ORNL-CF-61-8-86

sEP t 9 1961

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 1 61-8-86 , 1

External Transmittal Authorized DATE: August 18, 196l COPY NO. gf SUBJECT: Thorium Breeder Reactor Evaluation. Part I. Fuel Yields and

Fuel Cycle Costs of a Tvo-Region, Molten Sal t Breeder Reactor TO: Distribution

FROM: W. L. Carter and L. G. Alexander

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 extrap- olates t o 0.65 millS/kWhrj at 4$/yr (DT 0.85 mills/kwhr.

25 years), the fuel cycle cost is about .c A l l systems were optimized with respect t o fuel cycle proc-

c - w. essing 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 fert i le 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 &2,5 t o f3$/yr variation i n fie1 yield but only a 20.06 mills/kwhr

I '

variation i n fuel cycle cost.

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.

0.2 mills/- to fuel cycle cost.

0.5$/yr, but fuel cycle cost is negligibly affected,

2. Using 900 barns instead of1200 barns for Pa-233.resonance integral has

3.

4.

Doubling the thorium inventory adds about 1.9$/yr t o Puel yield and

Five m o l e $ ZrF4 i n LIF-BeFe-W4 Fuel salt decreases fuel yield about

\. -3 NOTICE

This document contai information of a preliminary notum and was prepared primarity for internal use at the Oak Ridge National Laboratory. It is subject to 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.

. Tw

I LEGAL NOTICE

f h i a npo 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, wi th roapoct to tho occurocy, completonoaa, or uaefulnoas of the Informotion contained in thi. ropwt, or thot the uao of

ony informotion, opparotua, morhod, or procoaa disclosed in this report mor not infringe privotoly owned righta; or

B. Asaunoa any l iobi l i t i ta 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, "prim 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 wi th tho Conmiasion,

or h is omploymont wi th auch controctor.

I

I

i - . -

LI

c

. g- , ,- e 3

t‘ e Y

W

- 3 -

43 43 45 47 47 50

53- 56 T7 58 59 59 59 60 62

FORklORD LJ 3

5 5 'As pa+k of the GRk, responsibility for guiding the ARC !PhermaJ Thorium

Breeder Reactor Program, an evriluation of types of reactors -capable of ef-' ficlent utilization~of thokum was lr@tiated at GRNL in July,1959. Included in this evskmtion were studies on the Aqueous Homogeneous Breeder 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 MSBR evaluation. 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 appreci- ation 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-ft- thick 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- stream and'10 ml $ gra;phite, Basic criteria of the study were that the reactor complex be capable of producing power at a, rate of lOOO$we and that chegical processing c . . . be carried out oi‘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 made of~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 GNU program8 for the IBM-704 and the Cornpone program' for the ORACLE. after attaining criticality

10 in these calculations, further computations were made using the ERG-~ program for the IBM-7th to determine the equilibrium conditions It is the equilibrium results that are reported here.

bd The MSBR is capable of breeding over a %tide range of operatlng conditions giving FudL yields* as high as about ?$/year for a doubling time of about ll.5 Aril-power years . 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 fuel 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.

mills/kwbr are predicted at Are1 yields of 1 t o 2$/year. When the fuel yield is k$/year, the fuel cycle cost is approximately condition, the income from sale of fertile material just offsets the asznral in-

3

9 J

In the reab of more econamical. operation, Atel cycle costs as low a s 0.65

mills/kvhr. A t t h i s l a t t e r

ventory charge. Calculations for-a representative set of operat- conditions 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 the reactor fiel. Eta values a t epithermal energies within&IO$ o f t h e values recammended for th i s study were employed In 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; fue ly ie ld wa,s lqproved about 0125$/year with a negligible lowering of the Asel cycle cost. inventory (140 tonnes vs 270 tonnes) decreased the fiel field about 2$/year w i t h

a corresponding tiecrease oi 0.2 mills/k~"nr in cycle cost. A representative 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

P.

P I

A lower thorium

ed t o a similar case containing no zirconium.

1.0 ~ O m C T I o n W A

The work on the Molten Salt Breeder Reactor (NSER) reported in this memo- Y rmdum i s a portion of a more complete study on +he& breeder reactors, which

includes the Aqueous Homogeneous Breeder Reactor (ABBR), the 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 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 inc laed in the reference memra?xlm. It is €ulvisable 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.

breeding potential and economic performance. t o neutron economy and is therefore associated with the cmposition and design of the reactor, 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. this cost asalysis because no reliable cost data are available; these are the capital investment in the reactor installation and waste disposal charges. 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. reactor fiel cycle and henceforth are referred t o as fuel cycle costs.

1

The MSBR was examined with the viewpoint of obtaining a relationship between Breeding potential is related directly

s Economic performance is determined by the annual charge on such -,

V

Two of the above charges have not been included in

In

It is observed that the r a i n i n g charges are concerned with the

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, a

such as reactor structural materials, are fixed by design requirements. A significant advantage in neutron economy is realized by controlling poisoning 2 from fission products by chemically processlng fuel and f e r t i l e streams for

- 9 -

their r e d . a t any desired poison level between that corresponding t o some practical minimum emd that of cmplefe burnout of fission products. It is customary t o identify fission product poison level in a reactor as the poison fraction, which is de-

fined as the rat io of neutrons absorbed in fission products t o neutrons ab-

sorbed in fuel.

It is a m e n t that the system in equilibrium may be operated

e 7

There is an inverse relationship of poison f'raction t o breeding and economic e, In order t o maintain h i reeding performance,. it is necessmy t o

e streams on a reiatively Frequent schedule at chen+cally process fuel and fer the expense of hi@ Fuel cycle cost, 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. 2.3 for a discussion 'of the chemical processing system.) In th is 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

On the other hand, less Frequent processing '

(See Section

c I

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 .

as plots of he^. cycle cost (mills/k~hr) versus fuel yield ($/year) and poison fraction .

The pertinent results are exhibited

. - *

i L,

2.0 DRSCRIFTION OF SYSTEH -. j .

2.1 E&ysical system -,

The molten salt breeder reactor 'examined,in this study is based upon the design of Mad&erson

I.- l3 '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 bye a 3-ft-thick blanket. Al-ft-thick graphite reflector surrounds the blanket on the sidesandthe ends. The reactor, heat exchanger'and circulating pump are 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. !Che reactor 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 some bypass 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.

.

,.-. bd

A

14

*

w

zc Y

UN C L A SStFlEO - 11 - O R N L - L R - D W G . 46040R

SECONDARY

’ C C O O L A N T

P

EXCHANGER

€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.l pOrtion 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 MM of 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 fertile- stream volume be kpt low. In fact, in 6ome cases 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 &%%a is tabulated in Table 2.2.

r.

v

4

- 13 -

Table 2.1. Molten Salt Breeder Reactor Plant Data(a)

General

Station electrical power, MwE lo00 Station net thermodynamic efficiency, $ 42.3 Number of reactors per station Thermal power per station, MwT

2 1182

F’raction of electrical power fed back into plant Geometry of core Moderator graphite Volume fraction of moderator in core Diameter of &ore, ft 7.66

0.03 cylinder (L/D = 1 )

- 0.773 Length of core, f’t Thickness of blanket, ft

7.66 3

Volume fraction of moderator in side blanket 0 010 I

Volume fraction of moderator in end blanket Reactor vessel material Reactor vessd thickness, in.

0 .lo IMOR-~ 1 375

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

Fud. Stream

Fuel carrSer

Density (12CJQ°F), lb/ft3

3 -75

63 mole $ LiF 37 mole B”F2

119.5 Fraction oithepmal power removed by

Mean heat capacl 0.544 Power density in portion of fuel s t r e w

fuel stream h 0.91

extermA t o reactor, Mwt/ft3 ._ 7.6 ~ 849

ation now rate, &/sec 178

Liquidus temperature, (b) OF

n Velocity (f’t//sec) of ~ z e l stream in

0

Core 20 s

End blanket 20

. .

Table 2.1 . Continued .

Heat exchanger data: tube, side shel l side Tube outside diameter, in. 0.. 375 Tube wall thickness, in. 0.029

Tube velocity, f't/sec 20 Flow rate, lb/hr 3.87 107 4.35 107 Fluid temperature % *F 1300 900 ELuid temperature out, OF 1125 1175 Pressure drop, psi 78 100 Mo. tubes per exchanger 8110 Tube length, f't u .13 Tube bundle dianeter, in. 69 Inside film coefficient, ~ t u / h r - f t ~ - * ~ 8020

Material IMoR-8 IEOR-8

Tube -1 coefficient, Bt;j/hr-ft2-mF 7080 Scale coefficient, 9tu/hr-fi2-*~ Outside film coefficient, Btu/hr-f'b2-'F

10, OOO

@,goo

/-

';id p.

v

0ver-U coefficient, ~ t u / h r - f i ~ - a ~ 2620 / v Out.side tube area, f"t 2 8320

Fertile Stream

Fertile stream carrier 67 d e 4 LIF 18 mole $ BeF2 1 5 d e $ ' f h F 4

Density (1200*F), lb/ft3 192 Mean heat capacity, Btu/lb-*F 0.32

f e r t i l e stream heat exchanger 0.09 &action of thermal power removed by

Fraction of fission power produced in fe r t i l e stream

station flow rate, &/sec 22.6

- 0.013-0.066

Mquidus temperature, (b) *F 932

Heat exchanger data:

Tube outside diameter, in. tube side shell side

t

0 375 - V

Tube wall thickness, in. 0.028

Material =OR-8 mm-8 z mbe velocity, ft/sec 14 .I

- 15 -

L,

P Y

f Y

u

Table 2.1. Continued

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 0Ver-u coefficient, ~ t u / h r - f t ~ - * ~ Outside tube area, ft2

tube siae she11 side

6 5.98 x lo6 4.48 x 10

1300 900 11s 1175 log 100

19 07 27 5550 5660

'1050

(a) A number of items in this tabulation are f r om a study by Spiewak and

(b) Temperature at which LiF precipitates. parSly.14

- 16 -

Table 2.2. Distribution of Fuel- and Fertile-Stream V o l u m e s i n the Molten Salt Breeder Reactor

Fuel stream i n

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

I

Fertile stream in ~

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

Volume Vofume per fraction station (ft3)

0.16 113 0.16 08.4 0 0

280.6 48.2 530 2

i.

95

{ 409

2470 3026 6000

cl

- 17 .

LI 2.3 chemical Reprocessing 6ys t em v

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

9- The processing operation consists of three parts: 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, 3

Fuel Salt Purification. The ael salt is purified in the fluoride volatility- BFdissolution 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 at which the reactor 1s permitted t o operate, whidh in t h i s investigation is a parameter.

The first step in purification i s t o fluorinate the molten salt vith ele-

mental fluorine t o volatilize m6. hydrogen t o produce UF4, which is recycled t o the reactor after dissolution in

This uranium hexafluoride is then burned in

t 1

the recovered carrier salt. flows f’ram the fluorinator t o the HF’ dissolution step. made betweenethe aaLt and the bulk o f t h e fission products. is dissolved in a 90s HF-10$ 5 0 solution leaving fission products, principally rare earths, as insoluble material. with recovered UF4, and recycled t o the reactor. 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.

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. 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

Uranium-f’ree salt, containing fission products, Here a separation I s

The carrier salt

The carrier salt is recrystallized, fortif ied In order t o purge those fission

,,

The fission

It is possible o achieve a kpecified poison

ombinations of the two cycle times.

Fertile Stream Processing;. The fertile stream is processed i n the fluoride volat i l i ty 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 ec t ly to the blanket without additional treatment. The m6 from the fluorlnator i s reduced with H2 t o Wk which is blended with UF4 recovered f’rom

F: 1

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

- 19 - L,

.? f

In the MSB at ferbile sitream cycle t i m e s less khan 100 days such a small

*action of fissions (4,6$) occurs in U e fertile stream that it is not necessary t o purif'y the sal% in a HF dissolution step. me flssion-product build-up is slow 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 this proce~s. PrOtactiniUm builds up in the salt unt i l I t s decay rate 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 YO~WZE giving Peirer neutron losses t o protactinium.

Fission Gas R e n m a L Fission gases are reaoved fromthe fuel 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.

2.4 Power Generatian Cycle

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

are given by Spiewak and Parsly.

The steam cycle of the TVA John &vier power plant was used as a model for

14

- 20 -

3 00 DESIGN BASES AND CaMpuTATIoEJAL METBODS a

3.1 Plant Size c

Based on a study by Robertson,2 it was assumed tha t f'uture pwer stations in the United States would have a capacity of the order of lo00 Mm. t h i s size was chosen for t h i s study. Also it was assumed that 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, at least two reactors &e specified for each station.

3.2 On-Site Processing

Consequently,

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 volati l i ty 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. was assumed that 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.

To avoid complicated calculations of startup and shutdown, it

3.4 Product Camposition

The product composition may vary between the l imits 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

the average composition of the entire system; i.e., reactor plus chemical processing systems. material are removed as product at the m6'vF4 reduction step. The product is an equilibrium mixture of uranium isotopes; viz., V-233, U-234, U-235, and U-236.

3.5 System Inventory

Calculated portions of the recovered spent fuel and of the bred

*

In order t o be consistent and unambiguous in the definition of fuel yield, the inventory should include all fissionable and potentially fissionable atoms 5:

(U-233, U-235, and Pa-233) in the entire system. 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.

Included i n the inventory is

Ld P

“;

-.

3.6 neutron josses.

Fission-product poisoning in-this reactor was based on a study by Burch, Campbell, and Weereq3 whckmade a study of the cumulative effect of 4.4 isotopes divided'into four &ups, This phenomenon Is 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 assumed that (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.

7 &her Fission Product Poisoning. Concernin@; the effect of the 44 fission-

product isotopes studied by Campbell, Burch, and Weeren, four groups were dls- 3

cerned and treated separately. The firstcomprised noble metals which were assumed to plajse out on the cold zones of the c+Alatlng system; the second comprised halogens which were assumed to 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.

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 assumed to vary inversely with the velocdty of the neutrons.

3.7 Nuclear D&a

Muclear cross sections for this study were compiled by Nestor.' Following the recommendations of Fluharty and Evans, 5 a velue of 2.28 was assigned to eta of U-233 at thermal energies. The resonance integrsl of Pa-233 was assumed to 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 composed of 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 GNU code,is limited in the number of elements for which absorptions

G = 4

i

Y

f

s

G

- 23 -

’ _

be ~calculated; consequently, hn&g needed calculations. A pseudo cross ,.

these elements saved space on the tape section for thefuel salt was obtained .-

by normaUzing the cross section to the basis of one atom of,L17 and summing the 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 assumed tobe g$?;gg at; $ 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 expressed as the atomic concentration of A. j thorsum in the salt. The L17 purity Is the s& as used in the fuel carrier.

3.8 Ruclesr Cslculations

Muclear calculations on the MSBR were performed using two different reactor codes: the 3bgroup GNU code8 -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 homogenized system using GNU, and a criticality search was made to 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 MSBR could 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 code lo 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 . 3 !Che disadvantage factors m&ntioned ,above were used-to relate the concen- ,

u tratlons of the homogenized readtor to those of the heterogeneous reactor. These fa&ors were determined In a lattice-cell, calculation by means of the Cornpone

- 24 *

program for the O R A W which yielded sets of 34-group disadvantage factors, one set for each region of the l a t t i ce cell . When employed in a 34-graup finite- reactor, 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

= volume of l a t t i ce ce l l vC

= volume of region j in the cell

fin = neutron flux in 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.

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 th i s capacity is within the region of Interest of this study. Weinrich*s

Ibel Stream Processing. The capital cost of the fuel stream processing

* 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 Labora- tory 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 volat i l i ty 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 Interest Rate ($/u 1 (&/Yr 1

Uranium inventory 159 OOo 4 Thorium inventory (as T ~ F ~ ) 2.p) 12.7 (b 1 FWL salt iwnetory (U excluded) 4 0 . p 12.7 (b 1 Fertile salt inventory (Th ex 45.6(') 12.7 (b 1

Thorium amortization (20-yr cycle) 27 2.6

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)

Breeding credit 158

( a ) ~h W u e at $=/kg plus $ 5 / ~ for preparation of s a l t sal.ution.

(b) Includes interest at 6$9 income taxes a t 4,6$, an8 local taxes and insurance at 2.14.

(c) Based on LiF at $44/kg and 3eF at $15.4O/kg plus $11 kg salt for

(a) Includes 14s interest on capital investment pius 155 for operation maintenance . * p r e p a t i o n . A t d c concentrahon of Ll is 99.99$ L i .f

- 26 - -.

‘. . . :

Table 3.2. -Cost Estimate of..Facilities-for Continuous . Processing of Molten Sa&t,Rreeder Reactor 374. Stream.

Weinrich's'dost Estimate Compared with Revisiori'kade by ORRL

?lkhk.s and Vessels mWeinrfch's Estimate ($1 Ow Estimate ($)

_ (Core Sslt Section)

Core saltholdtanks 19,500 39,ooo Core salt fluorinators W5q . 33,- UF6 chemicsl traps 311900 . . 63,800 uF4-m6 'reduction tower . 8,600 17,200 Vibrators, filters, burners, etc. 5,c)Qo 10,ooo HF dissolving tank 24,000 em RF evaporators 1Ol;m' W,w3 _ HF condensing tower a&= 4'6,200

HF storage tank 41,300 82,600

KOHscrubtower 3,500 - 7,m Mistikllaneous storage and utlllty tanks 20,000 20,000 Sub-Total 294,400 547,300 Installed cost (4.35 x cost) 397,440 ?38,8~

Coolers (Core Salt Section) :

m6 gas coders 3*@ 3,- ‘-

Reduction tower vent cooler 3,m 3,ooo HF' vapor desuperheater 9,600 91- HF condensing tower vent cooler 6,000 6,Qoo Circulating RF cooler 48,ooo %ooo CXrculatiugH20 uer 2,400 2,400

Sub-Total 72,6ot3 .72,-

InstsXl.ed cost (ml.10 x cost) 79,860 79,860

kessels and Tanks (Rlanket Salt Section)

Rlanketsaltholdtsnks 9,800 Blanket sat. fluorinator 51500 Fertile stream

m"6 chem.icEil trap .aSE processing esti-

Sub-Total 18,goo mated separately.

IustsXLed cost (el.35.x cost) 25,500

- 27 -


Weinrich's Estimate ($1 ORNL Estimate ($1 - - Coolers (=-et a t Section,)

w 6 codier . . 2,400 Installed cost (la.10 x cost) 2,640

Miscellaneous 4ulpment

Ausps 46, Ooo 69,700 Agitators 6, OOo 6, Ooo Filters Freon refrigeration system

20, ax, 160,000

Fbel reconstitution system M,m 80,000

Electric heating f'urnaces 148,OOO 148,OOO Pipe heating equipment 60,000 6%OOo F2 + gas supply systems 20, OOO 20, OOO

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

Sub-TOtaJ of' instelled cost of major equipment

A t t e n d a n t Facil i t ies

20, OOO 20, OOO 534, OOO 589, OOO

720,900 796,140

Special instnunentation Genera instrument at ioi Panelboards and alarms

I n s t U e d cost (d.40 x cost)

Piping, painting, ScaffolBs

Sub-Total

etc., Installed Cost

Special plplrq &nerd pip%(") muipment footings and foundations (b 1

60,000

160,OOO 24,000 .

224, OOO

Pipe insulation I

- 28 - Table 3.2. Continued

Weinrich's Estimate ($) ORE& Estimate ($)

Ruipment insulation 20, OOO 20, OOO

Electrical. distribution, lighting, etc 144,OOO 144,000

Remote operating equipment 75, OOo 75, OOo

Field testing and inspection 25, OOo 25,000 Operating and safety supplies 15, OOo 15, OOo

Sub-Total. 725,000 1,767,800

Painting (c 1 28, OOO 36,300

meight ('1 2UEE 48,400

TOTAL INSTALLED COST

Contingency (e ) 23-79580 901,650 TOTAL DIRECT MA!FEFUALS AND LABOR 2,393,400 4,508,300

Fees and Expenses , I Contractor's f ie ld expense (f,d U9,670 2,254,150 V

I 359, OOo 0% e 1 I Contractor's overhead fee

Purchasing and shop inspection (3 1 119,700 225,400 Engineering and design (i 1 478,700 901,f-

Estimated Cost of Additional Facilities

Sampling fac i l i t i es Ventilation Waste removaL C e l l s and buildings Laboratory

Crane Mock-up ce l l

2

- U

TOTAL ESTIMATED PLANT COST 4,970, Too 9,849,500 ci

- 29 -

‘cd ‘: ~+ie 3..2. CoIltinued I Footnotes

\ \ , f (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 Sum of contractor's field--expense.and overhead,fee taken by CBNL as 505 of tots3 direct materials.and,labor .costa -v . .

b) ' Estimated by.Weinrich'as 15s of to&IL direct materisJ.s and labor cost.

. . .'I. 6) Estima$ed as 20s of total direct matgrlals and labor cost.

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

plants. T&e portional t o

extrapolation the 0.6 power

- 30 -

was made by assuming that the of the processing rate. mis

,"--

capita3 cost is pro- id method of extrapolating

cost data has been found reasonably accurate when applied t o the chemic& in- dustry 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 is economic t o change from continuous t o batch processing methods. In t h i s study it was assumed that the lower l imit of continuous processing wauld occur around 7 ft3/day, which corresponds t o a fuel cycle time of 75 days. (The fuel stream volume was constant a t 530 ft . I When the fiel cycle time is 75 days, the ex- trapolated cost curye (Fig. 3.1) indicates that the capita3 Investment is about $5 million. 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. the above figure dims a premium of $1.6 million over Weinrich's estimate.

3

Furthermore, it was felt that the Investment i n ch Plat ~ o u l d

A batch processing plant was estimated by Weinrich to cost $3.4 millionj

Fertile Stream Processing. The fe r t i l e stream is processed only in a

fluoride volati l i ty step and, therefore, requires much less equipment than the accompanying fuel stream processing plant. stream plant a s an i n t eg ra part of his 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.

fe r t i l e stream a t a ra te of 20 ft /day, the same basis upon which the fuel

stream processing plant was designed. capital investment would be about $1.8 million dollars.

i n Fig. 3.2j the remainder of the graph was obtained by assuning the cost was

Welnrichu included the f e r t i l e

This tabulation presents values that are apaicable t o a plant processing

For th i s ra te it was estimated that the

3

These values were plotted

proportional t o the 0.6 power of the processing rate.

8

UNCLASSIFIED ORNL- IS-DWG 54.794

. I, Continuous Processin$ I

3- 2 5 IO 50 IO= 500 .

FUEL STREAM CYCLE TIME (days)

FIG. 3.1 CAPITAL .COST OF FUEL STREAM PROCESSING PLANT FOR A MOLTEN SALT BREEDER REACTOR’ ‘,

- 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 inat or UF6 chemical traps UF@lF4 reduction towers Vibrators m6 gas coolers Reduction tower vent cooler

A M P S

Filters

Agitators Freon refrigeration

20, OOO

20, OOO

43,000 12, OOO

10, OOO

Bred material reconstitution 8,800 Electric heating furnace's Pipe heaters

F2 compressor

Installed cost (4.35 x cost) Sub-Total

21900

304,290

225,400

Attendant Facilities

40, OOO

Panelboards and alasms 15,000 instruments 30, OOo

Sub-Total 85,000 Installed Cost (d .40 x cost) 119, OOo

Installed Cost of PIpirg, Insulation, Paint- , etc.

. /--

Specid piping 3,000 u General piping (..J,oo$ of major equipment -cost) 225,400

major equipment cost) 33,800 EQuilpnent footings and foundations (=I.?$ o f

- 33 -

L, I .


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 (4s of major equipment cost) Sub-Total

Total installed cost Contingency (1.2946 of to ta l installed cost) TOTAL DIRECT MATERIALS MID LABOR Contractor's f ie ld m s e and overhead

(=%$I of total direct materials and labor)

materials and labor)

direct materials and labor)

7 Engineering EKUI design of to ta l direct

Arrchasing and shop inspection of to ta l .

Addition82 Facilities Shared with Fuel lCjalt Processing Facil i t ies

sawling. Ventilation C e l l s and buildings Laboratory

-

Mock-up c e l l

Estimated Cost ($1

1,200

31 OOo

~ , m 6,800

u,m 3,700 2,200

9,ooo 320,100

743,400 185,800 929,200

464,600

185,800

46,500

1,748,200

I

UNCLASSIFIED OANL - LR-DWG 54793

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

30 I I I 1 I 1 1 1 1 I I I

I O

2

I 500

FERTILE SALT PROCESSING RATE (ft'lday per station

" 35 -

c, i 4.1 Reactor Size

- ? For engineering reasons it was decided that the A d be a cylinder

having a height equal t o the diameter.

Core Size. In determining the core s ize of the MSBR it was necessary t o f+x certain reactor progez-ties. velocity in the core, and the temperature.ri6e of the fuel In i ts passage through

the core &re arbitrari ly chosen.

In- this study the thermal power, fuel stream

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

U stream velocity p = stream density c = heat capacity f *action of core cross sect i cupied by flzel stream AT = temperature rise

P

When the appropriate numbers are substituted in this equation, a core diameter of ab0ut~7.7 ft is obtained. eter was one-half of the to ta l core power for the station, givlng two reactors for the ;Lnstallation. .%is agrees with the decision that the t o m stst ion load vauldl not be committed t o a single reactor,

The power used in obtaining this diam-

nanket micknes;. The cylindrical core blanket on the sides and on each end. of the blanket was fixed at 3 ft on both ends an8 on the side. This thickness was

Based on previaus the thickness

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 10 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. Bd

- 36 -

4.2 GNW Calculations

The calculations for the MSBR Were made a s indicated in the f l

The basic nuclear calculations were performed on the mul Fig. 4.1. dimensional, 34-group GNU program for the IBM-704. 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 epi- the& group for the energy range 5.5 m-0.6 ev, and a thermal group.

8 The equilateral, cylindrical .

Input Data. Input of specifications of reactor geometry, dimensions, and the c densities of the several d e - ments i n the system. region i n which it appeared. through the core of the MSBR, the core concentrations included the sum of fuel stream and t h a t portion of the fertile stream concentrations. 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

The concentration of each element was homogenized over the Since a sm8U fraction of the fertile stream passes

The i n i t i a l values

elements considered in these calculations are given i n !Table 4.1.

Output Data. The GNU program provides a cr i t ical i ty search by which either a dimension or one or more concentrations are varied unt i l the iplicat ion constant differs *om unity by less than some small specified Bs1ount. calcdxM.ons the reactor was made cr i t ica l by varying the concentrations of protactinium and thorium i n the core. stream carrier is'flxed by the salt composition, this I s equivalent t o varying the volume fraction of fert i le stream i n the core, data were the fractions of neutrons involved in absorption and fission reactions for each nuclear species i n each region of the reactor.

4.3 Cornpone unit c a Cdcaat ion

concentrations of the heterogeneous reactor. a homogenized system.) The core of the MSBR was visuaLized as being composed of a number of cylindrical, unit 9

unit cel l contained s i x regions (see Fig. 2.1): annular fuel zone, graphite tube, fertile stream passage, and graphite moderator.

In these

Since the thorium density in . the fertile-

Mditiond. useful autput

The second step in the nuclear calculations was t o determine the atomic (The GEJU progrrtm could treat only

diagrammed in Fig. 4.2. The inner fuel zone, graphite tube, (I

c

Lid

4 ! 1

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 - U N C L A S S I F I E D OR N I - L R - D WG. 5 86 4 5

GRAPHITE MODERATOR 1

FERTILE STREAM OUTER FUEL TUBE (GRAPHITE) ANNULAR FUEL ZONE INNER FUEL TUBE (GRAPHITE) INNER FUEL ZONE

I

I

I , I

I

14 2.375 6 .03 2 . 6 2 6 .65

r6 .36 11.07 1 r6

~

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 for o l t e n S a l t . B r e e d e r R e a c t o r C o r e .

- 39 -

c, Table 4.1 . nements Considered in Budear Calculations of the Molten Salt Breeder Reactor

Th

pa-233 u-233 U-234

u-235 u-236

NP

xe-135 Fissium

Sm-149 Graphite Fuel-Stream Carrier (a 1

(63-37 m o l e Q LIF-B~F~) /

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

(a) Uranium excluded in nuclear properties of salt

(b) content excluded in nudear properties of salt

9 The cel l was examined using the Cornpone code for the ORACLE t o develop a set of 34-group disadvantage factors for each region of the unit ce l le These disa&vartege factors, defined in Section 3.8, were then used i n subsequent Cornpone calculations t o determine the concentrations for the cri t ical , heterogeneous reactor.

The Cornpone code treated the unit ce l l as &p 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 In the prelhdnary GNU calculation. plication constant of the cell.

I n addition t o the disadvantage factors, the code calculated the nnilti-

6.4 Cornpone Finite Reactor Calculation

The disadvantage factors were used in a finite reactor calculation using

the Cornpone program t o determine the c r i t i ca 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 its volume fraction i n the region. with each concentration was specified as 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. in the k-th region is computed by the double summation

The set (or sets) of disadvantage factors t o be used

For exampleo the absoqrbions i n the i - th element in the j-th stream

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

"Vkp Di,j,k centration, and v;(Au) is the absorption cross section.

is the homogenized atomic con- i,3,k is the disadvantage factor, N

It should be pointed out that disadvantage factors were applied on ly to 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. 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 cr i t ical i ty was achieved by varying the thorium. concentration in the core,

The code does not make a "search" on any of the input

Reaction Rate Coefficients, Element absorptions i n each region of the

i,j ,k' which reactor were used t o compute sets of reaction rates coefficients, C are defined by Eq. 4.

The symbols have the same definition as given above for Eq. 3 . summation is computed by the Cornpone code, the calculation of C

forward.

Since the double

is straight- i,3& -

,3,k to The coefficients are properly disadvant the use of Di reflect the heterogeneity of the system. 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 th i s f’raction is multiplied by the tota,l number of neutrons born per unlt time in the reactor, the product is the absorption rate hy element i in stream j in region k. This l a t t e r quantity is very useful in c&culating the equillbrium state o f t h e reactor as discussed below.

a l s have the useful property

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 the fission coefficient.

4.5 EQuilibriam Reactor Calculations (ERC-51

4uilibrium Reactor Calculations were next performed on the c r i t i ca 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 the fallowing input information: times, process holdup tines, and cr i t ica 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, crit icali ty, and conservation of neutrons; these equatiom used the absorption and fission reaction rate 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.

fuel and fertile stream volumes, volume fractions, process cycle

*

- 42 - - W Output Data. -The program calculates the equilibrium stream concentrations

and neutron absorptions i n both mel and fertile streams for the elements l isted

in Table 4.1. all uranium isotopes, thorium and protactinium. t o sell a product that has the same composition as the system mixture, the fractions of recovered fuel and fertile streams that are directed t o sales are calculated. fissions i n the fertile stream and the inventory of fissionable material reserved for a possible 3O-day shutdown o f t h e processing facil i t ies.

Also the inventories and mass processing rates are computed for Since the sales philosophy is

Additional values calculated by ERC-5 code are the fraction of

%e program offers the option of attaining cr i t ical i ty by adjusting the U-233 and U-235 concentrations in the fuel stream or by adjusting the volume fraction of thorium in the core. 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.

In these calculations the second option was employed. S igh t

I n sane instances it was desirable t o specify the fraction of neutrons that would be allowed for losses in xenon, me1 fission products, and leakage. condition could easily be treated on the ERC-5 code by specifying a fictitious atomic concentration and absorption rate coefficient tha t gave the desired ab-

This

sorptions. their amaunts are controlled by predetermined processing rates. trolled by the reflector design.

Xenon and flrel stream fission products were treated this way because Leakage is con-

<

I

1

I

1 -

5.0 rn STREAM m s o n FRACTIOR CALCULATIONS

2.1 Poison Fraction

The to t a l poison fraction generated by fission products i n a reactor includes the contribution t o neutron losses from fuel I , stream plus fe 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 to ta l fissions and to sim- plie the calculations, the to ta l poison fraction was assigned t o the fuel stream.

~y definition,

c Poison fraction = 1

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

= volume fraction of fuel stream i n core, f1,l f = volume fraction of fuel stream in en8 blanket,

2 182

= average effective neutron nux in core, neutrons/cm -sec, @l 2 * average effective neutron flux i n end , neutrons/cm -see, @2

2 E effective absorption cross section for i-th atom, cm , = to ta l fission rate in react

a =i

Ft V = neutrons born per fission, -

I) = neutrons born per neutron

V = to ta l a e l stream volume, c

The atomic concentration, can be expressed in terms of known quantities Equathg the production rate by considering the steady state o f t h e i- th isotope.

t o the sum of ell removal rates, there obtains

- 44 -

Ni,lEi Ti

- r = (f 4 Ftyi v %,A + %,l 1,l 1

The v4ue of N from Eq. 6 can be substituted into Eq. 5 to i,l ~~

in

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

-1

yi in the chemical processing schemes),

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

Ti

i

= efficiency of removal of i-th isotope in chemical processing,

= 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. as T and Tld, the values being characteristic of the chemical behavior of an atam in processing. accomplished in 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. restriction that Tld must be greater than T1. be several times larger than T1.

In MSBR processing, Ti assumes two values, identified

1 T1 refers to those fission products whose removal is

The time Tld is in&@.?ndent of the time Tl; there is, however, the In the economic cases, Tld will

The total poison f’raction attributive to the fuel stream is the solution to EQ. 7. Through this equation the total poison fraction I s related to the

L-

cycle times T1 and Tld and thereby to the capital investment in the processing plant and the replacement cost of the fuel salt. to optimize these costs for a given poison fraction by the appropriate choices

Furthermore, It is possible

of T1 aad Tld. l lh ls optimization was made in this study. c .

- 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 m. 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 are known except the term 60. 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 = th # (8 1 Ft cc

Qc V/DC I

in which u: is the thermal microscopic absorption cross section for carbon asd Dc is its thermal disadvantage factor. above in E q . 5. *

necessary to multiply both sides o f Eq. 8 by the the& absorption cross

The other quantities were defined

If it is desired to treat fission product6 absorbers, +it is only

section, u?, to obtain the absorpticm rate. The 4gf v y so obtained nay be used in Eg. 7 in camputlng the poison *action. pessimistic - but more realistic - comgutation sorptions and In same msnner adjust the th flect %hese resonances. An effectifre‘cro~ dlculated for each

On the other hand, a more ude the resonance ab- os6 sections to re-

fission product by includhgthe res in the following manner:

(9) eff =i

- 46 -

where, c 2

(RI)I = resonance integral for 19th nuclide, em

-1 zf = macroscopic fission cross section in reactor, cm , -1 6Ct E: slowing down power in reactor, cm ,

f = fraction of total fissions occurring at thenaal energy,

V p: MzPiber neutrons born per fission.

Both sides of 4. 8 can be

are computed by the GMU code for the IBM-704.

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

eff ,eff r:

'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 Wterms of EQ. 7. These are

(u 1 eff eff @th,l ui th

which is substituted for the term O1 @:, and

which is substituted forathe term O2v; . The solution of E q . 7 revised by 4 s . 11 and 12 is the desired poison fraction.

The value of " for this reactor was 0.6013.

i

-- 47 *

L, i

r

:

.

Resonance Integrals. Values of resonance integral.6 have not been reported 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. When EL calculation was made, the method for tiflnite'dilutlon described by Dresner 17 tf+s used.

5.3 Fission Products Included in Poison I&action Calculation

The fission pimducts’used in the ‘&oison fraction cs3culations were those ‘3 recommended by Rurch, Campbe& and Weeren. 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 Srnetsls that 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.

5.4 Gas:&dsg '.

and Effective Yield ~. ,’

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

- 4 8 -

Table 5.1. Fission Product Nuclides Included In Poison Fraction CalcuJ-ations

Nuclide Thermal

Atoms removed by plating on wall6 or by gas sparglng m-103 Mo-95

stable n

1% 13.4

n

n 0

I f

&=io1 2.46 M0-n 2 Ru-102 1.2 RU-104 Ob7 Mo-100 0.2

1-129 11 stable

. n

1-13 600 O.WWO-7

0.029 1000 0.064 101 0.05 13.1 0.062 12.2 0.042 26 .? O.Ol8 15.8 0,065 6.3 0.029 25 0 . a 25 0.0025 o.oCu3

n w

1-127 6

*-93 4 W 0.063 Br-81 2.6

&-9l 1.5 0 . 666x10~~ 0.059

W-lZ o.162d06 0.OOOl

. Atoms removed by precipitation in HF-dissolution

Gd-E~-155(~) .7 d o 5 0.0003 1 3 2 6 Sm-149 .5 ~ l 0 5 0.007 0.2lgxlo S m - E i l - 1 ~ 7000 0.0033 33.15

stable 0 . 0 ~ ~ - 3 1512 0 . o y 37.4

Ell-153 400

0.0021 2850 Ed-143 290 sm-152 1-50

Nd-145 52 Pr -141 11 Nd-146 9.8

tu

0

Rm-147 60 o.845xlo4 0.015 20%

n 0.022 lO(C)

stable . 0.029 3 0 (s 0.0% 16

8.8 4 08 3 1.8 1.4 0.6 2.9

I f

n 0.38 Om 16.1

- 49 -

t *

5

Table 5.1. - (Continued)

Nuclide Thermal Decay Yield Resonaac

(barns ) (sec -1) (barns) Section

cross Constant

Atoms removed by fuel

0.25x1d stable 130 O,149xlOa6

stable 84 30 15 2 2 0 ~32X10-9 1 0.7 0*3

n H

(I

stable w n

41.9 193 1396 198 375 0 347

0.0021 8.6

10 "7 c )

(a) Considered together because cross sections and/or yields are about the same.

(b) Yields are Wusted t o reflect gas sparging of gaseous precursors on a 6-minute cycle.

(c) Assumed value of resonance integral since no data for calculating available .

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

9 'decay

'decay 4- 'sparge

where the terms des iaa te the decay rate and the sparge rate. yield then becomes

The effective

- (13 1 "decay

+ Effective yield re: (actual yield) 'decay 'sparge

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

In 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 th 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 in 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. 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 as fissick product absorptions per neutron born.

5.6 Fission Product Resonance Absorptions Incluiied in Poison Fraction Calculations

%e curves represent

A second set of curves, Hg. 5.2, was constructed fromthe solutions of 4s. 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

/..

i

- 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

Mgures 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 several compatible values of TI an8 Tla were chosen, and the tdal f i e l cycle cost was calculate& for each pair of values. The cycle time, 5, Influences the fuel cycle cost thrcwgh the capital investment in 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, Ta, (Section 6.2.1) and the optinnlm cost and corresponding cycle times were determined.

-- id

t

- 53 -

- 54 -

- 55 -

- 5 6 -

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,

1. 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. vaue of epithennal fission cros6 section 0f U-233 6. addition of &Fk t o stabilize fiel salt

The investigated paraneters were:

poison fraction i n fuel stream

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. 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. fission cross sections of U-233 (Item 5) , it vas necessary t o establish new I 8 cr i t ica l conditions for the reactor using the GmU code before the ERC-5 calculations could be made.

The remaining four items were examined for

Homer, i n making the calculations for severd values of epithermal

Reactor properties that were held constant during the parametric study were the f’uel stream volume (530.2 f t3) and station power (2364 Mwt). The effective carbon-to-urmium ra t io 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. 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.

These values are the actual C:U ra t io divided by the t h d

- 5 7 -

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 rate coefficients might no longer apply, it was.necessaryto repeat the Cornpone unit ce l l l a t b n s with new concentrations t o develop new sets of reaction rate coefficients (see Fig. 4.1).

finite reactor calcu-

Several items i n the neutron balance were specified for all of the parametric studies. leakage and fuel processing. spectively 0.0008, 0.0043, 0.0016, and 0.0022 neutrons lo s t in fuel.

These were the neutron losses t o corrosion products, delayed neutrons,

r neutron absorbed !l%e values adopted for these quantities were re=

Corrosion product losses were estimated f'rom the equilibrium concentration of corrosion products of INoR-8. of WeJker?' Leakage losses were estimated frm design considerations; it was

felt 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

Delayed neutrons were calculated by the method

Pertinent characteristics e MSBR on which equilibrium cd.culations wereamade are l isted in Table

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 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'rac- tions of" each stream sold as product a t h e fraction of to ta l fissions occurring in the f e r t i l e stream.

he contribution of inaividual items t o the

- 5 8 -

\ .-.

Table 6.1. Characteristics of M8BR id

me- +mer, M 2364 -

ors in station 2 mennodynamic efficiency 0.823

265.1 Fertile stream volume per reac 3000 Valume fraction fuel stream i 0.16 Volume fractian fuel stream in end blaalret 0.16 Volume fraction fertile stream in side blanket 0.90

Volume fraction fertile stream in end blanket 0.74 Volume fraction graphite in side Xlanket 0.10 Volume frclction graghite in end blanket 0 010 m@l stream holdup time in reprocessing, days 1

stream volume per reactor, ft

eam holdup time in reprocess%ng, days 1

llhe option used t o achieve c r i t i ca l i ty in the EEiC-5 CCiLculations was

variation of the volume fraction of f e r t i l e stream in the core. Since the volume fraction of f'uel stream was Flxed at 0.16, the addifion or subtraction of fertile stream vas made at the expense of removing OT adding moderator. Consequently ever these small variations in CtU r a t io Bid not significantly affect the neutron spectntm.

CtU ra t io in the core varied slightly from case t o case. How-

!&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 . ref lects the most economic rate for the chemical processing for the chosen values of poison fraction and fertile stream cycle tfme. optwzea . We fuel s a t discard cycle time has also been

1

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 just optimistic specid cases of the resonance cy,

- 59 -

b ansorpyon calculatlolls. For the resonance absorption cases the fuel stream i 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

Resonance Absorption 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+-233 concentration 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 ‘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 .mean q 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.'

‘.

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

u of fertile stream in the core decreases about 13s. . _

,--

Losses t o samasium and other fission products increases by 0.09 neutrons u consuming about 708 of the breeding gain. Higher isotopes of uranium build up because of slower processing rates at the higher poison *actions with the accompanying decrease (- 0.005) in the mean wCLue of 1). U-236 and Ep-237 increase by about 0.004 and

em Inventory_

Neutron absorptions by

ory of fissionable materials, which des U-233, U-239, and -0233, for the equilibrium reactors is presented in Fig. 6.1; a detailed breakdown of the inventory is given in !tables 6.2 - 6.3. asd l/v absorption cases the inventory does not,change very fast wi& poison fraction over most of the range of poison fractions. the poison fraction a sharp upturn in the inventory is observed. because of the increased holdup in . the chemical processing plant at these fast processing rates. a poison fraction of 0.02 is about equal t o the rate for the l/v absorption cases

For both resonance absorption

However, at low values Qf

This occurs

The proccessing rate for the resonance absorption cases a t

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. days, the t o t a l fissionable inventory increases about 505, or fram around 860 t o 1240 kg.

in the 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 a c t the longer cycle times causes the c r i t i ca l mass of U-233 in the core t o decrease. creases for increasing f e r t i l e stream cycle t ine at constant poison f’raction, the purge rate of U-233 becomes smaller because less U-239 is routed t o sales; hence the U-235 inventory in the core builds up.

As the f e r t i l e stream cycle time increases from 35-20

The increase is attributed almost entirely t o increase of U-233 inventory

However since %he breeding gain also de-

%e thorium inventory for this series of calculations was maintained constant a t 270 tomes.

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 volati l i ty process, it builds up un t i l i t s decay rate is exactly equal t o the U-233 production rate. Therefore the Pa-233 inventory will change in direct u proportion t o the breeding ratio. For th i s system the inventory is in the raage 100-110 kg.

. 1

C

L b

h

- 62 -

6i1.4 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 becomes a significant part of the total cost; at a 350day blanket cycle

-time about 308 of the cost may derive from fertile stresm processing‘

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$, are not an appreciable portion of the fuel cycle cost, being only 2 - 45 of the total. On the other hand, fuel carrier replacement charges become a'slgnificant factor especially at the lower values of poison fraction 1 because of the high salt discard rate. At very high poison fractions (0.065)~ this contribution is only about 4$ 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

'.

a

- 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 f’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 corr- spending 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,com- b&&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 proc- eising . The result is a lowering-of the fuel yield.‘ 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.

U NCL A 9 9 I Fl E b ORNL- LR- DWG ~ 1 6 4 9

Ld 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 at 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. 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 fuel cycle time determined the capital investment in the processing plant3 the fuel salt discard cycle time determined the replacement charges for the fud sat. A plot of fuel 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.

. For each selected value of the fuel stream

When the optimum fuel salt discard cycle time is entered into Fig. 5.2, the opkimum fuel stream cycle time is found.

Optimum values of the 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 as a function of the fuel yield in Fig. 6.4. lated 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

*el cycle costs are high, am must be processed at a relatively fas t rate. as

Fuel yields of the order of 7$/yr can be attained

The curves have been calcu-

1

Likewise cated by the lower values of the poison fraction.

a f'uel cycle cost of around 1.7 mili~/lrwhr.

In the lower range of fuel yields cycle costs of 0.75 t o 0,80 mills/kwhr can be attained at require fertile stream cycle times of 150 - 200 days.

elds of 1 t o 2$/yr. These conditions

1 The curves could have been * extended in the lower regions of fuel yields by performing calculations at higher

values of the poison f ionj however, uncertainties in basic data, e.g., cross ~ sections and resonance integras , would lend daubt a s t o whether such a system

would have a positive breeding gain, Lj

- 66 -

I.

m

, Y

9.

- 6 8 -

v-

In t h i s series of calculations fuel stream cycle times ranged from 12 - 84 id days, and fuel salt discard cycle times ranged that fa l l on the envelope of the curves.

om 145 - 1.550 days for the case

' The dashed envelope curve has been drawn t o indicate the estimated maximum . performance of t h i s reactor. It might be possible t o extend the envelope of the

family of curves out th i s far by modifications i n the C:U ra t io 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 th i s study; nevertheless, it is believed that the chosen C:U r a t io is near the optimum.

Along a l ine of constant f e r t i l e stream e t i m e , the fuel cycle cost drops rather sharply from i ts 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. i n i t i a l drop i n fuel cycle cost, the breeding credit is also decreasing, but the in 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 in i t i a l drop in fuel yield is not a s fast as for the fuel cycle cost. processing time becomes long (increasing poison fraction) the savings on the processing plant and fuel salt discard are not so effective in lowering the fuel cycle cost and the rate of decline decreases. celerates 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

During th i s

Eventually, though, as the

Meanwhile decreasing breeding gain ac-

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. of. the fertile stream processing plant t o the fie1 cycle cost decreases with increasing cycle time while the inventory charges increase. In i t ia l ly the savings

The contribution

essing plant overweigh the increased inventory charges resulting i n a net lowering o f t h e to ta l fuel cycle cost, A t the long cycle times, however, the inventory charges overbalance the lower plant costs and the fie1 cycle cost

i n i m and begins t o rise. F

bsi ecrease in fie1 yield along a l ine 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 that higher U-235 concentrations at the

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 which 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 0.0l l - 0.065 neutrons absorbed i n fission products per neutron absorbed In fuel. fertile stream cycle time parameters of 35, 50, 75, 100, 150, and 200 days and is broader than that considered in Section 6.2 for the resonance absorption cases, When fission products are treated a s 1/v absorbers, the poisoning effect is not a s great as when the resonance absorptions are included, and it is therefore possible t o extend the range of calculations t o lower poison fractions before intolerably short fiel stream cycle times axe reached.

This range af poison fractions was applied t o

The equilibrium calculations were made us- the ERC-5 code for the IBM-704 by varying the fert i le stream cycle time and by using ficti t ious fissium concentrations and reaction rate coefficients a s mentioned in Section 6.2. These calculations were performed at the beginning of th i s thorium breeder study before a campilation of resonance integral data became available, &d it appeared that treating the fission products as 1/v absorbers was the best approach t o the

problem. optimistic upper limit t o the fuel fuel. cycle cost.

merefore, the results reported below should be considered as an Id and an optimistic lower l i m i t t o the

6.3.1 ~ ~ a n o m i c Performance

The optimized fuel cycle costs obtained. by the procedure are 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. bound t o the MSBR fuel cycle costs3 more real is t ic performance is tha t repre- sented by Fig. 6.4 in which the best available resonance absorption data were

Consequently it is believed that these curves represent alower

For the conditions of Fig. 6.5, fuel yields as high as about 8$/yr at a

4-/ 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 at about 4$/yr fie1 yield. ’ The

UNCLASSl F IED - 70 - ORNL-LR-DWB 68652

i

curves rise linearly after attaining their minima-because of the.lnfluence of the constaut charge for the batch fuel stream'processlng plant. Batch operation becomes effective 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 comments made above 'in Section 6.2.2 and [email protected] repeated here,:

. : :-../ 6.4 Effect on Reactor &rfbrmance of Varying; Thorium Inventory

In order to study the breedingperformance of the MSRR over 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 stre& cycle t&&.was 20 days, and the fuel sslt discard time -.I i was lm days. Tpe thorium inventory Xas varied by adjusting the fertile stream -. volume In the.ra&e 2CCC-,~9COC ft3 per St&Ion; .ThiS particular series of,calcu- latlons 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 various fuel yields., How- 1 ever; the dependency of 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 ' (tonnes

--: Relative'I&l Yield :-

oloo) ~..; , ,"l,o

Relative Fuel Cycle Cost

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 in- z

fluence 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 reactor used‘in this study when the blanket thickness is 3 ft. In the corresponding 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& $.lculated. Ac-

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

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 MSBR containing 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 maximum performance of the MSBR .I at this low thorium inventory. The solid outer curve was then drati to indicate the

,

6m+

f

- 73 -

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 ie1 Stream Processing Cycle Times

c

Thorium fnventory, tonnes

270 140 270 140 FM. Yield, $/yr Diff . Fuel cycl e cost, m i l l s / k ~ h r Mff

2;o 0.3 1.7 0.83 0.63 0.22

2i5 0.7 1.8 0.84 0.63 0.21 3 i6 1.7 1.9 0 -87 .0.66 0.21

446 2 -7 1.9 0*95 0.75 0.20 5i6 2.7 1.9 0.95 0.75 0.20 6.6 4 .7 1.9 * 1.n 1.37 0.20

estimated limit of m.ximum performance of the MSBR when the thorium Inventory is optimized with respect t o fuel yield. improvement in the reactor can be found by a slight variation in the C t U ratio. The CrU ra t lo (- 5000) in th i s reactor was not optimized, but th i s value is believed t o be near the optlmuzu.

!the outer curve a lso assumes that a slight

me 180-tonne thorium case is also plotted in figs. 6.7 and 6.8 for optimized. Aiel cycle times and for a raage df poison fractions fram 0 . a - 0.065.

6i3 =feet of Value of q-233 on 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 set of cross section values that is used. case because approximately 304 of to ta l fissions occurs at epithermal energies.

This is the

Reported epithermal values of 3.233 apparently agree within about average or "recmended" set of values

10% of an 4

1 . 5

1 . 4

1.3

c v) 0 1.0 0

w J 0 $ 0 . 9

J W 3 LL

0 . a

0 . 7

0 . 6

0.5

- 74 - UNCLASSIFIED ORNL-LR-OW0 69090

I I I I I

I 2 3 4 5 6

a

F U E L Y IELD ( % / y r )

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

. . . . .

4

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 In the fie1 yield, as measured by the horizontal distance between corresponding points on the curves, between the recommended curve and the high and low epi- tXermaJ. 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. cgJ. difference between corresponding points on the three curves. Is approximately

This effect is measured by the verti- %is difference

0.06 mills/kwhr from the recommended eurYe.

The various vaues of q-233 are tabulated in Tclble 6.6. .

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. 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. 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.

The column headed

The columns

The other values headed q(RPI) are presented for camparison since these

6.6 Effect of Value of Pa-233 Resonance Integral on MSBR Performance

A second nuclear constant value of the resonance integra have been mentio 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. resulting curyes f o r the reactor performance are plotted in Fig. 6.7 and 6.8.

accurately is the e6 of 600, 900, and 1200 barns

r the value of this integral. As menti0ned.b Section 3.7,

m e

. _ ,

..

- 78 -

Ta;ble 6.6. Group-Averaged Eta Values of U-233

Group mergy (ev) q (-10%) q (MTR) q (+lo$) q(m1

1 4 x lo6 - 1 x 10 3 .os 3.39 3 *729 3-39 2 2 x lo6 - 4 x 10 2.583 2 -87 3 .I57 2 .m 4 3 105 - 1 LO 2 0259 2.51 2.7a 2.51

7 6

6 3 1 x lo6 - 2 x 10 6 2 ;42l 2.69 2.959 2.69

2 -133 2.37 2.607 2.37 5 1 105 - 3 10 5 6 3 x10 0.1 x 10 5 2.052 2.28 2.508 2 0 2 8 4

7 1 lo4 - 3 10 2.025 2 -25 2.475 2.25

9 1 103 . 3 LO 3 2.025 2.25 2.475 2.25

4 4 ' 8 3 103 - 1 x 10 2 .m6 2.24 2.464 2.24

10 4.00 - m XI. 150 - 400 12 100 - 150 13 go - loo 14 80 - go 15 65 .. 80 16 50 - 65 17 45 - 50 18 37 - 45 19 33 - 37 20 30 33

1.9 1.72

_' h u

1.78

1.54 1.68

1.77 1.93 1.96 1.7 1.97 2.06

I

21 25 - I 30 1.944 2.16 2 376 1.9. 22 20 - 25 1.944 2.16 2 376 1.93 23 17 - 20 1.944 2.16 2 0376 1.75 24 13.5 - 17 1.953 2 017 2 387 1.88

25 10 13.5 1 a 9 5 3 2.17 2.387 2 0 0 6

26 7.5 - 10 2.032 2.28 2 1.91

27 5.5 - 7.5 1.863 2.07 2.277 1.96 28 4 5.5 1.962 2.18 2.398 1.99 *

29 2.5 - 4 1.845 2.05 2 0255 1.96 30 1.4 - 2.5 1 .n9 1.91 2 .lol 2-07 *

31 0.8 - 1.4 2.12 2.23 2.342 2.23 32 0.6 - 0.8 2.29 2.29 2.29 2.29 33 thermal - 0.6 2.28 2 0 2 8 2.28 2.28

34 thermal 2.28 \ 2 0 2 8 2.28 2.28

--_ u

- 79 -

The improvement that the lower a u e o f t h e resonance integral makes in tlie MSBR performance is hardly appreciable. 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. ever of all the neutron absorptions in Pa-233, approximately 805 occur at epithermal energies.

The fuel yleld, measured by the

How-

!Che comparative N-barn resonance integral calculations were performed using the ERC-3 code with adjusted values of the reaction rate coefficients, which are defined above in Section 4.4. data from a Cornpone f in i te reactor calculation giving the absorptions in Pa-233 as a function of energy. rately for the thennal absorptions and the epithermaJ. absorptions, t he epithermal value being decreased by the r a t io of the resonance integrals, i.e., 900:1200. The two values for the coefficients were summed t o obtain the to ta l reaction rate coefficient as shown by E q . 14.

The adjustment was made using output

The reaction rate coefficients were calculated sepa-

reaction rate 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 in core, atoms/cm 3

A similar expression was used t o calculate C(Pa) for the blanket, and these adjusted coefficients were used in the equilibrium reactor calculations.

6.7 Effect on MSBR Performance of Adding ZrF4 t o Fuel Salt

Recent developments in fuel technology for the Molten Salt Reactor Experiment

I

(MSU3) have indicated that f ie1 s tabi l i ty i s enhanced by the addition of nominal amounts of ZrF4 t o the fuel salt. Zirconium acts as a “getter” for oxygen

preventing the union of oxygen and uranium which results in the precipitation G of uranium oxides. neutron poison t o the system.

The inclusion of zirconium, however, adds an a d d i t i o d f

- The effect on breeding rat io and fuel yield of adding 5 mole $ ZrF4 t o the

fiel salt was calculated for the ranges of values representative of the MSBR. The results are plotted i n Fig. 6.9. breeding rat io and fuel yield resulting fYosn the addition of ZrF4 as a S'uncticm of these quantities for a reactor containing no ZrFd. the detrimental effect of the ZrF4 is more pronounced for the reactors that have ini t ia l ly poor breeding performance. curve at fuel yields of the order of 2-3$/yr suggests that adding ZrF4 t o a low performance reactor can just about destroy i ts breeding potential.

The curves show the per cent decrease in

As m i g h t be expected, I

I

In fact the steepness of the fuel yield

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 fert i le stream cycle time of 50 days. according t o fuel strew cycle time and fuel salt discard time. difference between corresponding points in Flg. 6.10 shows that Zr decreases the

f'uel yield about O.5$/yr; whereas, the f'uel cycle costg measured by the vertical difference, is almost negligibly affected. The effect on Are1 cycle cost i s small because the effect shows up through the loss in breeding credit which is

The results were optimized The horizontal

not a large portion of the to ta l -el cycle cost.

5

.

- 81 . UNCLASSIFIED

”

7.0 CO~CLUSIO~S

The molten salt reactor offers considerable promise 8 8 a breeder i n the Th-U cycle, which use thorium in the form o f t h e oxide or metal , is in the simplified chemical processing method. fluoride volat i l i ty 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 ight of current technology, resigned t o the more complicated and ex- pensive Thorex process. , 8

The principal advantage of this system Over other breeding systems,

The molten salt system is able t o use the relatively si@e

The MSBB is capable of excess f i e l yields up t o 7$/year when operating 80s of the time. A t this hlgh yield, the Arel cycle cost i s about 1.65 mills/kwhr. A t lower fiel yields the fuel cycle cost is considerably improved, dropping t o perhaps 0.65 mills/kwhr for f ie1 yields of l-2$/year. However, yields as low as th i s constitute margiaal operation because uncertainties i n nuclear data introduce uncertainties of about a per cent in 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 is approximately

1

0.9 mills/whr.

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 fuel yield) and about 308 of the cost at the low processing rates. Another item that 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 l c r w 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 . 3 Thorium inventory for the 6000 f't fertile stream amaunts t o 8.179 of the cost.

Since the thorium inventory is constant (270 tomes), its 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 the processing rate of the fuel stream over most of the range of poison fractions studied. processing plant begins t o contribute importantlyto the inventory. factor increasing the inventory is the fert i le stream cycle time. 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 f’rom 1.346 t o 6.6$ of the to ta l fissions. The fissionable inventory accounts for about 19% of the f ie l cycle cost at a fiel yield of 1.5$/yr and for about 446 at a yield of 7$/yr.

However, a t the fast fuel stream cycle times the holdup in the chemicai The principa

In going from

c

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

fuel Yield m Y V-Y 8s much as * 2.5 t o * 3$/yr for variations of value of the epitherma3 cross sections of U-233 f’rom the set used i n these calculations.

Equilibrium reactor I calculations for a representative set of operating conditions 3,ndicate that the ~

lo$ i n the

The inclusion of 5 mole $ZrF4 i n the fuel salt t o enhance s tabi l i ty de-

creases ,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.

There are two ways of improving the breed- performance o f t h e MSBR. These

are (1) determining the optbum C:U ra t io and (2) increasing the thorium inventory in the blanket.

approximately 900 used i n these calculations is near the optimum and that only a very slight improvement might be expected by changing the reactor camposition. The most significant improvement in the MSBR breeding performance can be made by increasing the thorium inventory i n the fert i le stream. Pa-233 competes with thorium for neutrons; hence the losses t o Pa-233 are inversely proportional.to the thorium concentration.

In regard t o the C:U ratio, it is believed that the value of

, , .

In the blanket, I

However th i s improved breed-

i.

I - L d

ing performance comes at 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 th 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., leakproof graphite-to-metal joints and Impervious graphite that permits minimum xenon

absorpt;ian. chemical processing, further demonstration of the fluoride vola- t U i t y process and the HF dissolution process is necess design inf'ormation.

*

t o supply adequate

- 86 -

8.0 ~ E N C E S

1. L. 0. Alexander, et al., Thorium Breeder Reactor Evaluation. Fuel Yields and me1 Cycle Costs in Five Thermal Breeders, EXL-CF-6i-3-9, March 1, 1961.

Part I.

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

3. 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.

4. 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. a. 5 66 (1960).

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).

7. E. 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.

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

9. W. E. Kinney, Cornpone - A Multigraup, Wt i reg ion Reactor Code, ORJL-2789, i n preparation.

10. L. G. Alexander, WC-5 Program for Computing the Eguilibrium States of TKO- Region, 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 Fuel Thermal Breeders as Related t o Cost and Feasibility, ORNL CF-60-5-93,

I May 12, 1960.

15. E. R. Payne and J. C. Moyers, Determination of Capital Costs of Steam Cycle Equipment and Over-all mast Efficiency for Three Breeder Reactors, unpublished data.

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 Insti tute Report, June 1960.

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

Part-I.

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

a

'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) 35 75 Case No. 1 2 3 4 c i&&n fraction 0.02 0.04

Volume fra&lon fertile stream ln core 0.068 Volme fraction grabhite in core ::z 0.772 Carbon: Uranium ratio in core WO lkel stremncycletime (days) mel saltdlscardcycletime (days) E5 $0 Fractionof fuel stream soldas'product 0.0035 O.Ol.04 7!kactioti of fertile stream sold as prodkt 0.aa2 0.0134 Fraction of fission in fertile stream 0.0139 oAl33

Wel stream composition (atoms/cm3)(10~ 24 th) ) U-233 o.aagiE-4 I 0.8855E-4 U-234 0.2516E-4 0.26l.5E-4 y& '

m-237 . $

y74mE~

OhlE-7 :z$iE

Fissi (a)' 0:2a4?E-6

Xe-135 b, "p carrier(c)

o.&gE-g O.lrgoE-g 0.208~4 0.2oap-1

Fertile stream co&position (atoms/cm3)(10'24)

Z-233 0.4012E-2 0.4O'U?E-2. O.Z@&E+ OL&lE-5

U-233 kE2g;

O.l4OlE-5 U-234 O.lOl3E-7 V-235 Fissium(~)

0.3444~10 0.3444~.10 0.18923-5 O.l&@E-5

sfn-1% 0.37163-g

z;zr(d) .

Note: See end of Table 6.3 for footnotes.

0.06

0.066 0.775 5050 a4 1550 O.oog4

:%I% .

0.02

oJJ7* 0.770 5055

if5 O.OOjl

0.0245 0.026g

o.ahE-4 0.8iW33-4 0.2729E-4 0.2554~-4 oof2z . yws-5

0.54703-6 O&E-7

o.iz&hg- O.l2/OE-9 0.20853-1 0.2085!%1

0.40123-2 O.l554E-5 0 ;136oE-5 0.9m-B 0.3444E-10 O.l?I&E-5

0.4&-2 0.16393-5 0.30323-5 o.rlllE-7 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

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

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

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

Case No.

Neutrons absorbed by listed element per neutron absorbed in fuel

Th Tn fissions

U-233 U-234 U-235

pa-233 x 2

U-236 NP-237 Xe-135(b) Sm-1% + Sm-149 Fissium Carbon

Th carrier Corrosion products Delayed neutrons Leakage Fllel processing

Fuel carry&]

Neutrons born per fuel absorption (?e) breeding ratio

!hue 6.2 - cont'd 1

0 . m 0 .mg 0.0ug 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 .a6 10w53

\

.h c

c G

Inventory per station (kg) Th In fertile stream 263,000 263, 000 263, ooo Th in processlng 7500 7500 7500 Total Thorium rl0,500 no, 500 270,500

U-233 in fertile stream 94.6 Pa-233 in fertile stre 108 U-233 in fertile stream processing 2 07 U-233 in &el streem in reactor 196 U-235 in fuel stream in 19.5 U-233 In external f'uel circuit 274 U-235 i n external &el circuit 27*2 U-233 in Atel processing 43.1 U-235 in fuel processlng 4*3 U-233 in fuel reserve €8.4 U-235 in fuel reserve 6-7 U-233 + U-235 in f'uel aUmp tanks s.6 Total fissionable inventory 896.1

92.2 ,105 2.6 195 20dr 273 28.8 10.1 1.1 67 -7 7-1 3 -7 854.9

89.5 102 2.6 195 21 -9 rn 30.6 6.1 0.7 66.9 7.5 s.8 845.6

4

199 108 2.7 193 19.9 269 V.8 42.4 4.4 67.0 6 09 s.0 991.1

3% 13 93 6.0

5

263,000 3500 266,500

1% 105 2.5 192

29-3 10.1 1.1 66.4

18.3 4.4

6

263,000 3500 265,500

189 102 2.5 192 22.2 267 31 .o 6.5 0.7 55.8 'f! 7.6 51.2 ' 937 5

I

3% 35.5 2.3

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&)

“tt

!Cable 6.2 - cont’d 1

0.076 0.132 0.023 0.166 0.69 0.261 0.030 0.456 0.034

Gross €bel cycle cost 1.832 Breeding credit 0.136 Met fuel cycle cost 1 e 7 0 2

Processing rates spent fi3/ay> 44. Thorium (kg day) 7500

39-7 Thorium replacement (&/day fie^ s d t replacement (&/by) Excess fissile atoms produced (kg/day)

I Fertile stream loading, (gm U-233 + Pa-233)/@ Th 0.7’7

2

0.073 0.132 0.022 O L 6 6 0.265 0.261 0.030 0.145 0.034

1.128 0.090 1.038

10.4 7500

0 075

0.072 0.132 0.022 0.166 0.207 0.2Q 0.030 0.040 0.034

0.964 0.647 0.917

39.6 18.9 0.076

0 *73

4 5 0

0.083 ’ 0.08 0 0.130 0.13 0.130 0.023 0.02 0.022 0.164 0.164 0.164 0.650 0.269 0.207 0.164 0.164 0.164 0.030 0.030 0.030 0.455 0-135

1.735 1.029 0 .lrl 0.088

0.034 0.034

1.608 0.941

3 44.2 6.a , 3500 3500 39 01 39.0 a7 17.3 0.203 0.072

1.2 1.1

c- m J

c

Table 6.3. Performance of a Molten Sa l t Breeder Reactor for Several Values of assion Product miss Fraction. Fission product

Resonance Absorptions Included in Calculations.

Fertile stream cycle time (days) Case No. 7 6 9 Poison fraction 0.02 0.04 0.06

Volume fraction f e r t i l e stream in core o.qo4 0.0678 0.0652 Volme fraction graphite i n core 0.770 0.772 0.775 Carbon: Uranium ra t io in core 50 5090 5090 ~ u e l stream cyde time (days) 12 56 86

salt discard cyde time (days) 1% 400 1500 Fraction of fuel stream sold a8 product 0.0029 0.- 0.0074 Fraction of f e r t i l e stream sold as product 0*02n O.Olg8 o.mq !&action of fissions in fertile stream 0,0348 0.033 0.035

fiel stream composition (atams/cm )(lo

100

3 -24 (h) )

U-233 0.8663- 0,8639- 0.86l4E-b

U-235 0.gOm-5 0-9533s5 0,1004E-4 0-236 Od.l.90E-4 0.14033-4 0 ~ 6 7 5 ~ 4

U-234 0.25353-4 . 0.2667E-4 0,2763i3E-L

m-237 (a) 0.663@-7 0 0 3 ~ - 6 0,5936~-6

> O.lr/OE-g 0.12713-9 0 . m - 9 Xe-135 Fissiurplb) carrier (C 0.208%-1 0.2085E-1 0.20853-1

200 10 ll 12

0.02 0.04 0.06

0.0698 0.770

U .8 150 0.0022 0 03-70 0.0647

5140

0.0673 0.773 5150 50 445 0.0062 0.0259 0

0.0647 0.775 5160 84 1500 0.0052 0.0132 0.0587

0.8346E-4 0.8340E-4 0.8334~4 0.26343-4 0 27l4E-4 0 2795E-4 0 9368E- 5 o 97813-5 0 e102lE-4 0.1327~4 0.15313-4 O.lTgE-4 I

0 *73m-7 0.334936 0.6l78E-6 W

0 .mOE-g 0.127.l.E-g 0.12723-9 0.20853-1 0.2085E-1 0.2085E-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

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

0.1%1~-5

Case No. Case No.

Neutrons absorbed by listed element per neutron absorbed in fuel

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

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

0-9325 0.0018

0.9915 0.9646 0.9853 0.mg 0.0018 0.mg 0.0~8 IO.aU2 0.0~6 0.a05

0.9046 0 .g109 0 9071 0.9030 0.0963 0.09%

0.939 O = ! w t 0,0933

0.0970 o.om9

0.0891 0.0929 0.0914 0.WS

0.0179 0.0954 0.0168 0.0133 0.01%

0.0861 0.0906 0.0120 O.Ol41

O.Wl3 0.0023 0.0003 0.0013 0.0024 0.00% O.OOg0 0.0050 0.0050 0.0050 0.0050 0.003

0.0003 0.0003 0.0003 O.OOO6 0.0005 0.00 0.0218 0.0418 0.0617 0.0234 0.0432 0.06 0.0286 0.0287 0.0288 0.0286 0.0287 0.02 0 0302 0 0302 0.0301 0.0302 0.0302 0.0301

0. 0191 0.019 0.0194 0.0190 0.0200 0 a95 O.OOO8 0.0008 O.OOO8 O.ood3 o.OO43 0.a3 0.0043 0.0043 0.0043 0.0043 0.0016 0.0016 0.0016 0.0016 0.0016 0.0016 0.0022 0.0022 0.0022 0.0022 0.0022 0.0022

2.2128 2.2~5 2.2101 2 .ulg 2.2l07 2.2096 1 . q 2 l 1.0487 1.0247 1.0682 1.042 1.0u5

O.OOO8 o.ooo8 '

c.:

Q'

Case No. 7 Inventory per station (kg)

Th in fertile stream 263,000 Th in processing 2600

' Tot& thorium 265,600 U-233 in fertile 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-235 in external fuel circuit U-233 In fbel processing

U-233 In fuel reserve U-235 in fuel reserve U-233 + U-235 in fuel dump tanhs T o t a l fissionable inventory

&el sat (fuel excluded) manket salt (m4 excluded)

U-233 In external fuel circuit

U-235 in fuel processing

263 108 2.6 191 20.1 267 28.1 42.0 4.4 66,2 6 09 50.6 1049.9

3,500 19'7,340

Doubling time (full power years) Fuel yield at 80% plant factor (Myear )

14.3 5-6

Table 6.3. - cont'd

a 9

263,000 2600 265,600

256 105 2.6 - 191 21.2 266 29.6 a *Y 1 .o 65.7 7.3 50 -7 1005.0

0.7 65.1 7.6

29,500 29,500 I¶, 340 191,340

20.3 39.5 3.9 2.0

10 11

c

507 107 2.5

20.8 184

257 29.0 41 .i 4.7 63.4 7-l 49.1 3-272 -7

31,400 196,400

18.3 4.4

494 104

257

9.6 1.1 63 .o 7.4 49.3

3O-3

1224.0

12

263, OOo 1300 264,300

481 10.l 2.4 184

62;7 7 -7 I

49.3 \D 1206.2 ul

55.0 1.5

/'

Case No.

f i e l cycle cost ( m i ~ . s / ~ ~ ) Uranium inventory Thorium inventory A ~ e l s a t inventory Fertile salt inventory

processing plant Filanket processing 'Barium amortization Fuel s a t replacement Fertile s d t replacement(g)

Gross fuel cycle cost Breeding credit

"tt

. Net fuel cycle cost

Processing rates spent ft3/aay) Thorium (kg I day) Thorium replacement (&/day) fie1 s a t replacement (@/day) Excess fissile atoms produced (kg/day)

7

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

1.689 0.125 1.564

44.2 2600 38.9 210 0.19

Fertile stream loading, ( e ~ n U-233 + Pa-233)/kg !Ch 1.4

Table 6.3. - cont'd

8

0,086 0.130 0.022 0.163 0.250 0.139

0.155 0.033

1.ooe 0.084 0.94

0.030

9.5 2600 38.9 74.0 0.135

1.4

9 10 11 12

0.035 0.130 0.022 0.163 0.207 0 0139 0.030 0.042 0 033

0.109 . 0.105 0.129 0.129 0.023 0.022 0.162 0.162 0.653 0.269 0.092 0.092 0.029 0.029 0.439 0.140 0.033 0033

0 .io3 0.129 0.021 0.162 0.207 I 0.092

0.041 0.033

0.929 8

0.853. 1.669 0.981 0.817

0.808 105% 0 903 0.780

6.2 4.4 09 10.6 6.3 2600 1300 1300 1300 388.9 38.8 38 -7 38 *7 19.6 m9 66.6 19.6 0.068 0.189 0.125 0.060 .

0.043 0 .u8 0.078 0.037

1.3 2.3 2.3 2.2

a ., n a,

I-

It;

i

kd

I

Footnotes for Tables 6.2 and 6.3

The element fissium is a conglomeration of fission products. A f ic t i t ious 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 is assigned t o the fie1 stream. tha t a gas purge of fuel and blanket solutions will maintain Xe-135 absorptions at 0,005 neutrons per neutron absorbed in f'uel.

Based on Li-7 atoms.

It has been assumed

The absorption cross section of the fert i le stream carrier was normXized t o the basis of one thorium atom.

U-235 in fertile: stream is negligible.

Includes thorium burned up in breeding plus thorium discarded on 20-year cycle.

Fertile salt is discarded on a 20-year cycle t o maintain blanket fission products at 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 is assunred t o contain L i that is 0.01 atom $ Li-6. Absorptions-are based on equilibrium Ll-6 concentration for this feed. -

Distribution

1-10. L, G. Alexander 50. G. M. Watson 11. S. El Beall t

12. E. S. Bettis 13. F. IF. Blanke 14. A. I,. Boch 55-56. Central Research Library 15. E. 0. B ~ h l m a n n 57-9. Document Reference Library 16. R. B. Briggs 59-68 . Laboratory Records

18. D, 0 , Caslpbell 17. W. D. Burch 69. ORNL-RC

19. W. HI Cam 20. w, L. carter 21. 0. I. Cathers EXTERNAL 22. R. H. Chapman 23. F. L. CuUer 70. H. w. ~ehnnan, AEC, 24. J. 0. Delene Washington 25. E. K. Ergen Brooks, Hasvard Uni- 26. D. E. Ferguson versity

28. W. R. G a l l 74. D. H. Groelsema, AEC, 29. H. E. Qoeller Washington

3. J. P. Hammond Washington 32. W. H. Jordan 76. L. Link, Argonne 33. P. R. Kasten 77. J. W. Miller, K-25 FI

34. B. W. Kinyon 78. R. E. Pahler, AEC, 35. J. A. Lane Washington 36. M. I. Lundin 79. B. E, Prince, AEC, 37. R. N. Lyon Washington 38. E. G. MacPherson 80. W. Robba, Brookhaven 39. W. D. Manly 81-82. F. P. Self, AEC-OR0

41. H. F. McDuffie Washington

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

2'7. A. P. Fraas 72-73. D. F. COP, AEC-OR0

30. W. R. G r i m e s 75. J. F. Kaufmann, AEC, *

40. W. B. McDonald 83. D. c. maas, m, 42. C. W. Nestor 84-98. TISE-AM=

Date post:	31-Mar-2016
Category:	Documents
Upload:	the-e-generation
View:	217 times
Download:	0 times

ORNL-CF-61-8-86

Documents