ORNL-2634.txt

 

 

 

2 Ao bl e

 

 

 

 

 

 

 
 

 

 

I.EGA}. N'oflcz'

 

          

. nmor the Comminion, rior uny porson ucting on beha!f bf the Cummusion.

HA. Makes - ‘any .warranty or repreunfoflon, Bxpress “or - implmf with rnpeci 4o tha accorccy, : R e
: S L ST compleuness, or usefulnoss of. tlu Informaflon comulned hr-thll uporf o cht ,the use of" S T T

: corrlrcctor of -the Commiuion to fhe a:fent thm s cH amployoe or conlracfor propares, l\andln

 

i op distribufos, or provldes access io, cny infcrmafion pursuant to. hls ""‘PIOYmem o Comracg -
Wm" fhe Commlssion. SR

 

 

 

 

       

   

 

 

 

 

 

 

 

 
 

Contract No. W-Th05-eng-26

REACTOR PROJECTS DIVISION

MOLTEN-SALIT REACTOR FPROGRAM STATUS REPORT

H. G. MacFPherson, Program Director

Date Issued

DEC1 1308

 

Previously Issued May 1, 1958, as CF-58-5-3

OAK RIDGE NATIONAL LABORATORY
Operated by
UNION CARBIDE CORPORATTON
- for the
U.S. ATOMIC ENERGY COMMISSION

ORNL-263L

 
 

O ——— R S e by e

 

 

 

ACKNOWLEDGMENTS

The following people wrote portions of Part 1, "Interim Design of &

' Molten~-Salt Power Reactor": L. G. Alexender, B. W. Kinyon, M. E. lackey,

H. G. MacPherson, L. A. Mann, J. T. Roberts, F. C. Vonderlage, G. D. Whitman,
J. Zasler. In addition, the following people made major contributions to the
reactor design and to the organization of Part 1: E. J. Breeding, W. G, Cobb,
J. Y. Estabrook, F. E. Romie, C. F. Sales, D. S. Smith, J. J. Tudor.

‘The authors of Part 2, "Properties of Molten Fluorides as Reactor Fuels,"
vere W. R. Grimes, D. R. Cuneo, F. F. Blankenship, G. W. Keilholtz, H. F. Poppen-

| diek M. T. Robinson. Substantial contributions were made by C. J. Barton,

C. C. Beusman, W. E. Browning, S. Cantor, B. H. Clampitt, H. A. Friedman,
H. W. Hoffman, H. Ifisley, S. Ianger, W. D. Manly, R. E. Moore, G. J. Nessle,
R. F. Newton, J. H. Shaffer, G. P. Smith, N. V. Smith, R. A. Strehlow,

C. D. Susano, R. E. Thoma, W. T. Ward, G. M. Watson, J. C. White.

Part 3, "Construction Materials for Molten-Salt Reactors,” was contributed
by W. D. Manly, J. W. Allen, W. H. Cook, J. H. DeVan, D. A. Douglas, H. Tnouye,
D. H. Jansen, P. Patriarca, T. K. Roche, G. M. Slaughter, A. Taboads, G. M. Tolson.

Part 4, "Nuclear Aspects of Molten-Salt Reactors," was written by
L. G. Alexander, the work reported there was done in collsboration with
J. T. Roberts.

- Part 5, "Equipment for Molten-Salt Heat Transfer Systems," was vritten
by H. W. Savage, W. F. Boudreau, E. J. Breeding, W. G. Cobb, W. B. MeDonald,
H. J. Metz, E. Storto. Other major contributors to the work reported were
R. G. Affel, J. C. Amos, J. A. Conlin, M. H. Cooper, J. L. Crowley, P. A. Gnadt,
A. G. Grindell, R. E. MacFherson, W. R. Osborn, P. G. Smith, W. I. Snapp, |
W. K. Stair, D. B. Trauger, H. C. Young.
Part 6, "Buildup of Nuclear Poisons and Methods of Chemical Processing,"”

"was vritten by J. T. Roberts, and contains contributions from G. I. Cathers,

D. 0. Campbell, and the guthors of Part 2.

The over-all editor of the report was A. W. Savolainen. The choice of
material included was the responsibility of H. G. MacFherson.

The authors wish to express appreciation to W. H. Jordan and A. M. Weinberg

" for their helpful guidance of the molten salt work.

 
 

gk

 

 

wrt

T

O 0 - O
.

CONTENTS

ACKNOWIEDGEMENTS

1.
2.

»

9.
9
9
9
9.
9.

0. @
© 10.1 Fuel Salt Reprocessing

PART 1. INTERIM DESIGN OF A POWER REACTOR

Introduction and Conclusions
General Features of the Reactor
Molten-Salt Systems

3.1 Fuel Blanket Circults

3.2 Off-Gas System

3.3 Molten Salt Transfer Equipment
3.4  Heating Equipment

3.5 Auxiliary Cooling

3.6 Remote Maintenance

5.7 TFuel Fill-and-Drain Tank

Beat Transfer Syeteme

Turbine and Electric System

Nuclear Performance

Procedure for Plant Startup

Reactor Control and Refueling

Accidents: Consequences, Detection, and Reqnired’Action

9.1 An Instantaneous loss of load from a Secondary
Sodium Circuit

9.2 An Instantaneous St0ppage of Sodium Flow 1in One
of the Primary Heat Exchangers-

- 9.3 An Instantaneous Reduction of the Heat Flow

Rate from the Reactor Core
4  Cold Fuel Slugging .~ = |
5 Removel of Afterheat by Thermal Convection
6 loss of Fuel Pump
7 Ioss of Electric Transmission Line Connection
. to the Plant
8'1’Ieak Between Fuel and Blanket Salts
9<
10
1

e

leak Between Fuel and Sodium -
. 1eak of Fuel or. Blanket Salt to Reactor Cell
Ieaks of Water or Steam to Sodium'”

'Chemical Processing and Fuel Cycle Econcmics

10, e}fBlanket Selt Reprocessing _
-10.3 " Cost Bases .

10.4 Chemical Plant Capital COSts
10.5 Chemical Plant Operating Costs
10.6 Net Fuel Cycle Cost

-iii_

11

s
VAN NO OV

18

22
2k

27
3k
3l
42
45

50
52
25
53

53
22
56
29
29

60

61
62
62
62

 
 

 

 

 

11,
12.

. 12.1 ' Alternate Heat Trensfer Systens

1.

o.

1.

- Construction and Power Costs

11.1 Capital Costs
11.2 Power Costs

fSome Alternates of the Proposed Design

.12 2 Alternate Fuels

63

68

69

T2

PART 2. CHEMIGAL ASPECTS OF MOLTEN~FLUORIDE-SALT REACTOR FUELS

Choice of Fuel ‘Composition

1 1 Choice of Active Fluoride
Uranium Fluoride '

' Thorium Fluoride

1.2 Choice of Fuel Diluents
Systems -Containing UFL
' Systems Containing ThF),
Systems Containing ThF) and UF)
Systems Containing PuFz

Purification of Fluoride Mixtures

2.1 Purification Equipment
2.2. Purification Processing

Physical and Thermal Properties
Radiation Stability
Behavior of Fission Products

5.1 F1351on Products of Well-Defined Valence

- The Noble Gases ,

'Elements of Groups I-A, II-A, III-B, and IV-B
.2 Fission Products of Uncertain Valence
«3 Oxidizing Nature of the Fission Processes

i\

PART 5. CONSTRUCTION MATERIALS FOR MOLTEN-SALT REACTORS

Survey of Suitable Materials
Corrosion of Nickel-Base Alloys by Molten Salts

2.1 Apparatus Used for Corrosion Tests

2.2 Mechanism of Corrosion

 Febrication of INOR-B

1 -Casting

2 Hot Forging

3 Cold Forming

4 Welding

5 3Brazing |

6 Nondestructive Testing

- iv -

s

Th

Th
75
75
75

82

90

%
%

91

91

93
100
10k
104

104

106
108
108

109

112

112
125

123
125 -

125

125

127
132

'l

 
 

 

 

 

o ~3 O W\

10,

2.

*

Mechanical and Thermal Properties of INOR-8

h.1 Elasticity
Plasticity

- k.2
| t Z Aging Characteristics

Thermal Conductivity and Coefficient of Iinear
Thermal Expansion

Oxidation Resistance
Fabrication of a Duplex Tubing Heat Exchanger
Ayailability of INOR-8

-Compatibility of Graphite with Molten Salts and

Nickel-Base Alloys -

- Materials for Valve Seats and Bearing Surfaces

Sunma.xry ofematerial Problems

PART 4. NUCIEAR ASPECTS OF MOLTEN-SALT REACTORS

- Homogeneous Reactors Fueled with P

1.1 Initial States
Critical Concentration, Mass, Inventory, and
" Regeneration: Ratio
'Neutron Balances and Miscellaneous Details
' Effect of Substitution of Sodium for Ii7T
Reactivity Coefficients
Heat Release in Core Vessel and Blanket
1.2 Intermediate States =
‘Withoiit Reprocessing of Fuel Salt
With Reprocessins of Fuel Salt

Homogeneous Reactors Fueled with U233

2.1 Initial States
2. 2 Intermediate States

Homogeneous Reactors Fueled with Plutonium

5 1 Initial States j?"" S '
» ,,jcritical Concentration, Mass, Inventory, and
folRegeneration Ratio C ,
:Neutron Balance and Miscellaneous Details

f 3 2 Intermediate States
7f3¢terogeneous Graphite-Mbderated Reactors

b1 Inttiel States

S fPART 5- EQUIPMENT FOR MOLTEN-SALT REACTOR I-IEAT TRANSFER SYBTEMS .

'—Pmnps for Molten Salts

1.1 Improvements Desired for Power Reactor Fuel Pump
1.2 A Proposed Fuel Pump

13k

134
134

- 144

1l

1b7
153

153
159
159 .

162
163
168

179

179
179
18k
185
185
185

191

192
- 192

192
199

199

199
199

201

. 202

206
208
209

 
 

 

ON 1 B W

 

 

 

- Heat Exchangers, fixpansion Tenks, and Drain Tanks 211
Valves 211
VSystem Heating | 213
Joints - 21k

' Instruments | 217

6.1 Flow Measurements . 217
6.2 Pregsure Measurements 217
6.3 - Temperature Measurements 219
6.4 Iiquid-Ievel Measurements 219
6.5 Nuclear Sensors ' 219
PART 6 BUILDUP OF NUCIEAR POISONS AND METHODS OF CHEMICAL PROCESSING
 JBui1dup of Even-Mass-NUMber Uranium TIsotopes 226
Protactinium and Neptunium Poisoning 023
 QFission.Pr5duct Poisoning 22k
. Corrosion Product Poisoning 226
* Methods for Chemical Processing 206
5.1 The Fluoride Volatility Process 228

- 5.2 K-25 Process for Reduction of UFg to UF) \ 251
5.3 Salt Recovery by Dissolution in Concentrated HF 2351

5.4 Rare Earth Removal by Exchange with Cerium 232
5.5

Radiocactive Waste Disposal

- Vvi--

236

 
 

 

 

 

STATUS OF MOLTEN-SALT REACTOR PROGRAM

PART 1

INTERIM DESIGN OF A POWER REACTOR

1. INTRODUCTION AND CONCLUSIONS

Thé general usefulness of & fluid fueled reactor that can operate
at high temperatures with low pressures has teefi recognized for a long
time. 'Therapplication'of the'molten salts to such a reactor system has
been discussed,l and the operation of the Aircraft Reactor Ex.periment2
demonstrated the basic feasibility of the system. Preliminary design
studies indicated that power reactors based on such systems would be
economicaiiy attractive. This study gives.a more detailed conceptual

~design and outlines operational procedures so that the problems of

handling a moiten-salt power reactor can be better visualized.
Particular attention has been given to the circulating-fuel system,
since this system and its associated equipment will be the heart of any

,molten-salt reactor plant. of perhaps lesser importance are the particular

reactor chosen for study (a_tfidéregion'homogeneous converter) and the par-
ticular heat transfer system (tfio sodium circuits in series). Alchough
later studies may indicate better ‘choices for the reactor and the heat
transfer system, those selected for this study are considered to be sound
and- to provide 8 good baS1s for estimating the cost of power from & molten-

: salt reactor._ -

 

13 Co Briant and.A. M4 Weinberg, "Molten Fluorides as Power Reactor |

:Fuels,“ Nuclear Bcience and Eng. g 797-803 (1957)

2% 5. Bettis, B. W. ‘Schroeder; Gs A. Cristy, E. W. Savage, R..G.

'fAffel, ‘and L. F. Hemphill, "The Aircraft Reactor Experiment - Design and
‘Construction," Nuclear Science and Eng. 2, 804825 (1957); W. K. Ergen;
A+ D, Callihan, C. B, Mills, and Dunlsp ScOtt,'“The Aircraft Reactor.
‘Experiment - Physics," Nuclear Science and Eng. 2, 826-840 (1957); E. ‘s.

Bettis, W. B. Cottrell, E. R. Mann, J. L. Meem, &nd G. D. Whitman, “"The
Alrcraft Reactor Experiment - Operation,” Nuclear Science and Eng. 2,
811-853 (1957). | =

L.

-

 
 

 

The reactor power station chosen for study has a gross electrical

capacity of 275 Mw and a net capacity of 260 Mw. Figure 1.1l shows an

isometric drawing of the principal portion of reactor plant, and the
most important of the reactor"statistics-arg'p:esented in Table 1l.1l.

It is estimated that this molten-salt reactor power station could be

built for 70 million dollars. At 4% per year interest and an 80% load
factor, the fiXed chargesj'including'fueleinventory rental, would amount
0 5.7 mills/kwh. Fuel and salt replacement costs of 1.7 fiills/kwh and
an operation and.malntenance charge of 1.5 mllls/kmh (1ncluding chemical
plant operation) lead to a total estimated pover cost of 8.9 mills/kwh.

‘The indicated power ‘cost must be considered-together with the state
of the technology of mblten'salts,‘of alloys for containingrthém, and of
engineering art for de31gn and construction of & reactor 1n order to
determine the emphasis that should be placed on gtudies of the system
in the future. Summaries of the current state of the technology of the
galts, metals, and components are'given in other parts of this report.
The fact of adequate solubility of uranium and thorium in the molten
salts and the strong position that is developing with respect to con-
tainment of the salts are characteristics that make the molfen_salt 8YyS=
tem unique among fluid-fuel systems. Although the materials studies are
not complete, the early results are so encoursging that plans should be

made now for the continued development of the molten-galt system.

The program visualized calls for carrying out the conceptual design
of an expefimental reactor during the fiscal year 1959 so that detailed
design could be started by July 1, 1959. The experimental reactor would
be designed to test typical comstructlon, operation, and maintenance
features of a large power reactor and could be completed by July 1, 1962.
After a'two-year operationai period, a very sound basis would exist for
decidlng whether or not to build large molten-salt power reactors. In
this proposed program, it should be noted that a substantlal part of the
materials compatibility program.would be complete before the major expendi=-

tures for an experimental reactor were made.

&

 
 

 

 

Fig. 1.1. Isometric View of Molten Salt Power Reactor Plant.

 

 
 

 

 

 

 

Table 1.1. REACTOR PLANT CHARACTERISTICS

Fuel enrichm:zny

" Fuel carrier

Neutron energy' |
Moderator

"?rimary'eoolant

Power
 Electric (met)
Regeneration ratio

Clean”(initial)
Average (20 years)

Blanket

Estimated costs

Total
Capitel
_Electric

Refueiing‘cyfile.at full power
.Shielding ,

Contrql_

Plant _effiéiency

Exit fuel temperature

> 90% U°3F, (initielly)

62 mole % LiF, 37 mole % BeFp,
1 mole % ThF),

Intermediate

Fuel solution circulating at
23,800 gpm

260 Mw
640 Mw

0.63

~4 0.53

T1 mole % LiF, 16 mole % BeF,,
13 mole % ThF),

§69, 800,000
§269/kw
8.88 mills/kwh

Semicontinuous
Concrete room walls, 9 £t thick
Temperature and fuel concentration

%0.6%

1210°F at approiimately 83 psia

 
 

 

 

 

 

 

T&ble 1.1,

Steam

Temperature
Pressure

Second loop fluid

Third loop fluid

Structural materials

Fuel eircuit

Secondsry loop

Tertiaxy loop

Steam boiler

Steam superhester and reheater

Active-core dimensions

Fuel eqnivalent diameter
Blenket thickness

Temperature coefficient , (Ak/k) /°F ‘

Specific power

Power density

Fuel inventozy

| Initial (clean)
Awerage (20 years)

uf,_fijritical mass clean
o up

( Continued)

1000 F with 1000 F reheat

1800 peia

 Sodium

Sodium

INOR-8

Type 316 steinless steel
5% Cr, 1% Si steel

2.5% Cr, 1% Mo steel

5% Cr, 1% Si steel

8 ft
2 ft

(3.8 £ 0,04) x 10~
~1000 kw/kg

80 kw/liter

600 kg of P37

~ 9% kg of o

) 26_7 kg of P27

Unlimited

 
 

 

 

 

2, GENERAL FEATURES OF THE REACTOR

The ultimate -power reactor of the molten-salt.type will probably
have & grephite moderetor, since & high breeding ratio is & mejor aim.
In order to obtain & breeding ratio as high as 1.0, it will be necessary
for the graphite to be in direct contect with the salt and with the nickel
- alloy «:cnrl:aa.:!.ner.3 Although it now seems probeble that graphite will be
| satisfactory for use in contact with the molten salt (see Part 3), the
‘technology is not considered to be far enough advanced to propose such a
-system for the initial reector. This considera.tion led to the specifica-
- tion of a homogeneous molten-salt reactor for this design sttjdy. |

A number of molten fluoride salts that are suiteble for & reactor
fuel are described in Part 2. The base salt chosen for the fuel solvent '
is 8 mixture of L:lTF and Be]‘2 in the mole ratio 62 to 37, respectively.
The Li7 and Be base salts iw.ve the moet desirable nuclesr properties of .
any of the possible salt cfitions. Beryllium, in addition to having &

low neutron s,bsorption cross section, sadds appreciably to the slowing- .
down power of the fluorime in the galt. Lithium-7 has the: lowest cross
section of the alkali fluorides. The exact percentages in the mixture
were determined &s & compromise of two physicel properties: the melting
point and the viscasity. The melting point increases &s the L1 content
increases, but the viscosity correspondingly decreases. The fuél mixture
is prepared by edding to the base salt small emounts of ThE), ahd‘"EUFu, the
ThFu being added to provide some regeneration of fissionable ‘materisl

in the core end the UF, being added to made the .reactor criticel. The
critical mixture cealculated for initia.l fueling of the proposed reector
hes the composition 61.8 mole %:1i Tp-36.9 mole % BeF,~1.0 mole % ThE), -
0. 3 mole % UF,+, with the u:ra.nium 93% enriched with lf?3 5

5 é

 

3. K. Ergen, A. D. Gallihen, C. B. Mills, eud D. Scott, "The -
Aircraft Reactor Experiment,” Fuc. Sci. and Eng. » 2 826-8110 (1957)

 
 

 

 

The selection of an 8-ft-dia core for this study was based primarily

~ on the criterion of critical inventory as indicated by nuclear calcula=-
itions covering core diameters of 5 to 10 ft. (Details of the nuclear
calculations are given in Part 4). The initial critical inventory for
S a U235 fueled reactor could be as low as 100 kg, which corresponds to
- & critical mass of about 50 kg. In actual practice, however, thorium
(that is, 1 mole % THF)) is added to the fuel to improve the Tegeneration
 ratio and thus reduce fuel costs, and the resulting initial critical in=-
~ ventory is about 600 kg. With thorium in the fuel, the 8-ft core is a
» feasonable choice that yields a good conversion ratio for a given invest-
_“menta Further, the 8-ft core provides sufficient volume for the average
.nower density in the core to be less than 100 w/cm3, which is well within

safe limits.' The gamms heaeting in the thicker parts of the core shell

~was also taken into consideration, and it was estimated that with the

8-ft core the heating in the core shell‘would amount to 12 w/cm.; which
is not expected to create significant thermal stresses.,

. It was decided that it would be worthwhile to include a blanket in
this reactor'system, despite the fairly high neutron absorption of the

core shell material, since the blanket would add between 0.2 and 0.3 to
the regeneration ratio and the increased saving in fuel cosis would amount

to about.$l,000,000 per year. Although the blanket adds some complications
to the reactor vessel, it offers compensations such as serving as a ther=
mal shield and as a convenient coolant for the fuel-expansion-tank dome,
which is subject to rather severe’ beta heating by the off-gas. The 2-ft-

, 'thick blanket Will allow less than 2% of the neutrons leaking from the
1»core to- pass through it without capture,_ The salt mixture Li7F-BeF —ThFu
| li"was chosen for the blanket and its composition vas selected as that which.
‘I_‘would give the highest ThFu content consistent with a melting point at
~ least 100°F below ‘the. lowest temperature expected in the blanket region.
fliThis specification led to the composition 71 mole % LiF, 16 mole % BeFp,
*13 mole % ThFh, which has a melting point of 980°F More recent chemical
data indicate that up to about 16 mole % THF), can be used'without in-

creasing this melting point°

 
 

 

 

 

 An alloy with the nominal composition 17 wt % Mo, 7 wt % Cr, 5wt % o
Fe, and TL wt % Ni, which is designated INOR-8, wes chosen as the struc-

‘tural material for all components of the reactor that will be in contact s
with molten salts. Details of the characteristics and fabricability of
.this alloy are presented in Part 3.

v}

- .The ehoice of the power level of this design study was arbitrary,
;since.the 8ft resctor core is capable of operation at power levels of
u§5£0_1900 Mw (thermel) without exceeding safe power densities. An
eleefirical generator of 275 Mw capacity was‘chosen, since this is in
| theieiie range that & number of power companies have used in recent years;
‘and & plent of this size could be justified in almost any section of the
United States. It is estimated that about 6% of the power would be used
in the etation, and thus the net power to the system would be about 260 Mw.

 

'Two sodium circuits in series were chosen as the heat transfer system
between the fuel salt and the steamn. Delayed neutrons from the circulating
fuel will activate the primary heat exchanger and the sodium passing
through it. A secondary heat exchanger system in which the heat will ‘
transfer from the radioactive sodium to nonradiocactive sodium will serve
- to prevenfi contamination of the steam generators, superheaters, and re-
heaters. A non-fuel-bearing molten fluoride salt is a possible alternate
choice for the radicactive intermediate coolant and has some advantage
~in that it is compatible with the fuel. In the design ‘adopted no dsnger
is expected to arise from mixing of the fuel salt and sodium, however,

and Therefore the cheaper sodium system is preferred.

| The fuel flow from the core is divided among four circults, so that
there are four primary heat exchengers to take care of the core heat
generation. This number of hest exchangers was based on maximum size Vs
thermal-strain considerations. Each of the four parallel heat tranefer
circuits originating in the fuel system trensfers the heat through two
sodium circuits to the steam generators. A similar single circuit is o 2
| providedete'rembve the heet generated in the blanket. |
 

 

 

 

Other 1inkages between the fuel and steam thst have been proposed
are a saltato mercury-vapor system and a salt-to-helium ges system.
The 1atter system is currently being studied. '

‘A plan view of the reactor plant 1ayout is presented in Flgure 1.2,
and an elevation viev is shown in Figure 1.3. The reactor and the pri-
mary heat exchangers are contained in a large rectangular reactor cell,
which is sealed to provide double contéinment for any leakage of fission
gases'and in which all operations must be cerried'out remotely after the
reactor has operated st poser. The primary heat exchangers are laid out

-0 provide an ineline heat exchange system. The rectangular configura-

tion of the plant permits the grouping of similar equipment with a mini-
mm of floor space and piping. The primary sodium circuits are thus
located in ome bay under a crane, and in the next bay'are the secondary
sodium circuits, the steam generators, superheaters, and reheaters under
'another crane. The plant includes, in addition to the reactor and heat
excnanger systems and the electrical generation equipment, the control
room,‘chemicsi processing equipment, end the fill-and-drain tanks for

" the liquid systems.

3.  MOLTEN.SALT SYSTEMS

3.1. Fuel and Blenket Circuits

The primary reactor cell, which encloses the fuel and blanket cir-

fl.ycuits, is a rectangular concrete structure 2k £t wide, 68 £t long, and
-:,}_70 ft high._ The walls are msde of - 9-ft-thick barytes concrete to pro-
p;.lvide the biological shield. Steel 1iners on both sides of the concrete
__J_fg:_wall form a buffer zone to ensure that no fission gos that may leak into
p:iithe cell can escape to the atmosphere and thet no air can enter the cell.
 An inert ‘atmosphere is. maintained in the cell at all times so that &
__5?7sma11 fuel leak will not lead to accelerated corrosion. Penetrations
'*a;;for pipes and‘for electrical and instrument lines are sealed on each side

‘of the enclosure, .

 

hB. W. Kinyon and F. E. Romie, Two Power Generation Systems for a
Molten Fluoride Reactor, Presented at the Nuclear Engineering And Science

Conference of the 1958 Nuclear Gongress (March), Chicago, Illinois.

~9-

 
 

oL

 

 

 

 

 

 

 

 

 

Fig. 1.2, Plan View of Molten Salt Power Reactor Plant.

 

 
 

L[l

 

 

 

 

 

 

 

 

 

 

 

 

Fig. 1.3. Elevation View of Molten Salt Power Reactor Plant.

 

 
 

 

 

Once the reactor has been at power, the radiation level of the
réactor and the fuel and blanket circuits will be so high that it will
not be possible to perform dirgct maintenance operations onlequipment
in these circuits. All equipment that might require replacement by re-
mote means is instelled in the shielded reactor cell. The alternative
of segregating each piece of equipment in a separate enclosure is more
costly in terms of space, shielding, piping, and fuel inventory. An
air lock is provided through which the crane and maintenance equipment
can be brought into the.ceil.

The principal items of equipmént in the reactor cell are the reac-
tor véssel, the fuel and blanket pufips, the fuel and blanket heat ex-
changers, heating and insulation equipment, and the reactor cell cooling
system, and, of course, there are many electrical and miscellaneous plumbing
lines. The reactor vessel and the fuel and blanket pumps ére a8 closely
coupled, integrated unit (Figure 1.4) which is suspended from a flange
on the fuel pump barrel. The vessel itself has two regions -~ one for
the fuel and one for the blanket salt. The fuel region consists of the
reactor core surmounted by an expansion chamber, which contains the
single fuel pump. The blanket region surrounds the fuel region and ex-
tends above the expansion chamber, and the blfinket salt cools the walls
of the expansion chamber gas space and shields the pump motor.

Four tangential pipes serve as ducts to return fuel into the lower
conical section of the core. The core shell, which should be as thin
as possible in order to reducé neutron absorption end to keep thermal
strains from gamma heating as low as possible, was chosen to be 5/16 in.
thick to provide adequate strength against buckling under conditions of
maximum pressure differential between the blanket and the core regions.

The floor of the expansion chamber is a flat disk, 3/8 in. thick,
which serves as & diaphragm to absorb differential thermal expansion
between the core and the outer shells. During reactor operation, the

 
 

 

 

 

bt A YA 1 1 e i s

 

 

" FUEL-RETURN

UNCLASSIFIED
ORNL-LR~DWG 28636A

 

 

 

 

 

 

FUEL PUMP MOTOR
MOTOR

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FUEL EXPANSION
TANK

 

 

 

— BREEDING BLANKET

 

     
 

 

 

 

Fig. |.4. Reactor Vessel and Pump Assembly.

-13 -

 
 

 

 

temperature difference between the corc and the outer shells will be
small, but during thé preheating and during power transients the dif-
ferehcé may be higher. The diaphragm will safely allow for differ-
entlal expansion corresponding to a temperature difference of 200°F
without undergoing appreclable plastlc strain.

'-The pipe ducts that enter the reactor_tangentially'will impart a
swirling action to the fuel and keep it turbulent near the wall. No
fluid flow patterns have(&et been obtained for a core of the shape il-
1ustrated; and flow tests; when made, may dictate changes in the shape}
The pump‘for circulating the blanket salt is not supported directly on
the reactor vessel, but iS'located.to the side and at a slightly higher
elevaticn,tc give full blanket ccverage of thc”reactor at all times.

The general requirements of pumps for the;fuel and blanket circuits
are-discusscd*in Part 5. This design study is¥based on use of a pump
of the type shown in Figure 5.2 of that section, which has a capacity
of 24,000 gpm. Based on a fuel volume of 530 £43 in the entire circu-

lating'system, the fuel will meske a complete circuit through the reactor, -

piping, and heat exchangers in 10 sec. A 1000-hp motor will be needed
for the fuel pump, and a shaft speed of 700 rpm will be required. As

indicated in Part 5, this pump incorporates advanced features not present

in any'mclten-salt pumps operated to date. Some consideration was given
to the use of multiple pumps, but since a single fuel pump simplifies
the top portion:of the reactor assembly, it was adopted for this study.

The blanket pump will be similar to the fuel pump, but of smaller
capacity. Sincé the heat generation in the blanket will be no more than
10% of the total heat generated, a pump of about 3200-gpm capacity is

requlred.

| Four primary heat exchangers are provided for the fuel circuit so
that each'heat exchanger will be of reasonable sizepl These heat ex=-
changers are designed to have the fuel on the shell side and sodium '__
‘inside the tubes. This arrangement is contrary to that which might

- 1l -

fl.

 
 

 

 

o e ol e e 0 1AL ST L W

. of power level is improved by keeping the high cross section Xe

intuitively be proposed because it might be expected that the fuel
volume would be lower if the fuel were inside the tubes. However,

the superior properties of sodium es & heat transfer fluid are not
realized with the sodium on the shell side, and therefore the over-all
system is most compact with the sodium.inside the tubes,

The heat exchangers (Figure 1.l) are of semicircular construction,
which provides for convenient piping to the top and bottom of the re-
actor. The blanket heat exchanger is similar in construction, but it
is scaled-down to be consistent with the smeller heat load. A more
detailed description of the hest exchanger is given in Teble 1.2 of
Section h
3.2, Off-Gas System

An efficient process for the continuous removal of fission~product.
gases ig provided that serves several purposes. The safety in the
event of & fuel spill is considerebly enhanced if the radiocactive gas
concentration in the fuel is reduced by stripping the gas as it is
formed. Further, the nuclear stability of the resctor under chigges
continuously at & low level. Finzlly, many of the fission-product
poisons aié; in their decay chains, either noble gases for a period of

time or end-their decay chains as steble noble gases, and therefore
the buildup of poisons is considerably reduced by ers removalo_

The solubilities of noble gases in some molten salts are given

 inFart 2, end it is deduced that solubilities of similar orders of
*V”’magnituae are likely to be found in the LiFwBeF, salt of this ‘study.
- :,i[It vas found that the solubility obeys Henry‘s law, so that the equi-
: 7g:librium solubility is proportional to the partial pressure of the gas
rd in. contact with thc salt. In principle, the ‘method of fission-gas Tes
,Vffimoval consists of providingAan»efficientumechsnism for contacting the
~ fuel salt vith an inert ambient ‘gas in which the concentra.tion of xenon
- snd krypton is kept very 1ow. f{,v'.:-f | | |

’

«l5

 
 

 

 

In the system .chosen; approximately 3.5% of the fuel flow is mixed
with helium purge gas and spreyed into the reactor expansion tank. The
mixing and spraying provide a 1arge fuel-to-purge-gas interface, which
promotes the estsblishment of equilibrium.fission gas concentrations in'
the fuel. The expansion tenk provides a 1iqu1d gurface area of epproxi-
mately 26 £4° for removal of the entrained purge and fission gas mixture.
The gas removal is effected by the balance between the difference in the
density of the fuel and the gases and the drag of the opposing fuel ve=-
looity.‘ The surface velocity downward in the expansion tank is approxi-
mately 0.07 ft/sec, vhich should screen out all bubbles larger than
0.008 in. in radius. The probability that bubbles of this size will
enter the reactor is reduced by the depth of the expansion tank being
sufficient to'ellow time for small bubbles to coalesce and be removed,

The liquid volume of the fuel expansion tank is approximately Lo ft3
and the gas volume is approximately 35 ft3 With a fuel purge gas,rate
of 5 cfm, approximately 350 kw of beta heating from the decay of the
fission gases and their daughters is deposited in the fuel and on metal
surfaces of the-fuel expansion tank. This 350 kw of heat is partly re-
,moved by the bypass fuel circuit and the balance is transferred through
?the expansion tank walls to the reactor blanket. o

| iThe'mixture of fission gases, decay products, and purge helium
1eaies the expansion:tank through the off-gas line, located in the top
ef the tank, and Joins with a similar stream from the blanket expansion
tank (see Fzgure 1.5). The combined flow is delayed approximately 50 min
in a cooled volume to allow & large fraction of the shorter lived fission
pflpducts to decay before entering the cooled carbon beds. The carbon
beds provide a holdup time of approximately 6 days for krypton and much

1oqger for Xenon,

The purge gases, essehtieily free from activity, leave'the carbon
'bed& to'join=the gases from the gas-lubricated bearings of the pumps.
The gases are then compressed and returned to the reactor to repeat the
cycle. Approximetely every four days one of the carbon beds that hes _
been’ 0perating at minus ho F is warmed to expel the Kr85 end other longu_
'11ved fission: products  _ -

",; | - . t - 16 -

-

'(’

 
 

-1 -

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

c - ~ C.C
UNCLASSIFIED
ORNL-LR-DWG, 28439
R _|,, =
wmw Hs R i
N K 13SCFM
b MOTOR
FUEL [~ j Qfimw& /95 scem_ (T \BLANKET
pump [ — Q_Jeume
| | /955ch HE —] ¥ 0.65 SCFM
] 1 | | [Beamwer exeansion
— ; - TANK D.ICFJf-/Z/O"F 5.3 SCEM
FUE gxpflfly/0” e 12I0°F —Lflzfl/w([?' BY-A4SS
;gfs'vg"/-" L ’MFJ—E’%-L 1 ez ‘YL— BLAVKET
»°F 2109
r el bwms . EeT
L8CES
L AES0F Le COA’E V
saszcem = S
/5940%1 L
6776 CFIM o s
P50°F |
| A cemsem 3 203CEM 1.63CEM
| 2I2F. | 62F —4OF
55 FT° 785 FT° '_,_—'|__5/_F'_d /6 FT8
x| % I\ x ¥ | X
| 4
=11 ot 5
' 7
25 0 755 FI° I6FT.
BEF 22F 4 62%, 40F,,
| - 4~ CLOSED BLOCK VALVE
S OPEN BLOCK VALVE
ouT T//v l/fl Tozrr loz/r Tw
800°F 105F 105F 00°F OF OF ~40°F ~dO'F
COOLANT COOLANT COOLANT COOLANT

Fig. 1.5.

Schematic Flow Diagram for Continuous Removal of Fission Product Gases.

 

 
 

 

 

 

" .3. Molten Salt Transfer Eguipmsnt

Ths fuel transfer systems are shovn echematically in Figure 1. 6o
Salt freeze valwes, described 1n Part Dy are used to isolate the indi- -

t-vidusl components in the fuel transfer linee and to isolate the chemical

'plant from the componente. in-the reactor cell. With the exception of
the reaetor dreining operation, which is described below, the 1iqu1d is
transferred from one veessel to enother by differential, gss pressure.

' 'By this means, fuel may be added to or withdrswn from ths reactor during
tpower operation. " '

- The fuel added to the reactor will have 8, high eoncentrstion of
U235Fh, with respect to the process fuel, so that additions to overcome
burnup vill rpquire transfer of only a small volume; sindlsrly, thorium-
 bearing molten salt may be added at any time to the fuel system. The
thorium, in addition to being 8 design constituent of the fuel salt,
~ may be added in amounts requirea to serve es & nuclear poison for ad-
| Justing the mean core temperature.

| When fuel is removed from the reector, it first goes to onhe of the'

withdrawal tanks. These tanks will be sized to serve ‘&8 holdup vessels
from which material mey be later transferred to the chemical blsnt.

The chemical processing plant is considered to be.an integrsl part'offg,'
the reactor complex; however, the chemical processing plsnt’is set apart

from the reesctor, is conteined in separate cells and has & sepsrste con=
trol center. As indicated elsevhere, the fuel-reprocessing cycle assumed

for this report requires an average daily withdrewal end addition of
gbout 2 ft3 of fuel. If a 30-dsy'holdup of fuel is required for cooling
"~ before chemicsl processing, the withdrswal vessels must provide & volume
of 60 ft3 . They will require both & heating and a cooling system simi-
ummmwmmuwmmmm1mmmmmmmmMu'
to. msintsin the temperature within reasonable . 1imits.

-"'18-

e
61

 

UNCLASS IFIED
ORNL-LE-DWG, 20630

S
LEGEND:
' — FUEL LIE
——— —-— GAS LINE

e ‘ _______.__..-—--—r«E-.r —g—.rozmow VALVE
1
L REMOTE OPERITED
—r'E"" |—£—-y ~42r—m/.x/£

{EVY-FREEZE WaivE
S —sueoly
V — VENT

|
i
| —— N FLOW
t DIRECTION

 

 

 

 

    
 

4

BY-A45S

      
 

   

  

REACTOR HERT

EXCHANGER

 

 

WITHDRAWAL

Lzas
ENRICHER TANK

 

 

 

 

 

e ——_—_—— e e

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4 ~ > m—gq

P T T T T T m et e e o - - A | [l
' I
I i Ha * Y : Y : I I ' | _
t .
: ! - [ |
} — : ]
|
: MF : [ l
. ADSORBER ! MAN FILL € DRAN SYSTEM
| I
I — *— i ~ALUORNATOR %

- - |
\ wrkaenm, | TERYR cuemmca processve ALANT . ]

Fig. 1.6 Schematic Diagram of Fuel Salt Transfer System.

 

 
 

 

 

For the main fuel drein circuii, mechanical valves will be placed

~ in series with the freeze valves to establish a stagnant liguid suitable

for freezing. Normally th:ese‘mecha,niéal valves will be left open.

. Draining of the fuel will be accomplished by melting the plug in the freeze
~ line a,nd,_aliowing‘ the fuel to drain by gra.vi'by._ By opening gas _pressgre |
equallization velves, the “liguid in the reactor will flow to. the dré.in tank,
_and the gas in the drain tank will be transferred to the reactor system.

Thus gas will not have to 'be added to, or vented from, the primary system

Two valves are 1ocated. in parallel in the fuel drain line g0 that a spa.re

path will be available in the event of failure or need for repa.irs.

All the liguid transfer lines will be equipped with heaters and
covered with insulatibn so that the system may be held at temperatures

a.bove the fuel melting point. Since the main drain line is at the 'bottom
of the rea.ctor, there will always be fuel in it. This line is kept from

freezing or overheating by use of a circulation bypass, as shown in Fi‘gure
.6; to keep the stagnant portion confined to the freeze valve area. This

'bypass prondes a certainty that the drain line to the freeze valve is

elways at temperature snd open for draining.

The blanket fluid transfer system is essentially the same as the
fuel system. A chemical processing plant will be provided for the blanket
salt, which may serve as & backup capacity for the fuel reprocessing

~ plant..

3.4. EHeating Equipment

The melting points of the process fluids used are all well above
room témperature. It 1s thus necessary to provide a means of heatihg ’
all pipes and equipment- conta'.ining' thege fluids. This will, in general,
be accomplishedj- by providing electric heaters to all pipes and equipflent.i
Inside the reactor cell, the heaters are 1ncorp<_>rated in removable 'a_s.- |
semblies that consist of the heaters and the insulation, as shown in
Figure 1.7. Outside the cell, conventional methods of insta.lling heaters

.. and insulation ere used,

i

)

 
 

 

 

L LIETING £ SPREADING

- MECHANISM ————

 

Figo 1 o7

  

 
 

e b e

 

 

345 Auxillary Coollng

Coollng is provided in the reactor cell to remove the. heat lost
through the pipe insulation and the heat generated in the structural
steel pipe and equipment supports by gamms -ray absorption. The heat |
is removed by means of forced gas circulation: through radiator-type |

'-space coolers. A cooling medium, such as Dowtherm, in a closed loop

rémoves heat from the space coolers and dumps it to a water heat ex-

- changer.

3.6.  Remote Maintenance

Provisions are made to carry out sll maintenance operations in

‘the reactor cell by remote means. While some small repairs may pos-
 sibly be made in the reactor cell, the principal requirement is to be

able to remove and replace by remote means all the necessary components
in the reactor cell. This will include pumps, heat exchangers, pipe, -
heaters for pipe and equipment, instruments, and even the reactor ves-

sel,

A prime requisite for remote maintenance is a reliable method of
making and breaking joints in pipe. This can be done, either by de-
veloping & remote cutting and welding process or by developing a satis=~
factory flanged pipe joint (see Part 5). (The development of a remote
welding process is underway at Westinghouse on the PWR project,)

All equipment and pipe Joints in the reactor cell are laid out so
that thsy;are accessible from above, Directly above the equipment is
a traveling bridge on which can be mounted one or more remotély.opera-
ted manipulators. At the top of the cell is another traveling bridge

for a remotely operated crane. At one end of the cell is an air lock

that connects with the maintenance area., The crane can move from the
bridge in the cell to a monorail in the air lock.

- 22 -

@

)
 

 

 When 1t hes been ascertained that a piece of equipment ghould be .
: replace&, the reactor will be shut dovm and drained, and the plece of
faulty equipment will be removed according to the following procedure.
The manipulator will be trangported by the crane from the main’bena.nce
erea through the air lock, to the cell, and pleced on the manipuletor
‘bridge. The manipulator will then be used to disconnect all instrue
| men'b, electrical, and service connections from the equipment and to
| unfa.sten the fle.ngee tying the ‘equipment to the system. The crane
will then remove the feulty equipment and tmnsport it to 'bhe meinte-n
nagnce area, The crane will then be used to move & epare 9ieoe of ejquip~
ment 1nto the cell for installation with the use of the ma.nipu]a.toro
After eompletion of the replacement, the manipulator will be removed
from the reactor cell by the crane, the air look will be closed, and
the rea.ctor will be ready for startupo Prelimimry t-ea'bs with & Genersl
Mills manipulstor have demonotruted ‘the fea.sibility of renotely removing
and repla.cing the rotn.tins assembly of e liguid metal pump, It appears
therefore that eat:lsfo,ctory t-chniques can be developed for remote '
‘maintenance.

Closed-circuit television equipnent 1s provided, for V:Lewing the
E maintemnoe operetion in the eello A A number of camers.s a.re mounted to
- ghow the opera.tion from different angles y and periscopes give a direet
view of the entire cell,

The mintenance area :Lo divided into hot and cold shop areas. The

B . e cold shop will be used. for generel repe.ir work on equiment that can be
‘{_‘_'jhandled d.ireotly. : 'I’he hot ehop area will be used. to (1) repa.ir minte-
. nsnce equipment that can not be handled directly; (2) to disassemble
o Zifailed equipment 10 de‘bemine 'l:he eause of fa.ilurej; (3) to prepare hot

' '-‘;i'equipment for disposa.l, that is )’ out or disassemble la.rge equipment to.

   

 

 

TR f-"“:f*/managea'ble size, plaoe 1n coffina, etc., end (h-) to repeir failed equip-n-
; ; B 'ment within the 1imits of 'bhat whieh ean be d.one with the equipment ree
s quired for the other hot-shop opera.tionse A completely equipped. hot '

 

- 23 «

 
 

 

 

 

-;shop capeble of making any and ell Trepairs to all eQfidpment does not
- eeppear to be economically adriseble for the anticipated maintenance
d"work for & single reactor plant. Although it is possible to remove |

and replace the reactor, it is & comparatively simple and rugged piece

of equipment with & low probebility of railure, and therefore & spare
reactor will not be provided. '

~ Maintenance of the intermediate sodium circuits will be done dia_

frectly. In case of an equipment failnre in one of these circuite, the
= loop will be drained. At the ‘end of & feirly short period, for rosidual
, fN Eh to decay, it will bve possible to remove the top slab from the second-
- ary cell end remove and repdace the faulty equipment by usins the building
| l_crane end direct maintenance procedures. Each secondery cell is ‘shielded,
d so that the addacent cells need not be drained to make a repair.

- ; 7. Fuel Fill-and-Drain Tenk

The main fuel flll-end-drein syetem.mnst neet the following wa Jor

design criterisa: . :
(l) A preheating system must be provided that is capable of main-

taining the drain vessel and its connecting plumbing at 1200°F .

(2) A relisble heat-removal system must be provided that hes euf-‘
ficient capacity to handle the Puel afterheat.

(3) The drain vegsel must be "ever=- safe” so that & criticel cone

jdition cannot occur when the fuel is drained.

The fuel dreining operation has not been considered as an emergency
procedure, that is, one which must be accomplished in a relatively short

~ period of time in order to prevent & catastrophe, There are,. however,_

other 1ncentives for rapid removal of the fluid from the fuel circuit.

- If, for example, there were a leak in the fuel system-it would‘be ine
- portant to'drain the fuel in order to minimize the cell contamination

and cross contaminafion of the systems. Further, rapid removal of the
fuel at the time of & shutdown for me.intenance would have an economic
edwantage in reducing “the power outage time.

e

v

¥

 
 

 

 

-A-cohsideration of these factors indicated that the maximum efter-
heat design load should be 10 Mw for & 600 Mw reactor that had been
operating'for one year and had been shut down for 10 min before the fuel
drain was started, No credit was taken for fission-gas removal during

'.operation, It was further estimated that 15 min would be required to

lremo?e the fuel from the reactor.

For the drain vessel design calculations, it was assumed that at
1200°F the fuel system volume would be 600 ££3. The design capacity of
the drain vessel was therefore set at 750 ft3 in order to allow for
temperature excursions and a residual inventory., An array of 12-in. -
dia pipes was selected as the primary contsinment vessel of the drain
system in erder,to obtain & large surface area-to-volume ratio for heat
transfer efficiency and to provide a large amount of nuclear poison ma-
terial (see Figure 1.8). Forty-eight 20ft lengths of pipe are arranged
in six vertical banks comnected on alternate ends with mitered joints.
The six banks ef pipe are comnected at the bottom with a common drailn
line that comnects with the fuel system. The drain system is preheated
end maintained at the desired temperature with electric heaters instal-
led in smallédiameter pipes;located.exially inside the 12-in.=-dis pipes.
These bayonet-type heaters can be removed or installed from one face of
the pipe array to facilitate maintenance. The entire systen is instal-
led in an ifisulated room or furnace to minimize heat losses.

The removel of the fuel afterheat is accomplished by'filling‘boiler
tubes installed between the 12-in.-dia fuel-containing pipes with water

- from headers that are normally filled. The boiler tubes will normally

"be dry and at the ambient temperature of about 950°F Cooling'W111 be
__'°accomplished by slcwly flooding or “quenchlng" the “tubes which furnish
. a heat sink for radiant heat transfer from the fuel-containing pipes

to the boiler tubes.. For the peak afterheat 1oad, about 150 gpm of

 water is required o 'supply the boiler tubes.

L

 
 

 

 

UNCLASSIFIED
ORNL-LR-OWG. 28535

 

Fig. 1.8, Drain and Storage Tank for Fuel Salt of Molten Salt Power Reactor.

26

vy
 

 

Thie fill-end-drein system satisfies the design criteria in that
it 1s always in & etendby condition, in which it 1s immedizstely evailable
for drainage of the fuel, it:can sdequstely handle the fuel afterhest,
end it provides double contaimment of the fuel. Heat removel is essen-
tially self-regulating in that the emount of hest removed 1s determined
by the rediant exchange between the vessels and water wall, Both the
water and the fuel systems eaxe et low pressure , and & double fallure
would be required for the two fluids to be mixed, The drein system temk
may be ee.sfly enclosed and sealed from the astmosphere because there are
no large ga.s-cooling ducts or cother major external systems connected to
it. A steinless steel trey will be placed below each benk of pi;pes to
catch the fuel 1if a leek develops. These treys will be cooled by water .
walls to prevent any possibility of meltdown and destruction of the cell.

A prelimina.ry eriticality caleulation wes mede in a drain tank as-
sembly without cooling walls, A multiplication constant of 0.2 was esti-
mated for a fuel containing 0.5 mole % ThFh_ and 0,125 mole % UF), at a
temperature of 125005‘.

4. EHEAT TRANSFER SYSTEMS

The intermediate heat transfer systems use sodium as the working
fluld to transfer heat from the fuel ’co the steam system, The latter
accepts the hea.t in the steam generators ’ the superhea’cers ’ a.nd the re-
heaters, A diagrem of -bhe heat removal 15 shmm :Ln Fig. 1.9. A speci-
fication thet the steam system components should be cempletely radiatidn-
o free dictated two sodium circui‘ts for each heat tra.nsfer pa.th.

Fbur systems :!.n pa.ra.'llel remove hea.t from the ree.ctor fuel;

o _f_-_a,dditiona.l system ha.ndles the power generated :I.n ‘bhe 'blanket sa.lt. Each

of the five systems :I.e separa.te and :l.né.ependent u;p to 'bhe point where
the superheated steam f.'l.ow paths ,join a.hea.d of the turbine. o

e

 
 

 

8¢

  
  
 
 
  

r"’}

o (BLANKET

il

 

UNCLASSIFIED
ORNL-LR-DWG, 28640

 

REHEAT
O/LER

A 220 PSS/ 690 F

FOORS/
/000 < ;;Q

. g ~—LPE/ P TURBINES

Y
LS HG

 

~—H PTURBINE

 

 

 

 

 

  

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

    
    
   
   

 

 

 

 

Y | J_’JEL "‘%’l"/’ 7 x\_"
A Y o - —__I~—TURBINE STOP
RReRL| | M BAMET i o ~—ATTEMPERATOR
A P M—EMERGENCY RELIEF
° ' ' COOLING WATER
(0807 - 2% 'CONDENSER
755 By X |8rAass ] SUPERHEATER
O | /%9055,’{_ 75t oo : ~—REDUCING STATION
9%
| srconpary X SUPER 1
Hé‘AT[A’C‘// 13CFS 1 HEATER ~CONDENSA 7'1." PUMP
28Ps/ |
isscrs]  (2EaM
NDER|——— I30F | {1800 Pst
O B s [P
' FEEDWATER
i HEATERS NV SE/?/[.S
A '
STEAM CHEST: o
\[j_ ~-DEAERATOR
1 X :
| - FEEDWATER PUMP
265/ 22pPs/ |- ' '
241 CFS [er/cFs 577CFs 70 BLIANKET
979 LY |
A 2T PR3 }7‘0 FUEL
3B8.E6CFS FEEDWATER
é % 4T HEATERS /A/&[/?/[.S‘
L apy 1 T~—8onR
SODIM _, #oF 1 | /,seqooc_; éff— —

 

 

Fig. 1.9. Schematic Diagram of Heat Transfer System.

 
 

Each primary sodiuwm circult includes & primery heet exchanger in
the reactor cell and a pump and & secondery heat ext:hanger located in a
cell adjacent to the reactor cell. No control of flow ratee is required,
so there are no valves, and constent=-speed centrifugal pumps ere used.
With a \pmnp‘ stopped, thermal-convection flow would be avaeilable to remove
afterheat from the reactor. The secondary heat exchangers are of the
U=tube in U-shell, counterflow design, with the sodium of the primary
eircuit in the tubes and the sodium of the secondary circuit surrounding
the tubes. In order for the sodium to be at a lower pressfire‘ than that
of the fuel in the primary heat exchengers, the pumps for the primarj |
sodium circuits are on the higher temperature legs of the circuits, The
essentlal characteristics of the various heat exchangers are described
in Table 1.2.

The secondary sodium circuits, except for the secondary heat ex=
changers, are outside the shielded area and thus are avallable for ad-
Justment and maintenance &t all times., Three paths are provided for the
sodium flow from a secondary heat exchanger: a steam superheater, a
steam reheater s 8énd & bypass line for control. Regulating va.lves autos
matically adjust the flow to sult the load conditions, so that at very
low loads most of the flow is through the bypass line, |

The three sodium streams are recombined in & mixer or blender, which
leads to & three-way valve. At design point, sbout one-third of the flow
returns directly to the pump suction a.nd two-thirds enters the steam gen-
era.tor a8 the driving streem of a Jet pump The Jet pump, located verti-
ca.lly along side the steam generator 5 is assisted by thezmal-convection

o i‘low ,mra in the Jet pmnp and downward in the bc::i.le::-° At 1ow power

r-.1.e1@r«eil.s ’ 'bh:l.s serves to ma.inte.in 8- good recirculation rate a.nd ensure good

: .,stability of control.‘ 'J.‘he three»way va.lve permits 'l',he steam generator

 

ci:rcuit to be isolated from the rems.:l.nder of the sodium 80 tha’c at zero
'_rpower the entire ’ooiler becames isothermal at the saturation temperature,
amd the pressure :l.s maintained a'b the desired level. '

- 29 -

 
 

| fuel ‘and Sodium to Sodium Exchangers

Number required
Fluid

Fluid location -
Type of exchanger

Temperatures

Hot end, °F

@®l1d end, °F
Change

AT; hot end, °F
AT, eold end, °F
AT, log meen, °F

-Og-

Tube Data

Material

Outside dlameter, in,
wall thickness, in.
length, £t

Number

Pitch (4A), in,
Bundle dlameter, in.

Heat transfer capacityé Mw

Heat transfer sxrea, ft

Average heat flux, 1000 Btu/hreft2
Thermal stress,* psi

Flow rate, £t3/sec

Fluid velocity, ft/sec

Maximm Reynolds modulus/1000
Pressure drop, psi

 

* [ch/(l-y)] [(max AT-of wall)/2)

‘ .
. , e

Table

 

 

 

l.2. Data for Heat Exchangers

Primary System

4
Fvel salt Primary sodium
Shell Tubas
U-tube in U=-shell,
countexrfliow
1210 1120
1075 925
135 195
o0
150
11T7.5
INOR=8
1.000
0.058
25.7
515
1.1
28
144
2800
175
9200
130 l‘l’ h‘6 ¢ l
10.8 19.7
%.5 523
40 15.5

Secondary System

4
Primary sodium Secondary sodium
Tubes | Shell -
U-~tube in U=-shell, -
counterflow
1120 1080
925 825
195 255
100
65.6

Type 316 stainless steel

0.750

0.049

21.5

1440

- 0.898

36

14l
5200
95

8200
46,1 - 33,6
13.9 13.2
270 166.0

10  14.8

 

 
 

 

 

 

 

(: ¢ 3

Sodium to_Steam Exehangers -
Number required )
Fluid -

Fluid loca.tion .
_ Type of exchanger

_ Te'mpera.tures

- Tg-

Hot end, °p .
Cold end, °F
Change , °F
AT, hot end, o
AT, cold end oF
AT, log mean, °F

Tube Datea

Materisl o
Outside dlameter, in.
Wall thickness, in.
 Length, ft ‘
-Number

Pitch (4), in.
Bundle diameter, in,

Heat transfer capacit
Heat Transfer aresa, f é
Average heat flux,

1000 Btu/hre ft2
Thermal stress, psi
Flov rate, ft3/sec

1000 1b/hr

Fluid velocity, ft/sec
Maximum Reynolds modulus/1000
Pressure drop, psi

 

Table 1l.2.

 

Steam Generator
.;_ o l
Secondary Water
sodium
Shell Tubes
- Bayonet,
countertlow
8as 621
O - 62
85 . ' 0O
- 119
20k
158

2.5% cr, 1% Mo Alloy -
S 2

0.180
18
362
2.75
55

82.2
2800

100
18.600
5T+5

5.6
200
5.7 (Jet pump)

410

* [aE/(J.-})] [(max AT of wall)/2]

¥

(Continued)

 

Superhea'ber :
4
Secondaxy Steam
sodium
Shell Tubes
U=tube in U=-shell,
counterflow
1080 1000
930 621
150 : 379
| 80
309
169

5% Cr, 1% Si Alloy

0.750
0.095
25
480
1.00
23
39.2
1760
76
9000
15.5
406
9.3 = 61
164 306

6.9 10.3

 

Reheater
, L
Secondary Steam
sodium
Shell Tubes
Straight,
counterflow
1080 | 1000
1000 - 6ho
80 360
- 80
360
- 186

5% cr, 1% 51 Alloy

0.750
0,065
16.5"
800
1.00
29.7
22.6
' 2200
35
' 5000
16t8
399
T.9 137
163 167
3.2 10. 4

 

 
 

 

 

A portion of the cooled sodium leaving the steam generator circuit
returns to the pump suction, which is constructed as a blender, end mixes
with the stream bypassed through the three-way velve, The centrifugal
pump is specified as two-speed, with the second speed being one-fourth
of full speed to give essentie.'l.ly one-fourth of the full flow so tha.t
the power output may be more easily regulated from 25% dovn to very low
levels.

The steam genere.tor , Fig. 1.10, consists of tubes suspended in the
flowing sodium. In this Lewis~type boller, the water flows through &
centrel tube to the bottom of each beyonet and boiling occurs during
upwerd flow in the outer emnulus. Baffles in the steam dome separate
the water and steam, and the water returns to & tray which collects it
for recirculation,

Just below the tube sheet and above the godium, & thermal barrier
and & gas space are provided to permit the tube sheet to he at the satu-
retion temperature and thus avoid thermal stresses. In addition, the
gas space will serve &s & cushion for the initial shock in the event &
tube ruptures and water leaks into the sodium.

The superheater 1s a U-tube in & U-shell, counterflow exchenger,
with the steam inside the tubes. As in the steam generator, there 1s a
thermal barrier between the sodium and the tube sheet &t the cold end
in order to minimize thermal stresses. The reheater consists of straight
tubes in a straight shell. With the sodium on the shell side, the tem-
perature difference between the shell and the tubes is sufficiently small
that this more economicel construction can be used. The reheaters are
" located mear the turbine’ to give a small pressure drop in the reheated
steam.

%R, H. Shannon and J. B. Shelley, "Double Reheat Cycle - Next Step?”
. _Power February 1955 p 98~99.

 

-32 -

 

 
 

e e G e o St . o e St s Ao A

UNCLASSIFIED
- ORNL-LR-DWG 27480

f——MMIER

|— WATER AND
= STEAM

TUBE
SHEET

 

 

 

 

 

 

 

 

 

 

 

e
N
i il ]
|
\!-ll

 

 

 

 

 

 

 

THERMAL fh
BARRIER l L

 

 

 

 

 

 

 

 

 

 

 

 

 

7
-
i

 

 

 

 

 

 

 

 

 

 

 

 

 

.10 Steam Generator, Lewis Type.

 
N —

 

 

 

 

5. TURBINE AND ELECTRIC SYSTEM

Steam is supplied to the 275-Mw-rated turbine at 1800 psie end 1000°F,

The single shaft of the turbine operates at 3600 ¥pm; there are three
exhaust ends, The turbine heat rate is estimated to be 7700 Btu/kwh

| for & cycle efficlency of 44,3%. The electrical genera.tor end station

heet retes eve, respectively, 7860 and 8360 Btu/kwh. With 6% of the
electrical generator output used for station power, the supply to the
bus ber is 260 Mv. These estimates are baged on Tennegsee Vs.lléy
Authority heat belances for & turbine of this type ,6 with athstnients

- mede for the modified steam conditions' end the different plent require-

ments of the molten-salt resctor system.

The condj.tions given above were selected to give the minimnn cost.
Increased cycle efficiency could be obta.ined with higher tanpera.tures

- end, higher Pressures, but the ilncrease in efficiency would be offset by

the :anrea.ses in equipnen‘b costs associa.ted with the higher temperatures
and pressures.

6. NUCLEAR PERFORMANCE

The nuélea.r behavior of the particular molten—sait reactor and fuel
processing cycle selected for this design study is presented in this
gection. (A more detalled parametric study of the nuclear performence of

. molten~salt reactors is givén in Part 4.) The reactor could utilize either

0233 or 1123 2 but, since U2 55 'is the only isotope presently available in

A qua.ntity, it was selected as the primary fuel, For comparison with other

reactors designed to use 023 5 8s a fuel, the performsnce of molten-sal'b
reactors fueled with *>7 is given in Part k.

 

6’.[‘he date used were for Units Nos. 5 and 4 of the Gallatin Steam
Plant, Gallatin, Tennessee.

7H. R. Reese and J. R. Carlson, “'I.'he Performance of Modern m:flaines "

Mech. Engr., March 1952, p 205,

 
 

 

Fbr the nuclear ana.‘l.ysis , the rea.ctor was conceptuslly resolve& into
& spherical core having & wniform temperature of 1180°F, a thin epherical
aore shell of INOR~8, & spherical snnulus of blanket fluid, snd & spheri-
cal reactor shell, A blanket thickness of 2 4t appeared to be sufficient
%0 prevent excessive losses of neutrons to the outside » énd a core vessel

‘thickness of 1/3 in, wvas used. A reactor shell thickness of 2/3 in, was

selected for the calcula.tions, but, in meny cases, the reactor shell was
neglected in order to shorten the calculetions.

The remaining independent va.z'-ia.bles of significence were the concen-
tretion of thorium in the fuel salt, the diameter of the corg, and the
fuel salt reprocessing rate. Of prineipal interest were the oorresponding

eriticel inventories of 02 5 and 02 35 end the regeneration ratio, In

Pexrt 4, the results of & paremetric study of the initial states are pre-
sentedj that is, the results are for "cleen" reactors, having no fission
fregments or nonfissionsble 1sotopes of uranium other than 02 38 present,
However, the optimum system could not be determined from such & study
alone; in pexrticular, the time efter startup when proceseing is initiated,
the method of processing, end the rate are important factors. The para-
metric study of various fuel reprocessing schemes that is under way at
present is deseribed in Part 4., This study 1s not yet complete bece.use '
the number of possible combinations of irdependent variables is quite
lerge. Therefore, a typleal get of cornditions, vhich may turn out to

be nearly optimum, was selec'bed for ;presen'bation.

A core diameter of 8 £t end a ’bhorim concentration of 1.0 mole g4
in the ruel eelt were select.ed a.s & reasoneble compromise between the

~ desire to minimize the - inventory oi‘ u2 35 a.nd to- maximize the regenera.-

tion ratio, The nuclear perfomance of the initia.l sta.te is set fourth

in T&ble l-5¢ .

A oonvereion ratio of 0 63 ie believed to be a.bout the maximmn tha.t

can be obtained in a. homogeneous molten rluoride sa.lt system with U255

es the fuel (see Part 4).  The perfoma.nce with Ue 35 would be subetan-

‘ tial]y ‘oetter, of course ’ a.nd regeneration ratios of 0.90 or higher could

555-

 
 

 

 

 

| T&blg_l;j. Initial Ruclear Characteristics of e Typicel

‘Molten-Fluoride-Salt Reactor

Core diameter: 8 ft
Power: 600 Mw (heat) |
Volume of external fuel eystem: 339 ft”
- U-235 inventory: OO04 |
' Regeneration ratio: 0.63 .

 

 

 

: Cation ' ‘
~ Inventory Concentration ~ Atom Density NEutron Absorption
(xg) __(mole %)  (Atoms/cm”) ___ Batlos*

| - ew0®
Core . | BRI ~ o

U-235 60k 0.25k 9,09 -

Fissions | ' S 0.729
n-y o U 0.27T1 ¢

U-238 k5.5 0.019 0.6Th . 0.039

'fl‘l 2100 100 . '32.0 ) 0.%1"

1 3920 61 1982 . (

Be 3008 .37 1183 (0.102

F 24000 - 47T ¢
Core Vessel 0.052
Blanket | | ' o | -

Th 30500 13 392 0.228

I 50% Tl 2139

Be 1460 16 : 482.2 - (0.021

F. 25100 l;671 ‘ \ -
.Ieakagé 0.004
‘Neutron yield, 7

1.8

 

% nbuirons absorbed per neutron absorbedfin flféfis. o

 
 

 

 

 

 

be obtained in the clean reactors. Further, as discussed in Section 2
above, the use of graphite to moderate the fluoride regctor may result -
in substantisl improvement.

The neutron balance is presented in terms of neutrons ebsorbed in
each element per neutron shsorbed in U235 Thus the sfim of the abéorp- |
tions in thorium end U2 give, directly, the regeneration ratio, end
the sum of all the ebsorptions givesM, the number of neutrons produced
by fission per neutron absorbed in U255 An examination of Table 1.3
shows that sbout one-third of the regeneration tekes place in the blenket.
The single, most importent loss of neutrone is to radia$ive capture in
U255; if other parasitic captures and leaskages could be reduced to zero,
the regeneration ratio would still be limited to 0.80 in this reactor
by rediative capture in 122, fhe other importent losses are to carrier
galt in the core and to the core vessel, which reduce the regeneration
ratio by 0.10 and 0.05, respectively. Losses to the blanket salt and to
leskage smount to less than 0,02 neutrons. | “

Of the neutrons lost to the carrier salt, the majority are captured
by fluorine, and the loss is unavoldable. The lithium is specified to A
be 99,99% L173 end the 116 content is estimated to be about equal to that
which would be in eguilibrium with the n- reaction in beryllium. Hence,
there is no point in specifying & lower concentration of I&sa The system
contains nearly 10,000 kg of purified Li7, Of the neutrons lost to the
core vessel, about ohe-third ere captured by the molybdemum; nickel cap~
tures account for mpst of the ramaining loss. Increasing the hardness
of the neutron spectrufi by increasing the thorium concentration tends to

decrease the absq;ptidhs in the carrier salt and in the core vessel, but .

this decrease 1s ‘moré than offset by the decline in N of U255 at higher:'
energies. |

The accumulation of fission fragments end nonfissionable uranium
isotopes tends to increase the inventory of U255 and to depress the re=-
generation ratio. Ehe production of U255 tends to ‘counteract these L

effects, N6verthe1ess, if the fission products are not removed, the

fiBTu

 
 

 

 

inventory of U-2? will increase repldly £ram 600 to 900 kg during the
first year of operation. The regeneration ratio will fall from 0.65 to
0.53 in the same period. About TO kg of 023 > will have accumulsted, of
which 85% will be in the fuel salt.

The nuclear characteristice of the system et the end of the first
year are presented in Teble L.k. As mey be seen, the increesing hardness
of the neutron spectrmn results in a decrease of losses of neutrons to
the fuel salt and to the core vesssl to O. 012, but this sa,ving is more
than offset by the corresponding decline in M to 1,78 (a.veraged over
all three fissionable :Lsotopes present). |

If the fission products were ellowed to &cemmzlate further, the
'5123 2 inventory would contlnue to rise. If, hawever, the fuel salt is
reprocessed continuously at the rate of one fuel volume per year {thus
holding the fission product concentretion constent), the 025 2 inventory
and regeneration ratio can be held stationary, as shown in Part L, Fig.
4,10, The continual incresse in the concentrations of nonfissioneble
uranium isetopes is compensated by the accumulation of 623 3. A neutron
balance for the system at the end of twenty years is giveri in Teble 1.5. |

As may be seen, 0256 is much more harmful than 0238 s 8ince it cap~

. tures 2.5 times as many neutrons and doez not form & fissionable isctope.

Despite these losses, however, the regeneration ratio does not decrease
appreciably, mainly because of the superior properties of 0233 s Wwhich
provides 4§0% of the fissions.

In sxmnary, once reproeessing t0 remove fission products 1is ‘begun,
nuclear performance of the gystem is stabilized to a satisfactory degree
for twenty years.‘ No provision for the rexhova]_. of the nonfissionable
isotopes of uranium need be made. | |

If desired, the trensients during the first year of operation cen’
be largely eliminsted by allowing the thcriinn congentration to decrease,
rartly through burnup end partly through withdrawal. Such & case is |
shown in Fig. 4.10 as a dashed line, in which the core reprocessing 1s

- 38 -

 
 

 

 

>

 

 

o)

Table 1l.k4.

Nuclear characteristics of & Typleal Molten-Fluoride-Salt Reactor

Power:
Load factor: 3
Volume of external fuel system: 539 ft

Core diameter:
600 Mw (heat)
0.8

8 ft

After Operation for One Year Without Reprocessing of the Fuel Salt

 

 

 

 

U-235 inventory: 890 kg
Regeneration ratio: 0.53
_; Neutron Fraction
Inventory  Concentration  Atom Dens%ty Absorption of
(kg) (mole %) (Atoms/cm”) Ratios* Fissions
x 1012
Core .
U-235 890 0.43 13.4
' Fissions . 0.618 0.861
n-y 0.262
U-233 61 0.029 0.926
Fissions 0.090 0.126
n-y 0.01k
Pu'259 608 0-003 00101 '
Fissions - 0.009 0.013
n.""'7 .. 0 0006
Th-23%2 2100 1.0 32.0 0.299
Pa-233 8.2 0.00k 0.125 0.005
1i~Be-F 0.080
U-234 1.9 0,0009 0.029 0.001
U-23%6 62.2 0.030 0.933 0.032
Np~237 4,2 0.002 0.062 0.004
U-238 57.9 0.058 0.860 0.0%6
Fission |
fragments 181 0,172 b 46 0.068
Core Vessel SN 0.0k2
 Blanket L -
 Th-232 30500 13 392 0.206
- Pa=233 - 5.5 0.0024 0.071
U-233 . B85 0.0037 0.110
Ii-Be-F. . e 0.010
Ieskage 0.004
Neutron Yield, 17 1.78

 

* Neutrons absorbed per neutron sbsorbed in U-235.

- 39 -

 
 

 

 

 

 

T&ble 1. 5
' After Ogeration for 20 !ears uith-

. Core- diameter.g

1oad fector: |
- _Volume of external fuel system‘
U=-23%5 inventory: '

0.8

 

- B-rt o
- Power: 600 Mv: (heat)

870 kg

339 £t°

Nuclear Characteristics of 8. Tygical,HbltenzFluoride-Salt Reactor
QWS vBsmoval'of Fission Products -

 

 

 

Feutron yield, n

............

Regeneration ratig: - 0_55_
- | . | " Newtron - Fraction
Inventory  Concentration Atom Density  Absorption of
(xg) (mole %) (A$cms/cm ) Retios* Fissions
o x 1019 |
| dore | ‘ , ‘
U-235 872 0.410 13,1
Fission - 0.407 0,550
n-y | | 0.182
U-233 312 0.152 4.85 | _
Fission 3 0.303% 0.410
n-)’ .. ) 0 0028
Pu-~-239 52,6 0.0k 0.778
Fission 0.03%0 0.040
n-7y | 0.022
Th-232 1.00 32,0 0.255
Pa-233% 7.32 0.0032 0.102 0,003
Ii-Be-F 0.073
U"23h‘ i 1201" 00058 1 87 00026
U-236 L8 0.210 6.72. - 0.1b47
Np-237 ; 0.015 0.471 0.019
U-23%8 -0.060 1.91 0.056
Fission , _ _
fragments 0.085 2.7T5 0.045
Core Vessel | 0.043
 Blanket
Th-232 30500 13 392 0.195
Pa-233 50 0.0021 0.064L5
- U-253% 33.0 0.01k0 0.hk22
- 1i-Be-F | : . | 0,009
1.8k

 

 

* Neutrons_dbsofbéd per nefitron absorbed in U¥255¢‘

- o -
 

 

 

 

 

 

 begun immedistely and the thori\m is removed at the rate of 1/900 per

day, in addition to the normal burnout at the rate of 1/4300 per day.

The critical inventory rises within one month to & maximm of 626 kg and
then falls to 590 kg at the end of eight months. At this time the re-
processing rate is increased to 1/560 ver day, end the thorium is returned
to the core. Thus, the thorivm concentration falls thereafter only by
burnout. The regeneration retio ie little different from that of the
previous case during the first two yeers, as indiceted by the dashed line ’
but it falls steadlly thereafter. The 112 3 inventory rises slowly » but
the UE5 5 inventory 1s stebllized at 200 kg after about six years. The
025 > inventory could have been stebilized &t the twe-year value by modest
withdrewals of thorium; however, the regeneration ratio would have fallen
faster and additions of U23 2 to compensate for burnup would have been
greater.

A necessary condition for the feasibility of a molten-salt reactor
1s the integrity of the core vessel. This member is exposed to high-
Intensity neutron and gama fields 'y end it is therefore subject to both

. radiation damage and thermal stress. With a preliminary estimate of the

hea.ting in a comparsble reactor having a pure nickel core vessela as a
basis and with ellowence made for such differences as diemeter snd com-
position of the fuel salt, the combined gamma and neutron heating in an
8-ft~die INOR-8 core vessel in a reactor heving 0.5 mole % ThF), in the

- Tuel salt and operating et & power level of 600 Mw of heat was estimated

to be not greatexr than 12 w/ m’ -of metal. The rate of heat relesase in

the 'blanket sa.lt was es’cimated te e not gree.ter than 50 Mw, exclusive
‘ ’of any contribution from fissions in the blanket » vhich may add up to
'a.nother 30 Mw. N |

 

BI..., G. Alexander and L. A Maxm, Firs’c Estimate of Genma, Hea.ting

in the Oore Vessel of a Molten Fluoride Convertexr, ORNIL~CF S wl2=57s

- 41 .

 
 

st

 

 

 

T. FPROCEDURE FOR PLANT STARTUP

The initisl startup of the plent will be accomplished in four steps:

(1)‘ prelimina.ry. checking of the systems, (2)_,prehea'_ting end £1lling of
~ the fluid circuits, (3) enriching to criticality, emd (4) operating et

lowpower. The integ'ity and proper functioning of the equipment will

be esta.blished insofar es possible, in the preliminary checking of the
system, This step will include, in addition to cleaning and leek testing,
checks of instrument snd alerm equipment settings end functloning, con-
tinuity and polerity of the electrical circuits ’ direction of rotation

of rotary élements , and opera.tion of vaelves and auxilia.:y gystems,

In the second step, the fuel, blanket salt, and sodium circuits
will be preheated to eabove the melting points of the various mediums

~and then filled. The preheating loads will be divided into menageable

sections that can be sutomatically monitored for hot and cold spots o0
thet thermal stresgses may be minimized. The systems will be filled at
temperatures as low as practical so that full advantage can be teken of
fluid circulation as & means of bringing the system to an isothermal con~
dition before enrichment. During the initial period of fluld circuletion
to establish the isothermal condition, the proper functioning of the flow
control equipment will be established. Also the cleanliness end metal-
lurglcal stability of the contaimment system will be evaluated by analyzing
samples withdrawn from the fluld systems. The operability of the fuel
system withdrawing and enriching equipment will be checked with barren .
salt; the high-temperature instrumentation will be checked; the draeining
and refilling procedures will be verified by testing; and remote mainten-
ance techniques will be tried out. This period of nonnuclear isothermsl
opera.tion at high temperature will also serve to familiarize the operating

 crews with the system and to establish their confidence in its operability.

The preheating and fillihg procedures will begin with the introq.uc-
tion of water to the steam generastors. The steem genmerators and the
gsodium systems will then be preheated to 350°F, which will produce &

- 42 -

 
 

 

 

pressure of 150 psi in the steam system, The éodim will then be pres-
purized from the sodium drain tenks into the primery end secondary sodium
systems, the sodium pumps will be started, and flow will be established.
Both sodium systems will thé_n be further heated to 600°F by using the
electric heaters end by making use of the fluid circulstion. Simulta-
neously the fuel and blenket circuits will be heated with electric heaters
to the same temperature. The pressure in the steam generators will have
risen to approximately 1800 psi end, before further system heating is
ettempted, & small loed will be imposed on the steanm generators to hold

the water temperaturé to 600°F as the rest of the circults are hested to
bigher temperatures. The steam genera.tors‘ will be loaded by bypassing

a small steam flow around the turbine, This load will be detexmined by

the smount of :éxcess power avé.ilable during the heatirig period from pumping
power end externsl heat sources in the systems. The load will be low |
relative to the design cepacity of the steem generators; there will probably.
‘be less than 1 Mw available for bypess steam 'generation in the five units.

At this jJuncture any increase in water temperature (above 600°F)
would result in overpressurization of the steem system, and if the steam
generators were allowed to evaporate to dryness end go to higher tempera-
tures, severe thermsl shocks would be imposed on the structures. when
water was again introduced. Theréfore, the sodium flow to the steam
generators will be reduced as the reactor systems »a.re elevated in tem-

: pera.ture. The flow in the secondary sodium loops will be reduced by

| lcwering the pump speed, wh:l.ch in turn will reduce the flow to the steam

o -'genera.tors ’ and the throttling valves will be mainpulated 4o refiuce the
f*'prcportion of the total flmr through the genera.tors and to ‘shunt the flow

| a.round the suPerheaters'_ .': o :" G

 

_ When the rea.ctor a.nd ;pr:!.mary sodium circuits have been preheated |
to 1100 Fy the sa.lts w:l.ll then be che.rged from the dmn_p ta.nks :Lnto the
process circuitry, and flcw w:l.ll be established. o ' ‘

-l|.3a-

 
 

 

 

 

 

 

 

At this time &ll the 'réa.cf.o;c ‘heat transfer loops will have been filled,
end the rea.étozj will be operating isothermally at 1100°F, The steem gen- o
erators will be running &t temperatures less than 625°F, and the super- .
heater end reheater sodium éiré.uits_ will be running et temperatures less
thsn 1100°F but greater than 625°F. The heat that is trensferred through
the systems by virtue of the 475°F gredient will be durqped in the steem
bypassing the turbine. This heet loed nmay. ve ira.riegi by éhanging the rate
. of dumping of the stéem and the sodium flow rate in the steem generstors.

When 1t hes been established that the plant is performing Se.tisf_a.q?- ;
torily and that the systems are tight and chemically cleen, the cfitical
experiment will be started. Fuel concentrate will be added to the resctor
'through the enrichment system. Approximately 38 ft3 of LiF-BeFE-UFu mixe
ture cénta.ining 2.5 mole % UF), will have to be added to the 530 £ of
carrier salt to achieve & fuel concentration of 0.15 mole % UF), » As the
concentrate is added, 1t may be necessary to withdraw fluid from the fuel #
system so that an adequate expension volume will be availsble in the | '
expansion tenk. The reactor will be titrated to criticality at llOOo‘F,"
and after criticality has been a.chievéd, & thorium~enriched salt will be
added to the fuel mixture. This will drive the temperature down, end
the oyerating temperature msy be finally trimmed by alternste additions
of fuel end thorium concentrate mixtures. By edding the thorium to the

- system as a last step, 1ts worth as e poison or chemical temperg‘.ture' sliim
may be evalusted before power operation. "

 

¥

A period of low power operation will fallow the criticality experi-
ment. By virtue of the negative temperature coefficient, the reactor
will be a slave to the demand loa.d; which will be imposed by _increa.s,ing" _'
the steam generation rate as & resulf of increasing the rate of sodium
flow through the steem system. Manuel memipulation of system comtrol -
velves will be required ;mt:!.l an epprecisble fraction of design power B
is obtained, sey, 30%. The rate at which the load may be incressed will
be determined bir the permissiblé rete of temperature change of the com-
ponents;: T o — o |
 

 

The turbine will be preheated by admitting steam thmugh ‘bhe turbine
control velves, a.nd when the turbine has been hesated and brought up to
speed, all the stemm will be directed 'bhreugh its normal path. At low
.power levels, it may be necessary to attemper the stesm so thet the tur-
bine temperature limits will not be exceeded

Normal plent resterts aefter power operation will follow the same
basic procedures, except that no criticael experiment will he required.
Since there 1s no control rod, close attention will have to be paid to
the fuel system filling rate and temperature so that nuclear trensients
will not be incurred.

8., RBEACTOR CONTROL AND REFUELING

The kinetics of circulating fuel reactors have been studied and
reported in a number of papers.9 A typicel velue for the temperature
coefficlent of reactivity for & molten-saelt reactor is «k x 10™7 (Ak/k)/o F.
This negative temperature coefficlent is sufficient to make the power
level in the reactor a slave to the applied load for all pormal. \ope‘re.-
tional power demand changes, without the use of control rods, As indi-
cated in the following section (Sec. 9), 1t keeps the reactor gafe from
excessive temperature excursions even under some rather a.dverse conditions.

The eritical temperature of the reactor graduelly deqreases during
opera.tion at ;power 8s & result of the burnup of fuel and buildup of fis-
~ sion product poisens. | At conste.nt power, all tempera.tureg in the heat
exchanger syst.ems decrea.se corresponding;ly, including, _in pa.rticular, the -
: temperature of the sod:l.mn retu:ming from ‘the superheater-boiler-nrehea.ter
systems This temperature must be maintained et ell times above an

 

L 9W. K. Ergen, Current Status of the Theory of Rea.ctor Inmamics, L
'—onm.-cr 55-7-157 (1§53); W. K. Brgen, "Kinetics of the Circulating Fuel
- Muclear Reactor,' Phys. Rev., 25, 702 {June 1954); J. A. Nohel, Ste.bilit.z
" of Solutions of the Reactor ‘Bouations, ORNL~CF 5%-9-25; Ws K. Ergen end -
"~ A. M, Weinberg, "Some Aspects of Nonlinear Reactor Dynamics,“ Fhysics xx,
k13 (1954); F. H, Brownell end W. K. Ergen, "A Theorem on Rearrangements
and Its Application to Certain Delsy Differential Equetions,"” Jowgmal of
Retional Mech. and Analysis, 3 565 (1954). ‘ B .

- 45 -

 
 

 

 

 

 

erbitrary minimm, determined by the melting point of the fuel, The tem-
perature of the return sodium is. therefore used es en indicator of the |
need for additional U25 > to restore the desired temperature level.

 The relation which gives the mass, AM, of fissionsble ma_.terial to
be added to the reactor for & given increase in the steady-state mean
core tempera.ture, Ty is given by the expression, -

¢AN[= ofM ZLT
vhere,

B = ____/ lkk
and,

ae —AE-I-{- /AT

For epithermal reactors, p has values between 2 and 10, usually greater
than 4, and can be obtained fram criticality experiments or by computew
tion. The reduction in the coolant return temperature vs the time ree
quired to burn up the corresponding mass (AM) of fuel, with constant
power generation of 600 Mw, 1s shown in Fig. 1.1l. For exsmple, if the
fuel inventory is 1000 kg of U23 2 , Bis 9, ails < x 1077 » @nd the sodium
return temperature can be allowed to drop 50 ¥, the reactor must be re-
fueled at intervels no greater then 13.5 days. On this schedule, the 023 5
addition required is 10.8 kg. The effect of bulldup of nuclear poisons
1s neglected in this calculation. In the flrst year of operation, the
increased inventory required to compensate for the poilsons requires more
frequent fuel additions. The calculation described gbove is typical cf
the conditions that exist after fuel reprocessing is initiated.

9. ACCIDENTS: CONSEQUENCES, DETECTION, AND REQUIRED ACTION
| The following diséussion gives the initial results of a study of - \'
- difficulties that may arise as & result of accidental occurrences in |

vaxrilous parts of the reactor system. Although no plausible accldents .
with inherently disastrous results have been postulated, the need for

- U6 -

 
 

 

 

  

UNCL ASSIFIED

(ge2 n 40 bY) dNNYUNB o
s | = O

 

- 5.:

 
 

| simula’c.or etudy is shown in Fig. 1.12, This diagrem segregates one of

 

 

9,1, An Instentaneous loss of Loa.d From & Secondsxy Sodium 01rcuit

T 10,

further experimental end design efforts to determine the most economical
way of handling some situations is epperent. | ‘

The tra.nsient beha.v:l.or of the remctor system has been snalyzed by
enalog computer techn:l.g_ues for several types of sudden changes :Ln the : 5
hea.t load on the reactor.lo The reac'bor flow disgrem essumed for the | -

the core heat transfer pa.ths for individual menipulation, s 80d lumps the |
others ‘together into one hest sump. The tempera.ture coefficient of re~
activity essumed was a4 x 10%7 /°F. o

 

This is the limiting cese of an accident occurring to only one of . .

the core heat transfer paths et the maximm distsnce awsy from the reactor. \

All temperatwes upstream of the feilure tend {0 become isothermal at

the new reactor outlet temperature, which is slightly lower then that
under full power. 'I'he. temperature change in the plping end heat exchangers
is rapid and emounts to 200°F or more. The heat exchangers, as designed,

will withstand the tempersture changes, but a complete stress analysis

of the plping layout should be made before such & reactor plent is built.

9.2. An Instantaneous Stoppage of Sodivm Flcw in One of the Primery Heet
Eb:cha.ngers

This case is similer to thet discussed above, except that tempera-
tures downstream from the primary heat exchenger drop quickly to a lower
isothermal temperature. During the trensient the mean core temperature
of the reactor rises to & pesk of at most 20° F sbove normal during & period
of time epproximately 10 sec sy while the outlet temperature d:rops auto-
matica.lly to its new value.

These two limiting cases ghow that there is no failure of & single
heat transfer path that can cause an excessive tempemture rise in the
reactor, '

 

E. R. Mann, privete commmication, ORNL.

- 148 -
-6{1-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

  

 

¢ .
o Pump PUMP
TN ~ |rransport
’:‘ Lag
T=L63 sec
l1210°F 20°F
| REACTOR
/ 600 Mw
1075 °F 925 °F .
Transport]
i el , Lag
3 can 3
40.2 ft>/sec. 13.4 ft7/sec. 46.! ft?/sec. T=1.7T5sec.

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

{lIst Order
Lag
T=2.5sec.|
1080 °F
150
Mw
HEAT
SINK
825 °F
Ist Order -
Lag
: 3
z= 2. 33.6 ft.°/sec.
PUMP 2.5sec. :

FIG. 1.12- REACTOR SIMULATION FLOW DIAGRAM

 

 

 

 
 

 

 

9.3; " An Instantaneous Reduction of the Heat Flow Rate from the fieactor Core

If,a,fuel pfimp should suddenly stop, the rate of heat removal from the
core would be quickly reduced to a fraction of that at\full power. The heat
removedfwouldfbe determined by thermal insulation iosses, byvthermal convec-
tion through the core fuel circuits, and by heat transfer to the blanket
~through the core vessel wall. The latter would be very significent if the
blanket pump remeined operative. During the first few seconds following &
~ sudden fuel pump faillure, forced circulation in the cbre circuits would per-

- gist és a resfilf of inertial effects but at a rapidly declining rate. The
sharply decreased circulation rate would result in a iarger fraction of the

delayed neutrons being reléased in the reactor core.

A limiting apprOXimationkof the effects of fuel ?ump stopping was studied
on the simulator. In the simulator studies, 1t was postulated that during
steady-state full-power operation, the heat removal was reduced instantaneously
to a small fraction of full power. It was further postulated that the flow of
fuel stopped instantaneously so that the fuel salt that was in the reactor
stayed there. The peak‘femfieratures which could be achieved if these condi-
tions could be met and the times to reach them are given in Fig. 1l.13 as
functions of the reduced heat removal rate. The temperature rise indicated
results from the continued fission power generation at subcritical conditions
from the gradual decay rate for the neutron flux end does not take into account
afterheat from fission product radioactive decay, which acts as an additional

heat source.

Thé curves in Fig. 1.13 should be used with caution; they are intended

" only to set upper limits on fihe temperature rise. The coasting effect from
the fuel's inertia and thermal-convection circulation will reduce the peak
temperatures markedly, but the relationships are complex and a more extended
analysis is necessary. The pesk temperatures are not a problem of themselves,
but their sudden appearance will cause thermal strains. The magnitudes'of
these strains and their effects on the integrity of the reactor requires

analysis, but no serious consequences are expected.

.50 -

 

 
 

 

 

 

 

 

o

 

 
 

 

 

 

 

9.k Cold Fuel Slugging |

If cold fuel is suddenly injected into the reactor eore when the

~ power level is very 1ow, the core may'become supercritieal on & fast

period¢ This can lead to the: poWer deneity exceeding the design level
before the mean eore temperature rises again to ite nermal operating

rangea Two gimulstor cases vere run te see whether these higher then
design level power densities ceuld 1ead to & serious temperature over-

shoot in the reactor._

One case eensidered involved suddenly 1ncreasing the power demand

at the beiler from 6 Mw to 600 Mw. The immediete effect is to 1ewer the N

core Inlet temperature to about 1000 F. The ayerage aore temperature 1s |
reduced'te ebeut 1050 F, The tempeneturee then rise asymptoticelly to
normal epereting levels with overshoot at moét of & few degrees. Ehe

- power level overshoots to about 900 Mw, but the oversheot in power hes

no practical significenee, A sudden load 1nerease at the boiler of this

‘magnitude is impractical to obtain, so that this is a limiting case in
- BO fer ag a suddEH applieatien of load is concerned. It must be eencluded
_ that "eold fuel slugging as & result of load.manipulation eannet lead

to any diffieulty,

A second caee was set’ up in en attempt to simlate. steppage ef fuel

rflbw, cooling ef the fuel in the héat exchangers to Just above its melting
‘_fpeifit, and then starting flow to. put a slug of very cold fuel in the =
reactor, 1In the starting condition of the simulator study, the reactor |

Was syberitical at a temperature greater than 1200°F, As the floW'was

-started end cold fuel was forced- into the reactor at the normal pumping

rete, 2 stepawiee increase in the reactivity of O h% vas artificelly -
inserted to place the reactor in & positive period. This insertion of_e_
positive period was intended to replace & condition of starting et very
low power,vsince the scaling limits of the simulator do not permit the
intradudtion of'initial'bcwer levels of lese than:B Mw. Under the simu-
lator eonditiens used, the reactor core. temperature again dyopped abeut
150 F and then rose asymptotically to the design tempereture with no .

 
 

 

 

 

perceptible overshoot. It is concluded that a fuel pump starting up
with cold fuel in the primary heat exchangers is unlikely to lead to
high temperature excursions in the reactor.

- 9. S. Rem0va1-of Afte;'heat by Thermal Convection

 

A survey 'éxamifiation of the capability of ‘the heat transfer system
for the removal of heat by thermal convection in the event that all

| pumping power is lost has been made, The temperature pattern of the

system reéluired for the removal of 4% of the design power by thexmal .

- convection is shown in 'Fig..‘ 1.14, This study shows in a preliminary way
- that thermal convection can remove enough heat from the reactor core so

that loss of power to the pumps in the redioactive areas will not neces-

 sitate the drainege of the fuel from the reacétor. A detalled system

enslysis may indicate thet slight modifications in leyout mey b: required
to accomplish this, however. |

9.6, Iloss of Fuel Pump

Any evefit which stops the forcéd ‘circulation of“fue}. shrough the -
primary heat excha.ngers req,uires tha.t steps be ta.ken to prevent freezing
of the fuel salt. The steam system is such a large, relatively low teme
perature heat sink that the fuel salt would be quickly frozer. if no action
were taken, There are two safety controls. First, fuel pump sfio;ppa.ge
or loss of power will cut the steam tb thé ‘fiurbine » 8nd reduce the turbine
output to a low level to handle ai‘terhea:b, etc. The second control,

,.triggered by a low tempera.ture :I.n 'bhe cold line of the primaxry sodium,
: ;will stop 'bhe sodium pmn;ps.

9.7. Loss of Elec'bric Transanission Line comection to 'l'.he Plant

In the even‘o of 1oss of electrica.l load on 'bhe pLa.nt s the turbine

" ‘,'”__-:-stop-valve wi].‘!. ad,just automatically *bo prevent turbine runa.way The

N "tu:r'bine ca.n be adJusted to & 5 to 10% re:bed loadll to supply lbca.l needs.
'.'-E[he r:I.se in pressure in the steam system resulting from closure of the

- -stop-valve will o;pen the emergency reliei’ va.lve until, to save purified

x>

 

llH L. Fa.]kenberry, TVA, pri'vate comunication.

T

 
 

 

 

 

 

EM,

A

 

- 54 -

£ PATTERN AT 4 7% HEAT -

 
 

 

 

 

 

 

water, the steam'bxpass valve is adjusted for the dumping of steam to

the condenser._ Simultaneously, or as soon as practiceal, secondary sodium
pump speeds will be reduced and valves controlling sodium flow to the ‘
boller, superheeter, and reheater and in the bypase_will adjust antomatig .
cally to glve design temperatures and pressures for the smount of after-
hesating being removed from the reactor. If this idling power exceeds

the power required to operate the plant it will be dumped to the condenser.‘-

The whole plant will be maintained in & condition ready 1o resume
its electrical loed as soon as it can be re-established | It is to be
noted that‘should the load loss-bewsufficiently prolonged so that the
'afterheet 1s not sufficient to provide power for local needs, the core
will generate fission heat automatically. In order to maintain the power
station:in a standby condition during a period in which, say, the elec-
tric generator equipment is inoperative and there is a simulteneous loss
of pover to the plant, emergency power generation equipment will be
needed. The emergency supply must have sufficlent capacity to operate
instruments, controls, feedwater punps end auxiliary equipment necessary
for control end removal of afterheat.

9.8, -leak Between Fuel and Blanket Salts

~ The free surface of the blanket salt 1is above the free surfece of
the fuel salt, and the blanket selt is more dense than the fuel salt,
Both the core and the blanket will have e common ges pressure over them, .
and both are on thejsucfiion'side of'the:pumps in their respective systems.
_,Under these conditions the blanket will always be at & higher static pres-
:sure than the’ core, and any 1eak between the fuel end blanket salts will
| drive the blanket salt into the fuel salt and lower the critical temperan

.ture of the reactor core. gj<

With such a leak fihe maintenance of ayatem.tem@eratures would re-
quire addition of fuel at a-rate in excess of that required for burnup
end’ fission product poisoning.. To mainxein the critical temperature |
| constant in a clean, 8-ft-die core with a fuel selt . contadning 0.75 mole %
thorium, ebout one atom of U256 must be added for three atoms of thorium

- 55 =

 
 

 

 

that lesk into ‘hl*e core from the blenket. Thus, 1f the fuel accounta-
bility is sufficiyently sensitive to detect a 10% excess fueling rate,
inleskage to the core of more then 165 cm? per day will be detected.

The fuel end blanlcet salts are chemically inert with respect to each -
other, and therefore no chemical effects of the mixing are ex_pected. |
if fission-product and heavy-element poisoning were to mask 'bhe excess
refueling caused by a blanket leak and prevent early detection, the lesk
would be detected eventual]y by cerresponding chenges in fuel end blanket
inventories s 85 indicated by the level ind.:l.cators of the respective
systems.

~ Once a lesk between the core and blanket was detected, the reactor
would be shut down and all liquid systems would be .drained. Replacement
" of the reactor vessel would be required, and this would be & 1engbhy

operation,

Complete rupture of the core vessel would lead e.utomatica:l.]y to &
suberitical cOndition. The core surge tank would £111 as the bls.nket
end core pressures tended to equalize.

9.9. Ieak Between Fuel and Sodium

The relative pressures in the fuel and sodium systems will always
be such that, in the event of a leak between the fuel and the sodium,
the fuel will enter the sodium stream. This errengement is used because |
the consequences of precipitation of urenium in the cir’cula.fiing fuel
system cannot be predicted with certainty.

The chemical cbnsequences of a leek of the fuel into sodium, such
as could occur in a primary hest excha.nger, have been examined on the
basis of thermodynamic data.la When the fuel is mixed with excess sodium,
the major constituents, except IiF, will be promptly and simultemeously -
réduced to their metallic states, according to the reactions:

 

12W.- R. Grimes, private eommication, ORNL.

- 56 -

v

C .
 

 

 

(1) -UFll_+Na.$NaF+UF3  AF = =36 kcal

(2) -UF5 + 3Na = U + 3NaF AF = =37.3 keal
(3) BeF, + 2Na =Be + 2NaF ~ AF = =29 kcal
(4) ThFy, + 4YNa<==Th + 4kNaF  AF = -hk keal

Of the fission products contained in the fuel, the alkaline earths, the
rare earths, end a considerable fraction of the alkali metals will remain
in the salt phase as fluorides, while Mo, Ru, Zr, cd, Zn, Sb and Sn will
be reduced to the metallic state.

The anions, particularly I]'5 T and Br87, which are important for
neutron detection because they are long-lived precursors of deleyed neu-.
tron emitters, will appear es halide ions. Accordingly, the salt mixture »
after reaction, will contain about 5% mole % NaF, 46 mole % IiF, and traces
(insofar as concentration is concerned) of fission product fluorides.
Such & mixture will have a melting point greater, probably, then 700 C.
(1300°F) » and accordingly will be solid at the normal temperatures in the
sodium circuilt.

- Metals such as Sn, Sb, Cd, Rb, Cs, and Zn should be soluble in molten
sodium, but all other materials introduced into the sodium by fihe fuel
are moderately high melting and will be sparingly soluble in the sodium.
Beryllium metal, which is present in relatively large concentrations a.nd

which appears to be relatively insoluble (< 100 ppm) in sodium, will
| probably be the first material precipiteted. - Sodium fluoride probs.bly

dissolves to the extent of 0.2 mole % in sodium at llOOoF, and this selt,
along with LiF, will exceed the solubility :l.n molten sodium and exist

as. separate solid pha.ses after relatively small quan'b:l.ties of fuel have
lea.ked into the sodium. '_' - o : ' o

Sodium iodide and sodium bromide a.re more soluble tha.n sodium fluo-
ride in molten sodium, and these precu:rsors of the delayed neutron enitiers
will, accordingly, be dissolved in the molten sodium until the NaF-LiF
mixture saturates the sodium end forms a seoond phase, - Since they are

‘-'57_-

 
 

 

 

 

more soluble in the salt phase than in the liquid metal' they could, in

, principle » then decrease in concentration in the molten metal due to their

| extraction into the solid salt phase., This extraction process is not

~ expected to be importent, however, since ‘the emount of solid salt phase

- will be small for a considersble.period, end extraction by a solid from
a liquid should be relatively slow. | M'bhemore s precursors of delsyed
neutrons present in the sodium a.rise only from the :t‘reshly leaked-in fuel,
and therefore the concentration of precursors wil_'l. not be appreciably .

_ affected even though the totel atomic: species concentration may be dimin
ished by extractions. '

Prompt detection of small fuel leaks into the primary sodium clr-
cuit poses a problem yet to be solved. Al cm3 /day leak will produce
approximately 0.5 n/ cm e geC by precursor decay at the secondary heat
exchanger, but it is doubtful that neutrons of such a source strength |
can be detected in the sodium cell. Likewise the neutron actiiration of
the primary sodium produces gama;ray activity which would tend to mask
fission fragnenn gamma activity. Detection of large leaks would ‘oe ‘aided
by comparison with the activity in the other similar secondary sodium
circuits, but the determination of the size of lesk that can be detected
has not as yet been mpde. . .

A welleagitated stoichiometric mixture of sodium and fuel salt will
result in & rapid temperature rise, estimated to be 1200°F under -adiabatic
conditions. It is difficult ’ however » for such conditions to ex_isn -in -
a practicel situation. Heat evolved from a small leak would be rapidly
carried away by excess sodium. For the larger leaks which could occur
from fatigue in bending or 'bension, the solids formed would in‘berfere .
with repid mixing. Some work has been done with a NaF-ZrFu base fuel and
NeX which demonstrated this smothering effect; further engineering tests
- 'will ‘be required to demonstrate safety with sodium and the present fuel
- salt in simuleted component equipment.

, ..5_8_..
 

 

 

As soon as & fuel to sodlum leak had been detected and the feulty
heat exchanger had thus been located, the reactor plant would be shut
down, the fuel and appropriate sodium circuits drained, and the heat
exchange:r replaced.

'9.10. _Leak of Fuel or Blamket galt to Reactor Cell

The presence of a small lesk from the fuel to the rebetor cell can
be defiected"by gas=sampling techniqfies. Its léca:l:ion will be more dif-
ficult to determine, 'The repair of such a leak would of course require
draining the fuel salt. The provision of an inert atmosphere in the
reactor cell will prevent rapid growth of lea.ks ca:used by salt-fluxed
oxidation. ‘

A gross leak or ruptu:re of either the fuel or bla.nke'b circuits is
a major accident. Means must be provided to vent reactor cell pressure
as it is bullt up by heat release from the spilled galt , and & suitable
noncritical emergency drain system that can handle afterheat on & one-
time basis must be availasble. Ways are known for doing both, but the
lowest cost way of accomplishing these disaster pmmw.tdéva a'teps. has not.
yet been determined.

9,11, Ieaks of Water or Steam to Sod.ium

The sodium in thermal contact with the water or steam is nonradio=-
active. The problem of leaks peWeen the water and sodium systems has
been faced by those engaged in the development of fast reactors, end
thelr studies and test results will be useful in determining heet ex-
changer desigl.

 
 

 

 

10. CHEMICAL PROCESSING AND FUEL GYCLE ECONOMECS

10.1. Fuel Salt Reprocessing

 

The system for chemical reprocessing of the fuel salt is e cambina-
tion of the ORNL fluoride volatility and the Kfi25 uranium.hexsfluoride
reduction processes. described in Part 6. 'The salt to be reprocessed is /

| transferred as described in Section 3, above, from the reactor circuit
%0 & holdup vessel on & convenient schednle, such as 2 ft3 once eaeh day

or 12 ft3 once each week. The holdup vessels provide containment during

~ the holdup period reqnired for decay of the short-lived activities and

act as a buffer between the reactor end the chemical plant so that the
operation of the reactor need not depend on the state of repair of the
chemical plant.

The fuel salt will be fluorinated in batches of 2 ft3 each, one
batch per day. After the uranium is removed by fluorination and collected
as'UFé on NaF pellet beds, the barren salt is transferred to waste storage.
The UFB'Will be discharged on a twice~per-week cycle from the NaF pellet
beds, which have a capacity of 10 kg of uranium. The volatility process
produces liquid UEB, in cylinders, which is subsequently fed to a reduc~
tion tower to produce UF), which is combined with fresh salt for return
to the reactor. The uranium losses in the chemical processing are sbout
0.1%, i.e., about 1 kg/year.

10.2. Blanket Salt Reprocessing -

Chemical processing of the blanket salt is physically much the same

as the processing of the fuel salt except that, after fluorination, the

salt is returned to the blanket system. Because of the much lower power -
density in the blanket salt, holdup for decay-cooling is not & problem.

~ Separate fluorinators for fuel and blanket salts, to prevent cross-

contemination, are assumed, as are separate NaF beds, to make possible

e the withdrawel of pure 0235 from the system i1f desired. The seme UFB

reduction tower will serve both fuel and blanket salt processing.

- 60 -

 
et b NS i s

 

 

A blanket selt processing rate that is about‘ the same as that for
the fuel selt is assumed, i.e., one 2 ft3 batch per day. Thus fluorina-
tion equipment of the same size will suffice. The uranium throughput

‘rate of the blanket salt processing system is , however, only sbout 10%

of that of the fuel salt. For convenience s the same size of NaF bed is

proposed for the two systems, elthough this meens that the UF; will be

discharged from the NaF bed in the blanket salt system only once every
other month,

The blanket salt processing rate is sufficiently fast to hold the .
U25 5 inventory in the blenket salt system to about 60 kg and to limit the
fissioning in the blanket salt to about 3% of the total.

10.3. Cost Bases

Fissioneble isotopes have been velued at $17/g in computing inven-
tory and burnup charges and breeding end resale credits. Capltalization
rates were assumed to be 4%/yr on fissioneble materials and 14%/yr on
everything else. The fuel salt was estimated to cost :';5].2'78/:13‘1;5 end the -
vlenket salt $2517/ft”, The varisble cost of fuel selt chemical process
sing is assumed to be equal +to the cost of buying new salt to replace
that process'ed.l The blenket salt is used over the life of the reactor
without excessive fisslon product buildup.

The fissionable materiel comsumption cost is based on feeding 93%
enriched 112 3 5 to the -core system to compensate for & regeneration ratio
of less tha.n unity. It is assumed that lf? 35 :Ls not availe.ble for purchese

- E=h m ecms\mic pz‘iue, a.l'bhough it would 'be worth a.pproximately twice &s
o .much as 025 5. 1n a.n mtemediate-neutron-energy mol'ben-sa.lt reactor due
L to its higher regeneration ra.tio et 1orwer cri'bical inventories. It is
o a.ssmned a.lso that isoto;pic re-enrichment of Ua 32 or 023 2 .elther by
| ga.seous diffusion or by excha.nge w:l.th & price penalty, 1s not econcmice.l
~ BO tha.t the molten-sa.lt power reactor must tolere.te the nonfissiona.ble
| 'uranimn isotopes a.nd 'c.he resulting 1ower regeneration rat:l.o and higher

U2 3 3—1125 2 inventory

- 6l =

 
 

 

 

 

10.4, Chemical Plant Capital Costs

The budgeted capital costs for the QRNL volatility pilot plant total

about $l,500 000 through fiscal 1959, This figure includes replacements

and.modifications, which should not be required in & second plent, end

1t also includes solid fuel element handling end dissolution facilities,

vhich would not be required in the molten-salt reactor plant. On the

other hand, the $1,300,000 does not include building and service facili-
ties, or any eqpipment for reducing UFg to UFh and, reconstituting fuel

salt. Aflditlons and subtractions considered the reference design .
chemical plant equipment and installation cost 1s. estimated to be $l,500 000,
The chemical plant's share of the total reactor capital investment is

_about twice this amount, when charges for bullding end site, design,_general~

expense, and contingencies are added. These capital costs are listed
with other capital costs in Section 11.

10.5. Chemical Plant Operating Costs

The ORNL volatility plant operating budget for three fiscal years
(1957-58-59) totals $1,368,000. The molten-salt reactor chemical plant
would have lower "unusual” costs (associated with development) than the
pilot plant, but higher "production-proportional” coSts, apd 1s estimated
to cost $500,000 per year to operate. To this must be added the cost
of replacing the fuel salt processed, or reclaiming it, if this can be
done for au equal or lesser cost. This is estimated to be 600 ft3 (ap-'
proximately one fuel system volume) per year at $1278 per ft5, a total
of $770,000 per year. A salt reclamation process might be expected to
reduce this considerably, although probably not more than by a factor
of 2, which nevertheless would save sbout 0.2 mills/kwh. The chemical
plant operating costs are listed with other operating costs in Section
11. '

110.6, Net Fuel Cycle Cost

 

For the purpose of estimating fuel cycle costs, values averaged over

the reactor lifetime of 1000 kg for the U253 U235 inventory and 0.5 for

thereffective breeding ratio were assumed. The nuclear heat power was

-62-
 

teken to be 640 My, the net electrical output 260 Mw, and the load factor
0.80. The net fuel cycle cost is estimated to be ebout 2 mills/kwh:

 

Ttem T - mills/kovh
¥ consumed o 2,260,000 S0 1.2k
Fuel salt make@ | 770,000 0.k42
R inventory 680,000 0,37

. 2,03

If‘ & comparison of total fuel cycle costs with thoée for a solid fuel
element power reactor are to be made , the chemical pla.nt capita.l cost
end the chemica.l plant oPerating cost should be added. These amounts
are as follows:

$/yr | mills/kwh
Capital cost ($3,000,000) 420,000 - 0.23
Operating cost | 500,000 0.28

They lead to a total mél cycle 'cost 61’ 2. Sy-mills/kwh.

,

11. CONSTRUCTION AND POWER COSTS

11.1. Capitel Costs

The information available in the preliminary design does not lend
itself to & rigorous cost ana.]ysis 5 hcwever, the power cycle has been:
| -sufficiently well-defined 'to pemit a segrega.tion of the major compoa
. 'knents in the plant.. The plant layout has progressed to the extent that

~ the over-a.ll size may be detemined. 5

 

 

| Des:l.gn studies of some of 'bhe fuel system auxiliaries have pemitted
a detailed cost brea.kdown The high-tempera’cure sodium pump | requirement.s
have been ascertained to the exbent tha.t manu:facturers of this equipment
*Vhave been a'ble to ma.ke prelimine.ry cost estimates. 'I'he fuel and blanket
. salt pumps were estima.ted by sca.lingup costs ‘of smaller pms that have
been febricated and tested at ORNIL for high-temperature reactor systems.

- 63 =

 
 

 

 

 

ORNL's experience in molten~salt and alkali metal heat transfer
equipment fabrication and procurement has been drewn on to estimate the
increased cost in producing reector-quality products.

In some cases individuel ccmponents were found to be too mumerous
for detailed cost analysis in the time availsble, and costs were essigned
to entire suhsystems on the bésis of general experience. The instrumen-

,tetion, electrical equipment and auxiliary systems were treated in this

menner, for exemple. | ' -

It has been assumed that the molten-salt reactor plent would be con-

-structed et a site similar to the one selected in a recent ORNL ges=cooled

reactor study. 3 Therefore, site ecquisition, imprevement, and structure
costs have been set at comperable levels,

 The capital cost summary is presented in Teble 1.6. It should be
noted that a 40% contingency factor has been epplied to the reactor por-
tion of the system. It is felt that there are a'rumber of uncertainties
in some of the larger reactor cost'packeges and a contingency factor of
this order is warranted. A T7.5% contingency factor was applied to the
remainder of the direct costs. |

The general expense or indirect costs charged to the plant represent
administrative, personnel, plant protection, safety and special construc-
tlon services which are largely incurred during comstruction and startup
0peraticns._ The design cost represents approximately 5% of the direct .
cost subtotal before the contingency factors were applied. This capital
cost sumary leads to & cost of $269 per installed killowatt of generating

capacity. | : . ¥

4

Table 1.7 presents a more detailed cost breakdown of the reactor

‘portion of the plemt. The major-emthents or items have been.listed .

end the materials of construction for a particular liquid system have
been indicated.

 

0me ORNI Gas-Cooled Reactor, ORNL-2500, Part 3 (April 1, 1958).

-6l -

O

 
 

 

10.
11.

-13A,

13B.
1k,
15,
16,

18.

‘Teble 1.6, Capital Costs
(FPC Account Numbers)

Lend and land rights

Structures and im@rovements

Reactor system (1ncluding chemical plant)
Steam system

Turbine-generator plent

Accessory électrical\equipment ,
Miscellaneous power plant equipment

Direct costs subtotal

T.5% contingency on 11,13B,1k4,15,16
409 contingenqy on 13A

Contingency subtotal

‘General expense

Design costs
| TOTAL COST

- 65 =

$ 500,000
7,500,000
20,232,000
3,750,000
11,750,000
4,600,000
1,250,000

 

49,582,000

2,201,000
- 8,093,000
10,294,000

 

7,500,000
2,450,000

 

- $69,826,000

 
 

 

 

 

 

I..

| II.

III.

‘Table 1. 7. Reactor System Capital Cost Summary
- (Section 13A of Capitel Casts)

'Fuel System (INOR-8)

Reactor core and blanket shell . $ 500,000

A,

B. One 24,000-gmm purp, pump shielding, - 81+5,ooo
- &nd motor : )
C. Four fuel-to-sodium heat exchangers - 672,000
D, System piping | | 100,000

E. Main fill-ond«drein system . _ 520,000

F. Off-gas system (includes blanket system) ‘- 568,000
G.' Enriching and withdrawal system | 100,000

exclusive of chemical plant B ' ;_

- Ho Preheating end insulation ' S 75,000
Blanket Circuit (INOR3S) |

A. One pump and motor | 350,000
B. - One blanket salt~towsodium hea.t exchanger ’

C. System piping - 20,000
D. Msin fill-snd-drain system | 120,000
E. Enriching and withdrawal system 50,000

| exclusive of chemicel plant |

. Preheating end insuletion - | 15,000

Intermediate Sodium System (stainless steel)
(4 fuel and 1 blanket circuits)

A.

B.

Fuel~toasodium systems:

1, four 20,000-gpm pumps end motors 960,000
2. four sodium-to-sodium heat exchangers hlS ,000
3. system piping \ 300,000
Lk, drain systems | 100,000
5. preheating and insulation 75,000
Blanket salteto-sodium system: '
1.  one 10,000~-gpm pump end motor 130,000
2. one sodimn-to-ssodium heat exchenger 45,000
5. system piping . 75,000
Lk, drein system | 30,000
5. preheating and insuletion . 20,000

=66 -

3,380,000

651,000

2,150,000
 

 

VI.
VII.
VIII.

X

XII.

 

IV.

Table 1.7. (Continued)

Secondary Sodium Circuits (Cr-Mo alloy -steel)
A. Fuel=to=-sodium-to-sodium systems:

1. four 15,000-gpm pumps and 2-speed drives

2. four sodium-to-water bollers

3. four sodium~to-steam superheaters

i, four sodium-to-stesm reheat exchangers

5. twenty remotely~operated throttling
valves

6. system piping -

7. fill-ande-drsein systems

8. hesting and insulation

B. Blanket salt~to-sodium-to-sodium system:

I. one 10,000~-gpm pump and motor
2, one sodium~to-water boiler

3« oOne sodium~to~stesm superheater
4, four throttling valves

5. system piping

6. fill-and-drain system

T. heating and insuletion

C. Sodium emergency drain system

Reactor Plant Shielding
(17,000 cu yd of concrete at $100/yd)

Mein Conteinment Vessel, Air Lock, Reactor
Support, and Cell cooling System

Instrumentation _
Remote Maintenance and Hendling Equipment

Auxiliary Systems
(helium, nitrogen, cranes, cooling systems)

SPare Parts:
- Pumps -

','B‘ ‘Heat exchengers tli"

Miscellaneous '

7 Original Inventories-',,‘

A Sodium (300,000 1b x_ $0. 20/lb) '
 B. Blanket salt (750 £t7 x 3.2 x 2517/1'155)
- €. Fuel salt (575 3 x 1.2 x 1278/ft.5)

” Ghemical Plant Equipment N

1,000,000

336,000
252,000

380000'

350,000
175,000
150,000

145,000
40,000
38,000

80,000
50,000
40,000

200,000

3,626,000
1,700,000

450,000

750,000
1,000,000
525,000

700,000
400,000
200,000

60,000
2, 260 000
880 000
1, 500,000

$20,232,000
e e —

 
 

1l.2. Power Costs

Power costs hawé been divided into three categories. These are:
fixed costs, operation and maintenance costs, and fuel cycle costs.
The fixed coste are the cherges resulting from the cepital investment
in the plent, This emount has been set at 144 per annum of the invest-
ment, which includes taxes, insurance, and financing charges. This leads
to an ennuel cherge of $9,776,000 or 5.37 mills/kvh. | -

The operation and maintenance costs are, in the main, dependent on
the ultimate relisbility of the reactor portion of the plant. The de-
velopment of practicel remote-maintenance techniques for the repair and
replacement of equipment in the redioactive systems'is also vitel to
| . assure reasoneble costs. No accurate determination of such costs cen
1 be made without further experience.

 

For the purpose of this report the operation and maintenance cost
breskdown given below has been assumed: | |

Amnugl Charge -

Labor and supervision $ 900,000
Reactor system spare parts |
Pumps 250,000
Heat exchangers ' 200,000
Miscellaneous 300,000
Remote-hendling equipment | 150,000
Chemical plant operation 500,000
Conventional supplies 400,000
| Total $2,700,000

This total cost results in an incremental power cost of 1.48 mills/kwh.
Net fuel cycle costs as discussed in Sec. 10 above amount to 2,03 mills/kwh.

 
 

 

The three categories add up &s follows: -

 

Annusl Cherge - mills/kovh
Fixed cost . $ 9,766,000 537
Operating and meintenance 2,700,000 - 1.48
Fuel charges 3,710,000 2.03
Total annuel charge $16,176,000
Total power cost \ 8.88

The difference in cost between having the :'eactor plent on standby and
heving 1t on the line is about 2 mills/kwh. |

12, SOME ALTERNATES TO THE PROPOSED DESIGN

12.1, Alternate Heat Trensfer Systems

The possibility of replacing the fuel-toésodim-to-sodium-to-steam
heat trensfer system with a fuel-*bo-gas-to-steém ‘sy'stem has been given
a cursory exsmination. The attractiveness of such & system' is based on
the replacement of two sodium systems in series with one gas system and
in heving the gas chemically competible with both the fuel and weter or
steam. Early estimates of the gas heat transfer performence indicate
that the fuel volume required to transfer an equivalent quentity of heat
would not be appreciably different from that requii-ed with the sodium
system. These estimates were based on use of & return gas temperature
below the melting point of the fuel, and the sa.fety of this procedure

" must be examined further. If the gas system were operated in the 300~
to 1!-00-1)51 range, the power required to circulate the gas eould:be.kept.at

a reasonable level. -

The gas system would permit & reduction in the nunber of heat ex-

j cha.ngers a.nd pumps, s.nd eliminate sodium va.lves. The dew point of the

gas would provide & rapid method of leak deteetion in cese of e steam-
to~gas lee.k A steam leak into 'bhe gas system would no-b heve the chemical
hazerd that exists with a steam-to-sodium leak. If the reactor were

- 69 -

 
 

 

 

operated inside a pressure shell, the fuel pressi;re could be maintained
slightly below the gas pressure to ensure that any leeks in the fuel
system would be inwerd. The small pressure differential required between
the fuel and the gas would permit the maximum fuel (gage) pressure to

be maintained at a level no higher than that required by the liquid-cooled
system. ‘

Gas _eooling would eliminate the need for the sodium handling eystems
with their sttendent preheating problems. These sodium facilities would
be replaced with gas storage and hendling equipment, Startup and shute
down procedures would be simpler with ges then with liquid cooling, par-
ticularly with respect to preheating end pert losd control, The amount
of secondary rediation shielding required with the liquid system would.
be considersbly reduced with the gas system because of the decreased in-
duced activity of the coolant. |

More studies of the gas cooling system are being made, and it is
spparent that the bulkiness of the gas system will present handice.ps; .
It 1s also probeble that a gas cooling system will prove more exjgensive.
A better comparison of the gas=cooled system with the sodium-cooled
system will result from a more detailed design study.

An elternate o the steam—cycle described above would be the
Loeffler boiler cycle. In this system (see Fig. 1.15) all the heat is
transferred to the steam in the superheater. A portion of the superheated
steam is recirculated by means of a steam pump to the boj.le_r,' where it
transfers heat to the water by direct contact to form saturated. steam.
With the seme steam conditions of 1000°F and 1800 psi, it would be neces-
sary to return approximately 2 1b of steam to the boiler for every pound
sent to the turbine. The advanta.ges of this system used in conjunction
with the molten-salt reactor are principally connected with the control
of heat flow. :I.'bh the sodimn—to-steam generator heat transfer system,
the steam genera.tor represents a 1arge ca.pacity heat sink &t a tempera-
ture more than 200°F below the freezing peint of the fuel bearing salt. |
To prevent freezing of the fuel, careful con’crol of flow in the seconde.ry

u70-i:

 
 

 

 

U e
NL- L'P Dsiz,ezoqzoz

~LF E/P TURBINES

 

o REHEAT
| BOILER

 

 

 

 

 

      
   
  
   
   

oy LIL
X FROM FUEL
FROM

BLANKET

N&r‘—w P TURBINE
i ~— AT TEMIFERATOR

N EMERGENCY RELIEF

e COVING WATER
~ CONOENSER

~—LDf-OUPERHEATER
~—FLOUCING STATION

1

 

 

!

 

1
Y

 

 

 

 

  
 

ary SUPER
HEAT EXCH, o HEATER C\-conoevsare PumiPp

 

 

1£

 

 

 

 

~SODION OFF INERT ST
o FEEDWATER
HEATERS W SERIES

. -DEAERATOR

q ~FEEDWATER PUAMP

 

 

p——

STEMNT PUIIE ~—f 7O BLAONKET

L
—-{} TO FUEL FEEOWRATER
—. HEATERS /N SERIES

 

 

 

 

Fig. 1.15. MOSPR WITH LOEFFLER STEAM SYSTEM FLOW OWIGRANM

 
 

 

 

 

 

 

sodium circult must be mainteined at low'pcwer operation. In the
loeffler boller eystem, the directly coupled heat sink is dry steem in-
‘stead of water, end control of the steam circuletion is believed to be

| easier then the control of the sodium flow. The elimination of the need
for two speed pumps and control valves in the intermedisate circuit would
therefore result in a system that would more fully exploit the inherent
self-regulation of the molten-salt resctor. ”

The elimination of the steam generator from & sodium circuit reduces
the need for sodium flow regulation. The reduction bf sodium eguipment
probebly would result in less frequent meintenance, Thus some of the
serious objections to having radiocactive sodium heating the steem would
thereby be leSSened and consideration could be given to the elimination
of one of the intermediate circuits. In addition, since minimm tempere-
tures would, &t design point operation, be sbove the melting point of
sultable fluoride salts, their use in place of sodium should be exemined.
If substitution could be made, chemical compatibility of the intermediate
- fluid would be markedly improved both with respect‘td the fuel and the
steam and these hazards would be lessened.

More detailed design comparisons will be necessary to evaluate,this
boiler system. Although the changes suggested above are plausible, the
detailed consequences must be snalyzed and felr cost comparisons made. |

12.2. Alternate Fuels

 

The substitution of U235 for U255 in the molten fluoride reactors
would result in substantial improvement in performance. Uranium-233 is
& superior fuel in almost every respect. The fission cross section in
the intermediate range of neutron energies is greater than the fission
cross sectlons of 0235 and Pu?59. Thus, initiel inventories are less, -
and less additional fuel is reqfiiréd to over-ride poisons. Alsc, the
n=-y cross section is substantially less, and the radiative capture
results in the immediate formation of a fertile isotope, P, e rate
of accumulation of 0256 is orders of magnitude smaller then with fl255
fuel, end the buildup of Np>2! and Pu->? is negligible.

—72.-"

O

 
 

 

 

 

Prelimina.ry and iné.omplete results from e parametric study of re-

- ac‘bors fueled with 0235 exe given in Part 4, Sec. 1.2, In a ty;pica.l

case in which the core diameter wes 8 £t and the aoncentration of ThF) -
wes 1.0 mole %, the initisl critical ma.ss was found to be on.'ly 87 kg of

‘ 0235 the inventory for a 600-Mw system was only 196 kg, the regenera.tion

reatio vas 0. 91, end the long—term performance was good. In another Befte
dle core system with 0,75 mole % ThF) in the fuel selt, the initial in-
ventory was 129 kg, and the conversion ratio was 0.82. After operation
for one yesr at & load factor of 0.8 end with no reprocessing of the core

 to remove fission products, the inventory rose of 199 kg, and the regene

eration ratio fell to 0.7l. However, if the reprocessing required to
hold the concentration of fission products constant wes stnrbed efter 1
year of operation, the inventory increased slowly up to only 247 kg
&f'ter 19 years and. the regeneration ratio TOSE sl:l.gh*tl;y to 0.73. Roughly
spea.k,ing, the eritiecsl :anentories regquired for the 0235 systems ere
about one-third those for the corresponding 1)23 2 systems, eand the burnu,p

.req_uirements are a.bout helf.

The ebove described case is not optimized for U255 __ Substa.ntial -
improvement can be obtained by using higher concentra‘oions of thorium
and correctly ma.tching the diemeter end processing rates.

-~ As discussed in Part 2, Pu:l?‘5 has eppreclable solubility in mixtures
of LiF end BeFpe It should be possible to maintein concentrations of
up to 0. 2 mole % sa.fely. - 'I'his is more than ample for clea.n systems
having dlemeters in the range from 6 to 10 £t with no thoriwm in the core.

A typical 8-ft-dia. core would have a. critical concentra.tion of 0,013

mole % PuF; end e regeneration ratio- ('l‘hFh in the blanket) of a.bout 0.35.

240

- The ei‘f’ect of accumulation of fission products and Pu on the cr:l.tica.l

concentration a.nd 'bhe effect of ra.re earbh fission products on. the solu-

bility of PuF. rema.in to be <’ie‘t'..srm:l.ne=.d° It does -eppear proba,ble, however,

3.

%hat e molten fluoride plubonium burner ha.ving unlimited burnup end

ex.hibit:l.ng substantial regeneration in -bhe blanket :I.s techn:l.ce.lly feasible,

- T5 -

 
 

 

 

‘of pentavalent uranium (UF., U2F9,', etc.) are not themmally stable

 

PART 2
- CHEMICAL ASPECTS OF MOLTEN-FLUORIDE-SALT REACTOR FUELS

1. CHOICE OF FUEL COMPOSITION

The search for a liquid for use at high temperatures and low ;preé-
sures in & fluld-fueled reactor led to the choice of elther fluorides
or chlorides because of the requirements of radistion stebility end solu~
bility of epprecieble guentities of uranium and thorium, . The chlorides
(based on the CJ.3 T isotope) are most suitable for fast rea.ct-o:r use, but
the low thermal-neutron ebsorption cross section of £luorine _mak_es ‘the
fluorides seem to be & wniquely desireble choice for & high-temperature
£1uld-fucled reactor in the thermal- or epithermal-newtron reglon,

1.1, Choice of Active Fluoride

Urenium Fluoride. Uranium hexafluoride is 8 highly volat:l.le com~
pound, and it is obviously unsuiteble as e componént of & liguid for use
at high temperatures. Thae compound UOLF, o whie.h ie relatively nonvolstile,
1s a strong oxidant that would be very difficult to contain. Fluorides
1 ena.
would be prohibitively strong oxldants even if they could be stebilized
in solution. Uranivm trifluoride, when pure and under an lnert amosphere ’
is steble even &t temperatures above 1000°C; 293 however, it is not so
steble in molten fluoride solutions.,h It disproportiona tes apyrecie.bly
in such media by the reaction, ‘ ' )

Y
uUF3‘———°-—3 UF, + U

 

1. 7. Kotz end E. Rebinowltch, The Chemistry of Ura.nimn ms-vm:f-s,

. MeGraw-H11l, 1951.

®roid. |
3¢, 3. Barton, W. C Wnltley, E. E. Ke’cchen, L. G. Overholser, and

" W. R. Grimes, Preparation and ProPerties of UF5, Oak Ridge National
_ La.boratory (vnpublished).

 

l*Se-e Reactor Hendbook, in press; material submitted by B. H. Clampitt 3

5. Lenger, end . F. lankenship, Ock Ridge Netionsl Leboretory.

S Th -

 
 

et b e ot

 

 

at temperatures below 800°c, Emall emounts of UF5 are p‘emiesible in
the presence of rela,tively le.rge concentramions of UFL and may be bene-
fic:l.el insofer as corrosion is concerned, It is necessary, hewever, to
use UFh es the major uraniferous compound :I.n the fuel.

Thorium Fluoride. All the normel eompounds of thori\nn axre quadri—
valent; accordingly, any use of thorium in molten fluoride melts must
be aS ThFl‘.I

1.2. choice of Fuel Diluents

The fluoride eomposit:l.ons that will be discussed here are 1:I.mited
to those viich have & low vapor pressure &t 'm?‘c and which have & melting
point no higher then 550 Cs Also, there ig ‘1ittle interest in u:anim .

concentrations higher then & few per cent for the fuel of thermal reaétors ’
‘end therefore mixtures with high UF, content will be omitted from this

discussien.

Of the pure fluor:!.&es ef mol’aen-;salt refic'bor interest, only ZBeF2
neets the melting point xequirement and 1t is too viscous for use in |
the pure state. Thus the fluorides of interest mre ternary or quaternary
mixtures containing UF), or TnF,. For the fuel,]_the relatively small
smounts of UF!; required meke the corresponding binary ‘or ternsxy. mixtures
of the diluents neerly controlling with regexd to physical properties .
such as the melting point. Only the elkelisfietal fluorides and the fluo- -
rides of bezyll:l.xm end zirconium have been given serious attention. ILead
end bismu'l;h fluorides, which might otherwise be ugeful because of their
low neutron a,bsorption, have been elimina,ted because they are reaflily

_reducea to the metallic eta:be by structural meta.ls such es iron a.nd

: Systems Gontaining UF)} Of 'bhe ternary aystems containing UFh

 

ftwo alkali-metal fluorides, o_nJ,v the LiFo-I{F-%UFu system, chown in Fig. 2.1,
gnd the LiF-RbF-:IJFh mrs'bem, ghown- in Fig. 2,2, have melting temperatures |
~ below 600 Ceat umnimn coneentratiene belew l@ mole % These two sys'cems

]e.nd the fourwcomponent systems LiFnNa.F-KFnUFh_, tor whic.h the alkali fluo-

ride ternsry diegrem is shown in Fig. 2.3, end LiF—-NaF—RbF-UFh are the

-75-

 
 

 

 

" UNCLASSIFIED
ORNL—LR - DWG 28633

UF, 1035
1000
TEMPERATURE IN °C. 950 ~LiF-4UF,
P= PERITECTIC
E=EUTECTIC
900
KF-2UF,— 850
P 765 800
4 70 ‘
70 ~-P 775
E 735~ ° <A -
650 A
TKF-6UF, ~ —7LiF BUF,
600 -
. ) —P &40
E740- 199 _ 55 Q
100 0

P 7674
2KF-UF, 5

   

 

LiF 845
E-4950

Fig. 2.1. The System KF-LiF-UF,.

76

€
 

 

 

 

 

 

UNCLASSIFIED
- ORNL-LR-DWG 28634
UF, 1035
1000
RbF - 6UF, _
TEMPERATURE IN °C
P=PERITECTIC 950 - LiF- “UF4
E=EUTECTIC
900
P-832
850
2RbF -3UF, _ 800
P-730 _ _
e 750 P775
RbF - UF, -
bF - UE, 650
7RbF- 6UF, - C 7LiF-6UF,
P-693 - 600
_ E-675
* ~P 610
P-818-
: 818 . 55
2RbF -UF, ~ 500
600
650 , E490
3RbF-UF, - 700 650 P 500
o 750
950 800 - 4LiF-UF,
' o ‘ 700
900 85_0
' 750
E-710-
: 800
750 . ' : - '-'-Ef;so AR )
ror LN\ VAV NN V) 1V N S Ny VNV v i
» . . -

 

 
 

 

 

UNCLASSIFIED
ORNL~LR-DWG. 1199

NaF

 

 

 

 

 

 

LiF

Fig. 2.3. The System LiF-NaF-KF. [A. G. Bergman and E. P. Dergunov, Compt. rend. acad. sei.
U.R.S.S, 3], 754 (1941).1

78

 
 

 

 

2

 

only a.va.ilable systems of IJZF‘,},r a.nd the alkali etal fluorides alone wh:l.ch
show low melting points et low uranium coneen'braficns.
Mixtures with melting points in the range of interest may be obtained

~ over relatively wide limite of concentration if Zth or :I!eJE'2 1s & con-
| ponent of the system. Phage relationships in the NaF-ZrF), end NaF-ZrFu

UF,’L systems are shown in Figs. 2.1 apd 2.5, The compounds ZrFk and UF,

| . have very similer unit cell ;pa.ra;mefi:ere:5 and are isemorphous. They form-
‘a continuous series of s0lid séolutions with 4 mininmm melting point -of
765°¢ for the solution conta.ining 23 mole % UF),.. Th:l.s minimm is respon-

gible for & broad shallow t-rough which penetrates the ternary dlagrem to
a.bout. the 1&5 mole % NaF s:l.tion. A continuous geries of solid solu-
tions without & meximm oy e minimm exigts hewe'@n a‘v}NaF' UF), &nd

‘5HaF ZrF) 3 in this aolution ‘geries the t@mperatuxe drops sharply with -

decreasing ZxF), concentratinn. 4 continuous solid-selutioq series without
8 meximum or & minimm a.lao exists between the isecmorpheus congruent com-
pounda TNaFs 6UF1L end TNaFs BZrFu; ‘the liguidus decreases with mcreasing
Zth content. These +two solid golutione sha.re & boundary curve over a

-considerable composition ra.nge ' The predomintnce- of - the yrimary phase

f:!.elds of . the three solid: salu‘bions presumably accounts for; the camplete
e.bsence of.&a -hernary eutectic in thie complex: system. ‘The liquidus . sur-

- face over the erea below 8 mole % UFu and between 60 &nd 45 mole % NaF

is rela:b:l.ve]y flat, ALl fuel compositions w:l.thin '!:his region have &ce"
ceptable melting poin’c.s. Minar e.dmtages m pharsical and thermal PTO=
:perties aceme from choosing mdxtures vith mininmm Zth eontent in this -

ccmposit.ton range.

The lWest melting ternary system.s which aenta.in UFh in the concen-

tration renge of generdl; :l.nterest are those gontaining BeF, and LiF or

 

‘NaF. Since BeF, offers the best cross section of &ll the useful diluents,
g such fuela are likely m be “of highest interest :m themal reacter designs,

gJo Jo K&tfi m Er Rabinmtch, -'Ihe
McGraw—iHill, w95 |

 

sigizy of Urentw, NNES-VIII-5,

-79-

 
 

 

 

 

UNCLASSIFIED
ORNL-LR-DWG-22105

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N
b ,
4471 - 4ONE
Y1179 . JoNL
T,
2\ 74172 - 4ONE
\4 ;
yd ¥117. 40NZ
= ;
T
= Y4122 - 4ONG ——
\Illlllunlulll_.lll _ —— ‘
,//11 Y417 - 4oNg
\
O O O O O O O O
QO O O o O O o O
o » © ~ © 0 < 0

(Do) HNLVYHIAWIL

80

90 Z’r Fa

80

20 30 40 50 60 70
ZrF, (mole To) |

10

NaF

Fig. 2.4, The System NaF.ZrF .
 

 

 

UF,
1035

 
  
   
    
  

NaF-2UF,
. 4

A

A\

\
/)
7NaF-6UF, ‘q |
5r~1<:r-'-3ur4
- 2Naf-UF, ATR~6 ‘ 800
AN AN
3NaF-UF, '

i

  

4 #=

675
700

    

o B '3N0F'-,2_-_Zr_F4)f . 3NaF-4ZrF, -
c o PSR TNaF-82rFy . “NoF-2rF,

. Fig. 2.5 The System NaF.Z:F ,.UF,.

81

UNCLASSIFIED -
ORNL—LR—DWG 19886

ALL TEMPERATURES ARE IN °C

765

7
50 800 A

 
 

 

 

 

 

The binary sys't;t-zrri';l.;:i.li'--:BeF2 shows melting points below 500°C over

' the concentration range from 33 to 80 mole % BeFy. The Id.F-BeF system

diegram shown in Fig. 2.6 differs substantially from previously published

| diagrems. é 1 It ie characterized by e single eutectic between BeF, a.nd
- 2LiF. BeF, that freezes &t 356°¢C end conta.ins 52 mole % BeF,. The com-
~ pound 2LiF. BeF, melts incongruently to LiF and liquid at 460%¢; LiF- BeF,
is formed by the reaction of solid BeF2 and solid 2IiF-BeF, below 274%¢,
~ The diagrem of Fig. 2. 7 reveals thet melting tmperatures below 500 C

ce.n be obtained over wide composition ranges in the three-compenent

‘ system I.iF—BeFa-UFll_.

The diagrem of the Na.F-BeFE system (Fig. 2.8) 18 simile.r to’ tha.t |
of the LiF-BeF system. The J.ac.k of & lcw-melting eutectic in the Ne.F--
UFh bina.ry' system is res;ponsible for melting points below 500 C belng .

' e.vailable over e considerably ema.ller concentratien interval in the Na.F-

BeF, -UFh system (Fig. 2. 9) tha.n in its LiF-BeF "'UFu counterpart

 The four-com_ponent system LiF-Na.F-BeFa-UFh hes not- been completely
diagramed It 1s obvious, however, from examination of Fig. 2.10 thet
the terna.ry ‘solvent IJ.F-Ne.FuBeFa offers e wide wriety of 1ew-melting

_com;poeitione, it has been established that considerable quantities (up.

to ‘at least 10 mole %) of UFI;. ‘can be tolerated in meny of these solvent
compositions without elevation of the meltigg point to a.bcve 500 Ce

Systems Gontaining A diagrem of the I&Fs-BeF -ThF ternary
L

 

' system, which is based solely on thermal data,is shown in Fig. 2.11. -
 Reeent studies in the 50 to 100 miole % LiF concentration range have

demonstrated (Fig. 2.12) that the thermal dats ere qualitatively correct.
Breeder reector blenket or breeder reactor fuel solvent ccmpcsiticne in
which the masimum ThFh ccncentra.tion is restricted to ‘thet a.va.ila.ble in
sa.lts he.ving less thsn & 550 c liquidus may be cbosen from an area of

 

61). M. Roy, R. Roy, end E. F. Osborn, "Fluoride Model Systems: IV,

.The Systems LiF‘BEFa and RbFE"BeF " J. Am, Cexrsm., SQG., 2;;’ 300 (195l|')o

TA. V. Novogelove, Yu. P. Simenov, end E. I. Yarembash, "Thermal

g.nd X-Ray Anslysie of the Lithimn-Beryllimn Flucride System," J. Phys.

Qem. R, 26, 1244 (1952).
- - 82 -

 
 

 

£8

TEMPERATURE (°C)

900 —

UNCLASSIFIED

ORNL—LR—DWG 16426

 

 

700 |——

200 L
UF

600

 

 

 

F— LiF + LiQuiD ———

  

 

 

\ |

 

 

IF + LipBeF,

P

e

/

BeFo + LIQUI

 

+ LIQUID

LioBe F4\\ \/

 

Ll2 Be F4 +

BeF, (HIGH

QUARTZ)

 

 

 

 

LiBeF,

 

 

 

 

 

LiaBeF4
+
LiBeF,

u-fO
L
@
3

 

LiBeF3 + BeF, (HIGH QUARTZ)

 

|

 

 

 

LiBeF3 + BeFp (LOW QUARTZ)

 

 

 

30

40

50 6

BeF, (mole %)

0 7

Fig. 2.6. The System LiF.-Ber.

0

80 90

Be Fz

 

 
 

 

 

e s i e

 

MOUND LAB. NO,
56~-14-29 (REV)

UF,

   
       
   
  

ALL TEMPERATURES ARE IN °C

E EUTECTIC
P PERITECTIC LiF - 4UF4
UFa| = PRIMARY PHASE FIELD

\
P

/\
LN\

7LiF - 6UF, .
P \fi
/ N\ \
&
4LiF-UF, ¢\

 
 
 
 

 

 

 

LiF

Fig. 2.7, The System LiF-BeF,-UF .
 

G8

- 900

TEMPERATURE (°C)

 

UNCLASSIFIED
ORNL-LR-DWG 16425

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

BeF, {mole %)

Fig. 2.8. The System NaF-BeF

©NaF +LIQUID
. S o® = ORNL DATA
700 t— HQ = HIGH QUARTZ
| LQ = LOW QUARTZ
600 {
_ /—————9
500 |— \ /
" a..“NQ?_BeFé; + NaF B'_ NUBeF3 Ber(HQ) + LIQUI
e ] +Liuio |
400 \ /"é}/ - BeRHQ) + B Naser
- a -'NozBleFt}' + B';NoBeF:,, Tt < \ -
L — ; _ P B-NaBeFy + LIQUID
300 - N1 -HeF. 4+ NaF - - NaBeF
oW a'- N_a_zBe.F4 +NaF + BeF,(HQ) + 8- NaBeF3
- B-NoBeFz +y-NapBeFs 1— BeF,(LQ) + B - NaBeF.
| I | / : °
| - Na,BeF, + NaF - —
200 L J- e e
NaoF 10 .20 30 40 50 60 70 80 90

2.

Ber

 

 
 

 

 

7 NoF-6 UF,

£

5Naf-3UF,
5NoF-3UF4~_ #
2 NoF-UF4-"""
P/

NoF 2 UF,

3NoF-UF, \
3 NaF-UF,
&

      
 
 

N

)

'u‘. —e00
N\ § 550 a

 
 

Q

a

UF,

   
  
   
     
   
  

N

7 NoF-6 UF,

R

'MOUND LAB. NO.
56—11-30 (REV)

ALL TEMPERATURES ARE IN °C

£ = EUTECTIC
PERITECTIC
UF, | = PRIMARY PHASE FIELD

v
I

 

 

NgF

 

86 ‘
— Q
\‘ N % =N
\ \ ‘&:h\_ \ - —. - \v/ - ‘x\\ E
E / £ £ Bef,
2NaF-BeF, NaF-BeF,

Fig. 2.9. The System NaF-BeF,-UF ,.

 
 

 

 

 

UNCLASSIFIED
ORNL-LR-DWG 16424

 

 

 

 

 

 

 

BeF,
542
DOTTED LINES REPRESENT
INCOMPLETELY DEFINED
PHASE BOUNDARIES AND o
ALKEMADE LINES ALL TEMPERATURES ARE INC
THE SYMBOL TC REPRESENTS
A COMPOUND WHOSE EXACT
COMPOSITION HAS NOT
BEEN DETERMINED
dTc
S~
280 <~ ]\\ 370
/ 4 ™~
356 — —~
(LiF - BeF,) 75
274
400,
350
450
a50 A \
459 Ag_, e @ Q (5 NoF:LiF -3 BeF,)
2 LiF -BeF, (NaF TE 2 NoF-Bef, 320
460 IF-BeF,) 595
500 24 N 570
550 600
600 /N 650
700
- 650 750
800
700 J\ 850
~ 900
750 700 . 649 700 750 800 850 900 950 NoP

Fig. 2.10. The System LiF-NaF-BeF,.

87

 
 

 

 

 

PWG, 18669

Thifs
1080°C

 

 

 

950°
900°
&s50°
8o e

 

€55

550%4 N\ 7505 '

. A
°'°°’ e/
A%A"o fua- "6.“\%. \ 63.:;,}&\ / A
B SO\ A\

LiF = Y g0
845°C LizBQF* LiBe F3 (?) 543 °c
475 °C Jeo®c

Fig. 2.11. The System LiF-BeF,-ThF .

 

 

88

 
 

 

 

 

i,

%

7 LiF-6 ThF,
45
40
35
o8
£
& P~595
& E-560
3 LiF-ThF,
E-570
2 _
© 550
15
650 00
10 700

  

550

  

UNCLASSIFIED
ORNL~-LR-DWG 27910

LiF-2 ThF, -
T E = EUTECTIC
P = PERITECTIC
LIQUIDUS
TEMPERATURES
ARE IN °C

   

 

 
  

 

 

: \ 750 e
5 800 e :'S‘*:e-_
| \ | S\ [450
5 10 5 20 - 25 30 / 35 40 45
: : = o ZLiF-BeF; “SP-490

BeF, (mole %)

| Flg 2.1'27. The Sy;tem_-L_fF-fieF_z-ThF 4' in the Conc'entrctiofi Range :50 te 100 Mole'% LiF.

89

 
 

 

 

the ;pha.se diagram (Fig. 2, 12) in which the upper limite of ThFu concen-
tra.tion are obtained in the compositiom '
75 mole % LiF-16 mole % ThF) -9 mole % BeF,
69 5 mole % IiF-21 mole % ThFh-9 5 mole % BeF,
- 68 mole % IiF-22 mole % ThFu-lO mole % BeF,

| Systems Containing ThF), and UFu The LiF-BeF --UFlF end the I.iF-]BeF'2
'131;115')_L terne.ry systems are very: simila.r, the two eutectics in the I.iF-BeFE

 

- 'ThFh_ eystem are at tempera.tures and compositions virtuslly 1dentical with

o those shown by the UFh-bea.ring system. The very greet similarity of these

two ‘oerns.ry systems end preliminary examination of the LiF-BeFe—ThFh-UFu

- quatemary system suggests that fractional replacement of UF,* by ThFu

will have little effect on the freezing temperature over. tb.e composition

SN

ra.nge of interest as ree.ctor fuel.

Systems Containing PuF5. The behavior of - plutonim fluorides in

 

molten fluoride mixtures hs.s received considera.bly lese study. Plutonium
| .tetrai‘luoride will pro‘oe.bly Prove very soluble, ae ha.ve UFA end ThFh,
in suite.‘ole :E'luoride-salt dilucnts , but is likely to prove too strong

an oxident to be cmnpatible with presently aveilable structural slloys.
The trifluoride of plutonium dissolves to the extent of 0.25 to 0.45

mole % in IiF-BeF, mixtures containing 25 to 50 mole % BeF,. There is
reeson to believe that such concentretione ere in excees of those reguired
to fuel & high-temperatm'c plutonivm burner (see Part 4).

2, PURIFICATION OF FLUORITE MIXTURES
‘Since commerclal fluorides ere availeble thet heve & 'low concentra~

tion of the usual nuclear poisons, the purifice‘bion process is deeigned
to minimize corrosion end to ensure the removel of oxides, c::yfluorides ,

; ;end. suli‘ur, rather than to improve the neutron econctmy Ihe rluoridcs
~ eve purified by high-temperature trestment with emhydrous HF end E, geses,
R fa.nd are. subsequently stcred in sealed nickel ccntainere under an atmos-

o phere oi‘ helium,

 
 

 

 

 

 

 

 

2.1, Purificetion Equipment

A schematic diagr-am of ‘the purificetion and storage vessels used
for preparation of the ARE fuel is shown in Fig. 2,13. The reaction
vessel, in which the ehemica.l processing is a.ccm@liehed, end the re~
celver vessel, into which the purified mixture 1is ultimately transferred,
axe vertical cylindrical containers of A-nickel. The top of the reactor
vessel is plerced by e charging port which is capped well sbove the heated
zone by & Teflon-gasketed flange. The tops of both the recelver end the
reaction vessels are plerced by short risers which temminate in Swegelok
fittings, through which gas lines, thermowells, etc., can be introduced.
A trensfer line terminetes neer the bottom of the reactor vessel end near
the top of xhe recelver; entry of. this tube is feffected threugh copper-
gasketed. flanges on l-in, -dia tubes which pierce the toye of 'both vessels.
This ‘branefer line con’oeine & filter of micmme‘ball:l.c sin'tered nickel
and a sam_pler which collec"ae & specimen of liguid é.uring -bransfer. Through
one of the risers in 'bhe receiver & tube extends to the receiver bottom;
this tube ’ which is eea.led eu‘beide the vessel, serves ee the mee.ns for
tra.nsfer of ‘the purified mixbu:'e to other eq,uiment. '

 

This essemtbly is connected te & menifold through which He, HE’ HF,
or vacuum ce.n be supplied te either vessel. Ey & eembina.tion of large
tube furna.ces , resistance hea.ters ’ “end lagging, sections of the e;ppa.re.tue
cen be brought independent]y to controlled temperatures 4n excess of
800 C.

, 2.2. Purifieation Processing , , | ,

N The re.w materia.ls ’ :l.n batches of proper eempoeition, are ‘blended

: and cha.rged into the ree.etion vessel. The material :Le melted and heated
._'to 700°%¢ under en e‘hnosphere of a.nhyd.roue HF to remove H20 with e minimum
of hydrolysis. The HF :I.s repla.ced w:l:bh Ha for e period of 1 hr, du;ring
whic.h the temperature is ralsed to 800°C, to reduce U5+ and UG+ to U

(1n the ca.se of simulated fuel mixtures) , sulfur cempounde to §°, and

| extra.neous oxida.nts (F.e Ty “for example) to’ lower velence states. The
hyd:rogen, as well es all eubsequent reegent geses, 1s fed et & rate of

- 91 -

 
 

 

 

UNCLASSIFIED
ORNL-LR-DWG 20157

  
 
 

TEFLON GASKETED FLANGE

~ 4~in~DiA CHARGING PORT

e 3/0—in~NICKEL
TRANSFER LINE

 

 

 

FILTER

 

SAMPLER

 

 

 

 

 

 

 

 

 

 

\

REACTION VESSEL

 

 

 

 

REGEIVER VESSEL—

 

Fig. 2.13. Diagram of Purification and Storage System. |

T 92
 

 

about 3 liters/min to the reacbj,on vessel through the reeeiver.. af.na transfer
line, and, eccordingly, it "bubbles up through the molten charge, The
hordmgen 1s then replaced by a.nlmsmus EF, vhich serves, during & 2= to

3whr :period &t 800 C, to volatilize E8 end HCL &nd to convert oxides

end owfluorides of urzmium and ziraonium to tetrafluorides &t the expense

of dissolutian of considerable N1F2 into the melt through reaction of HF

with the’ conts.iner. A fined 2h- to 30ehr treatment et 800°C with H, suf-

fices to mduce this IGI:!.F2 &nd the contalned FeF to insoluble metals.,

At the conclusion of the purification treatmfsnt & pressure -of heliwm
above the sa.l't. in the reactor vessel ig used to force the melt through
the trensfexr line with its £ilter 8nd sampler :I.nto the receiver. The
metallic iron and nickel a.re left in the reector vessel or on the sintered
nickel filtér. The purified melt is permitted to freeze under an atmos-
phere of heliwm in the receiver vessel. | |

'3, PEYSICAL AND THERMAL PROPERTIES

The melting points, heat capaclties, snd equations for density end
viscosity of & range of molten nixtures of possible interest as reactor
fuels are presented in Teble 2.1, aand thermal~conductivity values are
gliven in Te.ble 2.2, '

The tenqperatures abave vhich the materiels &re cm;pletely in the .
liquid state were detemined in phase eq_uili.bz-d.m studies, The me‘thods
used included (l) thexmal. analysia, (a) differential-themal enalysis,

| (3) q_uench:!ng fram high-tean;perature eq,uili‘brimn states, (ll-) vieual obser-

va.tion of the ‘melting ;pmaess, end ( 5): ;pha.se se;pematien by f:l.ltration
et high tm;peratures. Mea.sureznants o:E‘ density were ma.de by weighing, :
with' a.n analytical ‘ba..‘!fi.nce, & pl'tmme'l: suspended :Ln the mol'ben nixture. -

- Enthal;pies y heats of fus:!.@n, and heat ca.zaacities were de-bemined :t‘rom
| 'measmts of heat liberated when sam.ples 4in eapsules of Ni or Inconel

were dm:pped fm ve.rious ten;pera‘hures :Ln'bo calorime-herss both ice c¢calo~
rimeters and 1a.rge copper-block calorimeters were used, Measurements

_95-7.

 
 

 

 

Table 2.1,

Melting Poin-bs » Hea.t Capaci'bias and Equa:bions for Density and

Visaosity of Typical Molten Fluorides |

 

Oomposition

o fmole %)

VLiFfBe.Fe -

(69-31)

IAPeBe¥peUR,
(67-30.5-2, 5)

‘L:!.F-BBFQ o
(50-50)

NaF~BaFp

(5743)

Ter-Ber,-Un,
NaR-ZrF),
(50-50)
W"Zfl’lr
(50-16-4)

- IAP-NaF-KF .

65125042

LiF-NaF-BeFp .

- (35-27-38)

ngs) |

. Melting
Point
(°c)

505

550 -

360

510

520 -

b5l

338

',2.15_u ‘hO | ‘0-65

2,27 37 0.52

7 3095 ' 93. - ‘ | Qo?é |

) 2,55 - 73 . :.  _0;h5i o

Liquid Density
- (gfee) ~ Heat m:pacity
P = A BT (°C (cal/gram) -

) B at T00%

| : Viscosity '( cefitipbsé)'
= A eB/T (°K) |

 

At 600°%¢
 x 107 | |

0118 62k 7.5
238 b0 o5T sk
6.(_)189.

61T o2ie

003116 5161; 128

250 43 o R 105

379 93 0.28 00700 k68 . 8%
oo ws e

222 M 059 0058 M 7.8

 
 

 

e A s 4 S 8t e+ 1 e et i

-y

Teble 2,2, Thermal Conductivity of Typleal Flusride Mixtures

 

. | | '.téaemal donduéfiivity
Composition o (Bbq/hr-'ftb"FE n
(mole %) | Soiid ~ Idqui

 

: MF"‘M"‘E (46.5‘-]_1.5-11-2) - ' 207 o ’ . 7 2.6

LiF-ReF-KF=UF), (44,5-10,9-43,5-1.1) 2.0 2.3

. NaF~ZrF) ~UF, (59-%-4) - : o5 L3
| NeF-ZrR=UF, (53.5-4006.5) - | L

NeP-BeF, (57-43) = | 2.4

| NeP-KP-UR, (46.5-26-27.5) ; 0.5

 

- 95 -

 
 

oi‘ the v‘iscosities of the molten sa.its were mede with the use of & capil-
. lary efflux apparatus and & modii‘ied Brookfield rota.ting-cylinder device ;
) agreement between the measurements ms.de by the two methods indicated that

 the mubers obteined were within £10% of the true values.

 

- Vol. 2, Engineering, 0-56’4-6 (195‘5)

 

 

Thermal conductivities of the molten mixtures were measured in an
‘apparatus similer to thet described by Lucks end Deem,® in which the
heating plate is moveble o that the thickness of the 1iquid specimen -
can be 'varied. ‘The uncertainty in these values is probably less then
:L-25% The varietion of the thexmal conductivity of & molten fluoride &
salt with temperature is relatively small. The conductivities of solid
fluoride mixtures were measured by use of & stea.dy-state technique in
' 'which heat was passed through & solid slab. |

| The vapor pressures of PuFy (vef. 9), UR, (ref. 10), and ThF), are
negligibly small at temperatm'es that are likely to be :practica.l for

| rea.ctor operations. Of the fluoride mixtures likely €0 be of interest
as diluents for high-tempe.rature reactor fuels, only ALF Fa BeFi (ref.
.’Ll), and Zth_ (refs. 12, 13, 14) have apprecia.ble VepoT pressures below
700 0.

 

80. F. Lucks end H., W. Deem, Apparatus for Measuring the Thermal
Conductivity of ILiquid at Elevated Temperatures; Thermal Conductivity of
Fused NaOH to 600°C, Preprint No. 563A3.'1., Amerioan Soe. of Mech, Eng

'Meeting, June 1956.

\ 9(}. T, Seaborg end J. J. Katz, The Actinide Elements, Netional Nucles.r
Energy Series IV-14A, McGrew-Eill Co., Inc., New York (1954). | .

lOW. R. Grimes, D, R, Cunco, and F. F. Blankenship, Reactor Handbook

llI.bid.

121{. A. Sense, M. J. Snyder, and R. Bo Filbert, JToy "The Vapor Pressure

of Zirconium Fluoride," ‘7. Physs Chem. , g, 995 (195k)-

| 13y, Fischer (Institut fir Anorganische Chemle, Technicshe Hochschule,
~ Hennover) personal communication. Date for equation teken from S. Lauter,

Dissertation, Institut fur Anorganische Chemie, Technische Hochschule 5
Hannover (191’1-8

S. Cantor, R. F. Newton, W. R. Grimes, end F. F.. Blankenship “Vs.por
Pressures &nd Derived Thermodynsmic Infoms.tion for the System RbF-ZrF '

ggs. hen., 62, (1958)
,\ .

 
 

 

.-l

Measurements of total pressure in equilibrium with NaF-Zth-UFh
melts between 800 and 1000°¢ with the use of en a.pparatus similer to that

- described by Rodebush and D:!.J:onl5 yielded the date shown in Table 2.3.

Sense et a._l.,ls. vho used & transport method to eveluate partial pressures
in the NeF~-ZrF) system, obtained slightly different values for the vepor

:'_pressures and showed that the vapor phase sbove these’ liquids is quite

complex, The vapor=pressure ve.lues obtained from both investigations

are less 'l:.han 2 mm Hg for the equimolarx Na.F-ZrF,_‘_ mixture at 700°C. Bowe
ever, since the vapor 1s nearly pure ZrFu, end since Zth does not melt
under low Pressures of its vapor, even this modest vapor pressure leads

to engineering difficultles; all lines, eq_ui:pment s 8nd connections exposed
to the vapor must be protected from sublimed ZrF), "snow."

Mea.surements made with the Rodebuah apparatus have shown ‘that the

~vapor pressure sbove liquids of gnalogous composition' decreases with

increasing size of the alkeli cation, All these systems show large
negative deviations from Racult's lew, vhich are a consequence of the 1
lerge, positive, excess, partislemolar entropies of solution of 'szlt'
This phenomenon has been interpreted qualitetively es an effect of sube

stituting nonbridging fluoride ions for fluoride bridges between zirconium

lons as the alkeli fluoride concentration is increesed in the mel'l;.l7
Vepor pressure dats obtaeined by the transport method for Na.F-BeF2

m:l.:d:,u:t:'esl8 are shown in Table 2. L, which indicates that the ve.por phases

are not pu:re BeF While pressures a.bove LiF-BeF2 st be eocpected to

 

15W. H. ‘Rodebush a.nd A. L. B:onn "lhe Vapor Pressures of Meta.ls,

A New Experimentel Method," Pays. Rev., 26, 851 (1925).

lsK. A, Sense, c. A Alexender, R. E. Bowman, ‘and.. R. B.. Filbert, JTe,

- “Va.por Pressure and Derived Informetion of the Sodium Fluoride~Zirconium
- Fluoride: System. Descrip“cion of B8 mthod for the mtemina'bion of Molecular
Conxplexes Present in the Vapor Phe,se," &ge. dhem., l 357 (1957) |

s, Gantor, Re F. Newton, W. R. Grimes, snd F. F. B]nnkenship, "Vapor

_ ':Preseures and - Derived Themodynsmic Infomatien for the Syetem RbF~ZrF ,"

mnen. y 96 (1958)

 

lBK. A Sense R W Stone, and R. B. Filbe:'b, Jr. ’ Ve.p_or Preseure

 

and Fouilibrium Studies of the Sodium Fluoride-Beryllim uoride Systen,
USAE% BTw 1156, s Me.y 2'7 1957

..97..

 
 

 

 

 

» -a'}' ‘

| Teble 2.3, Vepor P'z.?essvtu-es‘ of Fluoride._Mixtui'és. Containing ZrFl@ -

 

| . '-Cdn:position'
- (mole %)

- 100

S 00
57 b3
o w
50 46
53 b3

UR),

A

{

. Yepor Pressure

B

TR |

12.5h2

T340

7.3T

x10°

11,360
7.289

TS5
7,105

.. . Yepor Preseure
Congtantg® - ..

et 900°%¢
(m'_Hs);. —

0.9 .
67

BT

28

S

 

*For the equation log P (mm Hg) = A -(B/T), where T is in %k, |
- Table 2.4, Vapor Pressures of NeF-BeF, Mixtures®

" . T c o (

 

Composition

mole

NeF

n

50

- 66 -

60

5

-’

59
»
‘._-'7855-1025 o '9;592 1.1667 : 9.089 .1.1063

- 25

‘Vamor Pressure Constants™*

iTHmperature' S ,
. Interval - RaF N BeFy

 

o TEs - & 5
S x10t  xaot

®OTT 0,55 1,09

eoa.gaa 110,06 1.085

796096 '7‘ | 952 1OT

NaJ+BeF,

A

9.TT
9.79.
- 9.82

 

B
b

x 100

1.206
1.187

- 1,187 -

Vh@or Pressure -
“at 800°¢

(rm Hg)

1.69 j
0.9%
oL

0.09

0,02

 

emnpned f’mm data o'btained by Sense et et al. (ref, 19)
FUr the equation lgg P Gmm Hg) = A.-Cafm), where‘T 1s in °K.

 
 

 

 be higher than those shown for Na.F-BeF mixtures, the values of Table 2, l;

suggest tha.t the "snow" problem with BeF2 mixtures 1s much- less severe
) than with Zth melts. B - '

, Physieel property values indica.te ths.t the mnlten flueride salts
exre, in general, a.deq_us.te heat transfer media., It is apparent y- ‘however, ’
:I‘:‘rom va.por pressure measurements and from spectrophotometrie examination
of . sns.legous ¢hloride systems that- such melts have een:plex struetures
" and are far from idesl solutions. | T

b, RAm:ATIoN STABILITY

When figsion of en active constituent eeeurs in & melten fluoride
solution, both electromagxetic radia.tions s.nd particles. of very high
energy and intensity originate within the fluid.. Iocal overhes.ting as
e. consequence of rapid slowing down of fissien fragnents by the. :f'luid
is probably of little coneequence in a res.ctor where the liguid is ferced

to flow turbulently and where re.;pid end, intims.te mixing oecurs.' Moreover, B

the bonding in sueh liq,uids is essentia.l]y cempletely ionic. Sueh & selu-

tion, which has nelther covalent bonds to sever nor & la,ttice to disrupt,

should be quite resistant to dszns.ge 'by particulate or eleetromagnetic o
radisation. ’

- More than 100 exposures to resctor re.die.tien of ‘va.rieus fluoride
mixtures conta.ining UFI;. in cépsules of I.nconel have been cenducted' 'in' |
these tests the fluid was not deliberately agitated. | The power level
~of each test was i‘ixed by selecting the U235 content of the teet mixture

Thermal neutron flwces have ranged from l(?oll to" 1011" netrt:ons/ em °sec

and power levels have va.ried from 80 to 8000 w/mn + . The ca;psules he.ve, o

in general, been exposed ot 1500 F for 300 hr, althorugh several tests

‘have been conducted for 600 to 800 h:r. A list of the materials that have )
been studied is presented in Table 2. 5e Methods of e:caminetion of the =

fuels aftexr irradiation have included (1) freezing-»peint deteminations 5
(2) chemicel a.lzlslysis‘9 (3) examination with & shielded petrographic B

- 100 -
 

 

 

 

Teble 2.5. Molten Salte Which Have Been Studied -
-+ " In In=Plle Cepsule Tests -

 

e o . Composition
PR o | System - . (mole %) o

 

Na.F-I;(FaUF#‘ o h§,5-26-é7.5
NeF-BeFy-UF), 256015
',fithBeFé%gEQ ; o N h7é5i~21 :
I.Na_li‘-geli’a-.m?# S 50-&6-# |
ner-zer, -3, . s
NeP-ZrF, VR, o - 535-40-65
Na?-Z?Fh-UFu - | - '50-112_3-_-2 .

| NaF-ZrF)~UF

3 70-48-2

 

< - - L0

 
 

 

 

 

microscope, (4) essay by mass spectmgra.;phy, and (5) emmina.tion by 8 |
gamme~rey spectroscope, The condition of the contedner was checked w:i.t.h R
& sh:l.elded metallograph. e L

No chenges in the fuel, except for the expected ‘bumup of 11255 nave -
been observed as e conseguence of irradistion. -Gomsj,on of the Inconel
 capsules to & depth of less than & mils in 300 hr was founds such corro-
“sion 16 comparsble to that found in unirradiated control specimens (see |
Part 3)e In capsules vhich suffered accldentel excursions in temperature
- %o e.bove 2000° F, grain grmrhh of the Inconel ocourred and eorros:l.on to
& depth of 12 mils was found. Such increases in comsion were glmost
cerba.in:ly the result of the serlous overheating rether 'bhan & conseq_uence
_ of the radiation f:l.eld

_ Tegts have a.lso been made in which the fissiening fuel is pun:ped

. th:mugh e aystan in which & thermal gradient is maintained in the flu:l.d.

- These tests included the Aircraft Reactor Equer:!.ment (described in Part l)

~ &nd three types of forcedecirculation loop tests. A large loop, in which
the pump was outside the reactor sh:l.eld, was operated in & hofizbntal o,

beem hole of the LITR., A smaller loop was operated in & vertical posi- '

'oion in the LITR lattice with the pump just outside the la.’s'bice. A

third loop wes operated completely within a beasmehole of the M.‘L'R -The

_operating conditions for these three 1oo;ps exe glven in Teble 2.6. .

~

The corrosion that occurred in 'bhese loop tests, which were of short
duration and whieh provided reletively mll temperature mdients ’
to a depth of less than b4 mils, end, &5 in the capsule tests, was come
parable to thet found in similar tests outside the radiation field (see
Pert 3). Therefore, it is concludsd thet, within the cbvious limitetions
of the e:qperience up to the present time, there is no effect of, ra.diation
on the fuel and no acceleration of corrosion by the radietion f:l,_eld.
 

Table 2.6, Descr:lptions of Inconel Foreedpc:i.rculation Ioops
: oPemtedmtheLIERandthem - '

 

Besis_mation
Horizontal . Vert:l.ca.l ... Horlizontal

 

NaF-ZrF), ~UF), mosition »

mole § S 62.5e12.5-25  63.25-12  53.5-M0-6.5
Maxtamum fission pover, w/cnn . koo - . 500 | 800
Totel pover, kv 2.8 . 0 20

.- ; Dilution factor* | ) - | 180 o T3 5
Maximm fuel tem;perature, °r ) 1500 - 1600 1500
Fuel ten;perature differential, - - . | .

oF ST e e 255

'Fu.el Beymlds mmber . 6000 . - o0 15000
Opereting time, e o | 645 32 kT
Time et full power, hr '_ s " - 235 2N |

% L P Y

. Bged e

e

Ratio of volume of fuel in system to volume of fusl in reactor core.

-105-

 
 

o ultima.tely assume are determined by the necessity for ca.tion-anion equi-

 

 

 

 

5. EEHAVIOR GF,FISSION PROIJUCI'S T NG

When fission of an active metal occurs in a melten eclution of ite
‘fluoride » the fission fragnents n:ust origine.te in energy states and ioni~-
zation levels very far from those nomally encountered. These fra.mnts ’ T
however, quickly lose energy through collisions in the melt and come to
equilibrim es common, chemical entities. - The valence states which they

, va.lenee in the melt and the reg,uirefizent the.t redox equilibrium be esteab-
lished smong componente ef the melt a.nd constituents ef the metallic eon—_
tainer. ' ' '

Structural metels euch as Inconel in contact with a molten fluoride .
‘solution ere not stable to Ey, UF 5» OF UF6. Tt is clear, therefore, that =
when fission of uranium as UFI; takes pla.ce ) the ultimate equilibritm mst
be such that four cation equivalents are fumished to satisfy the fluo-
ride ions released. Thermochemical data, from vhich the stebility of
- fission.-product fluorides in complex dilute solution could be predicted, S ]
are la.cking in many ceses. No precise definition .ef +the ;ve.lence stete‘:_ : . 3
~ of all fission-product fluorides can be given; it is, accordingly, not o
certain whether the fission process results in oxidation of the’ centaifier
- metal as & consequence of depoeiting the more noble fission products in |
_ the metallic state.

 

5.1. Fission Products of Well-Defined Valence

The Noble Gases. The fission products krypten end xenon can exist
only as elements. The solubilities of the noble gases in NaF-ZrF) (53
47 mole %), aF—Zth-UFh (so-hé-h mole %),19 and LiF-NaF-KF (46e5-11.5- -
L2 mole %) obey Henry's lew, incresse with increasing temper&ture ’ decrease
with increa.sing atomic weight of the solute » end very a;pprecia.bly with
composition of the solute., The Henry's lew constants and the heats o:t‘ -
solution for the noble gases in the Na.F-Zth end LiF-NaF-KF mixtures are
given in Ta.ble 2, 7. The solu‘bility of kry:pton in the NaF-ZrFu mixture
'_a.p;pears to be about 3 x 10"8 meles/em ctm.,

 

 

, 19W. Re Grimes, N. V. Smith, and G. M. Watson, "Solubility of Noble- e \w)
Gases in Molten Fluorides, I. In Mixtures of NaF-Zth (53-4T7 mole %) AR .
- &nd RaF-ZrF) ~UF), (50-746—1& mole %)," J. Fhys. Chem. in press).

-.-10&- S
 

 

 

| Teble 2.7. So_lubmties et 600°¢ end Hests of Solution for Noble =
. Ga.aes :Ln mlten muoride Mixbures ' L

 

--In NeF-ZrR),
(53=KT mole %)

Gas _;':A g

© In AP-NaRkF
_ (46, 52150 mole $)
Heat of Heat of

" Soluticn - . golution

 

L x 1078

Helium 21.6 £1

Neon 11.5‘ + 0.5

Argon' ° 5,1 £ 0,15
Xenon 1,94 % 0.2

(keal/mole) | K* |

X 108

(keal/oole)

6.2 - 11;5 1;6.7 o :3.0 |
7.8 . o B 2 0.2 8.9
"_-':48."2- T
ll;l

 

Henry s law eenstant in moles of gas per cuble centimeter of solvent

“per a‘lnnos;phere.

- 105 -

 
 

 

 

 

Eng., 2, 841 (1957).

 

 

~The positive heat of solutlion ensures that blanketing or sparging

of the fuel with helium or argon-in a law—tempera‘bure region of the _ L,
reactor cannot lead to difficulty due %o decreased solubility and bubble | _
formation in higher temperature regions of the. system. w&na.ll-_-soa.le in.- o .

- plle tests have revealed that, as these solubility date suggest, Xenon
&t low concentration is retained in 'a.*sta.gnant melt but is readily removed
by sparging with helium. Only & very mall fra.otion of the qnticipated |

Xenon polsoning was observed during operation of the a.ircraft Reactor

Experiment, even though the system contained no speclal apparatus for
Xenon removal. 20 It seems certain that krypton and xenon isoto;pes of

rea.sonable half-life can be readily removed from all practioal molten.. -
salt reactors, : S :

Elements of Groups I-A, II~A, III-B, and IV-B. The fission products
Rb, Cs, Sr, Ba, Zr, Y, and the lanthanides form very stable. fluorides' |
they should, accordingly, exist in the molten fluoride fuel in their =
ordinary valence states. High concentrations of ZrF) end the alkali a.nd |
alkaline esrth fluorides can be dissolved in LiF=-NaF-KF, LiEmBeFa, or . _—
NaP-ZxF) mixtures at 600°C. The solubilities &t 600°C of YF, end of |
selected rare earth fluorides in Na.F-Zth (5347 mole %) and LiF-BeF,
(65-35 mole %) are shown in Teble 2,8, For these materials the solubility
increases sbout 0.5%/°C end increases slightly with increasing atomic

number in the lanthanide series; the satura.ting phese is the simple tri-

fluoride. For solutions containing more than one rare earth the primary
phase is a solid solution of the rare ea.rth trifluorides; the re&tio of
rare earth cations in the molten solution is virtua.lly identical with

- -the ratio in the precipitated solid solution, Quite high ‘ournups would

be required before a molten fluoride reactor could saturate its fuel with

eny of these fission products.

 

20, s, Bettis, W. B. Cottrell, E. R. Maun, J. L. Meem, and G. D. -
- Whitmen, "The Aircraft Reactor Eqperiment - Opera.tion " Nuc. Sci. and .

- 106 -

 
 

 

 

 Teble é.fiB.

Salu'b:!.lity of YF3 a.nd of Some Rare Earth Fluorides

 Tn FaF-zrR end in I.iF-BeFa et 600°

- L

Solubility (mole % MF

C

)

 

“In ReFeixR,
(S'MB mole %) |

346
2.1

2.5

In I.iF-BeF
(62-58 mole %)

R

 

- 107 -

 
 

 

 

 

[ '5 Fission Products of Uncertain Valence

The valenoe states asamned by the nonmetallio elements Se ’ Te ’ Br,

w '\ end T must depend stmngly on the oxidstion potential defined by the con-

ta.iner end the fluoride melt, end the states are not at present well de-

T : fined. The sparse thermochemical data suggest thet if they were in the.
s ';pure state the fluorides of Ge, As, Nb, Mo, Ru, Bh, P4, Ag, Cd, Sn, end

w Sb would be reduced to the corresponding metal by the chromium in Inconel.

v

| While. fluorides of some of these elements ma:y be sta‘bilizeé. in dilute
N ‘molten solution in the melt, it is possrbie 'bha.t none of this group exists
o es compounds in ‘the equilibrium mixture. An epprecisble ;. end ;proba.bly

1arge , fraction of the niobium and ruthenim _produced in the Aircraft
Rea.ctor m;periment was deposited in or on the Inconel walls of the fluid
circuit, e detectable » but probably smell, fraction of 'bhe mthenium was

volatilized, presumably as Rqu, from the melt. '

'5 3. Oxidizing Na.ture of the Fission Prooess

The fission of UFu would yield more eguivalents of ca.tion than of

- anion if the noble gas isotopes of half-life greater than 1o min were

- lost and if all other elements formed fluorides of their lowest'teported
- valence sta.te. If this were the case the system would, ;presmably, re=

ta.in cation-anion equiva.lence by reduction of fluorides of the most noble
" ?-fission producto to metal and perheps by reduction of some U,'H' to .

Only about 3.2 ca:bion eq_uive.lents result, however, if all the elements
of uncertain valence stato listed above deposit as ‘metals; since‘_ the

enlons released total at least 4, the deficiency must be alleviated by

oxidation of the container, The evidence fram the Alreraft Reactor - -

R '_ -Experiment, ‘the in-pile loops, and the in-pile capsules has not shm
the fission process to cause serious oxidation of the conta.iner, it is
" possible that these experiments burned too little uranium to yield sig-
| " nificant results. 1 fission of UF, is shown to be oxidizing, the detri-
o mental effect could be’ overoome by deliberate end ocoasional addition -
' -  of & reducing s.gent to create a sms.ll end steble concentration of soluble
- UF, in the fuel mixt.ure.. |

./‘ .

- 108 -

CETad
.....
 

PART 3
CONSTRUCTION MATERTALS FOR MOLTEN-SALT REACTORS
1. SURVEY OF SUITABLE MATERIALS

A molten-salt reactor system requires structural materiels which will

~ resist corrosion by the salts, and evaluation tests of verious materials
in fluoride selt mixtures have indicated that nickel;-b_a.se ‘elloys are, in |
lgeneral s Superior Vt_b other commercial alloys for the 'cpnte.imnent of these

salts under dynamic flow _conditiens. In order to select the ‘alloy best
suited to this application, en extensive progrem of corrosion tests was

| carried out on the gvelleble, commercilel, nickel-base alloys, particulerly,

Inconel, which ty;pifies the chmim—containing alloys, end Hestelloy B,
which is mpresentative of the molybdenm~conta1ning alloys. :

Alloys eontaining epprecieble quantities of chromiuvm are atta.cked
by molten selts ma.inly by the removel of chromium from hot-leg sections
through reaction with UFL;’ if present, and with cther axidizing m;purities
in the salt. The remova.l of chromiwm is accompanied by the formation of

. subsurface volds in the metal. The depth of vaid formation depends strongly
‘on the operating tan;peratures of the sys'tmn end on the cemposition of the

selt mixture.
On the other hand Hastelloy B, in wh:l.ch 'bhe chrmiwn is replaced with

o ‘molybdenum, shcws excellent con:patibility with fluoride salts even a.t tem-
| '_pera:bures in excess of 1.600°F. X tmfortunately, He.stelloy B cannot be used
| *'a.s & structural material 111 hig}z-»tempera.ture systems 'because of :Lts age-

: hardening characteristics and its poo:c oxidation resistance. |

The :Lnfomation gained :m the testing of Hastelloy B m Inconel led

'to 'bhe fieveloment of en- a.lloy, desi.@a‘hed INOR-B which combines the
o better properties of bo’c.h aJ.'Loys for moltennsalt rea.ctor construction. |
The e.ppimate com;positions of the -bh:nee a.lloys ’ Incenel - Hastelloy B,

end INOR-S ‘axe given in Table 3.1.,

“1.09—

 
 

 

 

 

Teble 3.1.

.o,
.

4

Compositions of Potential Structural Materials

 

Quentity in Alloy (wt %)

 

 

Molybdenum

Mengenese

| Carbon. -

‘ Silicon .

Sulfur
Copper

Cobalt

Nickel

15-18

D) 0.8 ()
'0,15 (msx) 0.04-0.08 |

COmponents Inconel - - INOR-O Hestelloy B
Chromium | a7 6-8 1 (mex)
Iron 610 5 (max) h-T

1'.0! (max)

005 (oex)

6,5 (mex) 0.35 (max) '. 1.0 (max)
0.0L (mex) 0.0L (max) 0.03 (max)
0.5 (mex)  0.35 (mex)
0.2 (max) ) 2.5 (max)

| 72 (min)  Belance Belance

 

« 110 -

 
 

 

 

 

 

INOR-8 has excellent corroeion resistance to molten fluoride ca.lts

| et temperaturee concid.era.bly ebove those expected in molten-selt reactor
service, further, no ‘measureble atta.ck has been observed thus. fa.r in |
tests at reactor operating temperatures of 1200 to 1300 F. The mechenical
. properties of INOR-8 &t opere.ting temperatures are superior to those of
many eta.inlese steels and are virtually unaffected by long=time exposure s
to salts, The moteriel is structura.:l.ly stable in the operating tempere-
ture range, and the oxidation rate is less then 2 mils in 100,000 kr.

. No d.ifficulty is encountered in febricating standard shapes when the |
.. commercial practices eeteblished for nickel-beee elloys are used., Tubing, |
o plo.tes ’ ba.rs, forgings, and castings of INOR-8 have been made suocesefully

by several major metel manufacturing companies ’ and some of these compenies'
are prepered to supply it on & commercial ‘basis. :Welding procedures ha.ve "

'7"'”,,been establ:l.shed, end & good history of reliebility of welds elciets.
: The materisel has been found to be ee.eily weldable with rod of the seme

- composit:lon,

Inconel is, of eourse ’ en a.'l.terna.te cho:l.ce for the primery-»c:lrcuit
| structural material end muech information ie eveilsble on 1ts compati'bil:lty
i with molten salts end eod:lum. Althou.gh probably edequate ’ Inconel does
" not bave the degree of flexibility that INOR-8 has in corroeion resistence
to different sa.lt syeteme, and :lte lower strength at reactor operating

- tenmera.turee wouléd require hee.v:l.er structural coc:ponents,

 

A coneiderable nucleer a.dve.ntage would exist in & reactor with ean

{ ‘uncaxmed graphite moderator expoeed to the molten ee.lts. I.ong-time

a_fffixPOSure of graphite to e molten selt reeults :Ln the ealt penetratinglf' -
- ’_fthe ava:l.leble pores ’ but :!.t is probe.ble that 'y w:i.th the "1mpermeable

: _e_"f}:_':typee of gra.phite now being developed, the d.egree of ea.lt penetretion -
‘,f{??_-encomtered cen be tolereted Ihe atta.ek of the grephite ‘oy the salt

~* and the carburization of the metal container Beem to be negligible if
' "f"":—the temperature 13 kept below 1500 F. More teets a.re needed to finelly

 

i esta.blieh the compati‘bility of gmphtte-salt-e.lloy systems- S - .

 
 

 

 

 

Finally y & survey bes been made of materials suitable for bear:mgs

}» i' and valve seats in molten sa.l'bs. cemets, ceramics and refractory me’cals |
| a.p'_pear %0 be promising for this a.pplicat:l.on a.nd e.re presently being 1n-
'vestigated. B N

2, CORROSION GF;:P!IGCEI—BAS#E ALLOYS BY MOLTEN SALTS

Aflaaratus Used for Oorrosien Tests , - :
Nickel-base a.lloys have 'been e:r;poseé. o rlowing mlten salts :I.n both
themal-convection loops and in 1oops containing pm:ps for forced circu-

.la.tion oi’ 'bhe salts, Ihe t‘nerma.l-convection 1oops are. é.es:l@ea es shown .

in Fig. 5.1, When the bottem end &n adjacent side of the loop a.re heated,
usua.lly mth clem5hell heaters 3y eonvection farces in the centained fluid

'este.blish flow rates of up to 8 f£i/min, depending. on the temperature aif-
-ference ‘between the heated a.nd unhes.ted portions - of the leop. The rorced-

circulation loops are designed as shown in Fig. 3, 2._ ‘Heat is a.pplied to
the hot leg ef this type of loop by ‘Adirect resista.nce heat:l.ng af the
tubing. La.rge t&mperature differenees (up to 300 F) are obtained by air

| eool:l.ng of the cold leg. . Reynold.s numbérs of up te 10 ,OOO are a.ttaina.ble

with 1/2-'.'.n. ~-ID tubing, anfl somefiha.t higher values ca.n ’oe ohta.ined with
smaller tubing. :

. ,‘2 2.- Mechanism of Oorrosion

Most of the date on ccrrosion ha.ve been eb'aained with Inconél, a.nd
the theory of the corrosive mecha.nism was worked out for this a.lley
The comsion of INQR-8 eecurs to & lesser degree 'but fo].‘!.ws a8 pattem

| similar to tha.t observed fo:r Inconel and presmnably the sme theory

applies. - | . .
The fomation of subsurfa.ce voids ia initiated 'by the oxidation of

. chromium along exposed surfaces through oxida.tion-reduction rea.ctions -
- with 1mpuri'b1es of constituents of the mlten flucride salt mixture.

As the surface is depletéd in chromim, chromium fm the interior dif- o
fuses dorwn the caneentration gra.dient towarfl. the surfa.ee. ~ Since diffug_:ion

 
 

 

 

SIX 6-in. CLAM-
SHELL HEATERS |

 

&

i

fimjm&flmmm
ey ey s

 

 

METALLOGRAPHIC
SAMPLES

 

 

HOT—LEG
5 CHEMISTRY SAMPLE

 

 

UNCLASSIFIED
ORNL-LR-DWG 23451

COLD—LEG
CHEMISTRY SAMPLE

METALLOGRAPHIC
SAMPLES

 

 

 

 

O

| Fig’f—23‘.'1.-gl)'i'a'g|\-am-§l?§:Stafid&i'a'-l'he_rmd_l%C_onvecfilon.Loop Showing Locations at Which
Metallographic Sections are Taken After. Operation,

113

 
 

 

 

 

 

pLL

 

 

     

 

 

 

 

 

]~ FReeze vawve
HEATER CONNECTING LUGS ]

FiLL TANK

Fig. 3.2, Diagram of Forced=Circulation Loop for Corrosion Testing.

\ UNCLASSIFIED
\ ‘ ORNL-~1LR-DWG 7414

 

 
 

 

 

i

 

.

oceurs by & vacancy process and, in this pe,rticular situetion is essen=-
tially monodirectiona.l, it is possible to build up an excess number of
vecancies in the metel. These precipitete in arees of dieregistzy, prin- .
cipally at grein botnxda.ries and impurities, o form voids. These voids
tend to a.gglomera.te end grov. in size with increa.sing time end tempera.ture
Exeminations have demonstrated that the subsurface voids are not inter-
connected with each other or with the surface, Voids of the sene type
have been found in Inconel after high-temperature oxide.tien teets end |

- bigh-temperature vacuum tests in which chromium was selectively removed.

The selective removael of chromium by & fluoride salt mixture depends

on various chemical reactions, for example:

1, mPurities in the melt,

FeF, + Cr == CrF, + Fe (1)
2. Oxide £ilms on the metal surfa.ce ’
2Fe 0, + 3CrFy, v 3010, + ll-FeF5 - (2)
3, Constituents of the fuel,
or 4+ 20F), v=* 2UF; + CrF, | (3)

The ferric fluoride formed by the reactlon of Eq. 2 dissolves in the melt -

and further attacks the chromium by the reaction of Eq. 1.

 The time dependence of void formation in Inconel, as observed both

-i:l.n thezmal—convection &nd forced.-circulation eyetems, indiceates that the
. etteck 1s 1nitia.lly quite ra.;pid but tho.t 11; then decreases until e stre.ight-
"line relationehip exiete between depth of void. forme.tion end time,. This

- effect can be explained in terme of the corrosion rea.ctions discussed

‘ - iabove._ The initie.l rapid attack found for both typee of- loops eteme from

. the rea.ction ot chromiwn w:l.th im.pur:l.ties in the melt (resctions 1 and 2)

- _-_and with the UF,+ constituent oi’ the salt (reaction 3) to esteblish =& quaei-

| equili’brimn amount of CrF :!.n the ee.lt.e At this point ette.ck proceeds o

linee.rly m.th time and oceurs 'by e mess trenefer mecha.nism which, elthough
jit ariees from e. different ca.use, is simile.r to the phenomenon of tempera.ture-

gradient mass transfer observed in liquid metel corrosion.

..115..

 
 

 

 

 

 

 In molten-fluoride salt systems, the driving force f'or mass transfer
is a result of & tempersture dependence of the equilibrium constent for
~ the ree.etion ’oetween chromium and UF), (Eq. 3). If nickel and iron are
| considered inert diluents for chromium in Inconel, ‘the process can be
' eimly described Under ra.pid circulation s B unifom concentration of _
UIE’IL s UF,, and CrF2 is maintained throughout the fluid, the concentrations
- must satisfy the equili'brimn constent: ‘

 

CxF UF. CrE. UF _
-KaéK'KNg 22 2 * .2 — : (1*')
: 7 sy N ’N2 L
Tor UR), cr’” UR,

where N represents the mole fraction and y the e.ctivity coefficient of
~ the indice.ted component.

Under these steady-state conditions, there exists a tempere.ture )

T, intermediate between the maximum and minimum temperature of the 1oop s
at which the initial composition of the structura.l metal is at eq_uilibrium
with the fused salt. Since KN increases with increasing tempere.ture,

the chramium concentration in the alloy surface is diminished at- tempera-
tures higher then T and is eugmented at temperatures lower than T. In
some melts, NeF-LiF-KF-UF),, for example, the equilibrium constant of
reaction 3 changes sufficiently with temperature under extreme tempera-
ture conditions to cause precipitation of pure chromium crystals in the
cold zone. In other melts, for example, NaF-ZrFk—UFu , the temperature
dependence of the corrosion equilibrium is small, and the equilibrium is
satlisfied at all useful temperatures without the formation of crystalline
chromivm. In the latter systems the rate of chromim remova.l from the
salt streem at cold-leg regions is dependent on the rate at which chromium

cen diffuse into the cold«leg wall. If the chromium concentration gre.dient

, tends to be small, or if the bulk of the cold-leg surface is held at a
- relatively low temperature, the corrosion rate in such systems is almost

- negligible,

 
 

 

 

* ‘-“:"."h" ".-“ *

It is obvious that saddition of the eguilibrium concentrations of

UF3 end che to nolten fluorides prior to circulation in Inconel equip=

' ment would minimize the initisl removal of chramiwm from the alloy by

reaction (3). (It would not, of course, affect the mess transfer process
which arises es a conseq_uence of the temperature dependence of this re-~
action.) Dellberate edditions of these materials have not been practiced
in routine corrosion tests because (1) the effect at the uranium concen=-
trations normelly employed is'sn1all and (2) the experimental and analytical
difficulties &re considerable. Ad.dition of more than the equilibrium
gquantity of U’F3 mey lead to deposition ‘of some ursnium metel in the eq_uip-
ment walls through the reaction

hUFl'_t=’—* 3UF, + r° (%)

For ultimate use in reactor systems » however, it mey be possible to treat
the fuel material with calculated quantities of metallic chromimn to pro-

" vide the proper UF, and: (,trF2 concentrations at starbup.

3

o According to the theory described above, there should be no great
difference in the corrosion found in thermale~convection loops and in
forced-ci'rculation loops. The data are in general agreement with this-

conclusion as long as the same mexXlmm metal-salt interface temperature
is present in both types of loop.' The results of many tests with both
types of loop are summerized in Table 3.2 without distinguishing between
the two ty'pes of loop. The maximum bulk temperature of the salt as it
left the heated seetion cf the loop is given. It ie known that the actual
metal-salt interface temperature was not greater than 1300 F in the loops

;with a maximmn salt temperature of 1250 F and was between 1600 and 1650°F
- "’for the loop with & maximm aelt temperature of 1500 F.

The data in Table 3.2 are grou;ped. by ty:pes of base salt, because

.,‘ "the salt has a definite effect on the measured attack of Inconel ab 1500 F.
" w*"-'f_\The salte ’c.hat contain BeF2 are somewhat more corrosive then those con= -
| [.taining Zth, _and the presence of L:LF, except in combination with NeF,
| eeems to accelerate corroeion. | :

- 117 -

 
 

J._...,.,_.,w....._.._.._

 

. et

‘Table 3.2. Summary of Corrosion Data Obtained in Thermal-Convection
f and Forced-Circulation ILoop Tests of Inconel and INOR-8 =
Exposed to Various Circulating Salt Mixtures ' : ‘

 

Depth of Subsurface

 

: o o : Maximum Selt Time of - Void Formation at
Constituents of  UFy or ThFy Loop Tempersture Operation Hottest Part of
" Base Salts . Content Material °r) (hr) - (in.)
| NaF-ZrF),  lmole % UF,  Inconel 1250 1000 <0.001
: - 1 mole % UF), Inconel 1270 6300 0-0.0025
4 mole ¢ UF), Inconel 1250 1000 10.002
4 mole % UF), Inconel 1500 1000 0.007-0.010
- 4 mole % UF, INOR-8 1500 1000 0.002-0.00%
| C 0. Inconel 1500 1000 0.002-0.003
NaF-BeF, 1 mole % UF), Inconel 1250 1000 0.001
0 Inconel 1500 500 0.004-0.010
3 mole % UF), Inconel 1500 500 0.008-0.014
1 mole % UF), INOR-8 1250 6300 0,00075
I1F-BeF, 1 mole % UF) Inconel 1250 1000 0.001-0.002
3 mole % UF), Inconel 1500 500 0.012-0.020
1 mole % UF), INOR-8 1250 1000 0
NaF-IiF-BeF, Inconel 1125 1000 0.002
Inconel 1500 500 0.003-0.005
3 mole % UF), Inconel 1500 500 ¢.008-0.013
NaF-IiF-KF 0 Inconel 1125 1000 - 0.001
2.5 mole % UF,  Inconel 1500 500 0.017
0 INOR-8 1250 1340 -0
2.5 mole % UF,  INOR-8 1500 1000 0.001-0.003
I4F 29 mole % ThF,  Inconel 1250 1000 0-0.0015
NaF-BeF), 7 mole % ThF,  INOR-8 1250 1000 0

 

 

- 118 -
 

 

 

At the temperature of interest in molten-salt reactors, that is,
1250°F, the seme trend of relative corrosiveness of the different salte
may exist for Inconel, but the low rates of attack observed in tests pre-
clude & conclusive decision on this point. Similexrly, if there is any |
preferential effect of the base salts on INOR-8, the small amounts of
attack tend to hide it.

As expected from the theory, the corrosion depends sharply on the
UFI‘_ concentration. Studies of the nuclear properties of molten-salt
- power reactors have indicated (see Parb L) that the UF,+ content of the
fuel will ususlly be less then 1 mole %, end therefore the corrosiveness
of salts with higher UF), concentrations, such as those described in
Teble 3.2, will be avoided.

The extreme effect of temperature is also clearly indicated in
Table 3.2. 1In general, the corrosion retes are three to six times higher
at 1500°F then et 1250°F. Thie effect 1s further emphasized in the photo-
micrographs presented in Figs. 3.3 end 3.4, which offer & comparison of
metallographic specimens of Inconel that were ex;oosed to similar' salts
of the NaF-ZrF,_'_-UFu system at 1500 F end at 1250 F. A metsllographic
- specimen of Inconel thet was exposed et 1250 F to the salt proposed for
fueling of the molten-salt power reector is shown in Fig. 3.5

The effect of sodium on the stmctural materials of interest has
also been extensively studied, since sodium is. proposed for use &as the
intemediate heat transfer meditm. Gorrosion pro'blems inherent in the

:.utilization of sodium for hsat transfer purposes do not involve so much
the deterioration of the mets.l surfaces as the tendency for components
: -of the container materiaJ. to be transpozted from hot to cold regions and
| j‘f to :E'orm plugs of deposited material in the cold region. As in the case
‘of the. corrosion by the salt mixture, the mass transfer in. sodiume |
_conta.ining systems is extremely dependent on the ma.ximtm system operating
_temperature The results of mmerous tests indicate that the nickel-base

S 'alloys such es Inconel and INOR-B are se.tisfs.ctory containers for sodium

 

- 119 -

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fig. 3.3. Hot-Leg Section from an Inconel Thermal-Convection Loop Which Circulated

the Fuel Mixture NaF-ZrF,-UF, (50-46-4 mole %) for 1000 hr at 1500°F,

120

 
 

 

 

i
|
|

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fig. 3.4. Hfit-Leg Section of Ihconel Therml?Convecfion Loop. Which Circulated the
Fuel Mixture NaF-ZrF,-UF 4 (55.3-40.7-4 mole %) for 1000 hr at 1250°F,

 

121

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fig. 3.5. Hot-Leg Section of Inconel Thermal-Convection Loop Which Circulated the Fuel
Mixture LiF-Ber-UF4 (62-37-1 mole %) for 1000 hr at 1250°F,

 

122

 

 

”

 
 

 

e A O 21l AR 1 L PR

&t temperatures below 1500°F snd that above 1500°F the sustenitic stain-
less steels are prefersble.

5. FABRICATION OF INOR-‘»BI

ole Casting
Normal melting procedures , such &s induction or electric furnace
melting, ere suiteble for preparing INOR-8. Specielized techniques, such

" &as melting under vacuum or consumeble-electrode melting, have aleo been

used without difficulty. Since the major ealloying constituents do not
heve high vepor pressures end ave relatively inert, melting losses are

- negligible, and thus the specifled chemical composition can be obtained

through the use of standexrd melting techniques. Preliminam;r studies in-
dicate that intricately sheped components can be cast from this material.

5-2. Hot Fbrging
The temperature range of forgeebility of INOR-8 is 1800 to 2250°F.

This wide range permits opera.tions such &s hammer end press forging with
& minimum number of reheets between passes end substantial reductions

without cracking. The production of hollow shells for the manufecture
of tubing has been accomplished by extruding Porged end drilled billets
et 2150°F with glass as & lubricant. Successful extrusions have been
made on commercial presses at extrusion ratios of up to 1k:1, Forging
recoveries of u;p to 90% of. the :I.ngot weight ‘have been reported by one

"4_vendor. A

- 5_3 GOld. jbrming

In the fully annea.led condit:l.on ’ 'bhe ductility of 'bhe e.llcy ranges

'between ll-o a.nd 50% elangation fer & 2-:I.n. gage 1eng1;h Thus , cold-:foming
AL _fo;pera.tions, such 85 tube reducing, rolling, end wire drawing, can be-
e accomplished with noma.l production schedules. The effects of cold form-
 ing on the ultinate tensile strenghh, yield strength, end elongetion ere
'showninFig.56 Sl - -

—,123-

 
 

 

 

{x10%) B UNCLASSIFIED
250 _ ORNL-LR-DWG 24088

 

—
/
/’/ ,

ULTIMATE TENSILE sms~> /
///<o.z % OFFSET YIELD STRENGTH

NP2

 

200

€0

 

STRENGTH {psi)

 

 

PER CENT ELONGATION (2-in. goge length)

 

 

 

 

 

 

 

 

 

 

 

 

 

.- ELONGATION
50 20
l{ \
] 0
0 10 20 30 40 50 60 70 80 90

REDUCTION IN THICKNESS (%}

Fig. 3.6. Work Hardening Curves for INOR-8 Annealed 1 :hr at 2150°F Before Reduction.

124
 

 

- Forgeability studies bave shown that veriations in the carbon con-
tent have en effect on the cold forming of the alloy. Slight variations
of other components, in general; have no significant 'effe'-cts. The solid
solubility of carbon in the alloy is ebout 0.01%. Carbon present in
excess of this amount precipitetes as discrete particles of (Ni, Nb)sc
throughout the matrix; the particles dissolve sparingly even a:b the high
annealing temperature of 2150 F. Thus, cold working of the .a.lloy causes
these particles to align in the direction of elongation, and if they are
pre'een'b in sufficient quentity, they form continuous stringers of carbides,
The lines of weekness caused by the stringers are sufficient to propagate
longitudinal fractures in tubular :products during fabrication. The upper
limit of the carbon content for tubing is sbout 0.10%, and for other pro-
ducts it a.Ppea.rs to be grea.ter than 0,20%. The carbon content of the
elloy is comtrollsble to ebout O. 02% in the renge below 0.10%.

3,14,  Welding

The parts of the reactoi' system are joined by welding, and there-
fore the integrity of the system is in large measure dependent on the
reliebllity of the welds. During the welding of thick sections, the
material will be subjected to & high degree of restraint, and consequently
both the base metal and the weld metel must not be susceptible to cracking,

. embrittlement, or other undesirable features.

 Extensive tests of weld specimens he.ve been made. The circuler- -
sroove test ’ which ‘accwrately predicted the welde.bility of conventional

o materie.ls w:tth knom welding chare.cterietics » was found to give relieble
. results for nickel-'base alloys.. In the circule.r-gmave test, an inert-

) j?;gas-shielded tungsten-are weld pass is made by fusion welding (i e., the
;'weld metal ‘containg no. filler meta.l) in & -circular groove machined into
L "9. plete of -t;he base meta.l The presence or absence of cracks in the weld
metal :I.s then observed. Test samples of twe hea.ts of INOR-8 &lloys, to-

) 7_ gether with sam;ples of feur other e.lloys for comparison , 8T shown in
F:Lg. 3 7. . As may be seen 'bhe restraint of the weld metal caused complete

| 'circmnferentiel cracld.ng in INOR-B heet 8281\L whic.‘n contained 0.04% B,

-3_256-

 
 

 

 

 

Fig. 3.7. Circular-Groove Tests of Weld Metal Cracking.

126

 

 
 

vhereas there are no cracks in INOR-8 heat 30-38, which differed from
heet 8284 primarily in the ebsence of boron. Two other INOR-8 heats

that did not contain boron similarly did not crack when subjected to

the circular-groove test.

In order to further study the effect of boron in INOR~8 heats,
several 3-1b vacuum-melted ingots with nominsl boron contents of up to
0.10% were prepered, slotted, and welded as shown in Fig. 3.8, All
ingots with 0.02% or more boron cracked in this test.

A procedure specification for the welding of INOR-8 tubing is
evaileble thet is based on the results of these cracking tests and exami-
netions of numerous successful welds, The integrity of & joint, which
18 & measure of the quality 6ffi.a_;_"-'weld, is determined through visusal,
rediogrephic, and metallogrephic exsminations and mechanical tests at
room and service temperatures. It haes been esteblished through such
examinstions end tests that Sound joints cen be made in INOR-8 tubing
that contains less than 0.02% boron.

Weld test pla:bes of the type shown in Fig. 3.9 have also been used
for studying the mecha.nica.l properties of welded Joints. Such test
plates were side-bend tested in the epparatus illustrated in Fig. 3.10.
The results of the tests, presented in Table 3.3, indicate excellent
weld metal ductility. For exemple, the ductility of heat M-5 materlal
is greater then 40% at temperatures up to and including 1500 F.

. Brazing Ll ,

Welded and back-brazed tubewto-tube shee'b Joints are noma.lly used
_:Ln the :f'abrication of hea.t exehsmgers for malten se.lt service. The back
- 'brazing operation serves to ranove the notch inherent :Ln conventional

'---‘tube-.to—tube sheet ,join'bs ’ a:ad ‘bhe braze materie.l minimizes the possi-

 

 

 

- _"b:l.lity of 1ea.ka.ge through a weld fa.ilure that might be created by thermal
'stresses in service., T

- 127 -

 
 

 

 

 

Fig. 3.8. Weld in Slot of Vacuum=-Melted Ingot,

128

 

 
 

 

 

UNCLASSIFIED
ORNL-LR-DWG 22383R

 

 

v e —— e —— — — m—— e ———

- 6¢l -

 

S ol o in.
/// S e | ' //// {

e ] —— | ]/\4 Sin- “ A
Flg 3,9. Weld Test Plate Design Showing Method of Obtaining Specimen.

 

 

 

 

 

 

 

 

 
 

 

 

 

UNCLASSIFIED
ORNL-LR-DWG 18462

HYDRAULIG CYLINDER

     
  
  

1

SPECIMEN
THERMOCOUPLE

BEND
QUARTZ ROD SPECIMEN

CONTROL
THERMOCOUPLE

HEATING ELEMENT
CONNECTIONS
{220v OR Ov)

QUARTZ MIRROR

  

1012345
SCALE IN INCHES

BEND FURNACE

Fig. 3.10. Apparatus for Bend Tests at High Temperatures.

130

 
 

Table 3.3, Results of Side-Bend Tests of As'-Welde'd INOR-8 and Inconei Samples

 

Test

Filler Metal

 

_ INOR-8 f(Hgat M=5)

INOR-8 (Heat SP-19)

Inconel

 

Tempera.turé Bénd Angle* Elongation®** in

Bend Angle | Elongation in

(deg)

Elongation in

 

- T¢T -

(°F)_
Room.. |
1300

1500

s | ko
0 | sk
s | M

 

>0 sk

>90 |, >40

(g | b ()

=;5v,..>90 |

590
5%
30
15

15

1/1L-»in. (%)
~h0

 

 

1/% in. (%)
Sho
CSuo

>ho

> 40

>ho

 

 

 

 

*

- Elongation recorded is that at outer fibre at time first crack appeared.

 

»

Bend angle recorded is that at which f:l.rst crack appeared..',, s

o

 

 
 

 

 

The nickel-base brazing alloys listed in Teble 3.4 have been shown
- to be satisfactoxy in contact with the salt mixture LiF-KFhNaF%UFh

tests conducted at 1500 F for 100 hr. Further two precious metal-ba.se
brazing alloys, 82% Au-18% Ni end 80% Au-ao% Cu, were unattacked in the
L:I.F-vKF-Na.F-UFll_ salt after 2000 hr at 1200°F. These two precious metel
alloys were also tested in the LiF=-BeF -UFh'mixture and again were not
attacked.

3.6, Nondestructive Testing
 An ultrasonic inspection techniqgue is eveilsble for the detection

of flaws in -pla.te ’ ‘piping, and tubing. The _water-imer_sed pulse-echo .
‘ultrasound equipment has been edapted to high-speed use. Eddy current,-
dye penetrent, and rediogrephic inspection methods are elso used as
 required. The inspected materials have included Inconel, austenitic

" stainless steel, INOR-8, and the Hastelloy and other nickel-molybdenum-
:I'baé';e alloys. ‘ | '
Methods ere being developed for the nondestructive testing of 'weld-
‘ments during initial construction and after replacement by remote means
in & high«intensity radiation field, such as that which will 'be"";‘presen‘b
if maintenance work is required after operation of a molten-salt . reactor.
Thé ultrasonic technique appears to be best suited to semi~automatic a.nd
remote operation, and it will probebly be the least affected by. radietion
of any of the applicable methods. Studies have indicated that the dif-
ficultié-s‘ encountered in the ultrasonic inspe'cticn of Inconel welds and
‘_!,';relds' of some of the austenitic stainless steels because the weld struce
| tures have high ultrasonic 'a.ttenuation are not present in the inspection
' of INOR-8 welds. The high ultrasonic attenuation is not present in INOR~8
welds because the base metal and the weld metal are of the same composi-
tion. The mechanical equipment designed for the remote welding operation
will be useful for the inspec’o:l.on operation. ' '

In the routine inspecticn of reactor-grade construction materials ’
a tube, pipe, plate, or rod is rejected if & void is detected that is
larger than 5% of the thickness of the part being inspected. In the

A 4

M
 

 

 

 

Table 3.4. Nickel-Base Brazing Alloys for Use in Heat
Exchanger Fabrication

 

Brazing Alloy Content (wt %)

 

 

 

Components ~ Kiioy 52 Alioy 91 . Alloy 93

Nickel 91.2 91.‘3' | 93.3

S1licon | 4.5 b5 3.5

Boron , 2.9 2 . 9 " ~-;l_-,9

Iron and l | " | o
Cexbon Balance Balance Balance

 

- 133 -

 
 

 

inspection of a weld, the integrity of the weld must be better than 95%
of that of the base metal. Typicsl rejection rates for Inconel and
INOR-8 eare given below:

Re;jectien Rate (%) o

 

 

Iten Inconel R INOR-B
Pubing 17 .20
Pipe 12 1
Plate

Rod 5 | 5
Welds i | i

The rejection rates for INOR-B are ex_pected to decline as more experience
1s geined in fabrication.

4, MECHANICAL AND THERMAL PROPERTIES OF INOR-8

L.1, Elasticity

A typicel stress-strain curve for INOR-8 &t 1200 F is shown in Fig.
3.1l. Data from similar curves obtained from tests at room temperature
up to 1400°F are summarized in Fig. 3.12 to show changes in tensile
strength, yield strength, and ductility as & function of temperature.
| The temperature dependence of the Young's modulus of this materlel is
illustra.ted in Fig. 3.13. -

4.2, Plasticity

A series of relaxation tests of INOR-8 at 1200 end 1300°F ha.ve
indiceted that creep will be an importent design consideration .for
reactors operating in this temperature range. -The rate at which
the stress must be relaxed in order to maintain a .eons’ca.fit elastic:
~ strain st 1300°F is shown in Fig. 3.1k, and similar data for 1200°F
ere presented in Fig. 3.15. The time lapse before the,ma.tei'ial |
becomes plastic is ebout 1 hr at 1300°F and sbout 10 hr at 1200°F,
The time péfiéd during which the materiel behaves elastically becomes
much longer at lower temperatures and below same fiemperatures , 88

- 13h -

e
 

 

 

 

 

Jooo

     

N
o

-'SfRESs ~Sy

N
O
3

   

    

ELONGATION. /0" INCHES Pen “UieH
Fig. 3.11. Stress=Strain Relationships for INOR-8 at 1200°F.

135

 
 

 

 

STRESS (psi x 1000)

10

100

90

80

70

60

50

40

30

20

10

UNCLASSIFIED

ORNL—-LR-DWG 28709

 

TENSILE
\

STRENGTH

 

.

 

 

 

 

% ELONGATION

 

 

 

 

 

T~

\YIELD STRENGTH

Mo

 

\

\

 

 

 

 

 

 

 

 

 

 

 

 

 

4

c

8

10

12

TEMPERATURE (°F x100)

149

te

18

110

{00

90

80

70

60

50

40

30

20

10

Fig. 3.12. Tensile Properties of INOR-8 as a Function of Temperature.

136

PER GENT STRAIN
 

LE1

'YOUNGS MODULUS

G N ~ UNCLASSIFIED
) e ORNL-LR-DWG 29238
" | | | | | | |

 

 

 

 

 

0 200 400 600 800 1000 1200 1400
A TEMPERATURE (°F)

Fig. 3.13. Youngs Modulus for INOR-8 as a Function of
Temperature. |

 

 
 

 

8tl

UNCLASSIFIED
ORNL-LR-DWG 28708

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e T
~~0.2 % CONSTANT e
N
24 -
N 1 |

= 20 N~

S \ ~_

o ~N0.1% CONSTANT €

< N\ NN

x 16 S

2 N

T 0.05 % CONSTANT ¢ N AN

— | | N

w 8 \\ \\

¥* S \\fi\
DISCONTINUED TEST S~ %
4 : | =3¢
0.1 0.2 05 10 2 ) 10 20 90 100
TIME (hr)
Fig. 3.14, Relaxation of INOR=-8 at 1300°F at Various Constant Strains.

 
 

 

6€1

C

(x103)
28

- o C

e | UNCLASSIFIED
L ORNL—LR—DWG 29237

 

24
20j':lf

16

STRESS (psi)

12

T T T T T T TTTTT] T T T TTT]

. 0.1% CONSTANT STRAIN | —

 

bunemimr 7328 b e e e S e e Ty S e b
. . .

_0.05% CONSTANT STRAIN

 

 

 

 
 

% DISCONTINUED TEST

 

L L LIl L L L L LLLL] L L L LLLLl

 

2 5 10 20 50 100 200 500 - 1000
| | TIME (hr) - -

'Fig. 3.15. Relaxation of INOR-8 at 1200°F at Various Constant Strains.

 
 

 

 

yet undetermined, the metal will continue to behave elasticelly indefi-
nitely.

It is possible to sumarize the creep dete by comparing the times
to 1,0% total strain as & function of stress in the date shown in Fig.
'3.16. The reproducibility of creep datae for this meterisl is indicated
by.the separate curves shown in Fig. 3.17. It may be seen that quite
good correlation between the creep curves is obtained at the lower stress
values. Some scatter in time to rupture occurs at 25,000 psi, & stress
vhich corresponds to the 0.2% offset yleld strength at this temperature.
‘Such scatter is to be expected et this high stress level,

The tensile strengths of several metals ere compared with the tensile
strength of INOR-8 at 1300°F in the following tabulation, and the creep
properties of the several alloys at 1.0% strain are compered in Fig. 3.18:

' Tensile Strength at

 

Material 1300°F (psi)
18-8 stainless steel 40,000
Cr-Mo steel (5% Cr) 20,000
Hastelloy B 70,000
Hastelloy C 100,000
Inconel | 60,000
INOR=8 65,000

The test results indicete that the elastic and plastic strengths of

INOR-8 &are near the top of the range of strengfih'prbperties of fihe several |

alloys commonly considered for high-temperature use. Since INOR-8 was
designed to avoid the defects inherent in these other metals, it is ap-
perent that the undesirable aspects have been eliminated without emy
serious loss in strength. |

h.3. Aging Characteristics

Numerous secondary phases that are cepeble of embrittling & nickel-
base alloy cen exist in the Ni-Mo=Cr-Fe-C system, but no brittle phase
exists if the alloy contains less than 20% Mo, 8% Cr, end 5% Fe. INOR-S,

- 140 =

-~

.
 

 

vl

UNCLASSIFIED
ORNL-LR-DWG 28707

  

105
1100°F
2
__10?
g | 1300°F
o
o
&
E | EXTRAPOLATED
10
1yr 10 yr
10~

' 10 100 1000 10,000 100,000
TIME (hr) TO 1% STRAIN

Fige 3.16. Creep Data for INOR-8,

 
 

 

 

¢yl

STRAIN (%)

10.0

5.0

2.0

1.0

0.5

0.2

0.1

UNCLASSIFIED
ORNL- LR-DWG ‘29236

 

 

  

 

  

 

— T T 1171 T 1T T T 1T -
— RESULTS FOR THREE SAMPLES 47 —
STRESSED AT 25,000 psi

B RESULTS FOR TWO SAMPLES o
B STRESSED AT 20,000 psi |

| L1 | | LIl |
10 50 100 200 500 1000 2000

TIME (hr) '

Fig. 3.47. - Creep—-Rupture Data for INOR-8.

 
 

 

 

 

STRESS (psi)

10°

102

: UNCLASSIFIED
ORNL-LR-DWG 29235

 

1T

____H

ASTELLOY g TESTED Iy . | —

T T T TIT T T T TTTTm T T TTTTTE

   

 

 

I

 

 

B -‘|7r-*'l_-|I|l|_"|*i | LU

 

 

10 -

20 50 100 200 500 4000 2000 5000 10,000
| "TIME- TO % STRAIN AT 4300°F (hr)

| Flg v'3.18.".-C6rfib0.ri_$6n' of the Creep Properties of Several Alloys.

143

 
 

 

which contains only 15 to 18% Mo, comsists principally of two phases: - u
‘ the nickel-rich s0lid solution a.nd & complex carbide with the approximate |

composition (N1, MO)SG. Studies of the effect of the carbides on creep . -
strength heve shown thet the highest stremgth exists when & continuous
~ network of carbides surrounds the grains, Tests have shown that carblde o,

‘precipitation does not cause significant embrittlement at temperatures
up to 11:~80°F. Agling for 500 hr at various temperatures, as shown in Fig.
3.19, impi'oves the tensile properties of the alloy. The teneile Pro= -
‘perties et room temperature, as shown in Teble 3.5, are virtually unaf-
‘fected by eging, |
- 4.4, Thermal Conductivity end Coefficient of Linear Thermal Expension
Velues of the thermal conductivity end coefficient of linear thema.l
expansion are given in Tables 3,6 and 3.7.

 

5. OXIDATION RESISTANCE

The oxidation resistance of nickel-molybdenum elloys depends on the

~ service temperature, the tempersture cycle, the molybdenum content, end

the chromium content. The oxidation rate of the binsry nickel-molybdenum

-~ alloy passes through & meximm for the alloy conteining 15% Mo, and ‘the

scele formed by the oxidation is N:I.Mooh end NiO. Upon thermal cycling

from above 1400°F to below 660° F, the NiMoO) undergoes & phase transforma-

tion which causes the protective scale on the oxidized metal to spall.

Subsequent temperature cycles then result in an accelersted oxidation

rete. Similerly, the oxidation rate of nickel-molybdenum alloys contain-

ing chromium passes through a maximm for alloys conteining between 2

and 6% Cr. Alloys containing more than 6% Cr are insensitive to thermal

cycling and the molybdenmum content because the oxide scale is predominantly -

 stable Cry0,.  An abrupt decrease, by a factor of about 40, in the oxida~

~ ‘tion re.te et 1800°F is observed when the chromium content is increased . .
 from 5.9 to 6.2%. )

- 14k - | ‘ &
 

 

(x1o3) : UNCLASSIFIED

- ORNL-LR-DWG 29234

 

 

 

 

 

 

 
  
 

 

 

 

 

0 '
l | ]
100 |— —
% 90
e
L 80
o
o
W 70
%
L, 60
(—n_.J eommmrm ANNEALED {1 hr AT 2100°F |
z SO ™ emem == AGED 500 hr AT TEST TEMPERATURE
~ 40 | —
| 0.2 % YIELD POINT
2 | 1 |
60
e e B I I
B
z 40 | ELONGATION =
= -—_‘—’
< 30 |— —
O
=
"_-'--.;1000;-;-._’f‘ 1100 1200 4300 1400

TEST TEM PERATURE (°F)

an 3, 19 Effec’r of Agmg on ngh Temperafure Tensile
Propemes of INOR 8 |

145

 
 

 

‘Table 3.5. Results of Roam-TemperaturéwEmbritfilemenx'Tests.df'IfiOR§8

 

 

 

 

 

- f | Ultififife‘Tenéile_ Yield Point at | Elonéfition
 Heet Treatment Strength (psi) |0.2% Offset (psi)] (%)
Annesled* 114,400 14,700 50
hnnealed and aged 500 hr . | .. .

at 1000°F 112,000 k2,500 53
pnnealed and eged 500 hr S T o
~ at 1100°F - 112, 600 - 4k,000 - 51
Pnnealed.and ased,BOO hr Lo ~f s ‘ L
at 1200°F | | 112,300 kX, 700 .51
pnnealed and a.ged 500 hr o o
et 1300°F 112,000 L%, 500 k9
Annealed and aged 500 hr c .-i
at 1400°F 112,400 13,900 50

 

 

 

 

* 0.0hsuin. sheet, ennegled 1 hr at 21009F and tested at & strain ,

rate of 0.05 in./min.

- 146 -

C
 

 

 

 

 

 

A A it i s

Table 3.6  Comparison of Thermal Conductivity Velves for INOR-8
~ &nd Inconel at Several Temperatures .
| | | | Therma;l. Conductivity /fitu/rt fisec-(oF/ft)7
Temperature (°F) : INOR-8 Inconel |
22 5456 9.kl
92 6.T7 9.92
. 572 o 11,16 10,40
| 752 - 12,10 10.89
933 | k.27 11.61
12 16.21. 12,10
1202 18,15

12,58

 

 

- 147 -

 
 

 

 

 

 

- Teble 3.7

Lot

'Coefficient of Linaar-Expansian of INOR-8 for Several

Temperature Ranges

 

70 - 400
70 - 600
70 - 800

70 « 1000

70 - 1200
70 - 1400

T0 - 1600

70 - 1800

- . | . R . | Goeff;gient of Linaar'Ex@ansiofi.
Tempereture Renge (%) . (in./in,-%F)

X 10'6 |

5.6
6,23

- 6,58
6.89
o3k
7,61 .
8,10
8.32

 

- 148 -

 
b*

 

 

 

The oxidation resistanée of INOR~8 is excél'lént » and continuous
operation et temperatures up to 1800°F is feesible. ~ Intermittent use
at temperatures &s high at 1900°F could be tolerated. - For temperatures
up to 1200° F, the oxidation rate is not measurable; it is essentia.lly
nonexistent after 1000 hr of exposure in static air, It is estimated
that oxidetion of 0.001 to 0.002 in. would occur in 100,000 hr. of opera-
tion at 1200°F. The effect of tempersture on the oxidstion rate of the
elloy is shovn in Table 3.8.

6. FAERTCATION OF A DUPLEX TUBING HEAT EXCHANGER

The compatibility of INOR-8 and sodium is adequate in the tempera~
ture renge presently contempla.ted for molten-salt reactor heat exchanger
opei'é.tion. At higher temperatures, mass transfer could become & problem,
end therefore the fabrication of duplex tubing has been investigated.

- - _Satisfa.-étory duplex tubing has been made that consists of Inconel clad
with type 316 stelnless steel, and components for & duplex heat exchenger

have been febricated, as shown in Fig. 3.20.

The febricetion of duplex tubing is eccomplished by coextrusion of

' billets of the two alloys. The high temperature and pressure used result
- in the formation of & metallurgical. bond between the two alloys. In
' Subsequent reduction steps the bonded cmosite behaves as one materiel,
‘The ra.tios of the alloys tha:b comprise the composite axe controllable to
| within 3% The unifomity and bond. mtegrity obta.ined in this process
‘ are :!.llustrated :I.n Fig. 3.2.1. '

'me problem of welding INORa-B--sta.inless steel duplex ubing is

- ,,be:l.ng studied. | Meriments have mdicated that proper selection of elloy
e _Z’ra:tios end weld design will a.ssure welds thet will be satisfa.ctory in
'h:l.gh-temperature serivce. o ' ' -

In order 0 detennine 1irrlw‘-.'.ht‘-z:c' in'berdiffusion ef 'bhe a.lloys wmlld
result in a continuous brittle ; - at the interface, » tedts were made
in the tanpera.ture renge or 1300 to 1800°F, As expected, & new phase

" ‘i-llbgfl‘

 
 

 

 

 

Teble 3.8

Oxidation Rete of INOR-8 &t Various Tempereitures*

 

 Test Temzjaera;fiure. o

(5F
2200
1600

Tt
100

0.00
0.25
0.8
0.52

2,70

. Welght Gain (mg/em)
iififiiifi&?"‘“‘%fif4666135

Cub:i.c or. 1ogarithmicA .

0,00

: 067**

Lswe

2,0%%
‘ 28.2**

| “Cubie. B
o Parabolic |

- .—Pa.ra.'bolic. a

. Linesr’

Shape of
Rate' Curv'e-

 

-

U .7 mg/cm = 0.001 in, of ofidation.

) - 150 =

% Exbra.polated :E'rom date. obta.ined after 170 hr et tempemture.

©

 
 

 

 

 

 

 

 

 

 

Fig. 3.20. Components of a Duplex Heat Exchanger Fabricated of Inconel Clad with Type
316 Stainless Steel.

 

 

151

 

 
 

 

 

 

 

l INCHES T

 

£
»
n
~

 

 

 

Fig. 3.21. Duplex Tubing Consisting of Inconel Over Type 316 Stainless Steel, Etchant:
glyceria regia.

152

»

 
 

 

 

T

4, i

 appeared at the interface between INOR-8 and the stainless steel which

increased in depth along the grain boundaries with increases in the tem-
pereture. The interface of & duplex sheet held at 1300°F for 500 hr is
shown in Flg. 3.22. Tests of this sheet shoved an ultimete tensile
strength of 94,400 psi, & 0.2% offset yleld strength of 36,800 psi, and
an elongation of 51%. Creep tests of the sheet showed that the diffusion
resulted in an incremse in the ereep resistance with no significa.nt loss
of duetility. - -

Thus , DO ma,jor difficulties would be expected in the construction -
of an BWOR-B--steinless steel heat exchanger. The construction experience
thus fer has involved on.ly the 20-tube heat exchanger shoim in Fig. 3.20.

7. -AVAILABILITY OF INOR-8

Two production heets of INOR-8 of 10,000 1b each and mmerous smaller
heats of up to 5000 lb have been melted and fabricated into verious shepes
by normal production methods Evaluation of these connnereial products .
has shown them toshave properties similar to those of the laboratory heats
prepared for material selection. Purchase orders are filled by the vendors

in one to six months, end the costs renge from $2.00 per pound in ingot

fomi 0 $10.00 per pound for cold-drawn welding wire., The costs of f,ubing,
pla.te s end bar produc-bs depend to & le.rge extent on the specifications |

of the finished produc‘be. o

1"""' e

8 COI»IPATIBILITY OF GRAPHITE WITH MOLTEN SAUIS
| AND NIQCEL—BASE ALLOYS '

If g:-aphite could be used as & moderater in direct contect with &

"_'-molten sa.l'b s 1t vould make poseible e molten-salt rea.ctor with & breed:l.ng
| :-f"ra.tio in excess of one (see Part Ly, Problems that might restrict the
‘usefu.lness of ‘this approach e.re possible rea.ctions of grephite a.nd the

., - 153h

 
 

 

 

o i P

UNSTRESSE

Interface

316 88

316 88

oos,
Interface

INOR-8

 

 

Fig. 3.22. Unstressed and Stressed Specimens of INOR-8 Clad with Type 314
Stainless Steel After 500 hr at 1300°F, Etchant: electrolytic H,SO, (2% solution).

154

 
 

 

 

fuel salt, penetration of the pores of the graphite by the fuel, and
carburization of the nickel«alloy container.

Many molten fluoride salts have been melted end hendled in graphite
crucibles, end in these short-temm uses, the grephite is inert to the
salt., Tests at temperatures up to lBQOoF with the ternary salt mixture
Ifii&r.F-ZrFll_-UF,L geve no indication of the decomposition of the fluoride and
no gas evolution eo long &s the graphite was free from a silicon impurity.

Ionger«time tests of graphite immersed in fluoride salts have shown
greater indications of penetration of the graphite by salts, end it must
be assumed thet the salt will eventually penetrate the availeble pores .

in the graphite. The "impermesble"” grades of graphite aveileble experi-

menta.lly show greater reduced penetration and & semple of high~density,
bonded, natural graphite (De Gussa.) ghowed very little penetration.
Although q_ua.utita.tive figures are not availeble, it is likely that the
extent of penetration of "1mpemee.ble" 'gre.phite gredes cen be tolerated.

Although these penetration tests showed no visible effects of attack
of the graphite by the salt, anslyses of the salt for carbon showed that
et 1500°F more then 1% carbon may be picked up in 100 hr. -The carbon
;picku;p appesrs to be sensitive to temperature, however, inasmuch as only
0. 025% carben was found in the salt efter a 1000=hr exposure at 1500 F.

| In some mstances cea.tings have heen found on the graphite after
exposure to the salt in Inconel containers » 8 1llustrated in Fig. 3.23.

A cross section through the eoating is shown in Fig. 3.2k, The coating

wa.s found to be nea.rly pure chrmnium the.t wa.e presmnably tra.nsferred from

. the Inconel conta.iner.

In the tests Tun thus fe.r, ne peeitive ind:l.cetion has been found

.-of ca.rburize.tion of the nickel-e.lloy containers expoeed to molten se.lts o
’--e.na a'aphite at the temperatures e.t present contemplated for power reacters s
( <1500 F) El.‘he carburize.tion effect seems to be q_uite temperature sen-

sitive, however, sinee tests e,t 1500 F showed cerburizetion of Eastelloy B
to & depth of 0. 005 :l.n. in 500 hr of exposure to N aF-Zth-UFh conte.ining

- 155 =

 
 

 

 

 

i
i
i
i

 

 

0.10 IN/DIV.
it

 

Fig. 3.23. CCN Graphite (a) Before .and (b) After Exposure for 1000 hr to NaF-ZrF .~UF,
(50-46-4 mole %) at 1300°F as an Insert in the Hot Leg of a Thermal-Convection Loop.  Nominal
bulk density of graphite specimen: 1.9 g/cm3, | |

156

 

 
 

 

; (b) after
in (a) are pores,

 

(a) Before exposure

The black areas

3.23.

in (b)

ig.

F

in
ilm on the surface

 

157

 

 

 

 

 

 

f

 

iC

-

 

ions of Samples Shown
ith salt,

illed w

 

Cross Sect

*

 

Note the thin metall

 

.24

3

exposure,

 
 

(b) the pores are f

Fig.

 

In

 

 

 

 
 

 

 

graphité. A test of Inconel and gra.ph:l.t'e in & theimal~convection loop
in which the meximm bulk temperature of the fluoride salt was 1500°
geve a maximum cerburization depth of 0.05 in. in 500 hr, In this case,

however, the temperature of the metal-salt interface where the ca.rburi- -

zation occurred was considerably higher then 1500°F, probably about -
1650°F. |

A mixture of sodium and graphite is known to be & good carburizing
agent and tests with it heve confirmed the large effect of temperature
on the carburizetion of both Inconel and INOR-B es shown in Teble 3.9.

Table 3.9. Effect of Temperature on Cerburization of
Inconel and INOR~8 in 100 hr

 

 

Temperature Depth of Cerburization
Alloy ) (°r) (in.)
Inconel 1500 _ 0.009
1200 | 0
INOR=8 1500 0.010

1200 | 0

 

Many additional tests are belng performed with é. variety of molten fluo-
ride salts to measure both penetration of the graphite and carburizetion
of INOR-8. The effects of carburizetion on the mechanical properties
wlll be determined.

-‘-158-:

o
 

 

 

9. MATERIALS FOR VALVE SEATS AND EEARING SURFACES

Neerly all metels, alioys, and.hard-facing materials tend to undergp
solid-phase bonding when held togethor under pressure in molten fluoride
selts et temperatures above 1000 F. -Such bonding tends to make the

startup of hyflrodynamio bearings difficult or impossible, and it reducea |

the chance of opening a valve that has been closed for eny length of
time. Screening tests in & search for nombonding materials thet will
stend wp under the moltenusalt enviromment have indicated that the most
promising:materiqls are TiC-Ni end WC~-Co types of cermete with nickel

or cobalt cofitents of less than 35 wt %, tungsten, and molybdenum. The
tests, in general, have been of less than 1000-hr duretion, so the useful
lives of these materials have not yet been determined., |

10. SUMMARY OF MATERIAL PROBLEMS

Although much experimentel work remains to be done before the con-
struction of & complete power reactor sysiem can begin, 1t is apparent
that considereble progress hes been achieved in solving the materiel
problems of the reactor core. A strong, steble, and corrosion-resistant
alloy with good welding and forming characteristics is evailable. Pro-
duction technigues have been developed, and the alloy has been produced
in commercial quantities by several alloy'vendors.r'Finally it‘appears
that, even at the peak operating temperazure, no serious effect on the
alloy oceurs when the molten salt it contains 1s in direct contact with

graphite.

-9 -

R

 
 

 

 

 

PART & -

Ik

'NUCIEAR ASPECTS OF MOI.TEN-;S'AL'I‘ REACTORS.

‘ The ability of certain molten salts to dissolve uranium and thorium salts -
in quantities of reactor interest made possible the consideration of fluid-fueled
reactors with thorium in the fuel, without the danger of nuclear accidents as a
result of the settling of a slurry. This additional degree of freedom has been
exploited in the study of mOlten-salt reactors. , ,

Mixtures of the fluorides of alkali metals and zirconium or beryllium, as
discussed in Part 2, possess the most desirable combination of low neutron
absorption, high solubility of uranium and thorium compounds, chemical inert-
ness at high temperatures, and thermal and radiation stability. The following
' comparison of the capture eioss sections of the alkali metals reveals that Ii7
containing 0.01% 116 has a. cross section at 0.0795 ev and 1150 F that is a
factor of & lower than that of sodium, which also has a low cross section:

 

 Element o - | Cross Section'(barns)_
’Ii7-(conta1ning 0.01% 116-) : o 0.073 |
sodiwm % - | | 0.290
Potassium o . : 113

Rubi dium o oo

Cesium . - 29

]

| The.capture eross section of beryllium 1is also'satisfactorily low at all
. neutron energies, and therefore ‘mixtures of IiF and BeF Y which have satisfac-
tory melting points, viscosities, and solubilities for UFh and ThFu, were
selected for investigation'in the reactor physics study.

| Mixtures of NaF, Zth, ‘and UFh were studied.previously, and such a fuel
was successfully used in the Aircraft Reactor Experiment (see Parts 1 and 2).
Tnconel was shown to be reasonably resistant to corrosion by this mixture at
1500 F, and there is reason to expect that Inconel equipment would have a life
of atlleast several years at 1200 F. As a fuel for a central-station power
Lipreactor, however, the NaE-Zth systemjhas several serious disadyantages. 'The

T

w 160 =

 
 

 

sodium capture cross section is less favorable than that of Ii?: More important,
recent datal indicate.that'the capture cross eection'of zirconium is qnite-high

in the epithermal'and.intermediate neutron energy ranges. In eomparison with

the IiE%BeFé syetem, the NaFerFu_systen'has inferior heat transfer characteristics.
Finally, the INOR alloys (see Part 3) show promise of being as resistant to the
beryllium salts as to the zirconium salts, and therefore there is no compelling :
reason for selecting the NaF-ZrFu system. I

Resctor calculations were performed by means of the Univa02 program Ocusol,

>

3

a modification of the Eyewash program,_h end the Oracle program Sorghum.. Ocusol
is a Bi-group, multiregion, spherically symmetric, age-diffusion code. The_grnup-'@"'
averaged cross sections for the various elements of interest that were used weré |
based on the latest availahlesdata.s Wheredata'were lacking, reasonable inter#:‘
polations_based on resonance theory were made. The estimated cross sections o
were made to agree with messured resonance integrals where available. Setnra-"

tion and Doppler broadening of the resonances in thorium as a function offoOné -

centration‘were estimated. Inelastic scattering in thorium and fluorine was

| taken into scecount erudely by adjusting the value of got ;P however, the Ocusol .

code does not provide for group skipping or anisotropy of scattering.
Sorghum is a 51-group, two-regionx zero-dimensional, burnout code. The
group~-diffusion equations were integrated over the core to remove the spatial

 

lM’acklin, R. L., Neutron Aetivation Cross Seotions with Sb-Be Néutrons »
Phys. Rev. 107, 504-8 (1957).

2U‘nivereal Antomatic Computer at NEW Ybrk University, Institute of
Mathematics. -

3Alexander, L. G., et al Operating Inetructions for the Uhivac Program
Ocusol-A, A.Mbdification of the Eyewash Program, 0RNL~CF-57-6 4 (1957)

l"Alex,e.ruier, J. H., e.nd. Given, N. D. “A Machine Multigroup Calcula,tion.-_ ’.

_The ‘Eyewash Program for Univac, ORNLF1925 (1955).

50ak Ridge Automatic Computer and Iogical‘Engine at Oak Ridge National

. 6Roberts, J. T. and Alexander, L. G., Cross Sections for the 0cusol~A
Program, ORNIPCF'57'6 5 (1957),,, , . _

“ 161 -

 
 

 

 

dependency. The spectrum was computed in terms of a space-averaged‘gronp flux .
- from group scattering and 1eakage parameters taken from an Ocusol calculation. B
A critical calculation reqnires about 1 min on the Oracle; changes in concen-
‘tration of 1k elements duringia.specified time can then be computed in about

1 sec. The major assumpticn involved is that the group scattering and leakaget
_'probabilities do not change appreciably with changes in core composition as

: burnup progresses. This asénmption has been verified to a satisfactory degree -
of approximation.; '” . | '

The molten salts may be used as homogeneous moderators or simply as fuel
carriers in heterogeneous reactors. Although, as discussed below, graphite-
moderated heterogeneous reactors have certain potential advantages, their
technical feasibility depends upon the compsatibility of fuel, ‘graphite, and
metal, which has not as yet ‘been established. For this reason, the homogeneous
reactors, although inferior in nuclear performance, have been given greatest
attention. ; - - o

A preliminary study _in‘dicated that, if the integrity of the core vessel
could be guaranteed, theunuclear economy of two-region resctors would probably
be superior to that of bare-and reflected one-region reactors. ‘The two-region
reactors were, accordingly, studied in detail. Although entrance and exit con-
ditions dictate othernthan a spherical shape, it was necessary, for the calcu-
'lations, to use = model comprising the following concentric spherical regions:
(1) the core, (2) an INOR-8 core vessel 1/3 in. thick, (3) a blanket approxi-
mately 2 £t thick, and (h) an INOR-8 reactor vessel 2/5 in. thick. The diameter’
of the core and the concentration of thorium in the coré were selected as inde-
pendent Variables. The primary dependent varisbles were the critical concentra-
tion of the fuel (U255 U?BB, or Pu 39), and the distribution of the neutron '
absorptions among the- various atomic species in the reactor. From these, the
critical mass, critical inventory, regeneration ratio, burnup rate, etc., can
be readily calculated, as described in the following section.

1. Homossflpbus REACTORSE’UEIED WITH 255
y/) . ,

While the isotope U233 would be a,superior fuel in molten fluoride salt :
reactors (see Section 2), it ig. unfortunately not available in quantity. Any
realistic appraisal of the immediate capabilities of these reactors must be
based on the use of: 0235 o

:?.182 -
 

 

The study of homogeneous reactors wes dirided into two phases: (1) the

* mapping of the nuclear characteristics of the initial (1.e., "clean") states

- ag a function of core diameter and'thorium concentration; and (2) the analysis
of the.subsequent performance of selected_initial states with various process-
-ing schenmes and,rates. The detailed results of these studies are givenfin_the
following paragraphs. Briefly, it was found that regeneration ratios- of up

to 0.65 can be obtained with moderate investment in U2557(1ess than 1000 kg)
and that;.if the fission products. are moved (Section 1.2) at a"rate such that
the equilibrium inventory is\equai to one year's production, the regeneration
ratio can be maintained above 0.5 for at least 20 years.

1.1 Initial States

A complete parametric study of molten fluoride sait reactors having"
diameters in the range of U to lO”ft and thorium concentrations in the fuel
ranging from O to 1 mole % ThFh was performed. In these resctors, the basic .
fuel salt (fuel salt No. 1) was & mixture of 31 mole % BeF,, and 69 mole % I4F,
which has a density of about 2.0 g/cm at 1150 F. The core vessel was composed
of INOR-8. Thé blanket fluld (blanket salt No. 1) was a mixture of 25 mole %

' ThF, and 75 mole % LiF, vhich has e density of about 4.3 g/em’ at 1150°F. In

order to shorten the calculations in this series, the reactor vessel was neg-

lected, since the resultent error was small. These reactors contained no fis-
sion products or nonfissionable isotopes of uranium other than U238

A summary of the results is presented in Table k.1, in which the neutron
balance is presented in terms of neutrons absorbed in a given element per neu-
tron ebsorbed in U255 (both by fission and the n-y reaction). The sum of the
. absorptions is therefore equal to n, the number of neutrons produced by'fission
7fper neutron absorbed in fuel. Further, the sum of the absorptions in 023 and

_thorium in the fuel and in thorium Ain the blanket salt glves directly the re-

"-:generation ratio.; The losses to other elements are penalties 1mposed on the

'~‘regeneration ratio by these poisons, i e., if the core vessel could be con-

'-'structed of some material with a negligible cross section, the regeneration

 

: 'ratio could be increased'by the amount listed for capture in the core vessel.
| The inventories 1n these reactors depend in part on the volume of the

~ fuel in the pipes, pumps, ‘and hest exchangers in the external portion of the
fuel circuit. The inventories Iisted in Table-h 1 are for systems having &

 
 

 

 

 

Table 4.1, Initial-State Nuclear Characteristics of Two-Region, Bomogeneous,

Molten—FluoridefSalt Reactors Fueled with U":"3 2

Fuel salt No. 1: 3L mole % BeF,, + 69 mole % LiF + UF), + ThF),
Blanket salt No. 1: 25 mole % ThF) + T5 mole % LiF

Total power: 600 Mw (heat)

External fuel volume: 339 £t°

 

Case mmber 1 2 3 y 5 6

Core diameter, £t 4 5 5 5 5 5

ThF, in fuel salt, mole % 0 0 0.25 0.5 0.75 1

025% in fuel salt, mole % 0.952 0,318  0.56L  0.721  0.845  0.938
%97 atom density* 33,8  11.3  20.1 25.6 30,0 33,5
Critical mass, kg of U>7 124 81.0 144 183 215 239

Criticel inventory, kg of U-2? 1380 501 891 1130 1330 1480

Neutron absorption ratios¥*

 

 

"% (gissions) 0.7025 0.7185 0,700k  0.6996  0.7015  0.7041
22 (a-7) 0.29T77 0.2815 0.2996  0.300%  0.2985  0.2959
Be-Li-F in fuel salt 0.0551 0,0871L 0.0657 0.060k  0.,0581  0,0568
Core vessel '~ 0.0560 0.0848 0.0577 0.0485  0.0436  0.0402
Li-F in blanket salt 0.0128 0.0138 0.0108 0.0098  0.0095  0.0090
Leakage ‘ 0.0229 0.0156 0.0147  0.,0043  0.0141  0.0140
1®® in fuel sart 0.0430 0.0426 0.0463  0.045L  0.0431  0.0kl2
Th in fuel salt 0.0832  0,1289  0.16l4  0.1873
Th in blenket salt 0.5448 0.5309 o0.4516 O.4211  0.4031L  0.3905
Neutron yield, 1 173 1.77 1.3 1.73 1.73 174
‘Medisn fission energy, ev 270 15.7 105 158 270 o5
Thermal fissions, % 0.052 6.2 0.87 0.22 0.87 0.040 -
a-y capture-to-fission retio, @ 0.42  0.39  0.43 0.43 0.43  0.4203
Regeneration ratio ' 0.59 0.57 0.58 0.60 0.6L 0.62

 

* Atoms ( x 10719)/end.
*% Neutrons sbsorbed per neutron sbsorbed in ez

-16&-'

4
 

 

Table 4.1 (continued)

 

 

 

Case number 7 8 9 10 1 12
Core diameter, £t 6 6 6 6 6 7
ThFy, in fuel salt, mole % 0 0.25 0.5 0.75 1 0.25
U°2% in fuel salt, mole % - 0.107  0.229 0.408  0.552  0.662  0.l1k
22 atom density* 3,80 8.3  14.5 19.6 23.5 %.05
Criticel mass, kg of U>? k7.0 101 179 oly3 291 79.6
Critical inventory, kg of U-2° 188 Lok TL6 972 1160 230
Neutron sbsorption ratios *» | |
U222 (£1ssions) CO.TTTL  0.7343  0.7082  0.7000  0.700%  0.T748
722 (n-7)  0.2229 0.2657 0.2918  0.3000  0.2996  0.2252
Be-Li-F in fuel salt © 0.1981 0.1082 0.0770  0.0669  0.0631  0,1880
Core vessel " 0.1353 0.0795 0.0542  0.0435 0.0388  0.0951
Ii-F in blanket salt 0,016+ 0.0116 0.009. 0,008L  0.007%  0.0123
Leakage 0.0137 0.0129 0.0122  0,0119 0.0116  0.0068
" in fuel selt 0.0245 0.0575 0.0477  0.0467  0.0452  0.0254
Th in fuel salt | 0.1%21 0.1841  0.21%2  0.2438  0.1761
Th in blanket salt 0.53L2 0.4318 0.3683 0.3378 0.3202 0.4098
Neutron yleld, n 192 182 1.75 1.73 1.73 1.91
Median fission energy, ev 0.18 5.6 38 100 120 0.16
Thermal fissions, % 35 13 3 0.56 0.48 33
n-y capture-to-fission ra:bio, a_ 0.28. . 0.36 . o.b O.h2 0.k2 0.29

::_:.'Regeneration ratio S 0. 56 - 0.61 . 0.60 - 0.60 - 0.6 0.6L

 

' ?3'i{:,.:,* _AtOms (x lO 19)/cm O R
**'Neutrons absorbed per neutron absorbed :l.n 023 5

 
 
   

165

 
 

 

 

Teble 4,1 (continued)

 

Case muber

Core diameter, f’b
ThF% in fuel salt, mole %

U"%in fuel salt, mole 4

U255 atom density*
Critical mass, kg of U235
Critical inventory, kg of U23 >

Neutron absorptibn ratios¥**
y=o2 (£issions)
v (n-7)

Be-Li-F in fuel salt

- Core vessel
Li-F in blanket salt
Leakage
PR in fuel salt
Th in fuel salt
Th in blanket salt

Neutron yleld, g

Medlan fission energy, ev
Thermel fissions, %

n-y capture-to-fission ratio, «

Regeneration ratio

 

* Atoms (x 1077)/cm’

 

13 1 15 16 17
8 8 8 8 8 .
0 0.25 0.5 0.75 1
0.0k7  0.078  0.132  0.226 0.349
1.66  2.77 k.67  8.05  12.4
48.7 81.3 137 236 36L

110 184 310 535 824
0.8007  0.7930 0.7671L 0.7362  0.T146
0.1995 0.2070 0.2329 0.2638  0.285k
0.4130 0.2616 0.1682 0.1107 0.0846
0,149  0.1032 0.0722 0.0500  0.0373
0.0143 0.0112 0.0089 0.0071  0.0057
0.,0084 0.,0082 0.0080 0.,0077 0.0074
0.0143 0.0196 0.0272 0.0368  0.0428 .

0.20L5 0.3048 0.3397  0.3515

0.4L073 0.3503 0,3056 0.,2664  0.2356
2.00 1.96 1.89 1.82 1.76
Thermal 0.10 0.17 5.3 o7

59 k5 29 13 5
0.25 0.26 0.30 0.36 0.40
0.h2 0.57 0.6k 0.64 0.63

¥* Neutrons absorbed per neutron absorbed in UE'5 5 .

-166-
 

 

 

 

Teble L.l (continued)

 

Case number

Core diameter, £t

in fuel sult, mole %
u25g in fuel salt, mole %
" atom denstty*
Critical mass, kg of U255
Critical inventory, kg of U-27

Neutron sbsorption ratios*s
U235 (fissions)
v">? (n-7)
Be-Li-F in fuel salt
Core vessel
Li-F in blanket salt
Leakage
Uzj’8 in fuel salt
Th in fuel salt
'Th in blanket salt
Neutron yield, q
Medisn fission energy, ev
Thermael fissions, %
n-y capture-to-fission ratio, @

Regenera:tion ratio

 

* Atoms (x 10 l9(/cm

18 19 20
10 10 10
0 0.25 0.5 -
. 0.033  0.052  0.081
1.175  1.86 2.88
67.3 107 165
111 176 272
0.8229 0.7428  0.7902
0.1T7L 0.2572 0.2098
0.5713 0.3726  0.2486
0.1291 0.0915 0.0669
0.011% 0.0089 0,0073
0.006L 0.0060 0.0059
0,0120 0.0153 0.0209
~ 0.2409  0.3601
0,303 0.26l7 0,2332
2.03 2.00 1.95
Thermel Thermal 0.100
66 56 W3
0.21 0.24 0.26
0.32 0.52 0.62

: .** Neutrons absorbed per neutron absor'bed .'m U255 '

167

21
10
0.75
0.127
k.50
258
k25

0.7693
0.2307
0.1735
0.0497
0.0060
0.0057
0.0266
0.4324

0,2063
1.90

0.156
30

0.30
0.67

22

10
1
0.205
7.28
k7
€87

0.7428
0.2572 -
0.1206
0.0363
'0.00L49
0.0055
0.0343
0.4506
0,1825
1.83

1.36
16

0.35
0.67

 
 

 

 

"volume of 339 ft5 externsl to the core, which corresponds approximatelyfto a
power level of 600 Mw of heet. In these calculations it was assumed that the
heat was transferred to an intermediate coolant composed of the fluorides of |
14, Be,'and Na before being'transferred_to'sodium metal. In more recent designs
-'(see'Psrt 1), this intermediate seltsloop has been replaced by a sodium loop
and the external volumes are-somewhst'less because of the-ifiproVed equipnent
designa.ndlayout. - - | o
 Critical Concentration, Mass, Inventory, and Regeneration Ratio., The |
data in Teble 4.1 are moyre easily comprehended in the form of graphs, such as
Pig. k.1, which presents the critical concentration in these reactors as a '
function of core diameter and thorium concentration in the fuel salt. The ,
_data points represent calculated values, and the 1ines are reasonsble inter-
polations. The maximum concentration calculated, about 35 x 10 19 atoms of ,
55 per cubic centimeter of fuel sslt, or about 1 mole % UFu, is an order of ,'
magnitude smaller than the maximum permissible concentration (sbout 10 mole. %).
The corresponding eritical masses are graphed in Fig. 4.2. As may be seen
the critical mass is.a rather complex function of the diameter and the thorium :
concentration. - The calculated points are shown here also, and the solid lines
represent, it is felt, reliable interpolations. The dashed lines were drawn
vwhere insufficienttnumbers of points were calculated to define the curves preQ
=cisely; however, they are'thought to be qualitatively correct. Since reactors
-having dismeters less than 6 ft are not economically attractive, only one case
with a b-ft-dia core was computed. ‘
 The critical masses obtained in this study ranged from 40 to 400 kg of
U255; However, the critical inventory in the entire fuel circuit is of more
interest to the reactor designer than is the criticel mass. The critical in- '
ventories corresponding to an external fuel volume of 339 ft5 are therefore
~ shown in Fig. h 3. Inventories for other external volumes may be oomputed
~ from the relation, | | |

6v
=M ‘(1~+ —Des) »
. n .

where D is the core diameter in feet, M 1is the“critical mass teken from /{f
Fig. L. 2 Vé is the volume ‘of the external system in cubic feet, and I 18

the inventory. in kilograms of U235 The inventories plotted in Fig. 4.3 range

_p_.-]s's'__
 

 

 

Fig. 4.1.
169

 

 
 

 

 

 

- 170

 
 

 

 

O © © O i o - 0 Ay -

171

 

 

 

 
 

 

 

from slightly above 100 kg in an 8- ft dia-core with no thorium present to
‘"1500 kg in a 5-ft~dia core with 1 mole % ThFh present.i

The optimum combination of’ core diameter and thorium concentration is, o fa

qualitatively, that-Which minimizes the sum of inventory - charges (including
charges on Ii7, ‘Be, and Th) and fuel reprocessing costs.' The fuel costs are .
directly related to the regeneration ratio, and this varies in a complex |
,_manner with” inventory of U255 and thorium concentration, as shovn in Meg. h k
It may be seen that, at a glven thorium concentration, the regeneration ratio '
(with one exception) passes through a maximum as the core dismeter is varied N
_between 5 and 10 ft. These maxima ‘increase with increasing thorium concen-
tration, but the inventory values at which they oceur also incérease. ,
T Plotting the maximum regeneration ratio versus critical inventory -generates .
the eurve shown in Fig. 4.5, It may be seen that a small investment "in U235
(200. kg) will give a regeneration ratio of 0, 58 that 400 kg will give a ratio ..°
of O. 66, and that further increases in fuel. inventory have little effect.
The effects of changes in the compositions of the fuel and blanket salts

_are indicated in the following description of the results of a series of cal-

' culations for which salts with more favorable melting points and viscosities

" were assumed. The fieF content vas raised to 37 mole % in the fuel salt (erl

. salt No. 2) and the blanket composition (blanket salt No..2) was fixed at 13
mole % THF),, 16 mole 4, BeF s end Tl mole % IiF. Blanket salt No. 2 is & some-

- what better reflector than No. 1, and fuel salt No. 2 8 somewhat better moderator.

A8 & result, at a. given core diameter and’ thorium concentration in the fuel salt,

fboth the critical concentration and the regeneration ratio are somewhat lower
for the No. 2 salts. o ’

| Reservations concerning the feasibility of constructing and guaranteeing
the 1ntegrity of core vessels in large sizes (10 ft.and over), together wvith
preliminary consideration of inventory charges for large systems, led to ‘the
conclusion that a feasible reactor would probably have a ccre dlameter lying
in the range betveen 6 and 8 ft. Accordingly, ‘8, parametric study-in thiS'range
with the No. 2 fuel and blanket salts was performed~ In this study the presence
| of an outer reactor vessel consisting of 2/5 in. of INOR-8 was teken into sccount.
| The results are presented in Teble h:aaand.Figs h'6 and 4.7.. In general)’ the
‘nuclear performance is somewhat‘better with the -No." 2 salt than with the No. 1
salt. = o
e .

AFa.
o

H
 

 

 

173

 
 

 

 

 

 

Fig. 4.5.
174 |

»
 

 

 

 

 

 

Table k4.2, Initial-State Nuclear Chara.cteristics of Two-Region, Homogeneous,

Fuel salt No. 2

Total power:

Extermnal fuel volume:

MoltenflFluoride—SaJ.t Reactors Fueled with U235

37 mole % BeF, + 63 mole % LiF + UF# + ThFh
Blanket salt No. 2: 13 mole % ThF) + 16 mole % BeF, + 7L mole % LiF

 

. '?*,'Regeneration ra'bio

Case number

Core diameter, It

in fuel salt, mole %
u25% in fuel salt, mole %
U255 atom density¥
Critical mass, kg of UE55
Critical inventory, kg of U255

Neutron absorption ratiog¥**
2 (fissions)
v (n-7)
Be-Li-F in fuel salt
Core vessel ‘
Li-F in blanket salt
Outer vessel
Leakage
P 10 fuel satt
Th in fuel salt
Th in blanket salt
Neutron ‘y‘ield, N '

 Median fission energy, ev S .
”'I'hemal fissions, b o
n-y capture~to-f:lssion ratio ; ‘d,fi;

 

 

 

0 59

* Atoms (x 10 9)/cm . 7" SRR LT : i
i.w** Neutrons absorbed per neutron absorbed 1;1 U255 L

0.8 0.8

175

600 Mw (heat)
339 £t
23 2k 25 26 27 28
6 6 6 6 7 7
0,25 0.5 0.75 1 0.25 0.5
0.169 0.310 0.423 0.580 0.084 0.155
5.87 10.91 15.95 20.49 3,13 5.38
T2.7 135 198 254 6L.5 106
201 540 790 1010 178 306
0.7516  0.71Th  O.70M+  0.6058 0.7888  0.7572
0.2u84  0.2826  0.2956  0.3042 0.2112  0.2428
0.1307  0.0900  0.0763  0.0692 0.2147  0.1397
- 0.1098 0.0726 0.0575 0.0473 0.1328 0.0905
©0.02lF  0.0159  0.0132  0.0117 0.0215  0.0167
0.002%  0.0021  0.0021  0.0019 0.0019  0.0018
0.0070  0.0065  0.0064  0.0061 0.0052  0.0050
0.0325  0.0426  0.0452  0.0477 0.021k  0.0307
0.1360  0.1902  0.2212  0.2387 0.1733  0.2565
0. 4165  0.3521 @ 0.3178  0.2962 0.3770  0.329k
- 1.86 '1 77 . 1. 7h 1.72 1.95 1.87
oo - 1ot S8 1 761 0223 0.5
."fé211Q‘ '*¥¥{7:‘;<7}‘ 2.8 0.8 ;;' h}f- 2l
S 0.33 "b'39"-:-*6ifié;; ok 037 0.3
058 ,o 57 0.62

 
 

 

 

Teble 4.2 (continued)

 

Case pnunber

Core diameter, ft

ThF) in fuel salt, mole %
U%3 4in fuel salt, mole %
U‘e35 atom density*

Critical mass, kg of U235
Critical inventory, kg of U255

Neutron absorption ratios#¥*
F? (fissions)
0"’ (n-7)
Be-Li~F In fuel salt
Core vessel
Li-F in blanket salt
Outer vessel
Leakage
®® in fuel salt
Th in fuel salt
Th in blanket salt
Neutron yield, q

. Median fission energy, ev
Thermal fissions, %

n¥7 capture-to-fission ratio, o
" Regeneration ratio

 

* Atoms (x 10'19)/cm5.

 

 

29 30 31 32
[ T 8 8
0.75 1 0.25 0.5
0.254 0.366 0.064 0.099
8.70 13.79 2.24 3451
171 271 65.7 103
Lol 783 149 233
0.7282 0.7094 0.8014 0.78LL
0.2718  0.2906 0.1986  0.2186
0.1010 0.0824 0.2769 0.1945
0.0644  0.0497 0.1308 0.0967
0.0131 0.0108 0.0198 0.0162
0.0016  0.0015 0.0017  0.0016
0.0048 0.0045 0.0045 0.0043
0.0392 0.044T 0.0177 0.0233
0.2880  0.3022 0.1978 0,30L43
0.2866  0.2566  0.3%240 0.2892
1.80 1.75 1.97 1.93
T.61L 25.65 51% thermal 0.136
11 L.3 51 28
0.37 0.4 0.25 0.28
0.61 0.60 0.54% 0.62

¥¥ Neutrons absorbed per neutron absorbed in U235.

176~

55
8
0.75
0.163
5.62
165
374

0.7536
0.2L64
0.1354
0.0606
0.0130
0.001h4
0.0042
0.0315
0.3501

0.2561
1.86

0.518

25
0.33
0.64

3l

8

1

0.254

9.09
267
60k

0.7288
0.2712
0.1016
0.0518
0.0105
0.0013
0.0040
0.0392
0.3637
0.2280
1.80

7.75
ll .

0.37

0.65
 

 

 

 
 

 

 

 

178

 
 

Neutron Balances andNMiscellaneous Details.:'The distributions of the
neutron captures are given in Tables 4.1 and 4.2, where the relative hard-
ness of the neutron spectrum is indicated by the median fission energies
and the percentages of thermal fissions. It may be seen that losses to Ii,
Be, and F in the fuel salt and to the core vessel are substantial, especially
in the more thermal reactors (e.g., Case No. 18). However, in the thermal

235

reactors, losses by radiative capture in U are relatively low. Increasing
the hardness decreases losses to salt and core vessel sharply (Case No. 5),
but increases the loss to the n-y reaction. It is these opposing trends
which account for the complicated relation between regeneration ratio and
critical inventory exhibited in Figs. 4.4 and 4.7. The numbers given for
capture in the Ii and F in the blanket show that these elements are well
shielded by the thorium in the blanket, and the leakage values show that
leakage from the reactor is less than 0.01 neutron per neutron absorbed in
U255'in reactors over 6 ft in diameter. The blanket contributes substantially
to the regeneration of fuel, accounting for not less than one-third of the
total even in the 10-ft-dia core containing 1 mole % ThF) «

Effect of Substitution of Sodium for 117 In the event that 117 should
prove not to be available in quantity, it would be possible to operate the

 

reactor with mixtures of sodium and beryllium fluorides as the basic fuel

salt. The penalty imposed by sodium in terms of critical inventory and regen-
eration ratio is shown in Fig. 4.8, where typical Na-Be systems are compared
with the correSponding Ii?Be systems. With no thorium in the core, the use

- of sodium increases the criticel inventory by a factor of 1.5 (to about 300 kg)
| eand 1owers the regeneration ratio by a factor of 2.~ The regeneration penalty

“fis less severe, percentagewise, with l mole % ThFh 1n the fuel salt, in an -

f41f8 ft-dia core, the inventory rises from 800 kg to. 1100 kg and the regeneration
:fiiflratio falls from O 62 to O 50._ There is scme doubt concerning the validity of
_-ffithe pcint representing the 10—ft dia core for the Na-Be system with 1.0 mole %
'3'5ThFh, the explanation for the apparently abberant behavicr may'be that sodium |

 

,;is relatively'more harmful in the 1arge, near-thermal systems. Details of
gthe neutron balances are given in Table h 3 ‘“f :J'”'jf.-'_ . ,

N Reactivity Coefficients. By means of a series of calculations in which
the thermal base, the core radius, and the density of the fuel salt are varied

 

independently, the components of the temperature coefficient of reactivity of

- 179 -

 
 

 

 

 

 

 

180

 

 

 
 

 

Table k.3, Initial Nuclear Characteristics of M-Regiog, Homogeneous,

Molten Sodium-Beryllium Fluoride Reactors Fueled with l.l23 >

Fuel salt: 53 mole % KaF + 46 mole % BeF, + 1 mole % ('th + UFh)
Blanket salt: 58 mole % NaF + 35 mole % BeF, + 7 mole % ThF),

 

 

 

 

‘Total power: 600 Mw (heat)
External fuel volume: 339 ft5
Case number 35 36 37 38 39 ko
Core diameter, ft 6 6 :) 8 10 10
in fuel salt, mole % 0 ‘1 0 1l 0 1
u'°-’3g in fuel salt, mole % 0.17%  0.70L%  0.091  0.465 0.070  0.282
U™ atom density* 6,17 249 3.28 165 247 1240
Critical mass, kg of U-0° 764 308 951 48h W2 10
Critical inventory, kg of U>2° 306 1230 215 1100 o3l 1170
Neutron sbsorption ratios** |
U255 (fissions) 0.T417 0.6986 0.7737 = O.70L1 0.7862 0.7081
1°? (n-y) 0.2583  0.30Lh  0.2265 0.2989  0.2138  0.2919
Na-Be-F in fuel salt 0.2731  0.1153  0.4755 0.1kl 0.6119  0.2306
Core vessel 0.1181  0.0476  0.1125  0.0392 0.0917  0.2306
Na-Be-F in blanket salt 0.0821  0.0431L  0.0660  0.0315 0.0495  0.2306
Leakage 0.0222  0.0182  0.0145  0.0116 0.0105  0.2306
U238 in fuel salt 0.0360  0.0477 0.0263  0.0484 0.0232 0.0467
Th in fuel salt 0.2418 0.3150 10,3670
 Th in blanket salt 0.3004  0.2120  0.2165  0.1450  0.1550  0.1048
Neutron yield, n 1.83 1.7% 1.91 1.73 1.9% 1.75
Median fission energy, ev 1.3 190 0.20 36 0.087
Thermal fissions, % 17 o2 34 1.h 4.1
n#‘capture&d-fission ratio, 0.25 0.’4—3 0.29 0.43 0.27 0.h41
' ’ 0.3% 0.5 0.2k 0.5. 0.8 0.52

_ Regenefation ratio ,

 

*  Atoms (x 10° lg)/cm

| ** Neutrons a‘bsor'bed per neutron absorbed in U235

181

 
 

8 reactor can be estimated as illustrated below for a core 8 ft in diameter
and & thorium concentration of 0.75 mole % in the fuel salt at 1150°F. From

the expre551on,

-k = f(T)pr)_)'

4where X is the multiplication constant, T is the mean temperature in the
core,. p 1is the mean density of the fuel salt in the core, and R is the core
radius, it follows that

ldk=1 @15) '+_1_<_a_5) d.R+l(_Lk) o, ()
kdf k |ao7T k | 9R aT 2p aT - -
| PyR \ p,T :

 

/R,T

'where'thetermJ%f(%%% ' represents the fractional change in k due
D:R

change in the thermal

irepresents the change due to expulsion of fuel from the core by thermal
expansion of the fluid, and the term -% gg
an increase in core volume and fuel holding

base for slowing down of neutrons, the tenn

represents the change due to

p,T capacity. The coefficient

%% may be related to the coefficient for linear expension, a, of INOR-8, viz:
dR =RC!.

~ arT

'Iikewise the term %T may be related to the coefficient of cubical expansion,

‘B, of the fuel salt:

7 = "PB .

 

From the nuclear calculations, the’components of the temperature coeffi-
cient were estimated, as follows: |

-(0.13 ¥ 0.02) x 10"5/°F

==
01“’
REILT,
O
s .
= s
i

. + 0.412 ¥ 0.0005

i
(V"
ol
v
-
3
)]

14 .

- 0.405 < 0.0005

W

=

ofu
S o
w

Y2

0

- 182 -

 

 
 

“tion of the resonances in thorium and U
- the effective widths of the resonances would be increased at higher temperatures,

 

The linear coefficient of expension,'a,'of INOR-8 wcs estimated to be

(8.0 t 0.5) x lofg/bF,T and the coefficient of cubical expansion, B, of

the fuel was estimated to be-(9.889 ¥ 0.005) x 19-3/bF from & correlation -

- of the density given by Powers.  Substitution of these values in Eq. 1 gives

’ld-k_ 5for,
.E.&__-(380-o.oh)x10/1=‘.

for the temperature coefficient of reactivity of the fuel. In this calcula-

tionm, the effect of changes with temperature in Doppler broadening and satura-

235

were not taken into account. Since~

the thorium would contribute a reactivity decrease and- the U235 an increase.

These effects are thought to be small, and they tend to cancel each other.
Additional coefficients of interest are those for U255 and thorium. For

the 8-ft-dia cores,

14 (0.17 N, (") x 10"15>

 

 

 

 

N(UEBS) J
k -3N(Uz§_5)N(Th) 24 Nc(U235) x 10712
and - | : | .' B |
- o - 5
N(Th) [ 2k \ = N(m) [k ax, (v*?)
k (om(m))L 235y ok IN(ED) sy ()
where | o | | |
(u2 %) . . ;, 08(',5 0. os9sn(m) %1072
"EETEET"' T

In tfi@se equations, N(0255) represents the atomic density of U255 in atoms,
per cubic centimeter, N, (0255) is. the critical density of 0235, and N(Th)
18 the density of thorium atoms.;,., -

 

7K’inyon, B. W., Private Communication, ORNL (1958)
8waers, W. D., Private. Ccmmunication, ORNL (1958)

- 183 -

 
 

 

 

 

 

Heat Release in Core Vessel and Blanket. The core vessel Mo'f-é._hibltéh{ |

salt reactor is‘ heated by gamne. radiatibn'emana.ting from the core and blanket

and from within the core vessel 1tse1f. Estima.tes of the gamma heating can |
be cbtained by dete.iled enalyses of the type illustrated by Alexander a.nd
Ma.nn 9 The gamma, -ray heating in the core vessel of a reactor with an 8-ft- d:la
core a.nd 0.5 mole % IhFh in the fuel salt has been estimated to be the following°

 

' Soureé' | S | Hea.t Réiea.se Rate (w/cm5)
| Badioaétive_decay in core | 1
-Fissio':n,’h n-.7 capture, and o S
1nelastic ecattering in core - - 5.2
n-y capture in core vessel o ks
" n-7 capture in blanket | - - 0.3
o | Total 11.k

Estimates of gaxma-ray source strengths can be used to provide a crude
estimate of the gemma~ray current entering the blanket. For the 8-ft-dia

‘core, the core contributes 45,3 w of gamma energy per square centimeter to

the blanket and the core vessel contributes 6.8 w/cm , which, multiplied by
_the gurface area of the core vessel, gives a total energy escape into the
blanket of 28.8 Mv. Some of this energy will be reflected into the core, of
course, and some will escape from the reactor vesael, and therefore the value

~of 28.8 Mv is en upper limit. To this may be added the heat released by cap-

ture of neutrons in the blanket. From. the Ocusol-A caleulation for the
8-ft-dia. core and & fuel salt conta.ining 0.5 mole % ThF), it was found that .

'0 176 of the neutrons would be captured in the bla.nketa If an energy relea.se .

of T mev/capture 1 assumed, the heat release at s power level of 600 Mw {heat) 7—
1s estimated to be 8.6 M7, The total is thus 47.4 Mw, or say, 50 ¥ 10 Mw,_ to .
allow for errors. .. | S | |

No allovance ves made for fissions in the blanket. These would add 6 Mo

- for each 1% of the fissions occurring in the blanket, Tiuis it appears that

the heat release rate in the blanket m:l'g‘h'f: 'ra,rige up to 80 Mw.

 

9Alexa.nder, L G., end Mann, L. A., First Estimate of the Germg, Hea,ting

~ in the Core Vessel of & Molten Fluoride Converter, 0RNI.-CF-57-12-57 (19575.

«-J.&h-»

 
 

 

1.2 Intermediate States

 

Without Reprocessing of Fdel'Salt. The nuclear performance of a homo-

 

geneous molten-sslt reactor changes during operation at power because of the
accumulation of fission:products end nonfissionable isotopes of‘uranium. It
is necessary to add U255 to the fuel salt to overcome these poisons;'and, as
a result, the neutron spectrum is hardened and the regeneration ratio decreases

235

because of the accompanying decrease in 1 for U7 and the increased competition
for neutrons by the poisons relative to thorium. The accumulation of the superior
fuel U233 compensates for these effects only in part. The decline in the regen-
eration ratio and the inerease in the critical inventory during the first year of
operation of three reaotors'having 8-ft-dia cores charged, respectively, with
- 0.25, 0.75, and 1 mole % ThFhare-illustrated in Fig. 4.9. The critical inventory
increases by about 300 kg, and the regeneration ratio falls about 164. The gross
burnup of fuel in the reactor charged with 1 mole % ThF), and operated at 60C Mw
with a load factor of 80 amounts to about 0.73 kg/day. The 2 burnup falls
from this value as U~ assumes part of the load. During the first month of
operation, the °? burnup averages 0.69 kg/day. Overcoming the poisons requires
1.53 kg more and brings the feed rate to 2.22 kg/day. 'The initial rate is high
because of the holdup of bred-fuel in the form of Pa->2. As the concentration of
this isotope approaches equilibrium, the £ feed rate falls rapidly. At the
end of the first year the burnup rate has fallen to 0.62 kg/day and the feed rate
to 1.28 kg/day. At this time U253 contributes about 12% of the fissions. The
reactor contains 893 kg of U255,770 kg of U233 7 kg of Pu 59 62 kg of U236

‘and 181 kg of fission products.: The Ua;g end the fission prodncts capture 1.8

,and 5 8% of all neutrons and impair the regeneratiOn ratio by 0. 10 units.r Details
Vfof the inventories and concentrations are given in Table h h._idfffl,,-f ._1'fi5, -
' With,Reprocessing of Fuel Salt. If the fission products were allowed to |

 

_.'ffaccumulate indefinitely, the fuel inventory would become prohibitively large

'nieand the neutron economy would become very poor. Ebwever, 1f the f1531°n Pr°"ii"

 

 

, ffducts are removed, as described in Part 6 at a rste such thet the equilibrium '
lfl~1nventoryis, for. example, eqpal 40 the first year 's production, ‘then the in--
‘.d:crease din- 0255 inventory and the decrease in regeneration ratio are-effectively
*'farrested, as shown in Fig. 4,10, The fuel-addition Tate drops 1mmediately
from 1.28 to 0.73 kg/day when processing is started. At the end of two years,

- 185 -

 
 

 

  

OILY Y NOILOYINTD I

P AMOLNIAN/

  

186
 

 

Table 4.4k, Nuclear Performence of a Two-Region, Homogeneous,
Molten-Fluoride-Selt Reactor Fueled with U25 5

and Containing 1 mole % ThF) in the Fuel Salt

 

Core diameter: 8 ft

External fuel volume: 339 £t
Total power: 600 Mw (heat)
Load factor: 0.8 .

 

 

 

 

 

 

| ",-_.,":,_—U-235 Feed Rate,

kg/day

L ,Regeneration Ratio

2 22?_§if”g:’ e

©1.28-0.73

Initial State After 1 year _
Inventory Absorptions Fissions Inventory Absorptions Fissions
(ke) (%) (R (ke) (%) (%)
Core Elements

Th-~232 2,100 20.3 2,100 16.7
Pa~23%3 8.2 0.3
U=-233 6L.0 5.9 12,5
U-234 1.9 0.0
U-235 60k 5504 100 893 49.3. 86.3
U-236 62.2 1.8
Np=-237 ‘ h.2 0.2
U-238 45.3 2.2 57.9 2.0
Pu-239 6.8 0.8 1.2
Fission fragments 181 3.8
Li-7 %,920 1.9 3,920 0.9
Be-9 3,008 0.6 5,008 0.5

F-19 214,000 3.2 2lt,000 3.0

Blanket Element e |
U-esza SO e e 8.7
, motal Fuel*“ " f' | *5fi;69§1'7§j7f;j17i;f'Wf;i." 965
U235 Burmip. Bate, o e |
o kgfaay .0 o9 o 0.62

 

187

 
 

 

 

 

. Teble kb (continued)

 

 

 

 

After 2 years After 5 years
Inventory Absorptions Fissions  Inventory Absorptions Fissions
(ke) (%) (%) (kg) (%) (%)
Core Elements
Th-232 2,100 16.3 o 2,100 15.4
Pa-233 .9 0.2 - 7.5 0.2
U-233 110 9.7 20,8 201 15.3 33.0
U-234 a 6.5 0. 271 0.k
U-235 863 443 7.4 818 36.9 64l
U-236 115 3.1 222 5.2 ‘
Np-237 0.8 0.4 1.8 0.8 ,
U-238 69.7 2.3 | 9.0 2.7 :
Pu-239 12.0 1.3 1.8 2h,3 2.0 2.9
Fission fragments 181 3.6 181 3.l ' A
Li-7 3,920 0.8 3,920 0.6
Be-9 3,008 0.5 3,008 0.5
F-19 2k, 000 3.0 2k, 000 3.0
Blanket Element
U-233 16 2L
Total Fuel 990 1045
U-255 Burnup Rate,
kg/day 0.58 , 0.47
U-235 Feed Rate, :
kg/day 0.50 0.45
Regeneration Ratio 0.53 0.54

 

-188-

 
 

 

Table 4.4 (contipued)

 

After 10 years

. After 20 years

Inventory Absorptions Fissions Inventory Absorptions Fissions

 

 

(kg) (%) (%) (keg) (%) (%)
Core Elements ,
Th-232 2,100 14.6 2,100 13.7
Pa-233 Tl 0.2 6.7 0.2
U-233 266 L7.6 38.3 322 18.8 4.0
U-23k 6k 0.8 12k 1.4
U-235 831 3345 58,2 872 31.7 54.9
U-236 328 6.7 1150 7.9
Np-237 2.6 0.9 3,2 1.0
U-238 10.8 2.9 12.9 3.0
Pu-239 373 2.4 3.5 52,6 2.8 k.1
Fission fragments 181 2.7 181 2.4
Li-T 3,920 0.5 3,920 0.k
Be-9 3,008 0.5 3,008 0.5
F-19 2k, 000 3.0 2l ,000 3.0
Blanket Element
U-233 28 33
Total Fuel 1,129 1,232
U-235 Burnup Rate,
kg/day 0.h1 - 0.38
U-235 Feed Rate, -
~ kgfday ' Ous 0.39
Regenerstion Ratio ':0.533 o 0.5%0

 

189

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Core Diameter - 8-0" Power— 600 Mw Load Factor- 0.8
08
.‘g
S
S 07
Q
$
o : .
\ .
t o] |
S o
é \\ ' ' - 1% 7./7/';\
T+~ L Decréasing 7'/7)
o4
0 /0 20
)
1000 //
//\ — 1% Thiy
800 T
— - l _ - T " ———Decreasing Thhy
600 P>
&
\ 0,
| S /% Th\
T 400
- _ Uzss
200 — L1
o
0 /0 20

Time of operation, years

Fig. 4.10. Long-Term Nuclear Performance of Typical, Two-Region,
Homogeneous, Molten Fluoride Reactor Fueled with U235.

190

 
 

 

the addition rate is down to 0.50 kg/day, and it continues to decline slowly
to 0.39 kg/day after 20 years of operation. The nonfissionable isotopes of -
uranium continue to accumulate, of course, but these are nearly compensated
for by the ingrowth of U772, As shown in Fig. 4.10, the inventory of N
actually decreases for several years in a typlcal case, and then increases
only moderately during a lifetime of 20 years.

The rapid increase in eritical inventory of U235 during the first year
can be avoided by partial withdrawal of thorium. In Fig. 4.10 the dashed
lines indicate the course of events when thorium is removed at the rate of
1/900 per day. Burnup reduces the thorium concentration by another 1/4300
per day. The U255 inventory rises to 826 kg and then falls, at the end of
elght months, to 587 kg. At this time, the processing rate is increased to
1/240 per day (eight-month cycle), but the thorium is returned to the core
and the thorium concentration falls thereafter only by burnup. It may be
seen that the U235 inventory creeps up slowly and that the regeneration ratio
falls slowly. The increase in U255 inventory could have been prevented by
withdrawing thorium at a small rate; however, the regeneration ratio would
have fallen somewhat more rapidly, and more U235 feed would have been required

to compensate for burnup.
2. FOMOGENEOUS REACTORS FUEIED WITH U=~

Uranium-233 is a superior fuel for use in.molten-fluoride salt reactors
in almost every respect. _The fission croes seetion in the intermediate range
of neutron energies is greeter than the fission cross sections of 0255 and

239 Thus initiel eritical inventories are 1ess, and 1ess edditionel fuel
is required to override poisone. Aleo, the peresitic oross section is sub-
;stantially less, end fewer neutrons are lost to radietive capture. Further,.,r
37%the radiative captures result in the immediete fbrmation of a fertile isotope,,~

.71125 e rate of eecumulation of U2 is orders of’megnitude emaller then
' iwith U 35 es a fuel, and buildup of Np237 and Pu 39 is negligib1e¢.fi,r'

The mean neutron energy is rather nearer to thermel in these reactors
than it is in the correeponding UEBS cages, COnsequently, ‘losses to core
vessel and to core salt tend to be higher. Both losses will be reduced sub-
stantlially at higher thorium concentrations. |

- 191 -

 
 

 

 

2.1 Initial States

Results from s parametric study of the nuclear characteristics of two—
region, homogeneous, molten-fluoride-salt reactors fueled with U 233 are giVen
in Tsble k.5. The core diameters considered range from 3 to 10 ft, and the
thorium concentrations range from 0.25'to'1 mole %. Although the regeneration
ratios are less than unity,.they are very good compared with those obtained
with U255. With 1 mole % ThF), in an 8-ft-dis t::ore,‘the.’l]‘?j'3 inventory vas |
only 196 kg, and the regeneration ratio vas 0.91. ~ - )

' The regeneretion ratios and fuel inventories of reactors of various
diameters containing 0.25 mole % thorium gnd . fueled'with U235 or U 53
compared in Fig. h.ll. The superiority of U235 is obvious.

2.2 Intermediate States

Calculations of the long-term performance of one reactor (Case 51, stle
4.5) with U233 as. the fuel are described below. The core diaemeter used was
8 £t end the thorium concentration was 0.75 mole % The changes in inventory
of U253 end regeneration ratio ate listed in Table 4.6. During the first year
of operation, the inventory rises from 129 to 199 kg, and the regeneration ratio
falls from 0.82 to 0.Tl. - If the reprocessing required to hold the concentration
of fission products and Np 57 constant is begun at this time, the inventory of
U235 increases slowly to 247 kg and the regeneration ratio rises slightly to
0.73 during the next 19 years. This constitutes a substantial improvement‘over

the performance with U235

3 'HOMDGENEOUS REACTORS FUELED WITH PLUTONIUM

| It may be feasible to burn plutonium in molten-fluoride-salt resotors.
The solubility of PuF5 in mixtures of- IiF'and BeFé is eonsiderably less then
. that of UF), but is reported to be over 0.2 mole %,10
for eriticality even in the presence of fission fragments and nonfissionable

isotopes of plutonium but prdbdbly 1imits ‘severely the amount of ThFh that ean |

 

0
.Beryllium Fluoride (in preparation), ORNL (1958).
0192 =

 which may be sufficient

Barton; C."J.;" Bolublldty and’ Sts.hil:lty of PuFz in Fused Alkali Fluoride- o
 

 

 

 

 

 

Teble 4.5. Nucleer Characteristics of Two-Region, Hamogeneous,
Molten-Fluoride=-Salt Reactors Fueled with U 3

 

 

" Core diemeter: 8 ft -
Externel fuel volume: 339 £t
Total power: 600 Mv (heat)
Load factor: 0.8
Case rumber y1 b2 'h3 Lk 45
Fuel and blanket selte* 1 1 1 1 1
Core diameter, ft 3 L 4 5 6
TuF), in fuel salt, mole % 0 o 0.25 0 0.25
U235 in fuel salt, mole % 0.592  0.158  0.233 0.106 0.048
U2 atom density¥* 21,0 6.09 8.26 3,75 1.66
Critical mass, kg of U-2° 64,9 22,3 30.3 26.9 20.5
Critical inventory, kg of U253 1620 248 337 166 82.0
Neutron absorption ratios¥¥¥ _
2 (fissions) - 0.875%  0.8706. 0.8665  0.8725 0.881k
U7 (n-7) | 0.1246 0.129%  0.1335  0.1275  0.1186
Be~Li-F in fuel salt 0.0639  0.1061  0.0860  0.1h72 0.3180
Core vessel 0.0902  0.1k0L  0.1095  0.1380 0.1983
Li-Be-F in blanket salt 0.,0233  0.023%  0.0205  0.0196 0.02L5
Leakage | 0.0477 0.0310  0.0306 0.0195  0,0160
Th in fuel salt © 0,1095  0.1593
Th in blanket salt 0.9722  0.8857 0.8193  0.7066 0.6586
- Neutron yield, n S22 2,19 2.18 2.19 2.21
/"’f:MEdian fission energy, ev r{fg,,€f }?ifh'i;fi”felu:——_ 19 2.9 0.35
. Themsal fissions, % . 0,053 8.0 . 23 16 38
‘fig;;n-y capture to fission ratiO,:__gfi';fifO,lhrt'f  67157 0.15 0.15 - 0.13
-:-E;aRegeneration ratio Z,iy-;jjޢ3: '-;97r,'57 O 89 .93 0.37 - 0.66

 

193

% Fuel'salt No. 1 3limole % BeFo + 69 mole % LiF + UFy + TuF), -
. Blanket salt No. 1: 25 mole % 'm‘u + 75 m°1e % ur
7'l¥*;§Atoms (x 10 19)/¢m ‘ |

o 5f*** Neutrons absorbed per absorption in 0233

 
 

 

 

 

Teble 4.5 {continued)

 

. Case number 46 47 48 Lg
Fuel and blanket salts¥ T 1 1 1
Core diameter;"ft ' 6 8 _ 8 10 -
ThF) in fuel salt, mole % O 0.25 0.25 1 0.25
U233 in fuel salt, mole % 0.066 0.039 0.078 0.031
U2 atom density®*  2.36 150 2.95  1.10
Critical mass, kg of U-2° 29.2 M.l 86.6  63.0
Criticel inventory, kg of U2 17 931 196 104
Neutron absorption ratios*¥¥ ,

U%?° (fissions) 0.8779  0.8850  0.8755  0.888L
v°2? (n-7) 0.1221  0.1150  0.1245  0.1119
Be-Li-F in fuel salt 0.2297  0.3847  0.1899  0.5037
Core vessel 0.1508 0.1406 0.0778 0.1168
Li-Be-F in blanket salt 0,0179  0.0141  0.0095  0.0108
Leakage 0.0157  0.0095  0.0090  0.0068
Th in fuel salt 0.1973  0.2513  0.5768  0.2852
Th in blanket salt 0.5922  0.h421l 0.334h  0.3058
Neutron yleld, 1 . 2.20 2,22 2.20 2.23
Median fission energy, ev 1.2 0.20 1.1 50% Th
Thermal fissions, % | 29 43 ol 50
n-y capture-to-fission ratio, « 0.14 0.13 0.1kL 0.13 .

Regeneration ratio ' 0.79 0.67 0.91 0.59

131
- 216

10

0.063
2,29

0.8781
0.1219
0.2360
0.0629
0.0071
0.0065
0.6507
0.2408
2.20

3.2
30

0.1k

0.89

51

10
0.75

- 0.0597

1.97

58.8
129

0.8809
0.1191
0.2458
0.1168
0.0187

- 0.0050

0.4903

2.21

0.68
34

0.1k

0.82

 

*'  Fuel salt No. 1: 31 mole % BeF, + 69 mole % LiF + UF), + ThF)
Blanket selt No. 1: 25 mole % TuF) + 75 mole % LiF
Fuel salt No. 2: 37 mole % BeFp + 63 mole % LiF + UF) + ThF
Blanket salt No. 2: 13 mole % ThF), + 16 mole % BeF, + T1 mole % LiF

**  Atoms (x 10"19)/cm5.
¥¥%X  Neutrons absorbed per absorption in U23j.

i

-194~
 

 

 

 

 

 

 

 

 

 

 

 

UNCLASSIFIED
ORNL~-LR~-DWG 24923R
{.0
233
/U
' /..Z.
Co0.8 —
-
<
o
Z l
o ¢
- U235
<
& ‘ _.__.L/
W \ >
o
L
o
0.4 —
o 200 400 600

~ CRITICAL INVENTORY (kg of U)

Fig. 4.11. Comparison of Regeneration Ratios in Molten-Salt Reactors
- - Containing 0,25 Mole % ThF4 and 'U_235' or U233 Enriched Fuel.

 

195

 
 

 

 

 

Teble 4.6. Nuclear Performence of a Two-Region, Hqgggeneous,
Molten-Fluoride-Selt Reactor Fueled with U-2
end Containing 0.75 mole % ThF) in the Fuel Salt

 

Core diameter; 8 ft

External fuel volume: 339 ft3
Total power: 600 Mw (heat)
Load factor: 0.8

 

 

 

 

Initial State _ After 1 year _
Inventory Absorptions Fissiocns Inventory  Absorptions Fissions
(kg) (%) (%) (kg) (%) (%)
Core elements
Th-232 1,572 22,2 1,572 19.1
Pa-233 9.4 0.5
U-233 129 45,2 100 199 k5.3 99.5
U-234 | 23.3 0.9 . :
U-235 1.9 0.3 0.5
U-236 0.1 0.1
Np-237 ’
U-238
Pu-239
Fission fragments 181 7.9
Li-6 3,920 6.5 . 3,920 3L
Be-9 3,00l 0.8 3,008 0.7
F-19 2k,000 4,0 2k ,000 3.5
Blanket element
U-233 8.6
Total fuel 129 210
U-233 Feed Rate,
kg/day 0.790 0.370-0.189
Regeneration Ratio 0.82 0.71

 

-196-
 

 

 

 

 

 

o

Taeble 4.6 (continued)

 

After 2 years
Inventory  Absorptions
(kg) . (%)

Fissions

(%)

After 5 years
Inventory  Absorptions Fissions

(kg) (%) (%)

 

Core elements
Th-232
Pa-233%
U-233
U-23k
U-2355
U-236
Np-237
U-238
Pu-239

Fission fragmefits

1i-6
Be-9
F-19
Blanket element
U-23%3
Total fuel

1,572 18.9
9.0 0.5
20k L9 98.5
L4,0 1.7
Solt 0.8 1.5
0.6 0.3
0.1 0.1

1817_ TeT

3,920 33
3,008 - 0.6
2k, 000 3.l

10.7“

220

1,572 18.3
8.9 0.k
276 43,7 95.6
89 31
17.7 2.3 LY
h.2 0.2
0.5 0.1
0.3

181 7.2
3,920 2.8
3,008 0.6
24,000 3.3

16.2

050 .

 

kefday

U-233 Feed Rate, - .o
- Regeneratlon Ratlo.

0.181

 

 

 

 

197

 
 

 

 

 

Teble 4.6 (_Cdntimzed)" |

 

After 10 years _

After 20 years

 

Fissions Inventory Absorptions Fiseions

 

 

" ‘Inventory Absorptions
 (ka) (® - (%5 (k) (%) (%)
Core elements s : | o -
Th-232 1,572 17.8 1,572 17.2
Pa-233 8.6 ok S8 0.k
U-233 2% k2.5 92.8 247 k1.5 9.5
U-234 132 b2 12 5.0 |
U-235 32.5 BT 1AW 1.8 9.0
. U-236 12.5 0.6 2k 1.1
 Np-237 1.7 0.2 3. 0.3
U-238 1.7 0.1 5.1 0.3
Pu-239 0.2 0.1 0.1 - 0.8 0.3 0.5
Fission. fragments 181 6.7 - 181 6.3
Li-6 3,920 2.5 3,920 2.1
Be-9 3,008 0.6 3,008 0.6
F-19 2,000 3.3 24,000 3.3
 Blanket element
U-233 22.2 31.6
Total Fuel 282 295
U-233 Feed Rate, | .
kg/day 0.171 0.168
Regeneration Ratio 0.73 0.73

 

.198-
 

 

 

the ingrowth or fission pruducts will neaessitate the addition of more Pu

 

be added to the fuel s'alt.' Th:ls limita.tion, coupled with the condition that

259 is an inferior fuel in intermediate reactors, will result in & poor neu-

| tron economy in camparisen with that of 0255-fueled reactors. Ebwever, the
advantages of handling‘plutonium in e fluid fuel system.may'make the plutonium- .

fueled.molten-salt reactor more desir&ble than other possible rlutonium-

‘burning systems.

3.1 Initial States

Critlcal Concentration, Vass, Inventoryj“and Regeneration Ratio. The‘results

of calculations of a plutonium-fueled reactor having & core dlameter of 8 £t and
no thorium in the fuel salt are deseribed belcw;, The eritical eoncentratiofi was
0.013 mole 4 PuF3 which 1is an order of magnitude smaller ‘than the solubility
limits in the fluoride salts of interest. The critical mass was. 13, 7 kg and

the critical inventory in a 600-Mw system (339 ft5 of external fuel volume) vas
only 3l.2 kg. :

The core wes surrounded by the Ii-Be-Th-fluoride blanket nixture No. 2

(15 per cent ThFL) 8lightly more than 19% of all neutrons were captured in
the thorium to give a regeneration ratio of 0.35. By employing smaller cores

. and larger investments in Pu 2% , however, it should be posaible to increase

the regeneration ratio substantially.

Neutron Balance and Miscellaneous Detalls. Details of the neutron econony

of a reactor fueled with plutonium are given in Tsble 4.7, Parasitic captfires
in Pu?’ are relatively high; n is 1.8k, campared.with e v of 2.9. The neutron

.spectrum is relatively soft; almost 60% of all fiasions are caused'by thermal
.neutrons, and, as a reault, absorptions 1n 1ithium are high.

z, 2 Intermediate States

On the basis of the averaga value of a of Pu 59, it is estimated that

Puaho will accumnlate in the system wntil it capturen, at eqnilibrium, about

rhalf as many neutrons as Pu 39. While these captures are not wholly parasitie,

-2kl

-inasmuch as the product, qu s 1s fission&ble, the added campetition for neu-

trons will necessitate an- 1ncrease in the eoncentration of the 39. Iikewise,
239

Further, the’ rare earths among the fiseion products may exert a cammon-ion
influence on the plutonium end reduce ite solubility. On the credit side,

.na'199~

 
 

 TABIE 4.7

Initial State Nuclear Characteristics of & Twpical
‘Molten-Fluord de-Salt Reactor Fueled with Pu 239

Core Diameter:
External fuel volume:
Total power:

Ioad factor:

Critical inventory:
Criticael concentration:

8 £t 5

339 £t

600 Mw (heat)

0.8 239
31.2 kg of Pu’ 219

0,013 mole %

 

Neutrons Absorbed per
Neutrons Absorbed in Pu

39

 

 

. Neutron Abéorbers

0.6%0

 

239 (fissions)

259 (n-7)" 0.372
116_and_1i7 in fuel salt 0.202
Be? in fuel salt 0.022
2 1n" fuel salt 0.086
Core Vessel - 0.086
Th in blanket salt 0.352
Ii-Be-F in blanket salt 0.02h
fRéactor vessel 0,00k
Ieakage - 0.003

NEutron Yiéid,:nA-\ 1.éh
Thermal Fissions, % 59
. 0.352

Regeneration Ratio

 

 

« 200 =

=

2

»

 
 

 

 

:"qi

however, is the 0235 produced in the blanket. If this is added to the core
it mey compensate for the ingrowth of Pu2h0 and reduce the Pu?39'requirement
to below the solubility limit, and 1t may be possible to operate indefinitely,

es with the U255-erled reactors.
L. HETEROGENEOUSV GRAPHITE-MODERATED REACTORS

The use of a moderator in a heterogeneous lgttice w1th molten selt fuels
is potentially advantageous. First, the approach to & thermal neutron spectrum
improves the neutron yleld, q, attainable, especially'with U235 and Pu 39
Second, in a heterogeneous syetem, the fuel is partially shielded from’neutrona
of intermediate energy, and a further improvement in effective'neutron yield, 1,
redults. Further, the optimum systems may prove to have smeller volumes of fuel
in the core than the corresponding flnorine-moderated homogeneous reactors and,
consequently, higher concentrations of fuel and thorium in the melt. This may
substantially reduce parasitic losses to components of the carrier salt. On
the other hand, these higher cancentrations tend to increase the inventory’in
the eirculating fuel system externsl to the core. The same considerations
apply'to fission products and to nonfissionable isotOpes of uranium.

Possible moderators for molten-salt reactors include beryllium, BeO, and

'graphite._ The design and performance of the Aircraft Reactor Experiment, a
‘beryllium oxide moderated eodium-zirconium fluoride salt one-region, U235

fueled burner reactor,has been reported (see Part 1) Since beryllium and BeO

and molten salte are ‘not chemically compatible, it was: neceSSery to line the

fuel circuit with.Inconel. "It is easily estimated that the presence of Inconel,

'or any other prospective containment metel in & heterogeneous, thermal reactor

would seriously'impair the regeneration ratio of & conyerteerreeder " Conse-
quently, beryllium and Beo are eliminated from consideretiona |
| Preliminary evidence indiCates thet uraniumJBearing molten salte may be

"compatible with some grades of grephite and that the presence of ‘the graphite’

will not carburize metellic portions of the fuel circuit seriouely.11 It

 

11K9rtesz, F., Private Communieation, ORNL (1958)

- 201 =

 
 

 

e i e, et M v s P

therefore becomes of interest to explore the capsbilities of thefgraphitee ,
moderated systems. The principal independent variables of interést.are-the 5

-~ core diameter, fuel channel diameter, lattice specing, and thorium cbncentrfiti@h.

k.1 Initial States

The nuclear'parameter study of graphite-moderated reactors has‘just begun

_and’dnly two cases have been calculated. For convenience in comparison with -

the IMFR, these first two "MSFR" calculations vere based on essentially the

same geometry and graphite-to-fluld ratio as thdse of the referéhce design'IMER;la

with molten salt substituted for liquid.metal.» One calculation was performed
with bismuth instead of salt as & check point. The three cases are sumarized
in Table 4.8, ' o

 

_ 2Babcock and Wilcox Co., Iiquid Metal Fuel Reactor, Technical Fbasibility
Report, BAW-2 (Del) (1955). |

“ - 202 =
 

 

 

TABIE 4.8

Camparison of Graphite-Mbderated.MbltenLSalt

 

 

 

 

 

 

| MR MSFR-1 MSFR-2
Totael power, Mw (heat) 580 600 600
Over-all radius, in. ' 75 T5 T2
Critical mass, kg of U-233 - - -~ 9.9 9.6 27, 7
Critical inventory, kg of U-233%% U867 T7.8 213
Regeneration ratio | 1.107 0.83 Y. 07
Core o |
Redius, in. 35 -3 - 34,8
Graphite, vol % | g L5 5 s
Fuel fluid, vol % | 55 55 55
Fuel components, mole % ~
Bi ~ 100
IiF 69 61
BeF2 31 36'5
ThF), 2.5
‘Unmoderated blanket
Thickness, in. 6 6 13,2
Composition, mole % . '
Bl Q0
Th 20 (Th) 10 (ThF,) 13 (ThFh)
LiF _ 70 71
BeFa : 20 - 16 -
~ Moderated blanket
' Thickness, in. e 36 36 ol
Composition, vol % | :
Graphite . . 666 66.6 100 -
Blenket fluidg®x . 33,4 334
NEutron dbsorption ratio*** - " .
" Thin fuel fluid LT e 04566
. U-233 in fuel fluid ST 04918 ';c 925 © . 1,000
" Other components of fuel fluid - 0.081 C0.328 0 0.106
fTh in blenket fluid .. 1.110 -~ 0.825 0.4
“U-255 in blanket" fluid L 0.083 0.071 ,
‘Other camponents of blanket fluid 0.040 0,092 . 0.03%8
. Ieaksge ... .. 0.012 ~ 0.008  0.01k
'NEutron yield, q ¢:'3 _ .}-,_1? [f__.__ . 2;2 e 2.28 . 2421

 

¥  With bismuth, the external volume indicated in ref 10 was used. The
molten salt systems:are caleculated for 339 ft5 external volumes.

¥¥ Dame as ummoderated blanket fluid.

*¥¥% Neutrons absorbed per neutron absorbed in.U-253.

..203 -

 
 

 

PART 5

EQUIPMENT FOR MOITEN-SALT REACTOR HEAT TRANSFER SYSTEMS

The equipment required in the heat transfer circuits of a molten;salt
reactor consists of the components needed to contain, circulate, cool, heat,
and control molten salts at temperatures up to 1300°F. Included in such
systems are pumps, heat exchangers, piping, expansion tanks, storage vessels,
valves, devices for sensing operating variables, and other auxiliary equipment.

Pumps for the fuel and blanket salts differ from standard centrifugal
pumpé for ‘operation at high temperatures in that provisions must be made to
exclude oxidants and lubricants from the salts, to prevent uncontrolled escape
of salté,and gases, and to minimize heating and irradiation of the drive motors.
Héat is transferred from both the fuel and the blanket salts to sodium in shell-
and-tube heat exchangers designed to maximize heat transfer Per unit volume and
to minimize the contained volume of salt - especially the fuel salt.

Seamless piping is used, where possible, to minimize flaws. Thermal
expansion is accomodated by prestressing the pipe and by using expansion loops
and joints. Heaters and thermal insulation are provided on all components
that contain salt or sodium for préheating and for maintaining the circuits
at temperatures above the freezing points of the liquids and to minimize heat
loésgs. Devices are provided for sensing flow rates, pressures, temperatures,
and liquid levels. The devices include venturi tubes, pPressure trfinsmitters,
thermocouples, electrical probes, and floats. Inert gases are used over free-
liquid surfaces to prevent oxidation and to apply appropriate base pressures
for suppressing cavitation or moving liquid or gas from one vessel to another.
| The deviations from standard practice required to adapt the various com-

ponents to the molten-salt system are discussed below. The schematic diagram
of a molten-salt heat transfer system presented in Fig. 5.1 indicates the |
relative positions of the various components. For nuclear oPeration,‘an off~
gas system is éupplied, as described in Part 1. The vapor condensation trap
indicated in Fig. 5.1 is required only on systems that contain Zth or a com-
parably volatile fluoride as a component of the molten salt.

- 20k -
 

-

 

 

UNCLASSIFIED
ORNL—LR—DWG 24291

/PUMP (INCLUDES EXPANSION VOLUME)

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

5 VAPOR TRAP
LIQUID LEVEL
INDICATING DEVICE—-—I | 7
3%
HIGH POINT |- - - - ~— - - SLOPE
b \
HEAT I ngaT
SOURCE ,
oo EXCHANGER
O—= - L COOLANT SYSTEM
FLOW
 MEASURING DEVICE
LOW (VENTURI)
POINT

 

 

 

—t ] >
o EE;ESSURE MEASURING DEVICES

SHUT-OFF VALVE
, \% ' - : - VENT VENT

LIQUID LEVEL

FILL AND DRAIN LINE INDICATING DEVICE

 
 

_VAPOR TRAP

' iy " ' EQUALIZER

VALVE X

 

 

 

 

 

 

 

 

 

 

DUMP TANK

 

 

———— PRESSURE |
~ REGULATORS —= _

 

 

 

INERT
GAS SUPPLY

 

Fig. 5.1. A Molten-Sait Heat Truhsfer System.

205

 
 

 

1. PUMPS FOR MOLTEN SALTS

Centrifugal pumps with radial or mixed-flow types of impeller have been_
used successfully to eclirculate molten-salt fuels. The units built thus far
and those currently being developed have a vertical shaft which carries the
impeller at its lower end. The shaft passes through & free surface of liquid
to isolate the motor, the seals, and the upper bearings from direct contact
- with the molten salt. Uncontrolled escape of fission gases or entry of unde-
sirable contaminants to the cover gas above the free-liquid surface in the
pump are prevented either by the use of mechanical shaft seals or hermetic
enclosure of the pump and, if necessary, the motor. Thermal and radiation
shields or barriers are provided to assure. acceptable temperature and radia-

- tion levels in the motor, seal, and bearing areas. Liquid cooling of internal
pump surfaces is provided to remove heat 1nduced by gamma and beta radiation.

The principles used in the design of pufips for normal liquids are appli-
cable to the hydraulic design of a molten~-salt pump. Experiments have shown
that the cavitation performance of molten-salt pumps can be predicted from
tests made with water at room temperature. In addition to stresses induced
by normal thermal effects, stresses due to radiation must be taken into’
account in all phases of design.

The pump shown in Fig. 5.2.was developed for 2000-hr durability at very
low irradiation levels and was used in the'Aircraft Reactor Experiment for
circulating molten salts and sodium at flow rates of 50 to 150 gpm, at heads
up to 250 ft, and at temperatures up to 1550°F. These pumps have been vir-
tually trouble-free in operation, and many units in addition to those used
in the Aircraft Reactor Experiment have been used in developmental tests of
various components of molten-salt systems.

The bearings, seals, shaft, and impeller form a cartridge-type subassembly
thatris-removable from the pump tank after opening a single, gasketed Jeint a-
bove the liquid'levelQ The volute, suction, and discharge connections form
parts of the pump tank subassembly into which the removable cartridge is in-
serted. 'The,upper portion of the shaft and a toroidal area in the lower part
of the.bearing housing are cooled by circulating oil. Heat losses during
operation are reduced by thermal insulation. |

- 206 -

‘(3

 
20T

. . 7 3 > - ‘

UNCLASSIFIED
ORNL.LR-DWG. 6218R

  
 
 
 
  
    
 
  
   
 
  
       
   

DRIVE SHEAVE
UPPER SEAL OIL RETAINER

- UPPER SEAL LEAKAGE DRAIN
BEARING CLAMP RING

SLEEVE
BEARING HOUSING
S SPACER SLEEVE
TOP SHAFT SEAL ASSEMBLY SHIELD
‘ SEAL FACE RING OlL INLET
BEARING (UPPER)
_INTERNAL SHAFT COOLING B.H. BREATHER

 LUBE OiL. GUIDE TUBE
HEATED GAS VENT AND
LIQUID INJECTION NOZZLE

O RESISTANCE
HEATER CONNECTION
LAVA SPACER

BEARING SPACER
BEARING (LOWER)

. ~SEAL RING
OVERFLOW

 
 
 
 
  

e b e L L 'SEAL FACE RING
| ‘ © L w o SUNGER
. LOWER SEAL ASSEMBLY—
o " SPACER AND HEAT DAM
OIL :DRAIN HEADER RING

  
  
   
    
  

  

RESISTANCE
HEATER CONNECTION

  
 
 

    
 
   

DRAIN
) COOLANT CONNECTION

*  PUMP BODY ASSEMBLY. L FLANGE
~ - CLAMP

THERMOCOUPLE. GLAND——— =

  
   
    
   
  
 
  
 
    
 
 
     

PUMP TANK
COVER FLANGE

 

STILLING WELL

LIQUID LEVEL

DISCHARGE ELBOW' PUMP  TANK

LEVEL INDICATOR FLOAT

FLEXIBLE CONNECTOR

IMPELLER

HOUSING VARIABLE INDUCTANCE LEVEL INDICATOR

ANTISWIRL
INLET BAFFLE

DISCHARGE LEG

Fig. 5.2. Sump-Type Centrifugal Pump Developed for Use in the Aircraft Reactor Experiment,

 

 
 

 

 

 

A

In a2ll the units bullt thus far nickel-chrome alloys have been used in
_the construction of all the high-temperature wetted parts of the pump to
minimize corrosion. The relatively low thermal conductivity and high strength
of such alloys permitted close spacing of the impeller and bearings and high
thermal gradients in the shaft. ' '

Thrust loads are carried at the top of the shaft by a matched pair of
| pfeloaded angular-contact ball bearings mounted face-to-face in order to pro-
vide the'flexibility required to a#oid.bindings and to accomodate thermsl
distortions. Either éingle—row ball bearings or a journal bearing cafi be
used successfully for the lower bearing. |

The upper lubricant-to-air and the lower lubricant-to-inert-gas seals
are similar, rotary, mechanical face-type seals consisting of a stationary
graphite member operating in contact with & hardened-steel rotating member.
The seals are olil-lubricated, and the leakage of oil to the process side is
approximately 1 to 5 cmB/day. This oil is collected in a catch basin and
removed from the pump by gas-pressure sparging or by gravity.

The accumulation of some 200,000 hr of relatively trouble-free test
operation in the temperature range of 1200 to lSOOoF with molten sglts and
liquid metals.as the circulated fluids has proved the adequacy of this basic
pump design with regard to the major prdbiem of thermally induced distortibns.
Four différent sizes and eight models of pumps have been used to provide flows
in the range of 5 to 1500 gpm. Several individual pumps have operated for
periods of 6000 to 8000 hr, consecutively, without maintenance.

1.1 TImprovements Desired for Power Reactor Fuel Pump ;

The bdsic.pump described above has bearings and seals that are oil-lubricated
and cdoled, and in some of the pumps elastomers have been used as seals between
parts. The pump of this type that was used in the ARE was designed for a rela-
tively low level of radiation and received an integrated dose of less than
5 x 108 r. Under these conditions both the lubricants and elastomers were
proved to be entirely satisfactory.

The fuel pump for s power reactor, however, must last for many years.

The radiation level aenticipated at the surface of the fuel is 105 to 106 r/hr.

Beta- and gamma-emitting fission gases will permeate gll aveilable gas space

above the fuel, and the daughter fission products will be depoéited on all

exposed surfaces. Under these conditiofis, the simple pump described ebove
would fail within a few thousand hours. |

- 208 -

C

 
 

 

 

 

Considerable improvement in the resistance of the pump and motor to
radiation can be achieved by relatively simple means. Iengthening the shaft
between the impeller and the lower motor bearing and inserting additional
shielding material will reduce the radiation from the fuel to a low level
at the lower motor bearing and the motor. Hbllofi, metal O-rings or another
metal gasket arrangement can be used to replace the elastomer seals. The
sliding seal just below the lower motor bearing, which prevents escape of
the fission-preduct gases or inleakage of the outside atmosphere, must be
lubricated to ensure continued operation. If oil lubrication is used,'
radiation may quickly cause coking. Various phenyls, or mixtures of them,
are much less subject to formation of gums and cokes under radiation and
could be used as lubricant for the seal and for the lower motor beariné;

This bearing would be of the friction type, for radial and thrust loads.

' These modifications would provide a fuel pump with an expected life of_the .
order of a year. With suitable provisions for remote maintenance and repair,
these simple and relatively sure improvemente would probably suffice for
power reactor operation. '

Three additional improvements, now being studied, should make possible
a fuel pump that will operate trouble-free throughout a very long life. The
first of these is a 'pilot bearing for operation in the fuel salt. Such a
bearing, whether of the hydrostatic or hydrodynamic design, would be com-
pletely unaffected.by radiation and would permit use of a long shaft so that

- the motor could be well shielded. A combined radial and thrust bearing just
below the motor rotor would be the only other bearing required. The second
improvement is a 1abyrinth type of gas seal to prevent escape of fission gases
up the sheft. There are no rubbing surfaces and hence no ‘need for lubricants,l
g0 there can be no radietion demege. “The third innovation is a hemispherical
gas-eushioned bearing to aet as ‘8 combined thrust end radial bearing. It
would heve the advantage of requiring no auxiliary lubrication supply, and it
would cedbine well. with the 1ebyrinth type of gas seal. It would, of couree,
be unaffected.by redietion, _;' :

1. 2 A Propoeed Fuel Pump

A punp design embodying these 1ast three feetures is shown in Fig. 5.3,
It is designed for operation at a temperature of 1200 F,Aaiflow rate of 24,000 gpm,
and s head of TO ft of fluid. The lower bearing is of the hydrostatic type and is

- 209 -

 
 

 

 

    
 

-

o _
Fig. 5.3. Improved Molten-Salt Pump Designed for Power Reactor Use. Operating

  

Temperature, 1200°F; Flow Rate, 24,000 gpm; Head, 70 ft of Fluid.

210
 

eI

 

 

 

lubricated by the molten-salt fuel. The upper bearing, which is also of the
hydrostatic type; is cusfiiohed‘by helium and serves also as a barrier against
passage of gaseous fission products into the motor. This bearing is hemi-
spherical to permit accomodation of thermally induced distortions in the
over-all pump structure. |

The principal radiation shielding is that provided between the source
and the area of the.motor windings. Iayers of beryllium and boron for neutron .
shielding and a heavy metal»for gamma radiation shielding are proposed. The
motor is totally enclosed in order to eliminate the need for a shaft seal. A
coolant is circfilated in the area outside the stator windings and between the
upper bearing and the shielding. Molten-salt fuel is circulated over the sur-
faces of these parts of the pump which are in contact with the gaseous fission

products to remove heat generated in the metal.
2. HEAT EXCHANGERS, EXPANSION TANKS, AND DRAIN TANKS

The heat exchangers, expansion tanks, and drain tanks mfist be‘especially
designed to fit the particular reactor system chosen. The design data of items
suitable for a specific reactor plant are described in Part 1. The special
problems encountered are the need for preheating all salt- and sodium-containing
components, for cooling the exposed metal surfaces in the expansion tank, and
for removing afterheat from the drain tanks. It has been found that the molten
salts behave as normal fluids during pumping and flow and that the heat transfer
coefficients can be predicted from the physical properties of the salts.

3. VALVES

The prdblems associated with valves for molten-salt fuels are the consis-

'tent alignment of parts during transitions from rocm temperature to 1200° F,

the selection of materials for mating surfaces which will not fusion-bond in

the salt and cause the valve to stick in the closed position, and the pro-

vision of a gas-tight seal.; Bellcws-sealed, mechanically operated, poppet

valves of the type shown in Fig. Se A have given reliable service in test systems.
A number .of corrosion- and fusion-bond-resistant materials for high-

temperature use were found_fihrough exfehsive screening tests. Molybdenum

 
 

 

 

"~ UNCLASSIFIED
ORNL—-LR-DWG 21292

1 SYSTEM

    
   

CERMET SEAT
AND POPPET

BELLOWS
PROTECTOR

BELLOWS

DUMP
TANK

 

c Fig. 5.4, Bellows-Sealed, Mechanically Operated, Poppet Valve for Molten=Salt
ervice. '

212
 

 

 

against tungsten or copper and several titanium or tungsten carbide-nickel
cermets mating with each other proved to be satisfactory. Valves with very
accurately machined cermet‘seats and poppets haue operated satisfactorily
in 2-in. molten-salt lines at 1300°F with leakage rates of legs than 2 cmB/hr.
Consistent positioning of the poppet and seat to assure leak tightness is
achieved by minimizing transmission of valve body distortions to the valve
stem and poppet. ' '

If rapid valve operation is not required, a simple "freeze" valve may
be used to ensure a leak-tight seal. The freeze valve consists of a section"
of pipe, usually flattened, that is fitted with a device to cool end freeze
e salt plug and another means of subsequently’heating and melting the plug.

L. SYSTEM HEATING

Molten-salt systems must be heated to prevent thermal shock during fill-
ing and to prevent freezing of the salt when the reactor is not operating to

‘produce power. Straight pipe sections are normally heated by an electric

tube-furnace type of heater formed of exposed Nichrome V wire in a ceramic
shell (clamshell heaters). A similar type of heater with the Nichrome V wire
installed in flat ceramic blocks can be used to heat flat surfaces or large
components, such as dump tanks, etc. In general, these heaters are satis-
factory for continuous operation at 1800°F, Pipe bends, irregular shapes,
and small components, such as valves and.pressureameasuring devices, are
usually heated with tubular heaters (e.g., General Electric Company "Calrods")
which can be. shaped to Fit the component or pipe bend. In general, this type
of heater should be limited to service at 1500 F. Care must be exercised in
the installation of tubular heaters to avoid failure due to a hot spot caused
by insulatioh in direct contact with the heater. This type of failure can be
avoidednby installing a thin sheet of metal (shim stock) between the heater

' and the insulation. ,"

Direct resistance heating in which an electric current is passed directly

:,through 8 section of the molten-salt piping has also been used successfully.

Operating temperatures of this type of heater are 1imited onlyiby the corrosion
and strength 1imitations of the metal as the temperature is increased. Experi -
ence has indicated that heating of pipe bends by this method is usually not
uniform and can be asccompanied by hot spots caused by nonuniformity of liquid
flow in the bend.

- 213 w

 
 

 

 

 

5. JOINTS

Failures of‘some system conponents may be expected during the desired

' operating life, say 20 years, of a molten- salt power-producing reactor, con-
sequently, provisions must be made for servicing or removing and replacing
such_components., Remotely controlled manipulations will be required because
there will be a high level of radiation within the primary shield. Repair
work on or preparations for disposal of components that fail will be carried
out in separate hot cell fac1lities. |

The components of the system are interconnected by piping, and flanged
connections or welded joints may be used, In breaking connections between a
component and the piping the cleanliness of the system must be preserved, and
in remaking a connection proper alignment of parts must be re-established.

The reassembled system must conform to the original leak-tightness sPecifica—
tions. Special tools and handling equipment will be needed to separate
components from the piping and to transport parts within the highly radio-
active regions of the system. While an all-welded system provides the highest
structural integrity, remote cutting of welds, remote welding, and inspection
of such welds are difficult operations. Speclal tools are being developed

for these tasks, but they are not yet generally available. Flanged connections,
which are attractive from the point of view of tooling, present problems of
permanence of their leak tightness.

Three types of flanged joints are being tested that show promise. One
is a freeze-flange joint that consists of a conventional flanged-ring joint
with a cooled annulus between the ring and the process fluid. The salt that
enters the annulus freezes and provides the primary sealy, The ring.provides_
a backup seal against salt and gas leakage. The annulus between the ring and
frozen materiasl can be monitored for fission product or other gas leakage. The
design of this joint is illustrated in Fig. 5.5.

A cast-metal-sealed flanged joint is also being tested for use in vertical
runs of pipe. As shown in Fig. 5.6, this joint includes a seal which is cast
in place in an annulus provided to contain it. When the connection is to be
made or broken the seal is melted. Mechanical strength is supplied'by clamps

or bolts.

- 21 o

-

 
 

 

 

UNCLASSIFIED
ORNL-LR-DWG 27895

»' l-*GAP (~Yi6 in)

V//’ N} SOFT-IRON OR COPPER SEAL RING

7

 

 

 

 

 

 

 

~— AIR CHANNEL FOR COOLING

 

 

AN

FROZEN-SALT SEAL

 

 

 

 

WELD OF FLANGE TO LOOP TUBING

 

 

 

 

 

—s— SALT FLOW

 

 

 

 

 

 

A
A

RING INSERT TO PROVIDE LABYRINTH
FOR SALT LEAKAGE

 

FROZEN-SALT SEAL

Z

 

- — NARROW SECTION TO REDUCE HEAT
— TRANSFER FROM THE MOLTEN SALT
IN THE LOOP TUBING

£ A 4

 

 

 

1 Y 0 % 1
e —
INCHES

 

 

 

 

7

 

 

 

 

 

 

© \__ARINLETTO COOLING CHANNEL

'f BETWEEN FLANGE FACES INDICATE REGION OF FROZEN-SALT
 SEAL; 2222 INDICATE REGION OF TRANSITION FROM LIQUID
| TO SOLID SALT

Fig. 5.5. Freeze-Flange Joint for 1/2-in.~OD Tubing.

215

 
 

 

 

 

 

 

 

UNCLASSIFIED
ORNL-LR-DWG 27898

 

i

 

 

 

 

 

 

SEAL MATERIAL INSERT
SHOWN BEFORE BEING

\\ FUSED TO FORM SEAL

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

{ 1/2 0 ‘/2. 1

WELD OF / T INCHES
FLANGE TO LOOP TUBING ~” -

Fig. 5.6. Cast-Metal-Sealed Flanged Joint.

 

 

 

216

 
 

 

 

 

 

A flanged joint containing a gasket (Fig. 5.7) is the third type of
Joint being considered. In this joint the flange faces have sharp, cir-
cular, mating ridges. The opposing ridges compress a soft metal gasket to
form the seal between the flanges.

6. INSTRUMENTS

Sensing devices are required in moltenJSalt systems for the measurement
of flow rates, pressures, temperatures, and liquid levels. Devices for these
services are evaluated according to the following criteria: (1) they must be
of leak-tight, preferably allewelded, construction; (2). they must be capable
of operating at maximum tempersture of the fluid system; (3) their acouracies
must be relatively unaffected by changes in the system temperature; (4) they
should provide lifetimes at least as great as the lifetime of the reactor;
(5) each must be constructed so that, if the sensing element fails, only the
measurement supplied by it is lost. The fluid system to which the instrument
is attached must not be jeopardized by failure of the sensing element.

6.1 TFlow Measurements

Flow rates are measured in molten—salt systems with orifice or venturi
elements. The pressures developed across the sensing element are measured
by comparing the outputs of two pressure-measuring devices. Magnetic flow-
meters are not at present sufficiently sensitive for molten-salt service
because offthe,poor'electrioal_conduotivitj of the salts.

6.2 Pressure Measurements . -

Measurements of system pressures require that transducers operate at g
safe margin above the melting point of the salt, and thus the minimum trans-
' ducer operating temperature is usually about 1200° F. ‘The pressure transducers "~
that are available are of two types (1) a pneumatic fOrce-balanced unit and
(2) a displacement unit in which ‘the pressure is sensed by displacement of a
Bourdon tube or. diaphragm. The pneumatie force-balanced unit has the disadvan-
teges that loss of the instrument gas supply (usually air) can result in loss ‘
of the measurement and that failure of the bellows or diaphragm would open the
process system to the alr supply or to the atmosphere. The displacement unit,

- 217 -

 
 

 

 

8ic

~ UNCLASSIFIED
ORNL-LR-DWG 27899

   
   

WELD OF FLANGE TO
LOOP TUBING

 

e ey —

INCHES
ANNEALED COPPER RING | |

Fig. 5.7. Indented=Seal Flange. |

RAISED TOOTH 4 b O 4 1

 

 
 

 

on the other hand, makes use of an isolating fluid to transfer the sensed
pressure hydrostatically to an isolatedzlow-temperature output element.. Thus,
in the event of a failure‘of-the_primary diaphragm, the trocess fluid would
merely mix with the isolating fluid and:the closure of the syetem would be
unaffected. | o

6. 3 Temperature Mhasurements

Temperatures in the range of 800 to 1300 F are commonly measured with
Chromel-Alumel thermocouples or platinum-platinum-rhodium thermocouples. The
accuracy and life of a thermocouple in the temperature range of interest are
functions of the wire size, and, in general, the largest p0381ble thermocouple
should be used. Either beaded thermocouples or the newer, magnesium oxide-

insulated thermocouples may be used.

6.4 Iiquid-Ievel Measurements

Instruments are available for both on-off and continuous level measure-
ments. On-off measurements are made with modified automotive-type spark plugs
in which 8 long rod is used in place of the normal center conduc¢tor of the
spark plug.  In order to obtain a continuous level measurement, the fluid head
is measured with a differential pressure instrument. The pressure required to
bubble a gas into the fluid is compared with the pressure above the liquid to
obtain the fluid head. Resistance probe and float types of level-indicators
are available for uee in liquid-metal systems. |

6.5 Muclear Sensors ,

Nuelear sensors for molten-salt reactors are. similar to. those of other
reactors and are not required to withstand high temperatures. Existing and
'well-tested fission, ionizatiOn, and ‘oron trifluoride ‘thermal-neutron detec-
tion chafibers ere available for installation at all points essential to reactor
| ioperation.; Their disadvantages of limited life can be countered only'by dupli-
 cation or replacement and provisions can’ be’ made for this. It should be
;pointed out that the relatively large, negative, temperature coefficients of
reactivity provided by'most circulating-fuel reactors, make these instruments
unessential to the routine operation of the reactor. ‘

w 210 w

 
 

 

 

 

PART 6
BUILDUP OF NUCLEAR POISONS AND METHODS OF CHEMICAL PROCESSING

~ Even though nearly pure 1?2 or 177 is used as the initial fuel
- for & reactor, undesireble products quickly build up as & result of
the fissioo.process. First, each uranium nucleus thet undergoes fis-
'sion'splits into two "fission product"” elements. The'fission products
are‘nuclear poisons to vanying degrees that depend on_their_atamio- _
number and mass-and on the mean neutron‘energy of the reactor. A sec-
ond source of poison is the even-numbered isotopes of uranium, which.
are not fissioneble. A certain emount of U2 is fed, along with the
177, even in highly enriched uranium, and ‘as the U727, with its high
neutron absorption cross sectiofi,»is burned out, the percentage of
U238 rises. Simllarly, a certain fraction, a, of the captures in the
fissionable isotopes results in radiative ‘capture of the neutrons, and,
instead of fissioning, the next higher uranium isotope (UEBLL 256)
is formed. It is necessary to examine the rates and extent to which
these updesirable constituents build up so that changes in neutron

or U

economy may be understood and so that desirable chemical reprocessing
rates may be determined.

1. BUILIUP OF EVEN-MASS-NUMBER URANIUM ISOTOPES

The buildup of U222, 12, 1220, and UP>® as nonfissionsble iso-

topic diluents in U233 and U235 is, as stated above, significant in |
fuel cycle economics. Although_U232 does not bulld up enough to af-
fect the neutron balence significantly, its hard-gamma-emitting
daughters are produced fast enough to be a biological hazard in the
handling of U255 and thus adversely affect the resale value of the
U253.l It has been assumed that & molten-salt reactor will process

 

1. A. T. Gresky and E. D; Arnold, Products Produced in Continuous Neutron
Irradietion of Thorium, ORNL-1817 (Feb. 6, 1956).

 

- 220 -
 

and burn all the U23 > it produces, hence nhe 0232 problem has not been
considered in detail.- :

| Rediative captures in Pa.25 > and. Ua5 > lead to isotopic contamina-
tion of the U= wi‘bh'Uzsl} . With no processing to separate these |
isotopes, end none seems feasible , the 0231} builds up until it is be-
ing produced and burned at the same rate. Based on cross sections
for neutrons at a velocity of 2200 m/sec, as taken from BNL-325
(cf 25 532, a 23 = 0.10, o ih - 92), & thermal reactor at steady state
would have ~ 58% as much U234 as U233 with 025 capturing ~9% as
many neutrons as 0253 In the epithermal mol‘ben-salt reactor de-
scribed in part 1 and hereafter called the reference-design reactor
(a 23 ny_ 0.16, o, 25 / o’eu. ~s 4.67), the steady-state 023lP concentra-
tion is ~75% of the U255 concentration, with U23 absorptions equal

to ~14% of the 023 3 absorptions. Neutron capture in 02 produces
fissionable 0255 but capture in thorium would be preferable ) since
U25 5 is a better fuel than 0255

Rediative capture in ye?? yields 1125 6 which may be considered
~to be an isotopic poison, since neutron absorption in U23 6 effective~
ly ylelds Np 21 instea.d of a fissionable isotope. In a thermal reac-
tor (o7 = 0. 19, a/ = 582, 5-26 s Te5), the v22% yould build up
until :I.t was present to the exbent. of ~15 times the amount of 023 >
with v 6 capturing ~16% as many neutrons as 2, Normally, in
any actual thermal reactor, resonence captures in U25 6 will reduce
the steady-state ratio of U0 to 0235 to less then 15. In the epi-
thermal reference-design rea.c'bor (a o 0.k, '0’25/“‘ 26_ ~ 1.3),
the U‘?5 would build up only to NSO% of the 1123 5 at steady state,
but ab that point the 1323 6 would be capturing N}O% a.s many neutrons
as U 3 5 . Isotopic separation of Ua3 5 and 023 6 may be feasible in a
power rea.ctor economy beca.use of the 1arge amounts involved and be-

 cause’ 11; is importan-b in a. breeder-converter economy. Separation in

 

a sepa.ra.te ca.sca.d.e would cost at least nine times as much as separa.-
tion of 6235 and 02 ) bu‘b by feeding the U 25 5 0256 mixbure into

- 221 -

 
 

 

 

existing cascades (either by adding top stages or accepting lower pro-
duction rates) less expens.ive processing probably could be achieved.
A study of the gaseous-diffusion problem is being made 3 and an . analysis

2,3

of its bearing on the nuclear fuel cycle has been reported At present 3

the government buy-back prices for 1.123 > include the same pena.lty for di-

- lution with U23 b and. U23 as for U258 It has been assumed in the ref-

erenee-design reactor that U23 build.up must be tolerated. . ‘l‘his is.
probably a realistic assumption, since the fuel becomes a mixture of
22 3 %1556, 8. The assumption would be pessimistic, however, for
another type of feasible molten-salt reactor which would burn U23 > in
the core and which would make 0233 only in the blanket that would be
sold externally at a premium price for another molten-salt reactor
which would burn only U 235 » since in such a case the IJ23 > ~could be
"traded in“ on fresh diffusion plant materia.l when the U25 content

warranted.

For a steady-state reactor operating with highly enriched 023 5
feed (e.g., 93% 0255 6% 0238 1% U23h) without any isotopic reprocess-
ing, the U2‘7’8 at steady-state will capture 6/93 as meny neutrons as the
U23 > fed. (as distinguished from the U 235 ‘ouilt up from U 235 via U25h)
In a thermal reactor (o’ 25 691+ o, 28 _ 2.73) at steady state there
yould be ~16 times as mich 112 as 1255, 1n actual thermal reactors
the U238 to-U 255 ratic would not get this la.rge because of resonance
captures in 0238 . In the reference-design molten salt reactor
(I 5"25/5’2 o 1. 5), the U238 can build up only to Jlo% of the U235
fed (thus in the molten-saltreactor, = is a worse isotopic con-
tamihant than "X 4n ‘amount, number of neutrons captured, and in
being a poison rather then a fertile ma.teria.l_). For the reference-design

 

2. E. D. Arnold, Effect of Recycle of Uranium Through Reactor and
" Gaseous Diffusion Plant on Buildup of Important Transmutation
Products in Irradiated Power Reactor Fuels, ORNL~210k (August 21,
1956).
3« J. O. Blomeke, The Buildup of Heavy Isotopes During Thema.l Neutron
Irradiation of Uranium Reactor Fuels, 0RNL-2126 (Jan. 11, 1957)

- 222 -

 
 

 

 

: 'In the referenee-design molten—sal'b react.or ebou‘b l% of ’the Pa

molten-salt reactor 1t was assumed that 0238 buildup would have to be
accepted. The rea.lism or pessimism of thie e.ssumption is about the
seme &s discussed for IJE5 in the preceding paragraph

2. PROTACTINIUM AND NEPTUNIUM POISONING

Neutron capture in pa’dd or Np239' has the same result as nonfission

capture in 1123 5 or Pu 59 i.e., & fiseiona:ble atom is effectively lost,
as well as & neutron. Although neutron loss to Np 231 does not involve
loss of & fissionsble atom, this loss can be more important than losses
to Pa®>> end Np-~” in reactors fed with highly enriched U-~”, Although
neutron capture by eny of these three isotopes yields a fertile atom,
at present prices for fertlile and fissile materials the gain is negli-
gible compared with the loss.

The average ratio of neutron captures to beta decays by 1’&'.25 3 in a
reactor is given to a good approzd.mation by:
| — (Pa)

OO’-I-6P(1+O.’)—('T -—(—-T'

where
P = reactor power level, Mw (thermal),
(1 + @) = average ratio of absorptions to fissioms in
fissile material,
R = regeneration ratio, |
M('Ih) - mass of thorium in syetem, K&

’I'he P a.nd o refer to 't.he whole syetem 'I'he other parameters can refer

M

'ei'bher to the whole system or 'bo the :r:‘uel e.nd ‘bla.nket sys'hems separately.

233

captures

a neu'bron before it can d.ecay to 0255

In a U25 3 025 5 breed.er-eonverl:er reaetor w:l.th highly enriched 1125 5

| 3‘me.keup Np239 poieoning is rela.tively unimporta.nt, ‘but, i the breeding-
convereion ra.tio ie poor, the Npe 7 poisen:lng ca.n 'become q_ui“be high 1f
'i'b is hot removed by chemical 'proceseing. 'I‘his :Ls especially 'brue in

resonence reectors 5 n which U"25 (en& henee N;p 37, ‘ot stea.dy state)

- 223 =

 
 

 

 

 

 

yields may be twice as high as in thermel reactors. In the reference-
design molten-salt reactor, processed at the rate of once per year, |
Np257 polsoning is zero.initially;:dbout'ogs% after one year, about
2% after 20 years, and sbout 2.5% at steady state.

3. FISSION PRODUCT POISONING

A 600-Mw reactor operating at a load factor of 0.80 will produce
about 183 kg/yr of fission products. About 22 atom percent of these
fission products have decay chains such that they appear as krypton
or xenon isotopes with half-lives of 78 min or more and thus are sub-
Ject to physical removal from a molten-salt reactor as rare gases by
purging with helium or nitrogen. These "removable" fission products
contribfite'dbout 26% of the total'fission broduct poisoning at 100 ev.
(Eor half-lives of 3 min or more, the yield'and poisoning percentages
-~ are 30 and 31, respectively. For half-lives of 1 sec or more the
velues are Ll and 38%, respectively.) The comparable poisoning per-
centage in a thermal reactor is much higher because of ‘the very large
thermal neutron absorption cross sectig? of Xe 35.' In a thermal re-

actor, however, burnout limits the Xe poisoning to & meximum of
about 5%, while in a resonance reactor ad jacent nuclei do not have
greatly differing cross sections, and burn-out is relatively ineffec-
tive in limiting total poisoning. Thus, to a first approximation, in
resonance reactors poilsoning increases almost linearly with time if

fission products are not removed.

About 26 atom pércent of the long-lived fission products are
rare earths. At 100 ev they contribute sbout 40% of the total fis-
sion product poisoning. The remaining ru52% of the fission products
contributes ~hl% of poisoning at 100 ev and comprises a wide variety
of elements, no one of which is outstandlng from the nuclear poisoning
point of view.

In a thermal-neutron U 255 burner reactor the fission product
poisoning, = /2 ,js appro:dmately eq_ual to the equilibrium Xel3 2
| poisoning (0-5%, depending on flux level) plus the equilibrium.Sm?hg

- 224 -

 
 

.poisoning (o1, 2%) rlus the ‘contribution from ell other fission pro-
~ducts. From data.presented in ORNL-2127 (ref. 4) for thermal 0255

burners operated at constant power and constant U‘?35 inventory, with
no fission product removel, the poisoning from “all other fission
products” is calculated to be ~3% at 100% burnup (i.e., when the
total amount of U235'burnea is equal to the "2 inventory), ~19%
at 1000% burnup, and ~51% at 10,000% burnup. - Thus it is possible,
although not necesserily economicel, to run & thermel, fluid-fueled
reactor for many years without processing thé fuel to remove fission
product poisons. The penalties for not processing would'be higher
U255 inventory charges and lower breeding-conversion ratios. At &
load factor of 0. 80, a 600-Mw thermal reactor burns sbout 218 kg of

35 per year (183 kg fissioned, 35 kg converted into U236), and
therefore with a L36-kg U-35 inventory the fission product poisoning
would incresse from O to 5% initislly end then to 20 to 25% after
20 years.

| Even in thermal reactors, resonance captures in fission products
make the poisoning somewhat worse than the nubers given above. The
megnitude of the extre poisoning depends on the ratio of the neutron
flux at resonance energies to the thermal neutron flux, which is de-
termined in part by the effectiveness of the moderator. In resonance
reactors, the flssion product poisoning is considersbly worse than in
thermal reactors because of the higher average fission product ebsorp-
tion cross—section relative to U235 " At & load factor of 0.80, a 600-Mw
reactor using 100-ev neutrons would burn about 275 kg of 0235 per yeer

— (183 kg fissioned, 92 kg converted into U256 - In such e reactor with
. an inventory of 550 kg of U25 the fission product poisoning would
'increase approximately 1inearly from zero, initially, to AJ52% after

2 years. S

 

-.EQE'J. 0. Blomeke and.M. F. Todd. Uraniumea'"J 

 

| Fission-Product Production
&8 & Function of Thermal—NEutron‘Elux, -Trradletion Time, and Decay.
- Time, ORNL-2127 (Aug.:.19, 1957). |

   

. 225 -

 
 

For U233-fueled reactors, the fission product poisoning is about
the seme as for U‘?35 in thermal reactors, but in the resonance region
| the higher U233 cross section reduces the poisoning effect‘by & factor
of 2 compared with U255. Thus & lOO-eV'breeder-converter burning half-"
end-half U235 and U235 would have a fission product poison level of
~6% if the fuel were processed at the rate of twice per (100%) burnup.

 

The reference-design molten-salt reactor has & medien fission |
energy of n10 ev, with ~10% of the fissions at thermal energles. It
may be considered, to a first approximation, that sbout one-third of
| the fissions are at thermal energy and sbout two-thirds are at an ener-

 

gy of 100 ev, for comparison'with the analyses presented above. At a
losd factor of 0.80, the 600-Mw resctor will bumn 105 kg of U and
n:125 kg of ye? per year, and it will produce ~ 183 kg of figsion pro-
ducts, m13 kg of u2 , and m3h kg of 11236 It should be processed at
a rate of three to four times per 100% burnup, and the total fission
product poisoning will be 6 to 8%.

L, CORROSION%PRODUCT POISONING

Chemical analyses of fuel mixtures circulated in INOR-8 and Inconel
loops have indicated that the principal corrosion-product poisons will
be the fluorides of chromium, iron, and nickel. These are relatively
light elements and,.per atom, their capture cross sections in the reson-
ance region are lower than those of the fission products. Further, ex-
trapolations of short-time tests indicate that the concentration of the
corrosion products will be much lower than that of the fission products.
Corrosion produot poisoning'has therefore been neglected.

5. METHODS FOR GHEMICAL PROCESSING

The "ideel" reactor chemical processing scheme would remove fission
products, corrosion products, Pa233,rnp237, and Np 239 as soon as they
were formed. After the Pa 255 and the Np239 had decayed to U233 and Pu 59
they would be returned to the reactor {or sold), along with the urenium
and plutonium that would also be recovered in the process. This ideal

chemical plant would have low cepital and operating costs, would hold

- 226 ~

 

 
 

 

up only smell amounts of fissioneble eand other high-priced materials,
‘and would discharge its waste streams in forms that could be inex-
pensively disposed of or, possibly, sold as by—prodnctssrlPresent
technology, however, . does not offer such an ideal pro cess for sny

reactor.

More practical goals for processing a molten-galt reactor are
(1) continuous removel of most of the gaseous fission products by
purging the fuel with helium or nitrogen ges; (2) an in-line removal
of rare earth, noble metel, or other. fission products by freezing-out
part of the selt streem, pleting out fission products on metallic
surfaces (either naturelly or electrolyticelly), exchenging the reare
earths for Zcerium, or scavenging by contacting the salt with & solid
~such as BeO to remove certa.in constituents of the salt by edsorption :
or exnhenge, &nd (3) continuous or batch removal of the salt from
the resctor &t an cconomically qptimum.rate to seperate the uranium,
plutonium, end selt from the remaining fission products and corrosion_
products by the least expensive method availeble. Present technology
does not make all these methods immediately eveildble, but there is
. reason to expect that continuation of the current development program
would make them eveilsble in the nineteen~sixties. ’

Operation of the Aircraft Reector Experiment end of molten-selt
" in~-pile loops have indicated that gaseous fission product removal
can be achieved a.n.d theat Ru, Rh, and P4 plate out on metal surfaces.
Provisions for degassing nre included in the molteneselt reactor,
but et present the possible reduction in fission product poisoning

- as a result of plating out in the heat exnhengers and the possible
/freduction of corrosion.by formation of & protective surface sre not

_f“;being considered in economic studies. Similarly, possible increases
“in fission prodnct poisoning as e result of the pdating out of noble

 

- metals in the core ere nnt being teken into eccount.

| Mnthods for removing-fission products from.the selt so that the
selt can be reused are being inwestigeted in current research and

- 227 =

 
 

 

 

development programs, but for the present 'mol‘beh—salt rea;ctoif s'tlidy'
it is assumed that only the uranium will be recovered from the salt
by the ORNL fluoride volatility process and that the salt will be
stored for future recovery of thorium and lithium. Adequate tech-
nology already exists for the preparation of fresh fuel stafting
with nonradioactive UF6 and fluoride salts, and methods for remote
reconstitution of fuel from recycled uranium (snd recycled salt,

1f possible) are being developed.

5+l.  The Fluoride Volatility Process

A program of development and pilot-plant demonstration of a
fluoride volatflity process for recovering uranium from molten
fluoride salts by oxidation with F2 to form UF6 is currently under |
wey. Recovery of the uranium in the slightly radioactive ARE fuel
(Na.F-ZrFu—UF,_l_) was completed in February 1958., and a demonstra‘bion» )
of the recovery of uranium from a highly radioactive STR fuel ele=-
ment (which, first, will Be_di’ssolved in NeF-ZrF) in the presence
of HF) is scheduled for fiscal year 1959. A blo"ck flow sheet of
the process as adapted for molten-salt reactor fuel i'eprocéssing
is shown in Fig. 6.1. The uranium-bearing molten salt is trans-
‘ferred to the fluorination vessel in batches. Fluorine, diluted
‘with N , is bubbled through the salt st 450°C until its U content

2
is reduced to ~10 ppm. The UF6, N, and excess F,. pass out of the

fluorinator through a 100°C NaF pellet bed, wh:i.charemoves the UF
from the gas stream. In the pilot plant the excess F2 is disposed
of by scrubbing it with a reducing KOH solution, but in a production
Plant the fluorine might well be recycled to the fluorination step.
The UF6 is desorbed from the NaF bed by raising the temperature of
the bed to LOO°C and sweeping it with more F,-N,. The UF, then

- passes through a second NaF bed and finally is collected in cold
traps at -40 to -60°C. The volatility plent product is ligquid UF
obtained by isolating the cold traps from the system, heating to

gbove the triple point, and draining into the product receiver.

228 -
 

1 in i it et b b

 

6¢¢

 

 

 

 

 

 

 

 

 

 

 

 

 

Ru
Cold
Trop

 

 

 

 

 

 

UFg
=1 Cold
Trap

 

 

 

 

 

 

 

- -
R . (Desorption)
LiF-BeF,(+U,Th,FP) N
from reactor -~ © . 0 L
Fluorination - | F,
Hold-up Vessel” o
Vessel : -
' ~450°C
(+Th, >99%FP)

~ to salt recovery,
storage, or waste

- disposal.

 

UFg
= . Product

 

 

NaF NaF
Bed . Bed

 

 

 

 

 

 

 

 

- NaF Waste

 

Absorption on first

bed at 100°C.

Desorption through

both beds at 100-400°C.

——— F, Disposal Ha

 

 

 

 

UFg — UF, Reactor

 

 

 

 

 

=

 

 

 

  

 

 

Powder Removal System I

 

 

 

1

URs
Product

Fige 6.1.  Proposed Uranium Recovery Flowsheet for Molten Salt Power Reactor

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

—
| !
| |
| |
| }
| i
1 | | Y
L__J
b
Sintered
Metal
‘Fiiters
Ha,F2
' Dns;iosal
Ckhemical
Trap

 

 
 

 

- Most of the decontemination in the voletility process is achieved

in the fluorination step, since most of the fission end corrosion pro=-
ducts remain in the salt. The volatile conteminants (Xe, Kr, I, Te, Mo,
most of the Ru, end part of the Nb end Zr) either pass through the NeF
‘bed while the UF¢ is retained or remein on the bed when the UF6 is de-
sorbed (Nb, Zr). The I, Te, Mo and Ru are removed almost entirely by a
cold trap, and the remainder is scrubbed out of the gas system along
fl_with the excess F2 The Xe and Kr follow the N to the plant off-gas- |

system. The Nb end Zr slowly'build up on the NeF bed, which is replaced_
vhen poisoned. Replacement is infrequent, however, because a micrometal-
lic nickel filter between the fluorinator end NaF bed removes most of the
Nb and Zr.

| The ARE fuel was processed in batches of 1.k ft3 of salt, each batch
containing w10 kg of U-2” (the capaclty of the NaF beds), at the rate of
two batches per week. The not-economically-recoverable uranium losses

in processing the ARE fuel were approximately 0.015% in the waste salt
and O. 075% on the NeF beds.

The reference-design molten-galt reactor includes a volatility plant
approximately the same size as the ORNL pilot plant. About the seame
uranium processing rate is planned (two 10-kg batches per week through
the NeF beds), but & higher salt throughput rate (2 £t°/day each of
fuel and blanket fluid) will be used because of the lower uranium con-
centration in the salt (compared with the ARE salt). This higher salt
throughput rate is within the capabilitlies of the present ORNL pilot
plant, which could fluorinate one 1.k ft3 batch per shift, if necessary.
For the molten-salt reactor processing plent, separate fluorinators of
gbout 2 £t5 cepecity each for the fuel and blanket salts will preveht
Cross contaminetion and require only one fluorinetion per day for each.

The possibility-of continuous fluorinstion and the consequent re-
‘duction in the size of the equipment are being considered. In addition,
information needed in the adasptation of the process to the particular
fuel and blanket salts to be used in the molten-selt reactor is being

- 230 -
 

 

obtained » @nd simpler means for reconstituting the reactor feed mater-
lals are being studied.

5.2. K-25 Process for Reduction of UFg to UFy

 

The continuous, reduétion of highly' enriched. UF6 to _UFM_ is__wgli
proved as a nonradioactive proc:eesss.5 The process, used and developed |
at K-25, is indicated in Fig. 6.13 the UF product from the volatility
process is the feed mate:ial. The reduction takes place in a U"E'6-F2-H2
flame in a Y-ghaped reactor. The Fa is added to give the proper fleme
temperature. The reaction products are UF,+ and I-IF-HQ gas. Micrometal-
lic filters are used to recover eny UFI;. wvhich may be entrained in the
exlt gas. A vibrator 1s used to shake free any UFh which clings to
the filter or tower walls. A chemicel trap with a t',‘z?:..'i'xclL or an NaF
pellet bed 1s used to recover any unreacted UF6 in the exit gases,
although the amount so collected is negligibly small in normal oper-
ation. The HF in the exit gas is either scrubbed with a KOH solution
spray or sorbed on an NaF bed. |

This process has not been used a‘b 8 high level of radioactivity.

,S-:ane the UF6 from the molten-salt reactor volatility plant will be

somewhat radiocactive » the major actiflty probebly being II23 T , & pilot
plant demonstration will be required, but no serious difficulties are
anticipated. For the molten-salt reactor chemicel plant, Vequipment

of the same size as that used in the ‘K—-25 facility is assumed. A fa-
cility of this Size would have excess capacity on a continuous basis,
but it is essumed that it would be operated intermittently, probebly
on ;thé once-or-twice-a-week schedule 'fv'.'sed.' for the discharging of the

UF from the voletility plent NeF beds.

| 5.3. -Salt Recovery by Dissolution in Concentrated HF

. Leboratory work ha.s"ti:éefi initiated on & process for. recavering LiF

 

5. §. H. Smiley and D. C. Brater, "Conversion of Uranium Hexafluoride

. to Uranium Tetrafluoride," Chepter in "Process Chemistry, Vol. II,"
’ Prog‘resx)s in Nuclear Energy Serieés, Pergenmon Press (to be published

- 231 -

 
 

 

 

and BeF from contaminated molten~-salt reactor fuel by dissolution in
| concentrated hydrofluoric acid (>90% HF, balance H 0) The urenivm

would first be removed by the volatility process. Thorium, corro:s_ion
products, and most of the fission products are insoluble in the sol-

vent and would be separated by filtration » centrifugation, or other.

solid-liquid separa.tion methods. | The proposed flowsheet is shown in‘_' |

Fig. 6.2.

Lithilm'_fluoride is quite. soluble in anhydrous HF end quite in-

soluble in H20. By itself, BeF is Just the opposite, elthough its

solubility in HF is significantly increased by first saturating the

HF with LiF. A small emount of water in the: EE‘ also lncreases the
2

elements end some fission products, including the rare earths, are

insoluble in both solvents.

 The solubility of the 63 mole % LiF-37 mole % BeF fuel carrier
salt mixture is greater than 100 g/kg in ~90% HF. In an experiment
in iwhich mixed fission products (rare earths, strontium, and c.es.iunjl)
were added to the salt mixture and the mixture was dissolved and
filtered, the salt was decontaminated from rare earths by & factor
of ~1000; that is, the activity remaining in the salt was 1/1000
of the activity of the mixture prior to dissolution. No decontami-

nation from cesium was obtained. The results for strontium were inecon-

clusive, but the decontamination_ appeared to be intermedisate between
that from rare earths and that from cesium.-

In further studies, fission product and heavy element solubili-~

ties will be investigated, as well as possible hydrolysis of the
s’_alt by reaction of the salt with water in the solvent. | It may'be_

found possible to avoid hydrolysis of the salt byusing aphydrous HF

or to overcome it by trea.ting the reclaimed. salt with HF.

' 5 h Rare Earth Removal by Exchange with Cerium

A study of the removel of high cross-section rare earths from

- molten-salt reactor fuels by exchenging them for a low cross-section

- 232

BeF solubility markedly. Fluorides end oxides of most of the heavy -

. -
 

 

 

 

UNCLASSIFIED
ORNL LR DWG. 28749

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.........--.-.—.——---UF’6
LiF- BeF, Sait Fluorinator
U,Th,FP ~450°C
from Reactor
Salt HF,H,0 vapors _ Solvent
Th,FP] Condenser
>90% HF, <10% H20
Dissolver
32°C
~ 0% Salt Sall
Solution Solid Solution Fiash
in HF-H, 0, Liquid Evaporator
+ Th,FP Solids Separotor
100-400°C
Continuous e e - =
Batch -
Th,FP Solids Molten
- Some Salt & Salt
Solvent
HF H, UF,, ThF,
HF
H,O
o _ » Fuel
~\éVcste ’r - Make-up a
_ ~vapord or. Purification
" Th Recovery . - LiF-BeFa-ThFa-UFs
: : = to Reoctpr

Fig. 6.2. Proposed Salt Reclamation Flowsheet Based on Dissolution in 90+%, HF

233

 
 

 

 

rare earth, possibly cerium, has also been initiated. The process
might be carried out in either of two ways. First, the salt might
be saturated with ~1 mole % cerium at an elevated temperature, end
then cooled to within ~30°C of the liguidus temperature of the pure
selt, at which temperature the rare earth golubility is only 0.2,
Thus ~80% of the cerium would precipitate and carry with it ~80%
of the fission product rare earths. After a solid-liquid separation,
~this process could be repeated, if desired, to give high-percentage
removal of fission product rare earfihs. Second, & perhaps better
way to'accomplish the same end wnuld'be to cool the core salt to
near its liquidus temperature and éontact it with solid CeFj, prob-
ably in a columnar bed of pellets. In principle, the fission pro-
duct rare earths could.be‘exchanged'for cérium to ahy desired extent
in this menner, depending on the ratio of salt to cerium used, ‘the
pellét size (determining the surface area), and the contact time.

The attractiveness of the éxchange process is enhanced by the
low ebsorption cross section of cerium compared with the sbsorption
cross sections of most fission product rare earths. This potential. -
advantage is reduced somewhat for e reasonsble processing rate, in
that, effectively, several cerium atoms are exchanged for one fission
product iare earth atom, since the ceriufi solubility is ~ 0.2 mole %
near the liquidus temperature, whereas the fission-product rare earth
concentration in the core salt is only ~0.04k% for a processing rate
of once per year (and proportionately less for shorter processing
perlods, as would be desirable given an economical processing method).

In the reference-design moltenésalt reactor the advantage is :educed'

further by the fact that for resonance-energy neutrons the cerium

cross section is not as much lower than the cross sections of the

other rare earths as it is at thermal energies (see Table 6,1). Thus

‘&t 100 ev the poisoning due to 0.2% cerium would be equal to that from | .
| fissiOnQProduct'rare earths for & processing rate of once per six |
months, iQe., the total would be equivalent to fission-product rare .
_earfih roisoning for a once-per-year yrccessing rate by the volatility

processe . ' - . | W
- 234 -
 

Table 6!1-.

Abgorption Crose Sectlons of Various Isotopes

at Seversl Energies*

 

Abgsorption Cross Section {barns)

 

 

 

Fission

Isotope Yield (%) At 0.025 ev At 100 ev’ At 25 kev
1t 6.6 8.4 10.9 0.050
gl 0 6.5 0.63 0.61 0.021
prtil 6.4 11.2 14,5 0547
c:ell"2 6.2 1.0 0.64 0425
mlb 5.9 280.0 12.9
Ndw* 5.1 4.5 7.9
Nalhfi 4.0 52,0 22.8
N’dll‘s 2.1 9.2 7.8
T 2.3 60.0 35.8
su T 0.09 R
.Ndll‘l's 1.8 5.2 39
smt*9 1.h 66,000.0 48,6
N+ 0.7 2.8 2.8
smt?t 0.5 10,000.0 5Lk
Sm-2 0.3 140.0 26.2 0.860
B o> 0.1 420.0 90.2
e 0.1 5.5 20.0 0,465
adt?”: 0.06 70,000.0 54.6
el 0,05 | 35,7
el 0.02  160,000.0 62.5
'Weigh‘bed average of S | -
ebove 2,100.0 12.0 0.5

| ”o.'—é'r | 0.61 - 0.069

Na.tura.‘l. Ce

 

#0,025-ev cross sections from BNL-BSES e.nd its Supplement No. J., yields
and 10Q~ev cross sections from P. Greebler, H, Burwitz, end M. L. Stom,
Nuclear Sci. and Eng., 2, 334 (1957); 25-kev cross sect:lons from R. L.
Macklin, ORNL 3 personal connnuni.ca:bion. -

- 235 -

 
 

e et T R L i

 

e e el kbl et b om it

| ‘The cerium exchange processing method would rasise the liq_uidus
temperature of the core sa.lt by 30 c or more , end remove plutonium :

and trivalent urenium slong with the rare earths. These disadvantages

* do not e.ppear to be serious at present, and the potentia.l advantages .

7- of this processing method for thermal (gra.phite-moderated) molten-sa.lt |
rea.ctors would seem to justify continued development work.

- 5.5. Radioa.ctive Ws.ste Disposal

At 8 load factor of 0.80, 600-Mw reactor will produce ~183. kg/y'r
of fission products, About 23 wt % of the fission products cen be re-
moved as Xe-Kr gases, but the remaining ~1hl kg/yr must be removed by
chemical processing The chemical processing waste streams include

fused salt 5 NeF pellets ’ 2 N2 s and I-IF--H2 gases, Most of the nongaseous

: fission products rema.in in the fuel salt residue a.fter fluorination and.
ms.y be stored in this form. Most of the rema.ining fission products are

removed by periodically flushing out the micrometallic filter between
the fluorinator and the NeF bed and the cold trap between the NaF bed
and the F, disposal unit. The NaF bed is replaced infrequently, if and
vhen it becomes poisoned with niobium and zirconium. Any remaining
fission products in the gas streams are scrubbed out with the F2 and
HF, or vented to the reactor off-gas system.

For optimum costs, high-power molten-salt reactors should have
moderately high inventories of enriched uranium and the fuel should
be proces_sed about tiwlce per fuel-inventory burnup, & compromise be- |
tween the rate-proportional cost of processing (assumed e,t present to
be equsl to the cost of buying new salt) and the savings due to im-

'proved regenera.tion ratio made possible by processing If a fission- |
-~ ‘eble ms.terial inventory of 600. kg or more and a fuel ‘processing rate
B V'of once per year or greater are assumed and it is considered that

the fuel selt will be discarded after its uranium is removed by the
voletility process , the wa.ste salt volume amounts to 600 f't5 per |
year or more (sbout 1 ft /Mw-year, or 1 £t / kg of fissionsble mater-

ia.l processed, or 3 to 4k £t /kg of total fission products) - This

- 236 -
 

 

 

volume is quite 1ow in comparison with the wastes of other power
reac'bors s for example, STR wastes are ~90 f‘b3 /kg of U23 2 processed
in the present agueous process and ~3 ft3 /kg of U23 > in the proposed
volatility process, and the figures are gbout 4 times larger on &
cubic foot per kilogram of fission products basis. |

- 237 =

 
 

 

.

- -

\O 03 OV Bl o -

—
o

11.

13.
1h,
15,

16,

17.
18.
19.
20,

22.
23,
24,
25,
26.

. D. S. Billington
. F. F. Blankenship

E. P. Blizard

. A. L. Boch

C. J. Borkowski

G. E. Boyd

E. J. Breeding

R. B. Briggs

C. E. Center (K-25)
R. A. Charpie

F, L. Culler

L. B. Emlet (X-25)
D. E. Ferguson
A. P. Fraas

J. H. Frye, dr.
W. R. Grimes
E. Guth

C. S. Harrill
H. W. Hoffman

A. Hollaender
A

. S, Householder
W. H. Jordan
G. W. Keilholtz
M. T. Kelley
J. A. Lane
R. S. Livingston

ORNL-263%
Reactors-Pover
TID=-4500 (1h4th ed.)

INTERNAL DISTRIBUTION

27‘.
28,
29.
30.
31.
32.
33-
3k,
35,
- 36.
37.
38.
39.
Lo,
Ly,
ho,
k3.
L,
k5.
ke,

h-48.
49-60.

61.
62-63.

H. G. MacPherson

W. D. Manly

J. R. McNally

K. Z. Morgan

J. P. Murray (Y-12)

M. L. Nelson

A. M. Perry

H. W. Savage

A. W. Savolainen

H. E. Seagren

E. D, Shipley

M. J. Skinner

A. H. Snell

J. A. Swartout

E. H. Taylor

A. M. Weinberg

C. E. Winters

Biology Library

Health Physics Library
Reactor Experimental
Engineering Library

Central Research Library
Laboratory Records Department
Laboratory Records, ORNL R.C.
ORNL - Y-12 Technical Library,
Document Reference Section

EX‘I'ERNAL DISTRIBUTION

6)4- Division of Research and Development, AEC, ORO
65-5T1. Given distribution a8’ shown in TID-hSOO (14th ed.) under Reactors-Power

" category

..239..