a licensable, commercially acceptable system. _ REFERENCES ... 1. R. C. Robertson, Ed Conceptual Destgn Study of'a Stngle-FZutd " - Mblten~SaZt Breeder Reaator, ORNL-4541 (June 1971).' . 2.7 P. N. Haubenreich and J. R. Engel '"Experience with the Mblten-Salt o 'Reactor Experiment," Nuel. Appl.:Tech. 8(2), 118" (February 1970). | . 73;f Ebasco Services Inc., 1000 Milte) Molten-Salt Breeder Reactor Con- o ceptual Deszgn Study, Final Report, Task 1 (February 1972) . | 4; L. E. McNeese and Staff, PTOgram Plan fbr Development of'MbZten-Sth ’ Breeder Reactors, ORNL-5018 (December 1974). RN 5. Molten-Salt Reactor Program, Sbmmannual PTOQTQSS Rbport fbr Perzod Ending August 31, 1971, ORNL-4728, pp. 178-83. 6. L. E. McNeese, "Fuel Processing," Chap. 11, pp. 33163 of The De- velopment Status of Molten-Salt Breeder Reactors, ORNL-4812 (August 1972). _ 7. G. T. Mays et al., Distribution and Behavior of Tritium in the Cool- ant-Salt Technology Fuctltty, 0RNL/TM~5759 (April 1977) 8. W. R. Grimes, "Molten-Salt Reactor Chemistry,” Nucl Appl Tech 8 137 (1970) | _ 9, L. M, Ferris et al.,r"Equilibrium Distribution of Actinide and Lan- _ thanide Elements between Molten Fluoride Salts and Liquid Bismuth ~ -Solutions," J. Inorg. Nuel. Chem. 32, 2019 (1970) w 10. ”F J. Smith and L. M. Ferris, Mblten—sult Redctor Program, Semi~ ' annual Progress Report for Period Ending Fébruary 28, 1969, ORNL- : 4396, p. 285. - .. _ . , e ! 11, L. E. McNeese, Engineering Development Studies for Molten-Salt Breeder ... Reactor Processing, No. 5, ORNL/TM-3140 (October 1971). 12, H. F. Bauman et al., Molten-Salt Reactor Concepts with Reduced Po- tential for Proliferation of Special Nuclear Mhtertals, ORAU/LEA (M) 77-13 (February 1977). 13. 14. 15. 16. 17. 18. 41 L. Brewer, Science 161, 115 (1968). D. M. Moulton et al., Molten-Salt Reactor Program, Semiannual Pro- ress Report for Period Ending August 31, 1969, ORNL-4449, p. 151. MSR Staff, The Development Status of Molten-Salt Breeder Reactors, ORNL~-4812 (August 1972). An Evaluation of the Molten-Salt Breeder Reactor, prepared for the Federal Council on Science and Technology R&D Goals Study by the U.S. Atomic Energy Commission, Division of Reactor Development and Tech- nology, WASH-1222 (September 1972). H. E. McCoy, Jr., Status of Materials Development for Molten-Salt Reactors, ORNL/TM-5920 (January 1978). . J. R. Keiser, Status of Tellurium—Hastelloy N Studies in Molten Fluoride Salts, ORNL/TM-6002 (October 1977). " had ] 37. 38. 39. 40. 70. 71. 72. 73. 74, 75. 76. 77. 78. 79, 80. 43 ORNL/TM-6413 ‘Dist. Category_UC-76 Internal Distribution T. D. Anderson ‘ 41. R. S. Lowrie Seymour Baron 42. R. E. MacPherson D. E. Bartine 43, H. E. McCoy H. F. Bauman 44, L. E. McNeese H. W. Bertini 45. H. Postma E. S. Bettis 46-50. W. A. Rhoades H. I. Bowers 51. P. S. Rohwer J. C. Cleveland 52. M. W. Rosenthal T. E. Cole 53. Dunlap Scott S. Cantor 54, M. R. Sheldon J. F. Dearing 55. R. L. Shoup J. R. Engel 56. M. J. Skinner D. E. Ferguson 57. 1. Spiewak M. H. Fontana 58. H. E. Trammell A. J. Frankel 59. D. B. Trauger W. R. Grimes 60. J. R. Weir R. H. Guymon 61. J. E. Vath W. O. Harms 62. R. G. Wymer J. F. Harvey 63-64. Central Research Library H. W. Hoffman 65. Document Reference Section J. D. Jenkins 66-68, Laboratory Records . P. R. Kasten 69. Laboratory Records (RC) Milton Levenson External Distribution Director, Office of Fuel Cycle Evaluation, Department of Energy, Washington, D.C. 20545 E. G. Delaney, Office of Fuel Cycle Evaluation Department of Energy, Washington, D.C. - 20545 C. Sege, Office of Fuel Cycle Evaluation, Department of Energy, Washington, D.C. 20545 S. Strauch, Office of Fuel Cycle Evaluation, Department of Energy, Washington, D.C. 20545 K. A. Trickett, Office of the Director, Division of Nuclear Power Development, Department of Energy, Washington, D.C. 20545 Research and Technical Support Division, DOE-ORO Director, Reactor Division, DOE~ORO H. W. Behrman, DOE-ORO W. R. Harris, Rand Corporation, 1700 Main Street, Santa Monica, CA 90406 S. Jaye, S. M. Stoller Corp., Suite 815, Colorado Building, Boulder, CO ' W. Lipinskki, Argonne National Laboratory, 9700 South Cass Ave., Argonne, IL 60439 81. 44 'R. Omberg, Hanford Engineering Development Laboratory, P.O. Box 1970, Richland, WASH 99352° 82. 83-193, E. Straker, Science Applications, 8400 Westpark Drive, McLean, VA 22101 | _ : For distribution as shown in TID-4500 under category UC-76, Molten—-Salt Reactor Technology " "