ORNL-CF-61-8-86 by The "E" Generation

HOME

sEP t 9 1961

HELP

UCN-2983 (3

11-60]

OAK RIDGE NATIONAL LABORATORY Operated by

UNION CARBIDE NUCLEAR COMPANY Division of Union Carbide Corporation

Post Office Box X Oak Ridge, Tennessee

61-8-86

External Transmittal Authorized DATE:

August 18, 196l

SUBJECT: TO:

Thorium Breeder Reactor Evaluation. Part I. Fuel Yields and Fuel Cycle Costs of a Tvo-Region, Molten S a l t Breeder Reactor Distribution

FROM:

W. L. Carter and L. G. Alexander

COPY NO.

ABSTRACT

The MSBR (loo0 Mwe station) is capable of giving fuel yields of about 7$/yr (doubling time = 14 years) at a Asel cycle cost of approximately 1.5 mills/kwhr. A t fuel yields of 1 t'o 2$/yr (DT r 100 t o 50 years), the fuel cycle cost extrapolates t o 0.65 millS/kWhrj a t 4$/yr (DT 25 years), t h e fuel cycle cost is about 0.85 mills/kwhr. A l l systems were optimized with respect t o fuel cycle processing times. The effects on breeding performance of uncertainties in the epithermal value of q-233, uncertainty i n v a u e of the resonance integral of Pa-233, variable thorium inventory i n f e r t i l e stream and inclusion of ZrF4 in reactor fuel were evaluated. These effects may be summarized as follows: 1. A U O $ variation in the epithermal value of q-233 from "recommended" value causes a &2,5t o f3$/yr variation i n fie1 yield but only a 20.06 mills/kwhr variation i n fuel cycle cost. 2. Using 900 barns instead of1200 barns for Pa-233.resonance integral has only a small effect on breeding performance; the lower v+ue increases fuel yield . about O.25$/yr and lowers fuel cycle cost about 0.01 mills/kwhr. 3. Doubling the thorium inventory adds about 1.9$/yr t o Puel yield and 0.2 mills/to fuel cycle cost. 4. Five m o l e $ ZrF4 i n LIF-BeFe-W4 Fuel salt decreases fuel yield about 0.5$/yr, but fuel cycle cost is negligibly affected,

.c c

- w.

\. NOTICE

-3

T h i s document contai information of a preliminary notum and was prepared primarity for internal use at the Oak Ridge National Laboratory. It i s subject t o revision or correction and therefore does not represent a final report. The information i s not to be abstracted, reprinted or otherwise given public dissemination without the approval of the ORNL patent branch, Legal and Information Control Department.

LEGAL NOTICE

f h i a n p o r t wos proparod os on occount of Govornmont aponsored work.

Noithor tho United Stotea,

nor the Commiaaion, nor ony person octing on boholf of tho Commiaaionr

A. Mokos ony

worranty or nprosontotion, orproasod w impliod, w i t h roapoct to tho occurocy,

completonoaa,

or uaefulnoas of the Informotion contained in thi.

ony informotion, opparotua, morhod, or procoaa privotoly owned righta; or

disclosed

ropwt, or thot the uao of

in t h i s report mor not infringe

B. Asaunoa any liobilitita with napoct t o tho uao of, or for domogoa roaulting from tho use of ony informotion, opporotus, nuthod, or procora disclosod in thla report. A# uaod in the above, " p r i m octing on boholf of tho Commisrlon" includoa ony omployoo 01 controctw of tho Commiaaion, or omployoo of auch controctor, to the oxtont thot auch omployoo or controctor. of tho Commisrlon, or omployoo of auch controctor proporor, diaaominotes, or provide. occeaa to, ony informotion pvauont to his omploymont or contract w i t h tho Conmiasion, or h i s omploymont w i t h auch controctor.

I I

-.-

-3-

. ,,-

e 3

43 43 45 47 47 50

53-

T7 58 59

t‘

e Y

59 59 60 62

FORklORD

3 5 5

'As pa+k of the GRk, responsibility for guiding the ARC!PhermaJThorium Breeder Reactor Program, an evriluation of types of reactors -capable of ef-' ficlent utilization~of thokum was lr@tiated at GRNLin July,1959. Included in this evskmtion were studies on the Aqueous HomogeneousBreeder Reactor @RR), Molten l3aLi Rzeeder Reactor (MS&i), Graphite-Moderated Gas-Cooled Reactor (GGRR), Ikuterium-Moderaied Gas-Cooled weeder Reactor ~(DGRR).and cansaian-Deuteri~-Ur~i~'Reactor .(C~). This report presents the results of the MSBRevaluation. A comparison, of sJ.l five of these r&actors has been presented in.:t~'prekus reports by this study group. The rea& is referred'to these reports.for an appreciation of the perfOIFma.EkCe of these s&eral sy8tems. These _. reports are: ~" .I L. G.~Al&&ider, et al., Thorium Breeder Reactor lksltitioi. Part I. Fuel Yields and S-Cycle Costs for Five !Phemal Breeders, GRBL-CF-. h-3-9, March 1, 1961. L. G,. Alexander, -'et ril., Thorium Breeder Reactor Evaluation. Part I. Ibl Yields and Fuel Cycle Costs In Five %&rmal Breeders, GRRL-CE'63.03-g (Appendices, Part I), March 1, 1961. :

‘-

:’

A two-region;.@olten salt breeder-reactor (MSBR)having core dimensions approximately 7.7 ft diameter by 7.7 ft high and surrounded on the ends and,sides by a 3-ftthick blanket was studied'for determinatiozi of is breeding perfor&nce &d fuel cycle. cost. The core compositionwas ~pproxinatelyl6vol $&fuel-bearirig sslt, 6.7~~ $ fertile s&earn and-e.3 vol. $ graphite; side blanket composition was w vol $ fertile 1- criteria of the study were that the reactor comstream and'10 ml $ gra;phite, Basic plex be capable of producing power at a, rate of lOOO$we . .. and that chegical processing c be carried out oiâ&#x20AC;&#x2DC;site, Two reactors were req&ed, producing steam at 1800 psia and lOwoF. The fuel~s~slt jjassed th&ugh the core and upper end blanket in'sme. two-pass, .,. bayonet tubes madeof~impermeable &aphite which are inserted ti &enings in the Graphite moderator. The region between the core and re&tor vessel dd the 'annuli between the fuel tubes aud moderator are filled tith fertile titeri&. To mU.mize inventory the f+uel stream pump and heat exchanger are mounted directly above the reactor corei.. The fuel salt was ir 63-37 mole $ mixture of LiF-BeFg cor$xining at equilibrium about 25 gm U per k$ salt, uf which about 18 gm was U-233 and the remainder was higher isotopes. The fertild'salt was a 67-18-15 mole $ mixture of LIF-BeF&BiF~. At equilibrium the fertile stream contained from n0 to 2400 gm U-233 plus U-235 per tome salt. The fuel salt was processed for fission product removed by the fluoride volatility process and the HF dissolution process, A portion of the fuel salt was discarded during each processing cycle for removal of fission products not renoved by HF dissolution. The fertile stream was processed by fluoride volatility Only; fission product accumulation in the fertile stream was maintained,& a tolerable level by discarding the fertile salt inventory on a 20-year cycle. In this reactor only 1.3 - 6.65 of the fissions occurred in the fertile stream. Nuclear calculations were performed using the 34~group, multiregion GNUprogram8 for the IBM-704 and the Cornpone program' for the ORACLE. after attaining criticality in these calculations, further computations were made using the ERG-~program10 for the IBM-7th to determine the equilibrium conditions It is the equilibrium results that are reported here.

b d 3

The MSBR i s capable of breeding over a %d ite range of operatlng conditions giving FudL yields* a s high as about ?$/year for a doubling time of about ll.5 Aril-power y e a r s . A t this high yield, however, a premium Atel cycle cost of approximately 1.5 mills/lrwhr is incurred prlncipaJ.ly because of high Fuel stream processing charges. (Ilhe fie1 cycle cost was optimized by determining for each fuel yield We most economic combination of Fuel stream processing cycle time and f u e l salt discard cycle time. The f"uel yield was made t o vary by assuming several values of the fuel stream poison fraction and the f e r t i l e stream cycle time. In the r e a b of more econamical. operation, Atel cycle costs as low a s 0.65 mills/kwbr are predicted at Are1 yields of 1 t o 2$/year. When the fuel yield i s mills/kvhr. At this latter k$/year, the fuel cycle cost is approximately condition, the income from sale of fertile material just offsets the asznral inventory charge. Calculations for-arepresentative s e t of operatconditions were made t o evalwte MSRR performance in the light of uncertainties in nuclear data (Value of q-233 and the resonance istegral of pcl-233), variable thorium inventory and addition of ZrF4 as a stabilizing agent for t h e reactor fiel. Eta values a t epithermal energies within&IO$ o f t h e values recammended for t h i s study were employed I n nuclear calculatibns giving a &e.>t o f3$/year variation i n fit+, yield; corresponding Fuel cycle costs were negligibly affected (i0.06 mills/kWhr). Reactor performance using a resonance integral of1200 barns for Fa-233, used for this study, was compared with that for a 9OO-barn value; f u e l y i e l d wa,s lqproved about 0125$/year with a negligible lowering of the Asel cycle cost.

A lower thorium

inventory (140 tonnes vs 270 tonnes) decreased the fiel field about 2$/year w i t h cycle cost. A representative a corresponding tiecrease oi 0.2 mills/k~"nr in calculation in which 5 ' e $ ZrFd was added t o the fie1 salt indicated that the Fuel yield wauld be lowered by about O.5$/yea.r and that the Atel cycle cost would be negligibly affected ed t o a similar case containing no zirconium.

1.0

~OmCTIon

W A

The work on the Molten Salt Breeder Reactor (NSER) reported in t h i s memormdum i s a portion of a more complete study on +he& breeder reactors, which includes the Aqueous Homogeneous Breeder Reactor (ABBR), t h e Liquid Bismuth Breeder Reactor (LBBR), the Gas-Cooled Graphite-Moderated Breeder Reactor (GGER), and the Deuterium-Moderated Gas-Cooled Ereeder Reactor (DOBR). llhe important results of the complete study on all flve reactors is reported in ORE& CF-6l-3-9 1 by Alexander e t it is the purpose of t h i s memorasdum t o present more detailed data and calculations on the MSJB than those i n c l a e d in the reference memra?xlm. It i s â&#x201A;Źulvisable for:the reader t o examine ORmL CF-61-3-9 in conjunction with t h i s memorandum i n order to make a comparison of the several thermal breeders and t o obtain information on the MSBRthat may not be repeated herein. The MSBR was examined with the viewpoint of obtaining a relationship between breeding potential and economic performance. Breeding potential is related directly t o neutron economy and is therefore associated with the cmposition and design of the reactor, Economic performance i s determined by the annual charge on such items as the capital investment in the reactor installation, capital investment i n chemical. processing plants, operation of these plants, inventory of valuable materials (e.g.9 uranium, thorium, f'uel carrier salt and f e r t i l e carrier salt), use of these materials, and waste disposal. On the other hand, income from bred, fissionable material in excess of that required to refuel the reactor i s credited t o the econ@c performance. Two of the above charges have not been included i n this cost asalysis because no reliable cost data are available; these are the capital investment in the reactor installation and waste disposal charges. In defense of amitting waste disposal charges, it might be said that since all wastes are solids the disposal charges Kf13. be & very small'fraction o f t h e total. charges. It is observed that the r a i n i n g charges are concerned with the reactor fiel cycle and henceforth are referred t o as fuel cycle costs. In order t o make a breeding system of the MSRR, it is necessary t o exercise control over those neutron poisons that are amenable t o control3 some poisons, such as reactor structural materials, are fixed by design requirements. A significant advantage in neutron economy i s realized by controlling poisoning

from fission products by chemically processlng fuel and f e r t i l e streams for

-, V

- 9 -

their r e d . It i s a m e n t that t h e system in equilibrium may be operated a t any desired poison level between that corresponding t o some practical minimum

It i s customary t o identify fission product poison level in a reactor as the poison fraction, which i s defined as the r a t i o of neutrons absorbed in fission products t o neutrons absorbed i n fuel. There i s an inverse relationship of poison f'raction t o breeding and economic e, In order t o maintain h i reeding performance,. it i s necessmy t o chen+cally process fuel and fer e streams on a reiatively Frequent schedule a t the expense of hi@ Fuel cycle cost, On the other hand, less Frequent processing lowers the fuel cycle costs but has an adverse effect on breeding performance. The &el cycle cost associated with each poison fraction can be optimized by the proper choice of fuel stream cycle time and Arel salt discard time. (See Section 2.3 for a discussion 'of the chemical processing system.) In t h i s study 8ll. fuel cycle costs have been optimized With respect t o Azel stream processing conditions but not With respect t o fertile stream processing conditions. The fertile stream conditions were included as a &meter study In which a series of fertile stream cycle times in the range 35-200 days were studied for each value of fuel stream poison Fraction in the'range 0,OU 0 . 6 5 . The pertinent results are exhibited as plots of he^. cycle cost (mills/k~hr)versus fuel yield ($/year) and poison fraction emd that of cmplefe burnout of fission products.

e 7

. *

2.0 DRSCRIFTION OF SYSTEH -. j 2.1

,.-. bd

E&ysical system -,

The molten salt breeder reactor 'examined,in this study is based upon the I.design of Mad&ersonl3 'and is pictured schematically in Fig. 2.1. The reactor is cylindrical with a core 7.66 f% in diameter and 7.66 ft high. The core is surrounded on the sides-and.ends byea 3-ft-thick blanket. Al-ft-thick graphite reflector surrounds the blanket on the sidesandthe ends. The reactor, heat exchanger'and circulating pumpare arranged in a compact, vertical configuration tominimi~e the fuel volume, Surge dhme for the system is provided in the chamber housing the pump impeller, '/ Reactor Core. !Chereactor core is made entirely of graphite formed by assembling 8-in. square prisms. The corners of adjacent prisms are machined to form vertical passages of circular cross section about 5 in. fin diameter. The fuel -salt passes through the core in tubes of bayonet construction which are : inserted into these machined vertical passages$ the fuel tubes are,made of impermeable graphite. The outer tubes (see Fig. 2,l) have inside diameters of 3.75 in. and walls 0.75 in. thick. They are joined to an INOR- metal header by means of flanges,- frozen-plug seals, brazing, or transition welds. These,, '. joints are,presumed to be substantially leakproof. The inner tubes have inside diameters of 2.4 in. and walls 0.25 in. thick. They are joined to the inner plenum of the metal header by slip joints; these joints need not be leakproof since somebypass leakage at this point can be tolerated. The reactor contains approximately 90 bayonet tubes. Sufficient clearance between the fuel tubes and graphite moderator is provided to allow for differential expansion between the moderat,or and the metallic Are1 plenum. tie1 s&t enters at ll25*F, passes down through the annulus in the bayonet tube, rises through the inner tube at 20 ft/sec, and exits at 13OO.F. It is collected in the plenum and passes up through a duct to the impeller of the pump from which it is forced through the tubes of the heat exchanger. After leaving the heat exchanger, the cycle for the salt is repeated. The sslt circulates at approximately 50,000 gpm, removing1070Mw of heat. The heat'exchanger contains approximately 8100 tubes (INOR-8) which are 0.375 in. in outside diameter and have 0.028 in. walls. The shell side of, the heat exchanger contains molten sodium. .

- 11 -

UN C L A SStFlEO ORNL-LR-DWG.

46040R

SECONDARY ’

CCOOLANT

EXCHANGER

zc Y

€TAL

SALT

Fig. 2 . 4 . M o l t e n S a l t

B r e e d e r Reoctor.

-12-

Reactor R&a&et. The major portion of the fertile salt circulates through the side and 'end blanketsi however, a sm6J.lpOrtion bypasses through the core in the passages between the fuel tubes and the graphite moderator. In its passage through the reactor the fertile salt temperature rises fram ll5O'F to 1300eF; this sensible heat is then removed in a sodium-cooled heat exchanger. The s6.lt circulates at approximately 3900 p and removes about 112 MMof heat. This is about 10s of the totti'reactor energy; however, only about 1.3 - 6:6$ of the reactor energy originates frcnn fissions in the fertile streasz. The heat exchanger contain6 approximately 1000 tubes (INOR~) which 6re 0.375 in. in diameter rznd'have 0.028 in. walls. Reactor Composition. The approximate volumetric ccnnposition of the reactor core is as follows: 16% fuel stream, 6.74 fertile stream, and 77.35 graphite. The volumetric composition of the side blanket is 90s fertile stream and 10s graphite. The top end blanket contains both fuel and fertile stream; the volumetric composition is 16% fuel stream, 74s fertile stream, and lO$ graphite. Additional data on the reactor and heat removal system are given in Table 2.1. I 2.2

Salt Composition

The fuel salt consists of a mixture of 63 mole $ LiF and 37 mole $ BeF2 containing sufficient UF4 (equilibrium mixture of t&nium isotopes) to make the 6yStem critical - about 0.35 mole $. The fertile stream ha6 a basic composition of 67-18-15 mole $ LIF-ReF2-ThF4. The equilibrium mixture of cour6e contain6 Pa-233, uranium isotopes and a small concentration of fission products. The uranium content of the fertile stream is maintained at ,a quite low level by the efficient fluoride volatility processdng method (see below)3 therefore,' it is not edxmely important that the fertilestream volume be kpt low. In fact, in 6omecases it is desirable to have a large exce66 fertile-Stream volume to decrease neutron losses by protactinium capture through the dllutlon effect. The distribution of fuel- and fertile-stream volume6 inside asd Outside the &%%ais tabulated in Table 2.2.

- 13 Table 2.1.

Molten Salt Breeder Reactor Plant Data(a)

General Station electrical power, MwE Station net thermodynamic efficiency, $ Number of reactors per station Thermal power per station, MwT F’raction of electrical power fed back into plant Geometry of core

42.3 2

1182 0.03 cylinder (L/D = 1 )

Moderator Volume fraction of moderator in core

graphite

0.773 7.66 7.66 3

Diameter of &ore, ft Length of core, f’t Thickness of blanket, ft Volume fraction of moderator in side blanket Volume fraction of moderator in end blanket Reactor vessel material Reactor v e s s d thickness, in.

Mean pressure in reactor, psia Diameter of core f’uel channds, in.

lo00

0 010

0 .lo

IMOR-~

1 375 <loo

3 -75

Fud. Stream

63 mole $ LiF

Fuel carrSer

37 mole 119.5

Density (12CJQ°F), lb/ft3 Fraction oithepmal power removed by fuel stream h Mean heat capacl Power density in portion of fuel s t r e w extermA t o reactor, Mwt/ft3 Liquidus temperature, (b) OF

0.91

0.544 ._ 7.6 ~

ation now rate, &/sec Velocity (f’t//sec) of ~ z e stream l in

849

178

Core End blanket

B”F2

. .

Table 2.1

. Continued

';id

Heat exchanger data:

tube, side

Tube outside diameter, in. Tube wall thickness, in. Material

0..375

Tube velocity, f't/sec Flow rate, lb/hr Fluid temperature % *F ELuid temperature out, OF Pressure drop, psi Mo. tubes per exchanger Tube length, f't Tube bundle dianeter, in. Inside film coefficient, ~ t u / h r - f t ~ - * ~

s h e l l side v

0.029

IMoR-8

3.87

IEOR-8

107 4.35

1300 1125

107

900

1175

78 8110 u.13 69

100

8020

Tube -1 coefficient, Bt;j/hr-ft2-mF Scale coefficient, 9tu/hr-fi2-*~ Outside film coefficient, Btu/hr-f'b2-'F

10, OOO

0 v e r - U coefficient, ~ t u / h r - f i ~ - a ~ Out.side tube area, f"t2

2620 8320

7080 @,goo v

Fertile Stream Fertile stream carrier

67 d e 4 LIF 18 mole $ BeF2 15de$'fhF4

Density (1200*F), lb/ft3 Mean heat capacity, Btu/lb-*F &action of thermal power removed by f e r t i l e stream heat exchanger Fraction of fission power produced in f e r t i l e stream Mquidus temperature, (b) *F station flow rate, &/sec Heat exchanger data: Tube outside diameter, in. Tube wall thickness, in. Material mbe velocity, ft/sec

192 0.32

0.09 0.013-0.066

932 22.6 tube side

shell side

-V

0 375 0.028

=OR-8

14.I

mm-8

- 15 L,

Table 2.1.

Continued tube siae

plow rate, lb/hr Fluid temperature, in, OF Fluid temperature aut, OF Pressure drop, psi No. tubes per exchanger Length of tubes, f% Tube bundle diameter, in, Inside film coefficient, 9tu/hr-ft2-0~ Tube w a l l coefficient, Bt;~/hr-ft'-~F Scale coefficient, ~tu/hr-ft'-~~ Outside f i l m coefficient, Btu/hr-ft2-"F 0 V e r - u coefficient, ~ t u / h r - f t ~ - * ~ Outside tube area, ft2

she11 side

5.98 x lo6 4.48 x 106 1300 1 1 s log '1050

900 1175 100

1907 27 5550 5660

(a) A number of items in this tabulation are f r o m a study by Spiewak and parSly.14

f Y

(b) Temperature at which LiF precipitates.

- 16 Table 2.2.

Distribution of Fuel- and Fertile-Stream V o l u m e s i n the Molten Salt Breeder Reactor

Volume fraction

Vofume per station (ft3)

Fuel stream i n

Core Upper end blanket Gwer end blanket External t o reactor Dump tanks and miscellaneous Total

0.16 0.16

113 08.4

280.6 48.2

530 2

Fertile stream in Core Upper end blanket Lower end blanket Side blanket External t o reactor .Total

{ 409 2470 3026 6000

- 17 . LI

2.3

chemical Reprocessing 6ystem

A flow diagram o f t h e chemical reprocessing system is shown in Fig. 2.2.

The processing operation consists of three parts: fuel salt purification, uranium recovery fromthe f e r t i l e stream, and helium sparging t o remove fission gases from the fuel salt. Fuel Salt Purification. The ael salt is purified in the fluoride volatilityBFdissolution process by punplng a side-stream of the circulating molten salt through the processing plant in a specified cycle time. The cycle time is a function of the poison fraction a t which the reactor 1s permitted t o operate, whidh in t h i s investigation is a parameter.

F: 1

The first step in purification i s t o fluorinate the molten salt vith elemental fluorine t o volatilize m6. This uranium hexafluoride i s then burned in hydrogen t o produce UF4, which is recycled t o the reactor after dissolution in the recovered carrier salt. Uranium-f’ree salt, containing fission products, flows f’ram the fluorinator t o the HF’ dissolution step. Here a separation I s made betweenetheaaLt and the bulk o f t h e fission products. The carrier salt is dissolved in a 90s HF-10$ 5 0 solution leaving fission products, principally rare earths, as insoluble material. The carrier salt is recrystallized, fortified with recovered UF4, and recycled t o the reactor. I n order t o purge those fission , products which are not removed in +he HF dissolution step, portiona of the fuel salt are periodically removed and fresh make-up salt le added. The fission products purged in this manner include mainly the alkali metals and alkaline earths such as Cs, Rb, Sr, Ba, Te, Se, mb, Cd, Ag, Tc, etc. The &el salt replacement cycle time depends upon the fuel stream cycle tlme and the poison fraction. It is possible o achieve a kpecified poison fraction with several combinations of fuel stream cycle time and fuel salt replacement cycle t h e as is shown in Figs. 5.1 and 5 The proper replacement cycle is determined by optimizing the fuel cycle CQ h respect t o several ombinations of the two cycle times.

Fertile Stream Processing;. The fertile stream i s processed i n the fluoride v o l a t i l i t y step only. The salt is circulated at a specified r a t e through a -fluorinator where contact with fluorine gas volatilizes Ups. The s a l t then returns d i r e c t l y t o the blanket without additional treatment. The m6 from the fluorlnator i s reduced with H2 t o Wkwhich is blended with UF4 recovered f’rom the fuel salt for recycle t o the reactor. Ekcess production is sold. .. .

. -. . .-

.~~

. ..

. . . ."

I..

.. _". .

. ._

^ .

..... . . .

UNCLASSIFIED OWL-U1-DWB 64l6e

FIG, Z : Z

SCHEMATIC FLOW DIAGRAM OF MOLTEN SALT BREEDER REACTOR FUEL a FERTILE STREAM PROCESS1NG

L, .? f

the M S B at ferbile sitream cycle t i m e s less khan 100 days such a small of fissions (4,6$) occurs in U e fertile stream that it is not necespurif'y the sal% i n a HF dissolution step. me flssion-product build-up enough that their level can be conveniently controlled by replacing the s a t on a relatively long cycle. A 20-year cycle has been specified in this study. It wllf. be observed that protacrt;inium is not removed from the. fertile s a l t in t h i s proce~s. PrOtactiniUm builds up in the salt u n t i l I t s decay r a t e is Just equal. t o the U-233 production rate, The effect of Pa-233 on the neutron economy I s contrcilled by adjusting the v d . e of the f e r t i l e stream, lasger Y O ~ W Z E giving Peirer neutron losses t o protactinium. Fission Gas R e n m a L Fission gases are reaoved fromthe f u e l and fertile streams by sparging w i t h helium. Xenon, w o n anb the halogens are expected t o be removed in this way, The aPf-gas is passedthrargh cham& beds where the fission gases are absorbed. H a l i u m I s recovered for reuse.

In *action sary t o is slow

Power Generatian Cycle The steam cycle of the TVA John &vier power plant was used as a model for the MSBR concept?* Steam conditions axe taken as 1800 psla and 1050.F; the condenser pressure I s 1.3 in. &I Addltioml data on power generation equipment 14 are given by Spiewak and Parsly.

2.4

- 20

3 00 DESIGN BASES AND

CaMpuTATIoEJAL METBODS a

3.1

Plant Size c

Based on a study by Robertson,2 it was assumed t h a t f'uture pwer stations in the United States would have a capacity of the order of lo00 M m . Consequently, t h i s size was chosen for t h i s study. Also it was assumed t h a t anyone building a plant of t h i s size would be u n a t o +stall. the entire load in a single reactor; therefore, a t least two reactors &e specified for each station.

3.2

On-Site Processing

On-site chemicd reprocessing was chosen for the station. This method lends itself t p better control and definition of in-process inventory. Reasonably reliable cost estimatesu are available on fluoride volatility plants for processing-core and fertile streams.

3.3

OK!rating Conditions A l l caLculations were made for continuous, steady-state

ation of the reactor complex. To avoid complicated calculations of startup and shutdown, it was assumed t h a t the reactors would be continuously fueled and processed, and that the operation had been going on sufficiently long for all fission products and heavy isotopes t o be in equilibrium.

3.4

Product Camposition

The product composition may vary between the limits of almost pure U-233 t o spent fuel.. However, i n a many-reactor system complex, the fixel yield (or daubling time) is unambiguously defined only when the product has the same composition a s t h e average composition of the entire system; i.e., reactor p l u s chemical processing systems. Calculated portions of the recovered spent fuel and of the bred material are removed as product at the m6'vF4 reduction step. The product i s an equilibrium mixture of uranium isotopes; viz., V-233, U-234, U-235, and U-236.

3.5 System Inventory In order t o be consistent and unambiguous in the definition of fuel yield, t h e inventory should include all fissionable and potentially fissionable atoms (U-233, U-235, and Pa-233) in the entire system. Included i n the inventory is any f'uel that is reserved t o 8 3 . l ~reactor ~ operation during shutdown of the chemical processing plant. In this study a 30-day fuel reserve was chosen.

Ld P â&#x20AC;&#x153;;

. 7

3.6

neutron josses.

Fission-product poisoning in-this reactor was based on a study by Burch, Campbell, and Weereq3 whckmadea study of the cumulative effect of 4.4 isotopes divided'into four &ups, This phenomenonIs discussed more completely In section 5.0. Xenon Poisoning,' It was assumed-that xenon could be continuously removed from the circulating fuel by gas sparging and'malntained at a level.such-that the neutron loss to xenon Is 0.005 neutrons per fuel absorption. Since there are so few fissions (46s) in:the 'fertile stream of the MSBR,all xenon losses were assigned to the fuel'stream. In choosing a value for xenon losses , it was assumedthat (a) neither xenon nor iodine is absorbed by the graphite .moderator or otherwise collects at the interface between salt and graphite phases1 or (b) if the pores and vacancies in the graphite are accessible to xenon and Iodine, the rates at which they diffuse into the pores are very much slower than the rates at which they are stripped from the circulate stream by the sparge gas. &her Fission Product Poisoning. Concernin@;the effect of the 44 fissionproduct isotopes studied by Campbell, Burch, and Weeren,3 four groups were dlscerned and treated separately. The firstcomprised noble metals which were assumedto plajse out on the cold zones of the c+Alatlng system; the second comprised halogens which were assumedto volatilize dur& the-fluorination step. A third gr&p'vhich is soluble in HP and therefore not removed in the dissolution step comprised 'the 6lkal.i metals (notably Rb and Cs), the slkaline earths (Sr, Ba, etc.) and's ~6cellancou6 gro&(!Ce, Se, Wb, Cd, Ag,,!k, etc.). This soluble group.is removed by replacement of the fuel salt on-some specified -. cycle. The final group Is the rare earths which are removed by precipitation 7 during; the.PPdissolution step. -The poison fraction Is thus not a simple function of the processing rate _ and is computed as described in Section'>.0 below. Fission product6 in the fertile stream are controlled entirely by the 200year throwaway-cycle of the thorium carrier since this stream Is not chemi, ,. tally proceked ,. . for fission product remov6l.. '_ .

-Fuel Processing Losses. Fuel processing losses are based upon laboratory and pilot plant data, which have indicated essentially quantitative removal of uranium from the salt by the fluoride volatility process. Consequently, losses that occurw3J.l be &most entirely In the UF@ 'JF4 reduction step. It Is believed that onlarge-scale operation these losses can be made quite sti --of the order of 0.014 of throughput.

G= 4

Carrier and Moderator Losses. Neutron losses-to the- carrier salts are based upon the use of a feed salt in which the lithlum component is present as gy.gg.at. $.Li?and 0.03. at. $ L16. A salt having a lower Li6 concentration would be desirable; however, it is questionable whether or not the premium price for such a salt is justified by the increased neutron economy. In this study the mono-energetic'capture cross section of graphite was taken to be 4;2 mb at 0.02!3-ev and was assumedto vary inversely with the velocdty of the neutrons. 3.7

Nuclear D&a

Muclear cross sections for this study were compiled by Nestor.' Following 5 the recommendations of Fluharty and Evans, a velue of 2.28 was assigned to eta of U-233 at thermal energies. The resonance integrsl of Pa-233 was assumedto be 1200 barns. Allowance for resonance saturation (self-shielding) was made only in the case of thorium, and here Doppler broadening was also taken into account. Epithermal'cross sections of other isotopes of interest were adjusted to agree with the resonance integrals tabulated by @toughton and Halperin.' !I!he composite 2200-meter cross section of fissid products (exclusive of Xe, Sm-15& and S&149) was taken as !jO barns per fission and assign+ to an artificial 'element called "fisslum." A'resonance integral of 170 barns per fission was assigned to fissium as suggested by the work of Nephew.7.. Howover, in the computation of poison fraction, resonance integrals had to be assigned to-each individual fission product. Available values were taken from nephew; if no value was reported, it was calculated from available data.

l31el Carrier and Blanket Carrier Cross SeCtions. The fuel. carrier, which is composedof a mixture of Li7, Li6, B&, and F atoms, is treated as a single, pseudo fuel-salt atom in the nuclear calculations. It is convenient to do this because the GNUcode,is limited in the number of elements for which absorptions

s G

’

be ~calculated; consequently, hn&g these elements saved space on the tape needed calculations. A pseudo ,. cross section for.- thefuel salt was obtained by normaUzing the cross section to the basis of one atom of,L17 and summingthe results. In the normalization the cross section of each atom was multiplied by the atomic ratio of that‘particular'atck to Li7.. The atomic concentration of the salt Is then pressed as the atomic concentration of Li'l.ln the salt. !Lhe lithium componentof the salt was assumedtobe g$?;ggat; $ Li7. The fertile-stream carrier was treated‘in a similar &er with the cross sections of each component atom norms&ized to the-basis of an atom of thorium. The fertile-stream carrier contains Li7, L16, .Be, F& and Th atoms.. The atomic concentration of the carrier is then A. expressed as the j atomic concentration of thorsum in the salt. The L17 purity Is the s& as used in the fuel carrier. 3.8 Ruclesr Cslculations

3 u

Muclear calculations on the MSBRwere performed using two different reactor codes: the 3bgroup GNUcode8 -for the E&J704 and the &%pone code9 for the ORACLR. The use of the'two‘hodes expedited theealculations. The reactor was first treated In spherictdl~gecmdzy as a homogenizedsystem using GNU, and a criticality search was madeto determine the critical con&ntration of thorium and protactinium In the core. The diameter of the equivalent sphere was faken .as l,Og times the cylinder diameter. -!&is +nformatlon vas then used in Cornpone calculations, also~ln spherical geometry, to determine the thorium concentration in the core of the critical heterogeneous reactor. Fe heterogeneity of the MSBRcould be studied on Cornpone through the use of- 'disadvantage factors"; disadvantage factors could not be applied to GIW. Sfnce all parameter. studies were to be made:on the equilibrium reactor, the critical reactor concentrations frcun the Cornpone cakulort~on #ere used as 5npu-t information~for an equilibrium reactor calculation us+ng the ERC-5 codelo for the IBM=-701). This cs&ulatlon determined the concentrations and elemental neutron absorptions in the critica& equilibrium reactor. A more.detailed discussion of .the nuclear calculations appears In Section 3.2 ,. ..' -of Ref. lW t . !Chedisadvantage factors m&ntioned ,above were used-to relate the concen, These tratlons of the homogenizedreadtor to those of the heterogeneous reactor. fa&ors were determined In a lattice-cell, calculation by meansof the Cornpone

- 24

program for the O R A W w h i c h yielded s e t s of 34-group disadvantage factors, one set for each region of the l a t t i c e cell. When employed in a 34-graup finitereactor, Cornpone calculation, the correctly "disadvantaged" absorptions of each element i n each region of the reactor were calculated. The disadvantage factor is defined by the following equation:

where vC

= volume of l a t t i c e c e l l = volume of region

fin

j in the cell

= neutron flux i n differential volume dV i n neutron graup n.

3.9 Costs of Mterials and Facilities and Interest Charges !&e basic cost data employed i n this study t o calculate fuel. cycle costs are given in Table 3.1. These data are believed t o be representative of the costs of MSBR materials and amortization charges. Ibel Stream Processing. The capital cost of the fuel stream processing plant was based upon a cost study by Weinrich,' who estimate&the capital charges for a plant t o process continuously about 20 ft3/day of fhel salt. A

pldt of t h i s capacity is within the region of Interest of t h i s study. Weinrich*s * data were reviewed by Chemical Technology Division personnel for copparison with more recent cost data and cost estimating practices at Oak R i Q e National Laboratory and, as a result, his data were adjusted upward. These data and the ORML revised figures are presented in Table 3.2. The ORB& estimate is approxiaately twice that of Weinrich. These estimates were made from f'unctional flowsheets representing the best available design information on the fluoride v o l a t i l i t y and HF dissolution processes. In optimizing MSBR systems t o obtain the most economic combination of fuel processing plant cost and fiel salt replacement cost, it was necessary t o extrapolate the QRML cost estimate i n Table 3.2 t o both smaller and larger

*W.

0. Stockdale, D. 0. Campbell, and

W. L. Carter.

- 25

Table 3.1. Items and Basic Cost Data Included in the Cycle-Cost of a Molten SaLt Breeder Reactor

Unit Value

($/u 1

Uranium inventory

Interest Rate

159OOo

Thorium inventory (as T ~ F ~ )

2.p)

(&/Yr 1

4 12.7 (b 1

FWL salt iwnetory (U excluded)

4 0 . p

12.7 (b1

Fertile salt inventory (Th ex

45.6(')

12.7 (b 1

Thorium amortization (20-yr cycle)

Nel salt replacement

40.3

Fertlle salt replacement (20-yr

45.6

2.6

Fuel stream c h d c a l processing plant

29(d)

Fertile stream chemical processlng plant

29(d)

158

Breeding credit ( a ) ~h W u e a t $=/kg

plus $ 5 / ~for preparation of s a l t sal.ution.

(b) Includes interest at 6$9 income taxes a t insurance at 2.14.

4,6$, an8 local taxes and

(c) Based on LiF at $44/kg and 3eF at $15.4O/kg plus $11 kg salt for p r e p a t i o n . A t d c concentrahon of Ll i s 99.99$ L i

(a) Includes 14s interest on capital investment pius 155 for operation maintenance

. *

26 -

--. â&#x20AC;&#x2DC;.

Table 3.2. -Cost Estimate of..Facilities-for Continuous . Processing of Molten Sa&t,Rreeder Reactor 374. Stream. Weinrich's'dost Estimate Comparedwith Revisiori'kade by ORRL mWeinrfch's Estimate ($1 ?lkhk.sand Vessels (Core Sslt Section) Core saltholdtanks Core salt fluorinators UF6 chemicsl traps 'reduction

uF4-m6

W5q

311900 .

Vibrators, filters, burners, etc. HF dissolving tank RF evaporators HF condensing tower HF storage tank KOHscrubtower Mistikllaneous storage and utlllty tanks Sub-Total Installed cost (4.35 x cost) Coolers (Core Salt Section) m6

gas

39,ooo

19,500

tower

Ow Estimate ($)

coders

Reduction tower vent cooler HF'vapor desuperheater HF condensing tower vent cooler Circulating RF cooler CXrculatiugH20 uer Sub-Total InstsXl.ed cost (ml.10 x cost)

8,600

5,c)Qo 24,000 1Ol;m' a&=

41,300

3,500 20,000 294,400 397,440

33,.. 63,800

17,200 10,ooo em _

W,w3

4'6,200 82,600 7,m

20,000 547,300 ?38,8~

: 3*@

3,-

3,m

3,ooo

9,600 6,000 48,ooo

916,Qoo %ooo

2,400

72,6ot3

79,860

â&#x20AC;&#x2DC;-

.72,-

79,860

kessels and Tanks (Rlanket Salt Section) Rlanketsaltholdtsnks Blanket sat. fluorinator m"6 chem.icEiltrap Sub-Total IustsXLed cost (el.35.x

9,800 51500 .aSE 18,goo

cost)

25,500

Fertile stream processing estimated separately.

- 27 Table 3.2.

Continued

Weinrich's Estimate

Coolers (=-et

w 6 codier

($1

ORNL Estimate

($1

a t Section,) . .

Installed cost (la.10 x cost)

2,400 2,640

Miscellaneous 4ulpment Ausps

Agitators Filters Freon refrigeration system Fbel reconstitution system Electric heating f'urnaces Pipe heating equipment

F2 + gas supply systems F2 compressors Sub-Total Installed cost (4.35 x cost)

46, Ooo 6,OOo

69,700 6,Ooo

20, ax,

160,000

M,m

80,000

148,OOO

148,OOO 6%OOo

60,000 20, OOO 20, OOO

534, OOO 720,900

20, OOO 20, OOO

589, OOO 796,140

Sub-TOtaJ of' instelled cost of major equipment

A t t e n d a n t Facilities

Special instnunentation Genera instrument a tioi Panelboards and alarms Sub-Total I n s t U e d cost (d.40 x cost)

60,000 24,000 160,OOO 224, OOO

Piping, painting, ScaffolBs etc., Installed Cost Special plplrq &nerd pip%(") muipment footings and foundations (b 1 Pipe insulation

- 28 Table 3.2.

Continued

Weinrich's Estimate ($) Ruipment insulation Electrical. distribution, lighting, etc

20, OOO

ORE& Estimate ($)

144,OOO

20,OOO 144,000

28, OOO

36,300

75, OOo

25, OOo

25,000

15, OOo

meight ('1

2UEE

48,400

Sub-Total.

725,000

1,767,800

23-79580

901,650 4,508,300

Painting ( c 1 Remote operating equipment Field testing and inspection Operating and safety supplies

TOTAL INSTALLED COST

Contingency(e ) TOTAL DIRECT MA!FEFUALS AND LABOR

2,393,400

Fees and Expenses , I

I I

Contractor's f i e l d expense (f,d

U9,670

Contractor's overhead fee 0%e 1

359,OOo

Engineering and design(i1 Purchasing and shop inspection (3 1

478,700 119,700

2,254,150

901,f225,400

Estimated Cost of Additional Facilities Sampling f a c i l i t i e s

Ventilation Waste removaL C e l l s and buildings Laboratory Mock-up c e l l Crane

TOTAL ESTIMATED PLANT COST

4,970, Too

9,849,500

- 29 -

‘cd

‘:

~+ie

I \ f /

3..2. CoIltinued Footnotes \

(4 Estimated . by Weinkh & 25s 03 major equimnt purchase price. Estimated by Oral as lOO$ of'major'equipment purchase price. , (b) - .., ~_, , _ ,“ _r :._ -Estimated as 155 of major equipment purchase price. (4 Est@ated as 3% of major 'equipment purchase price. (d) Estin&ed as 4$ of major eq$pment,purchase price. ,' (4 Estimated by Weinrich,as &O$ of,total installeh cost. Estbated by.as 25s of t0tsi instsJ.led cost. (f)

-1 I. Estimated by Weinrlch as 5s of-t&al . -,

direct

materials and labor cost.

.(d Sumof contractor's field--expense.and overhead,fee taken by CBNLas 505 of tots3 direct materials.and,labor -v..costa . b) Estimated ' by.Weinrich'as 6) Estima$ed as 20s of . total.

15s of to&IL direct

materisJ.s and labor cost.

direct.'I. matgrlals and labor cost.

Estlmated~.as s of tot$l,direct ,, ,-materials and labor 'cost.

- 30 ,"--

plants. T&e extrapolation was made by assuming that the capita3 cost is proportional t o the 0.6 power of the processing rate. mis method of extrapolating cost data has been found reasonably accurate when applied t o the chemic& industry as a whole and t o plants which process nuclear reactor materials.

There I s a limit, however, t o the extrapolation in the region of low processing rates because at some low rate, which nay not be w e l l defined, it i s economic t o change from continuous t o batch processing methods. In t h i s study it was assumed that the lower limit of continuous processing wauld occur around 7 ft3/day, which corresponds t o a fuel cycle time of 75 days. (The stream volume was constant a t 530 ft3. I When the fiel cycle time is 75 days, the extrapolated cost curye (Fig. 3.1) indicates that the capita3 Investment is about $5 million. Furthermore, it was felt that the Investment i n ch Plat ~ o u l d not be sensitive t o M h e r increases in the cycle timej cons million value was assumed t o apply t o all plants having cycle time6 greater than 75 days. A batch processing plant was estimated by Weinrich t o cost $3.4 millionj the above figure dims a premium of $1.6 million over Weinrich's estimate.

fuel

Fertile Stream Processing. The f e r t i l e stream is processed only in a fluoride volatility step and, therefore, requires much l e s s equipment than the accompanying fuel stream processing plant. Welnrichu included the f e r t i l e stream plant a s an i n t e g r a part of h i s fuel stream plant design and did not make a complete separate breakdown of the costs. However, it was possible t o prepare a cost estimate for the f e r t i l e stream plant by extracting specific items from Weinrich's estimate and including allocations for instruments, buildings, etc. ORNL priclng procedures were applied t o prepare the estimate given in Table 3.3. This tabulation presents values that are apaicable t o a plant processing f e r t i l e stream a t a r a t e of 20 ft3/day, the same basis upon w h i c h the f u e l stream processing plant was designed. For t h i s r a t e it was estimated that the capital investment would be about $1.8 million dollars. These values were plotted i n Fig. 3.2j the remainder of the graph was obtained by assuning the cost was proportional t o the 0.6 power of the processing rate. 8

UNCLASSIFIED ORNL- IS-DWG

I, Continuous Processin$

IO=

500

FUEL STREAM CYCLE TIME (days) FIG. 3.1

CAPITAL .COST OF FUEL STREAM PROCESSING PLANT FOR A MOLTEN SALT BREEDER REACTOR’ ‘,

54.794

- 32 Table 3.3. Cost Estimate of Facilities for Continuous Processing of Molten Salt Breeder Reactor Fertile Stream Estimated Cost ($) Tanks and Vessels

S a t hold tank muor inator UF6 chemical traps UF@lF4 reduction towers Vibrators m6 gas coolers Reduction tower vent cooler

20, OOO 20, OOO 43,000 12, OOO 10, OOO

AMPS

Filters

Agitators Freon refrigeration Bred material reconstitution Electric heating furnace's

8,800

Pipe heaters

F2 compressor Sub-Total Installed cost (4.35 x cost)

21900 225,400 304,290

Attendant Facilities

40, OOO instruments Panelboards and alasms Sub-Total Installed Cost ( d . 4 0 x cost)

30, OOo 15,000 85,000 119,OOo

Installed Cost of PIpirg, Insulation, Paint- , etc. Specid piping General piping (..J,oo$ of major equipment -cost) EQuilpnent footings and foundations (=I.?$ o f major equipment cost)

u /--

3,000 225,400 33,800

- 33 I

Table 3.3.

Continued Estimated Cost

Pipe insulation Equipment insulation ELectrlce.l.distribution Binting (48 of mador equipment cos$) Remote operating equipment Meld testing and inspection 0perating.and safety supplies Freight of major equipment cost) Sub-Total

1,200

31OOo

~ , m 6,800

u,m 3,700 2,200

(4s

9,ooo 320,100

Total installed cost Contingency (1.2946 of t o t a l installed cost) TOTAL DIRECT MATERIALS MID LABOR Contractor's f i e l d m s e and overhead (=%$I of total direct materials and labor) Engineering EKUI design of t o t a l direct materials and labor) of t o t a l . Arrchasing and shop inspection direct materials and labor)

743,400 185,800 929,200

464,600 185,800 46,500

Addition82 Facilities Shared with Fuel lCjalt Processing Facilities

sawling.

($1

Ventilation C e l l s and buildings Laboratory Mock-up c e l l

1,748,200

UNCLASSIFIED OANL LR-DWG 54793

FIG. 3.2.. CAPITAL COST OF MSBR FERTILE STREAM PROCESSING PLANT

1 I1111

FERTILE SALT PROCESSING RATE (ft'lday per station

500

c, i

4.1 Reactor Size

For engineering reasons it was decided that the having a height equal t o the diameter.

d be a cylinder

Core Size. In determining the core s i z e of the MSBR it was necessary t o f+x certain reactor progez-ties. I n - t h i s study the thermal power, fuel stream velocity in t h e core, and the temperature.ri6e of the fuel I n i t s passage through the core &re arbitrarily chosen. me'dlameter'of the core is related t o these quantities by EQ. 2. d =

(%up cpqf)(bT) 4

where

= core thermal power per react stream velocity p = stream density c = heat capacity P f *action of core cross s e c t i AT = temperature rise U

cupied by flzel stream

When the appropriate numbers are substituted in this equation, a core diameter of ab0ut~7.7 f t is obtained. The power used in obtaining this diame t e r was one-half of the t o t a l core power for the station, givlng two reactors for the ;Lnstallation. .%is agrees with the decision that the t o m ststion load vauldl not be committed t o a single reactor,

nanket micknes;. The cylindrical core blanket on the sides and on each end. Based on previaus the thickness of the blanket was fixed at 3 ft on both ends an8 on the side. This thickness was sufficient t o reduce neutron leakage t o a tolerable level. As given in Table 2.1, the sidle blanket is 90 vol $ fertile stream and 1 0 val $ graphite. c

Reflector !&ic 6s. The reflector C t o be a 1-ft-thick block of graphite 'surrounding the side end blankets. %e over-all reactor dimenSiOn6, excluding the reflector, are 13.66 f t diesleter by 13.66 ft high.

- 36 4.2

GNW Calculations The calculations for the MSBR Were made a s indicated in t h e f l

Fig. 4.1. The basic nuclear calculations were performed on the mul 8 dimensional, 34-group GNU program for the IBM-704. The equilateral, cylindrical . reactor was treated as an equivalent sphere having a diameter 9% greater than the

cylinder; the 34 groups of cross sections consisted of 32 fast groups, an epithe& group for the energy range 5.5 m-0.6 ev, and a thermal group. Input Data. Input of specifications of reactor c densities of the several d e geometry, dimensions, and t h e ments i n the system. The concentration of each element was homogenized over the region i n which it appeared. Since a sm8U fraction of t h e fertile stream passes through t h e core of the MSBR, the core concentrations included the sum of fuel stream and t h a t portion of the fertile stream concentrations. The i n i t i a l values of fuel, moderator, and fission product concentrations used in the GNU calculations were based upon concentrations previously developed for the experimental gas-cooled reactor (KTCR) and from previous molten salt reactor studies. The elements considered in these calculations are given i n !Table 4.1. Output Data.

The GNU program provides a c r i t i c a l i t y search by which either

dimension or one or more concentrations are varied u n t i l the iplicat ion constant differs *om unity by less than some small specified Bs1ount. In these calcdxM.ons the reactor was made c r i t i c a l by varying the concentrations of protactinium and thorium i n t h e core. Since the thorium density i n . t h e fertilestream carrier is'flxed by the salt composition, t h i s I s equivalent t o varying the volume fraction of f e r t i l e stream i n the core, Mditiond. useful autput data were t h e fractions of neutrons involved i n absorption and fission reactions for each nuclear species i n each region of the reactor. a

4.3 Cornpone unit c a C d c a a t i o n The second step in the nuclear calculations was t o determine the atomic

concentrations of t h e heterogeneous reactor. (The GEJU progrrtm could treat only a homogenized system.) The core of the MSBR was visuaLized as being composed diagrammed in Fig. 4.2. The of a number of cylindrical, unit 9 unit c e l l contained s i x regions (see Fig. 2.1): inner f u e l zone, graphite tube, annular fuel zone, graphite tube, fertile stream passage, and graphite moderator.

Lid

icol concentrations

minimum fuel cycle

t t I

L-

- - ---

a -’--

a c)-

FIGURE 4.1 COMPUTATIONAL PROCEDURE FOR MOLTEN SALT BREEDER REACTOR

UNCLASSIFIED ORNL-LR-DWG 58644

- 30 UNCLASSIFIED

OR N I - L R D WG. 5 8 6 4 5

GRAPHITE MODERATOR FERTILE STREAM OUTER FUEL TUBE (GRAPHITE) ANNULAR FUEL ZONE INNER FUEL TUBE (GRAPHITE) INNER F U E L ZONE

, I

2.375 2.62

6.03 6.65

.36

11.07

14 1

I ~

Fig. 4.2 U n i t C e l l C o n f i g u r a t i o n f o r

.Breeder R e a c t o r Core.

olten Salt

- 39 c,

Table 4.1

.ofnements Considered in Budear Calculations the Molten Breeder Reactor Salt

pa-233

xe-135 Fissium

u-233 U-234 u-235 u-236

Sm-149 Graphite Fuel-Stream Carrier (a 1 (63-37 m o l e Q LIF-B~F~)

Fertile-Stream Carrier (b 1 (67-18-15m o l e $I LIF-BeF2-ThF4 )

(a) Uranium excluded i n nuclear properties of salt (b)

content excluded in nudear properties of salt

The c e l l was examined using the Cornpone code9 for the ORACLE t o develop a set of 34-group disadvantage factors for each region of t h e unit c e l l e These disa&vartege factors, defined in Section 3.8, were then used i n subsequent Cornpone calculations t o determine the concentrations for the critical, heterogeneous reactor. The Cornpone code treated the unit c e l l as

infinite cyLinder having zero

net current a t the outer boundary. Input information for the calculation was the stream concentration i n each region and the thickness of each region. The stream concentrations used were those developed I n the prelhdnary GNU calculation. I n addition t o the disadvantage factors, the code calculated the nniltiplication constant of t h e cell. 6.4

Cornpone Finite Reactor Calculation

The disadvantage factors were used in a finite reactor calculation u s i n g t h e Cornpone program t o determine the c r i t i c a l concentrations of the heterogepeous reactor. As in the GNU calculation the reactor was calculated in equivalent spherical getmetry.

- 40 -

Input Data. The calculations were made on a 3-region model core, blanket, and reflector. The concentration of each nuclear species, which were obtained from the GNU calculatioa, was homogenized over a region according t o i t s volume fraction i n the region. The set (or sets) of disadvantage factors t o be used with each concentration was specified a s w e l l as the dimensions of the region. The machine calculation multiplies each homogenized concentration by the appropriate disadvantage facter so that a l l properties that are dependent on the atomic density of that element are weighted by the relative flux t o which the nuclei are exposed. The concentrations are thus "disadvantaged" t o reflect the heterogeneity o f t h e system. For exampleo the absoqrbions i n the i - t h element in the j-th stream in the k-th region is computed by the double summation

where Fk(Au) is the group-mean flux in the homogenized Increment of volume

is the homogenized atomic conis the disadvantage factor, N i,3,k centration, and v;(Au) is the absorption cross section.

"Vkp

Di,j,k

It should be pointed out that disadvantage factors were applied o n l y t o

element events (absorptions and fissions) occurring in the core. I n the blanket and reflector, element events were calculated as that@ the disadvantage factors were unity for all lethargy groups. Output Data.

The Cornpone program determines the fractions of absorptions

and fissions of each atomic species i n each region and the multiplication constant of the reactor. The code does not make a "search" on any of the input information; hence it is necessary t o rerun the problem with adjusted input if the multiplication constant differs *om unity by more than a prescrfbed small amount. In these calculations c r i t i c a l i t y was achieved by varying the thorium. concentration i n t h e core, Reaction Rate Coefficients, Element absorptions i n each region of the reactor were used t o compute sets of reaction rates coefficients, Ci,j,k' which are defined by Eq. 4.

The symbols have the same definition as given above for Eq. 3 . Since the double summation i s computed by the Cornpone code, the calculation of C is straighti,3& forward. The coefficients are properly disadvant the use of Di ,3,k to reflect the heterogeneity of the system. a l s have the useful property that, when multiplied by the stream atomic concentration and the volume fraction of the stream in the consMered region, they give the f’raction of neutrons involved in absorption ‘hteractions with the I-th element i n the j-th stream in the k-th region. Rarthermore, i f t h i s f’raction i s multiplied by the tota,l number of neutrons born per unlt time in the reactor, the product i s the absorption r a t e hy element i in stream j in region k. This l a t t e r quantity i s very useful in c&culating the equillbrium s t a t e o f t h e reactor as discussed below.

In a calculation similar t o that described by Eq. 4, sets of fission rate coefficients were developed for elements that had a fission cross section. These fission coefficientsmre used in as entirely analogous manner t o the absorption coefficients t o describe element fission events in the streams and regions o f t h e reactor. A l l comments about the use of the absorption coefficients apply t o t h e fission coefficient.

4.5 EQuilibriam Reactor Calculations (ERC-51 4uilibrium Reactor Calculations were next performed on the c r i t i c a l Cornpone reactor by means of the EBC-5 code” for the IBM-704. This program integrated the reactor w i t h the Rzel and fertile stream chemical processing systems and cam-

p t e d pertinent equilibrium properties of the system.

Input Data. The equilibrium calculations required t h e fallowing input information: fuel and fertile stream volumes, volume fractions, process cycle times, process holdup tines, and c r i t i c a l concentrations$ reactor power, poison fraction override, f’uel reserve time, and recovery efficiencies associated with Are1 snd fertile stream processing. !Phe ERC-3 code solved a system of equations based upon conservation of mass, criticality, and conservation of neutrons; these equatiom used the absorption and fission reaction r a t e coefficients calculated fromthe Cornpone data. All neutrons were accounted for, including those absorbed in fertile materials, moderator, carriers, &c., and those leaking out of the blanket or l o s t as delayed neutrons.

- 42 -

Output Data. -The program calculates t h e equilibrium stream concentrations and neutron absorptions i n both mel and fertile streams for the elements l i s t e d in Table 4.1. Also the inventories and mass processing rates are computed for all uranium isotopes, thorium and protactinium. Since the sales philosophy is t o s e l l a product that has the same composition as the system mixture, the fractions of recovered fuel and fertile streams t h a t are directed t o sales are calculated. Additional values calculated by ERC-5 code are the fraction of fissions i n the fertile stream and t h e inventory of fissionable material reserved for a possible 3O-day shutdown o f t h e processing facilities.

%e program offers the option of attaining c r i t i c a l i t y by adjusting the U-233

and U-235 concentrations in the fuel stream or by adjusting t h e volume fraction of thorium in the core. I n these calculations the second option was employed. S i g h t adjustments in the amaunt of thorium i n the core had a negligible effect on the carbon-to-uranium ratio and hence on the neutron spectrum.

I n sane instances it was desirable t o specify t h e fraction of neutrons that would be allowed for losses in xenon, me1 fission products, and leakage. This condition could easily be treated on the ERC-5 code by specifying a fictitious atomic concentration and absorption rate coefficient t h a t gave t h e desired absorptions. Xenon and flrel stream fission products were treated this way because their amaunts are controlled by predetermined processing rates. Leakage is con< trolled by the reflector design.

1 I

5.0

rn STREAM m s o n FRACTIOR

CALCULATIONS

2.1 Poison Fraction The t o t a l poison fraction generated by fission products i n a reactor includes the contribution t o neutron losses from fuel stream plus f e r t i l e stream fission products in both core and blanket regions. of f e r t i l e stream fissions is a small portion ofthe t o t a l fissions and to simplie the calculations, the t o t a l poison fraction was assigned t o the fuel stream. I

definition,

Poison fraction =

c 1

where, %,l

f1,l f

182

@2 a =i

= atonic concentration of i-th fission product, atams/cm 3

= volume fraction of fuel stream i n core, = volume fraction of fuel stream in en8 blanket,

= average effective neutron nux in core, neutrons/cm 2-sec,

, neutrons/cm2-see, effective absorption cross section for i-th atom, cm2,

* average effective neutron flux i n end E

= t o t a l fission rate in react

= neutrons born per fission,

= neutrons born per neutron

= t o t a l a e l stream volume, c

The atomic concentration,

can be expressed in terms of known quantities by considering the steady state o f t h e i-th isotope. Equathg the production r a t e t o the sum of ell removal rates, there obtains

- 44 Ftyi - r =

%,A

L-

Ni,lEi

%,l (f1,l 41

The v4ue of N i,l from Eq. 6 can be substituted into Eq.

5 to

Symbols

not previously defined m e

i-th Isotope (for sone nuclides this number had to ed to account for the existence of a precursor isotope in the chemical processing schemes),

Ai = decay constant of i-th isotope, sec-1

Ei = efficiency of removal of i-th isotope in chemical processing, Ti = cycle time for I-th isotope in chemical processing, sec. The quantity EI/Ti in Eq. 7 expresses the removal rate of the i-th isotope in chemical processing. In MSBR processing, Ti assumes two values, identified as T1 and Tld, the values being characteristic of the chemical behavior of an atam in processing. T1 refers to those fission products whose removal is accomplished i n the EJ? dissolution step (see TEtble 5.1); therefore, TI is the actual fuel stream cycle time through the chemical processing plant. Tld is associated Kith those fission products whose remOval I s accomplished by discarding a portion of the uranium-free fuel salt each time the fuel stream is processed. The time Tld is in&@.?ndent of the time Tl; there is, however, the restriction that Tld must be greater than T1. In the economic cases, Tld will be several times larger than T1.

The total poison fâ&#x20AC;&#x2122;raction attributive to the fuel stream is the solution to EQ. 7. Through this equation the total poison fraction I s related to the cycle times T1 and Tld and thereby to the capital investment in the processing plant and the replacement cost of the fuel salt. Furthermore, It is possible to optimize these costs for a given poison fraction by the appropriate choices of T1 aad Tld. l l h l s optimization was made in this study.

- 45 5.2 Solution of Poison Fraction Equation ’

The total fission product poison fraction was conveniently calculated using; FT-8 and PF-9 codes for the ORACLE &ich solved 7. Detailed knowledge of cross sections as a f’unction of energy for the individual fission products was not available; however, reasonably reliable thermal cross sections are known. It was necessary therefore to relate fission product absorptions to absorptions in another element for which more extensive cross section data are available. Carbon was chosen for the reference element.

In Eq. 7 all of the terms

From previous GNU or Cornpone calculations a reaction rate coefficient, Cc, for carbon can be computed as the quotient of total carbon absorptions at all energies in a region and the homogenized concentration of carbon atoms (see Section 4.4). Using this quantity an effective thermal flux can be computed as

4$-F=

are known except the term

Ft cc th Qc V/DC

60.

(81

in which : u is the thermal microscopic absorption cross section for carbon asd Dc is its thermal disadvantage factor. The other quantities were defined above in E q . 5. * absorbers, +it is only If it is desired to treat fission product6 necessary to multiply both sides o f Eq. 8 by the the& absorption cross section, u?, to obtain the absorpticm rate. The so obtained nay be On the other hand, a more used in Eg. 7 in camputlng the poison *action. pessimistic but more realistic comgutation ude the resonance absorptions and I n same msnner adjust the th os6 sections to reflect %hese resonances. An effectifre‘cro~ dlculated for each fission product by includhgthe res in the following manner:

4gf v y

eff =i

(9)

- 46 -

where, 2

= resonance integral for 19th nuclide, em

(RI)I

zf 6Ct

-1 = macroscopic fission cross section in reactor, cm

-1 slowing down power in reactor, cm

= fraction of total fissions occurring at thenaal energy,

MzPiber neutrons born per fission. are computed by the GMU code for the IBM-704.

Both sides of

4.8 can be multiplied by o'iff from Eq. 9 to obtaln t-\

eff ,eff 'th 1

When the subscripts 1 and 2 axe inserted to denote core region and end blanket region respectively, two eqressions are obtained for insertion as the W t e r m s of EQ. 7. These are eff eff @th,l ui

(u1

which is substituted for the term O1 , :@

which is substituted forathe term O2v;

The solution of E q . The value of

7 revised by 4 s .

and

11 and 12 is the desired poison fraction.

for this reactor was 0.6013.

-- 47 *

Values of resonance integral.6 have not been reported ResonanceIntegrals. for all. 44 nuclides of Table 5.1. The ones reported by Rephew were used, and, for the unavailable values, asstied or calculated values were.used. WhenELcalculation was made, the method for tiflnite'dilutlon described by Dresner17 tf+s used. 5.3

Fission Products Included in Poison I&action Calculation

The fission pimducts’used in the ‘&oison fraction cs3culations were those recommendedby Rurch, Campbe& and Weeren.‘3 Forty-four nuclides that would make an appreciable contributlo~ to the poisoning were chosen; these are listed In Table 5il. The Isotopes of xenon are not included in this tabulation because the poisoning from xenon (prlmirlly Xe-135) Is so large that it is treated separately, and a special processing method (gas sparging) must be employed to bring this value . within tolerable limits. Hence the poison fractlok calculated by Eq. 7 for the fission products In Table 5.1 ekludes any xenon contribution.

The kk.fisslon products are divided Into three groups which classify the elements more or less according to their ch&cal behavior in the system. The first group contains the &&ls that are noble relattve to nickel and might be expected to be reduced and plate out on the uaU.6 of the system. Also included in this’ group ,are the iodines and ,bromine, that are probably removable by gas sparging and hence may behave like xenkn. The noble metals and the halogens are treated as If they are removed from the fuel solution on a very fast cycle and thus contribute little to the poison fractAop. The second group contal& the rarerearths that are removed by precipitation In the RF-dissolution process and are thereby controlled by the fuel stream cycle time. ‘This Is the time refeked to above as T1. The ,. third group contctins the alkali and alkalipe earth Srnetslsthat are soluble In the, RF’-dissolution process and are removed by discarding the fuel salt on a specified cycle. This’ cycle, time is iaetitifiea above as Tla. '. and Effective Yield ~. 5.4 Gas:&dsg ,’ Fission product nuclldes which’are daughters of gaseous precursors will have ,..I “effective” yields that are snaller than their actual fission yield because the ‘gas sparging operation removes a portion of the parent atoms before decay. The fraction of gaseous nuclldes of a particular species which undergo decay before being sparged is I

-48-

Table

Nuclide

5.1.

Fission Product Nuclides Included I n Poison Fraction CalcuJ-ations

Thermal

Atoms removed by plating on

m-103

Ru-102 RU-104

Ob7

0.064

1000 101

0.05

13.1

0.042

O.Ol8

26 .? 15.8 6.3 25

2 1.2

M 0 n

0.2

1-13

600

O.WWO-7

11 6

stable

1-127

2.6 4 1.5

Br-81

*-93 &-9l

0.062

Mo-100

1-129

0.029

13.4 2.46

&=io1

or by gas sparglng

stable

Mo-95

wall6

0,065 0.029 0.a 0.0025

n w

o.oCu3

.6 6 6 x 1 0 ~ ~

0.063 0.059

12.2

. Atoms removed by precipitation in HF-dissolution W-lZ

Gd-E~-155(~) Sm-149 Sm-Eil-1~

Ell-153 Ed-143 sm-152 Rm-147 Nd-145 Pr -141 Nd-146

o.162d06

0.OOOl

.7 d o 5 .5 ~ l 0 5

7000 400 290 1-50 60 52

stable tu 0

o.845xlo4 stable

9.8

8.8 4 08 3 1.8 1.4 0.6 2.9

0.0003 0.007 0.0033 0.0~~-3

1326 0.2lgxlo

33.15 1512

0.oy

37.4

0.0021 0.015

2850 20% 30

0.029 0.0% 0.022

0.38 Om

16 lO(C) 16.1

- 49 t

Table 5.1.

Nuclide

Thermal cross

Section (barns )

- (Continued)

Decay Constant (sec -1)

Atoms removed by fuel 0.25x1d stable 130 O,149xlOa6 84 stable n 30 H

15 2 2

0 ~32X10-9

1 0.7 0*3

stable w

Yield

Resonaac (barns)

41.9 193 1396

198

375 347

10 c) 0.0021

8.6

(a) Considered together because cross sections and/or yields are about the same. (b) Yields are W u s t e d t o reflect gas sparging of gaseous precursors on a 6-minute cycle. (c)

Assumed value

available

of resonance integral since no data for calculating

(a) Esccelpt as indicated by footnote (e), values are From Hephew (reference 7 ) or calculated by method of Dresner (reference 17).

'decay 'decay 4- 'sparge

where the terms d e s i a a t e the decay r a t e and the sparge rate. yield then becomes

The effective

Effective yield

re:

(actual yield)

"decay 'decay + 'sparge

(13 1

For example consider Sr-89, a daughter of I@-89, under conditions for which the average spa.rging time of the fuel stream is siX minutes.

Effective yield of Sr-89

I n t h i s example the effec ,ive yiela of Y-89 would be the same. Where applicable, effective yields based on a six-minute sparge cycle were used in poison fraction calculations i n t h i s study.

5.5 Fission Products

a s l/v Absorbers

A series of calculations was made using the poison fraction code for the

ORACLE t o establish the poism fractions associated with a large number of combinations of fuel stream cycle time, !li,and fuel salt discard time, Tla. !%e i n i t i a l caLculations were performed considering the fission products t o be l/v absorbers, and the results are plotted i n Fig. 5.1. %e curves represent the solutions of Eqs. 7, 11, and: 12 i n which the resonance integral term, has been amitted. Values along the abscissa o f t h e curves have been divided by eta so that the poison fraction is-expressed a s fissick product absorptions per neutron born.

5.6 Fission Product Resonance Absorptions Incluiied in Poison Fraction Calculations A second s e t of curves, Hg. 5.2, was constructed fromthe solutions of

4 s . 7, ll, and 12 t o reflect the influence of fission product resonance absorptions on the poison fraction. Resonance integrals of the Individud. fission products given i n Table 5.1 were used. A t the sane values of Tl and Tla the

/..

-s-

effect of including the resonance absorptions I s t o appreciably increase the poison f'ractlon over Its value when the fission products were consldercd t o be l/v absorbers. A comrparison of W S . 5.1 an8 5.2 shows that for camparable cycle times the inclusion of resonance absorptions increases the poison fraction by a factor of 2.5 3.

Values dong the abscissa of Fig. 5.2 have also been aivlaed by eta in order t o express the poison f ' r a d l a on a "per neutron born" basis.

5.7 Use of Figs. 5.1 and 5.2

3.1 and 5.2 were used in optimizing the fuel cycle cost at a chosen poison fraction. Along a line of ccmstant poison fraction in these figures Mgures

several compatible values of TI an8 Tla were chosen, and the tdal f i e l cycle Influences the cost was calculate& for each pair of values. The cycle time, fuel cycle cost thrcwgh the capital investment i n the processing plant; the fitel salt discard cycle t i m e , reflects the replacement cost o f t h e fuel carrier. 'ithe calculated fuel cycle costs were plotted as a Aznction of the fuel salt discard cycle time, (Section 6.2.1) and the optinnlm cost and corresponding cycle times were determined.

Ta,

-id t

- 55 -

-56-

The equdlibrium reactor was studied t o determine the effects of variations in certain reactor characteristics on the nuclear perfoxmace and economics of

the system,

The investigated paraneters were:

1. poison fraction i n fuel stream 2. f e r t i l e stream cycle time 3. f e r t i l e strean volume 4. value of resonagce integral of Pa-233

5. v a u e of epithennal fission cros6 section 6. addition of &Fk t o stabilize fiel salt

U-233

The last three items perhaps are not rlghtly classified as parmeters since they are not independent characteristics. However in the cases of the resonance integra and the epithemal fission cross section, the ranges of uncertainty in measured values are supflciently broad t o have significant effects 011 reactor performance. Item 6 was introduced because recent fuel s a t studies have indicated a need for ZrFq t o inhibit oxide precipitation of Azel atom.

The two major parameters i n this study were the &el stream paison fraction and the f e r t i l e stream cycle time. The remaining four items were examined for variations i n these two major parmeters. The studies were made on the equilibrium state of the reactor described i n Section 2.0 using the ERC-5 code” for the IBM-704. Homer, i n making the calculations for s e v e r d values of epithermal fission cross sections of U-233 (Item 5 ) , it vas necessary t o establish new 8 c r i t i c a l conditions for the reactor u s i n g the GmU code before the ERC-5 calculations could be made. I

Reactor properties that were held constant during t h e parametric study were the f’uel stream volume (530.2 f t 3 ) and station power (2364 Mwt). The effective carbon-to-urmium r a t i o was calculated for each equilibrium reactor and varied only slightly fram case t o case because of slightly different equilibrium conditions. ’The r a q e of C:U ratios for a U of the calculations was 5020 t o 5230. These values are the actual C:U r a t i o divided by the t h d energy disadvantage factor (0.879) for carbon i n the core, !Phe variation from case t o case was caused by small changes i n the volume fraction of f e r t i l e stream in the core for changes in Arel stream poison fraction and f e r t i l e stream cycle time.

-57-

I n the equilibrium calculations it was assumed that the absorption and fissYon rate coefficients-were not appreciably affected by s d l changes in concentrations (- 109) in important elements such as U-233, U-235 and fissium, and by much larger changes (- lOOO$) in minor elements such as U-234, Pa-233, %e, etc.. In cases in which the equilibrium calculations significantly changed the concentration t o the extent that the reaction r a t e coefficients might no longer apply, it was.necessaryto repeat the Cornpone unit c e l l finite reactor calcul a t b n s with new concentrations t o develop new sets of reaction r a t e coefficients (see Fig. 4.1). Several items i n the neutron balance were specified for all of the parametric studies. These were the neutron losses t o corrosion products, delayed neutrons,

leakage and fuel processing. !l%e values adopted for these quantities were re= spectively 0.0008, 0.0043, 0.0016, and 0.0022 neutrons l o s t r neutron absorbed in fuel. Corrosion product losses were estimated f'rom the equilibrium concentration of corrosion products of INoR-8. Delayed neutrons were calculated by the method of WeJker?' Leakage losses were estimated frm design considerations; it was f e l t that t h i s number wauld be small because of the small &moullt of fissioning in the blanket. Fuel process* losses were discussed above in Section 3.6.

6.1 , Results of Equilibrium Reactor Calculations Pertinent characteristics wereamade are l i s t e d in Table

e MSBR on which equilibrium cd.culations

Representative results of the equilibrium reactor calculations are>given in Tables 6.2 and 6.3 in the Appendix. mese results include the equilibrium atomic coicentrations of mador isotope the fuel and fertile-streams, a neutron balance Tor he contribution of inaividual items t o the

fuel cycle cost, t h Fraction of fertile stream in .the core for 'the just critical'reactor, and several Item of less significance such as the f'ractions of" each stream sold as product a t h e fraction of t o t a l fissions occurring in the f e r t i l e stream.

-58\

Table 6.1.

me-

.-.

Characteristics of M8BR

M ors in station mennodynamic efficiency +mer,

stream volume per reactor, ft Fertile stream volume per reac Valume fraction fuel stream i Volume fractian fuel stream in end blaalret Volume fraction fertile stream in side blanket Volume fraction fertile stream in end blanket Volume fraction graphite in side Xlanket Volume frclction graghite in end blanket

m@lstream holdup time in reprocessing,

days

eam holdup time in reprocess%ng, days

2364 2 0.823 265.1

3000

0.16 0.16 0.90

0.74 0.10 0010 1 1

llhe option used t o achieve c r i t i c a l i t y in the EEiC-5 CCiLculations was variation of the volume fraction of f e r t i l e stream i n t h e core. Since the

volume fraction of f'uel stream was Flxed at 0.16, the addifion or subtraction of fertile stream vas made a t the expense of removing OT adding moderator. Consequently CtU r a t i o in the core varied slightly from case t o case. However these small variations in CtU r a t i o Bid not significantly affect the neutron spectntm.

!&e fiel stream cycle times reported in Tables 6.2 asd 6.3 are the optimized cycle times. Each time has been so chosen that . r e f l e c t s the most economic rate for the chemical processing for the chosen values of poison fraction and fertile stream cycle tfme. We fuel s a t discard cycle time has also been optwzea

In these two tables the results are for reactor systems in which the flssion product res&ce absorptions were included i n the poi& fraction caclculation. Results of calculations in which fission products were assumed t o be l / v absorbers are not included because these are j u s t optimistic s p e c i d cases of the resonance

cy,

- 59 -

b i ;

ansorpyoncalculatlolls.

For the resonance absorption cases the fuel stream cycle time and fueLsalt discard times are respectively in the range612-8k days and 145~15p'days for poison fractions from 0.02 - -0.06> neutrons absorbed in fassion products per-neutron absorbed In fuel. For the l/v absorption cases, the corresponding cycle times are.ir, the ranges 12.5-735 days and 400-8100 days for poison fractions from O.Oll - 0,065. The procedure for determining the optimum fuel cycle times was referred to above in Sectlon 5.7 and Is discussed further in Section 6.2. 6.1.2

Meutron Balance

ResonanceAbsorption Cases; A'portion

of each of Tables 6.2 and 6.3 shows the distribution of neutron absorptlons in the reactor. Ekgnining the neutron balance of-!&ble 6.; fok increasing poison fraetioti, one finds.that losses to protactuuk f6J.l about lO$ partly because the P6+-233concentration decreases about 7$'due to decreasing breeding ratio. The decrease in losses to Pa is also partly due to a decrease In &he volume~fraction of â&#x20AC;&#x2DC;fertile stream $n the core by .about 74. This Is significant because about 609 of the captures in k-233 occurs in the core. .A similar effect Is observed for the'tabulatlons of Table 6.3 for other fertile stream cycle times. Concurrently neutron losses to samarium and other fission products increase by 'about O.&f5 neutrons. This Is more than half of the breeding gain and results in tire than one-half the production of excess fuel. e The longer cycle times at the higher poison fractions effect less~pur&ng of higher uraniwn isotopes. !L!he consequent b&d-up 'of U-235 causes -a decrease.in the .meanq of the-system by about O&Ok. Neutron losses to u-236 .and Up-237 increase'about 1.3.fold asd lo-fold; respectikly9 or by about 0.006 and 06002 neutrons.' â&#x20AC;&#x2DC;.

L/v iisorption Cases. The . neutrcm balances for these calculations are not presented in this memozxndum because they are of limited interest. Since the raxig~ of pOison fraction6 (0,Oll A d.065) is gr&ter~than that covered in,the resonance absorption cases* more variations might be expected.in the elementrr3 absorptions. ,The es&lanatlon.of the -.trends, however, is-the same as given above. For example,.,Pa-233,absorptions decrease about 14s over the-range of poison fractions because it6 concentration decrease6 about 7$, an&the volume fraction of fertile stream in the core decreases about 13s. . _

,--

Losses t o samasium and other fission products increases by 0 . 0 9 neutrons consuming about 708 of the breeding gain. Higher isotopes of uranium build up because of slower processing rates a t the higher poison *actions with the accompanying decrease (- 0.005) in the mean wCLue of 1). Neutron absorptions by U-236 and Ep-237 increase by about 0.004 and

em Inventory_ ory of fissionable materials, which

des U-233, U-239, and -0233, for the equilibrium reactors is presented i n Fig. 6.1; a detailed breakdown of the inventory i s given in !tables 6.2 6.3. For both resonance absorption asd l/v absorption cases the inventory does not,change very fast wi& poison fraction over most of the range of poison fractions. However, a t low values Qf the poison fraction a sharp upturn in the inventory is observed. This occurs because of the increased holdup i n . t h e chemical processing plant a t these fast processing rates. The proccessing r a t e for the resonance absorption cases a t a poison fraction of 0.02 is about equal t o the r a t e for the l/v absorption cases at a poison frac%ion of O.(u1.

The largest effect on fissionable inventory is observed for the vetriation in f e r t i l e stream cycle time. As the f e r t i l e stream cycle time increases from 35-20 days, the t o t a l fissionable inventory increases about 505, or fram around 860 t o 1240 kg. The increase is attributed almost entirely t o increase of U-233 inventory in t h e f e r t i l e stream which rises over >fold. Uranium-233 inventory in the Fuel stream decreases about 6$; concurrently the U-235 inventory increases about 746. Increased fissloniag i n the blanket act the longer cycle times causes the c r i t i c a l mass of U-233 in the core t o decrease. However since %hebreeding gain a l s o decreases for increasing f e r t i l e stream cycle t i n e at constant poison fâ&#x20AC;&#x2122;raction, the purge r a t e of U-233 becomes smaller because l e s s U-239 is routed t o sales; hence the U-235 inventory in the core builds up. %e thorium inventory for this series of calculations was maintained constant a t 270 t o m e s . Protactinium-233 inventory is not very sensitive t o changes in poison fraction or f e r t i l e stream cycle time. Since Pa-233 is not removed from the system in the fluoride volatility process, it builds up u n t i l i t s decay r a t e is exactly equal t o the U-233 production rate. Therefore the Pa-233 inventory will change in direct proportion t o the breeding ratio. For t h i s system the inventory is in the raage 100-110 kg.

6i1.4

62 -

Fuel Cycle Cost

A breakdown'of the fuel cycle costs for representative cases is given in Tables 6.2 ani 6,3. These costs were calculated using the basic cost data given in Table 3.1. The largest single contribution to the f'uel cycle cost is the charge for the fuel processing plant which contributes up to 40%of the total cost. At the faster blanket processing rates, the fertile stream processing plant cost also becomesa significant part of the total cost; at a 350day blanket cycle -time about 308 of the cost may derive from fertile stresm processingâ&#x20AC;&#x2DC; Total inventory charges on fissionable materials, thorium? fuel carrier and thorium carrier account for 40 - 55s of the fuel cycle cost. Individually,' the thorium carrier (- 200 tonnes) contributes most to the inventory charges, from 15.0 20s of the fuel cycle cost; fissionable inventory contributes about 7 .- 13%; thorium inventory contributes 12 i 16s; and fuel carrier inventory contributes the least, only from1.5 - 2.55.. Thorium amortization and thorium carrier replacement charges, which are amortized at 2,6$, are not an appreciable portion of the fuel cycle cost, being only 2 - 45 of the total. On the other hand, fuel carrier replacement charges becomea'slgnificant factor especially at the lower values of poison fraction because of the high salt discard rate. At very high poison fractions (0.065)~ this contribution is only about 4$ of the total fuel cycle cost; whereas, at low poison fractions as much as 25%of the cost is due to salt discard. 3reeding credit is an item of the fuel cycle cost that is directly proportional to the breeding gain and fissionable inventory, The high fuel yield reactors (6.8$/year) have breeding credits of about 0.13 mills/kwhr; in the very highly poisoned reactors (fuel yield Gl$/year,), the breeding credit is only about 0.04 mills/kwhr.' Although allowing the fertile stream cycle time to increase from 35 to.200 days lowers the breeding credit through increasing the fissionable inventory, the effect is not so pronounced as that caused by allowing the poison fraction to increase. Included A series of equilibrium reactor calculations was made for a range'of fuel stream poison fractions from 0.02 - 0.065 neutrons absorbed in fission products '.

63 -

per neutron absorbed in fuel. This range of poison fractiotis was applied to fetiile stream cycle time parameters of 35, sr 75, 100, 17 and 200 days. The rat?ge of p&son fractions was established after a few preliminary cslculatlons toilnclude reactors with quite favorable'breeding potential as well as those that are approximately 'hold-your-own" systems. Poison fractions lower than 0.02 were not considered because the required fast processing rates result In high fuel cycle costs wlthout appreciable increase in breeding gain. At a poison fraction of O.O65,.the MS& shows 'a'small, positive breeding gain; at higher poison fâ&#x20AC;&#x2122;ractlons it Is doubtful if the system will breed. These two parameters could very conveniently be treated in the EBC-5 code since they are items of input data. The poison fkactlon as such does not app& in EEX!&>input; however, the desired poisoning effect can be obtained by using fictitious fission product concentrations and fictitious reaction rate coefficients. The net effect of additional~polsons is to decrease the breeding gain aud corrspending breeding credit. For each comb&nation of poison fkactlon and fertile stream cycle time the code calculated equilibrium atomic concentrations, inventories, neutron absorptiona by elements, thorium concentration in the core aud processing rates. 6.2.1

'&el

Cycle Cost Optimization

The fuel cycle costs were. optimized for each value of the fuel stream poison fraction and each value of the fertile stream cycle time. For each,comb&&ion of these two parameters the f'kael stream processing cycle which ga& the lowest total fuel cycle cost was determined.' The &rocedure is discussed be$ow. l!kel Yield Versus foison.Fraction. fiti' yields caiculated for the equilibrium reactok were plotted as a function of.the poison fraction (Fig. 6.2). The : newly linear relationship'lndicates that the fuel yield is Inversely proportional to the poison f'raction. The plots begin to curve in the region of low poison fractloi because the required fast chemical processing rates begin sig- . nificantly to increase the fissionable inventory through holdup in fuel proceising . The result is a lowering-of the fuel yield.â&#x20AC;&#x2DC; The effect of i,ncreasing fertile stream cycle time is to decrease the Ike1 yield for a given poison iPr;action.' This occurs because of Increased fissioning in the fertile stream and the accompanying Increase in fission product co&e&ration plus an inventory increase.

UNCLA 9 9 I Fl Eb ORNL- LR- DWG ~ 1 6 4 9

Fuel Salt Discard Time a s a Function of Fuel Cycle Cost. As discussed i n Section 5.0, each value o f t h e fuel stream poison fraction can be attained by operating the reactor a t several values of fuel stream cycle time and fuel s a t discard cycle time, and there i s some canbination of these times for which the fuel cycle cost i s a minimum. For each selected value of the fuel stream poison f'raction, several pairs of compatible values of these two cycle times were chosen frm Fig. 5.2, and the total. fuel cycle cost was calculated for eaah pair of values. me f u e l cycle time determined the capital investment in the processing plant3 the fuel salt discard cycle time determined the replacement charges for the fud s a t . A p l o t of f u e l salt discard cycle time versus fuel cycle cost at constant poison fraction gave the curves exhibited in Fig. 6.3. The minimum of each curve represehts the point of most economic operation; the corresponding fbel cycle cost and Azel salt.discard cycle time are read directly. When the optimum fuel salt discard cycle time i s entered into Fig. 5.2, the opkimum fuel stream cycle time i s found. Optimum values of t h e fuel cycle cost,

stream cycle time, and fuel salt

discard cycle time are given i n Tables 6.2 and 6.3 6.2.2

Economic Performance

The optimized *el. cycle costs obtained by the above procedure have been

plotted a s a function of the fuel yield in Fig. 6.4. The curves have been calculated for a plant operating 80%of the time. The most favorable fuel yields are obtained at the shortest fertile stream,cycle 3 however, the corresponding am must be processed a t a *el cycle costs are high, Likewise cated by the lower values of the poison fraction. relatively f a s t rate. as a f'uel cycle cost of around Fuel yields of the order of 7$/yr can be attained

1.7

mili~/lrwhr.

cycle costs of 0.75 I n the lower range of fuel yields elds of 1 t o 2$/yr. These conditions t o 0,80 mills/kwhr can be attained at require fertile stream cycle times of 150 200 days. The curves could have been extended in the lower regions of fuel yields by performing calculations at higher ionj however, uncertainties in basic data, e.g., cross values of the poison f sections and resonance i n t e g r a s , would lend daubt a s t o whether such a system would have a positive breeding gain,

L j

- 66 -

I .

-68-

In t h i s series of calculations fuel stream cycle times ranged from 12 84 om 145 1.550 days for the case days, and fuel salt discard cycle times ranged t h a t f a l l on the envelope of the curves.

The dashed envelope curve has been drawn t o indicate t h e estimated maximum

It might be possible t o extend the envelope of the family of curves out t h i s far by modifications i n the C:U r a t i o in the core and by optimizing the f e r t i l e stream cycle time. These refinements t o the calculations were not made i n t h i s study; nevertheless, it is believed that the chosen C:U r a t i o i s near the optimum.

. performance of t h i s reactor.

e t i m e , the fuel cycle cost drops Along a l i n e of constant f e r t i l e stream rather sharply from i t s maximum value principally because of decreased charges on the fuel stream processix plant at the longer fuel stream cycle times and lower fie1 salt replacement charges for the longer salt discard cycle time. During t h i s i n i t i a l drop i n fuel cycle cost, the breeding credit is a l s o decreasing, but the i n i t i a l loss of breeding credit is far overridden by the savings on the processing plant and fie1 salt replacement mentioned above. Consequently the i n i t i a l drop in fuel yield i s not a s fast as for the fuel cycle cost. Eventually, though, as the processing time becomes long (increasing poison fraction) the savings on the processing plant and fuel salt discard are not so effective in lowering t h e fuel cycle cost and the rate of decline decreases. Meanwhile decreasing breeding gain accelerates the loss i n fuel yield. Ultimately at higher poison fractions than shown on the graph, a complete loss of breeding gain would necessitate rising because fuel would have t o be purchased. ne of constant poison fraction i n Fig. 6.4, the fuel cycle cost i s affected principally by changes i n capital charges on the fertile stream processing plant and in fissionable inventory i n the fertile stream. The contribution of.the fertile stream processing plant t o the fie1 cycle cost decreases with increasing cycle time while the inventory charges increase. I n i t i a l l y the savings essing plant overweigh the increased inventory charges resulting i n a net lowering o f t h e t o t a l fuel cycle cost, A t the long cycle times, however, the inventory charges overbalance t h e lower plant costs and the fie1 cycle cost i n i m and begins t o rise. ecrease in fie1 yield along a l i n e of constant poison fraction is not large and is due primarily t o increasing inventory of fissionable material i n the There is also the adverse effect t h a t higher U-235 concentrations a t t h e

bsi

longer fertile stream cycle times have on the mean value of q -.an effect that lowers the fuel yield through lower breeding gain.

6.3 Poison Fraction Studies in w h i c h Fission Products were Considered t o be l/v Absorbers

A series of equilibrium reactor calculations was made for a range of fuel

stream poison fractions fram 0 . 0 l l 0.065 neutrons absorbed i n fission products per neutron absorbed I n fuel. This range af poison fractions was applied t o fertile stream cycle time parameters of 35, 50, 75, 100, 150, and 200 days and i s broader than that considered in Section 6.2 for the resonance absorption cases, When fission products are treated a s 1/v absorbers, t h e poisoning effect i s not a s great as when the resonance absorptions are included, and it i s therefore possible t o extend the range of calculations t o lower poison fractions before intolerably short fiel stream cycle times axe reached. The equilibrium calculations were made usthe ERC-5 code for the IBM-704 by varying the f e r t i l e stream cycle time and by using fictitious fissium concentrations and reaction r a t e coefficients a s mentioned in Section 6.2. These calculations were performed at the beginning of t h i s thorium breeder study before a campilation of resonance integral data became available, &d it appeared t h a t treating t h e fission products as 1/v absorbers was the best approach t o the problem. merefore, the results reported below should be considered as an optimistic upper limit t o the fuel Id and an optimistic lower l i m i t t o the fuel. cycle cost.

6.3.1

~ ~ a n o m Performance ic

The optimized fuel cycle costs obtained. by the procedure a r e plotted as a f’unction of the fuel yield in Fig. 6.5. These fuel cycle costs and fuel yields should be regarded a6 rather optimistic values since considerable neutron economy resulted from the assumption that the fission products behaved. a s l/v absorbers. Consequently it is believed that these curves represent alower bound t o the MSBR fuel cycle costs3 more r e a l i s t i c performance i s t h a t represented by Fig. 6.4 in which t h e best available resonance absorption data were

4-/

For the conditions of Fig. 6.5, fuel yields a s high as about 8$/yr at a f’uel cycle cost of about 1.3 mills/kwhr were obtained; a mininnun fuel cycle cost of about 0.66 mills/kwhr was obtained a t about 4$/yr fie1 yield. The ’

- 70

UNCLASSl F IED ORNL-LR-DWB 68652

curves rise linearly after attaining their minima-because of the.lnfluence of the constaut charge for the batch fuel stream'processlng plant. Batch operation becomeseffective in the region of,fuel yields of 5 - B$/yr:at poison fractions of 0.025-- 0.03. The behavior of the curves for'variations iu'polson *a&ion and fertile stream &y&le'time can be explained by the'sarne commentsmade above 'in Section 6.2.2 and w-@ot.be repeated here,: . : :-../ 6.4 Effect on Reactor &rfbrmance of Varying; Thorium Inventory In order to study the breedingperformance of the MSRRover a wide range of operating'coiditions, the thori& inventory was '. varied in the range 100 - 400 tonnes.. fn a few representative calculations.for vhich the fertile stream cycle time was, 35 days, the fuel -.Istre& cycle t&&.was 20 days, and the fuel sslt discard time i was lm days. Tpe -. thorium inventory Xas varied by adjusting the fertile stream volume In the.ra&e 2CCC-,~9COCft3 per St&Ion; .ThiS particular series of,calculatlons was made at an early stage of the study, and the combination of cycle times is not optimum with respect to fuel cycle costs eat the various1 fuel yields., However; the dependencyof fuel 'yield and cost-on thorium inventory is only weakly affected, if at all;.by 'choice of cycle times. Tfierefotie the behavior exhibited by the selected~cases is typical &d may be used as a guide in selecting thorium inventories. !Die &qoFtant results of these calculations are given in Iable.6.4. : I_ -- L‘. : i. ,: . _i . . . Table 6,4. Dependency of Fuel Yield and Fuel Cycle Coston , Thoktuin Invehtory in a Molten SaltRkeeder Reactor ' _ :, '. .I , ._ .,. : .-. -!liw.riuul IzkIltory ' --: Relative'I&l Yield :- Relative Fuel Cycle Cost (tonnes oloo) ~..; , ,"l,o 1.00 140 1.2 ..I" ,. i.03. :-180 . 1;3 l.op‘ 2-p-J) ‘1.4 .' . . ', . :. <'_ 1.20 40@ i ~ -1.4 :- -. : ', _, 1.39 ,. -

- 72 -

As the thorium.inventory.increases, losses to.Pa-233 decrease, and there are gains in respect to mean eta and.&236 absorptions. Breeding gain increases, but at a decreasing rate. Meanwhile fuel inventory in the fertile stream rises. As a result, fuel yield rises rapidly at first, and tnen more slowly as the influence of increasing inventory overrides that of breeding gain. The cost increases steadily, however, being driven upward by increased charges for thorium anduranium. The-fuel yield reaches a point of negligible improvement at 2i'O : tonnes of thorium, and this thorium inventory was used for further studies reported above in Sections 6.1. - 6.3. One hundred tonnes of thorium is not sufficient to fill the blanket of the In the correreactor usedâ&#x20AC;&#x2DC;in this study when the blanket thickness is 3 ft. sponding c&culations, no adjustment was made for the greater leakage that would result from a thinner blanket. Thus' for the case in the above tabuiation, the fuel

yield

should be less

and the tie1

cycle

cost higher

th&

cordingly the .140-tonne case was selected as ti representative for further study.

$.lculated.

Ac-

lo%thorium

case

For the 1400tonne :thorium fertile ,stream, a series of calculations was made to optimize the fuel cycle cost and the fuel yield at a representative fertile. _ stream cycle time (50 days). A comparison between these results and those of the corresponding 270-tonne thorium case are presented in Table 6.5. The results of Table 6.5 were obtained for optimized fuel stream cycle times and fuel salt discard times, and in all calculations the resonance absorptions of fission products were included in the poison'fraction calculations. The two cycle times are longest for the low fuel yields and shortest for the high yield cases. Although some slight trend with fuel stream cycle time is observed,,the rule can be formulated that doubling the thorium inventory adds _ about l.g$/yr to the fuel yield and about 0.2 mills/k&r to the fuel cycle cost regardless of the fuel stream cycle times. The performance of the MSBRcontaining 140 tonnes and 270 tonnes of thorium is graphed in,Fig. 6.6. The solid curve which is drawn through the calculated points is the envelope curve of Fig. 6.4. The dashed curve is an estimated curve, based .on the few 140~tonne thorium cases, for-the maximumperformance of the MSBR.I at this low thorium inventory. The solid outer curve was then drati to indicate the

* z

- 73 6m+

Table 6.5.

Molten Salt Breeder Reactor

Dependency of Fuel Y i e l d and Fuel Cycle Cost on Thorium Inventory w i t h 5O-Day Fertile Stream Processing and mimized f i e 1 Stream Processing Cycle Times Thorium fnventory, tonnes 270

140

270

Fuel cycle cost, m i l l s / k ~ h r

FM. Yield, $/yr

Diff

0.3 0.7 1.7

1.7

0.83

1.8 1.9 1.9 1.9

2;o

2i5 3 i6 446 5i6 6.6

2 -7

2.7 4 .7

1.9

140

Mff 0.22

0.84

0.63 0.63

0 -87

.0.66

0.21

0*95

0.20

0.95

0.75 0.75

1.n

1.37

0.20

0.21

0.20

estimated limit of m.ximum performance of the MSBR when the thorium Inventory is optimized with respect t o fuel yield. !the outer curve a l s o assumes that a slight improvement in the reactor can be found by a slight variation in the C t U ratio. The CrU r a t l o (- 5000) in t h i s reactor was not optimized, but t h i s value is believed t o be near the optlmuzu. me 180-tonne thorium case is a l s o plotted in figs. 6.7 and 6.8 for optimized.

Aiel cycle times and for a raage df poison fractions fram 0 . a

6i3 =feet of Value of q-233 on

- 0.065.

Performaace

Uncertainty in the measured values of the e p i t h d fission cross sections of t?-233 can cause considerable variation in the calculated performance o f t h e MSBR, depenaing on the s e t of cross section values that is used. This is the case because approximately 304 of t o t a l fissions occurs at epithermal energies. Reported epithermal values of 3.233 apparently agree within about 10%of an average or "recmended" set of values 4

- 74 -

UNCLASSIFIED ORNL-LR-OW0 69090

1.5

1. 4

1.3

c v)

0 0

1.0

w J

0.9

J W 3

0.a

0.7

0.6

0.5

FUEL YIELD (%/yr)

Fig. 6.6 Performance of a Molten Salt Breeder Reactor with Varying Thorium Inventory

. ..

Several calculations representative op&ating condition were made using the GNU and EEC-5 programs for the IBML704 to determine the effect of these q variations. The plots of Figs. 6.7 and 6.8 show the effect for cases in which fission product resonance absorptions were included and for cases considering fission products to be l/v absorbers. There is considerable variation I n the fie1 yield, a s measured by the horizontal distance between corresponding points on the curves, between the recommended curve and the high and low epitXermaJ. eta curves. The deviation froxi the recommended value is f 2.5 to

39&/yr in

Fuel yield.

In fact, using the m r e pessimistic values of q makes

it difficult to attain fuel yields of a s much a s 4$/yr even at fuel cycle costs as hlgh as 1.6 mills/kwhr.

On the other hand the choice of high or low epithermal q-233 does not have a strong influence on the fie1 cycle cost. This effect is measured by the verticgJ.difference between corresponding points on the three curves. %is difference Is approximately 0.06 mills/kwhr from the recommended eurYe. . The various vaues of q-233 are tabulated in Tclble 6.6. The column headed q(MTR) contains values f r o m experiments performed at the Materials Testing Reactor and are the values recommended by Nestor' for use in this study. The columns headed q(+ 10s) and q(- 104) contab the extreme values used to obtain the two curves of Figs. 6.7 and 6.8 for comparison with the reconrmended vaues. Values in all energy groups differ by 10s *om the MLX values except in group 3l. where the difference is only 5% and in groups 32 through 34 where the values were not changed. The other values headed q(RPI) are presented for camparison since these are recent data fr& -a study by Yeater.19 The RPI set of values are. not thought to be more reliable than the MTR values; however, In the energy range 30 ev 1 kevthe RPI &ata represent the only measurements that have been made.

6.6 Effect of Value of Pa-233 Resonance Integral on MSBR Performance A second nuclear constant

accurately is the value of the resonance integra e6 of 600, 900, and 1200 barns have been mentio r the value of this integral. As menti0ned.b Section 3.7, a a u e of 1200 b s chosen for this s However, in order to determine what effect a lower value of the resonance integral would'have on reactor performance, severaJ. calculations were made at representative operating conditions using a 900-barn resonance integral for camparison with the 1200-barn cases. m e resulting curyes f o r the reactor performance are plotted in Fig. 6.7 and 6.8. ._,

- 78 h

Ta;ble

Group

1 2

3 4 5 6 7 8 9 10

XI. 12 13 14 15 16 17 18 19 20 21

22 23 24 25 26

27 28 29 30

31 32

33 34

q (-10%)

q (MTR)

q (+lo$)

q(m1

3 .os

3.39

3-39

2.583 2;42l

2-87 2.69

3 *729 3 .I57

2.51 2.37

105

20259 2-133 2.052

4 10 104

2.025 2.m6

1 x 107 4 x 106 2 x 106

105 - 1 105 - 3

2 x lo6 1 x lo6

6.6. Group-Averaged Eta Values of U-233

mergy (ev)

4 x lo6

6 LO 105

3 x104 0.1 x 1 lo4 3 3 103 1 x 1 103 3

4.00 150 100

go 80

65 50

45 37 33 30 25 20 17

2.5

2.025

2.959 2.7a 2.607 2.508

2.51 2.37 20 2 8

2.24

2.475 2.464

2.25 2.24

2.25

2.475

2.25

2.28 2-25

-m

. -

150 loo

30 25 20

13.5 10

7.5

2.5 1.4

1.4

thermal

5.5 4

0.8 0.6

1.93 1.96 1.7 1.97 2.06

45 37 33

thermal

1.77

0.8 0.6 -

1.54 1.68

go 80 65

1.9 1.72 1.78

400

13.5 10 7.5 5.5 4

LO3

2 2.69

1.944 1.944 1.944 1.953 1a 9 5 3 2.032 1.863 1.962 1.845 1 .n9 2.12 2.29 2.28 2.28 \

2.16 2.16 2.16 2017 2.17 2.28 2.07 2.18 2.05 1.91 2.23 2.29 20 2 8 20 2 8

2 376 2 376 20376 2 387 2.387 2 2.277 2.398 2 0255 2.lol 2.342 2.29 2.28 2.28

1.9. 1.93

1.75 1.88 20 0 6 1.91 1.96 1.99

1.96 2-07 2.23 2.29 2.28 2.28

--_

- 79 The improvement that the lower a u e o f t h e resonance integral makes in tlie MSBR performance is hardly appreciable. The fuel yleld, measured by the horizontal difference between corresponding points, is increased by perhaps O.25$/year; the fuel cycle cost, measured by the vertical difference, is lowered by about 0.Ol mills/kwhr. lihe small effect on performance is understandable Men it is considered that losses t o Pa-233 account for only about 0.5s of the neutrons born per fuel absorption using the 1200-barn resonance integral. However of all the neutron absorptions in Pa-233, approximately 805 occur at epithermal energies. !Che comparative N - b a r n resonance integral calculations were performed using the ERC-3 code with adjusted values of the reaction r a t e coefficients, which are defined above in Section 4.4. The adjustment was made using output data from a Cornpone f i n i t e reactor calculation giving the absorptions in Pa-233 as a function of energy. The reaction r a t e coefficients were calculated separately for the thennal absorptions and the epithermaJ. absorptions, t h e epithermal value being decreased by the r a t i o of the resonance integrals, i.e., 900:1200. The two values for the coefficients were summed t o obtain the t o t a l reaction r a t e coefficient as shown by E q . 14.

reaction r a t e coefficient of Pa-233 neutrons absorbed by Fb-233 at thermal energy per neutron born in core neutrons absorbed by Pa-233 at epithermdl energies per neutron born i n core homogenized atomic concentration Pa-233 i n core, atoms/cm 3 A similar expression was used t o calculate C(Pa) for the blanket, and these

adjusted coefficients were used in t h e equilibrium reactor calculations. I

6.7 Effect on MSBR Performance of Adding ZrF4 t o Fuel Salt Recent developments in fuel technology for the Molten Salt Reactor Experiment (MSU3) have indicated t h a t f i e 1 s t a b i l i t y i s enhanced by the addition of nominal amounts of ZrF4 t o the fuel salt. Zirconium acts as a â&#x20AC;&#x153;getterâ&#x20AC;? for oxygen

preventing the union of oxygen and uranium which results in the precipitation of uranium oxides. The inclusion of zirconium, however, adds an a d d i t i o d neutron poison t o the system. The effect on breeding r a t i o and fuel yield of adding

5 mole $ ZrF4 t o

the

fiel salt was calculated for t h e ranges of values representative of the MSBR. The results are plotted i n Fig. 6.9. The curves show the per cent decrease i n breeding r a t i o and fuel yield resulting fYosn the addition of ZrF4 as a S'uncticm I

of these quantities for a reactor containing no ZrFd. As m i g h t be expected, the detrimental effect of the ZrF4 is more pronounced for the reactors t h a t have i n i t i a l l y poor breeding performance. In fact the steepness of the fuel yield curve a t fuel yields of the order of 2-3$/yr suggests that adding ZrF4 t o a low performance reactor can just about destroy i t s breeding potential. The &-containing salt used i n the calculations had the composition of fuel

solution proposed for the MSRE:

70-23-5-1-1 m o l e $ LiF-BeF&kF4-ThF&JF4.

I n order t o determine the effect of 5 m o l e $ ZrF4 on a representative performance curve, the results of Fig. 6.9 were applied t o equilibrium reactor calculations for a f e r t i l e stream cycle time of 50 days. The results were optimized according t o fuel strew cycle time and fuel salt discard time. The horizontal difference between corresponding points in Flg. 6.10 shows t h a t Zr decreases t h e f'uel yield about O.5$/yr; whereas, the f'uel cycle costg measured by the vertical difference, i s almost negligibly affected. The effect on Are1 cycle cost i s small because t h e effect shows up through the loss in breeding credit which i s not a large portion of t h e t o t a l -el cycle cost.

G f

- 81 .

UNCLASSIFIED

”

7.0 CO~CLUSIO~S The molten salt reactor offers considerable promise

a breeder i n the

Th-U cycle, The principal advantage of this system Over other breeding systems, which use thorium in the form o f t h e oxide or metal, i s in the simplified chemical processing method. The molten salt system i s able t o use t h e relatively si@e fluoride v o l a t i l i t y process plus the HF dissolution process for uranium recovery and decontamination; whereas, breeders which employ Th& or thorium m e t a l are, in the l i g h t of current technology, resigned t o the more complicated and expensive Thorex process. ,

The MSBB is capable of excess f i e l yields up t o 7$/year when operating 80s of the time. A t this hlgh yield, the Arel cycle cost i s about 1.65 mills/kwhr. A t lower fiel yields the fuel cycle cost is considerably improved, dropping t o perhaps 0.65 mills/kwhr for f i e 1 yields of l-2$/year. However, yields as low as t h i s constitute margiaal operation because uncertainties i n nuclear data introduce uncertainties of about a per cent i n fuel yield into the calculations. A t a fuel yield of 4$/yeax, the point a t which the income from sales just balances the annual charge on fissionable inventory, the fuel cycle cost i s approximately 0.9 mills/whr. 1

lZle largest contribution t o the fiel cycle cost i s made by the fuel stream processing plant which a c c k t s for about 41s of the cost at the high processing rates (high f u e l yield) and about 308 of the cost at the low processing rates. Another item t h a t makes a major contribution t o the cost at the high processing rates I s the fuel salt discard, accounting for sllghtly m r e than 208 of the total; howwer.at the lcrw processing; rates, salt &iscard accounts for only about

474. %e f e r t i l e stream processing plant cost amounts t o only 12-135 of the total. Thorium inventory for the 6000 f't 3 fertile stream amaunts t o 8.179 of the cost. Since the thorium inventory i s constant (270 tomes), i t s cost is a larger portion of the cost for those cases that have the most favorable fuel cycle costs. The same i s true for the thorium carrier which accounts for 10-22s of the fuel cycle cost over the range me1 yields from 7$/yr t o l$/p.

- 84

Fissionable inventory (including Pa-233) is only slightly affected by t h e processing r a t e of the fuel stream over most of the range of poison fractions studied. However, a t the fast fuel stream cycle times the holdup i n t h e chemicai processing plant begins t o contribute importantlyto t h e inventory. The principa factor increasing the inventory is the f e r t i l e stream cycle time. In going from a f e r t i l e stream cycle time of 35 days t o 200 days, the fissionable inventory increases f r o m about 840 t o 1280 kg. A t the same time, fertile stream fissions increase fâ&#x20AC;&#x2122;rom 1.346 t o 6.6$ of the t o t a l fissions. The fissionable inventory accounts for about 19% of the f i e l cycle cost at a fiel yield of 1.5$/yr and for about 446 at a yield of 7$/yr.

The breeding performance of the molten salt reactor i s especially sensitive t o t h e value assigned t o the epithennsl fission cross section of U-233 since about 305 of the fissions occur$ at epithermal energies. Equilibrium reactor

calculations for a representative s e t of operating conditions 3,ndicate that the

I ~

2.5 t o 3$/yr for variations of lo$ i n the value of the epitherma3 cross sections of U-233 fâ&#x20AC;&#x2122;rom the s e t used i n these calculations. fuel Yield m Y

V-Y

8s much as

The inclusion of 5 mole $ZrF4 i n the fuel salt t o enhance s t a b i l i t y decreases ,the fuel yield about O.s$/year; however, the concurrent Are1 cycle cost is neaigibly increased. llhe MSBR already suffers from having relatively high neutron absorbers i n the molten salt carriers, as compared with graphite and heavy water in other breeder reactors, and the addition of a& other atom with appreciable cross section can only lower the breeding performance.

, I

There are two ways of improving the breedperformance o f t h e MSBR. These are (1) determining the optbum C:U r a t i o and (2) increasing the thorium inventory in the blanket. In regard t o the C:U ratio, it i s believed that the value of approximately 9 0 0 used i n these calculations i s near the optimum and that only a very slight improvement might be expected by changing the reactor camposition. The most significant improvement i n the MSBR breeding performance can be made by increasing the thorium inventory i n the f e r t i l e stream. In the blanket, Pa-233 competes with thorium for neutrons; hence the losses t o Pa-233 are inversely proportional.to the thorium concentration. However t h i s improved breeding performance comes a t the expense of additional charges for thorium and fertile

salt inventory, and the net effect on the fuel cycle cost w i l l be an increase.

Above a 27O-tonne thorium inventory, which was used in t h i s study, the increased breeding; credit is insufficient t o offset increased thorium and fertile salt inventory charges. The molten salt reactor conceived for this study necessarily includes some elements of design which are perhaps beyond current technology, e.g.,

leak-

proof graphite-to-metal joints and Impervious graphite that permits minimum xenon * absorpt;ian. chemical processing, further demonstration of the fluoride volat U i t y process and the HF dissolution process is necess t o supply adequate design inf'ormation.

- 86 8.0 ~ E N C E S 1. L. 0 . Alexander, et al., Thorium Breeder Reactor Evaluation. Part I. Fuel Yields and me1 Cycle Costs in Five Thermal Breeders, EXL-CF-6i-3-9, March 1, 1961.

R. C. Robert'son, Sizes of U. S. Steam Electric Plants, ORNL CF-!59->130, 26, 1959.

W. Do Burch, D. 0. Campbell, and H. 0. Weeren, Processing Methods, Fission Product Poisoning, Fuel Cycle Costs for Fluid Fuel Reactors, ORNL CF-60-4-1. April 1960.

C. W. Nestor, Multigroup neutron Cross Sections, ORNL CF-60-3-35,

March 15,

1960

5. J. E. Evans and

R. 0. nuharty, "Evaluation of Low-Ebergy Cross-section Data for U-233," Nuc. Sci. 5 66 (1960).

- - a.

6 . R.

W. Stoughton and J. Halperin, "A Review of Cross-sections of Particular Interest t o Thermal Reactor Operation, " Nuc. 6ci..&I 5 100 (1939).

A. Nephew, !B.mrmal and Resonance Absorption Cross-Sections of the U-233, U-235, and Fu-239 Fission ProdUCtS~ORNL-2869, Jw. 16, 1960.

7. E.

8. C. L. Davis, J. M. Bookton, and B. E. Smith, A Multigroup, One-Dimension Diffusion Program f o r the IB4-704, OMR-101, NQV. 12, 193.

9. W. E. Kinney, Cornpone i n preparation.

- A Multigraup, W t i r e g i o n Reactor Code, ORJL-2789,

10. L. G. Alexander, WC-5 Program for Computing the Eguilibrium States of TKORegion, Thorium Breeder Reactors, ORNL CF-60-10-67, Oct. 20, 1960. 11. Weinrich and Associates, Process Designs and Estimated Costs of Chemical =ants for Processing Molten Salt Reactor Fuels, a report t o the ChenicaL Technology Mvision of the Oak Ridge National Laboratory, June 1959.

12. H. 0. MacPherson, e t g.,Interim Report on Fluid-Fuel Thermal Breeder Reactors, ORNL CF-&34I. TRevised), March 15, 1960.

13. E. G. MacPherson, Molten Salt Breeder Reactors, ORNL CF-59-12-64 (Revised) an. 12, 1960. 14. I. Spiewak and L. F. Parsly, Evaluation of Ekkernal Holdup of Circulating I

Fuel Thermal Breeders as Related t o Cost and Feasibility, ORNL CF-60-5-93, May 12, 1960.

15. E. R. Payne and J. C. Moyers, Determination of Capital Costs of Steam Cycle Equipment and Over-all published data.

mast

Efficiency for Three Breeder Reactors, un-

16. L. 0. Alexander,

O a k Ridge National Laboratory, unpublished data.

17. L. Dresner, Tables for

CaanpUting Effective Resonance Integrals, Includi m e r Broadening of Nuclear Resonances,. ORNL CF-55-9-74 (Septr. 19, 19%).

18.

C. S. Walker,

Reactor Controls,

(PWL

CF-fl-1-1 (Jan. 5, 1 9 3 ) .

ilg. M. L. Yeater, R. W. Hockenbury and R. R.

Fullwood, Eta of U-233 from 1 ev t o 800 ev, Rensselaer Polytecbnic Institute Report, June 1960.

;20. L. G. Alexander, et al.,

Thorlun Breeder Reactor Evaluation. Part-I. Fuel Yields and Fuel Cycle Costs in Five Thermal Breeders, ORNL CF-61-3-9 -(Appendices, Part I), March 1, 1961.

121.

J. W. Miller, Evaluation of a Deuterium-Moderated Gas-Cooled Breeder Reactor, ORNL CF-61-3-2, March 1, 1961.

' I

9.0 APPENDIX

T&&S 6.2. Performance of a Molten Salt Breeder Reactor for SeveraJ. Values of Fission Product Poison Fraction. Fission Product Resonance Absorptions Included in Calculations.

?&rtilestreamcycletlme(days) Case No. i&&n fraction Volume fra&lon fertile stream ln core Volme fraction grabhite in core Carbon: Uranium ratio in core lkel stremncycletime (days) mel saltdlscardcycletime (days) Fractionof fuel stream soldas'product 7!kactioti of fertile stream sold as prodkt Fraction of fission in fertile stream Wel stream composition (atoms/cm3)(10~ 24 ) th) U-234 y& m-237 . Fissi (a)' Xe-135"p b, carrier(c) Fertile

0.068 0.772 WO

0.066 0.775

oJJ7* 0.770

5050 a4

5055

$0 O.Ol.04

1550

if5

O.OOjl

0.0134 oAl33

O.oog4 :%I% .

0.026g 0.0245

o.ahE-4 0.2729E-4 oof2z .

0.8iW33-4 0.2554~-4 yws-5

0.54703-6

O&E-7

E5 0.0035

0.aa2 0.0139

y74mE~

' $

0.8855E-4

0.26l.5E-4

See end of Table 6.3 for footnotes.

0.06

4 0.02

:z$iE 0:2a4?E-6

o.&gE-g 0.208~4

O.lrgoE-g 0.2oap-1

o.iz&hg0.20853-1

O.l2/OE-9 0.2085!%1

0.4012E-2 O.Z@&E+

0.4O'U?E-2. OL&lE-5 O.l4OlE-5 O.lOl3E-7 0.3444~.10 O.l&@E-5

0.40123-2 O.l554E-5

0.4&-2 0.16393-5

0 ;136oE-5 0.9m-B

0.30323-5 o.rlllE-7

kE2g; 0.3444~10 0.18923-5 0.37163-g .

OhlE-7

(atoms/cm3)(10'24)

Z-233 U-233 U-234 V-235 Fissium(~) sfn-1%

Note:

::z

0.2516E-4

stream co&position

z;zr(d)

2 0.04

o.aagiE-4

U-233

1 0.02

0.3444E-10 O.l?I&E-5

0,3444E-10 y5f&-;

0:2062&9 0.4OUE-2

0.0679 0.772 5065

Eo 0.0089 0.0179 0.025-7 0.87UE-4 0.2647E-4

6 0.06 0.0653 0.775 5075 7a 1700 o.om5 0,oog

060244 0.86&E-4 0.flfjlE-b o.!mm-5 0.16373-4 o .5332~-6

0.12703-g 0.20853-l

0.4OI.2E-2 0.15963-5 0.29553-5 0.2584~-7 0.3444E-10 0*3491E-5

0,4012E-2 0.15503-5 0.2871&5 0.2433-7 0.3444~10

!hue 6.2

Case No. Neutrons absorbed by listed element per neutron absorbed i n fuel Th

0 . m

Tn fissions

0.mg

pa-233 x 2 U-233 U-234 U-235 U-236 NP-237 Xe-135(b) Sm-1% + Sm-149

0.0ug

Fissium Carbon

Fuel carry&] Th carrier Corrosion products Delayed neutrons Leakage Fllel processing Neutrons born per fuel absorption (?e) breeding ratio

0.9168 0.0892 0.0832

0.0w3 0.OOM 0.00p O.ooo1 0.0207 0.0266 0.0302 0.0200

O.OOO8 0.0043 0.0016 0.0022

2. a 6 10w53

- cont'd

4 Inventory per station (kg) Th I n fertile stream Th in processlng Total Thorium

U-233 in fertile stream Pa-233 in fertile stre U-233 in fertile stream processing U-233 in &el streem in reactor U-235 i n fuel stream i n U-233 In external f'uel circuit U-235 i n external &el circuit U-233 in Atel processing U-235 in fuel processlng U-233 in fuel reserve U-235 in fuel reserve U-233 + U-235 in f'uel aUmp tanks Total fissionable inventory

263,000 7500

rl0,500 94.6 108 2 07

196

19.5 274 27*2 43.1 4*3 â&#x201A;Ź8.4

6-7 s.6

896.1

263, 000 7500

no,500

92.2 ,105 2.6 195

20dr

273 28.8 10.1 1.1

67 -7 7-1 3-7 854.9

263, ooo 7500 270,500 89.5

102

2.6 195 21-9

30.6 6.1 0.7

199 108 2.7 193 19.9 269 V.8 42.4

263,000 3500 266,500

263,000 3500 265,500

1% 105

2.5 192 29-3

10.1 1.1

189 102 2.5 192 22.2

267 31 .o 6.5 0.7 55.8 7.6

7.5

4.4 67.0 6 09

845.6

991.1

937 5

66.9 s.8

66.4

51.2

s.0

1393 6.0

18.3 4.4

35.5 2.3

'f! '

!Cable 6.2

Case No.

fiel cycle cost (mills/-) Uranium inventory Thorium inventory Fuel salt inventory Fertile salt Inventory Fuel processing plant Blanket processing Thorium amortization fie1 s a l t replacement Fertile salt replacemen&)

0.076 0.132 0.023 0.166 0.69

“tt

Gross €bel cycle cost Breeding credit Met fuel cycle cost Processing rates spent fi3/ay> Thorium (kg day) Thorium replacement (&/day fie^ s d t replacement (&/by) Excess fissile atoms produced (kg/day)

Fertile stream loading, (gm U-233

+ Pa-233)/@

- cont’d 2

5 ’ 0.08 0.13

0.261 0.030 0.456 0.034

OL66 0.265 0.261 0.030 0.145 0.034

0.072 0.132 0.022 0.166 0.207 0.2Q 0.030 0.040 0.034

0.083 0.130 0.023 0.164 0.650 0.164 0.030

1.832 0.136 1e 7 0 2

1.128 0.090 1.038

0.964 0.647 0.917

1.735 0.lrl 1.608

44.

10.4

44.2

6.a

7500 39-7

7500

3500 3901 a7

3500

0.7’7

0.073 0.132 0.022

39.6

0075

0.455 0.034

0.02 0.164 0.269 0.164 0.030 0-135 0.034

0 0.130 0.022

0.164 0.207 0.164 0.030

1.029 0.088

0.941

39.0

18.9 0.076

0.203

17.3 0.072

0*73

1.2

1.1

cTable 6.3.

100 6 0.04

7 0.02

Volume fraction f e r t i l e stream in core Volme fraction graphite i n core Carbon: Uranium r a t i o in core ~ u e stream l c y d e time (days) salt discard cy d e time (days) Fraction of fuel stream sold a8 product Fraction of f e r t i l e stream sold as product !&action of fissions in fertile stream

o.qo4

0.0678 0.772 5090

0.770

12 1%

400

0.0029 0*02n 0,0348

. 0

200

9 0.06 0.0652 0.775 5090 86 1500 0.0074

O.Olg8

o.mq

0.033

0.035

0.86l4E-b 0,2763i3E-L

0.gOm-5

0,86390.2667E-4 0-9533s5

Od.l.90E-4

0.14033-4

0.663@-7

O.lr/OE-g

0.02

0.04

0.06

0.0698

0.0673 0.773 5150 50 445 0.0062 0.0259

0.0647 0.775 5160

0.770

5140 U .8 150 0.0022 0 03-70 0.0647

1500 0.0052 0.0132 0.0587

-24 (h)

fiel stream composition (atams/cm )(lo

U-233 U-234 U-235 0-236 m-237 (a)

Performance of a Molten S a l t Breeder Reactor for Several Values of assion Product miss Fraction. Fission product Resonance Absorptions Included i n Calculations.

Fertile stream cycle time (days) Case No. Poison fraction

Xe-135 Fissiurplb) carrier (C

)

0.86630.25353-4

0 27l4E-4

0 2795E-4 0 e102lE-4

0 0 3 ~ - 6

6 7 5 ~ 4 0,5936~-6

0.8346E-4 0.26343-4 0 9368E- 5 0.1327~4 0 *73m-7

0.15313-4 0.334936

O.lTgE-4 0.6l78E-6

0.12713-9

0.m-9

0 .mOE-g

0.127.l.E-g

0.12723-9

0.2085E-1

0,1004E-4 0~

0.8340E-4

o 97813-5

0.8334~4 I

W >

0.208%-1

0.2085E-1

0.20853-1

0.40123-2

0.15&~-5 0.7519-5 O.lrn2E-6 0.3W~-lo 0.83903-5 0 J7olE-8 0-48793-9 0 AOI.23-2

0.40323-2

0.1%1~-5 0.7320E-5 0 -96233-7 0.34&E-10 0-798673-5 0.16583-8 0.47633-9 0.40123-2

Case No. Neutrons absorbed by listed element per neutron absorbed in fuel !El

0.9915

Th fissions Pa-233 X 2 U-233 U-234 U-235 U-236

0.mg 0.0~8 0.939 0.0914

Sm-ls + Sm-149 Fissium Carbon RSL carri r ( i ) carrierfdji) Corrosion products Delayed neutrons

mel processing Neutrons born per fuel absorption (:e) Net breeding ratio

c.:

0.9646 0.0018 IO.aU2 O=!wt

0.WS

0.0861

0.0906

0.0120

O.Ol41

0.003

O.Wl3 O.OOg0 0.0003 0.0418 0.0287 0 0302

0.00% 0.0003 0.0218 0.0286 0 0302 0.0200

O.OOO8 o.OO43

a95

0.0008 0.a3

0.9046

o.om9 0.0954 0.0168 0.0023 0.0050 0.0003 0.0617 0.0288 0.0301 0.0191

0.9853 0.mg 0.0~6 0.g109 0,0933 0.0891 0.0133 0.0003 0.0050

O.OOO6 0.0234 0.0286 0.0302 0.019

0-9325 0.0018

0.a05

9071

0.0963 0.0929 0.01% 0.0013 0.0050 0.0005 0.0432 0.0287 0.0302 0.0194

0.9030 0.09% 0.0970 0.0179 0.0024 0.0050 0.00 0.06 0.02 0.0301 0.0190

O.OOO8

O.ood3

O.OOO8

o.ooo8

0.0043 0.0016 0.0022

2.ulg 1.0682

2.2l07 1.042

1.0u5

0.0016 0.0022

0.0016

0.0022

0.0043 0.0016 0.0022

2.2128 1. q 2 l

2.2~5 1.0487

2.2101 1.0247

2.2096

Q' - cont'd

Table 6.3. Case No.

Inventory per station (kg) Th in f e r t i l e stream Th in processing Tot& thorium '

U-233 in f e r t i l e stream Fa-233 in fertile stream U-233 in fertile stream processing U-233 in fuel stream in reactor U-235 in fuel stream in reactor(e) U-233 In external fuel circuit U-235 in external fuel circuit U-233 In fbel processing U-235 in fuel processing U-233 In fuel reserve U-235 in fuel reserve U-233 + U-235 in fuel dump tanhs

263,000 2600 265,600

263,000

263

2.6 191

256 105 2.6 191

20.1

21.2

267 28.1 42.0 4.4 66,2 609

266 29.6

108

a *Y 1.o

1049.9

&el s a t (fuel excluded) manket salt (m4 excluded)

3,500 19'7,340

29,500 IÂś, 340

14.3

5-6

263,OOo 1300

265,600

T o t a l fissionable inventory

Doubling time (full power years) Fuel yield at 80% plant factor ( M y e a r )

2600

65.7 7.3 50 -7 1005.0

50.6

264,300

507

494

481

107 2.5

104

10.l

2.4 184

184 20.8

257 29.0 41.i

0.7 65.1 7.6

4.7 63.4 7-l 49.1

3-272-7 29,500 191,340

196,400

20.3

39.5

3.9

2.0

18.3 4.4

257 3O-3 9.6 1.1

63.o 7.4 49.3 1224.0

62;7 7-7 49.3 1206.2

31,400

55.0 1.5

- cont'd

Table 6.3. /'

Case No. f i e l cycle cost ( m i ~ . s / ~ ~ )

Uranium inventory Thorium inventory A ~ e sl a t inventory Fertile salt inventory processing plant Filanket processing 'Barium amortization Fuel s a t replacement Fertile s d t replacement(g)

0-w 0.130 0.023 0.163 0.641 0J39 0.030 0.440 0-033

Gross fuel cycle cost Breeding credit Net fuel cycle cost

"tt

0.035 0.130 0.022 0.163

0.139

0.207

0.109 0.129 0.023 0.162 0.653 0.092 0.029 0.439 0.033

0.105 0.129 0.022 0.162 0.269 0.092 0.029 0.140 0033

12 0 .io3

0.129 0.021 0.162 0.207

0.092

1.689 0.125 1.564

1.ooe 0.084

0.853. 0.043

0.94

0.808

1.669 0.u8 105%

44.2

9.5 2600 38.9 74.0

6.2 2600

4.409

10.6

6.3

1300

38.8 m9

38-7 66.6

0.189

0.125

19.6 0.060 .

2.3

2.2

Fertile stream loading, ( e ~ nU-233 + Pa-233)/kg C !h

1.4

0,086 0.130 0.022 0.163 0.250

0.033

2600 38.9

00139 0.030 0.042 0 033

Processing rates spent ft3/aay) Thorium (kg day) Thorium replacement (&/day) f ie1 s a t replacement (@/day) Excess fissile atoms produced (kg/day)

0.030

0.155

210 0.19

388.9

0.135

19.6 0.068

1.4

1.3

0.981

0.078 0 903

0.929 0.041 0.033

0.817 0.037 0.780

38*7

I-

Footnotes for Tables 6.2 and 6.3 The element fissium is a conglomeration of fission products.

f i c t i t i o u s reaction rate coeffidient and concentration were assigned t o fuel stream fissium as explained above i n Section 4.5 t o achieve the desired poison f'raction. The concentration of fertile stream fissium was calculated by ERC-5 using reaction rate coefficients developed *om GNU and Cornpone output. All Xe-135 i s assigned t o the f i e 1 stream. It has been assumed t h a t a gas purge of fuel and blanket solutions will maintain Xe-135 absorptions a t 0,005 neutrons per neutron absorbed i n f'uel. Based on Li-7 atoms. The absorption cross section of the f e r t i l e stream carrier was normXized t o the basis of one thorium atom.

U-235 in fertile: stream i s negligible. Includes thorium burned up in breeding p l u s thorium discarded on 20-year cycle. Fertile salt is discarded on a 20-year cycle t o maintain blanket fission products a t a tolerable level. The concentrations are written with the letter "E" used t o denote the exponent, e.g., read 0.88923-4 as 0.8892 x 10-4..

Replacement salt for Arel and fertile stream carrier i s assunred t o contain L i that is 0.01 atom $ Li-6. Absorptions-are based on equilibrium Ll-6 concentration for this feed. -

It;

Distribution

L, G. Alexander 11. S. El Beall 12. E. S. Bettis 13. F. IF. Blanke 14. A. I,. Boch 15. E. 0. B ~ h l m a n n 16. R. B. Briggs 17. W. D. Burch 18. D, 0 , Caslpbell 19. W. HI Cam 20. w, L. carter 21. 0 . I. Cathers 22. R. H. Chapman 23. F. L. CuUer 24. J. 0. Delene 25. E. K. Ergen 26. D. E. Ferguson 2'7. A. P. Fraas 28. W. R. G a l l 29. H. E. Qoeller 30. W. R. G r i m e s 3. J. P. Hammond 32. W. H. Jordan 33. P. R. Kasten 34. B. W. Kinyon 35. J. A. Lane 36. M. I. Lundin 37. R. N. Lyon 38. E. G. MacPherson 39. W. D. Manly 40. W. B. McDonald 41. H. F. McDuffie 42. C. W. Nestor 43. L. F. Paxsly 44. A. M. Perry 45. C. A. Preskitt 46. I. Spiewak 47. J. A, Swartout 48. A. Taboada 49. R. Van Winkle

1-10.

50.

G. M. Watson t

55-56. 57-9. 59-68 69.

Central Research Library Document Reference Library Laboratory Records ORNL-RC

EXTERNAL

70. H.

72-73. D. 74.

75. J. 76.

77.

78. R.

79. B. 80.

~ehnnan,AEC, Washington Brooks, Hasvard University F. COP, AEC-OR0 H. Groelsema, AEC, Washington F. Kaufmann, AEC, Washington Link, Argonne W. Miller, K-25 E. Pahler, AEC, Washington E, Prince, AEC, Washington Robba, Brookhaven P. Self, AEC-OR0

81-82. F. 83. D. c. 84-98.

maas,

Washington TISE-AM=