ORNL-CF-56-8-204.txt

 

tla 5

 
  

(CF-56-8-204(Del.) |

- UNCLASSIFIED |

REAcToRs-powm .
4 || uwitep staTEs ATOMIC ENERGY COMMISSION |

| -FUSED SALT FAST BREEDER A i
g 'Reactor Design a.nd Feasibility Study |

By L
- J.J. Bulmer ', e
~ E. H. Gift -~ o .
S R.JLHOW
A. M. Jacobs T e
E Kofiman LT e e
L e -'_j_R.L McVean
| ey T RA.Rossi

 

i
%
. fi“""flfi . - 4 "
VoL e .
ot Ceau e
ot i

 

: -“August 1956 |
'_QOak Ridge School of Reactor Technology
‘ f.f"_;Oa.k Ridge, Tennessee L

 

 

 

 “Technical Information Service Extension, Ok Ridge, Tenn. |

L
L

 

 

 

 

 

 
 

T TR g

 

 

 

 

© Date Declassified: March 6,405%. . .. . . .

f

I.EGAL NOTICE

= ‘This report was prepared as an sccount of Government spomsored work. Neither the -
- United States, nor the Commission, nor any person acting on behal oi the Commission.

_ " A. Makes any warranty or representation, express or implied with respect tothe - |, C L
accuracy, completeness, or usefulness of the information contained in this report, or that | - L
" the use of any information, apparatus, method, or process disclosed in this report may

not infringe privately owned rights; or : )
e B. Assumes any liabilities with respect to the use oi, or for damages resulting from_ 1

the use of any information, apparatus, method, or process disclosed in this report. L) BT \ o

o -As used in the above, “person acting on beha.lf of the Commission includes any em-

: ployee or contractor of the Commission to the extent that such employee or contractor .

~ prepares, handles or distributes, or provides access to, any information pursuant to his
Aemployment or contract with the Commission. . o '

 

 

 

 

 

| This report has been reproduced directly irom the best available cOpy. c - -

~ Issuance of this document does not constitute authority for declassification of
Ce classified material oi the same or similar content and title by the same au-_ o
L thors._ -

; Since nontechnical and nonessentiai prefatory material ha.s been deleted, the .
L first page of the report is page 5. : :

L Printed in USA Price $i.00. Available irom the Oifice of Tecimical Services,
Department of Commerce, Wa.shington 25 D. C

 

AECTechnicaiin!o:maflonlefliceEmmion s
. Mlidgc. Tennessee
 

1

(3

v

CF-56-8-204(Del. )

OAK RIDGE SCHOOL OF REACTOR TECHNOLOGY

Reactor Design and Feasibility Study

"FUSED SALT FAST BREEDER"

Prepared by:

J+ J. Bulmer, Group Chairman
E. H. Gift

R. J. Holl

A. M. Jacobs

S. Jaye

E. Koffmaen

R. L. McVean

R. G. Oehl

R. A. Roesi

OAK RIDGE NATIONAIL IABORATORY
Operated by
Union Carbide Nuclear Company
O2k Ridge, Tennessee

' August 1956

 
 

 

 

 

 

 
O

e

9

 

FPREFACE

In September, 1955, a group of men experienced in various scientific
and engineering filelds embarked on the twelve months of study which culminated
in this report. For nine of those months, formal clessroom and student
leboratory work occupied their time. At the end of that period, these nine
students were presented with a problem in reactor design. They studied it for
ten weeks, the final period of the school term.

This is & summary report of their effort. It must be reslized that
in so short a time, a study of this scope can not be guaranteed complete or
free of error. This "thesis" 1s not offered as a polished engineering report,
but rather as a record of the work done by the group under the leadership of

"the group leader. It is issued for use by those persons competent to assess

the uncerteinties inherent in the results obtained in terms of the preciseness
of the technical dats and analytical methods employed in the study. In the
opinion of the students and faculty of ORSORT, the problem has served the
pedagogicel purpose for which it was intended.

The faculty Joins the authors in an expression of appreciation for
the generous assistance which various members of the QOak Ridge Nationsal
Laboratory gave. In particuler, the guldence of the group consultants,

A. M. Weinberg, R. A, Charple, and H. G. McPherson, is gratefully acknowledged.

Lewis Nelson
for

The Faculty of ORSORT

 
 

 

 

 

ACKNOWLEDGEMENT

We wish 'lto.»express our appreciation to 'our group advisors Dr. Alvin
Weinberg, Dr.' Robéft Charple and Dr. H.-_G. MacPherson for their constant
aid and helpful snggestions towvard the completion of this project. We
would especially like to thank Dr. C. J. Barton ,for our fused chloride

equilibrium data; Dr. Manly and his group for the fused chloride corrosion

tests; Dr. J. A, Lane and Mr. W. G. Stockdale for their sid in our economic

evaluation; The _Atomic Power Dévelopnent Agsociates for their muclear E.al-
culations; The Argonne National Lé.boratory for their UNIVAC codés; New |
York University and Nuclear Development Corporation of America for their
a:ld.in the UNIVAC operation; and Dr. E, R, Mann and Mr. F, Green fofi' their
assistance on the reactor simulator at the Oak Ridge National Laboratory:.
In addition we would like to tha:'flff the mAny other members of the Osk Ridge
National Laboratory at X-10 and Y~12 who gave freely of their time and
valuable suggestions.

Lastly, we wish to thank Dr. L. Nelson and the ORSORT faculty for

their continuing help and the remaining personnel of the Educational Division

for their efforts in the completion of the project.

-8~

0

<

 

 
 

TABLE OF CONTENTS

E; Preface

Acknowledgement

F) ]

Abstract
- Chapter I. Introduction
1.1 Problem
! l.1.1 Purpose

+}

1,1.2 Scope
1.2 Evaluation of Fused Salts

1.2.1 Advantages of Fused Salts

1.2.2 Dipadvantages of Fused Salts
_ 1.3 Results of Study
. | 1.4 Description of System

1.4.1 Core |

1.4.2' Blanket

1.4.3 Control

1.4., Chemical Processing

= Chapter 2. Preliminary Reactor Design Considerations

- | 2.1 Selection of Core Fuel

A 2.1.1 Criteria for Selection

. 2.,1.2 Fuel Properties

%

2.2 Selection of Blanket Material
2.2.1 Oriteria for Selection

~9-

Page No.

B

 
 

TABIE OF CONTEKNTS (CONT.)

2.3 Reactor Coolant System
2.3.1 Internal Cooling
2.3.2 External Cooling
2.4 Materials of Conmstruction
2.4.1 Core System
2.4.2 Blanket System
2.4.3 Reactor Components
Chapter 3. Engineering
3.1 Genersal
3.1.1 Engineering Properties of Fused Salt, Sodium
Coolant and Blanket Paste
3.2 Reactor Power
3.3 Design of Heat Transport System
3.3.1 Circulating Fuel Heat Exchanger
3.3.2 Circulating Fuel Piping and Pump
3.3.3 Blanket Heat Exchanger
3.3.4 Blanket Heat Removal
3.3.4.1 Parameter Study of Blanket Heat
Transfer System
3.3.5 Blanket Piping and Pump
3.3.6 Sodium Piping and Pumps
3.4 Salt Dump System
3.5 Core Vessel and Reflector Heating -

3.6 Moderator Cooling

-10-

Page No.
34

35
35
36
37
38
38

65

75

t)

G

T

 
v

N

f

i

 

_TABIE OF CONTENTS (CONT,)

3.7 Once~Thru Boiler

3.8 Auxiliery Cooling System

3.9 Turbo-Generator

Chapter 4. BHNuclear Considerations

4.1 Swmary of Study Intentions

4.2 Calculation Methode Based on Diffusién Theory
4.2.1 Bare Core Multigroup Method
4.2.2 Reflector Sevings Estimate
L.2.3 UNIVAC Calculations

L.3 Cross Sections
4.3.1 Energy Groups
4e3.2 Sources of Data |
4.3.3 Calculation of Capture Cross Sections
4e3.4 Tabnlation of Cross Sections

4Le4 Results of the Parameter Studies
belod Preliminary;Analyfiia

 4e4e2 Bare Core Tefi Group Parameter'8£udy-.
4.4.3 Reflector Control |
4uhok Effect of a Moderator Section in the

| Blanket Reglom = |

4.5 Final Design

Chapter 5. Controls | |
5.1 General Considerations

5.2 Delayed Neutrons

83
83
83

oL
91
92
ok
ok

96

EE B

 
5.3
5¢4
5.5
»5.6

TABIE OF CONTENTS (CONT.)

Tempereture Coefficient of Reactivity
Reflector Control
Simulator Studiles

Startup Procedure

Chapter 6. Chemical Processing

6.1

6.2

6.3

Process Flow Sheets
6.1.1 Core Frocessing
6.1.2 Blanket Processing
6.1.3 On-Site Fission-Product Removal
6.1.3.1 Off-GAs System
6.1.3.2 Precipitation of Fission Products
6.1.3.3 Distillation Removal of Fisslon Products
Consideration Leading to Process Selection |
6.2.1 Processes Considered
6.2.2 Process Selection
6.2.3 Purex Modifications for Core Processing
6.2.4 Alternate Blanket Process
Processing Cycle Times
6.3.1 Gemeral
6.3.2 Effect of Fission-Product and Transuranic
Buildup
6.3.3 Economics and Process Cycle Time Selectlion

6.3.3.2 Core FProcessing

-12-

Page No,

16

122
130
132
132
132
135
136
136
137
138
139
139

150
150
1h2
1%L
14k

a1kl
146

- 1h6

3

 

0
o

)

L

#h

 

TABIE OF CONTENTS (CONT.)

6.3.3.3 Blanket Processing
Chapter 7. Shielding
7.1 General Description
7.2 Reactor Shielding Galculaiions
7.2.1 Neutron Shielding
7.2.,2 Gamma Rey Shielding
Chapter 8. Economics
8.1 General
8.2 Capital Costs
8.3 Life of Equipment and Anmual Charges Due to capital Costs
8.3.1 Power Cost Due to Capital Cost
8.4 Fuei Inventory Charges
‘8.5 Processing Cost Sumary )
8.6 Credit for Ereeding
8.7 Operation and Mbintenance
‘8.8 Cost Summary
Ghaptef 9. Recommendations for Fufure Work
. 9.1 Gemeral |
9.2, Engineerifig N
.9.3 Materisals
| 944 Chemical Processing
9.5 Reactor Gbntrol
9.6 Economics
Appendix A. Engineering Calculetions

-13-

Page No.

149

153
153
154
15k
15k
158
158
159
160
160
161
162
162
163
163
16%

16
165

166 |
167
167
168
169

 
 

TABIE OF CONTENTS (CONT.)

A.1 Circulating Fuel Heat Exchafigef

; A.2 Circulating Fuel Piping and Pump

% A.3 Blanket Heat Removal Caloculations
A./ Blanket Heat Exchanger

A.5 Blanket Piping and Pump

A,6 Sodium Piping and Punps

A.7 Core Vessel and Reflector Heating
A.8 Moderator Cooling Calculations

A,9 Steam Boiler Caleulations

 

Apperdix B. Gamma-Ray Shielding Calculations
B,1 Sources of Gammas |
B.1l.1 Prompt Fission Gammas
B.1.2 QGission Product Gammas During Operation
B.1l.3 Capture Gammas

B.1.4 Imelastic Scattering Gammeas

B.2 Attenuation of Gamma Rays
Appendix C. Experimental Tests
C.1 Summary of Melting Point Tests
C.2 Petrographic Analysis of Salt Mixtures
C.3 Summary of Chemical Analysis of U’Gl3

0.4 Corrosion Tests

References

~1l-

189
196
196
196
196
197
197

EEE S S

212

)

 
e o 1 i e .

4

o

 

ABSTRACT

An externally cooled, fused salt, fast breeder reactor producing 700 M4

of heat has been désigned utilizing plutonium as the fuel in a mixture of

the chlorides of sodium, magnesium, uraniun and plutonium. Depleted uraniwm
is used as the fertile material in a blanket of wraniun oxide in sodium.

Nuclear calculations have been performed with the aid of the UNIVAC for
multi-group, malti-region problems to obtain an optimm muclear design of
the system with the chosen fused ssalt.

Steam temperature and pressure conditions at the turbine throttle have
been maintained such that the incorporation of a conventionzl turbine-generator
set into the system design is boasible.

An economic anslysis of the system, including estimafiéd chemical pro- |
cessing costs has been prepared; The analysis indiecates that the fused salt
system of this étudy has an excellent potential for meeting the challenge of
economlic nuclear power. |

It was not learned until the completion of the study of the severe (n,p)
cross section of the chlorine-35 isotope in the range of enmergies of in-
terest. This effect was smplified by the large number of chlorime atoms pre-
sent per atom of plutonium, The reéult was considered serious enongh to
legislate against the reactor. o

It was determired, however, that the chlorine~37 isotope hadré high

~ enough threshold for the (n,p) reaction so that it could be tolerated in

this reacter. The requirement for the chlorine-37 isotope necessitates an
isotope separation which is estimated to add 0.5 mils per kwhr. to the cost
of power. The power cost would then be 7.0 mils per kwhr. instead of the

6.5 mils per kwhr. reported.
-15-

 
 

CHAPTER I INTRODUCTION

1.1 FROBLEM

1.1.1 Purpose

The purpose of this study was to assess the technical and economic

Ifeasibility of a fast breeder-~power reactor, employing a fused salt fuel,

based on & reasonable estimate of the progress of the fused salt technology.
Fuel bearing fused salts are presently receiving consideration for high
temperature applications and in addition have been proposed as a;possible

solution to some of the difficult problems of the fast reactor.

1.1.2 Scope

A major consideration was an initial decision to devote the group
effort to & conceptual design of complete reactor system insteéd of con~- |
centrating on parameter studies of the reactor or the heat transfer and power
plant at the expense of the other components. This philosofihj necessitated
overlooking many small problems that would arise in the detailed design of
the reactor andipower plant but provided a perspective for evaluating the
techfiical and economic feasibility of the entire reactor éystem'instead of
only portions of it. | | N |

At the outsel of the study it was determined that a breeding ratio
significantly less than one would be ofitained from an interaa11y'éob1ed machine.
It was theréfore decided to further restrict the study tb an externaily_cooleq,
circulating fuel reactor in whicfl & breeding ratio of at least one was ob-

tainable,

=16~

t)
.

 
{»

&

 

1,2 EVALUATION OF FUSED SALTS
1.2.,1 Advantsges of Fused Salis

The fused salts enjoy practically all the advantages of the liquid

fueled, homogeneous type reactor. Among the more prominent of these are:

1.

2.

3.

be

The large negative tempersture coefficient which aids in
reactivity control; |

The elimination of expensive and difficult'toperform fuel
element fabrication procedures;

The simplified charging procedure uhich‘provides a means of
shim control by concentration charges;

The higher permissible fuel burn-up without the attendant

mechanicel difficulties experienced with solid fuel elements.

In eddition, the fused salts display a superiority over the aqueous

homogeneous reactor in these respects.

1.

‘2.

Lowsr operating pressure due to the much lower vapor pressure

of the fused salts;

Higher thermodynamic efficiency due to the operation at

higher temperature.

1.2.2 Disadvantages of Fused Salts

There ere several disadvantages which are attendant upon the use

1,

of fused salts for the application reported upon here. Of these, the most
prejudicial 4o the success of the reactor are:

The corrosion problem which is so severe that progress in
this application awaits development of suiteble resistant

materiels;

 
 

2. The lerge fuel inventory required because of the externsl
| fuel hold-up; | | | |

3. The poor heat transfer properties e:éhfbited' by the fused
salts;

k. fThe low specific powers obtainsble in the fused salt fast

reactor system compared to the equeous homogeneous reactors.

1,3 RESULIS OF STUDY |
“ The fingl design is & two rég:i.on rea_ctor with & fused salt core and a

uranium oxide powder in sodium blanket. The fuel domponent 13 plutonium
with a totael system mass of 1810 kg. The reactor has a total breeding ratio
of 1.09 exclusive of chemical processing losses.

The reactor produces 700 MW of heat and has & net electricél output of
260 M7. The net thermal efficiency of the system is 37.1 per cent. The steem
conditions at the turbine throttle are 1000°F and 2400 psi.

The cost of electrical power from this system was calculated to be 6.5
mlils per kwhr. This cost included & chemical processing cost of 0.9 mils
per kvhr. based on & core processing cycle of five years a.nd & blanket pfo-

cessing cycle of one year.

=18~ -

1

 

 
0

3

W

9

6

 

1.4 DESCRIPTION OF SYSTEM

The fused galt fast reactor which evolved from this study is an externslly
cooled, plutonium fueled, powersbreeder reactor producing 700 megawatts of
heat with & net electrical output of 260 megawatts.

1.4.1 Core

The core fuel consists of a homogensous mixture of the chlorides
of sodium, magnesium, uranium and plutonium with mole ratios of 3NaCL, 2MgCl,
and 0.9Pu(U)613. The urenium in the core fuel is depleted and is present for

the purposes of internal breeding. The atom ratio of UZBS/Tu239

at startup
is 2 to 1.
The core container is a 72.5 inch I, D., nearly spherical vessel tapered
at the top and bottom to 24 inches for pipe comnections. The core vessel
is fabricated of a % inch thick corrosion resistant nickel-molybdemm alloy.
The fuel mixture enters the core at 1050°F and leawas.at 1350°F, where-
upon it is circuleted by means of & constant speed, 3250 horsepower, canned
rotor pump through the external loop and tube side of a sodium heat exchanger.
Sodium enteré,this core heat exchanger at 900°F at a flow rate of 45.5 x 10°

1bs/hr. and leaves at 10509F<__
1.4.2 Blagket
Separated from the core by & one inch fiplten lead reflector is a

stationary blanket of depleted u:anium present as & paste of uraniun oxide
powder in sodium under a 100 ps1_pressure.:chated within the.blanketiis 8
stainless steel clad zone of graphite 5.1/8,1nches:thick. The presence of
the graphite incresses the neutron moderation and results in a smaller size
blanket. |

=19-

- .

 
 

' Blanket cooling is obtained by passing sodium through tubes located |
throughout the blanket, Sodiwm is introduced imto the blanket at 1050°F at
a flow rate of 7.6 x 10° 1bs/hr end leaves at 1200°F. The blanket sodium,
vhich is considerably radioactive, then enters & horizontal sodiur to sodium
heat exchanger and heats the inlet sodium from 900°F to 1050"1-‘.' The sodium
from the blanket heat exchenger is then manifolded with the sodium from the
core heat exchanger and passes to a straight through boiler. At full load
conditions, thé feed water enters the boller at 550°F and 2500 psi at a flow .
rate of 2.62 x 10° 1bs/hr and produces steem at 1000°F and 2400 psi which

passes'to a conventional turbine generator electrical plant.
1.4.3 Control

The routine operetion of the reactor will be controlled by the
negative temperature coefficient which is sufficient to offset reactivity
fluctuations due to expected differences in the reactor mean temperature.

Reactor shim required for fuel burn-up will be obtained by variation

in the height of the molten lead reflector. Approximately one quarter of

~ one per cent reactivity will be available for shim by the increased height

of the lead. When fuel burn-up requires more reactivity than is available

~ from the reflector, compensating changes willl be made in the fuel concentration

and the reflector height will be readjusted.

~ In the event of an excursion, provisions will be nade to dump the entire

core contents in less than 4 seconds and in addition, to dump;the lead re-

flector. Dumping the reflector would provide a change in reactivity of abofit"

1.6 per cent,

-20-

 
 

 

o

A

€

1.4.4 Chemical Processing

Chemical proeessifig of the core and blanket, other than removal
and absorption of fission gases, will take place at a large central processing
facility capable of handling the throughput of about 15 power reactors. The
chemical process for both the core and blanket will embody the main features
of the purex type solvent extraction process, with_different head.énd treat-
ments required to make each'material adaptable to the subsequent processing

- steps.

Core processing will take place on a five year cycle whereas the blanket
will be processed bieannually. The plutonium preduct from the chemical process

is finally obtained as the chloride which can be recycled to the reactor.

3 P

 
 

 

CHAPTER II
PRELIMINARY REACTOR DESIGN CONSIDERATIONS

2.1 SELECTION OF CORE FUEL

 

| _One of'the objectives of this project was ihe toeough investiéation ef
alfueed salfi fuel system. Preliminary discussions resulted in the decision
that a core and blanket breeding system would be investigated,
A fused chloride fuel appeared the most pramising of the fused salt
systems. The core fuel eystem studied was @& fused Na Gl Mg 012, UCIB and

: _PuGIB salt. The results of preliminary nnclear calculatione gave the fused

salt compcsition as 9 mols NeCI 6 mols Mg012, 2 mols UClz and l.mol of PuGl3°

The uranium is 0238

2.1.1 Criterias for Selection

The principal properties that the core fuel system should possess

1. Low parasitic neutron sbsorption eross section.
2. Low moderating power and inelastic scattering.
3. Liquid below 500°C.
Lo Radiozctively stable.
5. Thermally stable.
6. Non-corrosive to the materiels of construction.
7. Low viscosity.
8. Appreciable uranium and plutonium content at temperatures of

- the order of 650°C,
9. High thermal conductivity.

| -22-

-

£t
.

 
 

 

 

For fast reactors, the choice of salts containing fissionable and non-
fissionable elements is 1imited to those in which the non-fissionable elements

have a low slowing down power and low cross sections for absorption end inelastic

scattering of fast meutrons. In genersl, elements of atemic wveight less than

twenty ere unsatisfactory because of their moderating effect. This eliminates

many of the common dilvents which contain hydrogen, carbon, nitrogen and oxygen.
Salts which are suitable nuclearwise are further restricted to'those which

are thermodynsmipally and chemically stable. The salts must be stable at the

operating tempersture/of the resactor, 675 C. Also the liquidus temperature

of the fused sslt m{zturs should be. b;}ow 500°C. This ie desirable so that more

; écmmon and chesper\ ructural msteria s may be used. The higher the temperature

of operation, the\pbre exotic are th# materials required. In addition & lower .

operating temperafhre tends to ret;fd cuerosion, The further very important 7"

requirement is ;fiet the diluents must dissolve the necessary quantities'of.
| /

uraniwm and plutonium to ensble the system to go criticel.

   
 

/

Based upon the aforementio ed'requirements the halide femily appesred ,
the most pr ising., Of the halides, - chiorides and fluorides were the initial
choices. 3ps bromides snd if ides were eliminated because of their high |

/
omine /has an avers,ge 6" at 1 mev of 30 mb and

ebsorption cross sections. /
fodins hss 8 G, of 105 mb &t this ?rergy. Chlorine -and fluorine have osptured
cross seétions of 0. 74 5 and 0,2 mb respectively.

Originally, it appesred*thatvthere were“available 3 possible fuel systems; -
one using chlorides, one fitilizifig fluorioss and & third using'a'mizture‘of-'
fluorides and ohlorioss. Ghlorides presented the obvious disadvantage of a
higher cspture cross seotion. The flourides were detrimental because of their
moderating effect. After a more thorough investigation, the fluorides were

23~

f
"’-

.

 
 

 

ruled ocut because of their prohibitively high inelastic scattering cross
sect_;on in the energy range of interest. Preliminery nuclear calculations
using fluorides showed the neutron .energy spectrun decidedly lowered.

Ultimately the mixed halides system of chioride #nd fluoride was eliminated
because of the high melting points of -f.he fluorides. This step was té.ken only
after it had been verified that & chloride fused salt system was feasible with
respect to the nuclear requirements of our resctor.

Once it was determined that the ffised chlorides would be used, great
effort was expended in the selection of the pa.rticfilar salts to use. One_ of
the most impo;'tant physical properties required was & low melting pd_int
for the salt mixture. It was felt that a ternary system would be most sultable.
A binary would have t00 high a melting point while a quat_ema.ry presented
many unknowns such as formation of compounds; and in general is too difficult
to hendle. | |

The core fuel system will utilize plutonium which is to be pcrodficed in
the blanket. Since there exists very meager information on plutonium fused
salts, it was decided that as a fair approximationl, many of the properties |
of uraniwm salts would be used. This eppears to be a valid assxmpfion for |
. physicals properties since plutonium and uranium salts form = so0lid solufiion.

As a preliminary step, possible diluent chlorides were reviéwed.- Keeping
the basic requirements in mind and reviewing whatever binary phasé Vd:lagrams |
~were available, the following salts showed promise ZrCI& PLC1, Mgclz,_ NaCl,
KC1, and CaCl,. ZrGJuwas rejected since it 1s expensive and might produce
the snow problem experienced in otherfused salt systems. PHCl, was 'réjécted |
s'ince‘it is very reactive with all known structural materials. From the fofir
remaining possibilities, the chlz and NaCl salts were saiected és dilnents.,’ |

,_ e - . |

B

 
 

**»°

[\

 

In addition to possessing many of the requirements,'they had the lowsst liquidus
temperature. As for the fissionable salt, the trichloride or tetrachloride

were the possibilities. PuCl3 andeCIB were selected because of the thermal
instability of the tetrachlorides. Hence the core fused salt system selected

is made up of NaCl, Mg012, U'Gl3 and Pu013. As was pointed out earlier, the
physical properties of our system were investigated using NaCl, Mg012 and 0013.

The PuGl3 is assumed to bé in solid solution with the U013.

2.1.2 Fuel Properties

Since the ternary properties of the proposed fuel were completely un-

vfinown, extrapolations of the known binary systems (shown in Figs..2¢1, 2.2, 2.3)

.....were made,

On the basic assumption that the ternary chloride system‘was a sifiple
one and containing none of the anomalous behavior of the known fluoride systems,
the pictured (Fig. 2.4) ternary diagram was drawn.”

To give some indication of the.mélting temperature to be expected in
our system, = series of melting point determinations was.undertakan. The data
récorded are sumarized below. '(The test procedure is descfibed in the-

Appendix C).

MgCl, NaGl wl, - Liouidus | Soldus
#1 38.6%-57.91% - 3.49% | 435°C - 420°C
#2 36.36%-54.548-9.108 - 432% 415%
#3  33.33%-50.014-16.66% 505°-44,0°C 405%

Sample #3 correspondsflto the composition of the fuel selected.
| -25-

 
 

 

 

ORNL~LR~Dwgo ~1812}

BINARY PHASE DIAGRAMs HaC]-U013

 

8§00
700 |

600

 

 

500
TEMPERAITURE

hoo

300 }
200 }+

100 +

 

 

 

1
0 20 4o 60 80 100
MOLE % UC1, |

Figure 2.1

-6~

 
 

 

N

&

900

800

700
TEMPERATURE
o
C

600

500

- Loo

 

ORNL~LR~Dwg,=181%5

BINARY PHASE DIAGRAM- MgCl,-NaCl

 

 

 

 

\_/

 

 

 

 

 

 

 

20 ho 60
MOLE % NaCl

Figure 2,2

27

 
QP

 

BINARY PHASE DTAGRAM~ MgGl,-UC1

 

800 _}

700 -

 

 

600 ~

TEMPERATURE °C

 

 

 

NgCl,

20

40
MCIE % UCl,

60 80 100
UC1,

Figure 2,3

92ToT="*3MI~YT~TND

 
N

‘\

 

ORNL=-LR=Dwg+=28127

- ESTIMATED
TERNARY PHASE DIAGRANM MQCIE:NaC]-UCl,

 

 

Figure- 2.4

=20=

 
 

 

 

In conjunction with the'melting point tests, a petrographic analysis was
cpnducted of the fuel mixture. On the Easis of this anélysis, neither NaCl,
Mg012, nor U013 were detectable in the solidfied fuel. There were two dis-
tinguishable phases preseht, one a colorless crystal and the other & brown
erystal, which was not as prevalent as the colorless one., The compositions
of the phases could not be determined. It was observed that the mixtfire was
very hygroscopic and was easily oxidized in air.

The remaining physical properties were estimated by analogy to the fluoride
systems which have been studied. Densities were calculated by the density
correlations of Cohen and Jones (3). Thermal conductivities, heat capacities,

and viscosities were estimated directly from fluoride data.

«30-

 
n

 

2,2 SELECTION OF BLANKET MATERIAL

A uranium dioxide-liquid sodium paste was selected as one of the pro-
mising blanket materials, Althbugh only & limited amount of work has been

done on pastes, the prospects for its use are very good.
2.2.,1 Criteria for Selection

The importent characteristics of a satisfactory blanket material
are:

1. Low cost.

2. High concentration of the fertile fiaterial.

3. Cheaply fabricfited.

4o Low chemical prccessing costs.

5. Good thermai conductivity.

6. Low neutron losses.in non-fertile elements.

7. Low melting point.

Natural or depleted urgnium were obvious cholces for the fertile materisl.
Either material is acceptable, the governing factor being the cost. At the
present tima, the coéf'of depléted uranium is conéiderably lesé fhan natural
uranium and was dhosen as the fertile material 1n the blankst.r' |

Several blankst Bthams were 1nvestigated The more prominent possi-
bilities fiere UOQ pellets in>molten sodium, 002 powder in molten sodium, canned
sol4d uranium, fused uranium salts and U02 slurries. - | |

Uranium dicxide pellete 1n.molten sodium appeared very pramising. U02 is;
unreactive with and vary slightly soluble in liquid sodiwm., Cooling could be
accomplished by liquid sodium flowing in tubes. It was estimated that approx-

~31~

 
 

 

imately 65% of U0, by volume could be cbtained. This blanket system vas re-

 Jected because of the high cost of manufacturing the pellets. It was estimated

that over 50 millions of pellebs would be required to fill the proposed blanket
volume of 100 cubic feet. \ | |
A solid uranium canned in stainless steel was investigated. ‘]‘hé ma..jqi'

edventages of this system is the high uranium concentration. This material was
rejected due to the high costs- of febrication. Typicel costs are sbout $9 per
kilogram for machining ura.nimnh and $7 per kilogram of uranium i‘or the addition
of the cladding material. | | | o

' Fused uranium selts would have been the logicel choice since fused sslts
were being used in the core. This would halve many of the problems confrorting
the desim. of the system such as corrosion, chemicel processing, etc. 'ihe
only fused salts which would glve a sufficlent concentra.tion of firdnitm in the

blenket were UCl, or '(JC:LI+ or a mixture of the two. Ucl3- has too high a melting

3
poiat, wkile UClh proved to be too corrosive. Even the UCL

(2)

3" UCllp. mixture was

felt to be too coxrrosive for e long life system, Hence this material was
eliminated.

A W, slurry was rejected due to the lack of knowledge of the properties
of the slurry ani the low uranium concentration due to engineering cdn.sidera-

tions.

The UOE-Na. paste was ultimately selected as the best available bla._nket material.

T™hie system has many of the features of the 002 pellet system w:l.th the omission |
of the cost of mamufecturing pellets. Although only & limited emount of work
has been done on pastes, the ocutlook is very promising. A an - Né. paste -offers
low febrication cost, ease of handling, high concentration of an a.nd good heat
tra._nsfer proPei'bigs « Froma ;pefsonal cénmunication with B.M. Abiaham of Argonne
Nationsl Leboratory, it was estimated that as much as 80% UO, by volume in

«32-

 
 

"

"

 

1iquid sodium is possible using a centrifugation process. We plan to use a
paste composed of 70% volume in the blanket system, The purpose of the liquid
sodiwm in the blanket is to improve the heat transfer properties., It is be-
lieved that Pu end U metal will be stable with liguid sodium and no reaction
occurs between Na and U0,. A major problem was the possibility of Na20
formation and its adverse corfosiva effects. This was solved by the addition

of corrosion inhibitors. A disgussion of this can be found in section 2.4.2.

 
 

\

2.3 REACTOR COOLANT SYSTEM |

The externally cooled system appears superior f.o the internally cooled
sys’ceni for a fused sall fast breeder reactor. In the exberhally cooled system,
the fuel mixtui*e is circuleted through a heat exchanger exbezfial to the reactor
vessel. The internally cooled system has heat transfer surfaces within the
reactor vessel; and heat is transferred from the fuel mixture to a fluid
coolant which In turn is cooled in an external heat exchenger.

2.3.1 Internal Cooling

A possible adventage of the internally cooled system is the lower inven-
tory of core fuel. However, due to the characteristically low heat transfer
property of fused salts, it wes calculated that elmost 50% of the core volume

wld be occupied by tubing and coolant in order to facilitate the required

cooling. The high percentage of tubing and coolant sffects this reactor system
in two ways; First the‘ parasitic capture is greatly increased and secondly,
the neutron energy spectrum is decidedly lowered. The above effects result in a

reduced breeding ratio in the core.

2.3.2 Externsl Cooling

The externslly cooled system was selected for use in the reactor system
investigated. The deciding factor in the choice was that a breeding ratio
of 1.20 was estimated in the externally cooled system compared to only
0.8 for the internally cooled system. This higher breeding ratio ie obtainable
because of sbout 15% greater blanket coverage, less parasitic capture and
highér neutron energy spectrum. Another factor in favor of external cooling

is the ease of replacement of equipment in case of a heat exchanger failure.

~3hu

(B

"

 
n

O

 

2., MATERIAIS OF CONSTRUCTION

The choice of materlals of construction in most reactor syst;ms is
qulite difficult because of the lack of corrosion data in the presence of
radiatidn fields. In spite of the lack of technological development, an effort
vas made to select the materials of construction for this reactor system.

The core vessel will be a nickel-fiolybdenum alloy, which is presently
in the development stage. For the other parts of the core s&stem such as the
primary heat exchanger and piping, a nickel-molybdenum alloy cladding on
stainless steel appears to be satisfactofy. The blanket system will utilize
stainless steel throughout. 'As far as the reactor components gp, it can be
generslly said that all the components in contact with the fused salt shall
be nickel-molybdenum clad or constructed of nickel-moly and all components in
contact with sodium are to be constructed of stainless steel.

Tests are now in progress at the ORNL Corrosion Laboratory to obtain
some datse on the corrosion of the fused salt of this system on nickel and

inconel at 1350°FF.

2.4.,1 Core System
Since the operating temperature of the fused salt shall be as high

as 1350°F, the choice of construction materials was severely limited. A further
limitation was imposed by the sbsence of corrosion data of fused chlorides |
bfl'structural metals."The.possibilit;es:which existed were inconel, nickel-
moly clad on stainless, hastelloy metals, or nickel-molybdenum alloys of the
hastelloy type which currently are under development

In selecting the best material, much dspendence was placed on the individual

chemical and physical préperties_of these possibilities with respect to the

=35

 
 

fused chloride fuvel,

The hastelloy metals were rejected due to the inability to fabricate the
material because of brittleness. Inconel was eliminated for the fibst part
because of its known diffusion of chromium from the 8lloy in fluoride salts.

In addition the corrosion data of inconel in the tampgrature range of intareét
is lacking. It is felt that these disadvantages overbalance the high tech-
nological development and good physical properties of inconel.

| The use of nickel-molybdenum alloy cledding on’staifiless steel sppeaxs very

favorable in the fused chloride system.
| It is expected that this alloy will not exhibit dissimiler metal mass

transfer and will be capable of being welded to stainless steels by use of
special equipment, On the basis that this slloy will have the properties as

described, it is being recommended for the core system.

204.2 Blarnket sttem

- The construction material for all equipment in contact with the
sodium such as ig present in the blanket will be stainless steel. Since the
blanket is to'be composed of a UO,-Na paste, it was feared that the sodium
would become contaminated due to the formatipn of Hazo in the presence of
free oxygen. At elevated temperatures, nazo is very corrosive; it reacts with
all the common metals, platimm metals, graphite and ceramics. The relative
degree of reactivity with the structural materials would be the following,
from the most attacked to the least: Mo, W, Fe, Co and Ni. In addition it
 1s believed that uazo would be strongly ebsorbed on.most‘metal'surfaees,

It 1s possible that since Nay0 is known to act as 2 reducing agent for
some metals and an oxidizing agent for others, the presence of some material

will reduce Nas0 to Na before it attacks the metal, Such & corrosion inhibitor

~36-

 
 

0

o)

n

)

 

 

would solve this dilemma, The two common reactor materials, uranivm and
beryllium could possibly serve as the inhibitor. Thermodynamically, each reacts
readily with Hazo to form the metallic oxide and free sodiwum. At 500°C the
free energy of formation for beryllium and uranium are -6 Kcal per mole and
-75 Keal per mole respectively.
The rate of these resctions has not been ihvsstigatedsexsspt indirectly
in & series of corrosion tests at KAPL5’6, These tests show that both Be
and U are corroded many times faster than any of the structural metals tested.
The metals included nickel, molybdemm, inconel, monel, 347 siainless steel
and 2-8 aluminum. Thus the addition of.either pure urenium or beryllium to
the UOZ-Ha paste should offer a high degree of resisfance to the possible corro- |
sion by the Ea20 which will be formed during irradiation. |
2.4.3 Reactor Components

In general, all cdmponeflts in contact with the fused chloride fuel will

be constructed of nickel moly alloy c¢led stainless steel, All reactor camponents
in contact with sodium will be constructed of stainless steel. |

-37~

 
 

 

CHAPTER 3 ENGIHNEERING
3.1 GENERAL

The reactor proper, as shown in Fig. 3.1, is a 120-inch 0.D. svhere

consisting of & blanket region and a core, The core is a 733-inch 0.D. sphere

with a %-inch wall designed to withstand a differential pressure of 50 psi.

The core inlet nozzle on the bottom and the outlet nozzle on the top are

_reinforged. The inlet has a series of screens to distribute the flow thru the

" core so that a scouring action is achieved.

Immediately outside the core shell is a one inch reflector of molten
lead in & -}-1nch stainless steel container. The f11ling or draining of the
molten lead is accomplished by pressurized helium, | |

The first blanket region is 2 3/4 inches thick and is followed by 5 1/8
inches of moderator, another 5 5/16 inches of blanket and finally 8
1nches of graphite reflectqr. The blanket iz a uranium dioxide;sodium‘paste |
and the moderator is graphite clad with 1/8 inch of stainless steel,

The reflector, blanket and moderator are cooled by molten sodium passing
thru 4-inch 0,D. tubing.

The core heat output is 600 megawatts, and it is removed by circulating
the fuel thru a single pump and external heat exchanger with a minimum of

- piping. The cooling circuit is fabricated using all-welded comstruction. The

fuel solution is heated to 1350°F as it flows up thru the core and is returned

‘to the core at 1050°F,

Any differential expansion will be absorbed in a pivoted expansion joint,
The'Blankat heat output is approximately 100 megawatts and it 1s removed
by eirculating molten sodium which enters the blanket at 1050°F and leaves
~38-

»

 
 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

0 l « y "
' » 0 .
39
- ORNL LR Dwg. 15226
SUPPORT - | 7 _[ N
HANGERS) ' 7= 20 FT. LONG '|
“Cq l
| % 3 I
' i 5 \bBLK l
s ' ‘ 99 _

A . v

'l;——o j : /’ ' C |
L7 _ 3~ 1

7 I 791

' | | 3,500 Tupes
S s .50" o}p,
g | o | somils WALL’
» . N / ‘
» _ . ' . :-_;1__ - ( ,I ' { ‘
1S70 TUBES B | | T 7 N - 4 2400,
50" OD. % SO wnls 3 PO 1 27,5006 A / —PIPE ¢
WALL x7.SFT SOuAD \ \+O0FT < AT riTTiNGS
’ ‘.‘ - | SUPPORTs \\ f T T, o ___.—’ v
| EXP. 3T,

 

FIGURE 371 | | 2" scaLe ¥ =10" Je"

 

 
 

 

at 1200°F thru tubes imbedded in the blanket as shown in Fig. 3.2. As
.the core, the blanket cooling-system‘haS’one punp, one heat exchanger, uelded
Piping and a pivoted expansion joint.

The combined core and blankst system has three solid leg supports on the
blanket, Constant load hangers will carry the remaining load at four lugs
provided at tfie upper core elbow, at the core heat exchanger and at each end
of the blanket heét exchanger; |

The basement floor of the reactor_building,-és shown in Fig, 3.3, will
have & series of duhp\tanks for the salt, The reactor floor and the main floor

will be constructed of removable. stesl panels, The reactor room and the base-

ment room will be below ground_lével and contained in a steel lined concrete structure.

The réactor building main floor will have television facilities-éndqa
remotely operated erane and will be enclosed in a 60 ft. diameter, one inch
thick steel shell. The steel shell is = éafety'measure and will prevent the
pollution of the atmosphere by radioactive materials in the event of an accident.
The steel shell, which will withstand 50 psi, will have two large airtight

hatches for equipment removal.

The blanket heat exchanger secondary sodium lines are siamesed with the
core heat exchanger sodium lines and the resulting 42 inch O.D., lines are con-
nected to the shell side of a once-thru boiler.

The U-shaped boiler and the sodium pumps are located in a shielded boiler

room between the reactor building and the turbo-generator portion of the plant.

-The layout of the turbo-generator and auiiliaries follows the convéntional
power plant design with two exceptions: an outdoor turbine floor with & gantry
crane and placefiant of the deasrator on the turbine floor because of the
élimination of the boiler superstructure.

d Q=

#fi

»

t

 
» o

EXPANSION
JOINT

-

BLANKET

|
’ 1
1
'
1

 

 

ORNL-LR-DWG 15032

 

 

COOLING TUBES

 

 

HEAT EXCHANGER
1570-1/2 in. OD TUBES
7.5 ft LONG, 0.050~in. WALL

 

 

 

 

 

COOLANT TUBE | |
. SUPPORT PLATE-—]

 

 

 

 

 

  

 

MODERATOR: C,

REFLECTOR: Pb

 

 

 

 

 

 

 

 

 

 

1000-HP
PUMP

18-in O. D. PIPE

Blanket Heat Transfer System,

Figure 3.2

 
#‘ L ' ORNL~LR~-DWG 15166

20 £t Qin.~»

  

75 ft Oin.

  
 
 

__EL 78 ft Oin.

  
  
 

EL 7O f¢

DEAERATOR

  

TURBOGENERATOR {!

  
   
 

CONTROL
ROOM

 

   
 
 

EL 35 ft Qin.

  

60~-ft O=in. DIA; #-in. WALL
SWITCH

GEAR

 

HEATERS

    

  

- BOILER

FEEODWATER
HEATERS

 
  
   

EL Oft

vl

8-in. Pipes.

  
 
  
  

 
  

e

  

   

 
   
 
  

CONDENSATE PUMPS

   

F PUMPS L L)
L

   

CIRCULATING LINES

42-in. Pipes

   

Elevation of Plant

Figure 3.3

 

 
 

o

13

"

reactor buildings and above the boller room,

 

The centralized control room is placed between the turbo-generator and
The stack, which is used for

the dispersal of reactor off gases after a sufficient hold-up time to reduce

the radioactivity, is placed near the reactor building.

 

3.1.1. Properties of Fused Salt, Sedium Coolent, and Blanket Paste

The éngineering properties of the fused Salt, godium coolant and
the UOp-Na paste blanket have been estimated by the follbwing methods. The
specific heat of the fused chloride salt as & quction of uranium concentration
(Fig. 3.4) was estimated using the method deseribed by W. D. Powers™. Correlations

vere not avallable for the properties of thermal conductivity or visocity of

the fused salts.,

—— The variation of the_density, specific heat, thermal conductivity, and

viscosity of sodium? are givefi in Figures 3.5, 3,6, 3.7 end 3.8 respectively,

The density of U0, was~taksn as 10.2 gm/bc.and'it'was_assumed that this re-

mained constant. The specifie heat4 of UOQIwas taken‘as:
= 19,77 +1.092 x 2077 - 4.68 x 10-5 T2 (Gal/mol c)
(Figure 3.9

The  thermal conductivity5 of UOQ is given in Figure 3. 10.
The properties of the paste were then daleulated using a'mixture of 70%

00,5, 30% sodiumiby volune.
@ = " Ora* ooz Coo, -
Cp = Wy,CPy, + Vo, CPyo,
¥ = V0,550, Ve B

vhere: V = Volume fraction

(Figure 3.11)
(Figure 3.12)
(Figure 3.13)

W = Weight fraction .
-t 3~

 
25

20

15

% UCL, in NaCl-
MgC1,? EUTECTIC

10

2}

0,19 0.21 0423 0.25. 027
SPECIFIC HEAT, BTU/1b. F

% UC]g VS, SPECIFIC HEAT

Figure 3.4

 

0,29

QIINTROD0GNN

Q2T Mg~ T~"TNHO

 
58

56

54,
DENSITY
(1v/rt3)

52

50

A8

46

400

 
 
 
 

€00

- -

Figure 3.5

800 1000

TEMPERATURE, °F

w

1200

DENSITY OF SODIUM

VS,

1)

TEMPERATURE

1400

 

1600

 

62T9T="*3ng=yT~INHO

 

 
  
     

Figure 3,6

SPECIFIC HEAT SOD
0.34 CF SODIUM

Vs. ‘
TEMPERATURE

0.33

I
>~
0.32 T
SPECIFIC
HEAT o
(BTU/1b- °F) %
031 E
o
0.30
0429

200 D400 600 800 1000 1200 1400 1600
S TEMPERATURE, °F '

 

 
50
4L8
46
- THERMAL ’
CONDUCTIVITY -~

(BTU/hr-£t- OF)

"

- 42

.40.

38

200

400

Figure 3,7

600 800
TEMPERATURE, °F

¥

 

1000

 

N «

THERMAL CONDUCTIVITY OF
SODIUM VS,
TEMPERATURE, °F

TETQT=*3nq-¥T~TNEO

- L*’-

 

 
VISCOSITY
(1bs/hr-rt)

1.4

1.2

1.0

0.8

0.6

0.4

400

 

#5

VISCOSITY OF SODIUM VS, TEMPERATURE

Figure 3.8

600 800 1000
TEMPERATURE, F

1200

»

 

1400

ZETYT~"3nq=YT~TNID

 
 

, f i
SPECIFIC HEAT OF U0, VS. TEMPERATURE

0.80

0.75

0.70
SPECIFIC HEAT
(BTU/1b- °F)

0.65

 

0,60

400 600 800 1000 1200 1400 1600
| | - TEMPERATURE, °F

Figure 3.9

 

£ETYT="3ng=YT=TNU0

 
S0

.——-——-—’

o~

 
 
 
  
  

Figure 3,10

THEFMAL CONDUCTIVITY (F U02
vs.
TEMPERATURE

 

4
THERMAL
CONDUCTIVITY
(BTU/hr-£t-°F )

-.0.('-

NETYT="3Ng~yT~TNI0

200 400 600 800 1000 1200 1400 1600
TEMPERATURE, °F

 
600

500

400
PASTE
DENSITY
(1b/£t3)

300

200

100

0.l

0e2

0.3

Figure 3,11

04 0¢5

VOLUME FRACTION UO:2

0.6

0.7

N

0.8

 

-‘[s-

SETRT="3Mg=T~INYO

 

 

 
0,100

0.095

 

SPECIFIC HEAT OF PASTE VS. WEIGHT % UO,

Figure 3,12

 

0.090 S
SPECIFIC v
- HEAT
(BTU/1b- °F)
0.085 %
L
=
0,080 ,c'.;
&
O
0.075 | _
0.88 0,90 0,92 0.94 0.96
. WEIGHT % U0y
i - ) ; . » "
' ¥
. v & ’ w .

 

 
(BTU/ hr-ft- °F)

THERMAL CONDUCTIVITY

53

Figure 3.13
50

    

40

30

20

10

0.1 0.2 . 0.3 0.4 0.5 0.6
VOLUME FRACTION U0,

1 )]

0.7

THERMAL CONDUCTIVITY OF UG,~Na PASTE
' VS.
UO, VOLUME FRACTION

0.8

 

-gs-

L ET9T="> Inq=gT~TN0

 

 

 
 

 

 

3.2 REACTOR POWER

The reserve capacity of an electric power system averages about 10 per
cent of the system load, To'mgke such a system relisble, no single unit should
exceed 10 per cent of the system capacity. Since most'of the systems in this
country are less than BOOOIMW'eapacity, turbo-generator units in excess of
280 M4 have not been built yet,

A reactor supplying steam for a single turbo-génsrating unit with a
system thermal efficiency of 40 per cent would be sized at 700 M{ of heat, or
also 260'MW net electric 6utput because of auxiliary power requirements of 20 MW,
A system larger than 700 MW of heat would require more than one circulating
fuel heat exchanger. Two fuel heat exchangefs would require manifoiding and
other flexibility provisions which would result in a gfeat increase in fuel
hold-up. Furthermore, too high a power level would involve such a large initial

investment that the risk of construction would not be Qarranted.

~5hm

M

 
4y

 

3.3 DESIGN OF HEAT TRANSPORT SYSTEM

Reference is made to Fig. 3.14 and Fig. 3.15, the Heat Balance Diagram
and the Salt, Sodium, Steam, and Condensate Flow Diagram, respectiwvely.

The optimm design was approached by careful selection of &eeign points,

Single wall tubing was assumed throughout wvhich is in sgreement with the
present trend of design. OSmsall leakage of water or steam into the sodium in

the boiler is not expected'to cause serious difficulty(lg).

Detection may
be accomplished by providing a gas collecting chamber and off-take in the
sodium return line. Build-up of NaOH in the sodium system should not be
difficult to follow and replacement or purification of the sodium can be under-
taken as it may eppear necessary. |

The influx of large amounts of water or eteem resulting from a major
failure would dangerously increase the pressure in the shell; and although
this possibility 1s remote, safety valves will be provided.

 Excessive fluid velocities result in erosion, corrosion, vibration and

increased preesure drop. Based on past experiences in the field ) the maximum
velocity was taken as J900/S" ft./sec., wh_ereg is the specific gravity of
the fluid. | o

Fluid-fuel reactors, especie.lly those with external cooling, are part-
icularly liable to be shut. down for repair or replacement of equipment (13 )
It is highly des 1_rable, therefore, that gll components be as s:l.mple end as de~
pendatle as ',.pos'si'ble jbut also able to be '_epeedi‘ij reple_eed or remotely main-
tained. It is eansiaerea undesireble to install valves in the large lines be-
tween the core a0 blanket heat exchangers and the p‘lmps to permit shut-off
of possible spare equipment or to regulate flow. These valvee would be large,
would operate at high temperatures and would handle corrosive fluids. It is

~55-

 
 

SE

 

 

 

 

  
  

 

 

 

 

 

 

 

 

 

   
  

  
      

PUMP
, AUX, Na
 4350% 3x1061b/hr |
| 1050°F ’
SALT
34.2 x1081b/hr
CORE Na
455 x 106 Ib/hr
800 Mw 900°F
00
riM & PUMP
e
- PUMP
r
1 1200°F 1050°F
Na Na
7.61x 10° 1b/hr 7.61 % 10° Ib/hr
PUMP PUMP
r

 

 
 

Heat Balance Diagram.

Figure 3.14

  
  
 

 

ORNL-LR-DWG 14973

" PUMP

550°F

Ho0

 

1000°F HEATERS

280 Mw
TURBINE

CONDENSEFQD

 

 

 

 

 

2.62 x 108 1b/hr )

 
o

\;

ORNL-LR-DWG 14985

= e e - -

STACK TO
ATMOSPHERE

CHARCOAL |
. BED

STACK TO
ATMOSPHERE

[oo----

    

 

 

~57-

 

STEAM JET VAC. PUMP

- - ———

[ o e o e  ——————— ———— e b —————— -

  

DEAERATING HEATER

 

 

 

- ——— i ——— - —— -

o ————— e

 

H31VM Q334

  

 

 

"HOX3 1VY3H ‘NNIJ
Q37000 Yiv
W3LSAS XNV

 

 

 

 

 

 

 

VENT

WWWWA

 

 

 

 

 

|
|
e G

 

 

:

 

 

 

 

 

 

PRESSURE
AND
SURGE TANK

 

 

PRESSURE
AND
SURGE TANK

 

 

 

 

 

 

S00IUM
2

 

 

 

 

 

 

 

 

 

 

" VENT

 

 

BED

 

 

 

| CHARCOAL

 

PRESSURE
AND
SURGE TANK

 

 

 

 

 

 

 

 

 

 

 

 

SALT DUMP TANK

 

 

BLANKET
bump

 

FILL PUMP

 

 

 

 

 

 

PRESSURE
AND
SURGE TANK

 

 

 

 

Salt, Sodium, Steam and Condensate Diagram.

T

Figure 3.15

 

 
 

 

‘more probable that these valves would fail before trouble is experienced at

thé héat exchangers or pumps.

Since no maintenance can be attempted with radiocactive fluids 1n the re-
actor and since it is not expected that any reactor part will-last‘five years
without requiring replacement or repair, provisidns will be made to inspect
21l components thoroughly at leaSt'evary two years, i.e., when fefilacing the

core heat'exchanger.

3.3.1 Circulating Fuel Heat Exchanger

To reduce external hold-up, small tube sizes are desirable in the

heat exchanger., The $-inch 0.D. tube size was selected as a practical

minimm. For sizes less than % inch, considerable difficulty would arise in
fabrication of the heat exchangers while the possibility of plugging would be
greatly increased. The wall thickness of 50 mils was assumed to provide
corrosion resistance for two years of useful life.

For the secondary heat transfer fluigd, a‘medium was required with good
heat transfer properties in order to reduce the external‘holdaup and with high
boiling point to permit operation at high temperature and low pressure to re-
duce capital_cfisfis.

Sodium,-lithiufi, NaK, bismuth, lead and mercury were considered as heat

trensfer media, Sodium was selected because of its good heat transfer pro-

perties, high boiling point, 1ow.coét,.ava11ab111ty, comparative ease of handling

and wide technological experience, The disadvantages of sodiwm arejits violent
reaction with water and the catastrophic_corrosion rate of Na,O0.
‘The following considerations were used to set the temperature limits for
the fluids entering and leaving the core and heat exchangers.
~58= -

"

n

 
 

t)

i

 

The coolant temperature is not to be less than the liquidus temperature
of the fuel, i.e. 870°F. .Thé temperatufes of the fuel end coolant leaving the'.
core and heat exchanger were set by economic, corrosion and engineering.con-
siderations, Low fusl outlet temperature would lead to excessive heat ex-
changer surface which would adversely affect the fuel inventory and increase |
the possibility of‘QQSB transfbr. High fuel outlet temperature would increase
the corrosion rate, require higher pumping power and increase the thermal
stresses, Low sodium outlet temperature Qould result in excessive thermal
stresses and lower thermal cycle effidiencya High sodium outlet temperature
would have the same result as low fuel outlet temperature.

The fuel outlet temperature was set af 1350°F to ensure reasonable equip-
ment life and the maximum temperature differential between fuel and sodium was
set at 300°F, which is in agreement with genefal design practices. |

The heat exdfianger is a single pass counterflow exchanger approximately
50 inches in dismeter and 20 feet long with 3500 tubes. All tubes will be
made from a corrosion resistant nickel-molybdemm alloy (ebout 80% Ni and 20%
Mo). The exchanger shell will be constructed of stainless steel with a 4 inch
Ni-Mo cledding. | |

The,remdfal and replacemenl of the core heét'exchangar requires remote

handling which is believed to be entirely feasible.
3.3.2 ;rculating Fuel Piging and Pump

The pipe size selected was 24-inch 0.D, with a one-inch wall thickness.

To reduce cost, the pipe material will be stainless steel clad on the inside

with & corrosion resistant Ni-Mo alloy.' Cladding thickness will be 4 inch to
provide & corrosion sllowance for five years life. To allow differential thermal
expansion, a pivoted expansion Joint 1s'pfo#ided.

==

 
 

.

‘A single pump arrangement was selected, because two circulating fuel
pumps would require two check valves, four shut-off valves and added proiisions
fdr flexibility. This wbuld increase the external hold-up and because of valve

stem 1eakage probabilities, would lower the system reliability. However, if

large, relisb le valves becone availéble; it might be advantageous to have the -

added flexibility afforded by‘fiultiple cooling systems. This is a matter for
further development. | |

'A canned-rotor.pump was selected instead'of a shaft-seal pump due to the
greatly reduced possibility of leakage. The fuel pump will run at constant
speed becafiée of its canned-rotor construction. A variable speed pump would

be preferable but this also requires further develoyment.
~ 3.3.3 Blanket Heat kxchanger

The blanket heat exchanger is a sodium to sodium exchanger constructed
of stainless steel and whose mean temperature difference iz 150°F. It has
1570 tubes of 4-inch 0.D, which are 7% feet long with 50-mil walls.

It was deemed necessary to have an intermediate loop on the_blanket system
due to the activation of the sodium coolant, Thus, in case of a sodium-water
reaction, only radioactively cool sodium would be ejected. The choice of
sodifim as a secondary blanket coolant was deemed advisable since the core second-
ary coolant and the blanket secondéry coolant, could be mixed,.thué necessitating

only one boiler and a slight amount of manifolding. For this same reason, the

 secondary sodium is designed to have a 150°F temperature rise through the heat

excfianger (QCOQF to 1050°F), thus matching the core sodium,
3.3.4 Blanket Heat Removal | o o |
- The breeding blanket is in the form of two separate spherical anmuli.

=8

»

1)

 
»

 

-

The first blanket region is 7 om. thick and hes 60 M{ of heat gemerated im 1it.
The second region is 13.5 cm. thick with 40 M{ generated in it. The heat flofis
by conduction through the paste to the wetted tubes where it is then cerried
awvay by convection in liquid sodium,

In blenket region 1, there are 940- % inch stainless steel tubes whose
centers lie on circles of radii 38.4, 39.1 and 39.9 1n¢hes. Each row contains
equal mmbers of tubes which have an effective length of 8 ft. Under these
conditions, the maximum possible paste temperature will be 1396°F which is well
below the refractory temperature of 18320F, |

In blanket region 2, there are 630- 4 inch stainless steel tubes whose
centers lie on circles of radii 51.8, 53.5 and 55.3 inches. Each row conteins
equal numbers of tubes which have an effective length of 10 ft. The maximum
paste temperature in region 2, under these conditions, will be 1468°F,

In region 1, the cooling tubes occupy less than 30 per cent of the avail-
able volume while in region 2 the tubes occupy less than 15 per cent of the

available volume.

3.3.4.1 Parameter Study of Blanket Heat Transfer System

For efficlent coolifig ofthe'bianket, we expect to match the
cooling tube density to the radial{diétribution of heat generation.

It will be assumed that thebasiccooiing tuberlattiée arrangémsnt éan
be simulated by conéentric cyiindérs."Tha ggneration rate in a cell will'be-
taken as constant and the_flg-Uoz'paéte will be.céfisidered stagnanfi. The pro-
perties of Na and ma.uoz' paste are §raphieany presentéd in SjectionB.l.l‘.

Taking a heat balance at any radius r where ri (r (ré |
GV(r) =-kp Alr) _o T |
o r |
T «Hl=-

 
 

 

where | . To = inside radius of tube

5 5 ™ ; outside radius of tube
v(r) =qf(r2 -r )L - r, = radius of cell
Alr) = 29r L
thus, & T = G 223 -7

 

 

or " S

T
2 » 0
(T-Tl) G [1-2 II.::A_;_-:t‘z-rl:2 / \
i L2 | J
| \

In the steady staté,

Q=0V =k & (To - Ty)
2 2
ke T2 ~—T1
r r22 In 2 - r22 -'r12-
' kP ry 2 y
(14)

basing over-all coefficient on inside tube area ,

U ok k, A, h
k=17 BIU _
hr—££-°F

In the Fig. 3.16, the value of U is given as a function of r, where rj

{3 treated as a parameter. In all cases, standard tube wall sizes were used.

For 1/2" 0.D. tubes with 50 mil wall

 

 

7 ReNa - 348,000 PrNa - 0.001&24
hy = 17,350 __ BTU
S hr, ft2 OF
Ay S _ ,2(.05) - .000303

kb, T 1o x 12 x J45
For :2[_4" 0D tubes with 65 mil wall

rReNa = 529,000 FPrp = 00424

h = 14,700 _ BTU
hr. £t °F

 

»

£

 
 

 

ORNL-LE~Dwg.~16138

   
 
   

1800
Figure 3,16
OVER-ALIL HEAT TRANSFER
~ COEFFICIENT IN PASTE
. BLANKET VS.

1600 EQUIVALENT TUBE CELL RADIUS

1,00

1200
OVERALL

HEAT

TRANSFER
COEFFICIENT
(BTU/hr-ft°F)

1000

800
600
- oo
200
0 , _ _
2.0 M 6,0 8.0 10.0
EQUIVALENT TUBE CELL RADIUS, (in.)
-63-

 
 

 

Ay & _  .310(,065) - .00041
kwAw 12 x 12 x .342

For 1" 0D tubes with 85 mil wall

Rey, = 720,000 Pry, = 00424

he = 12,000 BTU
Na = 1%s —_—t
hr. £t OF

b & .415(,085) = .000535
KA 12 x 12 x .458

3.3.5 Blanket Piging and Pump

The total pressure drop in the blanket is 145 ft, of head. Thié
includes the losses throfighthe blanket tubes, four plemum chambers, 10.67 ft.
of 18.inch 0.D, pipe, four elbows, one expansion joint, heat exchanger tubes,
blanket tube sheets and heat exchanger tube sheets.

The blanket sodium pump is a rotary pump with a capacity of 18,700 Gm
of sodium ageinst a 145 ft. head. With a2 pump efficlency of 70 per cent, the
motor required for the pump is a nominal 1000 hp.

The blanket is filled by pumping the UO,-Na paste into the blanket vessel,
under a helium pressure of 100 psi., prior to the reactor start-up. The blanket
will be completely filled-and any expansion of the paste will be taken up in
the blanket expansion tank.

To empty the blanket, part of the paste will be forced out, using 100 ps{.
helium., Pure sodium will then be used to dilute_and vash out the remainder
of the paste. When enough sodium is added, the paste will aésfime the properties

of a slurry and will flow'quite easily.

. -{61'_-

8

 
 

 

3.3.6 Sodiwm Piping and Pumps

The pipe sizes selected are 18-inch 0.D, for the blanket heat ex-
changer and 42-inch 0,D, for the main lines to the boiler. The piping materisal |
will be stainless steel. Sodium valves located in the lines will be plug~type
with freeze seals.

Canned-rotor pumps were selected in preference to the electro-magnetic
pumps because of their higher efficiency. The total flow is 114,000 Gmm, which
requires at least four pumps with 28,500 Gm. and 65 ft. head.

Provisions are made to drainland storé ell sodium in the event of a shut-
down. A one~foot thick concrete shield surrounds the sodium system including
the boiler. The sodium pumps will be shielded so that they can be drained and

replaced individually without denger to personnel.

“b5=

 
 

~

3.4 SALT DWMP SYSTEM

A 821t dump system is provided consisting of two valved drain lines, one

for the core and one for the heat exéhangerv and p:lping.- The lines are 12 inches
| and 8 inches, respectively, and are sufficient to drein the entire system in
four seconds.

The dump tanks will have a combined capacity of 10 per cent in excess of
the total circulating fuel volume., The tanks will be compartmentalized to
keep the fuel suberitical; and cooling provisions will be provided to remove
decay heat. Electric heating elements will be included to prevent the fuel
from solidifying.

The fuel will be removed from the dump tanks by a 5 Gpm., 130 ft. head
punp either back to the core or to a contalmer to be shipped for processing.

The tanks, piping and pump will be comstructed similarly to the main eir-

culating fuel system, i.e., nickel-molybdenum alloy cled stainless steel to

 

provide an allowable corrosion resistance for 10 years of useful 1ife,

fi

"

 
 

3.5 CORE VESSEL AND REFLECTOR HEATING

ih this system, as in most reactor systems, the internal generation of
heat in the core vessel due to gamma and neutron interactions with thelmetal
was found to be appreciable. The energy sources considered for this calculéfion
were prompt fiséiohigémmas, decay product gamfias, and neutrons of energies
greater than 0,12 Mev, .The inelastic scattering gammas in the fuel and the core
vessel were estimated as negligible with respect to the magnitude of the con-
sidered sources. These sources gave a gamma spectrum as shown in Fig. 3.17.

Using this integral spectrum and assuming it to be unchanged in space
we applied the Infiegral Beam Approximation method(IS) (Appendix A~7). The heat
generation rate in the core vessel and reflector was calculated as a function
of position. The gamma absorption coefficients of the fused salt (Fig. 3.18)
and of the nickel-molybdemum alloy (Fig. 3.19) were computed fdr use in this
calculation. The gemma heat generation rate as a function of position is
shown in Fig. 3.20.

The heat generation due to neutron capture, elastic scattering, and in-
elastic scattering were caloulated using_the-integral fluxes from the Univac
caleculations with the general equation' | |

G = Z (E) TF (E) E 5 B (Galculations :ln Appendix A—‘?)

where.ETE) = throscOpic cross-section for the specific interaction

¢(E) = The average flux = JQQE,r! d3
jd3r

E = Average neutron energy

3§ = The average energy  transferred/ interaction.

The sources ylelded a total averaged heat generation rate of 9.65 x 1013

Mev/cmB-sec in the core vessel and 1,77 x 1013 Mev/cm’-sec in the lead re-
-67-.

 
1.0

2.0

 

25

GAMMA SPECTRUM VS. ENERGY

3.0 4.0 5.0
ENERGY, Mev.,

- -

7.0

 

4

6€TQT~* 3~ T~TNHD

 

 
GAMMA ABSCRPTION CCEFFICIENT COF FUSED SALT

VS. ENERGY
1.3 n .

 

1.1

0.9

GAMMA
ABSCORPTION -
COFFICIENT

oML 0,7

-69=

0.5

0.3

ONTGT=*2nq~T-7NI0

0.1

1.0 2.0 3.0 ) - 4.0 _ 5.0
ENERGY, Mev.

 
0435

0,30
ENERGY ABSORPTION
CGEFFIGIENTl
/’-0‘; cmo-

0e25

0420

0.15

1.0

/O

Figure 3,19

GAMMA ABSCRPTION CCEFFICIENT VS. ENERGY

2.0 3.0
ENERGY, Mev.

40

5.0

n

 

THTT=*2Mq~TT-TII0

 
")

 

ORNL~LR~Dwg. =182

GAMMA HEATING RATE IN CORE VESSEL AND REFLECTI(R

1014

HEAT
GENERATION

FATE3
(Mev/cm’ -sec)

10

10

 

1.0 2,0 3.0 4s0
DISTANCE, CM.

Figure 3.20

-Tle

 
 

flector. It was found that approximately one third 6f the total heat genersation
in the core vessel was due to gemma ;nteractions. Using thesé averaged heat
generation rates a maximm temperature rise of 109.3 °F was estimated for the
core vessel (Fig. 3.21). Since such a température rise was believed to cause
abnormally high thermal stresses, it was decided to coal the lead reflector.

This gave a maximum temperature rise in the core vessel of 29,2°F (Fig. 3.22).

This was estimated to yield permissible thermal stresses.

In all these calculations the core vessel was taken to be 1.3 em. thick;

and the leed reflector, 2.5 cm, thick.

In order to maintain the 29,2°F tempefature rise in the core shell and
to minimize the thermal stresses, it was postulated that both surfaces of the
core vessel be maintained at the same temperature of 13509F and that heat be
removed from the reflector to accomplish this., It was also postulated that
both surfaceé of the reflector are at 1350°F. Using these conditions it was
found that 5.2 x 1013 Mbv/me-sec will be removed from the lead reflector.

Q = 5.2 x 10'2 Mev/emPsec = 3.80 x 10° BTU/Ar. = 1.11 Mi. |

Using a row of blanket cooling tubes we have & sodium flow of 83,500 1lbs/

hr. through 17 1/2-inch OD tubes with 50 mil walls, The heat transfer cal-

culations show that this is more than adequate to tfansfer the heat.

(Appendix A-7).

72

5

it

 
 

 

 

ORNL=LR=Dwg « «18143

WITHOUT COOLIMG IN LEAD REFLECTOR

TEMPERATURE VS, DISTANCE THROUGH CORE
VESSEL AND LEAD REFLECTOR

120

100

80

TEMPERATURE
°F

60

Lo

20

 

-1 <0 2,0 3.0 L.o
DISTANCE, cm,

Figure 3.21

73

 
 

 

50

25

| 0
TEMPERATURE
o
F
25
50
75

WITH COOLING IN THE LEAD REFLECTOR

ORNL~LR=Dig.=181kk

TEMFERATURE VS. DISTANCE THROUGH CORE VESSEL
AND LEAD REFLECTCR

1.0

2,0
DISTANCE,

Figure 3,22

“Th-

Clhe

 

4.0

Ll

 
»)

 

3.6 MODERATCR COOLING

The heat generation in the moderafior due to fast neutron moderation is |
3.5 x 1012 nev. which yields a heat generation of 6.15 x 1018 mev, or
QC--SGO. ) ' ) SGO.
3,36 x 105 BIU__ in the entire volume, With this heat generation rate, & sodium
hr.
flow of 1.13 x 10° 1bs. is required to maintain the maximm temperature of the
- ,

 

graphite at 1325°F, The sodiun flow rate is accomplished in 25 4-inch cooling
tubes with 50 mil walls.

3.7 ONCE-THRU BOILER

The once-thru bojler is well éfiited to the high temperature reactor plant,
since load conditions can be controlled by varying the flow of water. If the
reactor follows its load demand well, it can be conirolled directly by the
turbine throttle, Thus, operation of the plfint is greatly simplified. However,
the once~thru boiler is not yet well developed and‘in this case.is operating
very near the burn-out point. This is perhgps one qf the weakest points in the
design., It definitely requires further atfidj and possibly another intermediate
sodiunm loop to lower the inlet sodium temperature to the boiler. This type of
boiler requires very puwre feed water of ieés.than-%-ppm. impurity pnésent.

The boilér is in the form of a shell afia tfibe, cqunter current,rone-pass
heat exchanger with the 2400 psi stesm on the tube eide. There are 2400 tubes
which are $-inch 0,D., 45 ft. long, with a 50 mil wall, The entire boiler will
be made of stainless stéel which_is.résiétant to attack b} both hot_sodium and |
super heated steam, The tubes ére in e tniangular lattice with a 1,11 inch
pitch vhich léaves suffiéientrdom for welding the tubes into the tube sheet.

The inside shell diamster is 4.9 f%., and fts wall thickness is ome inch
which is sufficient to hold the sodium. The overall shell length is 50 ft.

=75

 
 

 were calculated using the Dittus-Beelter equatibn

which includes two 24 ft. plemums., The boiler was made into a U-shape in
order to reduce the size of the boiler.roam; o

The design vas ecoomplished by breaking the boiler ipto three distinct
fégions-a sub-oboled reglon, a boiling region and a superheated regidn. Thié'
is bnly an approximation as it 1s'mainly a philosofihical point as to where sub-
cooled boiling ends and net boiling begihs. The heat transfer coefficlents
(14), and a method of J. A,
Lane(16) was used in the bolling region.

In calculating heat transfer coefficients, use was made of inlet velocities
only. This is clearly an underestimate, and the excess aurfe.e‘e. ghould account
for the resistance of the scale to heat transfer.

At part-load operation, this boller tends to produce steafi di higher than
design temperature. The steam temperature to fhe turbine will be maintained

constant by attemferation end veriation of the boiler feed water température.

The part-load operating characteristics of the boiler are given in the following

table.
Teble 301
Boiler Characteristics at Part-Load Operation

Fraction of Full Load ¥ 3/4 1400
Steam Outlet Temperature 1080 F 6 1067 F 6 1000 F 6
Water Flow Rate 1.23.x 10~ lbs/hr 1,88 x 10° 2,62 x 10
Sodium Inlet Temperature ~ 1085°F 1082°F 1050°F
Sodium Outlet Temperature 1010°F 6 970°F 6 900°%F o
Sodium Flow Rate 53.2 x 10° Ibs/hr 53.2 x 10° 53.2 x 1
Over-all Coefficients f -

Sub-cooled Region 1000 1160 1275
Length Sub-cooled Region 2,1 £%, - 3,16 455
Iflflgth Boiling Rfigion 4.6 £t, ' 7.5% 9.47
Over-All Coefficients | o I o

Superheat Region 560 705 826

Length Superheat Region 38 f£t. 3 304
| ~76= |

N

"

 
o)

3.8 AUXILIARY COOLING SYSTEM

If the electric load 1s dropped to zero, 1t becomes necessary to remove
delayed heat from the reactor core and blanket., An auxiliary cooling system
1s provided for this, comsisting of a separate sodium circuit, a sodiwmm-to-air

heat exchanger and a pump.
3.9 TURBO-GENERATCR

A tendem-compound, triple flow, 3600 rmm. turbo-generator with initial
steam conditions of 2400 psig. and 1000 F was selected. Since & straight-thru
boiler is being used, there is no reheat. The latter generally is not too
desirable for nuclear power plants because of the attendent complicated controls.

The feed water cycle will consist of six heaters with the deserator in
number three place. The final feed water temperature is 550°F, Three condensate
and three boliler feed pumps ere specified to insufe the relisbility of the unit.

The thermal efficiency of the cycle is estimated to be 40 per cent. Auxil-

lary power requirements are estimated to be seven per cent,

-77=-

 
 

 GHAPTER 4 NUCIEAR CONSIDERATIONS
4.1 SOMMARY OF STUDY INTENTIONS

| At the onset of the project, tfio codling syfitems for a fused salt reactor

were considered, One was an internally cooled system in which the coolant,
- lquid sodium, was passéd through the core of the reactor. The other was an
externally cooled reactor in vhich the fuel was circulated through a heat ex~
changer enternal to the core. It was felt that the 1arge fuel_inyentory of
a»fast reactor would be increased to & prohibitive amount in the circulating
fuel system, Howévar, eerly calculations showed, that because of the large
amount of parasitic sbsorption, the total inventory of the internally cooled
system was about thé same as that of the circulating System. Poorer bianket
coverage, more parasitic capture and lower spectrun caused the internally cooled
system to have a breeding ratio estimated to be about 0,8 compared to an esti-
mate of about 1.2 for the circulating system., The lower spectrum would also
increase the fission product pbisoning. For these reasons and since the only
advantage attributed to the internally cooled system, lower 1nven£ory, did not
exist, 1t was decided to conduct parameter studies solely for mixed chleride
fuels in an externally cooled systen,

Preiiminary analysis (sec. 4.4.1) indicated that power output per mass
of plutonium increased with increased power. A core power of 600 M{ was choéen
ag 1t is the upper 1limit imposed by existing electric power distribution systems,
Engineering considerations ylelded & minimum external hold-up volume for the
removal of 600 Mi. This volume is so large that it remains essentially constant
over a wide variation of core sizes.

With the external hold-up volume constant a study was carried out on system

7S

¥

 
 

o)

 

mess and breeding ratios as a function of composition of the mixed chloride
fuel, It was realized very early in the study that, at the concentrations of

 the plutonium and uranium chlérides involved, the breeding ratio*was higher

and the critical mass about the same when U-238 was used as a diluent instead
of the other chlorides. The salt of composition 3 Eacl,,2 MQGJQ.and 1 Pu (U)
Cl3, which is the highest concentration of Pu () 013 in the mixed chloride
commensurate with melting point requirements was, therefore, used in the pare-
meter study with variation on the ratio of plutonium to uranium, The analysis
was carried out employing a ten group, one dimensional diffusion theory'method
(sec. 4.2.1) on the bare core system to find the bare core radius, breeding
ratios, and flux energy spectrum, Blanket cross sections were then averaged
over this spectrum to obtain an approximation of reflector savings on eritiesl -
core radius, rientzz has shown the validity of diffusion theory calculations
for fast reactor systems with dimensions greater than 30 em,

| Since the blanket material chosen has a low uranium density, an effort
was mede to lower the neutron spectrum in the blenket to increase the plutonium
production density and decrease the blanket thickness. Position and thickness
of = graphite moderator section, placed in the blanket region, were varied

to study results on btreeding ratio and concentration and distribution of plut-
onfun production as well as the effects reflected back into the core. The
Argonne National Lab. RE-7 code fbr ‘the UNIVAG (Sbc. 4+2.3) was used for this
study employing 13 energy;graups and 7 spatial regions,

V'Refléétor control is possible-forra.high core leskage resctor, such as
in theipresent«design._Aubriaf'study.was performed on the effect of changing
the level Qf & mdlten.lead refiéctcr adjaaent-to_%he core vesscl., These cal- -
culations were then performed mere sccurately employing a 10 energy group, 3
spatial region code on a digital computer. -

B

rp—

 
 

 

be2 CALGULATION METHODS BASED ON DIFFUSION THEORY

 4.2.1 Bare Core Multi-Group Method

The neutron diffusion equation in a bare reactor for the jth energy

- group 1s

1 B%5% 5t ste Ths|giay,

.Béi  Au
gy Jsit ) s, ww{’fg o,

wvhere Zf“ Z:t"“‘, Zj 27 iz the maoroscmpic cross section for removal
from the jth group by inelastie acattering,--az; is assumed to be the cross
section for elastic moderation out of the jth group, F%; is the fraction of
the fission spectrum born in the jth group, and P(1 — j) is the fraction of
1nelasti§a11y scattered neutrons in the ith group which are degraded to the
Jth groub on an inelastic collision.

The calculation of the bare system criticality was therefore reduced to
e tabulation of neutron events with an iteration on the geometric buckling;
B, until & neutron balance was obtained over all energy groups. The calculation
begins with the introduction of one fission neutron distributed over the fission
spectrum. In the Pirst (highest energy) group this is the only source of
neutrons so that the events in this group can be tabulated, Group 1 then firo—
vides the balance of the source for the second group through scattering, hence
the events in the second group can be determined., This procedure was continued
for each lower emergy group. At the conclusion of thé lowest energy group
tabulation of events, the total capture of each element, the mumber of fissions

 
 

£7

1,

 

in plutonium and uranium, and the mmber of neutrons which leak out of the
bare system were found by summation over all energy groups. A new radius was
chosen and the calculation repééted until the neutron production and loss were
equal.

In the calculation just described, st eriticality, the source of neutrons
for each energy group multiplied by the average velocity of that group times
the average time spent in that group is proportional to the flux of that specific
energy group. That is,

flau ~ N; V"J.Tf

Note that - '
T e . S
vhere ;S}ffl=:§%%2r + ¢§£2fl+.‘§:j: *'jgégf s
u #
| | , /
hence
’ ' P ’ a

4.2.2 Reflector Savings Estimate

The flfix energy spectrum obtained for the bare core was assumed, for
the reflector savings estimate, to be the equilibrium blanket flux energy spee-
trum, Averaging blanket paramaters over this spectrum and assuming an infinite
blanket the reflector savings was found to be insensitive to the bare core |
radius and bare core spectrum over the range of interest. For the study of
systen mass, bre;ding.fatids-and flux ensrgy spectrum as a function of the plut- |
oniwm to uranium ratio, the reflgctof savings on the bare core‘radius.was assumed

to be fi‘constant.

- 4e2.3 UNIVAC Calculations
In order to obtain a better representation of the effect of the blanket

 
 

 

énthe core and to gain information on the desirability of a moderstor section
in the blankbt'region,thg Re=7 Argonne National Leboratory code for the UNIVAC
was employed, The iteration in this code was performed on the fuel to diluent
ratio rather than the core radius. The optimm system core radius from the
previous perameter study and seven regions (core, core vessel, lead reflector,
first blanket, moderator, second blanket and graphite reflector) were used.
Extra lower energy groups were employed because of the lover energy spectrum
in the blanket. |

The input information, calculation procedures and resirictions of the BE-V
code are covéred in reference 23. The results of the problem consisted of
the criticel fuel to diluent ratio, the criticality factor, the fiséion source
at each space point, the integral of the fission source over each region, the
flux at each space point in each energy group, the integral flux over each

region in each energy group, and the net leakage out of each region in each

energy group.

=82«

 
 

 

4.3 CROSS SECTIONS

403.1 Energz GI'O“EQ

For the UNIVAC calculations thirteen energy groups were employed.
These are presented in Table 4,1, sectlion 4.3.4. The last four groups were

combined into one group in the bare ten group paremeter study.
Le3.2 Sources of Data

A1l total end fission cross sections as well as the (n, gamma) of
uvranium-238 and the (n, alpha) of chlerine were obtained from BNL-325. The
capture cross section of plutonium was calculated using values of =¢ employed
in réfbrence 2. 7The inelastic.scattering cross section of uwranium and plut-
onium were obtained through & private commutication with L. Dresner of (RNL,
These values were based on the experimental wofk of T, W, Bonner of Rice Imstitute,
M. Welt of LASL end R. 0. Allen of LASL. The sources of other imelastic scatter-
ing cross sections are references 25, 26, and 27. The spectrum of inelastically
scattered neutrons was tsken, for all elements, to be Maxwellisn in form with
the temperature of thé distribution give-n by the equation e:J-E- -~ C’, vhere E
1s the initial neutron energy, b fifisfaSSumsd to.be 20,7 Mevl and constant and
¢ was teken aé 0,08 Mev for high energy meutrons and extrapolated to zero ét
the threshol&. In reference 28, this form is used and gives good sgreement for
incident neutron energies of 1.5,'3 and 14 Mev. |

Mossured values of the transport cross section of carbon, iron, lead and
uranium~238:wbre obtfiined fram-refbrénce 29. A&ditionfil values for these elements
and all transport cross sectio;s‘fq: the other'elements were caleculated using
the angular distribution of scatterea neutrons obtained from refbfenee 30, Cap-

| -83-

 
 

ture eross sections for elements other than uranium and piutonium vere cal-

culated using the method deseribed in Section 4.3.3.

4e3.3 Calculation of Capture Cross Sections

Because of the lack of experimental determination of capture eross
sections at the emergies of interest (.001 to 10 Mev), a theoretical, energy
dependent equation employing persmsters which can be estimeted with some
accuracy was normalized to data by Hughes31 of capture cross ‘sections at 1 Mev,

The equation employed is that appearing as equation 4.2b in reference 32,

U-:;(E) 2‘”*' 2.0 + 1
1+ | |%? l—'/za'

where the functions, I p are g:lve_n by

 

| 1/2
0 s h ; exp (-ipa/h)
P
1 (&) /2 (a"'1+ Py exp ( -tpaf)
(r )
J ¥ w)d +12 Le1f- L :
{—p-% 8 T—% i exp ( -ipr/a )

The penetrabilities

I 2-_‘1‘:

 

Ig |2 for,Q equal 0 to 6 were calculated to be

'Il 2 |I°\2 beos_1_ 4 sin _1_ 2+ cos_;_-bsinl 2 y b=
' - b b , b - -

I,

 

 

 

  

(3b ~1) cos_l_+3b sin __1.__] [Bbcos | = (3v2-1) sin __1._]
b b o

-84~

 
 

 

 

2 (156°=6) boos_1_+ (15b2-1) sin 1
| o b

 

 

+ @51)2-1) cos 1 - (15b2-6) bsin 1
' b

o

2 | - =2
|I° |2 (1-45b2+105b4) cos % + (105b2-10) bsin 1 |
b

| | | . 2
+ ElO%z-lo) beos 1 - (1-45b° +105b%) sin .l]

o
n

b , . b

|5 | ? |% |2 (15-420b° +945b%) boos 1 +(1-1056%+ 94564) sin 1
| b b
2 s g A
(1-105b+ 945b7) cos 1 - (15-420b° +945b") bsin ;_
b

%6 |

|I | {[—14— 210b - 4'725b1*+10395b ) cos 1+ (21—1260%#10395b4)bsin 1] 2
b b

+ B21-E60b2+10395b ) beog 1 - (-1+ 210b2—4725bl’+10395b6)sin %] 2
Note that p/h = 2,2 x 10 on (E/ev)% and
pa/M = 3.23 x 10~4 at /3 (E/ev)%, 1f a = 1.47 x 10713 A1/3 cm
For nuclei where the level spacing has been exp_eriméntally
determined and the relevant energy staf.e of the compound nucleus is not in
the contimmm,D (the level spacing) was dbtazine_d as an average of data from
reference 33, If the relevant state is in jihe contimnm,then D(7 Mev) was
determined from the experimentsl dg.ta po_ints_lin. Fig. 3;5 of reference 32, and
the eqfiation DaC exp(-BE'%) was used with C equal to 10° ov (for 1light nuclei)
and B evaluated from the 7 Mev‘da.ta. "E, is the excitation energy of the epprop-
‘riate compourd nucleus. | | |
The parameter J 2n/h is obtained frcm complex potential well theory and
is plott.ed as a function of atomic we:lght in reference 32.
The equation for the capture cross section was then nofmalized to Hughes'

=85~

 
 

 

' 2
1 Mev cross section data by solving for l:. | ';‘ and Jn were considered to be

energy independent.

4e3.4 .Tabulatién of Cross Sections

 

Table 4.1 lists the énerg’y groups and fission spectrum used in the
thirteen group calculations. A1l the cross sections used in these studies ere

tabulated in Table 4.2 The spectrum of inelastically scattered neutrons

(assumed for all elements to be that of uranium-238) is given in Table 4.3

 TABIE 4,1
Egé gy g;ggpg and Figsion Spectrum
| Fraction of fission
Group Number : Energy Band peutrons born in band
1 oo - 2,23 Mev 0.346
2 2.23 - 1.35 Mev 0.229
3 1,35 ~ 0,498 Mev 0,301
4 0.498 - 0,183 Mev 0.091
5 | 0.183 - 0.067, Mevw 0,025
6 | 0.0674 - 0,0248 Mov 0.006
7 0.0248 -0,00912 Mev 0.002
8 9120- 3350 ev -
9 3350-1230 ev " | -
10 | 1230- 454 ev -
11 454= 300 ev - -
I N 300 -5 ev L

13 '5-0"ev ' -

-86-

 
 

TABLE 4.2

Fission Cross Sections (pawms)

Group Number &239 U 238
1 2.0 0.55
2 2.0 0.40
3 1,75 .02
4 1.65 -
5 1.8 -
6 2.0 -
7 24 -
8 3.2 -
9 4.0 -
10 7.5 -
11 11 -

12 40 -
13 60 -

-8

 
 

continued
Group Mumber Pu
1 .06
2 .10
3 .13
L .20
5 36
6 .60
7 .89
8 1.3
9 2,2
10 49
11 7.0
12 25
13 45

* Assumed values.

Group Number

o M O~ W N

.18
o27
40
57
«70
.90
1.0
2,0
30
2,0

«0007

.0019
.0045
.0097
022
050
Jq1
.19
.33
3.6

.0003
40004,
«0007
0014
.0025
. 00438
.0088
.016
026
042
.16

. 0025
«0059
«015
.038
+061
062
.062
063

017
.037
.085
.21
.51
.76
1.0
2.0

.02
02
«02
.02
.02
.02
.02
02

.02
« 02
04
10

 

¥ Inelastic scattering cross

section for removal from group.

«S8w

1)

 

 
 

continued

Group Number

O 00 N W W N

B B b

Fu
7.0
6.2
6.3
8.1
11
13

15

17
16
26
32
79

13 120

x 3

 

 

Transport Cross-Sectlons® (bams)“

T
6.5
5.9
5.7
7.3

76

9.5

Otr = Otet “H5 Os

Group Number

V. 0 3 O v &~ WD

Fu
055
«055

- «039.

+060
082

.091

098
.10
082

I

+060
. 065

053

072
.098
11

11

11

c1
1.9
1.8
1.5
1.7
2,1
3.0
2.5

3.5

3.6

4.0

45
12
20

»

Table 4.2
Na Mg
1.9 1.3

2.2 2.1
3.9 3.1
4.0 6.8
3.6 6.2
L8 3.8
5.5 3.8

20 3.4
30 3.4
3.2 3.4
3.2 3.4
3.2 3.4
3.3

3

2.0
2.2
2.1
2,6
3.2
4ob
5.7
8,0
7.4
10

5

3.8
3.5
3.4
5.5

[
[

E BE B EEE

9

1.3
1.3
2.8
4ol
3e4
365
3.6
3.8
3.8
3.8
3.8
3.8
3.8

1Q

2,0
2,8
4.0

405‘ |

4o5
4.6
4ob

4.6

4.6
4e6
4o7

: 4'8

Elastic Scattering Removal Crossesection* (barns)

c1

32

31
.13
012

13

o 17

A
a9

«20

Na

42
.51
.38
.34
.30
AL
47
1.7

Mg

027
oS

.28

.55
51
031
31

28

«28

Fe
WU

.18

.088

12
.16
.20
.28
026

Fb
095

+096
.051
.058
096
.10

010

ki

0
037
.38
41
.56

242

b2

43
046
46

c
.51
63
ol
63
71
oT1
o713
73
013

 
 

 

10 .12 .13 .22 027 ‘ 027 ‘ 039 011 046 073

 

11 .28 31 .58 .65 .65 .89 - .26 1.1 1.8
12 .29 .07 .16  .065 .065 .091 .026° .11 .18
% _
0 elamod = 057 3
AU
T

TABIE 4.3 INELASTIC SCATTERING SPECTRUM

 

To 2 3 4 _5 _6 7

From | | - |
1l 04/, 364 377 <157 .058 —
2 - <197 438 268 073 .02/,
3 - 47388 22 W03
4 - — — ST .300 L126
5 | - - — - 703 «297

1))

 
 

L. RESULTS OF THE PARAMETER STUDIES

For an externally cocled sjrste;n, the_#;mdmim power which can be re-
moved is proporticnal to the volume of the holdeup in the external heat ex-
changer. The system mass of plutontum 1s proportiomal to the total of the
system. Hence, an increase in the power removed at a given core volm results
in en increase in the ratio of power to the system mass of plutonivm, There-
fore, the lowest inventory cost ig cbtained with the maximm power out=put.

Engineezji.ng considerations yielded an external hqld.-fu;p volume of 3510
liters for a core power of 600 Mi, which was considered to be the maximm de-
girable. Witk this external volwme constant, & preliminary anslysis was pere
formed to minimize the mass of plutonium. One ten group, bare core caleulation
wvas performed with a wranium to plutonium ratio of wnity in order to obtain a
typical core spectrum. This spectrm was uséd to avemgé core parameters for a
"one-speed” parameter study ef system mass of pltrbnoimn variation with core gize.
The "one~speed” bare core criticalit_y equation is Ly

1] e [ 9% 5
vhere oY a—_gfl a-;dq,__ D= [i/3 S%Pl j_s WR, A nd E Z is the average
macroscopic capture Cross sebtion of the diluents other thanuranimn-238. In
terms of the bare core mass of pfl.utonimn, 9 andtheba.re core redius, R, this

equation ‘becomes |

M . w | — NoP | = "8_ .7"'",28 / o 1 (
s dn gu., 0l [ S 0]

 
 

 

where' - xé 6;—4‘7(1)4" | - ;(_% )+‘(-):€28_ 6':.-“ (])u- I)

| Az is the atomic weight of the zth element
f' is the atam fraction of the zth element in the ealt
o ’q: is the density of the salt in grems Per cu‘bic centimeter
H, is Avagadro's Number times 10'24' |
G 'e sre in units of barns |
M, is in units of grame
R 1s in units of centimeters
considering a reflector savings ofAR the core mass becomes

= Mc [ R—AR ]3

The system mass of plutonium, Ms’ is thus, for en external volume of V.,

M .

’ ’ ’ MS - MI c v
— ot LT R3 ©
3

 

With an external volume of 3.51'x 106 cc, thesé equations mmerically yield
M, = 1,25 x 1048+ 0,232 |R-2R [?, 2.05 x 20" +1,11 x 10°
_ | R R
This equation is plotted as the predicted results on Fig. 4.1.
The reflecter savings, AR, wvas determined from a blanket reflection

coefficie_nt which was obteined by averaging blanket parameters ever_ the core

flux energy spectrum, The reflection coefficlent was found to be 1nsenait1‘.ve-‘

to core radius. Thus & typical AR of 18 cm was used for all cages.

bede? Bare Core Ten Group Peremeter'Studx

For the reasons stated in section 4.1 the study was limited to con~
gideration of a salt of composition 3NaCl, 2Mg012, ( 1 ) PuClB, and
. ' ' l1+x

92~

1)

 
¥/

 

 

( _ x UC1,, where x is the ratio of uranium to plutonium, N(28)/N(49). OCal-
ciifiiions were performed for various values of x to obtain bare core eritical
mass, core flux energy spectrum, internal breeding ratio, and the met core
leakage which was used to obtein the maximum external breeding'ratio.

The feflected core critical mass variation with the reflected core radius
is plotted as Fig. 4.2. Note that the equation for M, in section 4.4.1 is of
the form

M = kR + kR
where‘kg/kl 1s about 10~7 so that for R less than 100 cm the deviation from
linearity should be less than 10 percent. This behavior i1s seen in Fig. 4.2
which is the result of multi-group treatment.

The system mass of plutonium obtained from the multi-group calculations
is given on Fig. 4.1 together with the prediction of section 4.4.1. It is
seen that the shapes of the two curves are similar and £hat the minimums fall
at the same reflected core radius, This indicates the validity of the assumption,
which was made in the preliminary analysis, that the parameters, when averaged
over the core spectrum, were insensitive to a change of core radius,

The system mass of plutonium and thé'breeding'rdtios are plotted as a
function4of x on Fig. 4.3. Core flux-e#ergy spectrums for x equal to 0 and 1
are given as Fig. 4.42 and x equal to 2 and 3 as Fig. 4.4b. The rapid 1ndre§se
of the system mass of plutonium as xrdecreasas frog_z vas considered to far
outweigh the advantages gccrued from the higher breeding ratio and the higher-
flux.energj spectrum., Thus the.Optimum system was‘¢hosen to ocdur with x equal

to 2.

 
 

 

4.h.3 Reflector Control
In & reactor with a high core leakage, control can be affected by

changing the fraction of the out~-going core leakage which is returned. Using
& molten lead reflector 1n vhich the level is varled, the 1argest contribution '
t0o control is due to the creatien of & void surrounding the core., This void
results in some of the neuwtroms reflected by the blenket, which ig now separated
from the core, to reenter the blenket directly. The change of refleetion co~
efficlent due to the separation of the bla.n.ket fram the core is calculated assuming
that the neutrons leave the blanket in e cosine epatial distribution. In terms of
the reflection coefficlent with no separation the effective coefficient with & void
surrounding the reactor core is given by

o te_Be

R+t (1 =)

where R 1s the core radius ani t 1s the thickness of the void shell., The
spproximete values of of , t and R used in the system were oC egual to 0.5,
t & 2.5 cm, and R equalto 92 cm. ‘Eor thege values, ocl is equal to 0.493.
Since the nev core leakage is epproximately one half the core neutron pro-
duction,

Ak -~ de £t 2 0,00k

k o
Atanic Power Deve

 

opuent Associates performed & three region, ten group
calculation to determine A4 k/k for the void control. These results give kfk
equal to 0.016.

bbb Effect of a Molerator Section in the Elanket Reglon
| To determine the effect of a graphite moderator section in the blanket
region, UNIVAC calculations employing seven spatiel reglons end thirteen energy
groups were carried out. For a constant total volume of moderator and blanket,

figl]--'-

43

 
 

 

sl

 

variations were mafe on moderator thickness snd position.

The core flux energy spectrum with no moderator present in the blanket
region was identical with that obtained with the thickest moderator section
used, considered at its closest approach to the coré. Therefbré, the only con-
siderations in choosing an optimum system were the concentration of plutonium
production in the blanket énd the totsl breeding ratio. These two considerations
are shown in Figs. 4.5 and 4.6.

The effect of a moderator section on the outer blanket flux energy spectrum
is shown in Fig, 4.7. The effective capture cross section of uranium-238 in
the outer blanket is 1.45 barns with the moderator section present and 0,68
barns when blanket materiel wes substituted for the moderator.

Over the range of moderator thicknesses considered (0 to 13em), the total
breeding ratio varied only slightly whereas the average concentration of plut-
oniun production increased by a factor of about 1,6 with the average concen-
tration in the outer blanket increasing by a larger factor. Thus the maximum
moderator thickness of thirteen centimeters and the minimum inner blanket
thickness of seven centimeters were chosen for the final system because of
higher1awerage;concentration'and more uniform spatial distribution of the plut—

onium production in the blanket.

 
 

 

 

4.5 FINAL DESIGN

The final aystem, based on the results of the UNICAC calcnlations, con=
gists of the seven spatial regions listed in Table 4.4.

TABIE 4., BEGION DIMENSIONS AND COMPOSITION

Region . Outer boundarz. (em) Composition
1. core 92 3 NaCl, 24gCl,, 0.6 UCl,,

0.3 PuCl,. 4= 2.5 gn/ee

2. core vessel 93.7 a,s,sfumed'to be iron for
nuclear calculations

3. lead reflector 96,2 1iquid lead

4. 3inner blanket 103.2 volume fraction UO2 = 0.50
volume fraction Na = 0.42
volume fraction Fe = 0,08

5. moderator 126.2 graphite

6. outer blanket 139.7 volune frection U0, = 0.54
volume frection Na = 0.44
volume fraction Fe g 0.02

7. graphite reflector 160 graphite

The detailed neutron balance sheet, normalized to one heutron 'absor‘bed,
in plutonium in the core, is given in Table 4.50) of Pu = 2.88 and V of 1°3%

2e5e

Q6.

 
 

i

‘region 1:

fission in Pu .

capture in Pu .

fissions

captures

captures
captures
captures
region 2:
captures
fegion 3:
captures
region 4:
fisslions
captures

| captures

captures

régiOn 5:
captures

regiohféz\
rfissions
captures
captures

captures

inU.,
intG.
in C1
in Na

in Mg
in Fe
in Fb

inU.
inU.
in Na,
1n-Fe |

inC .,

1n‘U.
InU.
1n'na
in Fe

TABIE /.5 NEUTRON BALANCE
neutron absorbed

*® @ o & & ¢ & ¢ o 0. 793
e o o o 8 o & & o 0.207
® & o & ¢ o o & o 0.048

‘0000000000.238

e @ ¢ & o s s o 0.111
e & & o '. ® e o oo 0.005
* & & o & & s 9 0. 011

* o - . * . - * . O.M6
L ® * * * ® . * ® 0.012

C e e e e e .. 0.023
e e e e e e .. 0,437
e e e e .. 0,003
e e e e . 0,04

* *® * * 2 2 o * » . 0‘m2

. | s o o o @ .'. Vo . 0.001 |
¢ & & & * @ o & @ Obm
e & o ® @ o & & o 0.005

00000000000014

neutrons produced

2.284

0.120

0.058

 
 

 

 

 

I

TABIE 4.5 (cont.)

" peutrons afigé;fiéafiikij;f‘ neutrons produced
region 7¢ | |
captures inC . . . . . . + + + o« » « » » 0,001
leskage . ; e o o o s o s s s s e o s o o« 0,055
totals for all regions « . o« o ¢ o o o o o o o 2464 . 2.46)

breeding ratio = 1,09

The spatial neutron flux distribution for each of the thirteen energy
groups is shown on Figs. 4.8, 4.9, and 4.10, These plots are for a core
vegsel thickness of 5.1 cm, and a lead reflector thickness of 5.1 em. These

238 in

values were subsequently reduced in ordér to Increase the fast fissions in U
‘the blanket ani to redfibt=the,parasitic captures in the core vessel and refleéfor.
Energy spectrums of the coré, inner blanket and outer regiéns are showfi
on Fig, 4.11. | |

The nfifiber of fissions occuring below lethargy u vs. u 1s ;lbtted as‘Fig. 4.2,

The total system mass of plutonium is 1810 kg. This extremelj higfi.vélue
1s primarily due to the low density of the mixed chloride-salt-ahd tdrfhé very
iarge external hold-up volume. Because of the low density and the lower thermal
conductivity of most low melting salts, this high inventory is an inherent
characteristic of fused salt systemé. Thé effect of %he high eitérnallhold;up
volume could possibly be improved somewhat by employing a salt with better héat

transfer characteristics,

 
 

m, o

 

 

ORNL LR Dwg. 15408

SYSTEM MASS OF PLUTONIUM VS, CORE RADIUS

Figure 4.1
3400

 
  

Bare core 10 group study
3200

   
 

Prediction

3000
SYSTEM
MASS

OF
PLUTONIUM '

2800 .
(kg. )

2600

2400

2200

2000

1800

L0 60 80 100 122
CORE RADIUS, cm.

=9
-_—

 
 

 

 

1600

"CRITICAL

MASS

(kg of Pu,)

1400

1200

1000

800

600

4,00

200

20

CRITICAL MASS OF )9 IN CORE VS. CORE RADIUS

;o

Figure 4.2

- 60 80
CORE RADIUS, cm,

~100-

A ———

ORNL LR Dwg. 15406

100

  
 
 

120

 
 

 

‘.". 4

}

 

ORNL LR Dwg.l5407

NUCLEAR CHARACTERISTICS PARAMETER STUDY ON A
3NaCl, 2MgCl,, 1 UClB(Pu013) SALT WITH EXTERNAL

HOLDUP VOLUME SPECIFIED AT 12} CUBIC FEET

1.6

Figure 4.3

BREEDING
RATIOS

 

3,00

3200 - 1.h

3000 1.2
KG OF 1,9
in
SYSTEM
: 2800 1.0
2600 0,8
2l00 - 0.6
2200 ; 0.l
2000 0.2
1800 | H0 0
0 1.0 2.0 3.0
| N(28)/N(L9)

-101-

 
 

 

 

Figure 4.4a o ORNL LR Dwg. 15405

CORE FLUX SPECTRUM VS. LETHARGY -

0.2l

 

6
LETHARGY, U

U 1n 10 Mev./E

0.2L

 

2 L 6 8 10
LETHARGY, U

=102~

 
 

®

p

 

CORE
FLUX SPECTRUM VS. LETHARGY

Figure A.4b

 

!
L 6
LETHARGY, U

=103~

ORNL LR Dwg. 15403

 
hhhhh

CAPTURE IN BLANKET
28/ ABSORPTION IN
49 PER MILLION CC.

 

/0 % " ORNL LR Dwg. 15410

CAPTURE IN BLANKET URANIUM PER ABSORPTION IN PLUTONIUM FER
MILLION CUBIC CENTIMETERS VS. THICKNESS CF MODERATCR SECTION

- WITH CONSTANT MODERATCR PLUS BLANKET VOLUME

-¥701-

6 8 10

4
THICKNESS OF GRAPHITE MODERATCR SECTION, cm.

2

Figure 4.5

 
 

 

 

' ‘ ? y ; 0 . 1
10 & ORNL LR Dwg. 15409
. TOTAL BREEDING RATIO VS. THICKNESS OF MODERATCR
~ SECTION WITH CONSTANT MODERATCR PLUS BLANKET
TOTAL BREEDING -
RATIO
1.10
1,09
10 12

6 8

 

0o 2 . 4
THICKNESS OF GRAPHITE MCDERATCR SECTION, em.

Figure 4.6

=601~

 

 
 

 

0.15

0.20

0.15

B

N
N

0.05

 

ORNL LR Dwg. 15412

INTEGRAL FLUX VS. LETHARGY

B R Y .16
- LETHARGY, U

.. Figure 4.7
«106~

 
 

2
L POSITION (.
em. )

Ffi@;xre 15-8

 

-

- ORNL
. LR Dwg, 15
s

 

~L0T-

 
=

 

 

0%

6
4l
LR Dwg. 15
ORNL

m

~80T~

 

L
.

Figure L.9

 
 

"

RADIAL

Figure 4.10

»

 

£ @

ofim. LR Dwg. 15117

 

IR

;60'[—

 

 
 

 

 

INTEGRAL FLUX VS. LETHARGY

g.154h 0}

 

 

0 L 8 12 16
| LETHARGY, U

 
 

'y

 

ORNL LR Dwg. 15411

FRACTION OF FISSIONS EELOW LETHARGY, U, VS.
LETHARGY, U.

1.0
0.8

0.6
FRACTIONS
O FISSIONS:

BELOW U

  

0.4

0.2

 

C 2.0 4.0 60 80 . 10.0
R IETHARGY, U, - =
Figure 4,12

 
 

 

CHAPTER 5 CONTROLS

5.1 CGENERAL CONSIDERATIONS

The control of a fast reactor is no mcrc difficult thac that of & thermal
reactor. Even though the prompt neutron lifetime is much shorter in a fast
reactor, the delayed neutrons are still the controlling factor. It is the nmumber
of delayed neutrons aveileble that determines the ease with thich the reactor
is controlled. In & plutonium fusled reactor there is less than one-half the
number of delayed neutrons thct'are availskie in a reactor using 238 for fuel.
Also; & circulating fuel reactor reduces the effective number of delayed neutrons
aveileble for control because- some are born in the loqp outside the core and are
lost to the system. Therefore, the main difference between the_control of & fast
end thermsl reactor is in the method of control.

One method of control is with the use of & neutron absorber. This method
is not generally satisfacgory for fast reactors_bccause of the low capture
cross sections for neutrons in the high energy spectrum. Ehis.requires that e
largc amount of ebsorber material be moved.in a relatively short tifie. Also,
the conversion ratio ln a fast breeder reactor is lowered.

Ancther mcthod of control is fiith the movement of fuel in the reactor. This
does not lower the conversion retio but does present the additibnal_pfoblems of
heving to remcve fhe heat generated in the qul rod and having to process
the rod. This method is not too practicable in e circulatflng fuel reactor.

The use of & movable reflector appears to be the most practicable method
of controlling & circulating fuel fast breeder resctor. This method has the
disadvantcéé of having to move a large mass of reflector material in e short
perioa of time. It also lowers the conversion ratio slightly. chever, this
method‘cf control was selected for the reactor under consideration in this

project.
"].12‘ ‘ »

4

 
 

 

N

-

»

5,2 DELAYED NEUTRONS

The control of & fast reactor with only pfampt neutrons available would
be extremely difficult because the avefagé lifetime of prompt neutroms im a
fast system is of the order of 10"6 seconds._when_delayad neutrons sre avail-
able, the average neutron lifetime in the system becomes approximately 10'2
seconds. This Increases greatly:the ability to control the reactor in a safe
manner., |

The ffaction of delayed neutrons emitted by the fast fission of plutonium- -
239 1s 0.0023 and of uranium-238 is 0.0176. From the nuclear celeculations
it was found that 5,7 percent of the total fissions sre from uranium-238 so
that the delay fraction, fi?; is 0.0032.. This is the value when the fuel is
not being circulated.

In considering = circfilating fuel reactor, it is obvious that a part of
the delayed neutrons will bé émitted outside the core and therefore lost to
the system, The fraction of delayed neutrons that are useful to the circulating
fuel reactor under steady state conditions can be calculated frafi the ratio of
the average concentration of delayed neutron precursors 1n the core to the
concentration of delayed neutron precursQrs in the core when fhe fuel o
is stagnant. This_fraction for the 1th delay group can be written as follows:37
ol - e}f';fit )(,_C-Nfa)

MY \e e TNy N

where )\71 i1s the decay constant, t3 is the time spent in th‘e,core by the cir-

 

culating fuel, and t, is the time spent outside the core by the fusl. The
average ¢ was found to be 0,519. Since ome dollar of reactivity = d. (3’ =
0.0017, the reactivity dollar has been deflated mearly fifty percent due to

circulation of the fuel.

<113~

 
 

 

., 38

 

 

 

O PeW N

'TABIE 5,1
DELAYED NEUTRONS FROM Pu?3?

T} (sec.) N (sec 1) Bi X3
53,7 0.0129 0.00009 0.462
22.9 0.0303 0.00062 0.462

6.11 0.1134 0.00045 0.464

2.1 0.3238 0.00088 0.480

0.40 1.7325 0.00028 0.709

0.15 4.620 0.00002 0.88,
TABIE 5,2 o0

DELAYED NEUTRONS FROM UR38

 

 

 

 

T+ (sec,) N [ B1 A4
1l 53.0 0.0131 0.00014 0.462
2 22,0 0.0315 0,00178 0.462
3 5.3 0.1308 0.00278 0.462
4 2.0 0.3466 0.00718 0.480
5 0.51 1.359 0.00419 0.657
6 0,18 3.851 0.00153 0.861

-114~

 
 

 

 

The lifetime of promptr» neutrons can be calculated by

 

L-__1 &
| '7))2__; = 0.5 x 10  seconds

[ $dE

vhere
Y S e4 dE
ad 5. J5; ddE
o JedE

In the region below prompt critical, the delayed neutrons determine the
average meutron lifetime in the system.

With circulating fuel’’

L= Z s é i _ + L= 0.018 seconds
i
At

With stagnant fue1>

L- Z gi + L = 0,039 seconds
i

=115~

 
 

 

 

5.3 TEMPERATURE COEFFICGIENT OF REACTIVITY

Thé change in reactivity_du_e_ to a change in temperature is of impbrtance
to the stability and control of the reactor. The largest contribution to this
coefficlent of reactivity is from the expansion of the fused salt. The following
derivation is for an approximate value due to the change in density .of the

salt. ,
DB*) | -
£ - n¥F I~ 5 o | (5.3.1)

where DB2 = leakage cross seetion

and S = total removal cross section (inciuding leakage)

DB*

probability of leakage
2

|- £8* = probability of non-leakage

K
D - > 3.
efine S, = ZR - DB (5.3.2)
Substituting (5.3.2) in (5.3.1) and rearranging we get

£ -t /O | (5.3.3)
{ 2_,./0-1- B*
IfD = L
3%

:
then ) 35, 5
*eE s ey (0

From preliminary core calculations it was found that 3). Zl; = B2 so that
small changes in 3; Zt. in the mmerator of (5.3.4) will not be affected very
much if 3 Zrzt +B? in the denominator is assumed to be a constant. (5.3.4) can

be rewritten és

A= C2 2, | (5.3.5)
and 2).. = lgt N)r;_ 6:1'— = NTG_r- (5-3-6)
and 5 = No ‘ - (5.3.7)

=116~

¥

 
 

 

Q)

 

Since Hr f. N

 

then N, = CoN | | _ (5.3.8)
Substituting (5.3.8) in (5.3.6)
2r~Cy NOT | (5.3.9)
Substituting (5.3.7) and (5.3.9) in (5.3.5)
where 03 - 0102 4y GT'_' |
Reactivity = dk -~ 2 C NdN = Z.ELM . (503011)
k C, N = N
and N«<}Q
so dk |  (5.3.12)
e 2 dp ,

From the curve of fused salt density vs. temperature (°F), it was found

that
d)a s-be? X 10'4 at

The average temperature of the fused salt in the core is 1200°F and the
average density is
| /‘5 = 2.5 g/om3 |
Hence i_kg_a‘zd = =3,3x 10"!* dT
and the temperaim‘_e coefficient of reactivity due to the expansion of the |
fused salt is negative and approximately | -
3.3 x 1074 per: °F
The above approximation was verified 'bj‘_a ten gfoup, three region machine
caleulation which féund the nsgative temperature_ coefficient of r'eactivity-

%o be 2.4 x 107% per °F,

Since there is mo experimental data on the ‘density of the fused salt being’
used in this reactor, it was felt that the high 'Ee'riiperahira densities as obtained
from theoretical calculations were not relieble, The temperature coefficient

«117-

 
 

 

of reactivity obtained using the theoretical densities appears to be on the

. high slde., Therefore, 3.3 x 10~% per °F wvas taken as en upper limit, The

lower 1imit used in simulator studies was 2 x 105 per OF., These values appear
to bracket the coefficients used in the design of gimilar reactors.

There are several other factors contributing to the coefficient of reactivity.
The expansion of the lead in a partially filled reflector due to a rise in’
temperature will give an increase in reactivity. A simple calculatidn was made
to determine the magnitude of this effect. It was assumed that the reflector
was & cylindrical shell 176 em high,

The'dhange in the density of lead due to a temperature changé was found
from Figure 5.140 to be

- 0.,00065 g/em>/°F

Therefore, /o= fi-0.00065 T
where T is the change in temperature from T,.

If the reflector is ome half full at 1200°F and the temperatufe is increased
so the reflector level will reise ome cm, the weight of lead will remain con-
stant, so

2 Tr x zkdfl .2.1Tr(‘L/—:+I)J(f7 0-000(95"7)

Rearranging, T =

 

jO
0 000 bs~ (4 h+t)

= 10,22 g/em’ at 1200F

so T & 177°F rise,

If the total reactivity of the reflector is 0,016, then the average re-
sotivity per cm of height is 0.9 x 1074 per cm, Therefore, a 177°F rise in
temperature will raise the reactivity 0.9 x 107*, The temperature coefficient

~ of reactivity due to ths expansion of the lead reflector is then approximately

-118~

 
10.6

 

 

10.4

 

10.2

 

Density
(g/cn3)

10.0

 

9.8

 

 

9.6

 

 

 

 

 

 

 

 

600

800

CEANGE OF DENSITY WITH TEMPERATURE C(F LIQUID IEAD

1000

1200

1400

Temperature (°F)

1600

1800

2000

Fige 5.1

SHTET="3Ad=YT-TNYO

61T~

 
 

=6

0.5 x 10 andvis'positive. 'This is considerably smaller than the lowest

value of the negative coefficient used for expansion of the fused salt.

- The Doppler effbct‘l is another source of varlation of reactivity with

température.' The overall effect is to increase resonance cross sections with

~ an ircrease of temperature. Thus, the fissions in Pu?3? will be increased

with‘increasing temperature, leading to é positive temperature coefficient of
reactivity, This positiva coefficient is in part balanced by the negative

reoefficient of reactivity arising from the increased absorption in the Pu239.

023§'1ntrodnces a negative coefficient of reactivity so with the proper

balance of the'twb materiaels, the positive éoefficient can be cahcelled cut.'

Tt was found in a U235 system that to obtain a negative temperature coefficient

of reactivity, the ratio of 0238 to U235 nnclei would have to be greater than
1.9. In the reactor being studied, the ratio of U2° to Pu?3? 1s 2.0. Although
no calculation was made for the Pu®-? system, it appears that 1f the temperature
coefficiénfi»df reactivity due to the Doppler effbdt is still positive, 1t will

be small compared to that obtained from the density change in the fused selt.

~1:20-

b

4

 
 

 

5.4 REFLECTOR CONTROL

The lead reflector will be used primarily'for shiu control‘to ccmpensate
for burn-up of the fuel. This will allow the additiun of fuel at fixed intervale
rather than continuously ifr concentration control were used. The cperating
level of the_lead rerlectcr at the beginning of a burn up period will be at a |
point where only 0.0025-of reacti?ity can be added by completely fiiling‘the.;{
reflector. This will ailcw for ebout ten days of operation between edditionejjii
of fuel,

The dumping of the lead reflectcr can be used for normal ehut dcwns of
the reactor. However, the operating temperature of the fused ealt must be .
naintained during ghut down either by decay'heat or by the addition of external
heat. This is tc prevent the reactor from going critical due to the negative L
temperature coefficient of reactivity if the temperature drops. The dumping ‘;ii
cf the fused salt will occur only as an emergency screm or when the reactor -
requiree maintenance. Dumping cf the lead reflector fcr shut dcwn will reduce o

greatly the consequent start up time.:

wl2)e

 
 

 

 

5,5 SDMULATOR STUDIES

Simulator studies were run to determine the stability of the system under
changing load conditions. The load demand fias varied from full load dofin to
1/6 loed in steps of 1/6. The load was then taken back up to one half load
and then to full load, Even though the load changes were msde much faster than
they could be changed in actual practide, the system proved to be veiy stable
under these conditions. This was becauée of the negative témperature coefficient
of reactivity and the large heat capacity of the system. The use of different
negativé temperature coefficients of reactivity only changed the time with
wvhich the system resporded to the load changes. o

Due to & lack of time, no method to hold the steam temperature at its
design point when the load was reduced was simulated. However, there are
several things that can be done, either wholly or in part, to maintain the steamx
temperature. The temperature of the boiler feed water can be reduced by reducing
the amofifit of.steam to the boiler feed water heaters or also the steam temperaturé
can be reduced by attemperation. The auxiliary cooling system could be used
to remove parf of the heat., This design calls for constant speed pfimps but'if
variable speed pumps were avalilable they could be-used'to regulate the steam
temperature. The temperature of the reactor could be varied by the reflector
ghim control but there is a lowér limit to prevent freezing of(the fused salt,

The following diagram shows the design temperatures of the various loops
- in the system at full load. .

122~

1,

 
o

 

 

1350°F 1050%F 1000°F

 

 

 

 

 

>
CORE ~ FUEL Na
—r () !
1050%F 900°F | 550°F

As seen in Fig. 5.2, the reactor power follows the load demand with
practically no overshoot with & negative temperature coefficient of reactivity
of 3.3 x 16”4. There is no noticable change in the mean fuel temperature as

the load demand 1s varied.

5

A negative temperature coefficient of reactivity of 2.0 x 10~ was use&
to obtain the results shown in Fig. 5.4. Even with this small coefficient,
the reactor is stable but requires more time to reach equilibrium after a load
demand change, | .
Fig. 5.5 shows the different temperatures obtained in the system whén
the loed is varied., This is.with no method of controllihg the steam temperature
in the simulator circuit. | |
The diegrem used to set up this reactor system on the simuletor is shown
in Figs. 5.6 and 5.7. I |

-123-

 
°F

Temperature -

Temperaturs ~ °F

1200

—
=
Q
o

1000

900

1300

Ny
3

b
3

 

 

./;17L. ORNL LR Dwg. 15413

L0 60 g0 100 120 140 160
' Time - Secs, o

-

 

 
 

 

1.2

1200

1,0

S

I, = emjereduoy

S

QO
e~

1300

ORNL=IR=Dwg » lu.m“_.rm

g
A

 

o
o

S

- oamgereduo],

1000

Fig,.

160

140

100

60 80
Time - Secs.

w0

20

120

 
1200

°p
B
3

Temperature -
"
o
S
o

900

1300

5

1 .
\]‘

Temperature - °F

B

20

40

Y

 

60

/a6

80

Time -~ secs.

100

120

ORNL LR Dwg. 1541l

140

 

160

1.2

Fig, 5.4

=9e1-

 
 

 

Temperature - Op

700

1400

1300

1200

1100

1000

900

800

 

 

 

ORNL~LR=Dwg ,=1811s7

TEMPERATURES VS POWER DEMAND

=127=

 

Fig. 5.5

 
 

-60V

0.123

 

 

 

 

 

0,889 1.0M 3 0.889
SODIUM CIRCUIT

SIMILATOR DIAGRAM

-128-

ORNL~LR=Dwg.~18148

 

 

7- +(T, T,

 

 

 

0,135

 

1.0M

Fig, 5.6

)

1

 
"

-+ 100V

~100V

0.320

 

 

 

 

 

 

 

 

 

 

ORNL=LR=Dwge=18149

+T1

0,222

 

  

 

 

0. 059

 

 

 

 

Passive //

Networks

o

 

REACTOR CIRCUIT

1.0M

v -

 

 

 

 

I | >l AN
- (Tl—T_Q) '

-11,25V

 

1.,0M

L e

- I> (T

 

1,0M

- 0,222
1.0M"

 

‘Tst.
) =50V
1.0M

=L 0.111

STEAM TEMPERATURE

CIRCUIT

 

 

COUFLING CIRCUIT BETWEEN FUSED SALT AND SODIUM CIRCUITS

1,0M

0.600 .

 

  

 

- =60V

1.0M

 

 

e AAAA—] L
. +L. - L7 L

LOAD DEMAND CIRCUIT

SIMULATOR DIAGRAM

-129-

e

- 1.0M

 

 

Pig. 5.7

 
 

 

 

5.6 STARTUP FROCEDURE

The following procedure is to be used when the core is empty and the re-

actor is to be sterted up. k

1. Bring blanket up to operating temperature by adding heat through | the blanket
fieét exchanger, |

2, Heat fused salt to operating temperature in dump tanks. |

3. With the lead reflector empty and the source in the blanket;_begin pumping
the fused salt into the core, stopping at ifitervals to check criticality.
With the source in the blanket, the multipli¢ation constant is not very
sensitive to the addition of fuel until the reactor becomes nearly eritical,
At this point, more care must be exercised as criticality is approached.
The concentration of Pu must be such that when the core is completely filled
and at operating temperature, the multiplicetion constant is 0,95. The
pumping rate is 5 gpm which is adding reactivity at approximately 0.0001
per seconq. If a positive period is detected while filling the core, the
dump valve will be opened automaticaily. It is estimated £hét‘the solenoid
will operate in about 30 milliseconds and the core will empty in J seconds.

4. After the core 1s filled, finish filling the fused selt loop and start
the fuel circulating punp. Add heat through thé mfiin heat exchanger to
keep the fuel at operating temperature. - -

5. Fill lead reflector to ofierating level, stopping at interwvals to check
griticality._ | |

6. Add Pu to bring reactor critical, This must be added in small amounts at
a point in the loop shead of the heat exchangefto obtain maximum difquiOn
In the salt before it enters the core. This dempens out the fluctuations
of the multiplication constant whiéhbécur whefi the richer fuel enters the

~130-

.. ‘

1)

 
.

core. These fluctuations must not be large enough to put the reactor on a
prompt critical period. S |

T I:t‘ the meen témperature of the reactor is below the .operating temperature
aefter it has gone criticael, continue to add Pu until the reactor reaches

the operating tenmerature; Then control the temperature level by reflector
shim during the burnup period.

 
 

 

CHA 6 _CHEMTCAL PROCESSING
6.1 FPROCESS FLOW SHEETS
6.1.1 Core Processing

The core processing flow sheetAz’ 43, 4 is shovn in Fig. 6.1. Both
the core and blanket cfiemical treatments employ a Purex-tjpe process as an
integral part of their processing-cycles. Since standerd Purex is a relatively
well-developed operation, 1t will mot be explained in detail and is shown as
a single block on the flow sheet. |

_The chemical process for the core is g;van in the following outline:

8. The fused salt is draimed from the core. After ®cooling" at the re-
ector site, it is transported to the processing plant. -

b. The solidified salt mixture is then aissplved in water using‘heat if
required. Proper precautions are employed to maintain subcritical conditions.

¢. Sodium hydroxide is introduced to precipitate the uranium, plutonium,
magnesium, and some fission products as hydroxides. After centrifugation, the
the filtrate solution of sodium chloride and éome fission~product chlorides
is discerded by approved waste-disposal techniques, provided the plutonium con-
tent is low enough. | | |

d. The precipitate is dissolved in acidie solution'buffered with emmonium
iofi.

e. Ammonium hydroxide is introduced to a pH of 5-6 to precipitate the

uranium, plutonium, and some remaining fission products as hydroxides. After

centrifugation, the filtrate solution conteining most of the magnesium is again
disdarded, i1f the Pu content is acceptably low,

f. The precipitate is dissolved in nitric acid solution.

“172-

IY)

 
 

 

'/33 - -

’fi'flcfof Caes Lwertiosl IFRoOLESS S

 Frouw)  IHEET
\//‘0 \/VQ.M

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Bencroe | Fosen | DumP Jank| Sowd | Drissacveg | Frespr7mroe |
(o E | Teer | Tar \

\ i Fr4TRATE
o P 1 P77 Na o7AHets

| ST A OH U, 1, % ’

HA/O.:? . | \ « SorwE ST % r

| Drssocviee |\ | Fecrrzmrog O /350LVER

- | | | wero p#, ByerEE
N /\/;s/

 

 

 

b

»/ O THELS

Al it OFO/COPENE
\ o \ 2% 2056

—eeT-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_@CE_K lZé.‘?iE.‘_’E _ E/’V‘i}; JBECIPITS? 7O 22 B sy % -
N ' 5
| /AP L ¥
/3370 | .!..
APeCODUWETS o
<o | d’b Coi 0 @12 7700 &
v
7z

7o AT
Figure 6.1 | COLE

 
 

 

 

frscron Cusmer Caemicor Firocessms

 

 

 

 

 

 

)

 

 

 

 

 

\\\\<fizjanuzaqfizfiz%mezubf \\\<<:562,

 

 

 

 

 

 

 

 

Bosrw *_/.?‘_(‘E%)i_ Fleecrmyrarop
IMPUCE
Ve XA
r

 

 

 

N Cucocmarron

 

 

e 7% ¢4,
7 7(745‘/-767‘46
CorL

;ZZcuu '~520557”
\C;:{.-OA’
?A"f’cr"'e Ua‘ "/%" - [V/;Pa(’,-yfa( (/a-‘ :;’“ 2 A/lt, - -?5176770&/ , '
Benwwsr | Posre - SIISI0N FLOOUETS e “\\\<gf?cjcz,fig

 

 

 

o,  Fa,
FSISION PROLUETS

 

Y

 

Dsssol vER

 

 

 

 

Aa)| Pueesx |
FRocess \i,"’f s0al

Podbuers
\U{A@ J;

 

 

 

Figure 6,2

=7eT-

TSTYT=* SMA=HT~TNO

 
 

 

g, A modified Purex process is used to oblain decontaminated plutonium
nitrate. Detalls of some of the modification will be discussed in & later
section,

h, The plutonium is precipitated from nitrate solution with oxalic scid.

i. The plutonium oxalate is refluxed in hexachloropropene for 24-48 hours
at 224°C, Impure anhydrous plutonium trichloride (containing carbon) results,

J. The plutonium trichloride is chlofinated with phosgene for eight hours
at 600°C to remove impurities. The plutonium is then in a form which oan be

returned to the feactar eore,

6.1.2 Blanket Processing

The blanket process flow sheet” 1s shown in Fig, 6.,2. The following
outline sumarizes the chemical processing scheme for the blanket:

8. The uranium dioxide-sodium paste is drained out of the blanket by use
of pressure and dilution with additional sodium, if needed.

b, After the paste 1s ®cooled" at the reactor site and then transported
to the processing plant, the sodium is evaporated from the wranium dioxide and
is recovered for re-uss..

¢. The uranium dioxide powder is.confiacted-with ethyl algohol to dissolve
the remaining sodium. This step may not be nacessary,.dependifig on the effiéiency
of the previous step. | |

d. The powder is then dissolved in hot mitric acid.

e. The standafd Purex-proceés is employad-fio ohtain'decéntaminated plut-
onium nitrate. |

f. Steps h through J in the core prbcessing autline are followed to ob-
tain plutonium in a form sultable for use in the core.

=135-

 
 

.

,,_,..‘.,_..__.w“.......__...._.__,w,

 

 

6.1.3 On-Site Fission-Product Removal

6.1.3.1 Off-Cas System

To make provision for the removal of fission-product gases, an
off-gas system must be included in the design of the reasctor complex. 1In
addition, it i1s suspected that some chlorine gas may be given off from the core
material, although the amount will probably be small., -

With the production of some chlorine assumed, the following outlineLB
desceribes an off-gas system on the bagis that some 9.4—yeaf krypton will be
formed and fhat'the reactor will not be located in a desolate region where dis-
persal techniques could be used. |

2. The gases from the circulating fuel loop are removed through a vent
at the top of the inlet plenum to the primary heat exchanger. No compressor
is required, since the core system is under pressure.

b. After passing through a filter and & cooler to remove entrained particles
and vapor, the gases go through a let-down valve. The salt from the filter
and cooler is returned to the core systenm.

¢. The gases next pass through an agueous or caustic scrubber and a silver
nitrate reactor to remove the chlorire.

d. The gases are dried at - 70°F to remove water vapor.

e. After passage through charcoal absorbent beds, &ll rare gases
are retained in the charcoal, If a carrier gas such as helium or nitrogen were
introduced subseQuent to the let-down valve, this gas-would‘then pass through
a CWS filter or equivalent and finally out of a stack;

Periodically, the éharcoal,beds would have tfi.be heated, and the rare
gases thus driven off would be stored in firessufe eylinders. However, if the

..]_36.. ‘

U

"

 
 

/r) .

 

amount of 9.4-year krypton were sufficiently small, buried pipes containing
charcoal at ambient temperature could be substituted for the refrigerated char-
coal beds and would provide sufficient holdup to allow decay of the rare gases

before release to the atmosphere.

- 6,1.3.2 Precipitation of Fission Products

After the reactor hasfibeen 15 operation for a time, ifi is poséibla
that certain fission products (presumably rare earths) will build wp fo con-
centrations exceeding their solubilities in the fused chloride core mixturé.
Thus, it is necessary to consider the removal of precipitating fission-product
chlorides. Lack of experimental data in this area requires qualitative treat-
ment of this problem,

Since fission~product concentrations will be building up rather slowly,
it seems reasonable that they could be kept below their solublility limits by
contimwous processing of & small side-stream from the circulating fuel loop,
This stream would be tapped off from the hot stream leaving the reactor core
end then passed through & small large-tube vertical heat exchanger cooled by
an»auxiliary'sbdium stream, The'chloride mixture would then g§-to one of tfio
filters in parallel where precipitates wofild be removed. Oné filternwculd be
on-stream vhile the solid material was baing removed from the other. |

Te 1nsure'effbct1ve removal of fission-product precipitates, 1t.wcu1d,be
necessary to cool the side stream to & temperature below the mifiimum in the
circulating fuel lodp.- In order to preiént ifitroductien into the éére of a
gtresm with cold spots, the'side stream would be returned to the inlet plemm
of the main heat exchanger, sllowing time for mixing. |

-137-

 
 

 

 

6.1.3.3 Distillation Removal of Fission Products | | .

The possibility arose of the contimuous paf%ialrremoval of low~
‘boiling fiSSidn-product-chlorides from the core mixture by distillation of

8 small side-stream, Again, lack of experimental date prevents quantitative

- treatment of this problem. However, 1@ is at least worthy of mention that it
might be possible to postpone for a long period of time the complete aqueous
processing of the core materialby'means of an on-site continuous distillation

»~

process.

-138- .

 
 

Y

6.2 CONSIDERATIONS LEADING TO FROCESS SELECTION
6.2.1 Processes Considered

Any attempt to select a chemiecal processing treatment for a reactor
systen ag broadly defined initially ag the onerdescribed in this report de-
pended upon the basic problems of core and blanket materials selection. Quite
naturally, every tentative cholce of materisls for either the core or the blanket
necessitated a preliminary investigation of the processing problems involved
in orderftc determine any excessive cost requirements or prohibitive operating
conditions, ‘

Although the nature of the reactor studiqd dictated the general type of

‘material in the core, there was a considerable degree of latitude in choosing

the blanket material, as indicatgd previously in the report. Under consideration
were uranium dioxide-sodium systems and canned solid uranium in addition to

fused uranium salts, Thus the chemical processes investigated included pyro-
metallurgical, volatilizetion, dlectrolytic, aqueous and other processes.

It became evident very early that any evaluation of most of the processes
considered would be hindered by two potent'f&ctors,,viz., lack of experimental .
data end non-existence of relieble cost data. Since the time to be spent qfi_
chemical processing during the course of the projest was limited, it was decided
that studies would be reétricted to those processes on which sufficient experi-

- mental and cost data were available to allow & rea}isticlappraigal.;_Uhfortunately,'

this decision almost automatically eliminated everything excépt agqueous processing,
The ebove decigion; however, was in 1line with the gemeral project philo-

sophy that the reactorsystemdesigned.would be ome for which a capability of

construction might reasonably be expected to exist in the next couple of years,

=139~ -
wm—————

 
 

 

In addition, it was felt that with the low fisslon-product capture cross séctions
in a fast reactor, it might be possible to process infrequently enough to make
aqueous processing économical by employing & centralized proceséing facility.

The vindicatioh of this idea appears later. (See Section 6.3.1). Actually,

this appfoach should be conservative,'since:economics will undoubtédly dictate
that the construction of any type of processing facility have no higher costs

than those estimated at present for aqueous plants.
6.2.2 Process Selection

Among the aqueous processes, the Redox and Purex processes have been
most widélj studied and are feasible for plant-scale construction. Data on
\othefaquéous processes are not widely available, and it seems unlikely that
much cost advantage could be obtained with any of them, Of the Redox and Purex
procésses, the latter is more economical; and thus, it was selected as the basis
for chemical processing to separate uranium, plutonium, and fission products.

" Before the cholce of Purex could be finalized, however, several problems bad
to be resolved.  Since thinking was'in terms of a centralized processing
facility, it was necessary to ascertain whether Purex could be adapted for use
with the core material in order to be able to employ the same basic prdceSs

for both 6ore and blanket, This problem'is discussed in Section 6.2.3; Further,
there remfiined the determination of economical head-end and tail-end treatments
sultsble for the materials in the blanket and the core. Treatments were faznd44 |

and are outlined on the flow sheets in Sections 6.1.1 and 6.1.2.

6.2.3 Purex Modifjcations for Core Processing
Information obtained on.the Purex process43 indicated that it would

be entirely fessible to adapt the process to handle feeds of high plutonium

140~

 
 

 

v

o

content., However, several major problems were apparent imuwdiately;
Processing of fuels with high plutonium content requires elose control

of the Pu concentration to maintain suberiticelity of all equipment, The high

concentration of‘fission producte in the feed poses both chemical handling and

radiolytic reagent degradetion problems, In fhe plutonium separation step,

the reductant (ferrous sulfamate) céncentration.mnst bs stepped up to maintain

the increased smount of Pu in a chemically reduced state. This in turn leads

to & requiremsnt for separating large quantities of ferric ion from plutonium.

The solution of these problems would reguire development work in the laboratory.
Much morerdilute concentrations of uranium_than in & standard Purfix wvaald

have to be employed in order to maintain the low plutonium concentrations
necessary for subcriticality. One msfihod for alleviation of this situation
vhich would bring the process closer to a standard Purex is to recycia uranium
from the uranium strip column. This uranium 1s in & dilute nitric acid solution
and could be used with additional nitric acid to dissoive the hydroxides from
the head-end_treatment, ylelding a solution closer ekin to the conventional |
Purex feed with respect to uranium'concentratidn. To improve this treatment,
a higher initial TBP concentration could be used in starting the Purex to'pro- '
vide a higher uranium concentration in the strip solution and this latter leution
could be further concentrated by evaporation. |

| .Longer fuel cooling times would lessen the decomposition of Purex reagents
by fission-product decay. The increased amount of ferric ion in the final
plutonium solution could be handled by increésing“the’iofi'éxéhéfiég*féciiitieh

employed in the standard Purex firocess;

 
 

 

 

 

~

6.2.4 Alternate Blanket Process

In addition to other processes mentioned, there was anoth.er44 con-
sidered which is worthy of mention in deteil, since it seems to have escaped
mention in the literature and since it has quite interesting possibilities.
The process depends upon the fact that plutonium dioiide becomes refractory at
a much lower temperature (500-60000) than doeajranium dioxide (about 110000)
and most fission-product oxides and hence can be exceedingly difficult to
dissolvé from an oxide mixture, As a result, if the proper dissolution con-
ditions are used upon an oxide mixture in which only the plutonium dioxide is
refractory, the uranifim dioxide and a good portion of the fission-product
oxlides can.be dissolved initially, The resulting matrix can then be dissolved
yiélding a solution which contains mainly plutonium with some fission products.
| Since no experimental data has been found on the decontamination obtain-
able, it is difficult to evaluate this process., However, it is interesting
to consider its use for fast reacfor bred material, where very high decontam-

ination factors are not essential, due to low fission-product absorption cross

sections, One possible limitation of this treatment is that the concentration

of PuO, in U0, must be high enough so that the Pul, will exhibit its own pro-
perties. |

A tentative outline for such a process 1s given as follows:

a. The sodium is removed from the urenium dioxide-sodium blanket material
by volatilization and alcochol dissolution as explained in Section 6.1.2.

b. An oxidation may be necessary to con&ert any plutonium metal to dioxide.

c. The oxide mixture is roasted at 500-600°C to meke the PuO, refractory.

Depending upon the blanket temperature, this may have already been accomplished

 
 

 

g

during irradiation. On the other hand, too high a blanket temperature could
cause all oxides to become refractc_:ry, destroying the feasibility of this
process,

d. The wranium dioxide and most fissifin-product éxides ere dissolved out
of the mixture flth hot concentrated nitric acid.

e, The remaining material, ma:lfily Pu0,, is dissolved in nitric acid
containing fluoride (about 0.1% HF).

f. The plutonium is precipitated as oxalate and treated as described in
Section 6.1.1. )

=143~

 
 

 

 

6.3 FPROCESSING CYCLE TIMES
6.3.1 General

 Thus far, little mention has been made of theyeconomic‘feasibility

of the chemical pfocesses already described. Actually, preliminary cost estimates
were made before final process selection to insure that there would be no
econamically unrealistic choice made, |

The determination of processing cycle tifies re#olves‘largely, though not
entirely, about the question of economics. For this reason, the mfijor_part of
this section will be spent upon economics calculationso Although fission-pro-
duct buildup 1s not as hermful in a fast reactor as in a thermal reactor, this
problem affects the economy of breeding and will be treated.

Before specific discussion is begun, it should now be mentioned and em-
- phasized that all process economics will be based on the use of a large cen-
tralized chemical processing facility. Previous economic studies46 bear out
that the production of relatively cheap electric power requires the minimiéation
of fuel proceséing costs through the existence of a central processing plant
treating the fuel from a number of power reactors. Thus, it should be borne in
mind that all cost figures are based on the assumed availability of such a

processing plant,

‘6.3,2 Effect of Fission-Product and Transuranic Buildup

Although a study of the absorption cross sections of fission-product
elements at high neutron energies indicates that poisoning effects will be
small in a fast reactor, it is important to determine the magnitude of this
47

effect and its influence upon the core process cycle time. P. Greebler at
KAPL has computed an average fission-product cross section from estimated

w1/ li=

 
 

 

n

 

resonance parameter distribution for 023 5 fission in intermediate spectra,
From & plot of Greebler's results, average values of 2 0‘5 for each group in
the final multi-group calculetion were taken &nd averaged over the flux in

lethargy space as shown in Table 6.1.

Table 6,1

Average Fissiog -~ Product Cross Ssctions

2 0p $ U |

Group_(3) (barns) (nomalizod) 200 dau

1 0.029 0.0574 0.00166

2 0.034 0.0780 0.00265

3 0.041 0.1862 0.00763

4 0,057 0.1854 0.01057

5 0.086 0.1809 0.01556

6 0.152 0.1359 0,02066

7 0.318 0.0980 - 0.03116

8 0.695 0.0359 0,02495

9 1.5 - 0,0210 0.03150

10 28 0.027 0.06356

1 5.8 © 0.0047 = __0.02726
T S 0237 barn-zfo_;:f

To achieve sufficient burnup to provid.e two fission-prodnct a.toms for

| eéach plutonium atom in tho core system would require a.bout nine years of operation,

with 600 MW core heat production, & core system' Pu invientory' of 1810 kg, end an
80% 1oad factor. If this situation existed, calculations indicate that the
maximm possible fractional decrease in the external breeding ratic would be

~145-

 
 

 

 

 

’2_0;- = 00237 - 00090 .
G 1+e) 0 21 (17 0.26)
pro#ided that sufficient additionsl plutonium is added to the core to maintain
a constantsintgrnal breeding ratio. Thus,\withjan'initial external breeding

ratio of 0.852 and & constant internal breeding ratio, the decrease in the

. external breeding ratio per year will be

0,090 x 0.852 = 0.,0085 per year
9.0

No adjustment of this value will be made to account for removal of fission-
product gases, since this effect will only be about 10% and since there is
uncerta;nty in Greebler's values.

To account for transuranic buildup, it was decided to make an order-of- |
magnitude correction only, due to limited time. Since it was felt that the
net effect of transwranic absorption and fission would be of the same order-
of-magnitude as the absorption in the fission products, the value for the de-
crease in external breeding ratio per yeer due to fission products was doubled
to include the transurenic effect. Thus, the figure for the over-all decrease
in external breeding ratio will be

2 x 0,0085 = 0,017 per year
6.3.3 Economics and Process Cycle Time Selection

Regarding the reactor site stockpile for replecement of fuel burnup losses,
it‘wasldecided to commence each quarter of the year with & three-month supply

end to allow this quantity to dwindle essentielly to zero befbre'replacement

- at the start of the following quarter.

6 3.3, 2 Core Processing
To determine the optimum.core processing cycle tims, an economic

balance was made betwean processing cost and 1088 1n breeding credit due to )

-.146- |

 
 

 

o

!\

 

fission-product and transuranic buildup., Neglected was the increased pluton-
ium inventory cost with time due to the extra amount required to maintein cri-
ticality as poisons build up. Preliminary calculations indicated that this
was & negligible factor, Chhhéeé in inventory coBtrdue to changes in cooling
time with varying process cycle time were also neglected.

Since the amount of uranium and plutonium in the core material remains
essentially constant with time, the weight of these materials ir the core systém
is

3605 kg of U + 1810 kg, of Pu = 5415 kg U and Pu.

At a cost of $62 per kg. for head-end and Purex treatment, the cost for this
part of one core processing is
5415 kg x $62/xg = $335,700. |

With a cost of $2000-per kg of plutonium for conversion of Purex plutonium
nitrate to plutonium trichloride, the cosfi'for this part of'ofie core proceasing -
is

1810 kg x $2000/kg = $3,620,000.
Thus, the total cost for one core proéeséing at any time is

$335,7oo-+¥$3,620,060 = $3,955,700.
On a basis.of one year, the processing cost per year is

$3,955,700
y

where y is the number of years 1n the core processing cycle time.

‘Each time the core is processed, the diluent salts anfi the uranium |
trichloride must either be replaced or recdvered. Since the goét involved 15.
quite small, it will be computed on the basis that the salts Qill be replaced
after each core processing. Cost figures obtained for sodium chloride, magnesium'

50

chloride and uranium trichloride are 12¢ per 1b°°, 15¢ per 1b°C, and $10 per kedl.

Using these mmbers, the cost of the core material (nmot including plfitonium

“147-

 
 

 

 

trichloride) is $1.73 per 1b of salt mixture. Thus, the cost for onme replace-
ment of this mixture is |
31,770 1b. x $1.73/1b = $54,900

The cost for salt replacement per year is then

- $54,900
Y

From the caloulation in Section 6.3.2, the external breeding ratio de-
creases by 0.017 per year. With an 80% load factor and anc( of 0,26, the re-
actor consumes 217.7 kg. of plutonium per year, using the conversion factor
that 1,0 gm. Pu fissloned = 1 MiD. With & total breeding ratio of 1.09, the
average breeding credit per year 1is | |

| 217.7 x E).o9 - 0.5 (0.017) y] x $15,000 o
where again y 13 the mumber of years in the core processing cycle time and
where the excess bred plutonium is sold back to the AEC for $15 per gram. The

results of selecting different values for y are shown in Table 6.2.

Table 6,2

Core Processing Economiec Balance

Processigg Salt Cost Breedi Net Cost Net Cost
Cost ($10°)  ($10°) Credit ?§1031 (8103) (mils/kwhr)

X

1 3955.7 5449 266.2 3744, 2.06

3 1318.6 18.3 210.6 1126.3 . 0.62

5 791.1 11.0 155.1 | 647.0 0,36
7 565.1 7.8 99.6 473 0.26

9 439.5 6.1 441 0L o022

1 359.6 5.0 -11.4 376.0 0.21
13 304.3 42 -70.0 3.5 0.2l :
15 - 263.7 3.7 =122.5 3899 - 0,21

148~

 
 

 

LiJ

»

 

The results of the cost analysis show that it would be most economical
to process the core about once every 10 years or 850, Howevef, the core §93861
will be replaced every five years, entailing a 1oné reactor shut-down period.
Thus, it was decided to process the core material during this period and avoid

extre inventory charges during cooling-down shipping and processing.

6.3.3.3 Blanket Processing

To seiect the economically optimum blanket processing cycle
time, & balance was made between processing cost and the inventory charge on
the plutonium replacing thaet burned during reactor operation. It is assumed
that the internal breeding ratio will be maintained constant at 0,238 and that

~ the external breeding ratio will decrease with time from its initial value

of 0.807.

With & total burnmup of plutonium in the core of 217.7 kg, per year and
an internzl breeding ratio of 0.238, the net amount of plutonium to be replaced
is

0.762 x 217.7 kg = 165.9 kg Pu/year

again neglecting the extra amount of plutonium which must be added to counter-

‘act the increase in poisoning with time, If an inventory of enough Pu for

three-months burnup 1is aequired at-the.begifining Qf'each quarfer; the initial
stockpile»every quarter will be |

165.9 x 0,25 = 41.5 kg Pu stockpile.

The average amount of plutonium inventory carried due to replacement of
burned_Pu i1s then | |

4105 x[oo5 (_g_.)"‘ 0'5 kg Pu

where 2 is the mumber of months in the blanket processing cycle time. The

plutoniur inventory ecost per kg per year is L% of $15,000. Thus, the total

«149=
/-

 
 

 

 

Pu 1nventory cost due-to_burnup replacement is

0.04 x $15,000 x 41.5 x [0.5 (_Z_)-t—o.fl,
v E T e 3 1

Since the blanket contaifis 15,000 kg of uranium, the blanket processing
cost per processing for the head-énfl and Purex treatments will be

115,000 kg U x $31/kg = $465,000 | ._
since the amount of uranium in the blanketl rémains essentiaily constant with
tims, Peor year, this cost is |

$465,000 x _1__%__ :

In addition, the charge per year for processing the_blankét plutonium from
Furex plutonium nitrate to plutonium trichloride will be
217.7 kg'x:[?.SSZ - 0.5(0.017) _Zé] X $2000/kg.
12

The total processing cost will be sum of the two above costs. The economic
balance as & function of time between blanket processing cost and plutonium

burnup inventory cost is shown in Table 6.3,

Teble 6.3

'Blanket Processing Economic'Balance
Total

Z_ ’ifii‘;fi%eorzp“ (ko) ‘ggsin ve(figgry) gfa::es@s:itfi) _Totel ($10%) (nils/ioihr)
12 103.7 62.2 832,2 - 894.4 0.49

24, 1867 112,0 5961 708.1 0.39

30 228.3 137.0 547,8  684.8 . 0.38

36 269.7 161.8 514.8 676.6 0.37

42 311.3 186.8 490.7 ~  6T1.5 0.37

48 352.7 21,6 4724 684.0 0.38

60 435.7 2614 454 7068 039

=) 50«

Q)

 
 

 

The results show that the economically optimum blanket processing cycle
time is about three or four years. However, as there is uncertainty as to

the corrosion effect that formation of Nazo in the blanket will have, a shorter

- cycle time is desireble. Thus, a blanket processing cycle time of two years

will be selected, since the increase in cost 1s only about 0.02 mil per kwhr.

-151

 
 

/S22

 

 

 

 

 
 

 

CHAPTER SHIELDIN

7.1 GENERAL DESCRIFTION
The shielding of the reactor will consist of steel and standard concrete.
In genersl, component shielding will be omitted; instead a compartmental shield
will be used for the entire reactor cell room, This philosophy of shielding |
was adopted in order to facilitate maintenance and replacement of equipmentQ
Since the entire reactor cell room will be underground (Fig. 3.3), the
surrounding earth will provide additional shielding, however this will not
be fiaken into account in the calculations. The ceiling or top of the reactor -
cell room will have a shield 1.75 feet thick made of stegl. 'The sides of the
cell will have a thermal shield consisting of 4 inches of steel and a bilological
ghield of 6 feet of ordinary concrete. The thermal shield will be made of
3 7/8 inches of carbon steel and 1/8 inch of stainless cledding on all sur-
faces which will require decontemination. Cocoling of the thermal shield was
not designed., However, it presents no problem, elther air or water cooling
may be used. The thermal shiela structure will slso providelthe foundation
for the steel containment vessel over the reactor plant. The biological shield
will be made of orfiinary concrete; all special concretes were rejected due to

52
their high costs.

~153~

 
 

in a leakage flux of 1.0 x 103, 2.1 x 103, 9.5 x 103, and 12.6 x 107

7.2 REACTOR SHIELDING CATCULATIONS

7.2.1 Neutron Shielding

The results of the Univac calculations were utilized in determining

the neutron leakage flux to be attenuated. Only the first four energy groups

- are of any concern with reSpect to shielding. 'Below are tebulated the energy |

groups andrthe neutron leakage.

Tnhle.7.1
leakage Neutron Ener
S - | (neutrons)
J (Group No, ) : : E (Mev) : "~ Lleakage ( sec )
0l oo =2,23 - 23 x 102
02 - 2e23=14.35 4 9 x 108
o4 0.50-0,18 - o 7 x 108

The leakage surfece area of the reactor is 2.3 x 105 cmz. This results
3 neutrons
per cm? per sec for energy groups 01, 02, 03 and 04 respectively, This'leak-
age flux is extremely small, hence the gamma rays will be the determining

factor in the design of the shield.

' 7.2.2 Gamma Ray Shielding

Four major sources of gammea rays exist in this reactor configuration.

These sources of radiation are the prompt fission gemmas, fission product gammas,

capturevéammas, and inelastic scattering gammas.

The number and energy spectrum of the prompt fission gammas per fission
is given in TID - 7004. By using the following equation, the mumber of gammas
in this reactor system is determined:

=154= o

 
 

.l

Ay

Noy=3.1x 10" (E) B(x)

The power density P(r), was taken as 167 watts per c.c. the average power density
in the core. H

During operation, gammas are given off by the fission products. The mmber
and enefgy‘spectrum per fission is given in_fID - 7004. Agaln the average‘
power density of 167 watts/c.c. was used.

Genmas are produced due to captures in the core vessel and the lead re-'
flector. The core vessel was assumed to be steel. Using the average thermal
flux in the core vessel as given by the Univac results, the cross-section'for'
capture and the photons of various energies produced by captures, the mmber
of gammas‘produced was calculated, Using the seme technique, the same was dome
for the lead reflector.

Inelastic scattering gammas are produced in the core vessel, lead reflector,
end in the blanket., Since the high energy gammas are the most difficult to
shield; inelastioc scatterings of only the two highest neutron\energy'groups
were calculated. A major assumption was made cbncerning the energy of the

gamme, produced. Sinpe very meager information exists as to the mmber and

energy of inelastic scattering gammas, it was assumed that the Ol neutron energy

group pfoduced a single 10 Mev gamma and the 02 group, & single 2,2 Mev gamma,.
It is réalized that this 1is aconservativg assugpfion.

.The energy spéctrum or'alligammfis produced will be approximated by four
energy groups; 2.0,.5.6,'7.0 and 10,0 Mev,  The phdtons produced»of_enérgy less

then 1.5 Mev were néglected;-all others were averaged inte the groups above.
A further approximation is that the source of gammas other then core ¥§'s, will

| be takefi as loceted at the outer surface of the lead reflector. The core was

assumed'to have self absorption and some attenuation is produced. Below are

<155~

 
 

 

1isted the number of gammas produced:

Se c—cm2

Table 7,2
Gamma Ray Sources

_ £§gotons
Source Ene Mev sec

o

 

 

1. Prompt Fission - Core

2. Fission Product - Core

3, OCapture - Core Vessel

Pb reflector
L. Inelastic Seattering

Core Vessel

Pb reflector

Blanket

2.0
5.0
2.0
2.0
5.0

7.0

10.0

7.0

2.0
10.0
2,0
10.0
2.0

10.0

3.1 x‘1013
3.3 x 107
1.2 x 1014‘
2.9 x 107
7.0 x 107
6.4 x 107
1,4 x 10

1.6 x 10

4.6 x 10
1.0 x 1010
2.9 x 10
L.6 x 10
7.8 x 10

- 3,7 x 10

Since the 7.0 Mev gammas were much smaller in mumber than those of other

energles, they were considered unimportant. This resulted in the following

total surface sources of gamma rays,

-156~

 
 

 

Table 7,3
Total Gamma Source

.Energy (Mev) | fiphr;)iorsl:c
2.0 1.5 x 10%4
5,0 - 3.3 x 107

10,0 3.7 x 1012

The gemmas are attenuated through the blanket of U02 and Na, the carben
moderator and reflector, the air(which was neglectedl and the shield of steel
and concrete. The spherical source was converted to a monodirectional infinite
plane source and the attenuation calculations were performed using the appro-
priate equations. For a detailed aenalysis of the shilelding calculations, see
Appendix B.

It was found that 4 inches of steel and 6 feet of ordinary concrete or
in the case of the top shield, the 1.75 feet of steel results in a radiation
dose less than the maximum permissable dose of 50 mr/hr., This dose of 50 mr/hr

_was tasken as the maximum permissable dose since rno one will be required to |
spend more than 2 houré per week in a rad1at1on area. This would give the per-
son & total weekly dose of sbout 100 mr/week which is one-third the maximum

pe:l'm:lssable; dose designated by the Atomic Energy Commigsion.

~157=

 
 

 

8,1 GENERAL

The reacior power cost of 6.5 mills/kwhr as presented below, comperes
favorably with conventional power cost. It must be realized however that, in
spite of effdrfis to be on the conservative side, there are a mmber of un-
certainties which when resolved might substantially change the total cost of
reactor power.

A considerable uncertainty exists regarding reliability., The design is

| basically simple, but the high negative coefficient of reactivity combined

 with large temperature fluctuations could result in frequent dumping'or the

core.,

The fuel and blanket processing cbsts were based on a large projected

centralized chemical plant and might be revised upwards in sctual experience.-

The cost.bf operation and maintenance are arbitrarily arrived at since

no experience is available. (See reference 56 and 57.)

~158-

¢,

 
ay

6

8.2

CAPITAL COSTS

The capital costs were predicated on the following assumptions:

1) Cost of material and fabrication

Structural Steel .20 $/1v
Stainless Steel 3,00 $/1b
Ni-Mo Alloy 10.00 $/1b
Heat Exchangers 30 to 50 $/ft2

2) Cost of Installation
Piping 100% of materials and fabrication
Vessels, Tanks and Heat Exch. = 50% of materials and fabrication
Pumps B 25% of materials and fabrication |
3) Overhead and Contingencies 40% of installed cost

Table 8.1 shows capital cost of equipment for the reactor portion of the

plent,
TABIE 8,1

EQUIPMENT LIST AND CAPITAL COSTS FOR THE REACTOR PORTION OF THE PLANT
ITEM DESCRIPTION " 'MFR. COST "~ INST. COST
Core Vessel - 73 1/270.D, 1/2" Wall: $24,,000 $36,000
Core Piping 24"0,D, 1% Wall - $25,500 - $38,300
Core Heat Ex. 3500-1/2" Tubes -~ §390,006 ©  $660,000
Core Pump 2750 GPM 140 FT Head $350,000 $437,000
Core Dump Tanks | 3 -

With heating = 250 FT° $.8,000 $72,000
Core Dump Piping - I - $10,000 $20,000
Core Injection Pump =~ 1 GPM 80 FT Head - $3,500 $4,400
Core Fill Pump 5 GPM 80 FT Head - $3,500 $4,400
Blanket Vessel 120"0,D, 1" Wall $162,000 $243,000
Blanket Piping. - -~ 20%I.D. /2" Wall = = $6,700 $13,400
Blanket Heat Ex. 1570-1/2" Tubes $83,200 $125,000
Blanket Pump 28500 GPM 176 FT, Head . - $300,000 $375,000
Blanket Rem. Eqpt. $50,000 $75,000
Blanket Fill. Eqpt. o . $75,000 $94,000
Sodium Piping 4210,D, 1/2" Wall $99,000 $198,000
Sodium Pumps 4-28,500 GFM 65 FT. Head  $1,000,000 $1,250,000

~159-

—

 
 

 

TABIE 8.1 (Cont.)

. ITEM DESCRIPTION
Boiler 2400-1/2" Tubes
Blanket Graphite
Blanket lead

Blanket Uranium

Remote Repl. Eqpt.

500 Stack

Instr. and Controls

Steel Shell 60 FT, Dia 1% Wall
Reactor Crane '
Reactor Building

Emergency Cooling

Sodium Dumping

Pressurizing and Venting System

MFR. COST
$410,000

$110,000

8.3 LIFE OF EQUIPMENT AND ANNUAL CHARGES DUE TO CAPITAL COSTS

1) Life of core heast exchanger 2 years.

INST. COST
$615,000
$50,000
$10,000
$126,000

-$400,000

$500,000
800,000
165,000
$30,000
$750,000
$500,000
$300,000
$400,000

Anmual fixed charge = Interest tax+4 depreciation = 6+6450 = 62%

2) Life of core, core pump, and core piping 5 years.

Annual fixed charge 6 +6 +20 = 32%

3) Life of reactor plant 10 years.

Annual fixed charge = 6 +6 +10 = 22%

4) Life of turbo-generator and general plant 30 years

Annual fixed charge: 6+ 643 = 15%

8.3.1 Power Cost Due to Capital Cost

1) Core heat exchanger $924,000 based on two years life,

0,924 x 10% x 0.62 w 0.314 mills/kvhr

1.82 x 10°

2) Core, core piping, core pump, blanket pump, and sodium pumps

$3,090,000 based on a five year life.

3,09 x 107 x 0,32 = 0.544 mills/kishr

1,82 x 107

-160-

1

4,

 
 

al

3) All other reactor parts $7,520,000 based on & ten year life.

 

9
7.52 x 107 x 0.22 _ ( 91 mi11s/kwhr.
1.82 % 10%

4) Turbo-generator and general plant at 105 $/kw.

1.05 x 10° x 0.15

3 = 2.25 mills/kvhr.
7 x 10

Total power cost due to capital cost:

8.4

4.018 mills /kwhr

Equivalent capital cost: 7 x 103 x 4.018 = $187 /kw.

0.15

FUEL INVENTORY CHARGES

1)

2)

3)

Plutonium inventory in system 1,810 kg.

Inventory cost at 15 $/gm. Pu 1,810 x 15 x 10° « §27.1 x 106

Inventory charge at 4% = $1,085 x 10°
Plutonium inventory (Average)

Supply to the core 104 kg.

Inventory cost7104 x15 x 10° = $1;56 x 10°

Inventory charge $62.4 x 103

Power cost dfie to inventory charges | |
Total inventory charge $ (1,085+62.4) x 103 :ll,llfl.l\» x 103

Charge per kwhr. 1.1474 x 109 = 0.630 milis/kwhr.
7 x 103 x 260 x 103 o

~161-

 
 

 

 

8.5 FPROCESSING COST SUMMARY

1) The salt will be processed every fivé years,
Cost of processing uranium and plutoniwm with salt 62 $/kg.
Cost of replacing salt: $1.73/1v. |
Total weight of uraniwn and plutonium in salt 5415 kg.
Plutonium processing cost at 2000 $/kg. |
5,415 x 62+31,770 x 1.73+1,810 x 2,000 = § 4,011,000,
Core processing cost: 4,011,000/5 = 802,000 §/year
2) The blanket will be processed every gecond year.
Cost of processing paste 31 $/kg. of uranium.
Plutonium processing cost: 2,000 $/kgz.
Total weight of uranium in paste 15,000 kg.
Total cost of blanket processing:
15,000 x 31+184 x 2,000 = $416,500.
3) Power cost due to processing:
Total processing charge: 802,000+ 416,500 = §1,218,500

Charge per kwhr: _%.285x 107 = Q.67mills/kwhr.
1.82 x 10

8.6 CREDIT FOR EREEDING

Breeding retio: 1.09
Plutonium gain per year: 10.3 kg.

Credit per kwhr: 10,3 x 15 x 10° _ o 065 mi11s ke,
1.82 x 107 -

"162"

 
 

)

‘..\.\_ ‘

 

8.7 OPERATION AND MATNTENANGE

We assumed & one mills/kwhr powver cost due to operation and maintenance.

8.8 COST SUMMARY

The cost of firocudingAelectrical power by the system reported upon here

is shown in Table 8.2.

ITEM
Capital costs

Fuel inventory
Processing

Credit for breeding
Operation and maintenance

Total cost:

~ TABIE 8,2

TOTAL POWER COSTS:

MILS /kwhr. | '
4,018
0.630
0,670
-0.085
1,000
6,233 mils/kwhr.

«163-

 
 

 

CHAPTER 9, RECOMMENDATIONS FOR FUTURE WORK
9.1 GENERAL

In order to determine better the technical and economic feasibility of
a fused-salt fast power reactor system, an extensive program of research and
development would be necessary. The following sections suggest areas in which
important contributions can be made toward the advancement of the fused-salt
reactor technology.

It i= realized that significant technical efforts in certain study areas
mentioned may currently be in progress. However, lack of knowledge of this

work prevents incluslon here. :

4

-164~

 
 

 

 

.

9.2 ENGINEERING

To increase the.feasibility of a fused-salt reactor for power production,
development progrems should be conducted to perfect valves and variable capacity
punps for use in circulating fuel heat exchange loops. To improve the basis
for the use of once-through boilers, it would be most helpful to have better
data for the prediction of_pressure drops and heat transfer coéfficients fbr
two-phase aqueous flow in such bollers.

To treat the problem of heating in a volume which has garma-rays being
produced in it and is exposed to a gamma—ray source, better analytical methods
correlated with experimental data are required.

Further information is needed on the feasibility of making the blanket
paste of sodium and uranium dioxide or other high solid content slurries,
Additional data on concentrations 6btg1nable would also be desirable, as waild
information on the characteristics of equipment used to achieve such high con-

centration.

.165-

 
 

9.3 MATERIALS

Progress.in the fused-salt techfiology necessitates extensive experimental
fiork‘on salt systems. Fhase diagrams for ternary and quaternary chloride
systems containing fuel are almost non-existent and are badly needed. Like-
fiise specific heats, viscosities, fhermal conductivities and other physical

properties of fused-salt mixtures are required to analyze possible reactor

'systems.

Informetion on the physical properties of high density oxide slurries

in sodium should be obtained. The corrosion caused by the presence of Na,0

in such a slurry should be investigated, as should possible remedial techniques

such as addition of anti-oxidants,

Both static and dynamic corrosion-rate data on fused salts in various
structural materiels, especially the new nickel-molybdenum alloys, should be
taken in the temperature range from 900+1500°F. The effects of mass transfer
in heat exchange loops made of these materials should be assessed experimen- -

tally with long-time tests. Scale coefficients of fused salts in different

| materials need to be determined.

!

-166-

4

4

 
)

9.4 CHEMICAL PROCESSING

To eliminate or reduce the requirement for aqueous processing, it would
be adventaegeous to investigate‘tha continuous or semicontinuous removal of
volatile fission-products chlorides by dis‘billa.fion from fused chloride mix~-
tures containing uranium trichleride. It might also be worthwhile to consider
fhe oxidation of 0013 and PuGl3 to UC1, and PuCl, to effect a gross separation
of fuel and fertile material from fission-product and diluent chlorides by
distillation of the more volatile tetrachlorides.

In the case of fused salt mixtures irradiated to 50 or 100% fuel burnup,
studies should be made to ascertalin the effects of high fission-product con-
centrations on mixture properties. Although precipitation and deposition
might occur, this might possibly be employed as a method for removing insoluble
fission-product chlorides from a side-stream which is cooled and filtered.

Experimental work should be done on the aqueous processing of fuels con-
taining high concentrstions of fission products and plutonium. Recycle of
a diluent uranium stream to-simplify the chemical and criticality problems
involved should be investigated. o

9.5 REACTOR CONTROL

Further calculations of the Doppler effect should be carried out to
determine whether it is positive or negative. A detalled study of possible
reactor accidents should be mede in order to define better the control problems

involved in the operation of a fused-salt fast power reactor.

=167~

 
 

 

 

9.6 EGCONOMICS

Many of the cost figures used in making the economic studies in this re-
| port are based on the arbitrarily standardized numbers. In addition, other
. figures have been assumed with rather'weak'basés, due to the lack of good cost
information. Thus, further information developed in the future or new AEC
decisions may change any or all of the cost figures.

In order to determine the feasibility of a fused;salt-reaptor system which
will be constructed and operated at some time in the future, it will be nec- .
essary to make'fiore vallid economic projections in time if any truly realistic
cost study is to be made., The ability to do this will depend largely upon
changes in the amount of govermment fégulation in the reactor fleld, which

are difficult to predict.

-168-

 
 

 

"}

P

“}

 

APPENDIX A - ENGINEERING CATGULATIONS

A.1 CIRCULATING FUEL HEAT EXCHANGER

; |
Total heat load™™ Q = 600 MW x 3.413 x 10° BIU_ = 2.05 x 10° BIU

 

 

MWER R
Fuel flow® Wp = Q = 2,05 x 107 = 34,2 x 10® b
At x G 300 x .2 hr
Fuel area at 20 fps fuel vel.,Ar Wf = 34.2 x :I.O6 - 3.06 ft2
S’V 155::20::36:!:103
Sodtwm flow Wy = __ Q  =2,05x107 =45.5x 105 1b
At x Cp 150 x .3 hr
Sodium area at 30 fps Na vel. 'A'Na. = wHa = 45,5 x 106 = 8.3 f-b

 

E_v" 50::3():t:36::1()3

2

2 _1.57 x .20

Tube area per cel& I r = 0628 inch2
2

Cell ar68? At+ANa x 0628 - 4.848.43 x 062 272 inch2

Ar 3. 06
Tube spacing‘a =] 272 z 792 inch

433
Tube clearancej «792 ~ 500 @ 292 inch which is adequate for welding.

 

Number of tubes, Ay o\ = 3,06 . 3’500_
wd, e 1255/144
P _
8
Prandtl ber for fu Pr = .
an , mmber for e], Blu _61x1 103 = 484

12 x 6.72

Reynolcls8 number for fuel,Re - S = 240 x 20 x 155 x 103 = 15.4 X 103
M

Fiu.sss.el’o8 mumber for fuel Mau = .023 (Re)‘8 (Pr)" ‘= .023 x 2190 x 1. 88 = 94.8

Heat transfer coefficient for f‘uellj =k Mua 93 = 2'{90 Btu/hr ft2°F
D .4_071'2_'
2
Heat transfer coefficient for tube wall,h = K = __ 12 = 2880 Btu/hr £t°°F

-169~
—

 
 

 

- Equivalent diameter for sodiwm Dy = LA = 4 X .19 = .99 inch

b X 25

K

Prandtl number for sodimg,Pr = C M_._., x 1 x 3.6 x 103 = 4o 35 x 10 -3
| —EZ 38 x 16%

Reynolds number for sodium)Re = DeVS = ,99 x 30 x 50 = 810 x 103

M 1x L53x 107

Nusselt number for sodium Nif = 7 +.025 (PrRe) -8 = 7+ .025 x 871 = 28, 8
Heat transfer coefficient for sodiwm h = _K_Nu = 38 x 28,8 = 13.3 x 10 Btu/hr
-9

/7 5

Oversl]l heat itransfer coefficient,

 

 

i=4 . % ol zl2s 11 . 1
Bk R 2,790 ' 2,880 ' 13,300

u = 1,1oo‘B~mhr ft2 Op

Mean temperature differencelfitm = Atf - Atna = 300 - 150 = 216°F

lndtf/ZtNa JQW %(5)—8'

 

required tube length L = Q where

UAAt.m
Aampm g 23,500 x gy = 458 £4%/1%

L = 2,05 x 107 e 18.8 ft

1, 100 x 458 x 216

- Pressure drop8 through 'bubes,APT = ¥ V2 S L

2g

Relative roughness = ,0001, and friction factor f = 028

£_
D

APT 028 x 400 x 155 x 18,8 = 15,200 1b/rt? = 98 ft
. 64 4 x 400/12

'Entrance and exit loss (K +K, ) V2 = 1.4 400 = 8.7 £t

8 61004

Total pressure drop through heat exchenger 106,7 £t

-170-

fi\

££2%F

iy

 
 

 

#)

90

L

 

Minimum allowable tube wall 1:hicls:neessl9

t = DP where P = 150 psi and from ASME Unfired Pressure Vessel Code at 1350°F

28
gllowable stress,s - 3,000 psiy',_‘,t = .5 x 150 = .012&%
2 x 3,000

Fuel hold-up in heat exchanger

2
VF = NT 7TQ_Z_ x L+VP/ where plenmum chambers volume

Vp = Af;Ag x2xL < 3,06-;13,58 x 2 x 75" = 12,4 £t

Vp = 3,500 x Zfic_fif_ x 18.8 +12.4 = 57.3+12.4 = 69.7 £t
X

Heat capaclty of heat exchanger

C = Vy S?c = 3,500 x I (.5% - 40%) x 17.9 x 498,12 = 1,840 Btu
P e Ak OF

4 x 14k

“17-

 
 

 

APFENDIX A.2 CIRCULATING FUEL PIPING AND PUMP

Min S.S. pipe wall thicknessl? ¢ = DP
- 25

whare P = 150 psi and from ASME Unfired Fressure Vessel Code at 13500 allowable
stress S = 3,000 pe:l

t =24 x 150 = .600" use 3/4% plate
2 x 3000

Total Pipe Wall .75 SS+ .25 Ni-Mo = 1,00"

- Fuel velocity in pipe Ve = W. 6

e, = 34e2 x 10" = 23.2 £.p.5.
FE, 155 x 2.64 x 3.6 x 10°

Equivalent pipe length: Straight pipe + 1 ell+2 tees +1 expansion
Joint = 3%, 50"+ 2 x 120+ 40 = 3331
Pipe pressure drop 4 P = _ffi S L
2g D

Re = DV.f = 22 x 23.2 x 155 x 10° = 980 x 10°
M 12 x. 6,72

Relative roughness € = ,00007 and friotion factor £ = .013
D
AP, - 013 x 380 x 155 x 535 x 12 = 3,180 1b/ft° = 20.5 £t
6Lod x 24

Developed pipe length: straight pipe +1 ell +2 tees +1 expansion Joint-+

+pump = 3% 3.15'+ 8'+2,5'+ 4" = 20.65'

Fuel holdup in piping V_ = 20,65 1T 222 = 54.3 f‘t3

P T
Total fuel hold-up core +exchangar +piping = 116,54+ 69.7 + 54.3 = 240.5 f 'f
Total pressure drop of core and external cooling system:

Core head loss+exchanger head loss +piping head loss = 12.4 +106,7+ 20.5

Hd =139.6 £t = 150 psi
6

| Punp horsepower = Wf x' Hd 3= 34,2 x 10" x 132,63 = 3,220 HP
‘Nx33x10 75 x 60 x 33 x 107

~172-

LY

 
 

¥)

90

| -

 

APPENDIX A.3 BLANKET HEAT REMOVAL

Region 1:
. For & core power of 600 M{ operating on a 80% load factor, there will be
175 kg. of plutonium bred per year. Thus at years end, jJjust before processing,
the largest amount of power will be produced in the blanket.
In blanket region 1 (nearest core), the breeding ratio = 0.401. Thus
there are 70.2 kg of plutonium in this region.
 Vq = Volume of blanket region 1 - (4/3)T I:(-SLO.’B.S)3 - (95.8)3]
= 8.67 x 10° cc.

’ (01 = Plutonium density of region 1 = 70.2/ 8,67 x 105

0.0810 gm/cc.

~ mean fission cross-section in region 1 averaged over the flux = 3.87 barns.
Zp1= (.PNO/A)J_E = (0.081 x 0.602 x 3.87)/ 239 = 7.9 x 107 cm
From Univac data: é. 1= 1.28 x 165 (From Section 4.4)

P; = Power in region 1 =V &l @ 1

9

= 8.67 x 10° x 7.9 x 107% x 1,28 x 10
- 8.77 x 10"7 fisstons/sec.
=253 M |
) Now from the muclear celculationss
. P = Power due to U(238) fission = 15.8 My,
fl_g"l‘dtal fission power region 1 = 40.8 MJ

For = cons.ervative calculation we u:lil take & t_otal power '1n regioh 1l afi .
60 Mv, This will :I.ncludé fissions, neutron moderation, and gamma heating.
Q =60 Ms =2.05x 1073 BIU/hr,. - o |
Wga = W/C,AT = 2,05 x 108/0.3 x 150 = 4.55 x 10° 1o fhr.
| .. <173~

 
 

 

 

 

-
"

TmuNaflwama=%Jfihv=mfix1&/flijx%w
- 0.826 £t |
Using 1/2% OD tubes with 50 mill walls

2/h = Trx 0,42/ 4 x Ul

afi = Flow ares per tube = TTD
- 0.000878 £t°
N = Number of tubes c Ap/ay = 0.826/ 0.00087¢
= 940 tubes, | |
Using three rows of tubes equaliy spaced we have 31, tubes per row'wifih the
tube centers on & 97.3, 99.3 and 101.3 cm radii. In the first row the tubes

are on 0,775" centers, 0.79" centers in the second row and 0.805% centers in

the third row. Applying section 3.3.4.1

Row r,  U(BTU/hr-£4-°F)
1 0,500 1285

2 0.445 u30

3 0.500 1285

Therefore U = Average U = 1330 B'I'U/hr-ff:-% |
The effective length of the tubes is eight feet.,

&, = Heat transfer area per tube = (Tx 0.4 x 8)/12
0.838 £t°

T Q/ UN &, =2,05x 108/ 1330 x 940 x 0,838 = 196°F.

' Thus the maximm paste temperature will be 196 °F above the sodium coolent
or an upper limit of 1396°F.

Region 2:
In blanket region 2, the breeding ratio is 0,404, Thus there are 70.7

kg, of plutonium at the end of a year of operatién. ;
<174~

 
 

 

"

(4/3) 7 Em9.3)3 - (115.3)37_]
2.56 x 106 cc.

N
n

6

= 70.7/ 2.56 x 10" = 0.0276 gm/cc.

= 10.0 barns
(PZNO/A_)o—fz = (0.0276 x 0.602 x 10)/ 239
6.96 x 1074 em™L.

MR
N N
i '

K
N
i

- 1.69 x 1014

P, = 2.56 x 106 x 6,96 x 10-4 x 1.69 x 1014

3.01 x 1017 fissions/sec.

® - 8.7 MW.

P = Power from U(238) fission = 1.1 Mw.

Due to the thickness of the blanket and thelafgelabsorption cross-gection of
U(238) for neutrons of low emergy, we took the total power of blanket region
2 to be 40 My, This should lead to a conservative design.

Q = OMw. = 1.375 x 108 BIU/hr

Wy, = 1375 x 103/ 0.3 x 150 = 3.04 x 108 1bs/hr.

A, = 3.04 x 10/ 51 x 30 x 3600 = 0.552 £t2.

t
N

0.552/ 0,000878 = 630 Tubes.

Using three rows of tubes, we have 210 tubes per row with their centers on
131.6, 136, and 140.6 cm, radii, In the first row_the.tfibes are on 1.55" centers,
1.6" centers in second row, and 1.66" centers in the third row.

Applying sectiom 3.3.4.1.

Row " r, U (BTU/r-£4-°F)
1 0.935 o
2 0.950 780
3 o965 780

«175=

 
 

 

Therefore, U = average U - 780'BTU/hr-fét9F
The effective length of the tubes is 10%.
8, =W (0.4) x 10/ 12 = 1.045 £°.
AT - 1,375 x 105/ 780 x 630 x 1.045 = 268°F.
Thus the maximum paste temperature will be 268°F above the sodium coolant or
an upper limit of 1468°F,

APFENDIX A.4 BLANKET HT, EXGR.

This is a sodium to sodium counter-flow heat éxchanger. Naj represents

the primary-blanket coolant and Na, represents the secondary coolant.

Q = 100 Ms = 3.41 x 10° Bty
m o
6
WNalg primery Na welight flow = Q . =341 %x 108 = 7.61 x 10 lbs
CpATy,, 3 x 150 hr

= tube fl W
&y ow &res = ¥Nay = 7,61 x 10° = 1,38 £t2

Nay VNa; 51 x 30 x 3600

where Vual was teken at its meximm "safe" value of 30 ft./bec.la

Using 0.5" OD tubes on a triangular pitch with a 50 mil wall

oy = flow sves per tube = _T 02 = I (4% = 0.126 1n.°= 0.000878 £t°

4 A
N = no., of tubes = AT = 1,38 = 1570
| “a_  ,000878

_ T
No. of cells = 2N = 3140

. ' . 6
Wy.o = secondary Na weight flow = Q = 7.61 x 10 lbs

CpATyay nr
A = secondary Na flow area = Wyoo = 7,61 x 105 e 4.22 ft%

ONa2 vy, 50 x 10 x 3600

-176-

 
 

0

 

In this case V, .. was taken at 10 ft./sec. in order to allow sufficient

Na2
space between tubes for welding in the tube sheets.

a_ = flow area per cell = As = 4.22 x 144 = 0,193 in.z_

 

& 28 3140
= 0433 &. - 0096
| ?
a = 0.815" on tube side \9
| 2
Reygy = P DeVy,4 | ‘ 9
AL
= 24 x 51 x 30 x 3600 = 348,000
12 x .53
Prya = Cp/4 3 x .53 =
‘Nal = = 00425 _ s”
374 . 4=z 0- IS
=7 1 (RePr)' = 7+_1_ (348,000 x .00425)
1\h'i\lal 20 40
= 15.5
- Bay = %}_g_ = 15.5 x 37., 12 = 17,300 BIU
lk .4 : hroft :
On shell side:
DB:AA :AX]. 1220.982“
2 ;_%L _
ReNaz""' .982x51x10x3600-218 000
12 x .53 |
Pryp = 0:00425 f
My, = 7+ (218,000 x .00425)*® = 13.5
- 40 | o o |
= 13.5 x 3.4 12 = 6,170 BT Basing over-all coefficient
"Ne .98 | hr.£t° °F R

on outside areau’ and a stainless steel tube.
2 = o + AO'S +1
U hNal Al kKAw

=177-

 
 

 

17,300 x .4 12 x 12 x .45 6170

 

U= 1,485 __ EIU

N

hr. £4° OF
MID = log mean temperatm-e difference = 150°F |
A = heat transfer area = Q = 341 x 108 = 1530 ft2
UAarT 1485 x 150

L = tube length = __ A = __1530 12 = 7.44 ft.
ND, 1570 x .5 .

Total ghell sres = 2N(As+l/2.At) = 2 x 1570 (193¢ ,096)
: 144 -

= 6,31 £t.2
Inside shell diameter = 2.84 ft. "
Using a 1" steel shell of 316 stainless steel
Outside shell diemeter = 3.01 ft.

AFFENDIX A.5 BLANKET PIPINGC TO PUMP |
Pressure drop in blanket tubes for 1/2" 0D, with 50 mil wall.
N, velocity taken as 20 ft./sec.

Mean tube length = 12.6 ft.

e = (EDeV = 51 x .4 x 20 x 3600 = 233,000
U 12 x .53

_E = relative roughness = 0.000L,
D .

f = friction factor = 0,017

APT-f V2 L = ,017 x20° x 12,6 12 = 40 ft.
T 2gDe ,2x322x.4 |

For the smooth tube bends in and out of the tube sheets, we allowed 1

velocity head loss.

2
APB - 20 = 6.22 ft,
, 2 x 32,2 _

-178=-

u

 
 

 

&)

4P core =AF, +APB = 46.2 ft.

 

We took the worst possible entrance and exit conditions - sudden expansion
- and sudden contraction.
K exp = 0,55 K cont = 0.37

for 2 entrance and exit plemums

° K = 2(K exp#K cont) = 2(.55+.37) = 1.84
. L AR, =K_Y¥_ =le 282 -17.9 4.
‘ Pregsure Drop in Connecting figigg‘

Inside pipe diam. = 16.3"
Na velocity = 30 ft./sec.

Using commercial steel pipe
Re = 1,36 x 30 x 2600 x 51 = 14.2 x 10°
e53

€ = 0,00009
D

£ = 0,012

total length = 10.67 ft. o

AP = .,012 30° x 10,67 = 1.1 ft,
P 2 x 32,2 x 1.63 ‘ T

- Pressure Drop in Elbows !4!

= 1.5 (designed so0) .-

»

- 0.00009

' o
mW M IW

179

 
 

 

4r, = K v a7 30 =o.4§t_.

2g 2 x 32.2

Pressure Drop in Heat Exchanger Tubes

Re = 4,4 x51x30x 2600 = 348,000
2x .53 = S

0.’00014 ’

ol
"

0,016

AP :f v2 _I_-l__"-'

.016 _302 x 7.4 x 12
HE 2g D

2 x 32.2 x 4

 

Pressure Drop in Blanket Tube Sheets

Area ratio = 0.364

Expansion Ke = 0.4 o
- 205 f'tu

A Pce = Ke V2 = o4 202-
2g 2 X 3202
Contraction Ke = .34
4 Pce - Ko 72 - o34 02 : 4.8 ft.
es a Heat Exch T eets

Area ratio = 0.245

Expansion Ke = 0.55

 

AP = Ke v = .55 __30° = 7.7 ft.
HEe 28 2 X 32,2

Contraction Kc = 0 37

AP :KC j___:oB? ____L___'__=51ft
EEc 2 2 x 32,2

‘AP = Total Head Loss = 144.7 ft. = 51.2 psi
Teking 80% efficiency of the pump |

HP = pump horse power = 175 x 7,61 x 106 = 960
60 x 33,000 '

Thus use & 1000 HP pump.
. =180=

e —

- -4905

.

 
 

»

n

APPENDIX A.6 SODIUM PIPING AND PUMPS

a) Pressure drop through heat exchanger

APHE =f v oL where De =099 inches and Re = 810 x 10°

2g D,
Relative roughness = 00006 and friction factor £ = .012
DPE = ,012 x 900 x 50 x 18,8 x 12 = 1,910 1b/ft° = 38.2 ft.
6.t x 0.99
b) Pressure drop through piping
6

AP, = f¥2 SL  where D_ = 42" and Be = 34.2 x 1
2 Do e |

Relative roughness € = .000045 and friction factor £ = .OL

D

AP, = 401 x 900 x 50 x,200 x 12 = 400 1b/ft° = 8 £t
6.4 x 42

¢) Pressure drop through boiler

AP, = £ SL  where Do = 2.22,Re = 1.28 x 10°
B EE 7,

Relative roughness €& = .000025 and friction factor f = 011
| D
4 | w
Pp = »011 x 400 x 50 x 50 x 12 = 924 1b/f%° = 18.4 ft
6ok X 2422 o
d) Total pmp heed
38,248 +18.4 = 64.6 £t

018

 
 

 

APPENDIX A,7 CORE _VESSEL AND REFLECTOR HEATING

Gamma Sources and Heating

/

The gamma sources were estimated by methods given in the Reactor Shielding

Hahdbook.(zo)

Sv (Photons) « 3.1 x 1010 N(E) P (r)
cm356e ~ 3sec )

The prompt gammas were estimated using the equation.

N(E) = Fhoton/fission of energy E (20)

P(r) = Power density (watts)
e
The power density was aésumed constaht and equai to the average of 167 watts/
em’ for temperature rise calculations.,
| The decay product gammas were estimated assuming an 1nf1nite operating
time and the averaged power density. This gamme spectrum was also found in
the Reactor Shilelding Handbook. (20)

These two sources yielded & volume source of gammes of 77.6 x 1012

hotons, of average energy 1.33 Mev, having the energy spectrum as shown in
cm3sec
Figure 3.17.

The gamma hesting in the core vessel and lead reflector due to this source
was estimated using the Integral Besm, Straight Ahead Approxim.tion.( 15) This

approximation ylelds the equation:

g G(; -Sl' (rg) Eg(r ) dn(=, ,B) /7 e -F /L{ei Arii
virs =

Y = Volume Source (photons/cm sec, )

R

P(E) e Energy distribution of photons

 

P .-fL(‘"z fl) = probability that a photon will be emmitted into solid angle d_n_

-182-

4

 
 

 

»

 

 

( 1 for spherical source.)
Iy
_ Sz
ey B L o #e E, P(E) 4
E

» 2 P(E) aE

AsSuming & spherical source the generation rate was found to be:
noo_ oM
¢ (r) = 8, B, 4 e-z J, AY;—#—Z /a Ly, [e-/fe;i A,

 

=y X

w =t
El = b//” € at

—ozide (fl ZAM+ZA E (A z Dr )]

t=b

With the use of this equation and the estimated energy absorption coefficilents

of the salt fuel, Figure 3.18 and stainless steel, Figure 3.19, and Pb(ao) the

gamme contribution to the heat generation was estimated. This heat generation
rete is shown eas a function of distance through core vessel and lead reflector

in Figure 3.20. The averaged gemma heat generation rate wns found to be 3.29 x 1013

Mev 13
P in the core vessel end 1.59 x 10™ Ebvsec in the reflector,

Neutron Sources snd Heating
The neutron sourcee were'taken to be the averaged integral fluxes
in & specific energy group,.as_given by the Unitec-calculations, ovef the'core
vessel and the lead reflector. Usiné these/averaged flnxes the average heat
generation rate was then calculated for the specific neutron interactions of
elastic_seatter, capture, and inelastic scetter. The genma source due to

neutron caepture and inelastic scattering has been negiected.

-183~-

 
 

 

 

The heat generation was calculated for neutron capture in each energy
group by:

st ®PEE So

energy transferred by neutron/incident = m
M+m

e

m

neutron mass

M = target mass

OE) = J6EY) &r
J &

£ - average energy of neutron in the group

ZA capture cross-section, em™L

c

For neutron scattering in each energy group the heat generation equation becomes:

i ZAS () ?6(5)5 cgs

CS g =1-(M-m N
M+ m
A gimiler equation was used for heating due to inelastically scattered neutrons:
- A - = o -
&= 5, @ @(E) E 4 ]
. [(.F
$1 = 1- 72

LI

. (21)

g = MNuclear Temperature

G

m
M

o

*E

Applying these equations over the five neutron energy groups having

average energies of 3.75, 1.82, 0.92, 0.3}, and 0,12 Mev respectively gave
total averaged neutron heat generation rate in the core vessel of 6.36 x 10 3

Mev and in the lead reflector of 1.83 x 1012 Mov

cm3 sec cm3 SeCe

~184~

i

L

 
 

»

&

 

 

Temperasture Rise in the Core Vessel

For reasons of simplicity in calculation the heat generation rates

in the core vessel and in the lezd reflector were assumed to be constants and

the geometry was taken as a slab.

The worst condition was assumed to be that at which the temperature in

the core and the blanket were equal,

weres
(1) et x=0,T; =0

(2) at x

n
o
%
<
N
n
o

(3) a‘bx_-;a,T]_-T
(4) at x =8, K, dT

The appliceble equations were at steady state:

 

 

# B2 TTE (6 -G)

With this in mind the boundary conditionms

Region

(1)
(Fe)

 

 

Region
2)
(Pb)

 

Region (1)
2
K, &0 = -G
1
ax?
Region (2)
32
K, a1, - -6,
dx?
The solution to this set of equations is then:
) '
e Tt
2K Ky
Ty = 52_.(3.'_’:_2. - By (b-x)
2K,

~185«

 

 
 

Blj 1 [Gz -2, 17 -2

T [ECERS

Using the previously mentioned average heat generation rates in the core vessel

-Gz)

 

and the lead reflector of 3.68 cal and 0.676 cal  respectively, the

cm3 sec . : cm3 sec

thermal conductivities and thicknesses of the core vessel and lead the equations

reduce to°

T = 946 x ~ 36,8 x°

1
T, = 46.9+22.4 x - 9.15 x°
(T = %)

These equations yield a maximum temperature of 109.3%F at a distance of 1.286
em into the core vessel.

Since it is believed that a temperature rise this great would cause uflduly
large thermal stresses in the core vessel it was thought that cooling of the
lead reflector would alleviate this.

Using the previously derived equations for temperature in the core vessel
and the lead reflector and assuming that the heat removal rate from the re-
flector could be assumed constant across the reflector and that.the_interface
temperature was equal to that in the core and blanket, a temperature distribution

could be calculated, Thus these equations are:
T, (x) = 48.9 x = 36.&%°
T, (x) = 25.8 X - 131 5% +126.9
We found that the totel heat removal rate was required to be 2.586 cal/'cm3 sec.

The maximum temperature in the core vessel had now been reduced to 29,2°F, 1%

was determined that this caused negligible thermal stresses.
-186- . .

i

 
 

 

 

=

 

The temperature distributions in each of the described cases are shown

in Figure 3.21 and Figure 3.22 respectively.

Reflector Heat Removal Calculation

 

Q= 380x106B‘I'U/hr

Wy, = N weight flow = _Q__ = 3,8 x 1o = 84,500 1bs/hr
o opAT .3 x 150

ap = flow erea per tube « T DI° = _m (,4)° = .000878 £t2
4 4 x 144 |

AT = total Na flow area = Whe = 84,500 = ,0154 ft
T 51 x 30 x 3600 -
Oy

N ~ no. of tubes = AF‘ = ,015, =18

2

 

Taking the outside tube wall temperature to be 1228°F as calculated in
section on core shell heating.

Qzduwfy (T) -Tp) = hy, A1 (T,

S

Where subscript W refers to tube wall properties and Tl énd T

Na)

5 &re tube
surface temperatures.

'L = length of tubing = Q__ 14 T
‘ 'Ir(Tl;_-,THa)- | lw Dw "By Dy

- 80 X 106 T 005 + ;_2 - 102 ft.
11785 l-12 x 5. 17,350 x .4

For the above calculation, we took the minimm wall temperature to maximum

 

Na temperature, This will lead to a conservative result,

The effective length of each tube is 8 ft. Thus, 13 tubes would transfer
the required heat and using 17 tubes will tend to reduce the core shell temp-
erature to a more conservative level

~18 7=

 
 

 

 

 

 

APPENDIX A.8 MODERATCR COOLING CATCULATIONS

Q= 5.1 x 10° BIO
hr

=51x 10 =1.13 x 10° 1bs
«3 x 150 hr

fiT = flow area per tube = 0.000878 ft2

Ap = e = 1,23 x 10° = 0208 £t°
V ' 51 x 30 x 3600 '
&
N - no., of tubes = ,0208 = 24
.000878

To find the maximm moderator temperature,‘ue apfiroximated the rectangular

cell by a cylindrical one of equal area

r2 - 4.22"
rl = 0.25“
ro wn 0.20"
G:Q:i..l_zi_lQ.é.:S.Zleol*BTU
v & hr £t

q = heat removed per tube = 5,1 X 106 e 2,04 x 105 BIU
25 hr

Using the method derived in section 13.3.4.1

4,22 - .25

 

from the ssme section

hife = 17,350 BIU

= 7.75

 

hr £42
hw = 12 x 12 = 2880 BIU
«05 hr £t2 OF
bfised on inside wall ares
_]; - 04 + 0_45 + 1
U .5 x 1112 .5 x 2880 17,350

-188-

BTU = 1112 BTU

 

= 0.00141

 

hr in;q§ h4ft2 op

4

 
 

 

 

= 710 BTU
hr £t2 OF
Ay = heat transfei' area per tube for 10 ft effective length
- LT x 10 & 1.045 £t2
12

 

For a Na temperature of 1200 (this is maximm)

T, = 1050 2,04 x 100 = U75
710 x 1,045

AFPENDIX A,9 STEAM BOILER CAICULATIONS

Q = 700 My = 2.39 x 10° BIU

 

hr
Wy = Q 22,39 x 107 = 53.2 x 10° 1bs
Oria AThe 3 x 150 hr

Water inlet conditions 550°F J 2400 psia
Steam outlet conditions 1000°F, 2300 psie

AH = Hoyy - Hyp = 165 - 549 = 916 PIU

Vg0 = = 2,39 x 10% = 2.62 x 10° s
H 916 hr

Using 1/2" OD stainless steel tubes with 50 mill wail thiékness
ap = flow-a.rea pertube = 127 in? = .00088 P2
Feed water inlet vel, = 7.5 fps
N - no, of tubes = H20 e 262 Aé .
T TH,0 VH,0 °T 46 x 7.5 x 3600 x .00088"
= 2400
Using a trianguler lattice
No. of cells = 2N =« 4800
Using a Na velocity of 20 fps
~189-

 
 

 

 

a, = Na flow area per cell = wNa

&a vNa N
e52.2x20% ____=ussmd 4 etn”

51 x 20 x 3600 x 4800 Q'
435 = W33 8% - J092

a =1,11"

Dividing the boiler into three distinct regions

1 ) _ . - L
1

 

 

———Na

 

 

 

 

 

 

i
- }
A0 — | sub-cooled | Bolling | superheated
l‘ : - i e Nl
Iy | 2 s f
Subcooled Hegting
As & first assumption, calculate without pressure drop
t, = 550°F t, = 662°F
Py = 2400 psie P, ¥ 2400 psia
H, = 519 BTU Hzcfism
1b ~1b
Qe = Wyog (Hy = By)) =Wy, C (T, = T,) = 4uk2 x 10° H
but T, = 900°F therefore T, = 92¢°F
MTDEC = (T - t ) - ( 1) = 66 - = Bozor
ln T, - % 1n 266 -
'1‘1 - tl

DBB = M _’ L x .‘A-ai" = 2,22"

257 _ -
c Pra%s 'ng = fl;g.gxzoxgg 0 = 1,280,000
//Na 12 x .53 |

2190w

 
 

 

Pr. .Op A  =,3x,53 = .00425

Ha&
kNa 37.5

Koy =7+ 1 (Re P.r)'8 =7+ _1 (1,280,000 x .001.25)"8 = 31.5
& 40 40 | |

bpe = Mg Ep. - 31,5 x37,5 x12 = 6380 BTU
Des 2.22 hr.ftz_op

Repog = 42X oA X 7.5 x 3600 = 190,000

12 x .2
Pryo =1
Mo = +023 Re.s P = .023'(190,000)'8 (1)'4 = 380
Ph20 M=M=uoo U
by hr.ft-,

t

basing over-all coefficlent on inside tube ares

 

 

 

 

 

1 = At + a4 + 1 = b 4+ hx(,05) +__1
U Thy A X, A, Byoo 6380 x .5  o12 X .5 3400
= 7.84 x 1074
U = 1275 BTU
ne, £t° °F
A = Qsc 'J='A,gg x 10°_ = 1150 £t.2
U__ (MTD) . 1275 x 302 —
sc sc | |
Isc = ____ Asc = 1350 x 12 = 4,55 ft.
¥r D, 24007 x o4 |
AP =f 72 L = .019 _7,5° x 4,55 x 12 = 2.26 £t
8¢ 2 g De 6&. ol

= 0,66 psi . This is negligible and does not reqfiiré'iterationgl

Boiling
Assuming no presgsure drop

=191

 
 

 

t, = 662°F ty = 662°F | 5

P, = 2400 psia P4 = 2400 psia
H, = 718 BTU '113=1101_1_3_'rg
- 1b 1b

Qp = ¥pyg (Hy - By) =Wy O (T3 - T,) = Uy Ay (MID)y

 

=262 x 10%(383) = 1.0 x 10° EIU

*

T, =T, 4 __Q = 298 + 1,0 x 10° = 991°F
CP WH& ' 03 X 5,32 x 10

4

)

Using the method outlined on Page 701 of Glasstone's Engineering16 | *

c(a1nb/a +1 1 q, + 413
At [g.l]_c_bL .11;] Y, %.1.13 (q/A) | '

¢ = 120 ¥

D 5/ ! J23
i 12 x 12 6380 —J /s 720743 \/a )

-4 0413
= 7.55 x 10 q + 0134 q )
| /s /s

This function is plotted in Fig, (A-1). From Fig, (4-1)¢

At = 266 (q /A) = 330,000

At =329 2 | | | .
A |

| 1 S
A =_Wfia Gp‘ {g 1£ b/a + ‘h}:} In ((fi;: + ,705 T:ixg c [(q/& )z-.587 -éjA )-.587] .
| | g, c* ‘ :

r

7 ¢ .3(7.55 x 107%) 1n 400 + 705 x 5,32 x 10" x ;3 (5.7 = 5.08 x 107%)
' ) 7.2

= 5.32 x 10
- S 330

= 2390 £t°

;’:192.:.-

 
 

.}

at

 

ORNL~LR=-Dwg, =18152

At =7,5%10"4 (q/a)+ .13h(q/a)0 113 ()

1000
800

600
500

00

300

200

100
80

60
50

- 20

 

10
3 4L 5 6 8 105 2 .3 L 5 6 8 106

a/A

Boiling Temperature Difference
<193 | Figure A=l

 
 

 

 

L =. Ap =2390 12 = 9,47 ft.
BDi 24,00 x .4 ‘

for plain H,0

2
AP = 1.37 psi = 198 lbs/ft

Ch20 = Y20 =262‘x106 x 1 =34 1bs__
T Ty T 300 | za1 | see

Using the Martenelli and Nelson Equation(s)

APy = AP, (A Prer +g§
TPF (TT'} r

198(4.7) + .18 344° =1590 1bs_
32,2 £t.2

= 11 psi

This may be neglected for iterative calculations

.

Superheat

Assuming a 100 psi pressure drop

_ o - o

t, = 662°F t, = 1000°F

Py = 2400 psia p, = 2300 psia

H, = 1101 BTU H, = 1462 BIU
1b 1b

Qgy = Voo (B = By) =y, C (T, - T5) = Ugy Agy (m)SH

= 2.62 x 10° x 361 = 0.94 x 10° BIU
Rey,q = sk X 5,12 X 2,42 x 10° = 610,000
12 x ,068 |
Pr_ . =1.8

“HoO | o
Fugoqg = .023(610,000)'8 (1..1:»8)‘4 = 1160

194~

11

 
 

bpo = 169L0L) 55 - 1390 —ETL
. br.£4.° °F

'On the Na side, the heat transfer coefficient is the same as in the subcooled |

region,

1 = L + _ ok (,05) + __ 1 = ,00121
U .5 x 6380 12 x 12 x .45 1390 |
U =826 __ BTU o
hr-£t2-°F L
(o), Ta8) = (T3 -%3) o am . 50 - 149%
. In T, - 1n 331
50

T3-t3

g

gy = 494 X 107 = 7650 £t2
826 x 149 ~

L=__ 7650  x12 = 30.4 ft.
24,00 T ._L (8)

for a compressable fluid

AP =6° (V, -V)) +£ _LG°
g ngarfi

G = W = 344 1bs/ft° sec.
A

V2= Outlet specific volume
V,= Inlet specific volume
f = friction factor = 015
rg= bydraulic radius = ,101 in,

2R’ 86 x 1291
AP = 3447 (.1966) + L005 x 30,4 x 342 x 12 _63.5 pst

32.2 4 x6hidix 0L x 3.5

Thus our assumption of 100 psi pressui'e 'drop is well within the acouracy of this
calculation,
Total length = L, + kg + Loy = 45 £,

- 195~

 
 

APPENDIX B

B. Gamme-Ray Shielding Calculations

B.l Sources of Gammas

B.l.1l Prompt Fission Gamas(53 )
' ’ | 10
' cm'3 - gsec

P(r) = avg. power density = 167 yatts
om>

 

Table Bl Prompt Fission Gammas
v N J 15

 

 

 

 

 

a'-finergy (Mev) N(E) ¥'s/fission P o3 - s
| 1.0 | L 3.2 1.66 x 16~
| 1.5 .8 4.15 x 10%°
2.3 .85 4ol x 10M2
3.0 15 7.8 x 10
5,0 .2 1.0 x 1000
B.l.2 Fission Product Gammas During Qgeration(s 3) |
Table B.2 Fission Product Gsmmas
Energy (Mev) | Byt —tesaes) Mg )
do 2.0 x 10%° 8.35 x 1002
.8  1.2x10M 2.5 x 107
1.3 2,0 x 201° | 2.57 x 1002
1.7 3.3 x 10°0 3.2, x 1072
2.2 | 2.1 x 10%° | 1.6 x 102
2.5 9 x10° .6 x 10™°
12

2.8 . 1 x10° | 6 x 1

<196~

4}

i1

 
s

 

b))

 

 

B.,1.,3 Capture Gammas
A, Core Vessel

The average thermal flux in the core vessel is 2.2 x 1(! ' netzlts
en” ~ sec
Ceptures = ¢ =
vessel 1s assumed to be iron
‘Thermal neutron }(_ N ¥) cross-section = 2,43 barns
The energy spectrum of gemmas 183(53 )

s+photons per 100 captures

0=1 Mev. 1-3 3-5 5.7 _T_
— 10 2 22 50

= A
> =(CAL1 g {7,8[56(,603) (2.43)
< = 0,204 cm'l '

Captures = ¢ = = (2.2 x.loll) (

10

«204)

= 4.5 x 107 captures

cm3 - 3ec

Number of photons produced.

- ‘ 9 -
1-3Mev 110 x4.5x100° & 4.5x 10 Y's
109 . e - sse

3-5Mov 2 2L x 45 x 10 = 12x100 =
5-7Mev 322 x 4S5 x100° = 1.0x100°0 n
100 -
TMev 1 50 X 45 x 108 = 2322000 =

100
highest energy ganima ~ 10,2 Mev

-Be Isad Reflector:

2

§ = 2.3 x 10™ peutrons

cn” -gec 197~

 
 

 

 

Captures = § =

Captures = (2.3 x 1012) ( 5.6 x 10-4)

2=10-%/ fir
(53)

Gamma Spectrum
photons/100 captures
7 at 6,73 Mev
93 at 7.38 Mev

= = (11,34) (,603)
(_'2%6"(7"__2 (_.017)

= .56 x 10~ ant

= 1.3 x 10° captures

3

cm -3ec

Number of FPhotons produceds

Assume g11 J's are at 7.38 Mov energy

then (' 's = 1.3 x 108 gammas

3

cm =S5ec

B.1,), Inelagtic Scattering Gammss

1,

Scatterings

Core Vessel

13

0g; = 1.7 x 107~ peuts

2
cn - -gec

2.9 x 100>

boz :
Cclh =-3SecC

221 = 92.9 x 10~ cm™t

-1

02 , 24,7 x 10~2 cm

Zy

from 01 to 02 neutron energy groups

neuts

;G;_r = 0,17 barns

only.

1)

 
 

Scattering of 01 grdup neutrons produces a 10 Mev ¥ , 02 group neutron gives
a2.2Mev § .
Number of gammas:
Ky =¢ =
- 13 -3y
10 Mev Ny = (1.7 x 10°) (92.9 x 107°)

=1.58 x 100°  X's

cnl -gecg
2.2Mev Ny = (2 ot> -3
. y = (2.9 x10™7) (24.7 x 1077)

= 7.16::10“ ¥ s
3

co” -3eC

2. Lesd Reflector

¢-01 = 7 x 10°° peutrons
en®-sec
¢gp =1.5x 1013 peutrons
cn®-gec
0l
%, = 52.4x107 e
%2 =160 x107 el

Number of J 's:

N ;-,in

@10 ¥ev Ny = (7 x 10-%) (524 x _1_0"3 )

=3.67x10" _¥1s
3 |

em”’-gec

(1.5 x 10°2) (16.0 x 1072)

! |
2.4 x 10 ! J.'s
-~ 3 .

 em”-gec

@2.2 Mev Nb;

=190~

 
 

 

3. Blanket (First half)

¢01 = 3.2 x 10]'2 neutrons
en®-sec

= 8,0 x 10-° neutrons
cm? " se ¢

zi’l = 39.8 x 1of3 P

o2

222 = 32,4 x 10~ e~

Bumber of gammas:

10 Mev Fy = (3.2 x 10%%) (39.8 x 2072

" ‘1{
= 1.27 x 10 's

cma-see

2.2 Mev Ny = (8.0 x 1012) (32.4 x 10-3)
= 2,6 x 10“ 's
3

cm -gec

Blanket (Second half)
11
601 = 1,7 x 10 peutrons

clnz-sec

1
& = Le3 x 10 neutrons

02 cmz-s ec

Zgl = 39,8 x 102 en™?
0 3

2

5 et

2 e 32.4x 10

Nunber of gammes produced:

1 _3
@10 Mev Ny = (1,7 x1C ) (39.8 x 107°)

= 6.8 X 109 a 's

3

cl” =gec

@2.2 Mev By = (4,3 x 10™) (32.4 x 107)

=1.4x100° ¥ 1g
' cmB-sec

-200-
—

t

hl

 
 

 

-~

 

The above sources of gammes will be broken up into four energy groups, 2,

5, 7 end 10 Mev, All gammas of enefgy below 1.5 Mev will be neglected, The

location of the source of all gamfias other than fission end fission product

gammas will be the outer surface of the lead reflector.

Surface ares of source

Sy

" For core

o

ITR? = 13r(96.8)>

=1,18 x 105 cm2
gammas, accounting for self absorption:
= S,A  (55)
=1
P

2,0 Mev ¥'s: p = .29 o

5.0 Mev J 's: p = ,30 emt
3 3 3

Core Vessel Volume = 751,1 cm”3 Pb Refelctor Vol = 1,4 x 10 cm”;

Blanket = 3.4

16 3

x 107 em”,

Converting all the volume sources to surface sources the following is obtaified:

Table B.3 Sources‘qf Radiation

_ , +Fhotons
Energy (Mev) @ __em< - sec

 

Source | o

1. Prompt Fission - Core ‘ 240 , 3.1 x 10l3
, 5.0 - 3,3 x10%?
2. Fission Product - Core | 20  1l2zx0t
3. Capture -~ Gore Vessel | 2.0 o 2.9 x 107
5.0 - 7.0 x 107

7.0 b x 107

| 0.0 1.4 x 205

Pb reflector 7.0 . lbx 106

Le Inelastie Scattering ~ - | o 9

Core Vessel 2.0 Le6 x 10
10.0 1,0 x 10%°

~20]1«

 
 

 

Tgble B.3 (Cond't)

 

. ~#£Fhotons
Source | Energy (Mev ‘ ___ggs:ggg____
Pb Reflector 2.0 o 2.9 x 10° )
10.0 Le6 x 10° |
2.0 7.8 x 1012
& -
Blanket, | 10.0 . 3.7x 1
Table B.,4 Total Gemma Source -
##Photons
Ener Mev cm? ~-_sec 2
2.0 1.5 x 1004 '
5.0 303 X 1012 i .
7.0 6.6 x 107
10.0 - 3.7 x 10
The 7.0 Mev & source will be neglected. - .

B.2 Attenuation of Gamma Reys

Since the source of gammas is at the outer surface of the reflector there

wh)

will be attenuation through the blanket, carbon moderator and reflector, and
the shield,
Blanket sttenuation coefficient:
Blanket - 55% U0, by volume ' T
45% Na
fi =z;; fflf*’J:z/uflQ
 Volume of U02 = .55 (3.4 x 106)

= 1,93 x’lO6 Q.B

f9= 10,3 g

- ~202-

4y

 
 

 

 

Weight of UO2 =1.,99 x .'l.O'7 grems

Mols of U0, = 7.4 x 10%
Volume of Na = 1,5 x 106 cm3
L= .83 e
CcC

1:3 % 10° grams
A

Weight of Na -

L

Mols of Ka 5.7 x.10

it

Total mols in blanket = 12.1 x 10%

Mol fraction of UO2

Mol Fraction of Na = 0,39

= 0,61

Taking only U and Na as effective

2 Mev : Mev 10 Mev
2
Na : p/p = 0427 cme/gm p./p = ,0272 cmafan p./,o = ,0218 cm /gm
' 2 2
Ut wp= .03 en/em  Wp = 0455 em/Em pjp = 0531 cu/em
‘Ener Mev Na | | u
2,0 0.0363 em 7, 0.5072 em ™t
5.0 | 0,023L em™, 04778 em ™
10.0 - P 0.0185 em™t, . 0.5576 om -
@ 2 Mev

EB =£P‘U +762‘_"Na‘ |
pp = (0.61) (.5072) + (.39) (.0363)
Fy = 0.323 |

= (61) (.477) + (.39) (,0231)
om0

oF !

w203~

 
 

 

 

Bg = (.61) (.5576) + (.39) (.0185)
= 0.357

For Carbon moderator and reflector

oy

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2 Mev B o= 043 en’lg -
5 Mev Wpo = 0270 *
10 Mev W= 0195 *
P=1.6 g/cc
2 Mev - g = O7L emt
5 Mev B = 043 omt
10 Mev p = 031 em™t z
Attenuation Within Reactor
Table B.,5 Gamma Attenuation Lengths in Reactor '
Energy Blanket Carbon
(Mev) (™) [t(em) | Rt len™) | t(em) | Bt
2 323 20,5 6.62 071 33 | 2.3 '
10 «357 20.5 7.3 031 33 1.0
Conversion of the isotropic spherical surface source to an infinite plane sources
= 96,5 cn |
/T _ fi_h l3=(17+8)x12'x2.54cm “
I_ r, | | = 760 cm
=204=

 
 

 

 

 

The infinite plane source which will give the equivalent dose at the out-
side of the shield iss

S(lnf plane) = z S (sphere)

Vo

Sa = 96,5 Sy
760

Sa = ,13 Sr

Infinite Plane Sources

2 Mev 13 (1.5 x 100%) = 2,0 x 10%3 _J's
cm2 - SecC

5Mev 13 (3.3 x 10%°%) =43 x 100 o«

12)

10 Mev 13 (3.7x 10 11

= A.8 x 10 f

Attenuation of gammas:

-5 e -Z(px)
g =s e P

A. For 4 inches of steel and 6 feet of ordinary concrete:

10 Mev Blanket - px% = 7.3

Carbon - px = 1,0
Steel (Fe) - px=2.38
Gonerete - b/p = .0229 en/g
| P= 2.3 gfee
 p=.0528 el
6 feet = 182 em
) | plvf- = 9,6
‘j‘:(p'x) = 7.3 + 1.0 -!?2.1. + 9.6
= 207.3
¢10 = (4,8 x 1011)6
= 839 x 10° Ehotons
cn"-gec 506,

-20 03

 
 

 

 

Blanket = px = 6,2
Carbon - p= = 1.4
Steel =~ pX =2,5
Concrete- px = 12,0

S(pa)= 22,1

bs = (43 x 1011) ( e722°1y

¢5 = 6.0 x 10° phgtons_
Cll =58C
2 Mev
Blanket px= 6,62
Garbon pxX= 2,3
Steel px= 3.3

Concrete px= 18,6 |
Z(p =)= 30.8

¢ = (2.0 x 1012) ¢~30-5
2

= 1.l photons (,o.1101116)

om“-sec
B, For steel shield (top)

Using 1,75 feet of steel (5344 em)

0 Moy
WP = 0300 en®

gn
B = 235 em™t |
Sfeel: px = (53.4) (.235) = 12,5
Blanket: p%fl.= = 7.3

Carbons pxr | = 1.0

= /f)_r=20.8

~206~.

L

“

 
 

 

= 550 photons
cmz-sec

Steel pxX= 1301
Blanket px= 6.2

Carbon px = 1,4

20,7
¢5 = 4.3 x 107 72047

= A90 pfiotons

cmz-sec

The 2 Mev ¢’'s are negligible
Dose (Unscattered)
A, Steel and Concrete Shield
Dose = (5.67 x 10°°) Ex) (/) (0 ) r/br
10 Mev |
Dose = (5,67 x 10™°) (10) (0.0162) (8.9 x 10°) x 10°
Dose = 6,0 .mr/hr | |
5 Mev | |
Dose = (5,67 x 10-5) (5) (.0193) (6 x 10° x 103)
Dose = 3.3 mr/hr, |
B. Steel Shield
10 Mev |

Dose = 6.0 x 550
| 890

Dose = 3,7 mr/hr

=207~

 
 

 

 

2 Mev

Dose = 3.3 x 490
600

Dose = 2,7 mr/hr N
Using the build-up factor of water for that of concrete,
(For concrete shield)
Doge (scattered) = B, (p x) Dose (unscattered)

B, (pv) %= 5.0
Dose = 5,0 (6.0 + 3.3) = 46,5 mr/hr

For steel shield
By (px) ~6
Dose = 6(3.7 + 2.7)
Dose = 38,4 mr/hr

=208~

-}

0

™y

 
 

 

 

 

APPENDIX C
EXFERIMENTAL TESTS
C.l1 Summary of Melting Point Tests

Since there was no afiailable data on the melting points of fhe proposed
ternary chloride compositions a series of tests were underfiaken to provide some
fragmentery data. ”

The tests were run in the standard apparatus. This consisted of a nickel
crucible in which the salt was placed, a nickel container (which was provided
with openings for a stirrer, thermocouple, and a dry gas atmosphere) into which
the crucible was sealed. All operations were done in a dry box,

Since there was some doubt that the MgCl, vas anhydrous it wes purified
by the addition of some'NHzgl and heated to its melting point, This succeeded
in remeving the water of hydration from the Mg012 without the conversion of the
M’gGl2 to MgO, This was determined by a peteographic anelysis,

The NaCl., was then mixed to the eutectic composition (60% mol NaCl) and

2
melted as a check on the accuracy of the equipment° The melting point was found
to be 43700 as compared with LSO G the literature valueo

Ueing the above outlined procedure melting points were then determined of
three salt mixtures having (1) 38.6% Mg c1,, 57.91% NaCl, 3 3.49% UC1 4 (2)
36.36% Mgclz, 54.5L% Nacl, 9.09% UC1 4 (3) 33, 33% MgCl,, 50,01% NaGl, 16,663 UCL 5
The deta is summarized below as:
Melting Point

 

| Sample | nguidus ' Solidus

(1)  135°% - 420%

(2) - 432°% £15°¢c

(3) 505°02 440°C% 405°C
-209-

 
 

 

 

C.2 Petrographic Analysis of Salt Mixtures
Petrographic enalysis of the salt mixtures were done by Dr. T. N. HMcVay
et the Y-12 plant. These;analyses afe given beiowe
Sample one: Eutectic of MgCl, -NaGl (40-60 mol %)
Main.pnase nelllorystallized One p above 1, 620 and the other below.
"There is mioroorystalline material present and this has a general in-
dex of refraction below 1.544 (NaCl)., This suggests hydration. X-rays

show neither Mgcl or Nacl.

Semple twos (36.4% Mgclz, 5445 % HaCl, 9.1% 0013)
Sample has colorless phase with brown crystals in it, Brown phase
“has p about 1,90. U'Gl3 is higher at about 2,04, All pheses are

microcrystalline, Semple oxidizes in air and is hygroscopic.

Sample three: (38.6% MgCl,, 57.9% NaCl, 3.5% U013)
Sample has brown compound noted above., Very small lath crysteals of
a brown phase are present., The sample oxidizes and is hygroscopic.

Conclusion: More Data required to properly identify phases.

These analyses show that for the-Mgclz-Hacl eutectic neither the Na¢1(nor
the MgCl exists as such. This is to be expected since the phase diagrem shows
compound formation on each side of the eutectic, The conponnde formed were
assumed to be the expected ones since there was no means of making the complete
- identification,

Both samples containing the eutectic mixture plus U’Gl3 also ehowed'compound
formation, This was assumed since none of the original salts were reoognized.
The salt mixtures were also checked for the preeence of UQIL. This was not found
present as such. o o -

=21 0=

-l

 
 

 

b

W

 

C.3 Summary of Chemlical Analysis of UDlB'

On a wt.% basis 68.8% of the UGL, showld be U'>, The chemicel analysis

3

of the UCl3 used for our teéta showed the 57.1% of the UCl, was U*B; This

3
indicates that the remainder of the U was in the tetravalent state.

C..4 Corrosion Tests

A series of 500 hour, see-saw_cépsule tests containing the chloride mixture
of 33.33%.Mg012, 50,01% NeCl, and 16.67% 0013 were initiated., The tests are
Seing run in capsules of nickel, and of inconel, The results of these tests
are not yet available, but are expected by September 1, 1956,

As an adjunct of this test the chloride salt mixture will be chemically

analyzed as & further check on the exact composition.

211

 
 

 

 

1.

2.
3.

b

5

13,

14.
15,
16.

17,

18.

19.

20,

REFERENCES
Blankenship, F. F., Barton, G. J., Kertesz, F., and Newton, R, F.;
Private Communication. . |

Barton, C. J., Private Communication, June 28, 1956,

 ORNL 1702, Cohen,.S. I., and Jones, T, N., "A Summary of Density Measurements

on Molten Floride Mixtures and & Correletion for Predicting Densitles of
Floride Mixtures",

Untermeyer, .,_"An Engineering Appraisal of. Atomic Power Costs®, Nationsl
Industrial Conference Board Third Annusl Gonference (1954)

TID-67, Weberg G Loy "Problems in the Use of Mblten Sodium 88 & Heat Trans-
fer Fluid®, (1949)

- TID=70, Webery, C, L., "Problems in the Use of Molten Sodium &s &8 Heat Trans-

fer Fluid®, (1951)

ORNL 1956, Powers, W, and Ballock, Ges "Heat Capacities of Mblten Floride
Mixtures®, :

Reactor Handbook, Volume II.
ANL 5404, Abrsham, B, and Flotow, H,, “UOZ—Ha Slurry®.

TID-5150, Stavrolakis, J, and Barr, H,, ®Appreisal of Uranium Oxides",

CF-53-10-25, ORNL, "Fused Salt Breeder Reactor", ORSORT summer project, (1953).

MIT 5002, "Engineering Analysis of Non Agqueous Fluld Fuel Rezctor®, Benedict,
M., ot al, (1953). o 4

 

McAdams, W, H,, "Heat Transmission®, 2nd Edition, McGraw-Hill,.(1952).
Alexander, L. G., ORSORT notes, "Integral Besm Approximation Method®,

Giésstoneg S.s; Chapter 1], ®Principles of Nuclear Reactor En ineérin ",
Ven Nostrand, (1955). -

ORNL 1777, Hoffmen, H, and lones, J., *Fused Sslt Heat Trensfer".

‘Reactor Handbook, Volume 1II,

_Roark, R. Jey ”Formnlae for Stress and Strein®,

%ID-?%OA, ¥Reactor Shieldin Desi Manual®, Chapter 3, Rockwell, T, et al,
1956

212w

-
$

L4 ]

 
)

8

 

22,

23,

2l

254

30,
31,
32,
33.

34,

35,

36'.' |

37.

3,

39.

 

~ REFERENGES (Cond't)

IRL-84, Bjorklund, F. E., "Spectrum of Inelasticelly Scattered Neutrons®,

(1954).

Okrent, D., et al, "A Sury e Theoretical and Experimental Aspects
of Fast Reactor Physics", Geneva Pgper A/Gonf &/P/609 USA, |

Butler, M. and cook, J. "Prelimin Information Relative to RE-6 7 8,
6, 28, URIVAC Multigroup Codes", ANL-5437

R-’259,'Project Rand, Safonov, G., "Survez of Reacting Mixtures Emplonng
Pu=239 and U-233 for Fue d H.O Dzo, C, BeO for Moderator®,

Walt, M, and Beyster, J. R., Phys. Rev, 98, p. 677,

Beyster, Jo Ro’ PhYB RGV. 98, po 1216 o
Teylor, E. L., Lonsjo, O., Bonner, T. We, Phys, Rev. 100, Pe 1Tk

CF-5%-7—165, ORNL, Dresner, L., 'Resume of Inelastic Sca.ttering De:te. for

ARCD=2904, Feshback, H., "Fgst Neutron Datg®, (1950). |

Langsdorf_, ANL, Private Commumication, subsequently pfibliahed as ANL 5567.
Reactor Handbook, Volume I (aecret edition). |

Weinberg, A, M., Higner, E., Notas (Unpubliahed)

Endt, P, M., and Kluyver, J G., Reva of Hod Phys., (1954) 26, No. 1.

Blatt, J. M., and ueisskopf. V. F.y ”!hgoretical HNuclear thsics" wney, (1954).

H].‘l‘-500, M, I.T., Goodman ot al, Ghapter I, "Euglear Problems of R”on-gueons
Fluid F‘uel Reactors®, (1952). R o

m.-5321, Okrent, D., "On the Application of Multigroup Diffusion Theo

to Fgst Critical Assemblies®, 119525. |

Cole, T. E., and Walker, C. S., Reactor Gontrols Lecture Notes, ORSORT, (1956).
chafpie, R. A., ot al, Chapter 7, % esg_in Nuclaar En neerin .FPhysics and
Mathematics“ London Pergamon Prpss, %19565. _ ,

MIT-5000, M.1.T,, Goodman, et al, Ohapter 5, "fluclgar Problems of Non-Agueous
Fluid Fuel Reactors®, (1952).

Liquid Metals Handbook, AEG and Dept. of the Navy, (1955).
-'213-

 
 

53.
54.
55.

56,

57,

REFERENCES (Cont'd)

Feshbach, H., Goertzel, G, and Yameuchi, H., "Estimstion of Doppler Effect
in Fast Reaetorg“ Kuclear Science and Engineering, Vol I, Ho. 1.

LA-1100, Ghemistgz of Uranium end Plutonium ’ (1947)
Flanary, ‘Je Res ORNL, Privsate Gommunication.

Case; F. N.; ORNL, Private communication.‘
Stockflale, W. Goy ORHL, Private Gommnnication.,

MIT-5005, M, I.T., “Processin of . S ent PaweréReactor Fuels", Benedict, oy
et 1, (1953). - | - |

Charpie, R, &, et al, Ghapter II, "Pro ess in NnclearéEn neering Ph ics
and Mathematies®™, London Pergamon Press, 119565

Ullmen, Jo W., ORNL, Private Gommnnication. S | ’
%011, Paint and Drug Beporter“ August, 1956 B |

Barton, C. J.s ORNL, Private Gommnnication, August 10, 1956

Lane, J, A,; "How to Design Resctor Shields for lowest Cost®, Rucleonics,
June, 1955, VoiTHi3Z'7FZEEEL"""""""""'""""""'""“"'

Toid , 20, o - T

ORNL 56-1-48, Blizard, E. P., “Shield Design®,

Glasstone, S;; "Prinei les_of Ruelear Reactor'Eh ineering® Ghepter°10,
Van Nostrand, (1955). | L .

Lane, J, A.; "Economlos of Nuclear Power", Gensva Paper A/Conf 8/P/L7T6 USA. .

 

Davis, W, K.y ¥Capitol Investment Regquired for Nuclear Energy," Geneva
Paper A/Conf 8/P/LT7 USL, e

_214_

LH