ORNL-CF-57-8-7.txt

 

EAS

N AND F

DESIG

 
 

 

 

 

OAK RIDGE SCHOQL OF REACTOR TECHNOLOGY 53'77 t%% h
T ims cxorument consist ol 1 9\3 Lo

o 2O of B3 A9pas; &Maaafiimifi
Reactor Design snd Feasibility Study

 

   
 

"HIGH PERFORMANCE MARINE REACTOR"

Prepapred by:

K. H. Dufrane, Group Leader
T. G. Barnes

C. Bicheldinger

W. D. Lee

N. P. Otto

C. P. Patterson

T. G. Proctor

R. W. Thorpe

R. A, Watson

 

August 1957

 
26,
28,
29,
30.
31.
32.

33.
34,

36.

37
38.

39.

40 '”ll'l .

 

A. M. Weinberg
J. A. Swartout
R. A. Charpie
W. H. Jordan
Lewis Nelson
D. C, Hamilton

 

E. P. Blizard

L. B. Holland

W. B, Kim'ley
W. R. Gall

A. P, Frags
J. A. Lane

P. R. Kasten
A. F. Rupp

E. 8. Bettis
C. E. Winters
R. B. Briggs
A. L., Boch
W. T. Purgerson

R. V. Meghreblian

L. R. Dresner
K. H. Dufrane
T. G. Barnes
C. Eicheldinger
W. D. Lee
N. P. Otto
C. P. Patterson
T. G. Proctor
R. W. 'I'hOl’pe

- A, Watson

ho-h3, REmp Library

4hohs.
4666,

O7-72. ORSORT Files
73-87. TIsE

88.

P. P. Eddy, Maritime Reac

 

 

 

Martin Skinner

Central Research Library
Laboratory Records

 

 

 
 

 

 

 

PREFACE

In Septenber, 1956, a group of men experienced in various scientific
and engineering flelds embarked on the twelve months of study which culmi-
nated in this report. For nine of those months, formal classroom and
student laboratory work occupled 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 a summsry report of their effort. It must be realized that,
in so short s time, a sgtudy of this scope can not be gusranteed complete
or free of error. This "thesis" is not offered as a polished engineering
report, but rather as & 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 uncertainties inherent in the results obtained in
terms of the preciseness of the technical data and analytical methods
employed in the study. In the opinion of the students and faculty of
ORSORT, the problem.has served the pedagogical 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 National

Laboratory gave. In particular, the guidance of the group consultant,
A. P. Frags, is gratefully acknowledged.

Lewis Nelson

for

The Faculty of ORSORT

 
 

W,

ABSTRACT

For marine applications & circulating fuel, fused fluoride salt reactor
gsysten appears to qffer'a sfibStantially reduced specific weight (1bs per shaft
horsepower) ovér'current and planned reactor systems, Such a weight reduction
would make nuclear power'feasible.forfsurface ships smaller than 7500 tons
displacement, the cufrent_mifiifiufi for.pfesent and proposed reactor systems,
as well as overall performance improvements for larger vegsels,

Keeping within the bounds of currently available technology and proven
practices, reactor-steam system capsble of developing 35,000 SHP with an
overall specific weight of approximately 6.4 1bs/SHP is indicated, The
partieulaf installation of this sysiem aboard a 931 class destroyer of 3-4000
tons displacement was found feasible, When this system is used in conjunction

with the.conventional steam systenm to provide‘fuel—oil for shielding, an
overall reactof plant weight of 5/ lbs/SHP is realized,

In addition, the future potential of this design concept was investipated
utilizing unproven but indicated feasibie teéhnology advancements, Speéific

weights on the or&er.of 54 1bg/SHP were ‘found possible in this power range;

 

 

 

 

 
ACKNOWLEDGEMENT

 

The sauthors wish to take this opportunity to express their appreciation
to the many people throughout Osk Ridge National Laboratory who so_freely
allowed us to infringe upbn their spare moments to gain the benefits of
their experiences and knoyledge.

The group's special thanks goes to A, P, Fraas and W. R, Chambers,
our group advisors, for their guldance and for selecting the study. Our
particular indebtedness to the many pecople of the ANP Project‘at ORNL is
attested by our numerous. references to their work.

Also we wish to acknowledge the personal aid received from specialists

 

v of the Bethlehem Steel Shipbuilding Division, Ingalls Shipbuilding Company,

Babcock and Wileox, and the Knolls Atomic Power Laboratory.

 
-6-

TABLE OF CONTENTS-

1.0 Summary, Descrifition and Conclusion

2.0

3.0

1.1
1.2
1.3
1.4
1.5
1.6
1.7

Introduction

Réactor

Fuel

Materials

Heafi Exchangers and Steam Generators
Potential |

Conclusions

Introduction

2.1
2.2
2.3
2.h

2.5
2.6

Need for a High Performance Marine Reactor

Ship Installation |

Design Philosophy

Reactor Comparison and Selection

Advantages and Dlsadvantages of Fused Salt Reactors

Additional Applications

OveralltPower Plant Description

3.1
3.2
3.3
3.4
3.5
3.6
3.7

. Introduction

Alternate Approsach

Reactor

General

Shielding

Weight Comparison of Nuclear and Conventional System

Hazard Evaluation

 

Page
15
15
16
18

20

20
21
22
24
2l
25

26
28
30
33
35
35
38
39

43

42

43
b5

 

 

 

 
L

 

Page

v 4,0 Fuel and Secondary Fluid W6
4,1 Fuel b6

4.1,1 Introduction hé

4,1.,2 Coupogition Wt

4.1.,3 Corrosion 51

4,1.% Physical and Thermal Properties 53

4,1.5 Nuclear Properties 54

4,1.6 Availability and Cost B

4,1.7 Fuel Addition 55

%g: 4.1.,8 Fuel Reprocessing 55
J k,2 Secondary Fluid 57
4.2.1 Introduction 57

4,2,2 Physical and Thermal Properties 58

4,2.,3 Disadvantages of Fluid 59

5.0 Material Selection 61

5.1 Structural | : 61

5.2 Moderator 65

5.3 Reflector 66

. 5.4 . Poisoned Modetator Section 66
‘ffi 5.5 Design Properties of Materials | 67
¢ 6.0 Reactor and Primary Heat Exchanger Design 68
6.1 Introduction 68

6.2 Reactor - Types Cofisidered and Selection 68

6.2.1 Internal Arrangement 69

6.2;2 Vessel Design 73

6.2.3 Structural Arrangement | 75

 

 
7.0

8.0

S~

6.2.4 Fuel Pumps
6.2,5 Pressurizes and Expansion Chamber
6.3 Primary Heat Exchangers
6.3.1 Design Criferia_
6.3.2 Basic Design
6.3.3 Parameter Study
6.3.4 Stress Considerations
Steam Generating System
7.1 Introduction
7.2 Molten Salt Cycle Selection
7.3 Steam Generator
" T.3.1 Types Consgsidered
T.3.2 Design of Selected Steam Generator
7.4 Superheater
7.5 Auxiliary Equipment
7.6 Part Load Operation
Reactor Analysis
8.1 Nuclear Configuration
8.2 Parameter Study
8.2.1 Cross Sections
8.2,2 Summary of Results

8.2.3 Control Rod Study

Page

 

76
76
78
78
78
83
8k
88
88
88
91
91
ok
97
98
102
110

110

2113

114
116

117

o~

a3

oy v
B

€

9.0

10.0

8.3 Nuclear Design
8.3.1 Criticality
8.3.2 Self Shielding
8.3.3 Burnup and Fisgion Product Poisons
8.3.h Prompt Temperature Coefficients
8.3.5 ZXenon Poison
8.3.6 Delay Neutron Loss
8.3.7 Excess Reactivity Requirements
8.3.8 Control Requirements
Shielding
9.1 Introduction
G.2 Neutron Flux Calculation
9.3 Secondary Salt Activation
9.4 Dose Tolerance levels
9.5 General Shield Arrangemenf
9,6 Primary Shield
9.7 Secondary Shield
Heat Balance and General Aspects of the Steam System
10.1 Introduction
10,2 Steam Requirements
10,3 Condensate and Exhaust Heat
10,4 Heat Addition in the Steam Generating System

10.5 Comparison of Efficiencies

Page
123
124

12k

129

129
131
131
132
133
135
135

135

141
142
143
145
146
154
154
154
157
158
158
w1 0w

Page
11,0 Overall Power Plant Particulars - 160 |
11.1 Introduction | 160 ¥
11,2 General Arrangement - 160
11.3 Power Plant Control | 165
11,4 Emergency Operation 171
11,5 Maintenance 178
11,6 Removal and Disposal of Volatile Fission Products _ 180
11,7 PFuel Loading | . | : * 182
11,8 Pumps, Valves and Blenders | 184
12,0 Modified Approach 189 «
13,0 Weight Summary | 191
14,0 Future Potential | | 196 )
-
Figuré No.

2-1

3-1
3-2
3~3
h-1

b2
4-3

Lok
5-1

5-3

6-1

1]

LIST OF FIGURES

Partial Summary of Current and Proposed Nuclear
Marine Installations

Flow Diagram
Reactor Diagram
Specific Weight Comparison in 1b/shp

Phase Diagram of the Three-Component NaF—Zth-UFh
System

Partial Pressure of ZrF, Based on the Assumption
that Only NaF and ZrF) %xist in the Vapor Phase

Fused Salt Fluoride Volatility Uranium Recovery
Process

The System LiF-NaF-BeF,

Design Curve for As-Received Inconel Tested in
Fused Salt No, 30 at 1300°F|

Comparison of the Stress Rupture Properties of As-
Received Inconel Tested at 1300, 1500 and 1650°F
in Argon and Fused Salt No. 30

Effect of Section Thickness on Creep-Rupture
Properties of As-Recelved Inconel Tested in Fused
Salt No, 30 at 1500°F under 3500 psi Stress

Estimate of Weight Per Power Ratio vs Primary Heat
Exchanger Outer Diameter for Various Tube Spacings

Proposed Steam Generator
Proposed Superhea ter
Proposed Basic Arrangement
Salt Viscosity

Friction Factor

Heat Transfer Correlatlon

Pump Equivalent Weight

Page

 

29
36
41
by

49

50

56
60

62

63

64

85
103
104
105
106
107
108

109
Figure No,
g1
g-2
8-3
Bl
8-5
8-6
817

8-8
9-14
9-1B
9=2
9=3
10-1

11=1
112

11-3

11-6
11=7

-]2=

Reacfivity vs Mass U=235

Reactivity'vs U~235 Concentration

Reactivity vs U=235 Mass

Radial Flux Distribution

Radial Power Distribution

Thermal Flux Distribution in Unit Lattice Cell

Percent Reactivity loss During Lifetime Due to
Burnup and Fission Product Polsons

Core Reactivity vs Cofitrol Rod Position
Neutron Flux Plot -~ Core té Primary Shield Lead
Neutron Flux Plot in Primary Shield Tank
Secondary Shield

Reactor Compartment

Predicted Steam Balance for Reactor Powered System
at 359,000 shp |

General Arrangement
Simulation Flow Sheet

Reactor Power and Temperature vs Time for a
Ramp, Change in Power Demand

Reactor Power and Temperature vs Time for a Step
Change in Power Demand

Reactor Power and Temperature vs Time for Step
Change in Reactivity

Output Steam Temperature vsg Load

Reactor Power and Steam Temperature vs Time for
a Linear Change in Power Demand

Page
118
119
125
126
127

128

130
134
139
140
144

149

155
163

172
173
17Th

175
176

177
Figure Nog
A<5,1

A=5,2
A=5,3
A=6,1
A=11,1

A"“ll 92
A=11.3

A=11,/

A=11,5

=1 3

InconelhDesign Data

Inconel Degign Data

Inconel Design Data

Moderator Rod Tefiperature Distribution
Analog Represgentation of Fuel Loop

Analog Representation of Salt Side of Primary
Heat Exchanger and Superheater

Analog Representation of Salt Side of Steam
Generator

Method of Generating Power Demand Voltages

Analog Repregentation of Control System

Page

 

20k
205
206
214

2hg

250

251

I252

253
~14e

APPENDIX Pago i
5el Materials Data ¢« o ¢ o o o o o 0 2 s o o o o o o 202
6.1 Jugtification of Moderator Material . . o . « , 207
6,2 Calculations for Final Design of Primary Heat
Exchanger o« o o s o o 6 06 6 06 6 o 6 0 o 0 0 o o 215
7.1 Steam Generating System o o o o o o o o o o o o 226
8,1 Three=Group Cross Sections. o o o o o o o o o o 2ko
8.2 Perturbation Technique o o o o o o o o o o o o 21
8,3 Burnup and fission Product Poisons . « o « . & 2h3
8.4 Prompt Neutron Lifetime . « o o o o o o o o o o 25 )
11,1 Degceription of Simulation Program o o o o o o 246 .
11,2  BExpansion Chamber Heating Calculations . , . . 248 -
13,1 Breakdown of Basic Reactor Powered System

Components Weigh‘ba © &6 B © & 6 © © B ° o 6 o6 o 255

 

 
=15

1,0 SUMMARY, DESCRIPTION AND CONCLUSIONS

1,1 Introduction

This report covefs a study of the feésibility of & high performance
marine reactor (HPMR) utilizing a circulating fuel, fused salt reactor concept.,
The definition of high performance as considered in this report is low
specific velght in terms of total power plant weight per shaft horsepower,
By significantly decreasing specific weight below that which is currently
found feasible with present and proposed systems, reactor installations on
a lighter class of ships is now possible. This wduld also offer potential
improvements for all heavier classes,

A design study was made for a reactor system of this type to power a
931 class destroyer of 3-4000 tons displacgmenta The reactor and steanm
generating equipment simply replaced one of the present boiler rooms on
this c¢lass ship and duplicated the steam conditions (950°F, 1200 psig)
supplied to the propulsion machinery. An overall specific weighi of 59 ibs/SHP
was achieved for the 35,000 SHP delivered per boiler room, This is comparable
with the presently installed oil-fired system including fuel.‘ This speci=
fic weight was achieved with a reactor afid steam generating equipment overdesign
of approximately 30%, Indications are that if time had allowed a reiteration
of the system size to the 10% overdesign factor used in most reactor systems,
a specific weight reduction to at least 54 1lbs/SHP would have been achieved,
These specific weights, which are approximately one half that of any planned
syatem, were brought about by obtaining a small reactor package to minimize
shielding and combining this with the production of high temperature steam

to give godd steam plant efficiency,
=] b

The initial basic study incorporated a single intermediate loop utilizing
gnother fused salt {also compatible with water) to transfer the heat from the
fuel to the steam generator and superheater. This prevented activation of
the steam and through the use of blenders the temperature of the salt entering
the steam generator was feduced substantially to decrease the problem of

" thermal stress, However,; this required that saconflary shielding be placed
about the large volume of fhe steam generating equipment,- It was found that
through the use of two intermediate loops the amount of secondary shielding
could be.redueed and the overall specific weight releaaed from 58 to the 54
1bs/SHP, The comparable reduction fdr the case with 30% overdesign is from
65 to 59 1bs/SHP. Unfortunately sufficient time was not available to allow
as detailed a study as that given to the single intermediate loop systen,

Considerable use and reference has been made of the ANP studies and
experimental work carried out at ORNL on fused salt reactors, This has
allowed demofistrated components and materials to be incorporated directly

into this plant,

1.2 Reagtor

In order to achieve the primary overall objective of reduced specific
weight it is desirable to keep the reactor size as small as possible in order
fio minimize n;t only reactor weight but that of the primary and secondary
éhielding as well, The compact reactor selected was cylindrical in shape with
the fuel circulating up through a central critical region and then down through
an ammular downcomer at its periphery.containing the primary heat exchangers
(fuel to secondary fluid). The core is moderated by cylindriecsgl rods of
beryllium oxide clad with Inconel that are gquisPaced throughout the core region,

A nickel reflector surrounding the core plus an additional blankeflblack to
=] e

thermal neutrons shield the primary heat exchanger and prevents excessive
activation of the secondary fluid,

Because of the inherent stability that has been demonstrated with
reactors of this type, poison rods are not needed for control but a single
réd is placed at the core centerline to provide for reactor shutdown, mean
témpérature change, and fuel burnup, |

The reactor and steam generating system were designed to produce 125 MW
which is a conservative overdesign of greater than 25%., This safety factor
is considerably larger than felt necessary but was brought about by the
necesgity of starting the reactor design before the details of the steanm
_ system became available,

An average core temperature of 1225°F with an 100°F difference across
the core was selected as a compromise of weight and thermal efficiency against
corrosion and thermal stress problems,

The neutron flux gpectrum is largely intermediate giving rise to a
fission distribution of 28% thermal, 63% intermediate and 9% fast,

~ The nickel reflector tends to hold up the thermal flux spectrum at the

outer edge of the core and helps %o prpvide the favorable pesk to average
power diétribution of approximately 1.4. The power density averaged over
the core is 360 watts/cmB. |

The reactor vessel itself is approximately 6,7 ft in diameter and 6,7
ft high.’ An expansion tank for the fuel is incorporated into the head design
along with provisions for removing Xenon and other fission product gases., Three
fuel pumps are also located in the reacfor head in a manner such that they may
be replaced aboard ship. The reactor head is removable by unbolting and cutting

a smell omega type seal weld, This allows replacement of the primary heat
=18~

exchangers afid inspection of the core agsembly, However, it is recomménded
that the reactor be removed from the ship prior to this operation in order

to reduce the remote hgndling costs and problems., Also the feasibility

of balancing the cost of discarding complete reactor assemblies against that
of the design and operation of a remofe handling facility should be thoroughly
investigated with the idea of reducing both overall costs and simplification
of the basic reactor design,

The primary reactor shield is made up of structural support steel along
with approximately 5 inches of lead and 39 inches of water, The shield
requirements are based mainly on the fission product and sodium decay gamma's
and the delay and fission neutrons in the outer annular region containing
the primary heat exchangers. These activities were found to be seversl
order of magnitudes greaterAthan the prompt gamma and neutron radiation from
the core,

The secondar; shield for the basic_study enveloped both the reactor énd
the steam generating equipment and incorporated a thickness ofJapproximately
4 = 6=1/2 inches of lead. This requirement is a direet function of the
activation ‘of the sodium ions in the secondary fluid as it passes through

the primary heat exchangers.

1.3 Fuel and Secondary Fluid

In the selection of a fuel for this system, in addition to simpiy
selecting a carrier for a critical amount of uranium, primary emphasis was
placed on chdosing one that had been proven acceptable, This included its
chem;cal stability, corrosion, nuclear, and physical properties. This selection

was rather easily made since a large number of salts have been investigated
=19-

by ORNL and only a few found promising enough to warrant additional testing,

A solution of sodium, zirconium and uranium fluorides was selected on
the basis of reasonable nuclear and physical properties and because it had
been used successfully in a reactor experiment (ARE)}, Also, extensive
investigations have been made on its corrosion and physical properties in
anticipation of its use in the Aireraft Reactor Test (ART), The vapor
pressure of this salt is typically very low so that at operating temperatures
the reactor vessel has to be pressurized only slightly to prevent pump
cavitation, Thé actual composition of the fuel selected; closely approximating
that of the ART except for exact uranium concentration, is 9% NaF, 453 23
and 6% UFA (mole percent),

Uranium will be added to the system in the form of (NaF), UF,, Pellots

4°
or dissolved golution of this salt would be added during operation of the
reactor to compensate for uranium burn-up and to override fission product

and corrosion poisons. It is anticipated that sufficient addition of fuel

may be made throughout the life of the reactor to eliminate the necessity

of rgplacing the original salfi loading.

A basic ground rule requiring chemical aompatabiiity of the fuel, sécondary
fluid, water and sea water was established. In view of this coupled with
corrosion, heat transfer, radiation and chemical stability requirements, the
selection of possible choices was narrowed down to a fused salt, Because of
the difficulties involved in preventing this salt from freezing in the steam
gonerator a low melting point was also a requirement, On this besis a solution

of sodium, lithium, beryllium fluorides (mole percentages of 30, 20 and 50%

respectively) with a melting point of 527°F was selected.,
20~

1.4 Materials
.Mbst fuséd salts are quite corrosive to the standard structural materials,
However, it has been found that alloys containing large percentages of nickel
offer the gdod corrosion resistance to the fused fluoride salts, BExtensive
testing at ORNL under the ANP Project has shown that Inconel and the nickel-

molybdenum alloys present the best combination of strength and corrosion
resistance, DBecause the procurement and fabricability of Inconel are better

defined at pregent it was seléeted for the basie désign although the corrosion
resistance of the nickel-molybdenum alloys is much superior,

Inconel was salso selectéd as the structural material for components in
the steam system within the secondary shisld because of its superior resistance
0 chloride gtress eorrfisiono

The complexity of mechanical design problems involved in a separate
moderstor cooling system fiade it undesirable and must be weighed against the
high temperature difficulties encountered with fuel cooling. A ceramic
moderator appeared to offer a reasonable compromise from the temperature
standpoint although most did not have adequate nuclear and/br physical pro-
perties to bé aecepté‘ble° Beryllium oxide has the best overall characteristics
at present as its fabrication and physical properties are reasonably well
known and its satisfactory behavior under nuclear radiation had been demonstrated

experimentally,

1,5 Heat gerg and Steam

 

The primary heat exchanger is a once-through counter-flow type with
the secondary salt on the tube side, There are 12 heat exchanger tube

. bundles with each tube bundle made up of 6 subassemblies for ease of fabrication
=21~

and inspection,

The steanm genérator.and superheater are of conventional design fitilizing
U-tubes to reduce the thermal stress problem. The high pressure water and
steam are located on the tube side to minimize the component weight., Several
other designs that offered potential weight decreages were considered but
were not incorporated.because the design was not as well proven,

A blender was placed in the secondary fluid upstream of the steam
generator, This provides a means of maintaining the salt in the boiler at
a lower temperature than that in the superheater by mixing a relatively low
temperature salt for the exit of the steam generator with the high temperature
sfiperheater salt, This cofisiderably reduced the thermal stress at this

point and offered a weight saving over the use of a salt-to-salt regenerator,

1,6 Potential .

The major objective of the atudy covered by this report was to design
a power plant incorporating ideas and éomponents that could be substantiated
by referencé to a reasonable amount of experimental development work and test
programs. However, there exist many new facets of fused salt technology
that appear to offer large potential but at presént are little more than
qualified opinions plus a small amount of experimental verification., Because
it was felt.that this potential was significantly greater than that existing
with other type.of reactor systems, the study was extendsd to incorporate
the most promising of these developments.,

Through the use of a new fuel that offers more self moderation, moderator
materials that allow the core to operate at a higher power densities, and

structural materials that offer improved corrosion resistance, the basic size
=22

of the reactor itself decreased from 6.7 ft diameter by 6,7 £t high to
approximafiely 5 £t diameter by 5.4 ft. high, To further decresse the size
of the heat exchangers and steam generating equipment, sodium was used in
the intermediate loops.

This study indicated that it would be reasonable to expect that a specific
welght reduction on the order of at least & lbs/@HP could be achieved in

the future with fused salt reactor systems,

1.7 Gonclusions

This design study of a circulating fuel, fused salt reéctor for é
marine power plant has shown that such a.system is technically feasible at
present, Reactor systems of this type not only allow overall perforggnce

improvements over current systems, but allow reactor installations to”be
considered for a lighter class ship, In addition, with the developmental

and experimental work accomplished in thig'field at ORNL; the construction

of this plant could proceed with a minimum of additional development work,
Also considering the potential of this'systefi with developments that are

novw in sight, it appears that considerable performance and weight improvements
could be expected,

The difficulties involved in handling the fluoride fuel and maintaining
it above its melting point have been satisfactorily overcome and proven out
in test loops and a reactor experiment, Algo, materials that will give
adequate resistance to thé’high temperature corrosiveness of the salts have
been found, although increased corrosion resistance would be desirable,
Although the fuel inventory required is considerably higher than for other

systems, this is partially offsat.by the elimination of the need for replacement

 
~23-

cores and holdup for chemical reprocessing, When the many important advantages
of this type of system are considered, they appear to more than offsget the
above, These include: higher temperature gnd overall thermal efficiency,
low weight and volume reéuirements, low pressure system, proven stability
allowing the elimination of numerous integral cbntrol rods, continuous poison
gas removal, fuel makeup as needed, etc,

In the judgement of the authors the ciroulating fuel-fused salt reactor
not only shows considerable performance potential over present and proposed
marine installations but it offers the most promising system applicable %o

a small ship installation,
=2/~

2,0 INTRODUCTION

2.1 Use for HPMR

Atomic weapons were not only the forefather of atomic power reactors,
but also the forerunner of a completely new concept of naval warfare, A
small ship utilizing missiles with atomic warheads could have the destructive
effectiveness of the largest warship of the preatomic era. If one could take
such a small ship and build it in large numbers, give it a high speed along
with an effectively infinite range, it clearly would present quite a formidable
weapon. A small ship with no refueling problems would also have many other
potential uses such as convoy and patrol duties in isolated areas, The
purpose of this report is to determine the feasibility of a reactor systenm
capable of being instslled aboard # small ship to give it the éffective
infinite range mentioned above. Also once the feasibility of an improved
high performance (lightweight) marine reactor is established for a small ship,
it likewise holds promise for considerable weight savings on larger vessels
and volume savings on submarines, |

For the purpose of this report a 931 class destroyer of 3500-4000 tons
displacement was selected for investigation, This ship is roughly half the
displacement of the smallest current proposed reaefior installation (Sec, 2.4
and Ref, 8), but considerably over the minimum size felt necegsary to contain
a crew for long durations. This ship contains two separate boiler and
machinery rooms utilizing steam at 950°F and 1200 psig to produce a total of
70,000 SHP, These steam conditions fortunately fell into the rangé'considéred
desirable for reactor installations of this type. With the machinery room

fixed, the boilers could be simply replaced by a nuclear system without com-
=25

promising the basic reactor design. This would considerably ease the redegign
of a conventional 931 class destroyer to nmuclear power as well as offer the
shipboard advantage of the crew being thoroughly familiar with the steam
plant, In addition, the logistic and shore maintenance problems would be

reduced because of the number of identical steam plants in service,

2,2 Ship Ingtgllation

For the purpose of thig study it was felt most feasible to replace only
one of the boiler rooms with a reactor system, This not only gives the
advantage of having the proven reliability of a completely conventional system
aboard ship, but would considerably reduce the total cost of the complete
installation, 4lso the performance penalty paid for utilizing both the reactor
and boiler systems would be very small if not negligible,

The difference in speed of this class ship between operating on the
reactor system along (35,000 SHP) and maximum power (70,000 SHP) ig approximately
4 knots, Obviously, this inefficient utilization of power is not warranted
except under emergency conditions. In addition, structural design problems
associated with vibration and noise along with their relsted detrimental
effects on submarine and aircraft detection equipment does not make extended
maximum speed operation appear feasible, As an additional point it should be
noted that if an average fleet apeed of as high as 20 knots is assumed, this
ship would be developing less than 1/3 of its potential reactor power and
zero conventional, Therefore the conventional steam boiler system can be used
to augmeht the reactor when emergency conditions exist apparently without
fienalizing the overall ship operation and offer large savings from both the

coat and reliability standpoint,

 
s

The total amount of fuel oil carried aboard is approximately 54% of its
original value, Since this is to be used only under special conditions and
not for crusing it is considered adequate. A typical combat problem was
not available for analysis, but it is felt that the endurance of the nuclear-
0il fired ship combinatién at maximum power would be substantially increased

over the conventional ghip,

20,3 Design Philosophy
- Because of both the relatively short time available for this study and

the limited experience in certain aspects of the field it should be realized
that a thorough investigation of a1l phases was not possible. In instances
where there appeared several feasible approaches, but with éach requiring a
considerable design effort o evéluateg a somewhat arbitrary choice utilizing
engineering judgement had to be made, These selections and the alternate
possible choices are discussed throughout the report, The primary objective
was to establish design feasibility for the small ship application,
Accomplishment of this with the selected design, indicates that with additional
study the possible alternates herein bypassed could either be incorporated
with a subsequent design improvement or simply rejected,

Many detailed problems concerning the steam system were not thoroughly
investigated as it was félt,that solutions to these wers well established,
.Major emphasis was placed on the really unusual problems concerning the reactor
and intermediated salt systems to obtain plausible solubions,

The basic design philosophy was to use materials, designs, and techniques
that have been established as feaéible and backed up és much as pogsible by

experimental verification, In cases where the restriction to present day

 
 

w2

technology eliminated alternate approaches, they were briefly mentioned for
possible future consideration., Attempts were made to fully utilize the
experience, knowledge, and background of the personnel at the Osk Ridge
National Laboratory and other industries,
Several basic ground rules were established early in the design study,
The first was that within the limits of the previous paragraph, the design
optimization would be on the basis of obtaining the lightest weight on g
1b per shaft horsepower bésis. Another was to prohibit the use of a fuel
or intermediaste fluids that were.not compatible with each other as well as
steam plant and sea wafier6 This ground rule was believed to be basically
desirable from the battle damage standpoint because of the severe punishment
that ships of this type can receive and still be operable, Also this com-
patibility offers obvious safety advantages in the steam generator design,
Advéntage was taken of the conventionsl plant fuel oil left on board
by using it for shiel@ing purposes. If a completely nuclear destroyer design |
is required; it appears that the weight advantage may not be as acceptable
as for the combined conventional and nuclear powered ship. However, because
of the narrow beam of this class ship, the reactor compartment can be rearranged
to utilize the salt and sea water at the sides to reduce the shielding
requirements, Because of the decreasea volume and especially the height of
the fused salt system it is possible to install the top shield deck of the .
reactor compartment at the water line, Advantage could also be taken of putting
the reactor compartments back-to-back to reduce the required shielding,
Unfortunately limited time prohibited detailed consideration of these arrange-
ments from being made for a completely.nuclear ship 'although an estimate was

made on the added shielding weight required for the proposed installation.
 

 

~28-.

 

2.4 Reactor Comparisons and Selection

In the process of 1nvestigating potential small ship applications and
selscting g 931 olass destroyer for thie study, it became obvious that g
nuclear power plant specific weight on the order of 60 1bs/SHP had to be
realized, A brief review of current and proposed nuclear ship installations
vas made and these all fell considerably short of meeting'this requirement,
These values, summarized on Figure 2-1, are to'be_considered only approximate
and neither the latest or the bost values, The lightest values found were
105 for the FIW and 90 lbg SHP for D1G, The FIW is a joint WAPDmBethlehem
Steel effort in which g 1arge portion of the detail design has been firmed
up. This design preduces approximately 80,000 SHP and is installed on g
14,000 ton ship which would normally be considered in the light to medium
crusier class, The D1G program is a KAPLmBethlehem Steel venture that ig
81111 in the early preliminary design stage with the specific.weight given
being only a design objective and not a design-substantiated valueo The
design power is to be 60 000 SHP with a total ship weight of roughly 7500 tons.
This size is in between what had been considered the destroyer class (2,5 =
4000 tons) and a eruigser class (12 - 18,000 tons) It is apparent that thege
”1nstallatzons do not offer much promise of g welght reduction to 60 1bs/SHP
for a 931 class installation,

In investigating the field in general for g lightweight reactor systemg
the Aircraft Nuclear Propulsion (ANP) Projects appear to hold similar require-

ments for low propulsion system specific weight, In addition, it seems

 

substantially improved at a small enough increase in overall weight to make g

ship application most feasibleo

 

 
-29..

 

 

 

 

FIGURE- 2— | SECRET
PARTIAL SUMMARY OF CURRENT
AND
PROPOSED NUCLEAR MARINE INSTALLATIONS

SHIPS SHAFT HP PROPULSION  SPECIFIC
SYSTEM WGT* WGT™

(LONG TONS)  (LBS / SH)

SUBMARINE S
NAUTILUS (S2W) 15,000 1100 160
S4wW : 6,600 690 230
SEA WOLF 15,000 1200 180
(SIR-52G) -
TRITON (SAR- S4G) 34;000 1700 110

(2 REACTORS)

SURFACE SHIPS

FIW 80,000 3700 105

DIG 60,000 2900 - 90
93! cLAss DESTROYER 70, 000 1800 58
{ NON NUCLEAR) {INC. FUEL)

* NOTE: THESE VALUES ARE ONLY REPRESENTATIVE AND NOT THE LATEST
OR BEST VALUES. '

 

 

 
T — i

- =30-

   

Two types of reactor systems were considered (1) the heterogeneous
gas cycle using high temperature ceramic.fuel elements under development
by General Electric at Bvendale and (2) the circulating fuel, fused salt L
system being developed at ORNL,
A pas cycle did not appear to be readily applicable at present for a
ship installation because gas turbines of the size réquired”had not been
developed, Alsgo, éven though bigh gas temperatures and correspondingly
high turbine efficiencies could be achieved, material limitations could

prohibitly limit the extended life required for a ship applicationo

 

The fused salt reactor concept appeared to readily adapt itself to
a steam generation application, The nominal reactor temperature could be
decreased several hundred degrees (OF) from the ANP design values for an
improvement of the corrosion pfoblemo This still wouid résult in an asmple
temperature margin to provide steamlwith BmAOOOF of superheat at desired
pressures,

These factors coupled with.the "at. hand" availability of fused salt

technology made this type of reactor appear to be most feasible at present,

2.5 Advantages and Disadvantages of Fused Salt Reactorg
Like any complex system, a fused salt reactor installation exhibits | ‘
both strong and weak points, In any reactor comparison, a relative weighing
of these must be made along with the determination that flo unsolvable weak
points exist., However; in considering a 931 class ship installation such a

comparison cannot easily be made because no othdr reactor configurations exist

  

that can satisfy the strict weight requirements. Therefore if a need for a

nuclear ship of this si axists, the fact that no unsolfiable problems apparently

  
 

 
=31

exists in above sufficient reason to proceed with a detailed study., Faftunatelys
& fused salt system offers many advantages over conventional reactor systems
that could make it highly competitive even for large ship applications, There-
fore the advantages as well as the problems of fused salt reactor systéfis are
discussed to establish its potential over other reactor types for future
comparisons,
R.5.1 Advantages

1, High temperatures are obtainable which give rise to high
overall plant thermal efficiencies,

2. Low pressure reactor system, Pressure required (< 100 psi)
oniy to provide pressure differential for fluid flow and to prevent pump
cavitation,

3. Inherent stability of this class reactors has been demonstrated,
Multiple control rods and control drive mechanisms are not required at a conml
siderable saving to cost, reliability and maintenance problems, A gingle
control rod which may be required to compensate for fuel burn-up or to prdvide
for désired temperature changés may be located outside of the reactor wvessel
and subject to relatively easy maintenance,

4o With this type of reactor design it would be possible to
overtemperature the reactor to obitain large increased in power output for
emergency conditions, Undoubtedly this would be at a sacrifice in overall
1ife of the system but extreme conditions could warrant this use,

5. The fission product gases may be continuously removeds; thereby
eliminating the Xenon override problem.and the excesgs reactivity requirements.,

6. The basic reactor is generally much more symmetrical and

smaller than other systems, thereby reducing the shiel&ifig problems as well as

 
=32

the overall size and weight,

7. Inexpensive fuel preparation. The reactor core is of simple
design and therelis nb fuel element fabrication and burnable poison costs,
The handling of U-235 is simplified as "apiking" of the salt fuel carrier
is required only aféer the readtor hag been filled,

8. Reloading o; refueling would generally not be required
during the life of the reactor., Tuel additions may be made during reactor
operation to compensate for fuel burn-ub and to override soluble fission
product buildup, Because of thig and'(5) the excess reactivity requirements
are considerably reduced and can lead to & reactor thét is inherently safe |
from power excursions, . | |

9. Chemical stability - No radiation‘damaée or fuel decomposition
problems. Explosive radiolytic gases are not formed, thereby eliminating
problems such as the recombination of H2 and 02 in water reactor systems,

10, The combination of (7) and (9) make it possible to utilize
high core power densities with the subsequent reduction in reacfior size,

1l, Chemical reprocessing is greatly simplified with & homogeneous
system giving a corresponding cost decrease,

12, Although not diredtly applidablé to a marine installation,
it should be noted that for breeding'purposes both thorium and uranium are
soluble over a large range of concentrations, This is not true for either an
aqueous homogeneous or a liqfiid nmetal éystem.

Re542 Disgdvantgges

1, Gorrosion problems are more difficfilfilthan for an aqueous or

sodium system but probably better than a homogeneous liquid metal reactor.

<+ High melting point requires that careful attention be paid

 
=33

to loading and Operational techni‘queso A molten state is required at all
times; however the successful operation of a reactor experimant (Ref, 6)
indicates that these problems may be SOlvedo

3, A high degree of leak tightness and reliability 1is required
for the core vessel and primary heat exchangerso Careful and tight quality
control and material inspection is requirefio

4; A high fuel investment in the reactor is required, Howéver,
considering the elimination of replacement cores and the cooling period
before chemical reprocessing the total investment may be comparable to
heterogeneous, solid fueled systems,

5. Poorer neutron economy is obtained than with aqueéus

homogenedus systems although newer types of fused salts offer improvements,

\tions

   

2.6 A@ditignalhfizilig
| Once the design feasibility of a fused salt reactof system is proven
1t offers considerable potential for both larger and smaller vesgels, If-
specific weights on the order of 60 lbs/SHP ean be maintained for smalier
power sizes many new opportunities are available for an even smaller ghip
application. In going to a larger size ship an overall specific weight of
60 should be more easily attained, This would offer either a weight reduction
or an increase in storage capacity of approximately 1600 tons for a ship the
size of FIW or 1000 tons for D1G,

A fused salt reactor also offers a considerable reduction in the size
of an installation, While important for any ship, this is even more important
for a submarine application, A preliminary comparison madé by KAPL in 1953

(Ref, 7) indicated the degign advantage of this type of system for a submarine

 
_ym

application, 4 current design would tend to give an even bigger advantags.
Incidentally, a potential non-marine application not pertinent to this
gtudy but of general-interest invélves_thé use of stationary fused salt

reactor plants for treeding and electric power production (Ref, 5).
_35 -

3.0 OVERALL POWER PLANT DESCRIPTION

3.1 Introduction

  

The utilization of only a single reactor system aboard a 931 class

 

destroyer offered gsome desirable flexibility as to the overall ship arrangg%-
ment; However, the numerous basic considerations required to establish | 4
the dptimum shipboard installations were somewhat beyond the gcope of this
report, With a cursory investigations, it at first seemed to be most
-advantagecus to replace the fbreward boiler room with the reactor and steam
generator equipmento This had the sizeable advantage of not requiring any
layout considerations or secondary shield penetrations for passage of the
propeller shafts from the other engine room, However, under detail design,
the size of the reactor compartment was reduced below that originally
contemplated and means of circumventing this problem became spparent, The

af't compartfiént location also offered the advantages of easier accessibility
and a better location of the fuel o1l tanks to maintain ship trim,

The basic system upon whi&h the major design effort was placed consigted
of the eirculating fuel, fused fluoride salt reactor incorporsting an integral
heat exchanger unit fo remove heat from the fuel, A secondary fluid, another
fluoride salt, is used to transfer the heat from the primary heat exchangers
within the reactor vessel to the steanm generating equipment, The steam is
then supplied to the conventional 931 class destroyer machinery room at a
temperature of 950°F and a pressure of 1200 psig, A simplified schematic of
this system is shown on Pigure 3-1, The detailed heat balance is discussed

later in Section 10 and presented in Figure 10-1,

 
IAIVA WINHL — @

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
      

 

 

 

 

 

 

 

 

 

 

 

 

 

¥43ON38 — V - - }-¢ 3¥Nn9i4
dwnd —(d)
4.89% , .
7 o 1u/dl 01929
¥3L1V3IH \\_ de204
H3ILVYMQ33d .
| MOLVYINIO
WVY3LS
mumzuozoo\x
¢ +
4 1sdgez , &
4,226
. MWK 6°C6
HOLOV3Y
——h,
, _ 431V 3H . | 4809
HS 000°€e § 1 3INISHUNL ¥43dns *
/q) . 01X 612 /g1 Ol x 612 o
1 i} L\%/’.
1Isd 0021 —4,066 4, 8€11 I

 

"W3ILSAS NOISTNdOHd — ¥01Ov3H 40 OJILVWIHOS d3I4I1dWIS
.‘..37..

The mean reactor temperature is 1225°F with a power output of 95,9 MW
required to supply 35,000 shaft horsepower to the ship's propellers. The
temperatures across the various equipment as well as the steam and salt fldg-
rates are given on Figure 3=1, The secondary salt syétemg which carries tfi;_
heat from the reactor to the steam generator; is broken up into two inter— ?}
‘connected ioopso The top loop supplies the superheater with a relatively |
flot salt, This is required %o reduce the su?erheater size becausé of the
low heat transfer coefficients characteristic of the steam side. The bottom
loop maintaing the éalt at a lower temperature to reduce the thermal stress
problems in the steam generator, This is desired here because the relatively
high heat transfer coeffioiénts on both the water and salt side would give
a large temperature drop across both the tube and header walls and hence
a high thermal stress, Blenders are used to interconnect the two loops as
indicated thereby allowing cold salt from the exit of the sfieam generator
to be réciraulated to reduce the resactor inlet temperature to the desired
value,

-.It should be noted that ihe reactor and steam generators were basically
designed to produce 125 MW, This overudesign.of approximately 30% was brought
about by the time lag involved between when the basic reactor had to be
selected and when the detailed steam conditions for the desired size ship could
be obtained, A 10% over-design safety factor (used for other marine reactor
applications) would be desirable but time did not permit a reiteration of the
work to thié size, An approximation was made to allow for this over-design
(Sec, 12) to .indicate the overly large weight penalty incurred. In replacing
the boiler equipment with this reactor system, it appears (Section 11,2) that

a substantial volume saving is also realized, While important for any small

 
ship, it should further emphasize that in this application the volume is
saved over the height of the boiler room (approximately 30 ft) making it also

available for missile storage,

3.2 Alternate Approach
The sodium component in the secondary fluid becomes highly activated as

1t passes through the primary heat exchanger due to both delayed neutrons
from fuel in the exchanger and fést neutrons from the core, Because of the
large amount of this secondary salt outside of the primary shield, it is
necessary to incorporate a relatively thick secondary shield aboubt all of
the steam generatlng equipment,, A more detailed design should consider the
possibility of reducing this activation somewhat through the use of poison
bearing materialsg:ioeog boron, in the_héat exchanger region, However, thig
does not appear to offer a large reduction in tfie secondary salt activation
because only approximately 10 = 15% of the activation results from thermal
neutrons, the remainder occurring because of the high intermediafie energy
flux and sodium resonance fieakso |

An slternate approéch that was briefly gstudied used two intermediate
fluids (both salt) in order to prdvide & non-redioactive salt in the steam
equipment. Although a penalty was paid due to increased pumping power require-
ments and superheater size, this was fiore than compensated for by a shield
welght decrease due to a smaller enclosed volume, Direct access was also
given to the steam generators and superheaters for maintenance° In addition,
1t now appeared feasible to keep_the secondary shield small enough to allow
an aft boiler room installation, if desired, without the complication previously

mentioned,
A

2-¢ 3J4dnoid

MW G2l
HOLOVIAY 3INIMVIN 3ONVNHO443d HOIH

 

ll8 -Ig >

 

 

 

 

 

 

 

 

 

 

NOILO3IS 40d
HO1LVH3IAOW
g3aNOSiOd

Saoy
HOLVHIAOW—

H0L03N434
TAMOIN —

H3IONVYHOX 3
AUVWIEC

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

g13iHS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

H38WVHO
NOISNVdX 3 13Nnd

a0y 04Y1INOD

S3INVA
43SN43ia
S

 

\
801.2 :’?:3 -\\\\‘
N

Saoy QT3IHS \\
NOI93Y

¥IONVHOX3 LV3H

 

 

\,g <

 

 

 

 

 

  
  
 
  
   
    

SL3ININ0 HIONVHOX3F
LV3H AdHVWIYd

 

 

 

13LNC dAnd

QAIAISSVIONDN
gyLee "OMO-U1-"INYO

 

nNne
3,3 Reactor

The overall éhape'of the baslie reactor 1s a cylinder approximately 80
inches high by 80 inches in diameter. Fuel is circulated up through a central
critical region equivalent to a cylinder 75 cm éiameter by 80 cm high and then
down through an annular downcomer around the periphery. (See Figure 3-2)

This outer region contalns the primary heat exchangers for transferring the
heat from the fuel to the secondary salt,

Gylindrical rods of beryllium oxide suitably clad with Inconel are
equispaced throughout the core 1o provide for moderation, The ends of these
rods are loaded with a poisoned material to reduce.end leakage and fissioning
~in the entrance and exit plena. A single control fod éhannel, approximately
. inches in diameter, extends through the center of the core region,

The sides of the core are enclosed by a nickel reflector blanket 6 inches
thick, This inelagtically scatlers some high energy leakage neutrons back
into the core to improve the criticality as well as to offer both compact
neutron and gamma shielding, To further reduce the neutron flux in the heat
exchanger region the reflector is in turn surrounded by a 5-1/2 inch thick
region of cylindrical rods containing a mixture of beryllium oxide plus
boron-l0, Boron bearing Inconel rods are placed in the interstices of these
cylinders for shielding purposes and to reduce the fuel located in this region,
A thin slab of material essentially black to thermal neutrons, boron carbide
in a copper matrix, then surrounds this region to completely absorb any neutrons
that are thermalized in its outer periphery,

Small passages are provided through the nickel reflector and the
cylindrical BeO-BlO'rods to eirculate fuel for cooling purposes., An 1/8 inch

annular gap is also located between the thermal shield and the reactor pressure
e

vegsel wall for cooling purposes. The lower temperature fuel from the heat
exchanger axit circulates through here and then to the expansipn chamber in
the reactor head., This minimizes both the head temperafiures brought about
by decay heating and "snow" formation from the fuel (Ref. Sec. 4.1.2).

Three removable centrifugal pumps which are located in the reactor head
profiide for fuel flow and pressurization of the system. These pumps are also
designed to facilitafie'removal of the gaseous fission pfoducts.

Additional details of the reactor design are incorporated into

Sections 6, 8, and 11,

3.4 General

A major disadvantage of a reactor of this type is that provisions must
be made tO'enSufe that temperatures are maintained above the fuel melting
point at all times, Althdugh accomplishment of this has been proven feagible
by both many loop tests and a reactor experiment; (Ref. 6), careful attention
to operational procedures are required, Although a complete freeze up is
not catastrophic from s nuclear sense, experience has shown that severe pro-
blems exist from a stress standpoint upon remelting, Becauss of this, dump
tanks are included between the double hull under the secondary shield com-
partment for emergency use.

The reactor and sgecondary salt pumps may all bé replaced through the
secondary shield, Sufficlent room also exists above the secondary shield so
that the pump drive and control drive mechanism motors may be accessibly located,
The lightest and most flexible system for pump drives appeared to be the steam
turbine, This offered the advantage of having variable speed characteristics

and also did not require the addition of generator sets to the system, It
o=

was planned to back up each of these drives with a small AC, motor, Those
would provide suitable'ciréulatibn under zero power standby conditions as
well as offering the safety advantage of having two irdeperdent systems
under emergency conditions. |

Control of the reactor systefi could be accomplished by varying the
sécondary salt fiow rate as demanded by the steam plant, This, as well as
an alternate approach of by-passing sécondary salt around the reactor,
is discussed in Section 11,3,

The reactor as illustrated on Figure 3m2 indicates a possible method
of unclamping the head to allow replaceménfi of heat exchangers and othér
internal components. Two different methods of connecting the heat exchangers
to the reactor vessel for disassembly purposes aré shown, Although shown
to be feasible on this drawing,_it 1s very questionable as to whether or not
the cost for this eage of disassembly 1s warranted from the overall maintenance
standpoint, Additional discussion on two different concepts of maintenance

is presented in Section 11.5,

3 . 5 Shieldin
The basic shield is designed to limit the maximum allowable dose to

15 mr/hr on the outside of the secondary shield, This would allow access

to the auxiliary engine room for 20 hours per week for maintenance on pump
drives, deaerators, feed and boiler recirculating pumps, etc., At an average
distance of 10 feet from the secondary shield, the limited access would be
increaged to approximately 30 hours per week, Unlimited access would be
allowed in the main engine room,

The primary shield is of laminated structure containing the equivalent

 
of 5 inches lead, 39 inches of water, and 1-1/2 inches of gstructural steel.
The secondary shield is designed to attemuate the decay gammas from the
activated sodium component of the secondary salt and gamma leskage from

the primary shield tank, This shield makes use of the fuel oil fequired

on board for the conventional system for shielding the forward, pért and
starboard éidea and approximately 4 - 6=1/2 inches of lead for the top and
aft seetiohs° An additional 1-1/2 inches of lead is located over the reactor
fuel pumps to eliminate streaming through the crevices required to drive

and replace these pumps,

3.6 Weight Comparison of Nuclear and Conventional System

A weight Breakdown for a 931 clags destroyer with a conventional and
a nuclear installation is given in Figure 3,3, To simplify the comparison,
specific weight rather than the actual weight of the cqmponefits is presented.
The actual shipboard weight (lbs) for the conventional system may be obtained
by multiplying through by 70,000 SHP for the total weight or 35,000 SHP for
the welght per engine room, This will also hold true for the fuel-oil weights
listed.

Using this information, it can be calculated that the conventional
total ship power plant weight is 1123 tons (long) wet, with 728.5 tons of
fuel oil, |

System No, 1 is considered to be the basic design upon which most of the
design effort was spent, It contains information a reactor and steam generating
system over-designs of epproximately 10% and 30%., System No, 2 used the
alternate épproach congisting of two intermediate fluids to allow placement

of the steam generators,; etec, outside of the secondary shield, and a similar

 
 

 

v vS €S 8¢ | 0'6S (%01)
909 269 (%0€) - 1viol
= N9IS3AYIA0
mm L L v'g2 110 13nd
2 : -
02 02 032 S, S3IYOLS B QVO1 "
2’ 2¢ 2% 2l SWILSAS LN3AN3IJ3ANI r8H
. 09 09 09 o' LNVId 01819313 98 4
92l S'v¥l 802 00 ONIGT3IHS NOILVIOVY 3
bl ovl €32l 08 AYINIHOVW LNVId ¥O1OV3¥ 0% 9
o 2Ll 2Ll 2Ll 9’61  AYINIHOVW NOISTNdONd WV3LS 878 V
IVILNILOd QIIIGOW 0IS¥E TWNOILNIANOD | | AHO93LVO"
TERENTY | H __
. )
dHS/gg7 NI SNOSIYVAWOO LH913M 0I1d4103dS =
n (e
5 ¢'¢ 3YN9ld 1
3
 

45

approximation of both a 10% and a 30% overdesign safety factor, These
designs utilized a portion of the fuel oil required for the conventionsl
steam system, The 1,7 lbs/SHP for fuel oil as listed, is the fuel oil above
50% of the compietely non-miclear destroyer capacity that is required to
maintain ship balance, A third design utilizing advanced material and
reactor technology and eliminating the ground rule of required fluid
compatability with water, achieved a further reduction in specific wéight

to 56,7 1bs/SHP without any fuel?oilzrequired for shielding., A comparable
value with the above utilizing fuel-oil would be 46,7 1lbs/SHP,

It is Tealized that in many cases the welght of a reactor system goes
up in proportion to the amount of design detail accomplished, However, this
general tendency would be reduced in thig study because the entire steam and
electric piant9 which accounts for approximately 1/3 of the total welght, has
been actfially detailed and constructed, In addition, an attempt was made to
apply conservative estimates to the various components to account for unknown
growth factors, A detailed weight breakdown, including the estimates made,

is presented in Section 12,

3.7 Hazard Fvaluation
A hazard study for marine application of thig type of reactor was

carried out by a pair of ORSORT students (Ref, 64). This evaluation indicated
that basing the major destruction of both the ship and reactor vessel, this
system was inherently as safe as any nuclear system, With a major catastrophe,

however, a more widespread release of fission products would result,

 
46

4,0 FUEL AND SECONDARY FLUID

L,1 Fuel

 

b,1.1 Introdufition

The chief'advantage of using s fused salt fuel is that high
temperatures may be obiained at low pressures. BSuch a system ig also
capable of high power density with accompanying small reactor size, and
low shield weight. Also, gaseous fission products way be rémoved. No
fuel element fabrication results in 1ohg life for core, and high fuel
burnup. Fuel may be continuously or periodically added as it 1s burned.
In addition, and by no means of least importance, fused salts do not react
violently with water. | |

For such a system, the fused salt fuel afid diluent must have g
reasonably low welting point, low neutron capture cross section, stability
at high temperaturés and in extended high fieutrdn, beta and gamma fluxes,
In addition, it is essential that the fuel.system be sufficiently non-
corrogsive to the confainer material that an acceptably long life and freedom
from maintenance may be realized.

The fused salt may or may not function as a moderator. In the reactor
herein described, moderation of fast neutrons is accomplished largely by
means of moderator rods dispersed throughoutmthe core, The design chosen
and fuel selected resulis in an epithermal'br intermediate reactor, rather
than a thermal reactor.

A large amount of fundamental as well as engineering reéearch has been
performed at ORNL toward development of fuels, and the selection of the fuel

known hereinafter as Fuel 30 was based on the results of several years of
47~

phase dlagram reséarch, dynamic and static corrosion testing, and in-pile
160p tests, The choicé of this fuel permits the fise of technology already
at hand, and does not require additional extensive fundamental research,
In addition, critical experiment data and actual reactor operational data
are availafile, where similar fuels were used or simulated,

While it appears desirable for moderating efficiency that a.fuel be
used which contains LiF and B§F2, the present technology of contaifiing such
fuels is nqt considered adequate, However, it is expected fihatgfuture
designs for_fused salt reactors will be possible as soon as research
currently in progress has been completed. Such research is now leading
toward development of very corrosion resistant nickel molybdenum alloys,
which show.extremely good prospects for future use in fused salt reactors.

k1.2 Cdmfiosition

The approximate composition of Fuel 30, as modified by the
criticality requirements of the particular configufation of the reactor,
isl(expressed in mol percent) 49% NaF, U5% ZrF), 6% UF). Zirconium fluoride

is made from hafnium free zirconium, Additional composition data are:

Comgosition

Mol % = Wt %
NaF | 48,7 17.9
ZrF), | 45,2 65.7

UFy, 6.1 16.4
48

1200°F - 12000F
Mol % Gus/Cm3 ) Atoms/Cm3 Atoms /Cm
Sodium 15,49 335 . 8.76.x 10°% 2,57 x 1021
- Zirconium 14.2& 1,221 8.30 xv1021 | 2,43 x 102t
Uranium 1,90 420 1.08 x 1022 .317 x 1021
Fluorine 68.36 1.43h _" 45,43 x 102t 13.32 x 10=1

Fiéure -1 is phase diagramlof the 3-component system, NaF-ZrF)-UF).
Tt will be seen from Figure -1 that the composition selected 1s in the
vicinity of fihe triple eutectic low melting composition, Also, if solid
fuel concentrate is added in the form of'NaEUF6, only compositions having
lower wmelting points than théfconcentrafe are formed as dissolution progresses,
Figure L4-2 shows Zth.vapor pressure for various mol percentages of
ZrF), as a function of temperature. It is apparent that this vapor pressure
is dependent on both ZrF) concentration and on temperature. The formation
of ZrF) acicular crystals ("snow")'has resulted from h%gh temperature treat-
ments of ZrF) -bearing salt mixes; This segregati0n can become a problem if
conditions are favorable for sfiow formation, According to our best information
(Ref, L49) snow formation should not Qe 2 problem if the waximum fuel tem-
perature is kept below 1350°F in the expansion chamber. The accumulation
of snow-like ZrF) crystals is most undesirable and may lead to the plugging
of passages or fouling of the expansion chamber, To further avoid this
cold surfaces in tfie expansion chamber should be eliminated, It is clear
thatthe use of a fuel devoid of ZrF), is desirable, Hofiever, corrosion

considerations dictate the selection of Fuel 30 at the present time.

 

 
-49- UNCLASSIFIED
ORNL-LR-DWG 33984

 

 

 

NaF

 

 

COMP. NO. 34
43 mol 9, ZrF4

 

ZrF4

Fig. 4-1 --Phase Diagram of the Three-Component NaF~ZrF4-UFrp
System.

 
GQ dduaasyax 8sg ‘aseyd .H_omm> 92Ul Ul 1sIXy VAIZ pue JeN LA[uoc
¥ey} wonydwnssy sy uo pesedq ¥ 417 Jo seanssaayg Ietlred -- g-3% ‘314

g 0T

0 §9¢

ORNL-LR=Dwg, =18170
UNCLASSIFIED

-50 -

 

Q09
T

 

 

2 ‘mmjpasdwaey

Q0L 008 006 o001 QS0

i i [ [
wm_auw ‘0INI0ICWR L (09040120

  

109

AmdBw ww ‘einsseyy

 

 
.,5]_-

4.1.3 .Corrosion
4.1.3.1 Introduction

As;a design criterion, it was hypothesized that all
design work should be predicated on the basis that the core vessel and all
other parts of the systew, which were subject to activation or to radiocactive _
contamination, would be apeoified of such materials and thicknesses as to
be able to withstand full power operation (125 MW), for a period of at least
10,000 hours without failure from corrosion by the fuel selected. Insofar
as 1t is posgible to predict, from dynamic and static corrosion research
at ORNL, this standard has been adhered to for the Fuel 30-Inconel-Secondary
Fused Salt System described. Final metal thicknesses were selected on the
basis of experimental results and personal experience {Refs. 39, h2,_hh, 45).
Fuel 30 and the seoondary‘NaF-LiFnBeFé fused salt mix wefe selected because
researohuand informed opinion showed that Inconel is a satisfactory container
for them at tha temperatures of operation antioipatad.

h.lf3.2 Corrosion Mechanism

 

The most cfitical location, as far as corrosion is con-
cerned, in this reactor is estimated to be the moderator cladding. The type
of corrosion to be expected is chromium depletiop, by diffusion and dissolution,
with hot leg-cold leg cycle accelerating mass transfer by solubility gradient.
The'chemioal reaction is UF) + Cro = CrfFp + 2 UF3.

Another possible source of trouble due to corrosion in fused salt Inconel
systems i1s a mass transfer buildup, or depositionzo% chromium in the cold leg
at a greater rate than inward diffusion can dispose of it. If such daposition
were localized, clogging of small passages might result. This type of buildup

was predicted for nearly all fuels tested., However, Fuel 30 was free from such

 
-52-

buildup after 1500 hours at 1500°F hot leg fiemperature in a thermal con-
vection ioop. (Ref, 58)

Dynamic hot leg-cold leg tests have shown that maximum initial attack is
about 5 mils in first thousand hours operation, and will average 2 - 3 mils
per thousand hours operafion at 1500°F. On this basis, 40 mils of Inconel
moderator cladding is expected to be sufficient for 10,000 hours full power
operation., It is tolhé hoted that the reaction which may be expected %o
proceed if BeO méééiéfibf.directly contacts fuel is

UF, + 2 BeOg=>2 BeF,+ U0,

2
This reaction would gradually concentrate the fuel on the surface of the
moderator rods. With unclad BeO, this deposition of UO2 on its surface

would greatly retard the reaction.

The nature of the Inconél corrosion is sfich that the corroded layer is
chromium poor, and characterized by unicellular voids. However, tests have
sfiown that even helium cannot penetrate the corroded layer. The strength
is greatiy lowered, but, bérring fracture and peeling of cladding, the UF), -
BeO reaction rate, even when entire thickness of cladding is chromium depleted,
1s controlled by rate of solid state 'diffusion of Be through the cladding.,

No great difficulty is expected on fhis point,

fhe corrosiofi rate in the heat exchanger tubes, based on extrafiol&tion
of 1500°F dynamic corrosion data with a 3000F hot-cold difference (Refs, 36
and 42) to 1200°F and 1000F differences is estimated as 10-12 mils maximum
pexr 10,000 hours .operation, on fuel side of tubes. Another favorable factor

1s that the fuel is already chromium rich (from contact with hot moderator

cladding) when it enters heat-éxchanger. This would tend to reduce the corrosion

t0 an even lower rate.

 
53

%.1.4 Physical and Thermal Properties

 

The physical and thermal properties of the fuel, as determined

by calculation, and by derivation from data contained in Ref., 4O are as

 

 

 

follows:
Density
Solid at room temperature (gm/cc) 4,09
Liquid ( ¢ - gnfcc, T = °C) | ©= k.03 - ,00095T
Liquid (63; Ts/f45, T = °F) Q- 253.0 - .0328T
Mean fiolumetric coefficient of liguid expansion per ©C 2,83 x 10~1+
Liquidus Temggrature about 525°C (977°F)
Ehthalpy, Heat Capacity
Solid (340° - 500)
Enthaipy (cal/gn) Ht - Ho'C= -12.6 + ,0215T
Heat capacity (cal/gm °C) Cp = 0.22
Liquid (5400 - 8940C)
Enthalpy (cal/gm) Ht - Ho®C=2,1+0,318T ~ 4,28 x 10-91°
Heat capacity at 1200°F Cp=0.26k
Heat of Fusion (cal/gm) Hl - Hs = 57
Thermal Conductivity
x (BTU/hr £t F) 0.5 (s0lid siab)
1.3 {liquid)
Viscosity
°F 1b/ft-hr £t%/hr
1100 - 23.0 | 0.098
1200 18.0 0.08k4
1300 14.5 0.069
1500 9.7 0.0k7
5k~

 

 

 

Prandtl Number 4.4 at 1100°F, 3.3 at 1200°F, 2.5 at 1300°F
Volume of Fuel in Core - : L.77 % 105 cm3

Total Volume of Fuel 12.7% x 105 cmd

U°3? Content of Fuel | 605 kilograms

4.1.5 Nuclear Properties

 

The use of Fuel 30 and Inconel cladding on beryllium oxide moderator
rods results ifi a rather large fuel concentration. Absorption cross sections
of the sodium atom is higher than is dgsi:able and #ery 1ittle moderation
is accomplished in the fuél. When testing'and_dgvelOPment work on nickel
molybdenum alloys éfid.fuels containing lithium.and beryllium has feen completed,
it is expected that critical mass-and.fuél concentration may be materially
reduced, For example, where use of Fuel 30 éictates.that 4O mils thickness
of Inconel éladding 59 used around moderator rods, use of nickel molybdenum
might permit a cladding thickness of perhaps 15 mils, with accompanying
neutron econouy and reduced fuel concentration. Incorporation of Li and‘Be
fluorides in the fuel would give shorter slowing down length and a smaller
size for the core. However, Fuel 30 and Inconel is the only system whose
technology is thoroughly tested and found satisfactory at this time.

h,1,6 Availability and Cost

 

Reactor grade NaF is commercially available at $0.20 per pound
and hafnium free ZrF) can be obtained at & cost of $3.50 per pound. To
prepares fuel mix.for the reactor, powdered salts are mixed and then treated
with hydrogen and hydrogen fluoride at 1500°F. This reduces the corrosiveness
by removing traces of sulfur, iron, nickel,-water, chlorides and other

impurities, Mixed, treated, fused 52 NaF - 48% ZrF) can be produced at

ORNL (Ref. 57) for a cost of'$7.50'per pound in thousand pound quantities,

 

 
=55~

- It was estimated that 20,000 -~ 30,000 pound quantities might be available
for $6.00 per pound , |

4.,1,7 TFuel Addition

The uranium burnup is compensafed by periodic additions of

(NaF), UF,. From the phase dlagram, Figure 4-1, it is noted that dissolubion
of this makeup salt in Fuel 30 proceeds so that only constituents of
consiétently'iower melting points result. (NaF)QUFh may be added as pellets
or powder &ireétly to the reactor., It may be melted and injected directly,
or it way beldissolved in a small quantity of fused salt solvent and injected
as,neéded.

The fuel concentration is dictated by the operational temperature and
amount of poisoning material in the reactor. As concentration falls or as
poisons bulld up, the reactor critical temperature decreasés. Fuel must
be.added when adjustment of the control rod can no longer maintain the desired
operating core temperature.

4.1.8 Fuel Reprocessing (Ref. 5)

 

Xe13? wi11 be_continfiously removed from the reactor, along with

a part of the 1135 precursor, and all stable xenon and krypton isotopes,

It is expected that rare earth fission firoducts will accumulate in the
salt mix; thelr solubility limits the problem to one of neutron polsoning.

Ruthenium, rhodium, and palladium plate out on metal surfaces.

Reprocessing of the fuel after several years operation will be required
to recover U7 from the spent, poisoned fuel before discarding radioactive
waste, The fluoride volatility process, which depends on the high vapor pressure
of UFg, is expected to allow uranium recovery with a minimum of effort. Thig

process is currently being perfected at ORNL, Figure 4-3 is a flow sheet for

 
~56-

 

 

 

 

 

 

$S8001J LI12A009Y wmiuea) LIT[I}RICA @prIoanig-31res pasnyg -- ¢-% "S1d

 

 

 

 

 

 

 

06061 OSMT-YT-TINYO
Ad313ISSVTIONN

 

 

Pj17- 40N FLSYM

 

 

 

 

D0.069-009

NOILYNIHON 14

1S ddsn4d

e

 

 

Yin-"447 - 40N

 

 

 

- .

 

 

 

(NOIld¥0S3q)

40N JLSYM
| Sy3gH0SaY
B
19N00¥d
°Jn a X
d34d a3d
; 40N 4DN
dvyl
v
o X
| dvdL |
F— 1 309 |
ny |
L = YSOdSIa A

<t

 

 

Y
4H + °H

 

J0069
NOILVNIEONTd
~OddAH

 

 

4H Yj47- 4oN
SNOYAAHNY

 

 

SINIWII3 73N N-4Z

 
57 =

the. fused salt-fluoride volatility uranium recovery process.

4,2 Secondary Fiuid Heat Exchange Medium

 

4.2,1 Introduction
On the basis of the following reasons, it was decideéd to

select a fused salt having the eutectic composition 50% BeF,, 30% NaF,
20% LiF, expressed in mol fraction percent, as the secondary fluid,

(1) The salt is non-reactive chemically with the fuel and with water.

{2) Leakage of the salt into the fuel would give a loss in reactivity
.(dué to L16 afisorption cross section) réther than an increase.

(3) Rather low conductivity and somewhat -high viscosity tends to reduce
thermal stresses in steam génerator and superheater‘tubes.

(4) Melting point must be reasonably below the critical temperature
of water, 705°F, in order to make steam generation feasible without an
additional transfer loop., The above ternafy eutectic composition was
selected from three compositions recommended by Ref. 56 because it possessed
the lowest melting point, 527°F.

(5) Intermediate loop prevents neutron activation of the steam and
consequent shielding of steam system components.

(6) Corrosiveness of this fluid toward Inconel is estimsted to be less
than that of Fuel 30 under identical conditions becauéé (a) no uranium is
present to datalyze the corrosion reaction and (b) lower wmaximum temperature,

i.e., 1150°F vs 1275°F (Ref. 42 and 45).
~-58-

4.2.2 Physical and Thermal Properties (Refs. 41, 51, and 56)

Comgosition

 

 

 

 

 

Mol % Wt %
NaF 30 30.52
LF | 20 - 12.56
BeF, 50 | 56,92
N Gfi/cm3(11000F)- Atoms /Cu3 (1100°F )
éodium' 12.0 - ) .329” | 861 x 10°%
Lithivm | 8.0 . |  .066 573 x 1072
Beryllium 20.0 _  - - .215 1.436 x 1022
Fluorine | 60.0 0 1.360 | 4.320 x 10°2

 

 

 

 

 

Melting Pt. - 527°F (275°C)

Density _____ 2.17 - ,000345 T(°c) -
Density at 655°F - 2.05 Gm/Cm3
Density at 865%F - 2,01 Gn/Cm>
Density at 1100°F - 1.97 Gm/Cmd
Specific Heat‘(Cp) - 0.57 cal/efi

Viscosity, Centipoises, estimated

at 620°F - 390
TO0CF - 200
8656F - 70

1100°F - 22

Thermal Conductivity - 2.4 BTU/Hr-Ft-CF

 

 

 
-5Q=

Viscosity, density and conductivity are given as predicted by responsible
ORNL personnel., Actual measurements are in process, but special eguipment'
needed was not available in time to permit determination before publication
of this report.

Bagis of deductions is the three component BeF_-NaF-LiF phase diagraum,

2
Figure k-4, On this diagram, composition selected is noted by T275: that
ig, ternary eubectic melting at 27500.

%.2.3 Disadvantages of Fluid

(1) A comparison of the heat e#changer volume needed tao transfer
125 MW using this salt and using sodium has been made. It was found that
the reactor pressure vessel size could be reduced congiderably by using sodium,
with a consequent shield weight reduction.
| (2) Melting poinf of the fluid, 5270F is so high that a shutdown of

the secondary system pumps would require that system be drained to prevent
freeze up of salt. Considerable care must be exercised to assure that
boiler feed water is preheated before introduction into boiler, or freeze
up may résult. |

(3) Pumping power is considerably greater with fused salt fluid than

with sodium, because of greater viscosity.
-60~

693 b8

 

   
    

    
            

066
40N j i ; T 417
Qt e
0.G3 (6499 - 411 - 4ON)
C6e¢ GG¢ 08t
498G -40N %409 - 41712
OrAY

(2498¢ -317-40NGS)— gie3

2ee3
b1Z
08¢ \ 5
2508, 40N (2498 - 417)

9Gg3
0Le4d

Do NI 34V SIUNLYHIJSWIL SNCINDH
SISHHLINIYVL NI

NMOHS SAONNOdWOD SNdINosans
J1L03L1¥3d =d

ONNOJdWOD AYVNY3L =021

2HS
2409 DI1D3LNT AHVNYIL =1

21L03LN3 =4

1SS OMO-YT-INFO
IVILNIQIANOD

 
wBlw

5.0 MATERIALS SELECTION

5.1 Structural Material

 

Once Fuel 30 had been selected, the results of several years testing
(Ref. 39) the dynamic and static corrosion resistance of structural materials
made perfunctory the selection of Inconel as the primary structural material..
This included 1ts use for pressure vegsel, moderator cladding, primary
heét exchanger structural material, pumps, and all other surfaces in direct
contact with the fuel e#cept the nickel reflectors, As indicated in our
discussion of the fuel, Sec. 4.1, developments in nickel molybdenum alloys
now underway are expected td change the fuel and container materials picture
in the foreseeable future. Nevertheless, this design study 1s based on
present technology and already proven systems. (See Sec. 4,1.3 Corrosion)
Some of.the results of corrosion research islpresanted as Jjustification for
metal thicknesses énd materials chosen., Figure 5-1 shows stress'elongation
and rupture curve for Inconel tested in Fuel 30 at 1300°F. Figure 5-2
shows temperature dependence of gtress rupture properties of Inconel in
Fuel 30. Figure 5-3 shofis effect of section thickness on creep-rupture
properties of Inconel tested in Fuel 30 at 1500°F at 3500 psi stress., Figures
5-1, 5-2, and 5=3 support choice of Inconel with Fuel 30, and thickness of
tubing and cladding specified., Our specification of 40 mils wall thickness
of primary heat exchanger tubes is based on research leading to Figure 5-3
and advice by informed ORNL personnel {(Ref, 42). Figure 5-3 indicates that
creep resistance of Inconel immersed in Fuel 30 at elevated temperatures

shows a remarkable improvement when section thickness reaches 4O mils.
-62-

000°0F 0006

 

C9ESE IMC-YT-TINYO
Q31dISSVIONN

(worydeD yjrm 10a098) 4 L0081 38 08 "ON
HeS pasnd Ul pajsa] [9UOOU] POATIVSY-~-SVY I0] 2AaIn) udiseg -- 1-¢ "8td

(44) INIL
000¢ 000} 004G 00¢ OO0/ 0g 0¢ Ol G

J¥NLdnNy | Y oL G o6 2

Yo b

% G0

000

000¢

0006

000 0}

000'02

(1sd) SS3IUIS

 

 
-63-

000 0¥ 000%

el OMO-41-TIN¥O
Q31AISSVIONN

Q002

(worided yjtm 38x09g) Qg "ON I[eS
pesng pue uoday ur 4 (0S91 PUB 00ST ‘00%T }® PRISSL 19U0DU]

PoATEO9Y -5V JO serlasdoad sanidny ssaxig ayi Jo uostaedwo) --
(44) IWLL
000} 00¢ 00¢ OO0} O¢ 0¢ O

Z-¢ "314

 

000}

000¢

000¢

0000}

000'02

(18d) 8S34IS

 
TIME (hr)

800

700

600

500

400

300

200

100

-84 -

UNCLASSIFIED

ORNL—LR—~DWG 17921

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fig., 5-3 -~ Effect of Section Thickness on Creep-Rupture Properties of
As-Received Inconel Tested in Fused Salt No. 30 at 1500° F
under 3500 psi Stress,

52 % 37 % —
S, 0
/' RUPTURE
45%,/ |
ey
/V’dr’
5 %o ®
159 f A~ L ——
T
6‘70 2c70 A
— 1 ]
o 20 40 60 80 100 120

SPECIMEN THICKNESS (in.)

{(Secret with Caption)

 

140 (x10™2)
~65-

Insofar as 1t is possible to predict from dynamic and static corrosion
data, Inconel thicknesses have been chosen so that, after design lifetime
‘has passed, sufficient sound void-free metal remains to provide stress
registance édequate for the barfiicular use involved. |

Inconel has also exhibited superior resistance against chloride stress
corrosion over most conventional materials, Because of the severe prohlems
that have been attributed to this in the steanm generating equipment of both
mobile and stationary filants, it is recommended that it be used for both

the steam and salt side of this equipment.

5.2 Moderator

Since all surfacés in contact with the fuel were of necessity Inconel
(except nickel) it was necesséry to choose a moderator of low neutron
absorption which would permit the reactor to go critical with a reasonably
small core volume, 'Consequently, after investigation (Ref. 43) BeO was
selected as the leading prfiven moderétor which could withstand the temperatures
‘expected. A cladding of 40 mils vas considered necessary, as previously
discussed in Sections %4.1,3 and 5.1,  Whi1e this thickness of Inconel cladding
does not make for neutron econony, or for low fuel loading, nuclear calculations
indicated that the réactor could be expected to. operate satisfactorily.

One inch diameter test pleces of BeO ceramic were exposed in the MTR
and showed satisfactory thermal stress resistance (Ref. 45, 55), The diaméter
of 3/4 inch selécted for this application was based on extrapolation of

these results to the higher energy deposition rate expected. See Appendix
6.1.
66

5.3 Reflector

Because of the poor moderating proPerties of the fuel and the somewhat
high thermal neutron capture cross sectlon of the -core, due to Ne and Inconel,
a8 fast neutron reflector constructed of pure nickel wasg choeen. Consequently,
calculations indicate a large percentage of epithermal fissions. The only
fair heat oonductenoe of nickel necessitates the circulation of a small
portion of the fuel through the reflector to equeiize temperature and lower
thermal stresses. ‘It is not considered necessary to ¢lad the reflector

for corrosion resigtance, which is satisfactory unclad,

5.4 Poisoned Moderator Region

 

An annular ring of boron beering, beryllium_oxide rods, clad with
Inconel, with interstices filied with borOn bearing Inconel.rods forme
the neutron shield, Calculations based on an everage thermal flux of_lo}o
show that helium generation over a period of iQ;OOO full power hours is
about .0l cm3 (STP) of helium per cmd of BeO, Since BeO may be about 6%
of theoretical density, no significant pfeeeure will be generated,

Between the beryllium-boron region and the heat exchanger region an
Inconel clad, copper-BhC cermet layer is interposed as a thermal neutron
abeorber, to prevent escape of thermal neutrons to the heat exohangere.
The beryllium-boron regionm, although heavily poisoned, containg a eource
‘of thermal neutrons due to thermalization of fast neutrons from the core,

Copper-BhC haslbeen satisfactorily fabricafed, cOntaining 25 volume
percent of BjC, to a fheoretical density of 95%, by cold pressing and hot

rolling (Ref, 36).

 
 

 

-67-

5.5 Design Properties of Materials
Appendix 5.1 shows the design properties of Inconel, beryllium oxide,

and nickel used in this . study.
~68-

6.0 REACTOR AND PRIMARY HEAT EXCHANGER DESIGN

6.1 Introduction
The basicbreactof design is conceived as being a pregsure tight
cy}indrical vessel cofitéining a circuléting fluoride fuel., A primary
objective of the design was to minimize its size and weilght in order %o
redfice its contributibn to the overall system specific weight. In addition,
& small reactor design is desirable because of the large effect it nay
“have in turn on the.size of both the primary and secondary shield,
The volume of the reactor is basically dependent upon; 1) the
| nuclear properiies of the fuel as it affecfs both the.éritical size and
limiting power densities, and 2) methods which can be devised to remove
the_fission heat from the circulatihg fuel, The establishment of an
-allowable critical size and fuel loading as well as other nuclear con-
siderations are discussed in detail in Section 8.0, The methods of
selection and optimizing a heat exchanger configuration are presented
later in this section, (See 6.3),.
Pogsible alternate solutions or approaches‘to the various problems
are discussed in the appropriate sections along with tpe reasons (either |
engineering or arbitrary because of time limitations) for the selections

made,

6.2 Reactor
The basic configuration, illustrated by Figure 3-2, is approximately
80 in. in diameter and 80 in, high. Its total net weight is calculated to

be 69,700 1bs (Appendix 13.1). Centrifugal fuel pumps located in the

 

 
 

-69-

reactor head are used to circulate the molten fluoride fuel up through a
central c:itical region, and then through an annular peripheral downcomer
which contains the primary heat exchangers. Heat is removed in this region
and the fuel is again circulateq up through the core.

6.2.1 Internal Arrangement

Calculations for the central core region were based on it being
qui#alent t0 a cylinder 75 cm in diameter by 80 cm high, This was modified
for design purposes to an octagon shape for a more even moderator rod spacing
and tapered ends to gain extra core volume. An optimum volume fraction
of fuel for the core was found to be 50% (Section 8.0).

Fuel cooled cylindrical beryllium oxide rods, clad with Inconel for
corrosion resgistance, were used for moderation purposes, These were
equispaced throughout the core on a triangular pitch under the. distance
between centers being defined by the rod size and the desired volume fraction.
Taper fittings were utilized at both ends of these rods to provide for the
proper area and flow distribution. These rods would be held in ‘the bottom
support plate by & bayonnet joint and left free to expand in an axial
direction to eliminate thermal stresses. A hollow ring is attached to each
rod at the end of the upper taper. This will maintain préper rod spacing
and still.provide a suitable flow passage; These rings may be interlocked
to prevent rotation and hence uncoupling of the bottom bayonnet Joint, but
still allow free axial motion,

The effective vertical boundaries of the core region are fixed by poison
material located in the ends of the moderator rods., This poison material,
beryllium oxide plus boron 10, also helps to reduce end leakage as well as

to cut down on fissioning in the entrance and exit plena,

 
~70-

A 40 mil cladding of Inconel is required around tfie beryllium oxide to
provide propér corrosion resistance for the 10,000 hr design life (Ref.
Section %4.1.3), Since this thickness is fixed and not a function of moderator
rod size, it is of considerable nuclear importance to use fewer large rods
rather than many small rods in order to reduce the total amount of Inconel
poison within the core., However, the maxifium gize 1s limited not by the
nuclear aspects such as self shielding of the fuel, but by thermal stresses
due to heat generation within the woderator material,

The feasibility of using beryllium oxide as a moderator material has
been satisfactorily demonstrated under cyclic reactor conditions in the
MTR (Ref. 54 and 55). Using this information, calculétions were made to
limit the design stresses for the present system to that found to be allowable
in the above tests (Appendix 6.1), This limited the moderator rod size,
without cladding, to approximately 3/4 in? for the preseni. Because no
indications were found in the MIR tests to indicate that higher power
densities could be allowed, this minimum size could possibly be increased
in the future when substantiated by additional test programs. Calculations
of the temperature rise across the boundéry'layar (EOOF) and through the
mcderfitor rod (143°F), also included in Appendix 6.1, indicate that a
maximum centerline moderator temperatfire.of 1491°F is to be expected. This
is well within the operational limits of this materiél and approximately
equal to that of the MIR tesis.

The moderator elements may be fabricated by inserting slugs of BeO
B/h in. in dlameter by 2 in, long into Inconel cans of sultable wall thicke
ness. In the MIR tests, improved heat transfer out of the moderator material

was reallzed by utilizing helium in the small clearance gap required between
-T1-

the slug and cladding. Stress calculations indicate that a shrink fit of
the cladding around the BeO could be used in conjunction with the above to
obtain further improvements.

A nickel blanket, approximately 6 in. thick, is incorporated around
the cylindrical side of the reactor core to offer advantages both as a
reflectaf and a shield, High energy leakage neutrons are inelastically
reduced to a lower energy level and scattered back into the core %o
improve the .core criticality and power distribution., Also because of its
close proximity t§ the core 1t acts as an effective shleld, from a weight
standpoint, for both:prompt gammas and neutrons. A detailed stress and
heat generation analysis was not made on the reflector, However, because
the reflector supports no load other than its own weight, it can be alloved
to operate at high'temperatures and in the plastic region so that thermal
sfiresses may be effectively annealed out. Fuel flow channels of approximately
2% by volume should be more than adequate for cooling the reflector.

In order to minimize the dctivation of the secondary fluid, it is
necessary to reduce the.neutron flux in the primaxry heat exchanger region
as much as possible. To help accomplish 4his; a region.containing BeO
to thermalize fast neutrpns and boron to capture the thermal neutrons is
included outside of the reflector. This region contains closely packed
3/& in, cylinders suitably clad with Inconel and 1s approximately 5-1/2 in.
thick. Small boron bearing Inconel rods are placed in the interstices of
these cylinders for additional shielding and %o reduce the fuel and hence
fissioning in this region., Sufficient flow areas will still exist within
the interstices of fhe large and small rods to provide for cooling.

To assure absorption of neutrons that are thermalized in the outer
~72-

edge of the above region a thin layer of boron carbide in a copper matrix
is then placed around the above region.. The feasibillty of using these
materials are discussed in Section 5.4,

A single control rod thimbie, approximately 4 in, in diameter, extends
through the length of the.caré. A clearance gap of 0,1 in, on the radius
is allowed between the thimble and the control rod to assure free operation.
To facilitate fuel cooling of the poison rod this gap would be filled with
either a salt or a liquid metal. A small reservolr éould be included in
the reactor head in order to keep the'fihimble full as the controi rod is
withdrawn. For the purpose of the control rod worth evaluation (Section
8.2.3) it was assumed that this gafi waé filled fiith sodium. Because of
t he small quantity involved it was felt that this wéuld not be a serious
hazard. |

A low point drain hole is located at the bottom centerline of the
reactor vessel to provide a place for both filling and locating an
emergency dump or‘blcwout valve. This is incorporated into the bottom
lateral support of the control rod fihimble.

A thermal shield is located Just outside of the heat exchangers to
reduce the gemma and neutron heat generation problem in the reactor vessel.
A small gap is placed between the thermal shield and core vessel to provide
. for cooling. Relatively cool fuel from the exit of the primary heat
exchafiger fiill flow up through this gap and into the fuel expansion tank
in the reactor head. This flow hag the additional advantages of providing
increased circulation through the head to remove decay heating and to
decrease the temperature in this region to help alleviate the snow problem

(Section &.1.2),
=13~

As mentioned previously the moderator rods are fixed only at the
bottom in order %o allow free expansion and thereby reduce the thermal
stress problem. For a similar reason, the remainder of the internal
structure, that is, the reflector, control rod thimble, and the basgket
supporting the poison rods, are suspended only from the reactor head.

The only exception to this is the primary fieat_exchangers which run
straight through the vessel. However as explained in Section 6.3.k4, the
thermal stresses obtained were found to be tolerable,

6.2.2 Vessel Design

One of the major advantages of a fused salt system is that
due to the low vapor pressure of the fuel, it is necessary to contain
dnly small pressures wlth the core vessel. With a minimum pressure of
30 psia required within tfie system to prevent pump cavitation and a
pressure rise of approximately 35 psi requiréd across the pumps to provide
fuel flow a normal design differential pressure of=on1y 50 psi is obtained,
Basilcally this would require a wall thickness of less than one-half of an
inch, However both because off design conditions would undoubtedly occur
and navy requirements of meeting 20 to 30 .g shock loads are required, this
thickness was increased to 1-1/2 in. using the ground rules proposed in
Ref. 11, but adapted to Tnconel,

“Although not shown in the reactor drawing, Figure 3-2, cooling coils
must be included in the head design to také care of internal heat generation.
The possibilities exist of using either the pressure drop across the fuel
pumps to force the flow of a small amount of Ffuel through suitably con-
structed cooling tubes or to circulate a small percentage of the secondary

salt,

 
7l

Two possibilities which exist as to the most economical method of
maintaining reactor installations of this type are discussed in Section 11.5.
Basically they affect the vessel design in two different ways: 1) the
vessel should be designed so that 1t may be reasonably feasible to assemble
and disassemble 1t seve?al times, or 2) the design should be simplified
with the idea that only one assembly wbuld be required. Because these
concepts were well beyond the scope of this study it was decided to present
a reactor design that could satisfy both. To accomplish this only the
feasibility of a system allowing disassembly had to be shown because this
was the most complex. The alternate solution, being of simpler design, |
was not illustrated as the Joints, flanges, etc., would just be changed to
welded structure. In both concepts as deséribed, it was felt to be-
desirable from both an economic and a Weight standpoint to remove the
reactor from the ship for any maintenance.

To facilitate easy removal, the head is simply butted against the
dore véssel and held by the use of a fianged Joint. An omega type seal
is welded across the Joint to provide proper leak tightness. This isg in
turn backed up by a steel "0O" ring both as a safety precaution against
possible fallure of the omega seal and to help prevent fuel from easily
- flowing into the ring. If fuel settled in the seal fing, it would éddl
to the decontamination problem upon reactor disassenmbly, Hofiever, corrosion
would not be a problem because the chieflsource of corrosion with Inconel
is8 with a dynamic system fiowing over a large temperature difference; Be-
cause flow is prevented in this region, the seal weld will remain at con-
stant temperature and corrosion would be limited to that caused by the

initial chromium solubility.
;75_

The reactor may be disaésembled remotely by removing the hold down
bolts and cutting the seal weld; Omega type seal welds of the type
recommended should not present a problem as they have designed for use
in pressurized vater reactors at pressures up to 2500 psi, Also mechanisms
for remotely cutting and rewelding these types of owega seals have been
developed for use in the marine PWR systems,

In order to be able to remove the reactor head and replace the primary
heat exchangers it is necessarylthat the secondary fluid inleit and exit“
pipes be detachable from the core vessel, Two suggested methods for
doing this are illustrated in Figure 3-2. The one in the reactor head
utilized a concentric tube-with the Jjoining weld being made approximately
12 in, off the reactor head for access purposes. This type of joint gives
good vrigidity but has the disadvantage of allowing only a limited number
of welds to be made. Also since it is a strength weld it wouid be more
difficult to mfike remotely., The bottom connection is fashioned after a
bridgeman closure which is used on many high pressure autoclaves. The
closure provides the structural strength while an omega seal weld similar
to that previously described is used to assure leak tightness. However,
the rigidity of this type of connection under side loads and thermal cycling
is not known.

Methods for the head closure and secondary pipe attachment were
not given detailed consideration bub are offered as one of many possible
solutions.

6.2.3 Structural Arrangement

Two possible solutions exist for supporting the basic structure,

however, a detailed study would be required to determine the optimum., The
16—

first of these as shown in Figure 9-3 simply rests the reactor on supporting
structure allowing free vertical expansion. Side play would have to be.
limited by guldes. A second approach which at first hand appears to be more

advantageous would support the reactor through a beefed-up section Just

. below the head flange. This would not only take the Welght of the head

 

and most of the internal structure of f the reactor side walls but also
simplify the ba51c installation.

6.2.4 Fuel Punips

Three'centrifugai pumps are located in the reactor head to
provide for fuel flow and to &id in the removal of the fisgion product
gases. These pumps have common inlet and exit plena and are sufficiently
overdesigned 50 as to allow almost full'poher'reactor operation in the
event of a'single pump failure,

To provide for.a lightweight and variable speed system (required to -
compensate fér pump failure) a steam turbine driven motor was selected
for tfiese pumps, A small_AC.eiéctfic_motor which could be clutched into
the drive sgaft would also be incorporated to maintain circulation under
zero power operation. Also becafise'if could be gwitched into the ship's
emergency electric power system, it would serve as a safely device in case
the steam flow to the turbines was interrupted.‘

A more detailed description of these pumps is given in Section 11.8,

6.2.5_ Pressurizer and Expansion Chamber

6.2.5.1 Pressurizer

It is necessary to provide & pressure of at least_
15 psig at the inlet of the fuel pumps to prevent cavitation., This pregsure

is appiied by means of bottled helium gas at startup. After startup, a

 
helium gas differential pressure of a few pounds is maintained at pump
shaft over that in the expansién chamber, 1o prevent escape of fisgsion
product gases along pump shafts. After initial filling, stable xenon
and krypton generation can be used to maintain pressure. Off-gas systems
to provide for poison gas removal are discussed in Section 11.6.
6.2.5.2 Expansion Chauwber

-Thé pumps are so designed as to cause a swirling motion
of fuel in the expansion chamber, so that equilibrium gas-liquid cdncentration'
is quickly reached; A small stréam of 11750F fuel 1s brought up to the
chamber through a passage between thermal shield and core vessel, ané
circulated thrdugh the chamber to remove heat generated in fhe chamber
by fisgion product decay and by fission, (See Appendix 11;2 for heating
calculations).

It is calculated that 150 kw is generated in gas, 157 kw is generated
in liquid due to fission, and 93 kw is generated in liquid due to decay
heat, It is obviously necessarfi that some heat removal system be incor-
porated to cool off the expansién chamber roof due to this and internal
heat generation as discussed in Section 602_02° Assuming that one-half
of the gas heat is absorbed by the roof, a cooling rate of 75 kw, or about
250,000 Btu/hr would be expected at full power. Allowing a 50°F rise in
temperature, this will require circulation through the head ofbabout
8800 1b of fused salt per hour,

A stream flow of 50,600 1b of fuel per hour is required to provide
cooling for liquid in the expansion chamber to prevent snow formation,
Arrangements have been made to bleed off a stream of fuel from the cool

region (1175OF) at the bottom of the reactor, so maximum fuel temperature
18-

in expansion chamber should be the same as maximum temperature in resctor,
that is, 1275°F with a conservatively estimated temperature rise of 1100°F,
Thus, maximum temperature of liquid in expansion tank will be about 750F
less than 1350°F,'maximum temperature at which'sndw problem may be neglected

using fuel 30 (See Ref, 50).

6.3 Primary Heat Exchanger

6.3.1 Desgign Criteris

The.design criteria for the?primary-heat'exchanger is the

same as for the system as a whole; that is, obtaining the lowest specific
~weight for the overall power plant consistent with a life of ten thousand
full-power hours, Méeting this goal required the 0ptimiéation of a com-
bination of several quantities which fary with heat exchanger design.
These are: heat exchanger weight, primary shield weight, pump weight,
and pumping horsepower.

Tfie variables of the heat exchangef design were placed, essentially,
in two categories: 1) those which could be fixed early in the study
dependent on the expefience of others doing similar work or due to the
limitations imposed by the rest of the systefi, and 2) those which were
varied in an extensive parameter study to determine the most favorable
union of these quantities,

6.3.2 Basic Design

The primary heat exchanger is of once-through, counterflow
design, The heat transfer surfaée is providefi by straight Inconel tubes
on a delta laftice which are contained.in an annulus surrounding the reactor

core,
..79_

The headers are segments of tori which have an elliptical cross section.
These headers circle the reactor core at the top and bottom of the heat
exchanger, each segment having a nozzle which penetrates the pressure vessel
and primary shield (éee Figure 3-2)}. To provide additional area on the
header surface, the major axis of the ellipse is longer than the width.
of the heat exchanger and is tilted'with respect to the horisontal,

The'secondary coolant flows through the tubes, entering at the bottonm
of the heat exchanger. The fuel flows on the outside of the tubes and
enters at the top of the exchanger;

The physical dimensions, flow rates, temperatures, temperature
differences; and heét transfer coefficients for the final primary heat
exchanger design are tabulated below. This heat exchanger would be capable
of removing 125 megafiatts ofbheat from the reactor. See Appendix 6.2

for calculational details.

Heat Exchanger Inner Diameter | 53.5 inches

Heat Exchanger Outer Diafieter 73.7 inches

Heat Exchanger Length 48  inches

Tube Inner Diameter - | .120 inches

Tube Outer Dianeter .200 inches

Tube Spacing .030 inches

Fue:l ¥Flow Rate 16.2 x 106 ibs/hr
Secondary Coolant Flow Rate 7.48 x 10 1bs/hr
Temp. of Fuel Entering Heat Exchanger lETSOF

Temp., of Fuel Leaving Heat Exchanger 1175°F

Tem§° of Coolant Entéring Heat Exchanger 10500F

Temp. of Coolant Ieaving Heat Exchanger 11500F
«80-

Mean Temperature Difference from Fuel

to Secondary Coolant | | 1250F
Outside Heat Transfer Coefficient - 1836 BTU/Hr—OFnF’G2
Inside Heat Transfer Coefficient 914 BTU/Hr-CF-Ft°
Overall Conductance | 37h BTU/Hr-"F-Ft°

Straight tubes rather than U-tubes were incorporated in the primary
heat exchanger because it would have been difficult to obtain as much :
heat transfer area in a given volume with U-tubes. Also, inlet and
outlet headers would have to be in the same end of the reactor, which would
further complicate the space problem. The wain advantage of a U~tube
exchanger would be the reduced 10ngitudinal thermal stresses, However,
as will be indicated in a later section, longitudinal thermal siresses
are not expected to be a major problem in this heat exchanger.

| The heat exchanger inner diameter and effective length were detérmined
by the reactor core design., It 15 necessary that the heat exchanger

tubes be nested closely about the reactor, and be sbout the sane length

as the reactor, in order to achleve fihe most compacf design,

Tube wall thickness was fixed at .00 inches, primarily because of
corrosion to be expected during ten thousand hours of operation. Although
no corrosion data are available at the temperatures encountered in the
heat exchanger, 1t has been predictéd that a maximum corrosion of twelve
mils on each side of the tube could be expected {see Materials Section h.2)}
This corrosion is of a penetrétive nature, with the maximum depth of cprrosion
being given for a few scattered displacements., Since it is unlikely that
penetrations on both side of the tube would line up, and because the dis-

placements are not interconnected, a large safety margin is realized in the

 
-81.~

twelve mil estimate. However, twenty-five mils were allowed for corrosion,
with the remaining fifteen being sufficient to contain the pressure and
thermal stresses.

The upper fuel temperature was set at 12750F to keep the corrosion
within acceptable limits. From examination of other proposed reactor
systems of a similar nature, a mean temperature difference between the
fuel and secondary coolant of 125°F was decided on. This is a compromise
value which will give both reasonable heat transfer and rermisgible thermal
stresses., Further investigation of this system should include an examination
of the effects of changing the temperature difference.

Many considerations were involved in the selection of a 1L00°F
temperature drop across each fluid circuit, It is desirable to keep
the temperature drop as large as-possiblg in order to reduce the flow
rates, and hence, pumping requirements. Also, it is necessary to keep
the temperature of the secondary fluid above the melting point of the
fuel which is 970°F, Using a mean temperature difference between the
two fluids of 1250F and a 100°F drop across each circuit, the lowest
temperature encountered in the secondary coolant loop will be 1OSOOF,
which should be safely above the fuel melting temperature.

Since the heat ekchanger must be capable of removing 125 megawatts

or 4,27 x 108

BTU/hr, choosing the temperature drops automatically sets
the flow rates.

In the final design, the tubes were spaced .030 inches apart on a
delta lattice. The delta lattice was chosen over a square lattice because

it permitted inserting more tubes of a particular size into a given space.

The .030 inch spacing was established by a parameter study which will bve

 
-82-

demonstrated in a later paragraph.

The tube spacing can be maintainéd by one of several wmethods, Most
of the present small-fube high performance heat exchanger tests utilize
flattened wire spacers, which are perpendicular to the tube axes. It
was for spaceré of this type that heat transfer and pressure drop cal-
culations on the fuel side of the heat exchanger were made. More recently,
some work has been done with helical spacers, wrapped about each tube.
Preliminafy results indicate thatlthis type of spacers will give about
the same heat transfér with a ldwér pressure drop.

To facilitate welding, it is necessary that the tubes be spaced at
least .075 inch apart on the tube header (see Ref. 67), This requires
that the headers have a surface area greater than the cross sectional
area of the heat.exchanger, but preferably will fif into the same annulus.
As described previously, this waé accomplished by méking the headers
elliptical in cross section and tiliting the ellipse with respect to the
horizontal.

The heat exchangér will be fabricated in bundles of approximately
six hundred tubes each, -This is approfiimately twicé the number of tubes
per bundle presently contemplated for the more ‘complex ART fuel-NaK heat
exchanger (see Ref, 36, Section 4.1). -Each bundle is to be tested individually
in order to simplify inspection and preclude the necessity of scrapping an
entire heat exchanger for a single tube-header Joint failure. 8ix of
these bundles will then be welded together and capped to make up one header
segment. There will be twelve such segments, each one having a nozzle

penetrating both the upper and lower heads,
~83=

6.3.3 Parameter Sfudz |

In the pérameter study tha£ was made, the variables were tube
outer diameter and tube spacing. For a given tube diameter and spacing,
a heat exchanger outer dlameter which would give the required mean tem-
perature difference of 125° fias determined by an iterative process.

For é selected tfibe size and spacing, an assumed outer diameter
of the heat exchanger was used to calculate film coefficients., Flow in
the tufies vas at all times leminar and an empirical equation for film

conductance during laminar flow (see page 232, Ref. 17),

_ K 1/3
hi = 1,75 "EI

was used. For flow outside the tubes an experimental correlation

0

(See Figure 7.6)

O

h = ell'T .u.ék_]g—. (Prf)oh‘(Re)o36

was used, which takes spacer effects into account. An expression was
derived to give the weight of the heat exchanger plus primary shield
- for each configuration. Pressure drops and pumping horsepower can be
determined from flow rates and heat exchanger geometry., Pressure drop

in the tubes was calculated from (See pages 45 and 50, Ref. 15)

where

 
-84

for laminar flow. For flow outside the tubes, an experimental expression

for friction factor

f = 5°T .
(Re) 50

was used, This.expression includes the effect of spacers and is given

in Ref', 13. The weights of the pumpé and drive motors were estimated

at 25 1bs/FHP. The weight of the machinery and equipment not affected
by heat exchanger design was determined, it was assumed that the steanm
generation equipment could provide steam for 35,000 shaft horsepower,
normal auxiliary équipment, and 600 pump horsepover, When the calqfilated
pumping horsepower was less than this, the shaft.horsepower was increased
by the diffgrence divided by the efficlency of the pumps,

The total weight was then divided by.the adjusted shaffi'horsepower
to, give the specifilc welght. Although this method will not give. the
exact specific weight of the power plant, it will indicate the configuration
vhich will give the lowest specific weight. Tube diameters were varied
from ,1875 inches to .25 inches and spacing from .020 inches to ,04O
inches. The results of this study are shofip in Figuré 6.1,

6.3.4k Stress Considerations

 

Exténsive thermal stress calculations were not wmade for the
primary heat exchanger. However, the thermal stresses due to the tem-
perature drop across the tube walls were determined, and also the stresses
which will be present due to tfie difference in longitudinal expansion of
- the pressure vessel and_heat exchanger tubes for several extreme cases

were calculated,
 

(S3IHONI) ¥3IL3IWVIQ ¥3L1NO

 

 

 

 

 

 

 

 

 

 

 

28 08 8. 92 b2 22 02
* | \.\\\QWI/ 8°8G
3 = TN
&
b A s¢ /
—
3 » 0°'6S %
, .
5 \ —— :
\ v
\t\\\“m_\oqmm ._.__zy/ 2°6G
b 66

 

 

 

 

 

 

 

 

 

SONIOVHS 38Nl SNOIYVA HO4 H3ILIWVIA H3ILNO
HIONVHOX3 LVIH AYVWIHA SA OIlVY
43IMOd 7/ LH9I13M 40 31VWILS3

1’9 34NOI

 

 

 

 

 

i
-86“ L

Due to the rather poor heat transfer characteristics of both fluids
in the heat exchanger, most of the temperature drop is taken across
the fluld films. With a small temperature drop across the tube wall,
the thermal stresses.are also quite small.

It was estimated that the contéinment vessel will be at an average
temperature of approximately 12250F, The inside of the vessel will be
cooled with fuel having a temperature of 1175°F and the average tem-
perature will be some 50°F greater than this due %o heat generation in
the vessel. The tube temperature can be thought of as being at an
average between the mean wall temperatures or about iléSOF. This gives
a‘temperature difference of 360F between the pressure vessel and heat
exchanger tubes, It was assuned that this difference in thermal elongation
would be taken up by mechanical elongation of the heat exchanger tubes
only. This was figured for several extreme situations, one in which the
tubes ran straight from one header to the other and were fixed at both
ends. In this case, the stress in the tubes remained below the yield
stréss. Another case was considered in which the énds of the tubes
were bent at right angles and then fixed to the header. In this case,
the difference in elongation was assumed to be taken up by deflection
of the étub ends. Maximum stresses will again remain below the yield
strenéth if the stub ends are at least .4 inch long.

In the event that detailed thermal stress calculations prove thét
straight tubes are untenable, the tubes could be wrapped partially around
the reactof to provide flexibility in ofder to alleviate these stresses,

The grestest pressure difference across the tube wall will be less

than 100 psi even if the pumps on either circuit should fail, A cal-

 

 
-87-

culation was made, assuming that the pressure inside the tubes was 100

psl and the necessary wall thickness came out to be only .01 inch inches.
For the headers, thermél stresses were not considered; however,

pressure stress calculations were made, again assuming an internal pressure

of 100 psi. For this condition, a wall thickness of .1 inch and end cap

thickness of .375 inch were determined.
 

-838-

7.0 STEAM GENERATING SYSTEM

7.1 Introduction

One of-the wa jor problems in adapting nuclear power to naval vessels
has been the development of a dependable steam generating system that will
deliver steam at conditions that are compatible with ‘the requiréments for
efficient steam turbine performance. In the design of the steam gefierators
the group endeavored to duplicate the existing steam conditions of +the
931 class destroyer, Thesé conditions are 263,300 1b/hr per boiler room
at‘950°F and 1200 psi. it was decided to replace one boiler room with
& nuclear_reactor-steam generating sygtem.' Other design criteria were
to keep the thermal stresses as low as possible, tb make the system as
light and compact as practical, and to have a realistic and somewhat
conservative system, |

In the preliminary analysis of the system it was decided that the
reactor power shquld be 125 megawatts, Therefore, the steam géneréting
equipment was designed to remove 125 megawatts of heat. When the actual
steam cycle data for the class 931 destroyer was received a complete
heat balance revealed that only 95.9 megawafts of heat was necessary to
supply the steam for the full power of 35,000 shp. This makes possible
the operation of the steam generator at lower temperatures and lower.zst's
throughout the system. Detailed calculations of the design speclfications

are presented in Appendix 7.1.

7.2 Molten Salt Cycle Selection

In selecting a cycle or rather a system for steam generation the

 
-89-

group was confronted with the problem of removing 125 megawatts of heat
from a molten éalt coolant., The temperature of this salt as it leaves
the primary heat exchangers is to be 11500F, and it is to reenter the
primery heat exchangers after losing only 100°F in temperature. There-
fore, to remove the full-power 125 megawatis 1t is necessary to circulate

6 1b/hr., One of the design criteria

the molten salt at a rate of 7.49 x 10
is to keep the temperature drop across tube walls below 1O00°F and since
the saturation temperature of water in the boiler is 572OF, then when
allowance is made for the bolling water film température drop and for
the molten salt £ilm temperature drop, it is found that the salt entrance
temperature to"the boiler should not exceed 800°F. The superheater is
to bring the steam temperature up from 572°F to 9750F, Because of the
high temperature drop across the steam film an entrance temperature
to the superheater of 1150°F and an exit temperature of 1126° would not
exceed the 100°F drop across the tube walls, Therefore, the problem
is to'lover the salt temperature from 1126°F to 800°F and then to raise
it from 73M0F, the boiler exit temperature, to 1050°F in order to return
it to the primary heat exchanger.

The two methods considered for meeting this problem were to use either
a regenerator heat exchanger between the boiler ang superheater, or blenderé°
In the regenerator system the .1126°F salt would enter and the 1050°F sait
would leave the hot end of the regenerator heat exchanger while 800°F
salt would leave and 734° salt would enter the cold end. This meant that
some 420 megawatts of heat would have to be exchanged in the regenerator.
Due to the rather poor heat transfer characteristics of the molten salt

and the low log mean temperature difference avallable, a tremendous heat
~50-

transfer area would be required. This in turn led to a very large salt
volume, high pumping power, and prohifiitive welght and size.

Attention was then turned tb a blending arrangement as a means-of
achieving the desired salf temperatures, It was found that blenders were
being considered by a fused salt power reactor group at ORNL, (Ref. T72).
After consultation it was decided that a blending system would be used
thereby allowing the system to consist of a separate hot and cold loop
{See Figure 3.1). The hot loop circulates the molien sali coolant from
the primary heat exchanger to the Superheétef, from the superheater
through the pump and blending apparatus, and then back to the primary
heat exchanger. On the discharge side of the hot loop pump, molten
salt at 1050°F is tapped off afid fed into the'éold loop, This flow
can Be regulated by means of a trim valve and serves aé the heatl source
for the cold loop. The cold loop‘contains the steam generator or' |
boiler and a pump. The hot salt ié fed into the cold1100p on the suction
side of the pumpvfrom'whence 1t trans#efses the stéam generator. From
the cold side of the boller the requisife amount of salt is tapped off
and fed into the suction side of the hot loop pump.. This completes
the path of molten salt through the circuit. The salt flow rétes in
the superheater and steam éenerator are to remain constant at 7.49 x

6

10° 1b/hr. The amount of salt bled from the hot loop to the cold loop at

full power 125 mw 1s 1.74 x 106

1b/hr. Schematic layouts of this system
are also shownlin Figures 7.3 and 10.1.
The fluid horsepower necessary to circulate the molten salt was

calculated to be 260 hp in the hot loop and 200 hp in the cold loop.

The calculated pressure drop in the superheater was 15.2 psi, in the

 

 
-91-

primary heat exchanger 20.5 psi, and in the steam generator 38.7 pai,
Since no finalized piping layout was attempted, the pressure drops due
0 line frictlon, bends, valves, entrances, etc. was estimated., It

is reasonafle to assume that these losses would not be as significant

as those in the primary heat exchanger, boiler, and superheater. There-
fore, the molten salt pumping horsepower should not vary greatly from
the above values,

In order to determine the optimum salt line size a short parameter
study was undertaken. Pipes with inner diameters from 7" to 17" fiere
investigated. The pressure drop per foot of pipe length was calculated
and from the resulting fluid horsepower a pump weight equivalent was
obtained. (See Figure 7-7). This was combined with the weight of the
salt per foot of pipe length and plotted against the various pipe
diameters. The results showed that the optimum pipe 1.4. would be

approximately 11".

7.3 Steam Generator

 

T.3.1 Types Considered.

 

In selecting a steam generator the genersl types considered
were (1) the flash boiler, (2) the once through boiler, (3) the natural
circulation, and (4) the forced circulation boiler. Each was given
serious conslderation and the conclusions drawn about each type follows:

(1) Flash Boiler: The only information found about flash boilers
was contained in Reports EPS-X-265, EPS-X-270, and EPS-X-288 by the MIT
Engineering Practice School at Oak Ridge. In these reports it vas

pointed out that the main advantages of flash boilers are that they
-92-

are capable of responding rapidly to 163& dgmands because of the small
amount of water contained therein and that it is probable that a high
capacity boiler of reasonable size could be constructed. Tn the past
the chief disadvantage of flash voilers has been tube burnout, but
this problem is absent in nuclear reactor applications where the coolant
is apt to be a moiten salt or metal. The chief reasons for not adapting
this type hoiler were foxr the most part pcinted.out in the above reports.
They were: (1) the need for high tube wall AT's in order to keep salt
from freezing on tubes, (2) the lack of nozzles that would glve an
adequate spray pattérn, (3) the need for very long tubes in order to
ensure dry steam at high loads, (4) the need for a method to insulate
the nozzle headérs from the heated tubes, and (5) the general feeling
of the group that, although flash boilers éhow'great promise, much more
developmental work is needed. |
(2) Once-Through Boiler: Once-through boilers have been used in
EuroPe for a number of_years and have recently come into their own_in
this country with the installatidn of the supercritical units at Philo
and Eddystone. They offer the advantage of boiling ang superheating
in a continuous passage thereby eliminating the need for heavy steam
drums. A once-through boiler also offers the advantages of rapid response
to load changes, compactness, and ease of arrangement. The principle
disadvantages are lack of water storage, the need for very high purity
water, and, especially for nuclear reactor applications, the thermal
stress problem. As was stated in Section 7.2 the stresses encountered
when boiling water at 572°F with a salt at 1100°F introduces intolerable

conditions. Therefore, it was necessary td separate the boiler and super-

 
93~

heater thus eliminating the chief advafitage of the once-thrgugh boiler,

(3) Natural Circulation Boiler: The natural circulation boiler
ils perhaps the conventional steam generator'for marine use. Unfortunately,
here the problem of arrangement is encountered, It ig necessary to have
a large drum at high elevation and a downcomer collector drum or drums,
Also the problem of whether to put the molten salt in tubes or let it
be on the shell side must be considered. Reports such as KAPL-1450,
"Review of SIR Project Model Steam Generator Integrity”, seem to indi-
cate that the best results for a liquid metal coolant such as sodium
would be obtained by placing the coolant in the tubes with the water
on the shell side. However, these experiments were all done uging
stainless steel, In the steam generator proposed in this report Inconel
ié to be used and with the obvious welght saving obtainable by filacing
the high pressure steam-water mixture in the tubes it was concluded
that the water-tube system was the more advaniageous,

In order to find some compact method of arranging a natural-circulation
water-tube boiler with a molten salt as a heat supplying wedium, a number
of different configurations were considered. The most promising appeared
t0 be a Lewis boiler which employs Fileld tubes. A Field tube is really
8 tube-within-a-tube. The inner .tube acts as an esgentially unheated
downcomer. The bottom of the inner tube discharges into the sealed off
end of the outer tube, The outer tube is heated and acts as thé riser.
High recirculation rastios are'obtained with this type boiler. Also, since
one end of the tube is free, there are little thermal expansion problenms,
The Lewls type boiler bhas seversl other advantages but its chief dis-

advantage would be the arrangement of a header sheet since it does have
 

~Qh_

the complication of a tube within-a-tube, The tubes must also be of the
order of 12' to 15' and 1t was felt that poséibly some other arrangement
would offer greater compactness.

(4) Forced Circulation.Boiler: The forced circulation boiler was selected
for the basic study because of its‘compactnéss and flexibility of arrangement,
Use could be made of a nearly conventional steam drum, and ‘the tubes could
be bent into a "U" shape to reduce the thermal expansion problem. The
steam output could be controlled by the circulating vater pump., The forced
circulation boiler is simple in design and principle and is well proven
in marine applications. |

7T.3.2 Design of the Selected Steam Géfierator

The collection of appropriate and adequate data for the steam
generating sys%em proved to be a task df no small proportions. The molten
salt was assumed to behave as a normal Newtonian fluid, Data is availsble
from experiments performed at ORNL giving the heat transfer characteristics
for heat exchangers, particularly delfia—array; This data was used for all
salt side heat transfer coefficients (See Figure 7.6). The molten salt
flow thréugh the steam generator is in the laminar region with Reynolds
Numbers of 200 to 300. This is due to this particular salt's high viscosity
in the temperature range to be used., The data for bolling water heat
transfer characteristics was hardly as easy %o get. Wide variances are
to be found in the literature for water boiling in tubes under pressure,
After consulting with a group of industrial boile: designers it was decided
to use a value of 6000 Btu/hr;ftg—oF for the heat transfer area calculations,
A value of 2000 Btu/hr-f+°-OF vas assumed for scale deposits collectirg on

the water side of the tubes. In the report, "Studies in Boiling Heat Transfer"

 

 
«95 -

(ucra Refiofit-Nb, C00-24), the conclusion was drawn that for water boiling
in tubes in the'pressure-range from 1000 psi to 2500 psi, the difference
between the tube fiall temperature and the water saturation temperature is
independent of the heat flux. Accordihg to this data the value of b, - tsat
that might be expected at 1250 psi was about 14.5°F, (See Appendix 7.1,
Section 5A). ' In McAdams (Ref, 17, page 393) the equation, by = %
10959/A%l/h, indicates clearly that the boiling film temperature drop is
P/90

heat flux and pressure dependent. These equations were used to determine

sat =

the temperature drop across the tube walls at points of maximum heat flux.
The McAdams equatlon gave the lowest film drop with a value of 9.6°F,

This in turn gave the waximum wall At of 85,20F at the maximum heat flux
of 172,000 Btu/hr-rt2,

It was decided to bring water into the steam generator at 5650F, seven
degrees below the saturation temperature. This water would be a mixture of
the recirculating water which is at the saturation temperature of 572°F
and the fesdwater which is at 486°F. 'The water'éntrance velocity into |
the tubes 1s 8 ft/sec., At the tube exit the dryness fraction is 0.11 which
corresponds to a SBV of 65 percent., This is approximately the maximum
steam by volume for this temperature and pressure that will still give
good wetting of the tube walls (Ref. 21).

A brief parameter study was undertaken to determine the most sultable
tube size and tube pitch. It was concluded from this study that in balancing
heat transferred against pumping power required, it should be possible to
g0 to smaller tube size than usually used in oil fired boilers, Upon

recommendation from ORNI pergonnel experienced wlth steam generators it

vas decided that g 1/2"'i,do tube was the smallest suitable tube. The tube

 
wall thickness compatible with the operating temperatures and pressures was
calculated to be 1/16", and from salt pressure drop congiderations the
closest tube pitech vas computed to be 3/4".

Once the tube size was set (the recirculation ratio and steam flow
rate are known), thé number of tubes necessérylto.carry the full power flow
rate could be determined. The steam flow rate for the 125 megawatt design
as calculated from a heat balance is 456,000 1b/hr. The number of tubes
for ‘the steam generator is then 2336,

The heat transfer area,fias’Qalcuiatéd_ih”two parts. The area necessary
to raise the water temperature to the sa£firation point was calculated. The
overall heat transfer coefficient in.the-watér heating region is 425 Btu/
hr-ft2~°F and the log mean temperature differencé 15.195°F; The heat trans-
fer area for this region is 3050 fte. This_tdtal'area of 3854 £12 made
necessary a tube length of 10.1 £+, | |

It was decided that the tubes should bé bent into "U" shape for reduction
of the thermal expansion problem, 'Calculations also showed that it would be
bestlto split the 2336 tubes into 8 bfindles. This would keep the salt
chketad vessels to a reasonable size and vall thickfiess,.and would reduce
The header thickness and weight, |

Of all the steam generatbr parts it was thought'that the headers would
present the greatest problem: It was concluded ithat there was no reason
why the tubes could not be run directly info the steam drum, The salt
Jacket could be attached directly to the drum or by'expansion Jjoints. The
tubes wouifi*%e "U" shaped and\"hfing" from the drum aé illustrated in Figure

7.1. A water header would be at the other end of the tube bundle. This

header could be of éither flat head or dished head design. The hot (800°F)

 

 
-97-

salt would be introduced into the jJacket just under the drum and heat baffle
would shield the drum from contact with the high temperature salt. The
molten salt would leave the Jjacket just below the water inlet header,

Four salt jackets containing tube bundles are attached to the bottom
of each of two drums. The tubes serve the purpose of risers., They dis-
charge their steam-water mixture into the drum where the steam is separated
by mechanical separators and scrubbers. The water leaves the drum through
the downcomers which are 1ocated‘on the bottom side of the drum along with
the salt Jackets., The water in the downcomers is at the saturation tem-
perature of 572°F. This water is blended with the 486°F feedwater and the
resulting water temperature is 5650F. (The séturation pressure at this
temperature is 1180 psia or approximately TO psi below the steam generator
operating pressure), This water is forced back to the water inlet headers
by the circulation pimp.

The design capacity of the steam generator is 456,000 1b/hr. The water
flow rate is 4,149,500 1b/hr and the salt temperature drop as it traverses
the steanm generator is 76°F. At the 95,9 umw power load the full-power steam
demand is 355,030 1b/hr and the salt temperature drop is 58,80F. At this

o
power the inlet the outlet tewperatures will be lowered to 761.80F.and 703 F.

7.4 The Superheater

 

1t was felt that the superheater design would be the most straight
Torward of the steam generating system. The superheater is toc %ake the
saturated steam at 572OF and heat it to 9500F. The 125 mw capacity of the
superheater was to be 348,000 1b/hr but the capaclty necessary for 35,000

shp is 263,300 1b/hr.

 
-98=-

Again, investigation showed that tubes of the smallest practical
diameter would give the best heat transfer characteristics, Tubes of 0.5"
0.D. and 0.4" I.D, were selected. A steam exit velocity of 100 £t/sec
was chosen as the maximum practical velocity, To carry the flow at this
velocity 722 tubes were necessary. It was decided to space these tubes
at a pitch of 3/4". This gave a molten salt velocity of 11.55 ft/sec
and a pressure drop of 1.33 psi/ft. The salt inlet temperature is 1150°F
at 125 mev and 1138.3°F at 95.9 mev, The exit temperatures are 1126°F
and 1120,4°F reépectively. The overéll heat transfer coefficient is 29
Btu/hr-£t°-F and the heat transfer area is 1070 P2, .This'gives-a tube
length of 11.4% ft, The maximum heat: flux was calculated to be 231,000
Btu/hr-ft2 and the maximum tube wall At was 80°F.

The superheater vessel ig "Uy" shaped with the headers at both ends,
(See Figure 7 2) The tube bundle runs through the #essel with the tubes
arranged in a delta—array. As is the case in the steam generator the
headers were considered %o presefit the greatest actual problem. Numerous
header arrangements can be deviged but the best seem %o be either a dished

or flat head,

7.5 Auxiliary Equipment and Arrangement
T7.5.1 The Steafi Drum and Desuperheater
The steam drums are an integral part of the steam generator
and contain the mechanical steam-water separators, the steam scrubbers, and
the desuperheater tubes. It was decided to.use the two conventional drums
of the class 931 destroyer boiler room with the attachment of the molten

salt jackets to their undersides and the replacement of the conventional

 

 
-99=

risers with 5/8" o0.d. tubes as described in Section 7.3.1. The drum méterial
is to be Inéonel but-the separators and scrubbers are to be of conventional
design. The drum diameter is 52.2" and the bottom shell thickness is 4,8",

The desuperheater will consist of tubes running through the saturated
vater in the drum. It is necessary to desuperheat 5340 1b/hr of steam when
running at 35,000 shp., Superheated steam at 950°F and 1235 psia will enter
the drum in the tubes and be cooled to 650°F,

The arrangement of the steam generating equipment around the reactor
and within the secondary shield would be as shown in Figure 7.3. It is
realized that the actual design of the steam and salt piping within the
éecondary shield requires careful analysis, which 1s particularly necessary
to keep stresses due to relative thexrmal expansion within reason. However,
neither Time nor talent permitted such an analysis for this study, and
therefore only a reasonable estimate would be made for the volume required
for this pluwbing.

As shown in Figure 7.3, there would be two identical galt and steam
systems. It would be desirable to have all pump driver accessible from
outside the secondary shield, It was therefore proposed that the secondary salt
pumps be mounted ofi the top and drive through the secondary shield. Similarly,
the water recirculation pumps could drive through the aft face of the secondary
shield,

T.5.2 Feedwater Heater and Other Components

Although most of the equipment following the superheater in the
steam generating system will remain unchanged, several components will no
longgr be necessary when the two furnaces are replaced by & reactor and at

least one new item must be added to the system.
=101~

-

to keep pump turbine weights down., If these pump turbines are assumed to
be expanding the steam to the same extent as the main feed pfifip turbines,
which also operate on superheated steam, the enthalpy contained in their

exhaust would be more than sufficient to make up that lost by the exclusion
| of the forced araft blovers and fuel ol pumps ,

In'ordérhté'maintain the deaerator saturation pressure at 18 psig at
full load, a somewhat greater quantity of auxiliary turbine exhaust_must
be bled back to the main condensers, This in turn wiil probably require
the addition of several condenser tubes to maintain the former condenser
.vacuum. Calculations show that at full power 12, 530 pounds per hour of
auxiliary turbine exhaust must be bled to the condenser, if a deaerator
pressure of 18 psig is to be maintained.

7.5.3 Feelwater Treatment

Just how much feedwater treatment would be necessary to
ansure long;term firouble-free service from the steam generators was one
of the many problems that the group did not have time to investigate.
The conventional destroyer supplies distilled wvater to the oil fired systeuw
and,no attempt was made to answer the question of whether or not further
treatment by ilon-exchangers would be necessary for the ;eactor heated
steam generators. However, a heat transfer coefficient of 2000 Btu/hr~ft2—°F
was included for scale. deposit;, which should be conservative. It should
be pointed out that the replacement of the oil-fired furnace by_é ffised salt
system makes the steam generation equipment relatively clean, and therefore
the use of ilon exchangers to further reduce the water impurities way be

Justified. Such a system should be a large lmprovement in cleanliness and

require much less maintenance than the conventional oil-fired boiler,
-102-

7.6 Part Load Operation

 

The method of achieving a certain steam rate for loads which are some
fraction of full power will depend upon the method by which the reactor is
held at part load. Presumably,,thé'flow rate of the molten salt coolant
and the sverage tempéiature of the reactor will remain the same, but the
temperature rise of the molten salt coolant as it passes through the primary
heat exchangers will vary according %o load.. Thus in the hot molten
salt loop, which inecludes the superhéater, pfimary heat exchanger, one
of the pumps and part of the-blending'abparatus;'the average " temperature
and salt flow rate will remain cOnsfant and fhe inlet and outlet temperatures
of the superheater will change apprppriatei& as will the amount of molten
salt that.is bled off as the heat source for the cold or'steam generator
loop. The'average'temperature of fihe steam.generat0r can be either railsed,
lowered, or held the same according to the émoufit'of salt bled from the
hot loop. It was fbund that in lowering the fiower frofi.las mw to 95.9 my
the aversge temfiératureuof the éfeam génerator could be iowered 380 thus

alleviating the thermal stress problem somewhat,

 

 
 

 

‘S 000°€9 - 1A 13M ‘183
"SE7 000°'CS - 1M ANG ‘1S3
L2681 -V3IYV HIISNVHL LV3IH

 

 

UNCLASSIFIED

ORNL=LR=-Dwg, =257L45

 

 

L1 -"a71 38nt
.8/% -'0'0 380l
89i1 - §3an1 "ON

 

 

HOLVHENIO
WV3ILS Q3S0d0oyd

 

"SHIAWOOINMOU 9 ' 3¥N91d

 

 

-103-

HOLVHINIO x«upmlxl\\

 

 

 

 

 

 

 

 

 

 

 

 

 

YOLVHINID ¥3d .wl\\fi

4d374N0 17VS

 

Qo |0 , O

 

 

 

 

 

 

NI H31VA
NOILVINONMIDI Y

 

   

  

-9 |

  

3744v8
1v3H

        

HOLVHINGD ¥3d ¢
L37INI LTIVS

~
1

o
=

 

 

™

WNY¥Y Wv3als

 

 

 

 

 

 

 

 

 

ns,2l

 

 
'S8 00€°1] = LM L13M 1S3

i u

‘'S87 0006 = LM AY¥Q 1S3

z1ld GES V3IYV WIJISNVHL LIV3IH
,004°0 =0’ 38nl
,00§°0 ='00 34anl

 

I9€ — $38N1 40 ON

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

P .
3 b6 |
—
_ § 1 u8l s3sni
LITINI 1IVS ¢
. I
S3IENL 13TLN0 LIVS | |
D [ @& ol Q
|
mmmazqmpmllflhw Hu
0 i
ln/h..._l. .
S8 fl
iz L AN og
N
A0
a5
Y
O

 

 

 

 

 

'L 3J¥nold H31V3IHY3IdNS J3S0d0¥d

 
 

 

 

A
<
o

f

 

GIFIHS ANVONOOIS 40 FA0YId MHINNI A

GNY 8% ARYYd ONILSIX3 l}’

WY3FLE QILVIHNIANS waf— T T T WEZivIAmaIdns oL T

WY3ILlS aaivIHE3disag ¥

       

 

 

 

 

 

 

 

 

 

   

 

[
e
[
8
-lm
$ SHILYIH U3LVYMAIAd OL —e—
: .
23
15 NI H3LVM Q334 —
S i
dWNd . 1
az_:.__._ofiomzl// ' _
Illllt.ll]cll‘ll!..!."..
+ am4
4 MOT38 NMOHS NILSAS HWVILS
S - - 3 dIHS
JAOGY NAOHS WILBAS LIVS o
+ (]

HOLOVIY
a3 1HS X
AN

AHYWiHd \ i

3Ivos 0%
MIIA NVd

         

 

JYIHNTNE AONY LS

. INYHAd SNILSIX3
13IHS ANVANOD3S NIHLIM .

- JILYWIHOS ONIdid

ANV LNIWIONVHNY
oI1SVE Q3IS0dOo¥dd

 

 

 

HOLVHANTAY

nVv3Ils YILVYIHHILANS

€1 34n9id

 

 

  

 

 

 
-106- OENL=LR~Dwg ,=25748
- UNCLASS IFIED
FIGURE 7-4
ESTIMATED SECONDARY SALT VISCOSITY
| TEMPERATURE °F

700 T8O 800 880 900 950 1000 1100 1200 (300 1400 |800 i800
| R i i T 1T

 

500

oo\
N

 

 

 

N

i \
0

o 100

a |
-

Z

i

o

|

>=

}....

wn

o

O

»

>

20 N

 

\

 

 

 

 

 

 

 

 

350 400 500' 600 700 800 900
TEMPERATURE —°C

 

 
 

100

 

ORN L-LR-Dwg, =257L9

UNCLASSTF IED
-107-

FIGURE 7.5

FRICTION FACTOR CORRELATION

EF. 25 TUBESIDE

F. 25 SHELLSIOE

-

S

N\

1000 10,000

REYNOLDS NUMBER
-108-
ORNL~LB=Dwg,~25750

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

UNGLASS IFTED
FIGURE 7.6
100 [ T T T T TT7 1 |
- HEAT TRANSFER DATA CORRELATION
Nu
—Pr -4 :
/|
Nu _ 08 | 4
Pr.a 0.023 Nre ""\3’ %
REF. MS ADAMS L7l
| 1
///
Nu _ 0.91
Bra =0.006 Ngre \;/
10 Nu 0.36 A
"4 20.47 NRe - .
Pr 4 B/ 6/'
> *—REF J.L.. WANTLAND
%\\
.0 . _
100 1000 _ 10,000

REYNOLD'S NUMBER
-109-

 

 

 

FLUID HORSEPOWER - FHP

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FIGURE 7-7
EQUIVALENT FLUID HORSEPOWER,WEIGHT AND
REACTOR RADIUS CHANGE FOR CONSTANT SPECIFIC WGT.
300 l
ASSUMING :
MNth=.25 STEAM PLANT EFFICIENCY
- MGEN.=.95 /
250 }—————T MOTOR =.90
: N pumMpP =.80 b / /
OVERALL PUMP weT, = 19103/ 1o y
GENERATOR WGT = 101bs., o Y
STEAM TURBINE PUMP = 51bs., ,
COMBINATION FH / /
/
200 ' / 7
A*'/ /
VERALL QQ\/
0 |
SPECIFIC WGT, <> /
150 LBS./SHD
100
d
| / EQUIVALENT CORE RADIUS WEIGHT
'/ BASED ON BASIC REAGTOR DIA, OF
50 | e L 72 INCHES & PRELIMINARY PRIMARY
/fl SHIELD WEIGHTS ONLY,
v, |
7,
0
0 4000 8000 12000 16,000 20,000
WEIGHT-LBS,
0 ' | ' 2 ' 3 . 4
REACTOR RADIUS CHANGE ~ AIN

 

 

 
=110~

8.0 REACTOR ANALYSIS

8.1 Nuclear Configuration

Figure 3-2 schematically indicafies the physical picture of the core
and reflector which will be described in detail below. General overall
nuclesar conéepts of this high pérformance physically small system are
modifications and combinations of advanced design ideas of ANP Technology under
consideration at the Oak Ridge National Laboratory, Ref. 36, Physically,
the core is a circulating, fused fluoride, uranium bearing salt flowing
through a beryllium oxide moderating matrix and incbrporating an inelastic
scattering metal reflectsf. Systeus of thesé-types feature high power
density and relatively high operating temperatfires.

Numerous nuclear advantages are manifested by these systems, The
released energy is easily extracted from the core in that i£ is generated
in and transferred out by the same fluid, Fluid fuel systems are, in
general, self regulatifig under small perturbatiofis awvay from nominai
operating conditions due to prompt vblume expansion within the fuel, Thirdly, .
'fiugh of the energy released in the fission process other than the kinetic
energy of the fission fragments is retained and collected through cooling
bmoderator, reflector and internal neutron and gamma ray shielding with
the coolant-fuel, 'Other advantages are low Operating pressures and the
relative ease of extfacfing volatile fission products,

Disadvantages include, large fuel inventory required from excess fluid
for component cdolihg, heat exchangers, pumps, core inlet and exit plena,

A relatively serious hazard is present in this circulating fuel system,

specifically, the containment of a corrosive, multicurie fiuid at high

 
-11l-

temperatfires. Requirements in maintaining the fused salt liquidous during
shutdown also burden these system,

Limitations were necessarily placed upon the nuclear designlto€meet
the requirements_of metallurgy, heat transfer and nuclear design, and to
narrow the breath of the étudy, With the choice of a fused fluoride fuel,
no point in the system can be at a temperature less than its solidification
value (in the order of 980°F) and no point should exceed & temperature
~ a% which rapid corrosion takeg place., All fluid surfaces should be Inconel
clad to & thickness which will withstand 10,000 full power hours of
operation. Power densities will be high, but not enough to induce. -
dangerous thermal stresses in all materials. Resulting limitations

regtricted the design to the following specifications:

Mean Core Temperature 11500F to 12500F
.Primary Fluid Surfaces 30 milslcladding Incone
or greater :
Maximum Power Density 1200 vatts/em> fuel
Core Dimensions Riéht ¢ilrcular cylinder

70 ¢cm in diameter 80
cm in height

8.1.1 Moderator Matrix
The moderating matrix consists of rodg comprised of beryllium
oxide, three-quarters of an inch in dlameter, clad in 40 mils of Inconel,
Radially, the rods will be close packed in a triangula?ly pitched array.
The pltch is defined by the selected volume fraction of_moderator in core,
Beryllium oxide "meat" extends the entire:length'of the core, and joining

on each end of meat will be, one and a half inches of boron-10, BeO ceramic

 

 
~112«

material to suppress fission in the core inlet and exit plena, Ends of
the elements neck down fofming thé plena and joining the structural members.
See Figure 3,2.

8.1.2 Reflector

An inelastlc scattering reflector is utilized in the systenm.

This choice has not been proven su?erior.to an elastic moderating material
such as beryllium oxide, but it is believed to contribute definite advantages
over BeO in this specific application,

The choice of a nickel reflector is based_upon the fact that this
material possesses excellent slowing down characteristics in the higher
neutron energy range, which is fif considerable importance in this inter-
mediate reactor. Secondly, relatively émall amounts of cooling will be
required for the reflector and therefore it will retain its desirable
nuclear properties to a large degree. Thié high atomic number material
will attenuate core gamma rays very strongly and thus reduce the required
gamma ray shieldifig. Also, fast nefitron leakage out of fhe reflector is
within acceptable limits and is only of minor concern in fast neutron shielding,

Time did not permit the detail investigation and comparison of systems
incorporating elastic and inelastic reflectors; intuitive reasoning lead us
to the nilckel reflector, Thé resulting reflector.is comprigsed of a 6 inch thick
cylindrical shell 29,6 inches in inside diameter surrounding the core, Cooling
annuli penetrate.the.nickel vertically through the reflector and coolant is
supplied from the 16wer plenum,. Estimated coolant required will occupy 2
rercent of the reflector’s volume, |

8.1.3 Fuel

Numerous types of fluoride salts are avallable,but in a large

 
-113-

ma jority, the existing data on their properties (corrosion, thermal and
mechanical) ave limited. Therefore, it Wés necessary to make the basic
criterion in the selection of the fuel depend upon the technology presently
avallable. " The resulting selection contained ZrF, Nak and UFM° (See
Section 4.1). Unfortunstely the nuclei constitubting this fuel lack some
of the more desirable nuclear properties. Namely, it contains nuclei of
high atomic number and thus has poor slowing down properties. Both zirconium
and sodium have significant absorption cross sections in the intermediate
energy range, although theilr thermal absorption cross sections are relatively
small. Due to the large volume fraction of fuel in the core, any added
neutron moderating material in the fuel will in general, reduce critical
mass and the average energy of the neutron number density in the core,
Although these changes will not be 1arge, they will be significant.

Other possible cations which could replace zirconium or sodium are
beryllium and lithium, Beryllium fluoride lacks corrosion compatibility
with Inconel although it would contribute appreciably to the neutron moderation
of the core., Lithium fiuworide also attacks Inconel and only isobopic lithium-7
could be considered due to high epithermal and thermal absorption cross

section of elemental 1ithium,

8.2 Parametric Study

An investigation into the nuclesr characteristics of the described core
containing various ratios of beryllium oxide to fuel were deemed necessary
in the selection of a feasiblé design. The principle objective of the study
was to determine the critical U-235 concentration in the fused salt? and to

minimize this value through varying moderator to fuel ratio. Secondly, the
 

~11h4-

total fuel inventory was to be minimized through the selection of a critical
fuel concentration and ffiel volume in the core, The range of investigation
was limited between 0.4 to 0.6 volume percent fuel by the power density in
the fuel at the lower limit and lack of sufficient neutron moderation in
the core at the upper 1imit; o

Group-diffusion methods were the means of analysis; specifically, a
3 group 3 region, one dimensional ORACLE diffusion code Ref. 60, Region
allocation were to: (1) control rod thimble, 5 cm in radius, (2) core,
cylindrical shell 32.5 cm thick and'ofitside radius at 37.5 cm and (3) the
nickel reflector 15.2h cm thick., Thé cofistituents of the regions are as
follows, measured in volume percent. Region 1; Ificonel 19%; Void 81%;
Region 2: Inconel, BeO, and Fluoride Salt - variables Region 3; Nickel - 100%,

8.2.1 Cross Sections | |

For the parametric study, the mean cére temperature was taken

as 1200°F. This condition results in a mean'néutron energy of 0.0795 cu at thermal
equilibrium with the core materials, assuming_no thermal spectrum hardening,
Energy boundaries for the three groups were'éelected on the following basis;
(1) Some existing data available for these_choéen boundaries, and (2) these
boundaries were suggested by the spectral distribution of fission from multi-
. group analyses of sifiilar reactors, Ref. 59.

Table 8.2,1

 

Group Energy Range | Lethargy Range
1 10 Mev - 0.183 Mev 0 -4
2 | 0.183 Mev - 1.4k ev b - 15.75
3 | 1.y ev -0 - 15.75 - o

 
-115=

Cross sectlons for energy degradation from one group to the next lower
group were defined by:

i

1 |
Op = Ogr, / AUL where

4 .
= i i
Oar jgaé *  Clie

These terus are defined as:

0} = average, transfer chss section from ith
X

4 group to the
(1 + 1)*" group cm

gi = average elastic scattering cross section for group 1. e
i

g = 8average inelastic scatiering crogs section for the ith
ie  growp, cn?

= averége slowing down cross section for the ith group. cn®
_§° = mean log energy loss pé% eiéstic scattering event, .

An spproximation in this method of Incorporating inelastic events in
the slowing down cross section is the assumption that each inelastic event
removes a neutron bne 1ethargy unit or the mean lethargy gain yper inelastic
event is unity. | |

Transport cross sections were evalusted in all groups as,

O_rmi = O'e:L (1- K )+ 0‘}_& + O‘ai
. where M =g%§r_
with the exception of Be0 in the thermal group where experimental datum
was incorporated. Ref., 63. Chemical binding effects upon neutron scatiering

were neglected and the assumption of free atom scattering was made throughout

with the above noted exception.
-116-

Fast and intermediate group absorption cross sections were taken from
various references. References 8 and 61. A large majority of these values
result from analytical apptoximatiéns to the energy dependent cross section,
Several are experimentally indicated values. Thermal absorptions cross

sections were taken from Ref, Bh,corr¢Cted for temperature and averaged over

 

 

the assumed Maxwellian spectrum, - - Bc/KT
| | 512-
F_a(kT) o (KD)l £/2 X ax
o a (2200 M/sec) "~ Ec¢/KT -XI
‘I Xe © ax

Ec is the upper thermal.group boundary
f(KT) is the non 1/V c0rfed£ion.
for Bc/KT = 18.1 |
‘Qzfigg%o W5y 0.50 £(KT)
Group one and two fiséion Cross secfiions were averaged over each
group from the values tabulated ifl Ref: 6l1. Group three fission gross

section was taken as

o £(XT) %15152) from Reference 3k,

L+ct)

A tabulation of all microscopic cross sections can be found in
Appendix 8.1 along with their references.

8.2.2 Summary of Results

 

Based upon the philosophy set forth as to the general concept of
the over-all design study, the following criteria were utilized in the

selection of a core design from the results of the parametric study, Of
-117-

ma jor importance is the power density within the fuel which must be main-
tained below 1200 watts per cmd of fuel to insure the reliability and
‘integrity of the beryllium oxide moderator rods. Thermal stress induced
in these rods through gamma ray and neutron energy deposition must be
maintained within safe levels., Also, as stated previously, it is desired
. to minimize the fuel concentration in the fused salt and to maximize the
utilization of the uranium investment. In addltion, the reactor should
be kept operabie with a maximum of thermal fissions, reducing both fuel
investment and control problems.

It was concluded on the bases of these data presented in Figures 8.1
and 8.2 and Table 8.22, case 2 (50.9 percent fluid volume in core) justify
the above criteria most satisfactorily. Case 3, with the decreased fuel
fraction, is eliminated automatically by its high“power density and wranium
¢oncentration in the fuel. Although case 1 (61 fiercent volume) indicates
larger safety margins with respect to power density along with an impercepti-
ble difference in fuel concentration compared with case 2, this system
-exhibits 20 percent less thermal fissions.

The twenty percent increase in thermai fissions with case 2, indicates

. lower average energy of the neutron number dgnsity in its energy distribution,
I% is believed this system will exhibit more control with an absorbing con-
trol rod than the faster systenm,

8./2.3 Control Rod Study

Reallzing the system under investigation would operate pre-
dominately in the intermediate energy range, control of the system must
be achieved through degradation in energy of fast and interwediate neutron

to thermal energies and result as a loss to the system through absorption.
 

-118. ORMLmLE=Dug ,=25752
UNCLASS IFIED
FIGURE 8-I

REACTIVITY vs MASS U?23®
MEAN TEMP - I200°F

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4
>
” /
~T / a
0 > _ 7
Yk
/
>~ / | / /
-2 F—— CASE | 1
z h 7
= /
:I) CASE 2/
u -4 / |,
= / <
— y / CASE 3
= / /
& -6 ‘ /
0 / /
a /
. //
/
-10 / /
40 50 60 70 80 90

MASS U®**® IN CORE (KGM)

 

 
ORNL~-LR=Dwg, =25753
UNCLASSIFIED

-119-

0g’

7304 WO/ . .N SWYYS
St ot s¢

og

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

NOILVYLINIONOD _,,N SA ALIAILOVIY

 

 

¢-8 JHN9IA

ALIAILOVIY LIN3OH3d

 
-120-

TABLE 8,2.2

RESULTS OF PARAMETRIC STUDY

¥See Appendix 8.4

 

 

CASE NUMBER
1 2 3
Volume Fractions) Salt 0.6108 0.5090 0, 4072
) BeO 0.3178 0.,4009 0.4840
In Core ) Inconel 0,0714 0.,0901. 0.1088
Mean Core Temperature 1200°F 12000F 1200°F
Uranium-235 Mass (KgM) 90 72.5 70.0
Multiplication Constant (K) 1.00187 0.99317 1.00666
U-235 Concentrgtion in
Fuel (gms/CM° Fuel) 0.k2khs 0.41030 0.49511
Core Fluid Volume (cm3) | 2,1204 x 105 L.767 x 105 1,513 x 107
Percent Fissions - Fasf _ 10.82 8.32 T.33
Percent Fisgions Intermediate 66.94 63.27 61.38
Percent Fissions Thermal 22,24 28.41 31.29
Average Powgr Density
(Watts/cm” Fuel) 589.5 7071 884.3
Peak Power gepsity
(Watts/Cm” Fuel) 831.2 990.4 1255,7
Prompt Temperatureo ' -5 -5
Coefficient §K/°F  -2.63 x 10 2,19 x 10 -1.75 x 1077
~ Prompt Neutron* -6 i ¢
Lifetime (Sec) 1.35 x 10 1.73 x 10 1.87 x 10" ’

 
121~

This requires the control element to contain both moderating and absorbing
materials, One centrally located rod was investigated and 5.4 percent
reactivity control (see Table 8,.2.3) was obtained with the element con-
taining the following materials: 19 percenfi by volume Inconel, Eh vercent
BeQ, L0 percent nickel with 1 percent by wt Blo and 17 percent void for
rod thimble clearance.

It has been concluded that 5.4 percent control is adequate as a
minimum value., (Ieave as a minimum, 3.7 percent shutdown marginc) There-
fore, only one control element would be required., However, the above
design was not considered adequate in that heat transfer across the
clearance gap was insufficient,

An alternate consideration immersed the rod in a bath of sodium,
filling the clearance gap, Also, the rod thimble would at all times be
filled with sodium and upon insertion, displaced sodium would be forced
into a small reservolr located outside the pressure vessel. 6.1 percent
control was achieved with this system and also 1 percent reactivity was
added to the system with no rod penetration. See Table 8.2.3 below.

TABLE 8.2.3
Void Filled Control Rod Thimble k = 0.98853
Sodium Filled Control Rod Thimble  k = 0.99585

Rod at Maximum Penetration with
Sodium in Gap k = 0,93786

The specified boron content was only a means of analysis, Due to
the damaging metallurgical instabilities resulting from high irradiation
exposures of boron in metal, it is recommended that the B-10 equivalence

of a Europium Oxide dispersion in nickel be used as control rod material,

 
-122-

This oxide also exhibits more reliable control during operating life in that
very large exposures arebrequired to obtain substantial burnup, hence only
& small loss in control will be obgerved upon long exposures. The following
date (Ref. 68) indicates thig phenomena ;
14000
e hr
fi

152 13Y .
H%ggfigb-E“ Bul53 fi20bb'EU15E,,£§¥%’)V

 

 

155
1E00b"‘ fu
my'%5_IPE 5156 1sab
13,0000 &
Natural Isotope Event Cross Section Per - Half Life
Abundance ‘ | Event (at 2200 M/s)
W77 R (n,y )Euto? 7200 b Stable
152 - |
(n #)Eu 1400 b Stable
Eul?? (h, y) 5000 b 13Y (p)
Eulo? 8 9.3 hr
52.,23% . pylS3 (n, ¥) 420 b Stable
Byt (n #) 1400 b 16Y ()
N 2 13,000 b 1.7 (g)
Eyl56 B | 15 4
18.8% p'0 (n, «) 4020 . Stable

In a high neutron field it takes 3.3 neutrons to be absorbed, on the
average, in an Europium atom before it is lost to the system; only one is

required in B-10.

 

 
-123-

8.3 Nuclear Design

Tabulated below are the resultant design conditions of the core,

Table 8.3.1
Power 125 MW
Core Volume 3.471 x lO5 cm3
Filler Volume Fraction o 0.5090
Be0 Volume Fraction - 0.4009
Inconel Volume Fraciion 0.0901
Mean Fuel Tempersture 1225°F
Hot Clean K 1.0275
Critical Mass | T1.75 Kgu U235 in Core
Excess Mass for 0. 4 2.75 percent 14,00 Kgm @35 in Core
Startup U-235 Capcentration 0,48528 gms U-235/cm3 fuel
Startup U-235 Inventory 605 Kgm U-235
Percent Fast Fissions - | 8.29
Percent Intermediate Fissions 63,87
Percent Thermal Fissions | 27.84
Prompt Temperature Coefficient -2.19 x 10  X/°F
Prompt Neutron Lifetime 1.92 x 10'6 Sec

Average Flux Over Core
Fast 1.33 x 10%0 neutrons/sec cmd
Intermediate 8.14 x 10lh " "

Thermal 1.97 x 1083w "

 
 

 

 

 

-12k-

Average Power Density 708 Watts/cm3 Fuel
Maximum Power Densify | 1040 Watts/cmS Fuel
Total Control Rod Reactivity Worth 6.18 percent SK/K
(_SM/M)

("8K/X) Core 7.1

Endurance | | 4000 Full Power Hours

0.5 percent Burnup
)
K/K) System 50

8.3.1 Criticality
Critical mass, and uranium concéfitration in fuel were obtained
through a series of problems performéd ofi the ORACLE simular to thoge
described in the fiarametric gtudy, Due to heat transfer considerations,
the mean core temperature was increased to 1225°F, Results iqdicatéd
a critical mass of T1.75 kegm U-235 under hdt, clean and unshielded con- -
ditions, graphical results are presented in Figure 8.3,
Radial flux and power spatial distributiofis are presented in Figures
8.4 and 8.5 respectively. Resulting flux distribution indicates the
reflector savings in the thermal and intermediate groups through the
gradient of the distribution near the reflector boundary,
8.3.2 Self Shielding
Disadvéntage factorg were obtained by diffusion theory methods
for the unit cell as defined in Figure 8,6. Both intermediste ang thermal
group factors were considered, although the intermediate factors were
insignificant, Thege effects were incorporated by defining effective crosg
sections and expressing the effect as a&n excess reactivity to the unshielded

Criticality calculations, A simple perturbation method was used to obtain

 

 
ORNL~LR-Dwg. =2575)
UNCLASSIFIED

-125-

 

 

 

 

 

 

 

 

 

(wbx) IYOO NI __.N-SSVW
002~
_ 00°1-
\ T
m
2
\ . o
m
000 5
2
m
D
\ o
ool
=
-
-
= 002
106070 - TANOONI
600%'0 - 028
060S°0 - 17VS
L ¢WND Ol x SILP'E-TNNTOA 3HOD 00¢
4. 6221 - dW3L NV3IW SNOILOVYd 3JANTOA 3HOD
SSVIN 21 SA ALIAILOV3Y
|

 

 

 

 

 

¢£-8 JYNOId

 

 
RADIAL FLUX DISTRIBUTION
RELATIVE TO AVERAGE CORE VALUE

 

 

-126-

ORNE~LR=Dwg, =25755
UNCLASSIFIED

 

 

N
| & \:»THERMAL AVERAGE CORE VALUES
 FAST ----1.33%10'® mrv
) INTERMEDIATE INTERMEDIATE -~ 8.14 X [0} v
1.4__...-’ ‘5_'_\ \\
1.2 \*\\
' N
\\
N\
A\
1.0 ‘\
2
o
w \
0.8 & |
QO \ REFLECTOR]
& W
N
063 . X
i \\
- \
= \\
S \
0.4 ° \ \\\
| CORE ___\ \\
\ 1\
0.2) ¥
\\\\
: _ \
O 5 10 5 20 25 30 35 40 45 50 55 chA
CORE RADIUS

FIGURE 8-4

 

RADIAL FLUX DISTRIBUTION

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
-127- ORN L~LR~ Dug, - 25756

 

 

 

 

 

 

 

 

 

///////////// /// ::
I

 

 

 

W

 

 

 

 

 

 

<
3 / S
= ///////////////
O
@™ 5 /
@ - )
= N
ln .
Q0 o / O
Wl o © : ol
o u_]m M W
hao | 3 =
o W | -
> _ o
E 80 <
[ / gfl:
ui |
O w
2 <« g
-
< W oo
— 1 > L o
w < @
@ o
O

 

 

15

 

 

 

 

 

 

 

 

 

 

 

 

/ /
// 7 /f/////f//

O

 

ALISNHG HBMOd

 
-128- ORNL~IR=Dwg, »25757
UNCLASSIFIED

 

 

CELL FLUX RELATIVE TO AVERAGE

FIGURE 8-6

THERMAL FLUX DISTRIBUTION IN UNIT

LATTICE CELL

REGION AVERAGES

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

BeO - 1.100
INCONEL~0.947
SALT - 0.833
1.4
.2 ]
.\\
.0 \-\
\\
\
0.8 ~
i)
: r
X
0.6 a =
(1p]
= J
> W w
0.4 _J < 0
- O o
> O =
(4 <
Ly T °
0.2 @ 0
0
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

UNIT CELL RADIUS-CM

 

 

 
-129-

these_effects. The development of this perturbation technique is presented
in Appendix 8,2..

8.3.3 Burnup and Fission Product Poisons

One advantageous feature of the large fuel inventory required

'for.this system is the extended endurance of fuel life. The large fuel
volume is circulated continuously thropgh the active core and burnup is
achieved homogeneously throughout the fuel, thus extending life by
approximately a factor of six over a stagnant fluid systeu.

Endurance in the order of 4000 full power hours is expected as the life
of the initial core loading. This represents felatively swall burnup (0.5%)
periodic additions of fuel are possible and would greatly enhance the l1ife-
time. It is believed the system will operate for the 10,000 full pover
hours of reactor life, with only minor additions of fuel %o the initial
loading.

In this analysis, non-volatile fission products were approximated as
equivalent to lOObbarns of added absorption in the thermal group and 10
barns added absofption in the intermediate group per fisslon event, It
is believed the above approximation results in an over approximation of
the non-volatile fission product poiéons°

Burnup and fission product poison effects upon reactivity were treated
Jointly as reviewed in Appendix 8;3° Results are presented in Figure 8.7.

8.3.4 Prompt Temperature Coefficient

 

Value of the prompt temperature coefficient as quoted
(-2.19 x 107 §K/°F) contains the effect of the volume fuel expansion and
the shift in the agsumed thermal Maxwellian spectral distribution with

temperature, Effects of doppler broadening in resonance absorptions were

 
 

| dNNYNE LN3IO¥3d
G2 02 Sl o'l c'0 0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

R
<8 _ _ I _ 0
Ky 00002 000°s] oon_..o_ 0006
0
mm S¥NOH ¥3MOod 11Nnd / o'l
i =
= \
5 \ oz
Ve -
m
yd 2
o 0O€ m
<
-f
mw p. _ Ob ®
e m
\ 3
/] 06 <
\\\ 1
\\\\ | -
\\ "
\\\\\ _ . 0L
—lo's
SNOSIOd LO2Naodd NOISSId ANV dNNdNg

 

oL 3Ina 3IWI1L3dITT OSNINYNA SSO7 ALIAILOV3Y LN3IOY¥3dd
L-8 34¥N9ld

 

 

 

 

 
-131-

neglected but are expected to be negative also, Ref, 71,
Method of calculating this coefficient.involved the investigation of
the multiplication constant of identical systems at two temperatures using

the 3G3R ORACLE code. Table 8.3.4 presents results.

 

TABLE 8,3.4
Loading Temp Reactivity
Kgm Op
72.5 . 1200°F 0.2550
72,5 1225°F 0.2003

8.3.5 Xenon Poison

As discussed previously (Sec. 4 ), the removal of volatile
flssion products can be achieved with relatively high efficiency in a high
temperature, liquid fuel system. Provisions for periodic removal of the
volatile matter are incorporated in the system desigg. of s
most concern in estimating xenon polsoning is the efficiency of iodine
and xenon removal which in = lafge extent is dependent upon their solubility
in the fuel at these temperatures. If there were no removal of these
elements, the steady siate poisoning is valued at -0,297 percent reactivity for
an average thermal flux of 1.97 x 1013 neutrons per cme—sec, Assuming a high
degree of removal, the poisoning is approximateiy as 10% of the steady state
value with no removai. Poisoning worth of xenon is evaluated as 0,03 per-
cent in reactivity.

8.3.6 Delay Neutron Loss

 

Circulating fuel reactors suffer from a loss of delay neutrons

in that a fraction of the delay neutron precursors undergo neutron emission

 

 
-132-

in the external fuel circuit. Loop time for the circulating is 1.127

seconds of which 0,287 seéonds are spent in the active core., For small

loop times compared to the delay times of the neutrons, a valid approximation
to the required excess reactivity required to compensate this loss is given

by: (Eef° 69)
* where

 

o
I

Fraction of delay neutrons emitted per neutron emitted from the

fisgion event.

T, = Trangient time the fluid spends outside the active core.
T, = Translent of complete fluid loop

Resultant worth in percent reactivity has been evaluated as -0.56 §X/k.
8.3.7 Excess Reactivity Required
Tabulated below in Table 8.3.7 are the excess reactivities

estimated as required to waintain criticality for 4000 full power hours.
Table 8.3.7

 

Condition SK/K

Self Shielding 0.0051
Burnup and Fission

0.0158
Product Poison
Xenon Polson G.0003
Delay Neutron Loss 0.0056
Temperature ( +32° AF) 0.0007

 

 

Potal 0,0275

 

 
 

~133~

14.0 kgm U-235 addition to the critical mass (71.75 kgms) are required
to produce this excess, Total fuel inventory for the initial startup
(k =1.0275) is 605 kem of uranium-235.

8.3.8 Control Requirements

~ Under nominal 0pération during initial startup (no burnup)

the control rod must suppress 1.65% in reactivity (burnup, fission pro-
ducts and temperature excess). This requirement places the control
penetration into the core as 30 cm. Control curve in Figure 8.8 indicates
core reactiv;ty s a function of rod penetration. In obtaining this
curve, we have assumed the axiai.flux distribution as cosine in nature

and the positional worth as a function of the flux-squared.
 

JH0O N} NOILVHLI3INId dOYy
0¢S Ov o¢ 0z

 

 

 

BN

ORNL~LR=Digr 4 =25759
UNCLASSIFIED

 

 

-13k-

 

 

 

 

 

 

 

 

 

 

 

 

NOILIONOD NV3TD 10H

NOILISOd Q0¥ TTOHLNOD sA >t>_.5*<mm m:
| |

 

 

 

 

 

8-8 34N9I

OG-

Ov-

O¢-

0¢-

O’l-

Ol

0°'¢

 

ALIAILOVIAY 3IHOD LIN3OY3d

 
..135_

9.0 SHIELDING

9.1 Introduction

The shielding of HPMR breaks down into two main calculational phases:
(1) a neutron physics evaluation of the core, reflector, poison rod
~region and heat ekchanger complex with a flux plot and sodium activation
in the secondary salt.as end result, and (2) the shielding of the resultant
radiations prodficed by the neutron captures and activation.

The following shield write up gives a look into thé general methods
used and assumfitions made, No attempt is made to give a detailed
analysis of the shield calculation complete with sample calculations; etc,

It is assumed that the reader is familiar with the reactor materials
and configuration, and reactor compartment arrangement from previous

sections (Sections 6 and 7).

9.2 Neutron Flux Calculatlon

 

The neutron flux in the faactor vessel was approached with two methods.
The primary one was with three group-three reglon ORACLE calculations.
The secondary approach was to utilize comparisons of the ORACILE results
with multigroup work done on the ART configuration. It should be stated that
both these approaches leave fiuch t0 beldesired, The 3G33 ORACLE code
is set up for low absorbing systems., In some HPMR regions the absorption
is close to the value at which the code will not accept the calculation.
Extrapolation of ART multigroup data to HPMR is a bit shakylas the con-~

figurations are quite different,

 
 

~136-

The 'cross section dats and programming methods ‘are the same as
described in’ Section 8, O The main difference between the reactivity
calculations done under Section 8.0 and the ORACLE shielding work was
in the choice of regions. Two geometries were piogrammed.. The first
took the core as region one, nickel reflector as region two and BeO-
Boron poison rod region as reglon three, This gave a three group flux
plot to the outside of the poison region with leekage currents into the
Inconel'eupport sheet, IThe second ORACLE calculation considered the heat
exchanger region as a slab, region one; with two inches of Inconel ag
a symmetric reflector, region. tfio. The absolute values of the three
group flux in the core can be determined from reactor power. The more
unorthodox problem of esteblishing the value of the flux in the heat
exchanger, a subcritical multiplying system with a delayed plus inside
vall leakage neutron source, was done with a two. group, one region hand
ealculation where only a uniformly distributed,emiseion of delayed neutrons
vas used as a source. This was checked with another, independent two
group one region hand calculetion worked out in a different manner, Both
calculations gave a fast flux of about 2 x 10%° neuts/cc-sec at the
heat exchanger centerline, Scaling up the ART multigroup flux taking in-
to account the increase in heat exchanger thickness of HPMR over ART
anfi summing intotwo group energy intervals gave a fast flux value of
roughly 2.5 x 1012'neuts/cc-sec.

Looking at the results of these three calculations a flux value of
2.3 x 1012 was chosen as a flightly conservative value of centerline fast
flux (group one plus groué two of three groups) due to delayed neutrons

and fission neutrons in the hest exchanger,

 

 

 
-137-

A three group flux plot through the reactor pressure vessel wall
was then constructed from the flux distribution given by the ORACLE calcu-
lations and the absolute centerline flux values found as described in the
above paragraph. From the centerline of the core through the BeO-Boron
poison region this was a straight normslization. In the Inconel support
sheet, thermal shield, pressure vessel, and the heat exchanger, exponential
attenuation was assumed with the same slope as in the nickel reflector and
BeO-Boron poison region, This is justifiable on the basis of ORACLE
results in which the fast neutron attenuation through the nickel and BeO-
Boron region was approximately the same, indicating the slope is not a
strong function of material, To the three group exponentials were added
the heat exchanger flufi values taking into account a subcritical multipli-
cation of 1.43 from an ORACIE estimate of K of 0,3, The resultant summation
is represented in Figure 9-1A by the flux to the outside of the pressure
vessel, The flux through the thermel shield was determined essentially by
diffusiop theory from the ORACLE computations, From the pressure vessel
to the outside of the secondary shield,removal cross sections and 1id tank
plots were used,

The fast and epitherual groups (¢1 and ¢2 on Figure 9-1A) were summed
together and removal cross sectioné used for attenuation through the one
inch of inner shield tank steel and five inches of primary shield lead.

ART 1id tank data were utilized from the lead water interface on out through
the water. The 1id tank ANP mockup was a reflected moderated configuration
and substituted one foot of beryllium in lieu of HPMR's 6 inches of nickel,

5-1/4 inches of BeO-Boron poison and 3/% inches of Inconel, The ART mockup

 
 

-138;

only allowed for a 4 inch versus HPMR's 10 inch heat exchanger and 4-1/2
inches of lead as againét HPMB'S 5 inches, However, it w;s the closest
mockup configuration with neutron flux data-avaiiable. The energy spectrum
emerging from the mockup lead and HPMR should not be too greatly different
}as the same main elements are @resent although in different volumé fractionso
The relation between fast and thermal flux was taken from work done in
Reference 73. However, -the thermal flux plot presented in:this report

does not compare with the shape of that of Reférence 73, as Figure 9-1B

is based on more recent information and does not contein the Hurwitz plane
to sphere transférmation. Lid tank and shield_fiank vater was 1.45% boron.
Thé flux pibt shown in Figure 9-1B then reflects the shape of the above
mentioned flux distributions combined with the absolute fast flux value

at the lead water interface aé determined by removal cross section
attenuation of the ORACLE based flux in the fihermal'shield. Absfilute
values of thermal flux were then known in the thermal shield and at the
lead water interface (by relation to the fast flux). The thermal flux
distribution between these points were estimated from inspection of WAPD's
FIW flux plot where a similar configuration existed.

As reported in Reference 73, the fast neutron flux relaxation length
approaches a constant as a function of distance 1ln water at water thicknesses
on the Order of 140 centimeters or greater. Lid tank thermal data only
went oufi 140 centimeters in water, From ART neutron shield work it was
assumed that the thermal neutroné :eadhéd an e@uilibrium state with the
fast flux at about 140 centimeters of water. In the shield design, as
shown in Figure 9-1B, exponential attenuation was used for fast and thermal

neutron flux beyond 240 centimeters radius..

 
GO L o aseo

_139_

HP ME
& | THREE Grour

NeuTroNn FLux PLoT

ORE.

To
PRiMARY SHIELD LEAD
Fie ©-1a

    

F:LJJ>< ‘}é}n%-sgc
O

o

N

i
io

i
10

 

e
o s 80 {20
Kapius -CM Core | Ecru..moL B.o-’ ch'r Exc.l-( Ilusut l
e e '}
SUPPO
“THeRmAL
sunr oy

P- -

 

 
 

 

] o
| -140- ORNL~LR=Diyz, =25761
1t
10 -

 

 

 

 

H P ME
Two GroupP

lom
NeutTroN FLux PLoT
IN
PrIMARY SHIELD TANK
Fia. 9-1b
IC;’ \F

 

 

1y L~
<

$

 

 

 

FLux H/cm’- sec
L~
/

.

 

 

 

 

 

 

 

 

 

 

 

1o \\\ \

N\ N
S

N\

 

 

 

 

 

 

 

 

Zo: B0 WO 180  1eo ITO

LEAD _ RoroTED WATER

 

P -kt T
des T e 1
Ao Tr .

nry, .o
50 R wdte B

Lo

G f T

8o

{20 200 2i0 220 280 240

e ,D,,l&TANCr_ Feowm % Core- Cm
-141-

A fast and thermal neutron flux. dlstribution was estimated through
the north head into the primary shield plug based on the behavior of the
ORACIE calculated horizontal flux through the heat exchanger. This vas

used to determine thickness of gshielding needed above the reactor.

9.3 §§condary Salt Sodium Activation

With the neutron flux situation estimated in the heat exchanger the
secondary salt sodium activation was determined as the gamma source
strength in designing the secondary shield.

The average flux in the neat exchanger was calculated by numerically
integréting the three group heat excfianger flux {as shown in Figure 9-1A)
in a radial direction and then dividing by the heat exchangér thickness.,

No credit was taken for flux distribution in an axial direction as thils
would have been at best a rough eétifiate. Neglecting axial flux distribution
was the conservative approach.

Sodium cross sectional data was taken from Reference 61 and was con-
verted into three group averages as done in SBection 8.0, This was checked
by independently three group averaging some earlier unpublished Curtiss=-
Wright multigroup cross sections. Both averages accounted for the sodium
regonance peak at 300 kev.

Knowing the atomic density of'sédium in the secondary salt, and the
three group averaged cross sectiofls and flux, the activation vas determined
by their product:

N = atomic density of sodium in secondary salt

it

(Wt % Na in salt) p_,. X .6023 x 1024
23

 

= 9.46 x 10°
-142.

A = activations/cc-sec

 

:No‘a$
g;;;gi § | -GE(Barné) A Percent
1 3.52 x 102 2.1 x 107" 7.0 x 200 1.66
2 1,70 x 1012 2.1 x 1072 337.5. % 10° 81,14
3 - 3.0 x 10°° - 2,53.x 107t | 71.6 x 100 17.20
| | - b16.1 x 100

This gave a value of h300 curies total activation when multiplied by the

volume of salt in the heat exchanger (13.5 cu £+) and divided by 3.7 x 1070,

9.4 Dofie Tolerance Levels

 The basis for the HPMR is a maximum dose rate of 300 mrem/week and
20 hour a week access time to spaces immediately adjacent to the reactor
compartment (guxiliary engine room aft and abofié and thé storage compartment
in old fuel oil deep tank forward of the reactor compartment), Ten percent
of this is maximum allowed fast fieutron dose, Reduced to terms of mrem per
hour, this is a total of 15 mrem/hr with not more than 1.5 mrem/hr in
neutron dose, This set a fast neutron flux limit of 15 neutrofié per cm2
per sec taking the predominant neutron energy at 0.5 mev became the fast
neutron source'is mainly from delayed fieutrons born in the heat exchanger.
A flux of 10 n/cme-sec at 0.5 mev is taken as giving one mrem/hr (AEC
Standards For Protection Against Radiation, Part 20 of Title 10 of the code

of federal regulations, February 28, 1957).

 

 
-143-

9.5 General Shield Arrangement

 

At this point the steam genersting equipment sizes had been firmed
up to the extent that a reactor compartment arrangement could be worked
out. This was done with compactness and minimum shielding weight as the
criteria with maximum use of the available fuel oil for shielding purposes.
The arrangement chosen placed the'reactorlwith primary shield tank forward
and steam generating équipment aft. This layout allowed a smaller primary
shield tank, as the steam generators helped shadow shield the gamma and
neutron leakage from the primary shield tank. Fuel oil attenuated this
leakage out the forward, port, and starboard sides of the shield tank.

The fast neutron dose determined the thickness of hydrogenous material
reguired to attenuate to tolerah¢e dose 1evel (Section 9.4), Approximations
of the gamma dose with simplified geometries and source energies determined
the predominant rad;ations and gave estimates of lead thicknesses., With
egtimates of secondary shiéld thicknesses the general shielding arrangement
shown in Figfire 9=3 was laid out with a judgement estimate of the best pPro~
portion of shield material in primary shield to shield material in secondary
shield,

In light of a last minute alteration (Reference 73) in the shape of
the fast neutron atitenuation curve in Watér {this change is incorporateé
in Figure 9-1B) a foot of polyethyene should be packed around the after
side of the shield tank at locations not shadow shielded by steam generating
equipment as shown in Figure 9-3, This will have %o be fitted around existing

piping as it was not allowed for in the original arrangement,

 
 

NOILVWI1S3 LHO13M ¥Od aqa3Isn

 

I
=
-t

T

 

HOL3MS OJILVA3HOS

 

G713IHS AYVANOD3S

 

INITAHLIAN0d

S3HON!I O Ol ¢+
NOYdd G3¥3advl #

 

 

¢-6 34N9Id

 

 

 

 

 

 

 

 

 

 

 

0 GL'0 | #Ov r
0 SL'O | L0 i
o ! ot H
) S0 g 9
0 51 S 4
0 Gl SS9 3
o2l ol G2 a
02l 01 TR 9
0-2l ol 09 g
08| 01 S99 v
INITAHLIAI0d | NOMHI | av31 ] LNIOd

 

 

 

SIHONI *SSINNDIHL
ONIQ3LHS

 

 

 

 

 

 

 
-145-

9.6 Primary Shield

The primary shield deslgned for adequate fast neutron attenuation and
with estimated gamma attenuation was then checked in more detail for adequate
gamma attenuation, All gamma sources listed in Table 9-1 were considered .
with the enmergy distribution indicated in the table and with & simplified
source shape most closely approximating the actual source geometry. Form-
ulations as given in Rockwell's shield design manual (Ref. 33) for lines,
disks, cylinders and truncated cones with uniform and exponential source
distributions were used, All dose vaiues below .0005 mr/hr outside the
secondary shleld were neglected. The gamma dose from fission products
in the heat exchanger, prompt gammas in core and heat exchanger, and water-
lead capture gammas in the primary shield tank were éonsidered in more
detail as described below.

The fission products constitute.an important radiation source as they
are rapidly circulated with a reactor cycle time of 1-2 seconds. This
invalidates nuclear data on gamma energies and decay times, Therefore, the
energy group breakdown presented in Reference Th was used which takes into
acéount k.9 of the roughly 5.9 mev total available. This difference is con-
sidered to decay before the fuel leaves the core. Saturation of long lived
fisgion products is assumed which is conservative in this case. The pre-
dominant enexrgy was found to be 3.2 mev for HPMR shield thicknesses.

"The prompt fission gamma dose was calculated by an energy integration
under the continuous fission spectrum from .1 to T.46 mev, A mean value
for the HPMR shield was found to be 2.85 mev by running a series of energies

assunming all proupt gammas at that energy.
 

 

~146-

The energy situation for the lead and water is firm at 7.0 and 2.23
mev, However the source geometry becomes an important factor, especially
in the case of water where the souice distribution (thermal flux) varies very
rapidly and cannot be completely fitted with a simple sum of exponentials,
Both radiations together contribute 80% to that dose outside the secondary
shield which comes from gaumma radiatiOfi ieakage from the primafy shield
tank. The geometry was handled by numerically integrating the dose con;
tributions from unit line sources into contributions from unit cylindrical
surfaces in the primary shield tank., These cylindrical surfaces of different
radii wére then numerically integrated into the total dose contribution
from the lead and water volumes. This method essentially gives an exact
geometrical representation to within the accuracy of Simpson's rule for
numerical integration. |

Three directions from the reactor vessel to the outside of the secondary
shield were considered. One horizontal shot out through the primary shield
tank and secondary shield to the aft face of the after reactor compartment
bulkhead, and a vertical computation through the north head, shield plug
and top hat were done in some detail, Another horizonial calculation for-
ward through the fuel oil shield tank was done for lead capture gammas in
detail with estimates for water capture and heat exchanger fission product

gammas, The resulis are tabulated in Table 9.1.

Q9.7 Secondary Shield

 

The activation of the sodium in the secondary salt required that a
secondary shield be placed around the steam generating equipment., The

overall dimension of this shield were established by the estimeted volume

 
147

requirements of the reactor and primary shield, the steam generators, and

the superheaters., A plan view of the arrangement of this equipment within
the secondary shield is shown in Figure 7.3, The resuliing shieid 1s box-
gshaped with internal dimensions of 23' x 24! x 15' high, (Figure 9.2},

The thicknesses of shielding required were then calculated for a maximum
dose of 15 milliroentgen per hour on the outer surface of the top and aft .
faces of the shield., It was assumed that fuel on water would be used to
aid in the attenuation of radiations from the forward and side faces of the
shield, as described earlier,

Except for direcily over the reactor, the amount of secondary shielding
required was determined meinly by the secondary salt activity. The primary
shield is relatively 5igh1y effective in shielding reactor sources, The
total activity of 1300 curies introduced into the salt in the primary heat
exchanger was assumed to be distributed in the ste§m generating equipment
in proportion to the ratio of the volume of salt contained in any particular
component to the total salt in the system, The individual volumeiric
gource strengths were then obfained by dividing the curies of activity of

the salt in a component by the volume of that component. Thus

% of activity, ~activity

location total salt curies decays/cmgsec
superheaters 20,7 890 2.0l x 107
steam generators 36.0 1550 l.58lx 107
salt lines 36.0 1550 6.33 x 10

primary H X 7.3 310 (not contrivuting)
~148-

......

This assumption was recommended by ORNL personnel working with similar
systems, and appears Jtsfified in view of the relatively long half-1ife of
godium (15 hr) compared to the secondary salt cycle time (10 sec)., It

was further assumeq thet the U-tube geometries of the superheaters and
steam generator could be replaced by a straight_cylindrical gources of
equivalent volume. The self-attenuation of these sources was determined
by homogenizing the salt and tube bundles within each cylinder, and by
computing mass attenuation coefficlents for the sodium gamwme ray decay
energies of 1,38 and 2.76 mev. The approximation of:replacing the
cylindrical sources by equivalent line sources was used, and Peeble's
correction was applied to calculations involving slant penetration through
the shield.

Using these aspsuuptions, the thickness of shielding required wes
calculated for eight points in the secondary shield and estimated for three
more. It was attempted to select points which would give an indication
of the shielding required for the areas receiving bath the largest and
the smallest irradiations., Time dld not permit a Qbre extensive situdy.

The - "hottest" points on the inner surface of the secondary shield
were found to be (1) on the aft face of the shleld near the primary heat
exchanger, (2) on the top face of the shield over the secondary salt pumps,
where salt lines are near the surface, afid (3) directly over the reactor.
The most lightly irradiated point appeared to be in the middle of the front
face. Polyethylene was added to the aft face of the shield to attenuate
fast neutron leakage wfiich could stream aft between the steam generators.
In estimating the lead thicknesses required, it was first assumed that

steel structure would be necessary to support the lead in the following
 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6%1
0 -b2 -
6 - b1 >
_ 1l » |
fi wC =l -
fi 1
n
€-6 3¥N9Id o _
. WOLL0OE 318Nn0Q
ONIGIIHS = %
ANVANOD3S
any
AHVYWINd P
ONIMOHS "
NOILO3S INITHILNID 7 7] 39VdS LNINdIND3
“ ONILVHINIO NV3ILS
ININWLYEYEWOD \ )
<x
HO10VH4AN 7. “/(I.II)I
o _— \ Xod” 21
\ —
¥ILVM L -prarys 4
ANVL gslvdog | A1Dpucaes jo yoyaxs
d¥vmyo4 10 e e \ J40WIYIS sag
73n4 T
| T— v ol T° NMOHS 3¥NLONYHLS
— gz 130ddNS GI3IHS ON 3ILON
N ——
e — 9M1d r
) Q13IHS =
qd v Mw g b T qd 3,9\ ;
_ SRR RIPRAX AR : R R R R R
. AN
84 u__
ONILIVHS : "
A70d ,9 1 g SIAINQ
g dNNd H3IA0
qad % | 1VH 401
%5 2

TR w
£9L62 "DMT-YT-"INEO
v

ML - | _@

  
-150-

amounts: (1) 1 in, on front and aft faces, (2) 3/% in. on side faces, and
(3) 1-1/2 in. on top face.

It appeared that the primary radiation reaching the forward face
fiould be from the superheaters, Therefore, it was believed advisable to
provide shadow shields for the superheaters directly réther than add lead
to the larger area of the front face.

A "top hat" of additional shielding is required directly above the
reactor, to enclose the comtrol rod drive and the primary fuel pumps. This
shields against neutron and gamma streaming and leaking through pump well
and control rod penetrations of the primary shield tank plug.

The resulting shield is shown schematically in Figure 9-2. For
reasons of shortage of time and case of weight estimation, the shield is
represented by slabs rather than by contoured thicknesses. This'assumption
is believed to be conservative, and hopefully counter balances the omission
of additional shielding for plumbing penetrations. The total estimated
wveight of the secondary shielding, including the steel wentioned above, is

estimated to be 456,120 1b,

9,8 Summary and Recommendations

The controlling radiation in this reactor is fast neutron flux. High
fluxes in the core inelastic scattered in the nickel reflector and elastically
scattered in the BeO-Boron poison region are multiplied again in the
relatively thick heat exchanger region to becoume a determining radiation
source., TFuel oil, which is a good hydrogenous fast neutron attenuation,
was used to shield this source on the front and §ides. Gammas born in the

core are firetty well stopped by the nickel reflector before they start, bul
-151-

figsion product gammas born in the heat exchanger from a fast cycling, high
power density fuel add up to a twenty percent dose_contributién outside the
sécondary shield depending on the length of time of operation. The high
fast neutron flux is also influential in its secondary effect of thermali-
zation and capture in the primary shield tenk lead and water. The 1.U45
percent borated water helps the water capture gamma dose by roughly factors
of ten to one hundred. Lead capture gamwa dose is roughly 17 percent, and
water captures contribute about 56 percent of the total dose outside the
polyethylene in the auxiliary engine room.
| Structural material activations were not considered fof the shutdown

condition as they were assumed to be masked by the sodium activation.

The resultant weights tabulated in the weight section are 6.4 lfis/shp

for the primary shield and 1k.l for a total of 20.8 1bs/shp.

in the primary shield tank the use of a two-infih thick cylindrical
ring of lead sbout 15 in. from the existing lead is recommended. This
would shield the lead and water capture gammas in the high thermal flux
region and offer a means of cutting down on the fuel oil required to shield
these secondary gammas.

If time had permitted another reactorlcompartment arrangement, sfiace
should be made for putting the 30 in, of polyethylene around the aftexr
side of the primary shield tank to eliminate the 18 in., on the after bulk-
head. |

A quick ‘look was taken at the shield weight for the case if no fuel
0il was used for shielding. Two‘reactor compartment arrangements were con?
sidered. One used additional lead and water shielding on the existing

reactor compartment bulkheads and the other used a larger primary shield
-152-

tank and no water or polyethylene on the bulkheads, Both systems were
designed to reduce radiation to the levels stated in Section 9.4 and gave
an additional shield weight of about 10 1fis/shp. This gives a total
shield welght for a two-reactof-all nuclear ship of roughly 31 1b/shp.
However, no advantage was taken for rearrangements'df machinery, Also
doge levels were reduced to same value on all si&es of the reactor com-

partment.
-153-

TABLE 9.1

PRIMARY SHIELD TANK LEAKAGE DOSE VALUES

OUTSIDE THE SECONDARY SHIELD

(mrem/hr)

 

Mean

Energy Auxiliary Room

Compartuent

Auxiliary Room Forward of

 

 

 

 

 

 

Source Type Mev Forward Bulkhead Above Reactor Fuel 0il Shield
Prompt 2,85 0.00132 0.020
Na Decay 1.38
2.76
Core Fuel Capture varied
Be Capture 6.00
Inconel Cap-
ture 8.37
Reflector Nickel Cap-
ture 8.37
Nickel Inelastic
scatter ~1.5
Heat Prompt 2.85 0,007 0.017
Excginger Pission Product Spectrum 0,145 1.032 0.2
North Head Na Decay 1,38 0.025 0.063
2.76
ture
Fuel Capture varied
P.V. and
thermal Nickel Cap~- 8.37 0.012 0.068
Shield ture '
Fe Capture T.2
¥
Shield b t . . * 0.
Tani Capture 7.0 0.123 0.33 12
H,0 Capture  2.23 0.408 0.99 0.5
0.72h 2,549 12.7

 

*Fuel oil is a poor shield for 7.0 Mev lead capture gammas.
-154 -

.10.0 THE HEAT BALANCE AND GENERAL ASPECTS OF THE STEAM SYSTEM

10.1 Introduction

Tn this section is included the schematic diagram, Figure 10.1, of
the steam flow and the heat balance, The diagram gives the steam flow,
the temperatfires, and molten salt fiofi for a reactor power of 95,9 mega-
watts. This is tfie_reacfior power necessary to supply sufficient steam
for the full power of 35,000 shaft horsepower.

Most of the equipment shown on the dilagram has received comment

and description in other sections of this report. In this secfiion some
brief additional comments will be made and é comparison of the efficlencies
of the oil fired steam generating system and the reactor driven steam

system will be undertaken.

10.2 'The Steam Requirements

The heat balance for the steam system was taken from an actual test
of a clags 931 destroyer. To drive the turbines at full power, 218,760
1b/hr of steam at 950°F and 1200 psig (h = 1470 BTU/1b) is needed. This
was the étarting point for the héat balance. In the previous sections the
pumping powers for the reactor ffiel, the molten salt coolant, and the
recirculating boller water have been calculated. It was decided to drive
these pumps with turbines using superheated steam in order o have &
smaller unit within the secondary shield., A turbine-pump efficiency of
604 is assumed and the pumping power was multiplied by a factor of 1.25.
The feed water pumpinglpower is not changed since the feed water rate is

the same as in the oill fired system. The feed water pumps are also driven
 

gel

SENIT SNILIINNGD LNVIO0D AUVUNDIAE
$4001 NIYN LNVI009 AHVONGOIR

ROILYINOYIDAY GNY
"eRlYH] JA4ONYET ONVYTD

NY3 LS AMVITIXOY

 

 

 

 

A — . —

e b it

 

 

 

     

 

 

    

 

 

 

 

 

 

 

 

 

 

817056 93X WH/BT 0DECIT

D184 OOL OL LS00E
ONILYMAO 2 QATIVLEHT Z
SaNAd 3133 NIV

  

 

 

    

LA WyILs |
Holow i Holow 3
aNiiwHI40 ®  d37IVLISNLE

BdNNd HIYHA
BILYM HEIHG

GdWnd #ILE0DE Qa3 RIVH

 

 

 

 

 

          

 

   

 

HILYH HEIHS

   

   

 

                

VT WH/ENT SZ6
3+ 507
PNILO3IN0A wmsgR /0 )

WYFLET WYALE t

 

 

      

 

 
    

                  

    

   

       

 

 

 

 

 

 

 

 

  

         
            

  

     

 

 

  

KY3ALs GIAVIHHIGNEI] —rr — oo 891=4 /567 £Z217C22
MH/SE7 G261 \
MY ILE GILYIHEILANSG e an s ]
. vl qazg | 37 00F _ @ 30z oaviem 2
Q333 ONY JLVSNIGNOD smmwmmmimmmme WH /88T OLe'S3T o ! e._fl anuveId0 ] M&fifim "
TaNAs3 a4/018 €IS PHILYEAYAG sie Y %z i Inaea MYSNIGNOD NIV
’ 2184 Ot o | W i e HiY NIYR EE gpspa o1
: de 983 w » =2 w“fl Quidd ahn- IRV
- ®
MALY3H u3LYM 433 S|a = s oo
= n u.T. ¥H/E8T D
w EdNnd NIYRD
= HiT1T1LELE 1 WOLAY NI HIDIN
9i5d oomw o»u._.mog u.__zu_m,..uum__%uN 2 p15e At arsan osee ) fl
UBTYLENE 2. dRnd Y 4 o »
e O srvexa savoxd T T TR s TN _sirealo
@ ) STNUEER G
x v ! M /GBY viE
- 2ol y s o om e e S
whrEEN £HOTD 40 WYTT OGNV
. ]\-l _
gdn 0007 =
s HILVIHUIA NS0 1V NOILYY O N T 4 ol
dHS 000SE LY [ B0 AdAl HEYT4 -
Y »EOL a9¥Ll5.Z-049 0QZI A o
W31SAS 02H43M0d ¥0LDVAE dOd i # w e
1 " MH/EAT 002 > T ORILYHIAS e WA
HOS JONVIVE WVALS (3101034d { NopvaaNz ¢ : Lvshiono?
WYl - - o
0L -~ "913 _. “ ky - in I _.,wxmmn; ot
i x - Wi -
: _ Hu/sgat o ® GNILY KAO
1 1 = - WY¥3LE ON[LNIWZNY o SHOLANPT MI¥
m|
# HHS -ane.hl—fi “ “ ; ABYIIXDY T
e la = I 0o HNILYHALD | “
HH/AT DIX ST =f o SHOLOAFY WY : 1 WH/EAT £8D .
VA W = ¥ nHH ¥
wn zr  Rn miMl | 3 ¥33NLE10 1 _ “
MO4 GANDISIQ MM G az% o . i 3 i EIRE
LY DNILYM340 HOLOVIK «0Zik ) ! MH/EBT O 4= 0¥8
i, o] | Ly i e g | e D e ey
004t M3LY2HHZINE 34 (56T % | 440 N¥IT ANV gy 1y oNILYHIdO
HILYIM ¥ L wwsen L AYNHALND * . uzoiuuzflw.wuuam
L i g bk - - — - =
. uzens L nLe 921 3= 9E9  DISd owil L1907 9%
I.h.th._luo.._lxliov» —— ' 515 OF1
m... , v ="Ell { .

POLSE COMI-YTI-TINNO

 

 

 

 

 

an/nla 0iwi 4s 056 0154 0OT!

—-—
sa1 09L'M3

HH/5A7 Ol8Z
. e —

oiRd 0901

 
-156~

by superheated stean. The superheated steam requirements for the reactor
system are summarized in Table 10.,1.
TABLE 10.1

SUPERHEATED STEAM REQUIREMENTS

T. Turbine and Turbo-generators = = = = = = = = = = = = = = = = 218,760 1b/hr
II, Pumps:

(1) Reactor Fuel - - - 150 P HP

(2) Molten Salt = - = = = ~ = =~ 550 P HP

(3) Recirculating Water =~ - - - 35 P HP

 

Total - = - - = 735 P HP
Steam Required 1.25.x oshs (BTU/hr) /hp x 735 hp 27,500 1b/hr
(170 - 1328) BTU/1b
TII. Feed Water Pumps = = = = = = = = = = = = = = = = = = === 11,700 1b/hr
Total Superheated Steam - - = = = = = = - = = 257,960 1b/hr

| Desuperheated steam is required in the galley, the air ejectors, feed
booster puwmps, lube oil pumps and condensate pumps. Desuperheating is
achieved in the steam drums by cooling superheated steam in tubes that
pass through the saturated water in the drums. The steam is cooled from
950°F, 1200 psig to 625°F, 1165 psig. The desuperheated steam requirements
are summarized in Table 10.2,
Table 10.2

DESUPERHEATED STEAM REQUIREMENTS

e
Air ejectors, galley, leaks, efc, = = = = = = = - = === === 3,371 1b/hr
Feed booster PUUP = = = = = =~ = = = = = = = = = = = = = = = == 895
Lube Oll pumps = = = = = = = = = = = = = = = = = = = = = === 300
Condensate pumps = = = =~ = = = = = = = = = = = = == - ===° 75

Potal = = = = = = = = = = = = 5,341 1b/hr
~157~-

10.3 Condensate and Exhaust Heat

The deaerating feéd tank collects exhaust from some of the suxiliary
equipment and it also receives the condensate. From the DAFT is drawn
the feedwater which supplies the steam generating system. In order for
the deaeration to be compiete, the pressure in the DAFT should not exceed
18 psig. The enthélpy of the saturated liquid at this ?ressure is 225 BTU/1Db,
This is the enthalpy sssumed for the feedwater entering the steam generating
gysten,

The DAFT is unable to handle the exhausts at full power steam flow so
it is necessary to run a portion of the exhaust directly to the condenser.
This excess exhaust 1s 12,530 1b/hr at full power. This is slightly higher
than the oil-fired systems 11,570 1b/hr; therefore, a small increase in
condenser capacity way be necessary to handle this additional flow.

A summary of the efihaust andlconéensate flows and their respective
enthalples as they enter the DAFT ig given In Table 10.3.

HEAT ENTERING THE DAFT

 

From: - w(ib/nr) n({BTU/1b)
Feed and Circulation Pumps - = - - - - 36,900 at 1,328
Feed Booster Pumps - - = = = = = = = 895 at 1,213
Lube O1l Pumps = = = = = = = = = = ~ 3b0 at 1,253
Condensate Pumps = = = = = = = = - - 775 at 1,218
Fresh Water Drain PUmps = = = = = = = 1,975 at 168
Condenser - - ; ----------- 220,1i3 at 102
Distillers « = = = = = = = = « - = - 2,300 at 148

HEAT LEAVING THE DAFT

 

Boiler Feedwater - - = = = = = - - - 263,258 1v/hr at 225 BTU/1b
~-158-

10.4 Heat Addition in the Steanm Generafing System | |

- The steam generating system must add sufficient heat to bring 257,960
1b/hr of water at an enthalpy of 225 BTU/lb up to steam at an enthalpy of
1470 BTU/1b plus 5,340 1b/hr of water at 225 BIU/1b to steam at 1263 BTU/1v.
This total heat addition is 3.272 x 108 BTU/hr or 95,9 megawatts,

In the reactor system a feedwater.héater is needed to do ‘the job that
an economizer does in an oil fired system. Feedwater from the DAFT ai 18
psig is raised to a pressure of T00 psig by conventional boiler feed pumps
and fed intolthe feedwater heater. Here, satuiatad sbeam at 1250 psia is
mixed with the feedwater to produce yater at 486°F, It takes 91,730 1b/hr
of saturated steam to achleve this. The 486°F water is now pumped to a
preggure of 1500 psia and let down by thrbttling t0 the boiler pressure
of 1250 psi. The feedwater heater forms an integral part of the steam
generating system and with the saturated steam used for the heating forms
a closed loop within the sysfem.

As has been stated in previous sections, the heat addition to the
steam generating system is by means of a molten salt. This salt drops
17.9°F in temperature in the superheater and 58.80F in the steam generator
at a flow rate of T.49 x 106 1b/hr. These temperature drops and flow

rates represent an input of 3.272 x 10° BTU/hr.

10.5 Comparison of Efficiencles
No attempt to compare thermal cycle'efficiencies will be made here
but only a simple calculation of the gross power-plant heat rate, For

the reactor system:
_]_59_

gross power plant heat rate

For the conventional oil fired system:

gross power plant heat rate

Heat Input
Shaft Horsepower

8
3,272 x 10 BTU/hr
35,000 shaft horsepower

9,34%0 BTU/shp-hr

3,108 x 108 BTU/hr

35,000 shaft horsepower

8,890 BTU/shp-hr

The reactor system does require more heat input because the additional

pumping power required for the molten salt and recirculating water is

greater than the power required for the fuel oil pumps and forced drafi

blowers.
™

~160-

11.0 OVERALL POWER PIANT PARTICULARS

11,1 Introduction

In order to determime the overall feasibility of a fused salt reactor
installed in a particular class ship, it is necessary to consider all of
the components in fihe complete system. A preliminary piping layout and
drawings of the steam gen;rating équipment are included in Section 7.0,
Rough sketches of the'primary:and secondary shields are also presented in'
the shielding section (9.0). To completely determine the suitability of
the resulting power plant, 1t is then'negessary to investigate the
installation as to its effect on the ship's overall construction, balance,
etc.

In addition, it ié also necessary 1o consider operating problems such

as control, emergency operation, and malntenance.

11,2 General Arrangement
A brief study of the possible layout of steamn generating components

within the reactor compartment and the location of the reactor comwpartment

 

in the ship was made with minimum shield weight as the major consideration.
No detailed optimization was attempted but rather judgement was used as to

the relative sizes and -location of the primary and secondary shield, Giveh
the decision of only replacing one oil fired plant with nuclear pover,

arrangements were worked up using fuel oll as part of the shielding. As

 

pointed out in the shielding section, fast neutrons are the primary radiation
problem in this system. A hydrogenous liquid like fuel oll takes the place

0f polyethylene and serves double duty as fuel for the oil fired plant.

 
-161-

Arrangement One

On first look, the best location for the reactor compartment seems
to be the aft fire room where accessibilitj for removal of reactor com-
ponents is done through the upper deck, However, preliminary estimates
of steam generating equipment sizes indicated a larger reactor compartment
than shown in the final design was required. To prevent propeller
shaft penetration of the aft reactor compartment, the compartment would
have had to move off centerline to such a degree that battle damage
stability problems would arise, Therefore, arrangement one (see Figure
11-1) was worked out with the reactor compartment in the forward fire
room, Provision for removal of the primary shield tank plug and reactor
vesgel could be worked out through a side port, as there is twelve feet
of clear height between the top of the secondary shield and the main
deck. If removal through the main deck is dictated, the bridge super-
structure would have to be removed.

Tn locating the exact position of the reactor compartment, use of
existing bulkheads and deep web frames were made, The forward boundary
of the compartment is existing bulkhead 63 and the after boundary is
deep web frame T5., Tying into existing main structural members winimizes
additional support structure in the nuclear power conversion,

A weight and moment study was made on arrangement one. The weights
and moments of the boiler plant were replaced with the reactor system
jincluding fuel oil shielding tanks, Fuel oll was distributed in the exist-
ing after fuel oll and ballast tanks to balance the forward mowent pro-

duced by the increased weight of reactor compartment over the boller com-
-162-

ponents. Results of this study showed that if Just the deep tanks aft of
the after engine room were used, a resultant trim of 1.94' by the stern
1s produced as compared with a 1.63' trim by the stern for a completely
conventional oil fired destroyer. This gives a total of 391 tons of
fuel oil compared with 728.5/2 = 364,3 tons per one oil fired plant. As
fuel is burned out of three after tafiks, the ship evens out, When the
stern rises to.the point where propeller emergency or sea keeping ability
becones a probiem, sea water ballasting will be needed in these empty
after tanks.

Even though the fused salt systém has a vertical center of gravity
4.5 feet below the boiler plant, the total nuclear powered ship C. G.
gtays about the same due to empiying 173 tons from the relatively low
fuel oil and ballast tanks forward. Since the total ship weight and
free surféces stay about the same, the free surface corrected wmetacentric
height (indication of ship stability) of about 3.2 feet stays about equal
to the conventional DD931l., Moment calculations showed that the exact
change in metacentric height was sensitive to more exact values of weilghts
and centers of gravity than could be calculated for the miscellaneocus items
in this feasibility study.

Arrangement Two

 

The finalized reactor compartment width was reduced to 23 feet. This
reduction allows the possibility of locating the nuclear plant in the after
engine room with onlylthree to four feet of off centerline required to
avoid the forward propeller shaft. This is shown in arrangement two (see
Figure 11-1). With fuel oil shield tanks on both gsides of the reactor com-

partment, the dangér of serious list if damaged is lessened. A detailed
£l

IVEANIQIANG

 

AT B IR

|
AUOALVId GNOT3IE

     

     

MHTL
wINIee s

 
      

1O "RENL <7

 

ik 40 3

   

u LD LAY
fomainn o4

 

to3l TON JYUNOLL

¥IAAOHLERD SEVYTD 1864 QC
HOLDVIY ANIMYN ZINVHHOIN3d RIIH

 
 

 

o

 

.

MOON INIDNT | 0N
1'OKr "WAS CRHNL

(I
]

FON N3 'ORHAL

 

ool

 

ARV
o
)

 

     

 

 

 

AL
aanie

 

nOON
AOY

      

         

 

 

SO I W ) me XIS
37t4Q0Nd QHVOBHN!

 

T ARIWIDNYUNY
AMYITIXOY $ HOLOYIY

 

¥ ININAONTHYY
ANYLUIXOY JHOLDYIH

 

ANDRLNYA RDY
NOLDYEY

     
    

Tanmniwsenns
ADLY TN

 

1
1 20vdil WOOM AMYIUXAY

4

 

AD v WOON

 

TYLLNGOIANOD
OLE 'OMI-uW]-THED

1
\
1

—
——
.
~164 -

damage stability evaluation should be made before arrangement two can be
recommended with certainty, but aside ryom damage contingencies, the
arrangement offers the advantage of putting the heavy, concentrated weight
of the reactor compariment near the C. G of the ship, thereby requiring
jess fuel oil, 303 tons, to balance the moments to give egsential the con-
yentional full load condition trim aft. The transverse stability gituation
is better than arrangement ope in that the forward tanks have considerably
less free surface than the after tanks which are =mpty under BPMR, arrange-
ment two, full load conditions. The resultant free surface correction is
.16 as compared %o 1.61 feet for arrangement one. These forward fuel oil
and ballast tanks have a lower cente: of gravity than the after tanks
used in arrangement one, but they hold less oil. Again as in arrangement
one, exact values of metacentric height cannot be calculated with any
confidence without a more detailed machinery arrangement, but 1t is indicated
that arrangeument two has bvetter stability than arrangeuent one and has
strong possibility for good improvement over an oil fired DD931.

In su@mary, two general afrangements were worked on., Both give the
big advanfiage of decreased space required. A detailed arrangement of
the auxiliary room was not worked out, but a relatively large amount of
the original fire room ig 1left both aft of and sbove the reactor com-
partment, In arrangement one, fourteen longitudinal feet of deep tanks
are freed for armament stowage or other use, Also a portion of the room
left for auxiliary space could be used for stowage.

Stability looks to be roughly about the same as & conventional DD931
with increased oil C.G.'s balancing a decrease in steam generating center

of gravity caused by elimination of uptakes and stacks and the design of
-165-

a compact, low reactor vessel, steam generator and superheater, and sur-
rounding shield. The free surface’correction can be éontrolled to some
extent by keeping the fuel o1l shield tanks full and under slight pressure,
As a feasibility project, the vertical moment study has indicated that a
detailed design with an eye to the stability problem, especially in an
arrangement of type two, cofild lead to an increase in metacentric height
which could be gladly used by the armament people to add migsile launching
and guldance systems topside;

Perhaps the greatest restriction in these arrangements is a lack of
flexibility in filling and emptying fuel oll and ballast tanks, Salt water
nust be used for trimming purposes which brings up contamination problems,
When fuel oll shield tanks are tapped, fuel oil or salt water must be pumped

into the bottom Lo maintain the shield and =eliminate free surface.

11,3 Power Plant Control

 

11.3.1 Introduction
Due to its negafive temperature coefficient of reactivity,

the fused salt circulating fuel reactor is self-regulating. That is; the
pover produced in the core of tfie reactor follows the power demanded by the
load with some characteristic time lag., The steady state mean temperature
of the fuel in the core remains constant since, in the absence of control
rod motion, burn up, and fission product build up, the reactor is critical
only at one temperature, Of course, other temperatures throughout the
system will véry with load, |

Bven though the reactor is self-regulating, there are several reasons

why a control system way be incorporated in the power plant design, First
~166-

of all, the transient response of the system may be poor, Fér example,
load changes ma& resuli in lafge temperature overshoots which, in turn,
cauge intolerable thermal stresses, A properly designed control system
can ifiprove transient response.

A control system may also be used to set up some desirable pattern of
steady state temperétures, pressures, and flow rates throughout the plant
as functions of power output. Such a pattern is called the plant program.
FPor the HPMR a constant steam temperature program is desirable. This
requirement is dictated by the fact that steam turbines for marine power
plants reguire esgentially constant steam condltions regardless of load.

11.3.2 Types of Control Systems

Several types of control systems seem to be possibilities
for establishing the constant steam temperature program. For example,
control rod position in the core may be varied as a function of output steam
tempe;ature. With é negative temperature coefficient of reactivity, con-
trol rod poéition determines the mean fuel temperature in the core and
thus fixed the level of temperatures throughout the system. Thus, ii
seems possible that steam temperature could be maintained at a constant
level by such a system,

Another system which strongly suggests itgelf is controlled by varying
the flow rate of'the inert salt in the intermediste heat transfer loop.

The rate at which heat is carried away from the primary heat exchanger
depends on the sglt flow rate and the difference between inlet and outlet
salt temperatures. Thus, if flow rate is varied with power output, the
steam temperature can be mainiained constanfi. This system has the distinct

advantage that the pumping power required decreases with decreasing load,
-167-

o

There is & resulting gain in efficlency which is lacking in the other control
systems, There 1s one other factor which should be considered here. In
any system in which the flow rate is varied, the possibility exists for
tra;sitions from turbulent to laminar flow and vice versa. Such transitions
usually result in large thermal shocks and are highly undesirable. In the
HPMR power!plant salt flow in the primary heat exchanger and steam genergtor
is laminar at design power and is well into the turbulent region in the
superheater., Thus, the flow rate can be varied over a wide enough range
to make control by this method feasible. |
A third possible control system involves & by-pass line across the
salt side of the primary heat exchanger, As the load is decreased a valve °
in the by-pass line i1s opened allowing a larger percentage of salt to
by-pass the primary heat exchanger. Thusg, returning cold salt is mixed
with the hot salt from the heat exchanger with the result that salt tem-
peratures throughout the'rest of the system can be adjusted to hold steam
temperature constant, A study to determine the optimum control system was not
attempted due to time limitations,
11.3.3 Simulation
It was decided to set ué an analog simulation study of the
reactor and power plant on the Reactor Controls Computer (Reference 30)
at the Osk Ridge National Laboratory. The study had two main objectives:

l. To determine the transient response and stability of the reactor
and power plant when subjected to changes in load, changes in
reactivity, and other perturbations,

2. To determine the ability of one particular type of control system

to maintain constant steam tempersture,
-168-

For details of the simulation, c¢ircuits used, etc., see Appendix 1l.1l.

A schematic diagram df the system which was simulated is shown in
Figure 11.2, Two heat transfer circuits are shown; each handles 62.5
megavatts at full power. Due to the limited number of amplifiers avail-
able on the computer Only'cirauit 1 was simulated in detail, In circuilt
2 as shown in Figure 11,2, the superheater and steam generétor vere
approximated by a single heat exchanger. Circuit 1 represents the arrange-
ment of components as visualized when the study was set up, It is not
markedly different from the arrangement finglly decided upon,

The control system chosen for simulation was the by-pass line across'
the primary heat exchanger;. This system waé chosen because it was relatively
easy to simulate and because it offeved good possibilities for control.

Not enough eguipment was availlable to simulate control by varying salt
flow rate. No control system was simulated in circuit 2.
11.3.4 Results -
Since the details of the_simulaticn are presented in the
appendix only the results will be indicated here.

In order to study the transient behavior of the reactor and powexr
plant a nuwber of runs were made with the control system inoperative. The
result of the first such run is shown in Figure 11.3 With the reactor
operating in steady state ét full power, the load demand was reduced linearly
%o one-half power (62.5 megavatis) over a period of 15 seconds. As can
be seen from Figure 11.3, reactor power followed the load demand and
stabilized at half-power with no undershoot. The mean fuel temperature
in the core reached a peak of about 1236°F and then returned to its steady

o
state value of 1225 F also without oscillation.

 

 
~169~

The results of the above test seemed to indicate that the transient
response of the system was completely satisfactory., As further verification,
it was decided to subject the plant to a wmore severe load change. In
this test the load demand was increased instantaneously from 10% pover
(12,5 megawatts) to full power, The results are shown in Figure 11.L,
Reactor pofier and temperatures throughout the system leveled out at steady
state values without_oscillation,' A number of other runs involving load
demand changes were made inciuding cases involving 25% and 50% overload.

In all cases the reactor and power plant éppeared tc behave as a critically
damped system; that is, reactor power followed load demand without
oscillation and temperature swings throughout the system were very mild.

Several runs were made to investigate the effect of gtep changes
in reactivity. The results 6f one-éuch test are shown in Figure 11.5 At
t= 0, a step change in reactivity of +0.2% was introduced. At t=70
seconds, a step change of ~0.2% was introduced.

All of the tests described above seemed to indicate that the transient
responge of the reactor and power plant was satisfactory. Therefore, phase
two of the simulation was devoted to the study of the control system,

As stated previously the purpose of the control system is to establish a
constant steam temperature program. The system which was simulated is a
by-pass line across the sal®t side of the primary heat exchanger. The
amount of salt flow through this line is determined by the steam temperature
by means of an elementary servo system of the on-off type. The salt flow
which could be by-passed through this line was limited, in one case, to

75% (1570 pounds/second) of the total flow and in another case to 90%
-170-

(1880 pounds/second) of the total flow. Figure 11.6 shows steady state
steam temperature as a function of load for these two cases as well as
the case where the control system is inoperative, Steam temperature is
held constant over the range of 60% power to 100% power for the 25% flow.
cutoff and over the range of 30% power to 100% power for the 10% flow
cutoff. The amount of salt which may be safely by-passed is probably
limited by the temperature difference across the salt in the primary heat
exchanger. A%t any given power level this temperature difference will
increage as the salt flow rate through the exchanger is decreased. No
study was attempted to determine the maximum tolerable temperature
difference, |

Also of interest is the transient response of the power plant and,
in particular, the steam temperature during load changes., Figure 1l1.7
shows the resulfs of a run in which power demand was decreased from full
power to half power in 15 seéonds. The steam temperature sitabilized at
its design point value (975°Ff* afterlafiout 100 seconds. The maximum
excursion of the steam temperature was almost 100°F, It should be noted
that.little attempt was made to optimize the control system., An optimum
system would undoubtedly improve this trangient response. It is interest-
ing to note that reactor power undershbots 1ts steady state value after
the load change. This is in contrast to its behavior with the control
system inoperative. Temperatures throughout the system also oscillate

slightly.

*mmis value was changed to 950°F in the final design.
~171-

11.3.5 Conclugions
The results of the simulation indicate that the kinetic
behavior of the reactor and power plant is completely satisfactory. These
regults alsoc demonstrate the feagibility of a control system to maintain
a constant steam temperature prograu, A more detalled study of all of
the possible céntrol systems is required to determine which is optimum,
Because of the higher efficiency obtalnable, control by varying salt flow

rate appears most atiractive at this tinme.

11.4 Emergency Operation

It is extremely important that any reactor installation subject to
battle damage be as inherently safe as possible. The demonstrated stability
of reactor systems of this type (Ref. 6) along with the elimination of
numerous integral control rods mekes it basically very desirable.

In addition, however, consideration has to be given to emergency
conditions, both major and minor, not only to establish an overall safe
system but to maintain operation if possible and to prevent damage to
the reacior.

A partial list of such considerations as applied to this system is
given below:

(1) Primary fuel pumps are over-designed sO that high power operation
can be maintained in the event of partial failure,

(2) Primary and secondary pumps are d:iven by steam (available from
both the reactor and conventional system) and backed up'b& an electric

motor which can be operated from emergency service.
-172-

ORNL=LR=Dwg , =25766
UNCLASSIFIED

 

 

 

 

avon

 

 

 

133HS MO1d

 

 

 

 

 

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HOLVHEINIO
WVY3lS
=
. " ¥IONVYHOX3
YW 1V3H
, AHVAINd 3405
M
.n@ -u.@ .u-.m
ua?fi $
_:W . :»@ 0 .pwm. Jw
= X ¢ = - /= avot
“e [s *m vy |0 g
d31v3HY3dNS
¥ 1IN0¥HI0 2 L1InJO¥HID

Y3I4SNVYL LV3H

¢l 3¥N9ld

434SNVYL LV3H

 
03S — 3NWIL

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

06 08 0L 09 oS ot 0§ 02 ol 0

[
D
T~ e L
e =

* ~——
MM ~<

1 S
mc ../il
= ~
& ™
o /.

N

£ GNVW3IG ¥3IMOd NI //

1

JONVHO dWVH V HOd4 3WIL SA
4N3IL ONV H3IMOd HOLlOV3IY
_cozi| dN3L 1304 NVIN —-— //
YIMOd HOLOVIY =~ ———
0S121 OGNVYHW3AG ¥3MOd \
\
05232 —= N\
....../ \
0o GE3! — - o p—"
0S¥ 2l
¢l 34N9id

06¢

Gl

00l

Gel

MW - d43IM0d
 

06 08 0L 09 oS

'03s — JNWIL

 

 

 

 

 

o+ o¢ 02 ol 0
2 ;.
B A ANVN3Q ¥3IMO0d NI /
LE ———39NVHO d31S V H04 IWIL SA 7
mw dW3IL OGNV H3IMOd HOL1oV3Y
32 =
25 m
& W ‘
-
- /
m \‘
© /
M 7
z /
. /
T % /
2
c Vv
2 S
I -~
-
e \
9 ...ll.....\..\\
Ifl_ R .I.l..llll...l.ll.ll
<31 | —
dW3L 1309 NVIN — - .
HIMOd HMOLOVIYH————m— - P /
§02/— anvw3a ¥amod T \
l....\ll_l_lll
|I.._..|\....l| /
22| e i
St 2l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Y 11-34NOid

 

 

G2

0G

w0
~
MWN — 43MOd

00l

G2l

 
UNCLASSIFIED

ORN L~LRe Dutgo =25769

«175=-

'93S — 3IWIL

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

oS! 0zl oll 06 oL oS o¢ ol 0 o
yad
7
-~
~ |
\.\\ :
- —— G2l
= .1...././ *
= ~
Z N Il
N
n
c \ _
m \
— \
= |
= ¥3IMOd ¥OLOVIY——- 062
3 JUNLVEIdWNIL 13Nd NVIW
c22i —
052 //
Slélinas0L=} 1V % 20- ANV O=4 1V %20+
005 | ALIAILOV3Y \
| NI S3IONVHO d31S ¥O04 INIL SA \
JHNLVHIdNIL ANV mmzom_mozqum | .
G2t ! _

St J4NOI4

MWN — H3IM0d HOL1OV 3y
OKNL=LR~Dwg. =25770
UNCLASSIFIED

-176-

09l

VOl ¥3MOd 11nd 40 LN3dD¥3d
00l 08 09 ov o<

ovlI

02l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

/.l-
/]Ir....
nnumew“;;;;; o N\ ///
\
..//
/...r
;;;:;f::,; 4///1
440100 %01 —WILSAS TOHINO) ======
4401N9 %GS2 —WILSAS “TOHLNOD
WILSAS TOHINOD N e o
|
avol SA
FUNLVYHYI4dWNIL WY3LS LNdLNO
9l 3IYN9IL

008
006
N
Ilm
m
) -
oool *
_ —{
m
=
0
o
00I1 2
nl.—
c
X
m
oozl |
2
00¢ |
ORNL~LR-Dwg, =25771

=177~
(do) 3¥NLVYIIWIL WV3ILS

03S— 3NIL

 

 

 

UNCLASSIFIED

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

oLl 0S| o<l oll 06 0L 0s o¢ oL o
NOILVHI4O NI W3ILSAS TTOHLNOD
‘ANVW3Ia ¢
¥3IMOd NI I9NVHO HV3INIT V
HO4 IWIL SA 3HNLVNILWIL
WV3LS ANV H3IMOd HOLOVIY o
\.\\\ ............_.._Ir : Omu
7 /!!!
/ ~ L
AL - .lnlllfl.llul..'.lh......l -
. S,
l/ .
< \
S : Sl
N
N \
N /
N
T 00l
\ \
\
\
\
GL6 e —— \l// -~ sai
§e0l ™ Funive3dnaL wvals ——— — S /
HIAMOd ¥OLOVIM -------
ANYWN3d 4IMOd — - —
S.0l (

 

 

L1l 3d4N9ld

MN—Y3IMOd
~-178-

(3) Fuel flow is in the direction of natural circulation which aids -
the fuei inertia in remofiing the peak afterheat immediately after the
afterheat without over-temperaturing the critical areas.

(4) Provisions are made to dump the fuel from the reactor into dump
tanks if necessary.

(5) Blow outlvalves are incorporated into the_system.which would allow
drainage of the primary or secondary fluids into dump tanks should the
system go sbove design pressure due to over-temperature or leakage from
the high firessure steam side., |

Inconel drain tanks will be located in the inner bottom of the vessel
directly below the reactor. There will be firovision made for circulation
of secondary salt in Inconel pipes throughout thg tank. Drain tank will
be maintained at llOOOF at all times after starbtup of reactor. Secondary
salt will be bled off from superheater loop, and will act as heater for
drain tank and as coolant to remove decay‘heat when hot fuel is Introduced
from reactor. Under emergency conditions if secofidary salt cooling becomes
impractical, water cooling will be made available. Thermal insulation will
separate the tanks from ship'’s bottom. Hot fuel may be returned to regctor
from drain tank by helium pressure,

Criticality calculations have not been made for this system to fix
size and dimensions: however, it is expected; from similar systems, that
any reactivity could be overcome by'use of poisons. Li in the secondary

fluid will contribute to this result,

11l.5 Maintenance

In considering the overall maintenance picture, because the power plant
-179-

is strictly conventional, only that perfiaining to the reactor system will
be discussed. Servicing of the steam generating equipment for the basic
system presents a problem due to the residual activation of the secondary
salt. However, since duup tanks can be provided in the hull double bottom
and if careful design attention is given to assure almost complete draifiage,
it is reasonable to expect that direct maintenance could be done after a
2-day cooling off period. The alternate system proposed using two inter-
mediate fluids not only offers a lighter system but would completely
eliminate this problem. Direct maintenance on the steam generating equip-
ment could be allowed at all tiumes.

Two entirely different concepts exist for overall maintenance on the
basic reactor itself, Since it is beyond the scope of this report to
evaluate these, both methods are simpiy presenfed with the recommendation
that a careful evaluation be made in the future.

In eitherlcase it is felt that the rveactor should be of sound design
and‘testeé sufficiently to ensure that any installation would be of reason-
able duration. Then because of the additional complexity and cost of having
remote handling equipment designed to perform in the confined space aboard
ship, it is recommended that the reactor be removed as a whole assembly
and work performed at shore facility. This may be done by remotely cutting
the two inlet and exit pipes feeding salt fo the steam generating equipment,
disconnecting the reactor from the supporting structure and then 1ifting
the assembly from the surrounding'shielding.

At this point thefe are two alternatives pértaining to further mdintenance:

(1) Have the reactor designed so that a complete disassembly by remote’

operation is possible, This obviously requires a more complex reactor vessel,
-180-

internal structure, and heat exchanger arrangement, Also a.means of remote
handling, seal weld cutting and welding, as well as remote testing and
inspection is reguired. |

(2) Utilize a completely unit basic design that cannot be teken apart
and reagsembled but has greater simplicity and hence more reliability.
Upon malfunction, the reactor would be taken from the ship and discarded
after the recovery of valuable material, i.e., BeO, There is strong reason to
believe that the cost of discarding the reactors that become faculty in
service would be more than offset by the elimination of the very complex

remote handling equipment, facilities, and personnel required from (1),

11.6 Removal and fiispoéal of Volatile Fission Products

There are two possible techniques by which removal and disposal of
volatile fissioh products can be achieved: 1) periodic removal and
disposal, andIE) continuous removal and disposal, Periodic operation is
recommended over the continuous cycle.

As pressure builds up in the expansion chamber due to fisslon produét
gas generation, a ?resSure relief valve permits excess gas to flow into
an originalk& evacuated disposal holding vessel. The vessel would be perhaps
copper tubing fabricated into a spiral form immersed in the fuel oil
biological shield at the fore side of the reactor space. Provisions for
storage of several such holding vessel would permit the holding of the
discharged off gas for sufficient time to cool radicactively, to levels
which would allow the vessel to be cast overboard without hazard to ship's
personnel. |

If expansion chamber has & free volume of 1 ft3 after operating fiemparature
-181-

is attained, gas may be allowed to accumulate for several days before dis-
charging to holdihg tanks. A helium purge would not be required under
these conditions, the heating rate (see Appendix 11.2) due to radicactive
decay in the expansion chamber, would be about 150 kw at equilibrium (where
production rate equals decay rate for nuclearly unstable gases, neglecting
neutrofi absorption loss rate) when reactor is operating at its rated
125 Mw of heat. At these rates of gas generation (approximately 0.2 moles/
day), pressure in the expansion chamber would rise from 20 o 50 psig in
about one week. At this {ime, excess gas would be bled off into the
.evacuated holding chambers. Initial heating rate in the holding chamber,
assumipg half the gas were removed approaches 75 kw. This would decay
rapidly and give a gamma source after 14 days of 5 x_lOll curies of 0.083
mev, 600 curies of 0.1 to 0.3 mev and 30 curies of 2.4 mev, as well as a
beta heating of 380 watts.l

As the alternate method, reéctor way be continuously purged with
helium at a rate, say, 1000 liters (STP) per day. A%t this rate, heat
‘generated by gas in expansion chamber would be about 30 kw, Ref., 50. On
~ leaving the expansion chamber, off gas would enter a cooled holding chamber,
After a specified holding time; chamber is continuously exhausted into
ship’s wake, An average holding time of 4 hr would permit decay of the
gas to about 0.2% of its initial activity in the holding chamber. During
periods when the ship lay at wet dock, off gas would be pressurized and
xenon and krypton retained on cooled activated charcoal beds., Beds would
be purged with hellum after ship was underway.

Calculations have been made on the activity of ship's wake during
-182-

continuous dischérge, assuming 4 hr and 48 hr holdup periods followed by
continuous exhaust of gas inito ship's wake at a speed of 26 knots, Results
are as follows: A 4-hr holitime might result in & meximum dose rate of
8 mrem/hr in the ship's wake at time of exhaust, and a resultant 660 gamma
events per second per cms of ocean water in a ribbon wake, 20 £t deep by
100 £t wide, Instantaneous turbulent dispersion is assumed for the ribbon
wake, after which diffusion, wave'actiofi, and ocean currents would govern
dispersion of wake. It is expecfed that the weke would remain somewhat
intact in calm 'ézeathér .:g‘or geveral hours. A 48-hr holdtime would give a
maximum dose rate of 0,02 m:em/hr, and 27 gamma event;/sec/cm3 of ribbon
wake.. |

These 1evels,bf radioactivity are not believed to be objectionable
from & biological point of view, when additional attenuation of dose rate
by diffusion and decay are considered. Nevertheless, a radigactive trail
would be quite objectionable, These trails could be readilj.identified
several hours after the vessel had passed. |

Atmospheric disposél has ‘been considered, but ls not considered

feasible due to biological arrangement and shielding considerations.

11.7 Fuel Loading

Tnitial core loading presents a rather difficult task in that the
primary fluoride salt melts at a relatively high temperature. By some
means, the reactor must be maintained at a temperature in the order of
1000°F prior to the fuel loading. A stepwise method of loading is suggested

below.

 
-183-

(1) Steam generating equipmeht, loop piping, and the reactor would
e igsothermally brought up in temperature to approximately 6500F, Heat
may be supplied by steam from the conventional oil fired boilefs, by
electrical heaters, or & combination of both, A meximum of 1000 KW of
electricity is available for this purpose from each of the engine rooms,

(2) The intermediate salt is then introduced into the system at
650°F and circulated throughout the secondary loops.

(3) The secondary system is brought up to approximately 9000F by
circulating superheated steam through both the steam generator and super-
hester, The reactor will likewise be brought up to this temperature by
circulating helium through the system and extracting heat from the primary
heat exchangers. |

(4) By either overtemperaturing the conventional steam system to
provide superheated steam at llOOOF or by using electrical heaters in
the secondary loops the complete system is brought up to 1050°F;

(5) A stripped fuel mixture (no uranium) at 1050°F is then introduced
into ‘the reactor and c;rculated by the fuel pumps using the auxiliary
electric drives.

(6) Fuel concentration is gradually increased by adding solid Na,UFg
until criticality is reached at lOSOOF.

(7) Additional uranium is added to bring the reactor temperature up
0 approximately 1100°F. Simultaneously the steam temperatfire feeding the
steam generators and superheaters 1ls decreased to the normal 9500F. It
electrical heating was uséd entirely, this steam would be applied to reduce

the temperature of the steam equipment.

 
~184 -

(8) The superheated steam would then be gradually reduced to a saturated
value of 572°F by mixing with saturated steam, By the use of the blenders
and pumps the separate loops to the superheater and steam generator would
then be adjusted to near their normal opersting temperature.

(9) Preheated feed water is now added to the steam generator and normal
operation of the system is established.

(10) Uranium concentrate 1s agein added to the reactor until the design
operating temperature os 1225°F is reached.

(11) The system is now operational and a load may now be applied by
withdrawing steam from the generator and superheater.

The entire procedure must be accomplished at a low rate of heating

to reduce thermal shocks. It is estimated that it would take several days

to accomplish this.

11.8 Pumps, Valves and Blenders
111.8.1 Pumps

The cholce of pumps for molten salt systems is limited to
gas sealed pumps. Electro-magnetic pumps are nofi.effective with fused salt.
Canned rotor pumps depend on lubricafion by the pumped fluid and are not at
present sultable for operation at 1200°F in a fused salt medium, Gas sealed
centrifugal pumps have been oPerated at ORNL for:durations up to 8000 hours
at 1200°F without bearing; seal, or other pump maintenance. The operation
of such pumps is now considered to be routine and trouble-free (Ref. 5).
These pumps may be so designed as fo accelerate removal of xenon and krypton
gas fission products into the expansion chamber vold,

Helium gas seals shaft mechanical geals and supplies pressure to the

 
-185-

expansion chamber so that inlet pump pressure is never below 15 psia. This
firessfire is required so that cavitation of the impeller is prevented,

Provisions are made for complete replacement aboard ship for both the
primary and secondary pumps. Hofievér, reactor shutdowns and use of remote
handling equipment are required., The pump drive motors and assemblies
will be located above the secondary shield to allow direct maintenance and
replacement.

All surfaces of the pumps in contact fiitb either the fuel or secondaxry
fluid will be made out of Inconel,
Fuel Puups

Three pumps powered by steam turbines will circulate the fuel in the
core and primary loop. These pumps inlet and exit to common plenum chambers
so that reactor operation under emergency conditions is possible with one
or two pumps. These pumps were designed to operate at 2/3 power under
normal conditions thereby éllowing almost full power operation if one pump

is lost, Pump specifications are based on calculations by ORNL personnel

(49,50):
Inlet diameter 8.8"
Hut to tip ratio 1./h
ITmpeller outside diameter 13.76"
Discharge height "
Height of volute | 8"
Mex. diameter of volute 18"
Punping head 60 ft.
R,P.M. 1150

Pump efficiency 85%
 

-186-

Pumping horsepower 50
Input horsepower per turbine 85
Discharge rate ' 3333 gallons per minute

Pump shaft will pierce shielding material and will be sealed by a
positive helium pressure. As xenon and krypton gas pressure builds up in
the reactor, a regulator valve insures that this positive differential
helium pressure is waintained to seal pump bushings and prevent‘radioactive
gas from leaking from the reactor. A 10 H.,P, electric motcr'wiil be
clutched to the same shaft as ‘the turbine to provide shutdown and emergency
circulation.

The direction of fuel circulation throfigfi the reactor is opposite
to that found in the ART (Ref., 36). However, because the maximum temperature
found in this system was significantly less than the AfiT, the design life
of the impeller is much more than adequate and the advantage of having
natural and forced circulation directions the same is realized.

.Secondary Loop Pumps

 

Two secondary loops are anticipated.  Each loop will have two pumps
powered by steam tfirbines. One pump which will circulate hot salt through
primary heat ekchahger and through superheaters will require 270 horsepower
input (160 pumping H.P.,) and will have an idling 25 H.P. motor on shaft
for sh;tdown circula;ion. The other pump will drive a fluid circult at a
lower temperature and connect fhrough blenders with the other citcuit; pro-
viding lower temperature fluid to the steam generator. This circuit will
require an input horsepower of 190 and punping horsepower of 115, A 20 H.P.

electric motor will 1dle on same shaft for standby use.
 

_187_

These two pumps effectively operate in series so that in general
stability would not be as severe a problem as under~-parallel operatiog.
However, because each pump contains an expansion tank, it is possible for
fluctuation in level between the two tanks to form a different sort of
stabllity problem. This can be eliminated by locating the two pumps in close
proximity so that either a short flow channel can connect the expansion
chambers or a common chamber is used, An alternate approach would be to
utilize two pump impellers on a single shaft along with a single expansion
chamber, Howe#er, because of the large overhang, it may be necessary to
use a hydraulic bearing fed with pressurized fuel as an end support.

11.8.2 Valves

(1) Dump Valve will be ball and socket type, located in
lowest part of fuel cfiamber. Upon opening the dump valve, fuel will flow
by graéity, aided by 20 - 50 pounds pressure in the reactor pressure vessel,
_into drain ténks. Stem would be Inconel, ball and seat belng faced with
Kentanium, a modified titanium carbide nickel cermet, Tests have been made
with this type valve, against up to 100 psig helium pressure during and
after many hours at temperatures up to lhOOOF, with satisfactory results;
Tests have been made withqthe valve in Fuel 30 for 2285 hours, with 32
open-shut cycles at a temperature of 1225°F, with a pressure differential
across the seat of 50 psi. Satisfactory results were obtained (See Ref,* 36),

A drain valve of similar construction will also enable emptying of
secondary fluid system.

(2) Flow Regulating Valves in the superheater and steam generstor loops

will be constructed of Inconel. It 1s believed that a gate valve or coaxial

 
-188-

cone valve will emable flow regulation without undfie pressure drop. 1t
will be necessary that the valves be designed éo as tb permit satisfactory
clearance and operation at the design temperature, Since only flow fegulation
is required, absolute shutoff is not necessary. Therefore, no difficulty
is anticipated in the design of satisfactory flow regulatory valves,
11.8.3 Blenders |

The blenders or mixers used to interconnect the secondary
£luid loops would be of a single "Y" type construction with the two legs to
be mixed feeding ihto'a single conduit. Because the temperatures of the.
filuids to be mixed did not differ by more than a few hundred degrees,
it was felt that & plénum or mixing chamber would not be required. However,

thermal sleeves would have to be used on the legs to reduce thermal stressges.
-189-

12,0 MODIFIED APPRCACH

A_detailed weight per shaft horsepower list of the components of
this reactor system shows that an appreciable fraction of the total
weight appears as secondary shielding. (See Section 13). A cursory
examination was made of the use of an intermediate heat exchanger and
a tertiary £luid to boil water and superheat the steam. It is believed
that the weight of the secondary shielding can be reduced from 1.}
1bs/SHP to a specific weight of below 5 1bs/SHP., (This includes addltional
punp weight, heat exchanger weight and the additional power required to
drive the intermediate circuit.)

The use of the tertiary fluid will eliminate the need for shielding
around the bulky superheater and steam generator and will greatly
facilitate their maintenance. The number of pumps handling radiocactive
liquids will also be reduced. Using a machinery arrangement much like
that of the basic design and using the same temperatures, power, and
Plows everywhere in the system except the salt inlet temperature and hence
the log mean temperature in the superheater, the calculations described
below indicate the weight of the secondary shield can be cut considerably.
However, since the basic design used the secondary shield to complement
the primary shielding (fiote bulkhead shield, Figure 9.2), the weight of
the primary shield will increase. Estimated increase is from 6.4 1lbs/SHP
to 13 1bs/SHP. Because of the reduced mean temperature difference in the
superheater, its weight will also increase; however, this represents only

a small percentage of the total weight.
 

 

~190-

The complete bank of intermediate heat exchangers is placed directly
forward of the primary reactor shield. It would rest ageinst the fuel oil
tanks which would be used as part of the secondary shield for one side of
the heat exchangers. The top of ‘the heat exchangers is covered with 1"
steel and 6" lead. The port and starboard sides are shielded with 1" steel
and 7" lead. In addition, 1" steel and 3" lead are used against the
primary shield and the fuel oil tanks.

The following conditions were used for the weight analysis,

Heat exchanger U tube, counterflow
Tube Size: outside diameter - 0.25"

inside diameter 0,17"
Tube length ' 16,7
Number of tube 3600
Nuwber of tube bundles ‘ 12
Dimensions of shells 6" x 15" x 100"
Temperature difference, tube side 100°F
Tempera ture difference, shell side | 100°F
Temperature difference, log mean | 100°F
Inlet temperature, tube side 1150°F

All structural materials are Inconel.

As indicated above, the heat exchangers were calculated using what
appeared to be realistic conditions. A further study with optimization of
the fluid horsepower required for the intermediate heat exchanger versus
the weight of the entire unit should result in additional improvements.

The heat transfer calculational methods are identical with those shown

in Appendixes 6.1 and 7,1, Weight results are shown in Section 13,

 

 
-191-

13.0 SPECIFIC WEIGHTS

Below are tabulated the specific weight breakdowns for the conventional
oil-fired syStem and the three reactor powefed systems considered in this
report. The categorization is that used by the Naval Reactors Branch,

Each category includes the following itenms:

A+ B . Steam Propulsion Machinery
C+D B Reactor Plant Machinery

E Radiation Shielding

F Electric Piant (In Machinery Space)

G Electric Plant (Out of Méchinery Space )
H+ d Independent Systeums

L Load and'Stofes

Thé system referred 1o as "Basic Design" is that in which an attempt
was made to utilize only presehtly available technfilogy. Also the steam
generating equipment is contained within the secondary shield, For the
"Modified" design an intermediate heat exchanger loop was incorporated %o
permit removing the steam generation equipmént from within the secondary
shield, In the "Potentlal" design study, materials and concepts of a more
advanced, but still technically feasible, nature were utilized. Also, for
this design, the ship was presumed to be entirely nuclear powered,

The basic and modified designs are based on a reactof and steam generation
system overdesign of 30%, while the pq?ential design 1s based on an overdesign
of only 10%, Since time limitations did not permit a reiteration for a
more realistic overdesign modified systems of 10%, an estimation of the

~ bower plant specific weights for an overdesign of 10%.were made as follows.
SPECIFIC WEIGHT (LB/SHP)

-192-

 

 

 

 

 

 

 

Category Conventional Basic Modified Potentisl
A+B 19.6 17.2 17.2 17.2
C+ D. 8.0 12,3 14,0 11.h4
E . 0.6 20.8 14.5 12,6
F4+C 5.6 6.0 6.0 6.0
H4 J 1.2 5.2 5.2 5.2
L 1.2 2.0 2.0 2.0
Fuel 0il 23,4 0.0 0.0 0.0
Total 59.0 63.5 544

58.9

 

 

 
~193 -

The design of the reactor and steam generation equipment was agsumed to
remain the Same, but the capacity of the remaining equipment was increased
sufficiently to give the appropriate ovefdesign value, The results of this work
gives a specific weight of 58.1 and 54.2 1b/shp for the basic and modified
reactor systems, respectively. As will be mentioned later in this section,
a fair comparlson of the three systems requires that the shielding weight
of the basic and modified designs be increased to make up for the shielding
done by the fuel oil. (See Section 9).

The tabulated specific weights for the conventiqnal, oll-fired system
were taken from a detailed ship weight breakdown compiled by the Bureau
of Ships for a DD931 destroyer. For the three reactor powered systems,
the equipmenfi weights not affected by #he reactor installation were also
taken from this table.

For‘the three reactor systems, the specific weight of the steam
propulsion machinery (Category A + B) is several lb/shp less than for
the convent;onal systen. This is true primarily because approximately
half the liquids in the reactor systems are accounted for under reactor
plant maéhinery (Category C+D), whereas for the conventional system all
liquids are accounted for under Category A+ B. Another factor which con-
trivutes to this lower value is the removal of forced draft fans and fuel
oll pumps. Finally, a portion of the insuiation, which for the conventional
system is totally accounted for under Cafegory A + B, has been included,
for the reactor systems, under Category C +D. For the conventional system,
all components which would be removed to make:way for the reactor plant
machinery were included in Category C+ D, For the reactor powvered designs,

the weight of the steam generation equipment was increased quite substantially
19k -

(25% of calculated weight) to account for the supporting structure and
other portions of this equipment for which no detailed weight calculations
were made., From examination of other reactor powered steam systems, 1t
was decided the two pounds per shaft horse power would be ample to cover
the weight of the control system and miscellaneous items, The specific
welght of the modified system's reactor plant 1ls somewhat greater than the
basic due to the additional.coolant loop and assoclated heat exchanger,

A slightly smaller specific weight over that of the modified system is
reglized fog the potential design since the application of more optimistic
concepts permits using a smaller réactor and intermediate heat exchanger
(Section 14.0). |

The gpecific weight of the electric plant for the reactor powered
systems was Increased slightly over the conventional system weight., This
increase was adjudged sufficient to provide for the electrical components
of the reactor control systenm.

A substantial increase in specific weight is indigated for independent
systems'(Category H+J), From a cursorj examination, it is apparent that
the machinexry ;equired to replace a reactor fuel pump while at sea will
weigh about two pounds per shaft horsepower. The weight of the offgas
system, fuel adding mechanism, and miscellaneous items, was estimated at
two pounds per shaft horsepower, also,

For Arrangement No. 1 of the reactor comparitment for the basic and
- modified reactor powered systems 1.7 1b/shp of fuel oll is carried in excess
of one-half the fuel originally on board the conventionally powered ship,

This fuel is nééessary'for trim of the vessel and also serves as shielding.
-185-

For the potential design, the ship is to be powered entirely by reactors
and therefore no fuel oil is carried.

Arrangement Wo. 2 of the reactor compartment would provide a trimmed
ship with the removal of enough fuel to provide a specific weight savings
of four pounds per shaft horsepower less than half the original amount of
fuel, Arrangement No. 2 is the one recommended in this report. There
is some question as to the advisability of providing the o0il-fired portion
of the power plant with less than its normal complement of fuel. For
this réason, the tabulated specific weights do not show a reduction for
fuel oil savings.

If the potential system were installed in conjunction with an oil-
fired boiler, the fuel oil inventory could be utilized as shielding, giving
a more favorable overall weight for this design.

If either the "basic" or "modified" design nuclear systems were used
for total ship power, instead of only half, an additional ten pounds per
shaft horsepower of shielding will be required. This is due to the fact
that the fuel oil carried for the oil-fired boller is placed in such =
way as to be equivalent fo approximately ten pounds per shaft horsepower
of shield. The overall power plant specific weights for an entirely nuclear
powered ship with a reactor and steam generation gystem overdesign of 10%
then becomes:

Basic Design Modified Design Potential Design

68,1 1b/shp 64,2 1b/shp - 54,4 1b/shp.
 

-1G6-

14,0 FUTURE POTENTTAL

The design philosophy taken on'the basic reactor system is %o use,
as much as posaible, materials and technology which ha#e been proven
feasible., It seemed desirable to illustrate the potential of the systen
by adopting materials and conditions which were more optimistic, but yet
feasible, as indicated by experiments and qualified opinions, Thus, =
cursory examination of a more advanced design was made to determine minimum
realistic specific wéight (weight/shatt horsepower) that may be achieved,
Also, since time had not allowed the optimization of many parameters in
the basi¢ design, an attempt was made to. select parameters which would

improve the performance of the systen.

14,1 Reactor Core

In order to help reduce the total weight of the system, it is
desirable to minimize the reactor core diameter. Four steps are taken
to achieve this. (1) Zirconium hydride was used as the moderator replacing
beryllium oxide thus increasing the moderation properties 6f the core,
(2) A beryllium, sodium, wranium fluoride salt vas selected to replace
the zirconium, sodium, uranium salt to improve neutron moderation. (3)
Because of improved materials a higher power density could be used. (k)
Use of a nickel-molybdenum cladding such as INOR-8 (Ref. 5) decreased the
poison in the core because nickel-molybdenum's cor:osion resistance pexrmits
a thimner cladding.

The geometrical afipearancé remains identical to that of the bagic design;

however, all dimensions are reduced. Preiiminary calculations indicate that

 

 
_]_97...

criticality.will occur at a uranium concentration which is approximately
that of salt mixture number 92 tested at ORNL (Reference 40). The control
rod is identical with that described in the basic design both in size

and materials., Conditions imposed on the. system sre:

Power 100 Mw
Power density (averaged over core) 1 Kw/cm3
Fuel (Approximate) | 38% NaF, 429 BeF,, 20% UF
2 b
(% by welght)
Moderator rod diameter, inches 0.5
Moderstor rod cladding thickness, inches 0.010 Mo

0.020 INOR-8

Volume fraction of fuel 0.5
Core diameter, cm 40
Nickel reflector outer diameter, cm 80
Core height, cm _ 78
Average fuel inlet temperature, OF 1100
Average fuel outlet temperature, °F 1300

One hundred megawatts was selected as the power necessary (with 10%
overdesign) to drive oné epgine room of a 931 class destroyer, The power
density 1is felt to be safe as;a result of an ORNL study (Reference 36) and
preliminary moderator rod stress calculations. The core temperature seems
to be modest from a corrosion standpoint., The use of INOR-8 will decrease
corrosion; however, lts fabrication is more difficult than that of Inconel.
The zirconium hydride moderator rods coupled with the improved moderation
properties of the beryllium fuel allow the size of the reactor to be reduced

conslderably, However, zirconium hydride goes through a phase change near
 

 

-198-

1100°F and thus cladding gstress may be a difficult problem, Hydrogen
escape from the moderator is resisted by the use of a molybdenum cladding
next to the zirconium hydride which’'in turn is clad with INOR-8 for

corrosion resistance,

14,2 Primary Heat Exchanger

Because of the relatively poor heat transfer properties of the
secondary coolant used in the basic design, 1t was felt £hat, although
initially a decision was made not to use sodium (See Section 2.3), it
should be investigated. The primary heat exchanger is placed Jjust outside
the poisoned shield rods just as in the basic design. It is é U-tube
design with the tube entering and leaving toroidal headers placed around
the top of the reactor. A baffle sheet is placed in the fold of the tubes
to effect counterflow_at all points.

The following is a list of the design conditions which were considered:

Tube diameter, outside | 0.55 inches
‘ Tube dismeter, ingider 0.50 inches

Number of tubes 2040

Tube length 7.5 feet

Fuel inlet temperature 1300°F

Fuel outlet temperature 1100°F

Na inlet temperature - 930 F

Na outlet temperature | 11300F

gtructural Material INOR-8

Although sodium is a very good heat transfer agent, the heat exchanger

considered in this study required as large a radial dimension as did the
~199-

straight through heat exchanger in the basic design because additional space
is required for a U-tube configuration. It was felt that it is desirable to
use U-tube design to avoid thermal stress and thus effect longer life,
Thermal stress becomes a more important problem when using sodium than the
fused salt as in the basic deslgn because, contrary to the case of the fused
salt, a larger portion of the temperature drop occurs in the tube wall,
Helically wound tubes are a definite possibility for smaller heat exchanger
volumes; however, thermal cycling causes a tricky configufation design pro-
blem especially when it is desirable to obtain a relatively long life.

The U-tube sodium heat exchanger required many less tubes and thus is
less expensive and easier to fabricate than the straight through salt heat

exchanger considered in the basic design.

14,3 Intermediaste Heat Exchanger

 

Because the core is sa high flux, high leakage wachine, there is con-
giderable neutron activation of the sodium., This activation causes a
secondary shielding problem which can best be minimized by the use of
intermediate heat exchanger (See Secti;n 12.0). For the afivanced design,
the intermediate heat exchanger is a counterflow U-tube sodium heat exchanger.
The shielding required iO in. of lead and 1 in. of iron on the top, port,
starboard, and forward sides. The primary shield was used on the aft face,

Heat exchanger parameters are

Na outlet temperature (tube side) 930°F
Na inlet temperature (tube side) 1130°F
Na outlet temperature (shell side) . 10300F

Na inlet temperature (shell side) | 830°F
 

 

-200-

fube outside diameter | 0.425 in,
Tube inéide diameter 0.375 in.
Number of tube bundles / 25
Number of tubes per bundle - 80
Dimension of shell (1 tube bundle) | 11.9" x 5" x 50"

1h,4 Boiler and Superheater

 

Because the boiler and superheafier are relatively small percentage
of the tdtal machifiery.weight, no attempt was made to design thié equipment,
Sodium has been used in thé Seawolfv(SiR) reactor system for steam generation
and will be used in the sodifim_graphite reactor (SGR) developed by Combustion
Engineering and Atomics International. The fihermal stress and chloride stress
corrosion on the water siflé éreafies an engineering problem, However, it is
felt that the use of Inconel as a structural mateiial and blenders o reduce
temperature'differénces fiay be a partial solution., For the weight of the
steam and water equipment, the same weights as were determined in the
basic design were listed. Because an ihtermediate fluid will probably
be used between the water and the.sodium, the disadvantage of the third
fluid will tend to counterbalance a reduction in weight caused by the

superior heat transfer properties of sodium.

1%.5 Primary Shield

The primary shield which was considered consisted of thé following.
One inch of structural steei plus six inches of lead were used just outside
of the insulation packed around the pressure shell. Next followed 15.7
inches of water and then one inch of steel and 6 inches of lead. Following

this lead is TO inches of water which is contained by a 1/2" steel vessel.

 
-201-

Primary shield weight is approximately 375,000 1bs. Dose rates at the
reactor shield face were approximately 10 mr/hr and 10 fast neutrons/cm2
sec. The most important gamma contributors were the water and lead capture

gammas and the fission product gammas released in the primary heat exchangers.

14,6 Caloulational Methods and Results
Calculational methods used in the advanced design are identical to
those illustrated in Appendices 6.1, 7.l and 8.2, The specific weight

results are given in Section 13.0.
 

-202-

 

- APPENDIX 5.1

Inconel Data

Composition | Wt Gms/cm3 Atoms /cmd
Nickel | . 79.5 | 6.405 6,57 x 10°°
Copper | e 016 016 x 10°2
Iron 65 .530 572 x 10°2
Manganese -.é5 - ,020 .02k x 1022
Silicon . - .25 020 - Okl x 10°°
Carbofl | .08 | . 0065 033 x 1022
Chromium | 13.0 1,060 1,229 x 10°°

Density at 1200°F - 8.156 Gm/cm3.

See graphs for data on'tfiermal condfictivity,_tensile strength, elongation,
yield.strength, wodulus of elasticity, coefficient of thermal expansion,
and hardness, Figures A-5.1, 2, and 3,

Beryllium Oxide

 

Theoretical Density  3.025 Gm/Cums

Density (965 theo,)0°C 2,904

Vol, Coeff, of Expansion 2.h3 x 10"5Iper °¢ approx.

Density at 1500°F 2,88 Gm/Cm3

Composition | Atdms{Gm EEZQEE Atoms/0m3
Be | 2.41 x 10°° 1.028 6.95 x 10°°
0 o x 10%2 1,852 6.95 x 10°°
-203 -

Modulus of Elasticity
o
at 68 F - 45
. -
at IW70F - 28 x 107 psi
o 5
at 2550 F - 12
Thermal Conductivity (see graph)

Specific Heat

32% - - - - - - 0.219
212%F - = - - - - 0.308
752°F - = = - - 0.420

472°F - - - - - - 0,492

Poisson's Ratio (up to 1800°F) ~ - - 0.35

Tengile Strength

TT = 750°F = = = = = = = = = - 15,000 psi
IYTOF = = = = = = - - 13,500
1830F = = = = w m - - e m 10,500
2010°F - = = = - = = = - - 8,000

References (10) and (20).

Density.(20°C) = =« = = = = 8,91 Gms/Cm3 6

Lin, Coeff. of Thermal Expansion - 7.4 x 10° per °F

Thermal Conductivity
200°F = = =~ @ = = = - 42 BTU/Hr-Ft-CF
1100%F = = = = = = - - 21 " -
1600°F - = = = = = « - 15 "o

Specific Heat = = = = = - - 0.11 Cal/Cm

Density at 1500°F - - - - - 8.88 Cms/Cu3

Melting Point - - = = - - < 2650°F

Atoms per Cmg at 1500°F - - 9.11 x 10°@
 

 

 

 

 

 

 

ORN L=LR=Dwg, =25772
UNCIASSIFIED

 

 

 

 

 

 

 

~20k4 ~

 

 

 

 

 

 

 

 

 

 

 

 

 

4o JHNALVYILW3IL
00+! 00£!l 0021 0Ol 000I 006 008 00L 009 00S OOt 0OOE 002 OOl O o
— T3NOONI NMVYHA 0109 oot 2
/ / z
7 o2 3
>
\ o
oyt -
L o
...l..l....l.l.ll.llll-l.....lll._ . -
13INOONI QIVINNY [ — &
o8l 4
X
n
3
i S
ogl 0’9 m—
......r/ \ x N
T~ % Obl = — 02 O —
a37704 po: f,/ = R
m m
ss INQuvH TTINIEE — og 08 X—
< > >
] 091 3 ——0'6~N—
=] < 0Ll »——0"0} > —
NOISNVdX3 40 LN310144309 @ o,
08l Ol ~»
-n

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Viva N9IS30d T3INOOINI

 

-GSV JHNOIL
1

-205 -

ORNL-IR“DWE . "2 57 ?3
UNCLASSIFIED

FIGURE A-5-2

 

| | :
INCONEL DESIGN DATA

 

 

 

 

 

 

 

 

0.1é

I

|

CIFIC HEAT BTU/# °F
o
>

.15

T
tSF’E
o

o

F-FT

 

 

pd

 

 

 

)
»
o

THERMAL

CONDUCTIVITY

P

/

 

/

 

o
o

/<_SP£ CIFIC
HEAT

e

 

o
o

v

 

 

@
o

//
/
L~

 

»
o

A

THERMAL GONDUCTIVITY BTU/HR-

 

 

 

>
2

400

800

1200 1600
TEMPERATURE, °F

2000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
-206~

ORVL~LR=Dwg. =2577h

UNCLASSIFIED

 

FELONGATION (% IN 2")

TENSILE MODULUS (10° PSI)

 

120

100

60

TENSILE STRENGTH (103 PSI)

4of

 

30

YIELD STRENGTH (103 pPsl)

257t

20}

10t

 

FIGURE A-5.3

INCONEL DESIGN DATA

 

™~

 

]

 

X/TENSH.E

STRENGTH /

 

N\
\

/

 

\

/

 

\

 

 

 

 

ELONGATION
N
— W
—~. S

 

 

 

 

 

YIELD STRENGTH
N (0.2% OFFSET)

 

 

\\\

 

 

 

 

 

. MODULUS/'/)\

N\

 

“

 

 

 

 

 

 

" -
ELASTICITY
N
000 1200 1400 6 00 (1800

TEMPERATURE (°F)

 

 
-207-

APPENDIX 6,1

JUSTIFICATION OF MODERATOR MATERIAL

 

Allowable Moderator Rod Size

 

The best way to justify a material selection is by 1ts satisfactory
performance under actual operating conditions. Beryllium-oxide has
suitably'withstood a preliminary evaluvation in the MIR under both high
temperature radiation énd cyclic operatiofi (Reference 54 and 55). While
the specimen. size and test conditions were not idenfiical to that proposed
for this study it is é reasonable first approach to calculate the critical
stresses that were withstood in this test and then apply these %o the
current case.

The lifiiting stress in the case of internal heat generation, especially
for a ceramic, is its tensile,

The specimens from the MIR test were 1l in, in diametef and were
satisfactorily exposed to & power generation of approximately 15 Watts/cm3
at a temperature of 1500°F, Utilizing the development given in Ref. 10,
Bquation LVIII, on the working curves of Ref. 3, the maximum tensile stress
is found to be 3000 psi. This will now be used as.the deslgn basis for
the selection of a maximum allowable moderator size for this study.

The energy release rate is approximately 197 mev/fission which may
be broken up roughly as follows:

Local deposition - 165 mev/fission - Fission product kinetic energy

5 mev/fission - Fission product decay beta particles
-208-

Non-local - 6 mev/fission_ ~ Fisgion product decay gaumas
5 mev/figsion - Prompt fission gammas
5 mev/fission - Prompt fission neutrons

11 mev/fission

To determine the energy deposition rate in the woderator, it is felt

quite conservative to assume that all of the neutron energy is absorbed in

the moderator,

On the basis of this it was felt juStifiable to increase the moderator

rod size to 0.75 inches diameter although additional testing would be
required to obtain definite verification,

Moderator Rod Temperature Digtribution -

By making the reasonably valid sssumptions of steady state, negligible

axial heat flow, and non-variance of the materials thermal properties with

temperature, the teuperature distribution through'a clad cylinder due to

internal heat generation can be found from the basic equation:

oo gtk = b (e )

r

By applying the proper interface and boundary conditions this reduces %o:

In the cladding - r, = r LT,

2
‘bc(l‘) ’b(rc)-}-. ——-——nc—-—-—— 1 = (....?r__) (q LYy _

C

r

T

 

Neutrinos (deposited at o)

qcili)

1
~209-

On the moderator - 0&1r < r,

m

b1t p 2
6 (r) = t(x,) +_f—%—m—— [1 - (.%.)2:‘

where: the subscripts ¢ and m refer to the cladding and moderator respectively

t(r) = temperature at point r

g''' = heat generation rate per unit volume
k = thermal conductivity

Ty = outer radius of the moderator

r_ = outer radius of the cladding

Applying these results to the case of a 0,75 in diameter cylindrical
rod of BeO with %0 mil Inconel cladding (k = 14,5 and 16 BTU/hr-ft-°F at
approximately 1500°F) the temperature distribution given on Figure A6.1
was obtained, It should be noted that a temperature drop across the interface
6f the moderator and cladding is not included at this point.and that the other
non-local energies were lost from the system, The possible gamma energy
absorption was not included because the relatively light welght BeO gives
poor gamme attenuation making this effect within the conservation of the
previous statement.

Maximum energy deposition in the moderator is then 2.% (l =55 ) of
that in the fuel, Considering the power distribution fofind acrogs the core
of 1.4 (Section 8.2.2) and a probable axial distribution. of the same order

gives an overall peak to average power of approximately 2.0, The maximum

heat generation in the moderator rods can then be calculated by:
~210-

Power density in the fuel = 700 watts/cm3
Fraction fuel volume = .50

Fraction moderator volume = 4O
Genefation rate in BeO = 2,.9% x 2.0 x 700 x L%gu = 50Iwatts/cm3

?he critical location for the moderator'materialois approximately at the
central region where the powef'density is & mgximum. Although the tem-
peraturelof_the fuel increases toward‘the exit of the core, the power
density decreases at a ra#e stfficient to cause a net reduction in the
moderatorltémperatfire.i It is estimateé later in this appendix that the
waximum surface temperature (location of maximum tensile stress) of the
moderator is 1410°F, C;omparingvthis with the MTR test information at 1500°F
and the basic strength characferistics of BeO_With temperature (Ref. 9),
it is appareht that‘no correction in the allowable stress should be made.

Utilizing the allowable stresé of 3000 psi and a maximum uniform heat
generation fafie of 50 watts/cmB, the limiting-cylindrical rod diameter based
on test data is calculated to be 0,6 inches (Reference 3).

- However, the folldwing considerations should be made before the rod
size is limited to this value:
| 1. The above calculations were fel£ to be conservative,

2. Actual tensile strengths for BeO of 9000 psi were measured (Ref. 9)

3. MIR tests which ran the ténsile strengths to 3000 psi gave

éatiSfactory performance,

Applying these results to the care of a 0,75 in. diameter cylindrical
rod of BeO with 40 mil Inconel cladding (k = 1L4.5 and 16 BTU/hr-ft- F
at approximately 15000F) the temperature distribution givefi on figure A6.,1

was obtained. It should be noted that a temperature drop across the inter-
-211-

face of the moderator and cladding is not included at this point,

Pemperature Rise Across Fuel Boundary layex
The temperature rise across'fhe bcfindary layer between the fuel and
the moderaior-rods waé caléulated by means of Reff lland 2, It is shown that
the overall temperature rise is geparable into the sum of two temperature
differences (l) Due to the temperatuxe drop required to remove the heat
generated within the moderator material and (2) Due to the temperature rise
through the boundary layer duse to decreased velocity and thus higher power
density. |
The followiné constants were found for.the reactor core operating at

125 MH with a temperature rise of 100°F and a mean temperature of 1200°F,

Flow Area | - =2,38 ft?
Hydraulic Diameter - 0546 £t
Velocity _ -g.1 ft/sec,
Reynolds Number | =2 X 10lL

Prandtl Number " =3.6

fuel Thermal Conductivity = 1.3 BIU/hr-ft-OF
Using Referxence 1, with an equivalent cylinder gives a Nugsault numbe$
of 100 for the (1) solution., (Use of the hydraulic diemeter analogy with
this method is indicated in Ref. 66).

Nu = 100= hd _ a
= - 9

KAt T Tk
For a rod of .830 inches (.75 + .08 cladding)

a/A = g, x Volume Assuming heat generation rate in

— moderator and cladding are equal.
-212-

at q"’m v d which upon substitution reduces to
U8 T T
100A '
At =701 q"'m where g'" = moderator heat generation rate in

vatts/cm3
Total temperature rise across the boundary layer at the center of the

core 1s then

AT = 701 q"& + .01l q;‘
q"s = 700 x 2.0 watts/cm> where 2.0 is the peak to average power
q"! = 50 Watts/me as discussed previously
then:
.gfiT = 5OOF — temperature increase of outside cladding above uwean

fuel temperature.

Temperature Rise Across Cladding - Moderafor Interface
Becauge this interxrface gap 18 quite small compared to the rod diameter
even at operating tewperatures (Sec. 6.2.1) it may be treated quite accurately

by an approximation to a flat plate as:

q/A = k AT
1
q/A _ qn!m Volume _ qll!:;:i a _ .750 qil!m

Area T
= 1510 ¢"' where q"' = wafits/cm3
Because & shrink fit of the cladding around the BeO appears 1o be

feasible a maximum of one mil clearance should be realistic. If the shrink

is made in a helium atmosphere k is estimated to be .1k BTU/hr-£4-CF (Ref, 20)
Ar _ 1510 @'y (,001) = M5°F for q"' . 50 watts/cmd
(.14) 12

 
~213-

 

Total Temperature

From Figure A-6,1 the temperature use through the cladding can be
estimated as

AT/q"' = .3k

AT — 17°F
Similarl& the temperature rise to the centerline of the rod (neglecting
the interface) is found o be: |

4T/q"£} ¥ 1.96

AT = 98°F

Combining the temperature rige across the boundary layer, cladding and
interface with an assumed maximum fuel temperature at the center plane of
1300°F {ave. temp.=§'1225°F) provides a maximum temperature at the BeO
éurface of 1412°F.

Similafly the maximum temperature at the moderator centerline is found

to be 1493°F.
 

ONIgavo dOLVHIQON
2
v7,, |
0’1l 860 $6°0 60 80 90 v 0O ¢ 0 0

//

-

_ / ol
IONVHD 31voS 310N — ] |

 

ORNL~LR=Dwg, »25775
UNCLASSIFIED

 

 

 

21k

O (3} -0y

 

 

 

 

¢Wo/ SLIVA
ALISN3Q ¥3MOd OSNIgavo =2 wib;

1w

= 2
gWo/ S1LIVM= ALISNIQ mob

R
43M0d HOLVY3IAOW = ,;

. 02
«SIP =SNIAYY ¥3LNO = 2

2 "b/1V

 

[

 

 

 

 

 

 

 

dvid T3NOONI ObO .
038 "ViQd ‘NI 06/

 

 

NOILNEGIYLSIO 3JYNLVHI4WIL QOH MNOLVNIQOW

 

 

I'9-V  2dn9id

 

 

 
-215-

APPENDIX 6.2

CALCULATIONS FOR FINAL DESIGN OF PRIMARY HEAT EXCHANGER

6.2,1 Basis of Design

The primary heat exchanger is designed to transfer 125 megawatts

or 4,27 x 108

Btu/Hr from the fluld fuel to the secondary coolant. Many
gquantities which would ordinerily be considered design parameters were, due 1o
the short time allowed for this study, given what seem to be reasonable
values and held invariant throuhgout the calculations.

The following calculations represent the final iteration of the

most promising combination of tube diameter and spacing as indlcated

by Figure 6.1,

6.2,2 Properties of Fuel, Secondary Coolant and Inconel

As given in Section 5.0.

- 6.2.3 Quantities Determined Before Employing Iterative Procedure

Heat Bxchanger Inner Diameter = 53.5 in.

Heat Exchanger Length = 48 in,
Tube Wall Thickness - .00 in.
Temperatures :
Fuel Entering = 1275°F
Fuel Ieaving - 1175°F
Coolant Entering == 1050°F

Coolant Leaving = 11500F
 

-216-

6.2.4 Quantities Determined by Iterative Process

Tube Outer Diameter = ,200 in,
Tube Spacing - 030 in,
Reactor Outer Diameter ~  T13.7 in,

Number of Tubes - k43,420

6.2.5 Flow Rates

| | | .
a, Fuel Loop Flow = .C A?T _ _h-QT X lgt Btu/Hr
x — -
P . 264 TH5-OF 100°F

_ 16.2 x 100 B
- R

 

 

 

4,27 x 108 ifi“
b, Coolant Loop Flow = —g g 7 - .
4 Biu o,
P 57 THE-OF 100°F

= T.49 X 106 LB
: Hr

6.2.6 Flow Velocities

a. TFor specific values of tube dismeter, tube spacing, and reactor
outer diameter, the approximate aumber of tubes contained in “the heat

exchanger are determined as follows:

 

 
-217-

 

 

 

let 4 = tube
0
outer

let S = SPace

thickness

diameter

& =

Dp - Dy

2

 

 

 

Number- of vertical rows of tubes

P-— [9:3 (do+s)} (D + Dy)

N =

Z

"3 (do+ 8)

Number of horizontal rows of tubes

Q:—Wfl—dofs.i- 1

Dy - D, - 58 - 3 do

 

X =

2

De - Dl - 5 = 3 do

+ 1

 

W:-_.

2{do +« s)

Total number of tubes

'W(DQ + D])} [

n = N??:

3 Ido-+ sj

D, = Dy - 58 - 3do

 

2 (do + s)

5

Y

 
~218-

w(Dy + Dy) .
TEVI @+ a2| (P2 Py -3 -0
Using the results of the final iteration,

nfi[wfi&7+5&wfi

2 Y3 (.200 + *'030)2jl 131 = 53.5 - 050 - .200

= 43,420 tubes

b. Coolant Velocity
} zc 7,49 x 10° 1b/nr _ .

¢ ffe E.23 1b/ft3] [43,11-20 "“1-&?24;)-080) Ft?l]
= 17,830 ft/hr = 1,95 ft/sec

¢, TFuel Velocity

 

v We 16.2 x 10° 1b/hr
toephe (208 w/e83][ I (73.7 - 53.5% - 43420 x .2%)
| 14 x 144

= 17,150 £t/hr = 4,76 f£t/sec

6.2.7 Fuel Side Hydraulic Diameter

d . 4 x cross sectional area
h = wetted perimeter

2
w [T5mr (7377 - 53.57 - 43h20 x .22]

 

“n
13~ (73.7+ 53.5 + 43,420 x ,2)

.00788 £t

I

 

 
-219-

6.2,8 Reynold's Numbers

a. TFuel Side Reynold's Number

5 _ %
of T
Fe

_ (208 1n/£47) (00788 £1) (17150 £1/nr)

 

18 1b/hr £t
= 1560
b. Secondary Coolant Side Reynold's Number
R - (Ocdivc
ec Mo

(123 lb/ft3)(;9%§9 £4) (17830 £t/hr)

53,2 1b/1t nr

 

= 412

6.2.9 Prandtl Numbers

a, Fuel Side

Pr_ - Cpf’flf _ (.264 BTU/1b °F)(18 1b/ft sec)
£ ki 1.3 BTU/hr-£4-CF

- 3.66

b, Coolant Side
-220~

 

6.2,10 Filw Coefficient

a. TFuel Side (See Figure 7.6)

0}+ -
o = W (prg)” (Re,) °

Q
- o G 6ot s

O
= 1836 BTU/hr-ft°- F
b. Coolant Side (Ref. 17, page 232)

. a.
hy = 1.62 XC (Re Pr __t )1/3
94 L

 

 

 

 

Ft 0 1
~ 1.60 (2:h BTgégr £8-F) (410 x 12.6 x 2220 ) /3
(55 Ft) e
o BTU
= 9 G ree-op

6.2.11 Overall Heat Transfer Coefficlent

 

 

 

L
U _ —
0= _
L+ - e in Yo
R~ In_ 2k L
_ 1
1 + .200 in

 

1836 BTU/hr-££°-°F  (,120 in)(915 BTU/hr-f+°-CF

.200 Ft) in '(,200 )
12 . 120

(2 x 14 BTU/hr-ft-"F)

 

 
~221-

1 ' BTU

 

> ~ 100182 + .000545 + .0030%) Hy-Fte-OF
- 37% —BIU
He-Ft2-"F

6.2.12 Total Heat Transfer Area

A=nwd L = 43420 x v x (.:..f...g_o £1) x (.;&.g. £t)

= 9090 Ft2

6.2.13 Mean Temperature Difference Required to Transfer 125 Megawatts of
- Heat

 

8
AT - 9 (b.27 x 10 BTU/Hr)

® AU (9090 F7)(37h BTU/Hr-Ft -OF )

= 125 4°F

6.2.14 Temperature Drops Across Surface Films and Tube Wall

a, Fuel Side Film
8

 

 

 

Ar . 9 _ b.27 x 30" BTU/Rr
o hA (1836 BTU/HreFt--°F) (9090 Ft°)
= 25.6°F
b. Coolant Side Film
d AT - 9 h.27x 10° BTU/Ar ,
| 1o A, (915 BTU/HEr-Ft2-OF) (2320 x 4 x 7 x 43420 Ft
‘ 12
= 85.5°F

c. Tube Wall

AT = 125.k - 25.6 - 85.5 = 14.3°%F
-222~

6.2.15 Approximate Average Surface Temperature of -Tube

a, Fuel BSide

+3
i

Inlet Surface Temperature + Exit Surface Temperature
_ (1275 - 25.6) + (1175 = 25.6)
: 2

= 1200°F
b. Coolant Side

Twc:rlnlet Surface Tewperature + Exit Surface Tempersture
- 2

_ (2050 + 85.5) -+ (1150 + 85.5)
2

1;8605

 

6.2,16 Approximate Average Tube Wall Temperature

+ T

 

0 Tye Y Tue 1200 + 1186
T e —
wa 2 2
= 1193°F

6.2,17 Friction Factors For Both Sides of Heat Exchanger and Reactor Core

a, Heat Exchanger, Fuel Side

F {Re ) (Ref. 13)

o 2

b. Heat Exchanger, Coolant Side

f

c (Ref. 15, page 50)

h
jo\
a|&=

: C
6h
= g5 = 155

 
6.2,18

223~

c¢. Reactor Core
f < .02k (Ref, 15, Figure 2-21b)
o
See also Appendix 6.2.18 for Reynolds Number in core

Pressure Drops for Both S8ides of Heat Exchanger and Reactor Core

a. Heat Exchanger, Fuel Side

 

 

- T e .
AP = fF o —F%) Ftc (Ref'. 15, page 45)
af 2(g 5 )
Bec

5

~ I (4.76)

= 0921 .00788)  2(32.2

- = 16,6 Ft

_ (208 1n/FE2)(16.6 Ft)

2
1l in_g
Ft .

= 24,0 psi

b. Heat Exchanger, Coolant Side

ap-+s B _(Ye)
F 4. 2g

1

i1

55 (b FE)  (4.95°7%°
120

2 x 32,2 Ft
5 Ft ey
sec
= 23,7 Ft
(123 Lb/FtB)(23,7 Ft)
- = 20.3 psi

 

(144 In® )
2
e
-22h -

¢. Reactor Core

 

 

 

 

' 2
1. Flow Area = A_ = .50 L ( 75 cm )
* 4 12 38 4 5 gy <2
Ft 7 In
2
= 2,38 Ft
2. Wetted Perimeter = P_= fi’I cm , 10cm
12 2 2,54 S0 cu
FE Mmoo Sk

+ 749 x .75 In) = 156 Ft.

See Bection 3.3 for dimensions of reactor core.

3., Hydraulic Diameter = dhr

- hox A,

 

 

 

- - 4 x 2,38
P 1
- 2
= 0610 Ft
' 6
“. Velocity . Wy _ _16.2 x 10 Iv/Hr
Cp * Ay 208 Ib/Ft> x 2.38 Ft°

= 32,700 Ft/Hr = 9.08 Ft/sec

5. Reynolds Number = Rer

_ Pfdhrvr (208 Lb/Ft3)(.0610 Ft) (32,700 Ft/Hr)
- }Jf o 18 LB/Ft-Hr

= 23,1
3,100 5
r r ( d ) 2g
: hr

Lo, 2
~ ook (3.9% Ft) (9.08 Ft7) ,
L0610 F% 2 x 32.2 Ft/Sec
6.2.19

~225-

3
= 2.0 Ft = 208 Lb/Ft2 x 2,0 F%
In
1k -
Ft
== 2.89 psi

Pumping Power Requirements for Both Sides of Heat Exchangexr and
Reactor Core :

a, Heat Exchanger, Fuel Side
b
] (W, Ib/Hr) (AP, F )

£ (1.98 x 10° %;%)

FHP

 

(Ref. 15, page 80)

6 Lb
(16.2 x 10° = )(16.6 Ft)

Ft-Lb
Hp-Hr

 

1.98 x 100

 

= 136 HP
b, Heat Exchanger, Coolant Side
prp o (1.9 X 10 Lb/Hr)(23.7 Ft)
1.98 x 10° Ft-Lb/Hp-Hr
= 96.0 HP

¢. Reactor Core

FHP = (16.2 X lOé Lb/Hr) (2.0 Ft)

¥ 1.98 x 10 Ft-Lb/Hp-Hr

= 16.3 HP
-226-

APPENDIX 7.1

STEAM GENERATING SYSTEM

i.- Heat-Transfer Calculations for Steam Generator
A, 'Incofiel Tube Data:
(1) size; 5/8 in, 0.D., 1/2 in, I.D.
(2)I Pitch; 3/4 in. delta array
(3) Thermal Conductivity = 11.3 Btu/hr-ft-oF
(h) - 8pecific heat = 0.124 Btu/lb;oF
(5) Density = 510 1b/ft3 |
B, Steafi Generator Inlet Conditiofis:
(1) Molten Salt:
(8) T = 761.8%F (95.9 MW); T = B0O°F (125 MW)
(b) w= T.49 x 105 1b/hr
(c) cp, 0.57 Btu/1b-"F
(8} p =130 cp (95.9 MW); p= 126 cp (125 MW)
(e) © = 127 1b/td
(£) k= 2.4 Btu/hr-£t-"F

(2) Water:

(a) T~ 564°F

(b) P - 1250 psia

(c) h = 567 Btu/1b

(d) w = 3,230,000 1b/hr (95.9 MW); w = 14,149,500 (125 MW)

(e) wvel, = 6.24 fé/sec (95.9 Mi); vel. = 8 ft/sec (125 MW)

(£) spec. vol. = 0.0221 ££3/1b
-227-

Steam Generator Outlet Conditions:
(1) Molten Salt:
(a) T = 703°F (95.9 M{); T = T24°F (125 MW)

6 lb‘/hr

(b} w=7.49 x 10

(c) 'cpa-_ 0.57 Btu/1b-°F

(d) p = 195 centipoises; ML 175 centipoises (125 MW)
(2) Waters

(a) T= 572°F

(b) P = 1250 psia

(c). h = 579 Btu/1b

() w= 2,874,970 o/br (95.9 M); w= 3,689,000 (125 MW)

(a) T 572°F

(b) P = 1250 psia

(c) h= 1181 Btu/hr |

(d) w< 355,030 1b/hr; w = 456,000 1b/hr (125 MW)

Water Flow Ares and Number of Tubes:

(1) Area. ¥ X (spec. vol.)_ (3.23 x 10° 1b/nr) x (00221 £43/1p)

=3.18 £t
vel, 3.6 x 103 sec/nr x 6.24 £t/sec 3

(2) Number of tubese Lobt8l Flow Area 3.184 ftz

- = 2336
Flow Area per Tube 66136) £42/tube

Salt Flow Characteristics:

(1) Flow area/Tube = 0,867 (Pitch)2 - _’Lf_ (cz)2

= 5 V@ 625 12
0.867 (‘—E‘) _g.. (_122_)

— 0.00126 ft°
-228~

Total Flow Area = 2336 tubes x 0.,00126 fte/tube.: 2.,

a0
W 7.49 x 10° 1b/br

(3) Hydraulic Diameter, De = hA a 4 (0.00126 £t )

Td (_“égi.ffi)

»

(4) Re veDe 2. 58 ft/sec x 127 lb/ft3 x 0.0308 ft
H 170 ¢p x 6.72 x 10” -l (1b-sec/ft)/cp

| Nu . |
(5) From Fig. 7.6; *;—573 = 2,15 (95.9 W) = 2.7 (125
r

127 1b/ft3 x 2.94 fta x 3.6 % 105 sec/hr

2.4

ol £t°

= 0,0308 %

=191 (95.9 MW)
258 (125 MW)

It

MW)

| ‘ ).
b= 2,15 x 2 [;C Flo- 235 (2.4) [50.57)(2.h2) 170:k 0.k
| K

d 0,0308

= 1015 Btu/hr-fta-oF = 1170 {125 MW)

1

=;_-—-—-_-—"20| 0 6
Rsalt 1015 00098

Inconel Tube Wall Regilstance:

3125 0.3125 )
Lot (ro/ri <; o 550

(1) R _ = = 11 3 — 0.000485

wall

Boiling Water Film and Scale Reslstance:

(1) Ren® (b 4 -
T h h
i scale boilin

| 1 1 |
= 1,25 (—5553 + 30—0(')—>= 0.000833

= 5,58 £t/sec
-229-

Heat~Transgfer Coefficlent for Weter in the Tubes:

e (6 24(2:2) 1.

 

 

D
(1) Res o 5 oos )(36;;10)3188::10 (95.9 M)
- = 2.4 x 10° (125 MW)
(2) h=0.023 & K pr0-4 g 0.023 O 1.35 (0.225)|°"" (1.88 x 10°)
= G.297 OU X

. 20
= 2700 Btu/hr~ft ~ F (95.9 MH) = 3290 (125 MW)

r
R _ e fr -, 1
(3) water = p, h + h ;)
L water scal

1 1
=1.25 3755 + 2000)
= 0,001087

Overall Heat Transfer Coefficient for Water-Heating Area:

(1) RTO tal™ 0.000986 + 0,000485 + 0.001087 = 0,002558
(2) U=_L_ _ m@%g = 390 Btu/hr-£t°- F (95.9 M)

Total
| = 425 (125 MW)

Over-all Heat Transfer Coefficient for Boiling Area:

(1) R = 0.000833 4+ 0.000k85 1 0.00Q986:= 0.002304

Total

(2) U = 430 Btu/hr-£t°-OF (95.9 M) = 460 (125 M)
-230-

K. Heat-Transfer Area for Water-Heating Region:

8
(1) apep = ¥ab = 355,030 1b/hr (579 - ¥72) Btu/lb= 0.373 x 10 Btu/hr

8
q 0.373 x 10 Btu/hr
Atsalt'“ 3

= . = 8.65°F
We T 7.49 x 10° Ib/hr x 0.57 Btu/hr-1o-"F ?
(2) at_ = 138.8°F (95.9 M) - 163°F (125 MW)

' 8
__a_ _ 0.373x10 o el
(3) A= gat = BE(00y = 00 (5.9 W)

2
Area = 80k £t (125 MW)
L., Heat-Transfer Area for Boiling Region:

(1) gq= wah = 355,030 (602) = 2.1k x 108 Btu/hr

(2) At (761.8 - 572) - (T2L.7 - 572)

 

 

 

 

1m 1 (189,8 = 1630F (95.9 M)
"\139.7 L
- = 195 F {125 MW)
(3) A= 3 o 218 x‘;bB Btu/hf _ 3080 £ ( i)
Uat 430 Btu/hr—ftdnoF x 163°%F 92.9

2
3050 £+ (125 MW)

M. Total Steam Generator Heat Transfer Surface:

~ Area = 3040 + 690 = 3730 £t (95.9 MV)

A= 3050 + 80k = 3854 ft2 (125 MW)

5 |
The area of 3854 £t 1s the design erea and gives a tube length
of 10.08 ft,
—,231—

II. Heat-Transfer Calculations for the Superheater
A, Inconel Tube Data:
(1) size: 0.5 in. 0.D,; 0.4 in, I.D,
(2) Pitch: 0.75 in. delta array
(3) Thermal Conductivity = 13.5 Btu/hr-ft-CF
(4) Specific heat = 0.133 Btu/lb-F
(5) Density = 507 1b/et
B. Superheater ‘Inlet Conditions:
(1) Molten Salt:
(a) T = 1138.1+°F'(95.9 Md): T= 1150°F (125 MW)
(b) w= 7.49 x 106 1b/hr
(c) c = o'.57 Btu/lb-dF
(d) H = 18.5 centipoise

(e) e = 123 1b/red

i

(f) k =24 Btu/hr-ft-oF
(2) Stean: |
(a) 7= 572F .

(b) P = 1250 psia

I

(¢) h = 1181 Btu/ib
(d) w= 263,300 1b/br (95.9 MH); w = 348,000 lb/hr (125 M)
C. Superheater Outlet Conditions:
(1) Molten Salt:
(a) 7= 1120.5°F (95.9 Mi); T = 1126°F (125 MW)
(2) Steanm:
(a) T 950F

(b) P= 1235 psia
~232-

(¢) h = 1470 Btu/1v

(d) w= 263,300 1b/hr (95.9 MW); w = 348,000 1b/hr (125 MW)
(e) vel. = 75.7 ft/sec (95.9 MW); vel, = 100 ft/sec (125 MW)
(£) spec. vol.= 0.650 £t°/1b |

Required Number of Tubes:
(1) Number of Tubes . ¥ X (spec. vol)
vel, x area/tube
(263,300 1b/hr) x (0.650 £15/1b)
(2.72 x 107 f£t/hr) x (8.72 x 10~ £4° /tube )
= T22

Salt Flow Characteristics:

2 2
_ 0.75 _ 0.50
(1) Fléw area/tube = 0.867 35 nip. G_Téé-

' 2
= 0,00203 't

(2) Totsl Flow Area = 722 x 0.00203 = 1,465 £2

 

 

 

 

. »
Cvel. = o 7.49 x 10 1b/hr . 11.55 £t
) CA (123 16/£t3) x (1.465 £4°) x 3.6 x 10° sec/hr ?7 e
(4) Hydraulic Radius, De = _l% = B__(g%gg:)_) = 0.0622 £%
L) ,F(_;i)
12
(5) Re=Y£De _ 1L.55 x o.ofizz x 123 710
6.72 (1077) 18.5
(6) From Figure .7.6; %g = 21 x 1‘-‘r0°1‘L
' 0.k
L 2.4 | 2.2 (0.57) (18.5)'] x 21
" 0,622 2.5
. 2 O
= 2050 Btu/hr-ft - F
R = 0.000487

gsalt
-233-

¥, Steam Heat Transfer Coefficient:
(1) Average spec. vol., = 0.500 ft3/lb

v _x spec, vol. (2,63 x 10 1b/hr) x 0.500 ft3/1b
G = Average Vel.= Area = D
722 tubes x (8.72 x 107 £+°/tube

G = 211,000 £t/br

0.0266 0.8 0.20

(2) From Ref. 21,}3x x G Xc XM
. _(6/12)0'2 4o

For these conditions,
0.20

CP XF —_— Ooll'o

0.0266
o.2

0.4
%)

0.8
h= x (211,000) x 0.40

 

=382 Btu/nr-£t°-F (95.9 Mi) = 478 (125 M)
T
o 1\ O.E 1
R steam = X (fi')~ 0. (382 ) = 0.00313

G, Heat Transfer Resigétance Through the Tube Wall:

v, 1n (zo/7;) _(0,25‘) <ln (0.25/0.20)

(l) R= K - 12 1305 = 0.0003&3

H, Over-all Heat Transfer Coefficlent:

(1) RTo”cal = 0,00313 + 0,000343 + 0.000487 — 0,00426

(2)' U= & = 235 Btu/hr-ft2~oF (95.9 MW)

=

O
U= 291 Btu/hr-fte- F (125 MW)
-234 -

I, Log-Mean Temperature Difference:

1120.5°F - 572°F = 548.5°F

Aty =
A 1138.4°F - 950°F - 188.4°F
(A% -At)
At, = e—B O
1m™ Ath
1n(

_ 548.5 - 188.}
]11(5 )
= 337°F (95.9 M) = 330°F (125 MW)

g Heat-Transfer Ares:

8
a 0.766 x 10 Btu/hr &7 o2
e _ . == 4 m
A= AW 235 Btu/hr-£t°-°F x 337°F 7o1 (95.9 1)

A= 1070 £t° (125 MW)
The heat-transfer ares of 1070 £4~ is the design area and wade
necessary a tube length of 1l.h ft.

IIT. Pressure Drop Calculations for the Steam Generator Loop

A, Head Loss in Salt Circuit

(l) In the steam generator heat-transfer region (see Fig. T.5):

JCPICIR

10.1 £4 ) (5.58 £t/sec)”
0.0308 £t/ \ 2 x 32.2 £t/sec?

h

il

0.25

i

= 42,7 £t
-235-

(2) In the salt lines:

(agsume 30 £t of ll-in. I.D. lines)

vel.= ¥ _ _ 3.Thx 10° 1o/hr

PR~ = o7 1w/7t3 x 0.66 ri2 - oo Tt/sec

veDe _ 12.4 (127)(0.918)
7 6.2 (107H)(270)

h = 0.03 —91-5) (J%fi—l,i—)-) 2.34 £t

(3) As a rough approximation it is assumed that thevKl for bends,

Re = = 12,600

entrances, valves, etc, is 4.0:

(4) Total Head Losses:
b= 2.7+ 9.6 + 2,34 = 54.6 £
Ap = 45.5 psi

(5) Pumping Power:

3.7h % 106 1b/hr x 54.6 £
~ (1.98 x 10° £4-1b/hr)/hp

= 98 hp
-236-

IV. Pressure Drop Calculations in Superheater Loop

A. Head Loss in Superheater Heat Transfer Region (Friction factor

from Figure T7.5):

h

s
N
ofer
St
N
9|4
mi o
S’

h

it

 

2
0 11,4 £5)/ _(11.55 ft/sec)
O 14.95 (0106_22 ft) (2 x 32°2 ft/sec2 )

= 18.6 £t
B, Assume 30 £t of salt line, 1l~in, I.D.:

vel., = 12.h £t

Re~ YeDe _ 12.h (123)(0.918)

Mo T g2 (10°4) 18,5

= 113,000

| 30 £% (12.) £t/sec)
0.018 (6T§187§%> ( 2 x 32.2 ft/sec? )

=
i

 

1.1 £t

l}

C. From Appendix 6.2, the head loss in the primary heat exchangers
is 23,7 f1.
D, Assume the total Kl due to bends, entrances, valves, etc,

is approximately 6.0:

K Q%é;)

6( (12.} £t/sec)? 2)

2 x 32.2 £t/sec

h

H

 

i

I}

b b £t

 
w237 -

E. Total Head Losses:
H = 18.6 + L.k 4 23.7 + 1h.b - 58.1 £%
F. Pumping Power Required:
p_ ___ 14.55 1b/br x 58.1 £t
(550 ft-1b/sec)/hp x 3.6 x 105 sec/nr
= 130 hp
V. Temperature Drops énd Maximum Heat Fluxes

A, In the Steanm Generator:

-g8lt  water
(1) Atmax== tmax - tsat

= 800°F - 5720F
= 228°F
(2) In "Studies in Boiling Heat Transfer" document No, C00-2k

(UCLA 1951) we find the empirical equation for water

boiling tubes,

(ti - tsat;)

1

loc 7. 123 = 35 logyy (P, )

i}

123 - 35 10g10_(1250 psi)_

14,5°F

(3) In McAdams (ref, 17) equation 1L-7 for water boiling in

T T EaéLialélifi
w ' Tsat T T p/900

(4) The results from (2) we used as a starting point to get a local

tubes is,

heat-transfer coefficient for boiling, The over-all local
Bo

 

~238-

heat-transfer coefficient vas computed at BOOOF. The
coefficiefit, h, for the salt was 1290 Btu/hr-£t-OF and

the wall resistance, R, is 0.000497, By a trial and

error method thé heat flux, temperatu:e dropy and over-
all.heat-transfer coefficient (local) was 752 Btu/hr-fte-oF.
Then:

(q/A)ma = UAt = (752 Btu/hr-r+2-CF) x 228%
X

2
= 172,000 Btu/hr-ft

 

At . 1.9 (172,000) /% 9.6%
sat = 12507500 = .
. ) | O
At 11 = (172,000)(0.000497) = 85.2°F
172,000 o
Atsalt = 1290 = 133.2°F

Total At = 228°F

In the Superheater:

(1) Aoy = 1126°F - 572 F = 554°F

(2) Steamn:

~ Q.0266 ' X GO,B % }10.2 X c
loc T (d/12)0.2 D

= 795 Btu/hr-r42-OF

(3) R, ;4= 0.000487

 

(%) R g11 = 0.000343
(5) v 1 propt
2} Uloe = = = 417 Btu/hr-ft - F

Total

 

 

 
 

=239~

(6) (q/A)max = UAt - 231,000 (Btu/hr)/fta
‘ | Q
(7) At 4. = 231,000 (0.000487) = 112°F

O
(8) A% o11= 231,000 (0.000343) = BO.F

231,000
(9) &t - = —3~%§-5——— = 36207

(10) Total At = 554°F

 

 

 
APPENDIX 8.1

~240-

3 Group Cross Sections

 

 

1200°F BE(KT) = 0.0795 ev
110 Mev —— 0,183 mev

Eiiiini Cog 5ie 5a - otr
Beryllium 0;768(61) See Ref 1 0.037u(61) :3.517(61)
Oxygen 0.322'%%) 0 0.00(61) 2,576(0%)
Fluorine o.3h7(6;) Neglected 0.00173 % 3.26]_(61)
Sodium o,é61(61)  Neglected 0.00021 %% 2.998'61)
Nickel =0.1é&(61) 0.743(62)  0,05(61) '3.607(61)
Zirconifim 0.135(6$) 0.765(62) 0.00017(63.) 6.15h(61)
traniun-235  0.745(61)  meglectea 1430 1.207(6Y) 7. 055(61)
Boron-10 Neglected Neglected 0,986 %) A
chromtn  0.131(%Y) 0.625(2)  0,050(6%) 3.373'6%)
1ron 0,104 0.665(52)  0,050(61) 2.819¢6Y)
ML en | EBMr MY
Beryitium  1.2200%%)  0.000198(6%) 50616
Oxygen 247'% Negloctea 3.573¢%)
Fluorine 37711 " 0,00364' %) 3.5u0(6)
Sodiun a97(81) 0,0220(61) 5. 6a9(61)
Mickel 580¢6L) o, 1066(61) 16.356(61)
Zirconium 1481 0.060(8) 7.08(62)
Uranium-235 Neglected 25f19(61) 16.36¢81)  o.9p08(61)
Boron-10 Neglected 91.9(61) 3.238(61)
Chromiun 0.22(6?) Same as Ni 5, 77(62)
Iron 0.32%%)  sane as mi 8.93(62)
Croup 3 = Thermal

-2 -

KT = 0.0795 ev

 

Atom or Molecule oca op | o

Beryllium Oxide 0.0102(34) 9_6(5)
Fluorine 0.005(3%) 5.8(84)
Nickel 2,3(3“) 19.6(3%)
Zirconium 0,09(3“) 6.3(3“)
Uranium-235 351.1(3%) 296.5(3)  357,1(3%)
Boron-10 2005(34)‘ 2005(34)

Chromium 1.45(3h) h.h(Bh)
Iron 1_265(3h) 12.2(3h)

APPENDIX 8,2

Perturbation Technique

The perturbation technique developed below is a'vefy simplified

approach in obtaining reactivity changes incurred through small

perturbations in cross sections, away from the critical parameters.

Upon making diffusion theory approximation to the current J

and assuming a2 solution for a bare one region three group system of

2 2
the form (7 + B )é = 0, one obtains the following steady state

equations.

(DlBl2 + 2g1 *Zrl)‘bl

-V (Zpry + Zephy + 2 p5f5) = ©

2 .
(DB, + Zap *2rp)02 = 2198y = ©

2
(D3B3 + Za3)¢l “51'2(#2 = O
-2h2-

The steady state solution is

 

 

 

 

 

 

Zey (_Zra ) ) 1, 2e 21
] - Za3  <rot+iap 2pitZa1 0 (B3LgT+1) (o +24p ) (Tal +2r1)
312%14-1- Yzl ‘322 T, + L
Dy LE. J) Ds
1T By vy "\ 7a
Define: -_-)Zfl

T ® )

7? o ))ZfE ' Erl
2 = {Z o+ 2pp) Car+ 2 1)
V 2 rg 2pp - 211

 

Now approximate the sfieady state solution as

N3
2
(B5"Ls

{Ble T, + 1 _-W.’Z'l} XBEZ’?;‘2+ 1}

Assume the system is perturbed by an amount 51?2 and Sns and

+ M
+ 1)

 

2 2

-~ -
—

 

neglect its effect upon-’t' and L, resulting in a multiplication con-

stant k + b k.
Then:

k + %k _ 3
X = 1 +f

 

 

 
-243~

and

Y73 + 51 p(B3°Ly" + 1)
Yo+ Mo (3321,32 +1)

-
-

 

APPENDIX 8.3
Burnup and Fission Product Poisons
Burmup:
fipon the burnup of one gram of U~235, 2.563 x 10°% atoms are

destroyed. Of this number, <§}/{yaj> is the fraction destroyed by

 

 

 

fission.
i=3 i=3
T _E ,<0“f> o E: ,
\ 0% :>>*' Y1 O /. 0 Yi
i=1 1 i=1

|

where Lpi = fraction of fissions in energy Group 1.
0"f>
0a i

1. <0“f/0“a> i Y4

 

Ratio of fission to absorption cross section for U-235
averaged over energy group 1.

i

 

1 0.8988 .083
2 0.6495 .639
3 0.8445 .278
o
< £ > = 0.7243
o
& gpectrum

i 21
Therefore one gram burnup of U-235 requires on the average 1.856 x 10

fissions, and one full power hour of reactor operation at 125 MW is
-2hh -

equivalent to T7.515 grams U-235 destroyed., Inventory of U-235 within

the reactor at any time T is expressed as "
M(T) = M(o) - 0.007515 ¥

M is kgms T in full power hours., Also the concentration of‘U-235 per

cm3 of fuel can be expressed relative to initial concentration as

e (T) - E,(O) 1 - 0.00;%i% T ~{

All cross section involving the fuel are written as functions of @

for example:

Sa3® {0.00165 + 0.45539} | °

Figsion Product Polsons

The additional sbsorption resuliing from non volatile fission
products are approximated by the following assumptions.

1 fission = 100 bérns equivalence of thermal polsons

1 fission = 10 barns equivalence of intermediate poisons,
Then the added wacroscopic absorption.cross section for the core

region are given as:

Core
AT, o(T) = 566 x 1001 cmt
Core - _ ' -
dZaE(T) . 566 x10 T — .

T in full power hours,
The worth in terms of reactivity are calculated as function of T

by the perturbation method described in Appendix 8.2,
-245-

APPENDIX 8.l

Prompt Neutron Lifetime

The following analysis is & relatively simple method of estimat-
ing prompt neutron lifetimes from multigroup constants for an unreflected

system. Method in part 1s similar to that presented in Ref. 70,

Define:

Ti = average time a neutron spends in the'ith energy group.

 

 

Ni

Yzi = Relative number of neutrons existing in the 1 4h group
o Ni = Fraction of neutrons born in the 1P group
J1i = Average neutron speed of the 3 group,
Then:
. _
Ti = i
Vi 2y
< | 2
= + D B,
Where :Eti ;Eai'* zgxi i 1
{j:i—ll .
Ni = Ni+ i (2x) ;
p=1 N3
= 2
ti
. J=l
Then the prompt neutron lifetlme over all energy groups, k in
nunber is:
izk
’ <fz ~ i i
= P
i1

o

1
-246 -

"APPENDIX 11.1

DESCRIPTION OF SIMULATION PROGRAM . -

The flow sheet of thé system which was simulated is shown in
Figure 11.2, Figures A-1l.1 through A-11.5 show the electrical
circuits or roadmaps which were used to represent the fuel loop and
heat transfer circuit 1. The roadmaps for heat transfer circuit
2 are similar to but simpler than those for circuit 1; they are
now shown., ‘The method for simulating reactor kinétics is well
known and is not repeatéd here, See Beferences 30 and 31. ¢

Figures A-11,1 ghows the rbadmap for the fuel loop. Amplifiers
1 and 2 represent passage of the fuel through the core. a%c 15 the
meanlfuel'temperature in the core. Amplifiers 3 and 4 represent passage
of the fuel through one primary heat exchanger. Oy is the mean
fuel temperature in the heat exchanger. Amplifiers 5 and 6 simulate
passage of-fuel through the other primary heat exchanger.

Figure A-11.2 shows thé'roédfiap for the salt side of the primary
heat exchanger and the superheater, Amplifiers 9 and 10 represent
passage of the salt through the primary heat exchanger. Terminals
marked C. S. go t0 the control system circuits which are shown in
Figure A-11.5. These wili be described later, Amplifier 11 generates
the salt temperature resulting when the by-passed salt mixes with the
salt from the primary heat exchanger. The box labeled ’ZS represents
8 time-lag device which simulates the transport delay in piping.

Amplifiers 12 and 13 simulate salt passage through the superheater,

 
-247-

Amplifier 14 again generates a mixed-salt temperature (See Figure 11,2),
Amplifiers 7 and 8 generate the coupling voltages between fuel and salt
in the primary heat exchangér. The pover transferred across the exchanger,
Ph is determined by the mean fuel and salt temperatures in the exchanger.
The timeiconstant”bf amplifier 7 represents, to some approximation, the
heat capacity of the tube metal in the exchanger;

Figure A-11.3 shows the roadmap for the salt side of the steam
generator. Amplifier 16 generates the mixed salt temperature going into
the steam generator. Awplifiers 17 and 18 simulate passage of the salt
through the steam generator.

-Figure A-11.,L4 shows the method used for generating the power demand
voltages. Ps and Pg are the power dewands from thé superheater and
steam generator respectively., The ganged potentiometers may be set to
any desired power demand. Amplifier 24 generates the output steam tem-
perature, The assumption is made that steam tenperature is §r0portional
1o the superheater inlet salt temperature and to the power extracted from
the superheater. Here an effect rather than a physical phenomenon is
being simulated, |

Figure A-11.5 shows the manner in which the control system was
simulated, A Brown recorder was used %o display the output steam tem-
perature Qvg. Limit switches were placed on this recorder in such s
manner that.when steam temperature varied from its design value byla certain
threshold setting a voltage of proper polarity was applied to amplifier
25 through a gain sefting potentiometer, Amplifier 25 integrated this

error voltage to give Ws5’ the flow rate fihrcugh the by-pass line. WSS

 
~248-

 

1s limited by the diodes in the feedback circuits around amplifier 25
to lie in the range of zero to 75 volits (zero to 1570 pounds/second flow

rate), The voltage representing W _ drives a multiplier and the

85
appropriate output connections are shown in Figures A-11.5 and A-11.2,
Weys the flow rate through the heat exchanger is generated by amplifier
o6 as the difference between Wsl (a constant represented by 100 volts)

and W Wsh'drives another multiplier as shown in Figure A-l1l.5.

g5°
APPENDIX 11.2

Expansion Chamber Heating Calculations

If fuel volfime is takenlto be 45 cubic feet, and fuel is pumped
into reactor at llOOOF, as in the ARE, and raised %o 1225°F average
operating temperature, | |

Density of fuel = 253.0 - .0328T(°F) 1b/£4t3

216,9

A

Density at 1100°F

Density at-l225°F = 212.8

Expansion of fuel = 32 £43 x b.1 lb/ft3== 18%.5 1v

18k.5 1b -~ 212.8 lb/ft3:= 867 ftB, volume of fuel in expansion
chamber at operating tempersture.

Assuming that all fission product gases are held in the expansion
chamber for 3 - 10 days, or until rate of generation just equals rate of
decay of all unstable gaseous fission products, ignoring loss due %o
neutron absorption, at 125 M{ power, heating rate 1ln expansion chamber
_due to gas decay is about 150 kilowatts

Decayed heatageneratlon rate is about 7-1/2% fission heat release.

 

 
d0071 TTdNd JHL1 40 NOILVLIN3IS3dd3d 9O0TTVNV

 

 

 

 

 

 

 

 

 

 

 

 

 

Fil-Vv 34N91td
YIONVHO X3
1V3IH AYYWIMd 3JHL NI LvS 3H1 8 73nd. 3HL N33ml38

ONITdN0D 081V ° ¥YILVIHYIANS 3IHL B IONVHOX3I LV3IH
AHYWING 3HL 40 30ISs 17vS 3IHL 40 3JAIVIN3SIHdIY 9DOTVNVY

gl 169
Iy
%l 1 @ ¥ ‘S'0

 

 

 

 

o %

 

 

 

 

 

 

ORNL~-LR-Dvg 25717

 
 

 

 

 

 

§
<
u™
o

¥

 

 

 

 

 

 

)

 

 

 

 

¢’li-v 3¥N9Id

 
 

JOLVY3INI9
WV31lS 3JH1 40 34IS 11vS 3HL 40 NOILVINIS3H¥d3Y 9S0TVNV

.fl.%@ _

 

 

J251-

 

 

 

 

 

 

 

 

298L ol

 

 

91

 

 

 

 

 

€1-v 34n9i14
JYNLIVH3IdWNIL WV3LS
1Nd1NO0 ANV AONVW3Q H3IMOd 40 NOILVHINIOD HO4d AHLINOYIO

 

 

 

 

 

A 001

 
 

-252-

 

 

 

 

<
o

 

 

A
£

 

 

 

Y A 001

 

sd Sd ~

 

PH-V 34N9Id
 

1
13!
M

G

09§93 o§S? o ¥'SD

 

WI1LSAS TTOHINOD
40 NOILVIN3S3IHd3H OO0TVNV

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

—l - YAAdILION
oleso alzeov)1'so
{
_ nw@ w@l n@
8°5°0
iiiii ¥IITdILINN
¥
Sm L eo
Age>%Sm >0

Gl

  

 

 

 

\I_I_ _m.s._.._..
¢S 1S. #S _ O
Mo MM / ool ~
i .
H304003Y
001—-O .«\«@
oNILL3S A OOl -M
NIVO |
Otigee
>oo_.+M

 

 

 

-V 3¥N9I1d

 

 

 

 
 

 
 

 

Instantaneous gammas and capbure gammas constitute about 6% more. As a
part of the delayed heating is in the gaseous phase, and an exact cal-
culation of gamma heating in liquid phase of expansion chamber is not
within the scope of this project; an estimate of the liquid phase heating
due to beta and gawma energy absorption in the expansion chamber is

taken to be 5% of the total Fission heat, divided by the total volume

in the expansion chambef.' Thus,

125 x 100 watts x .867 ft3 x .05 = 120 KW
45 £43

 

In the absence.of flux data for the expanglon chamber, a figure
of 130 XKW is assumed for fission heating in the chamber. The ball park
estimate is founded ofi information obtained from Mr. Lackey of ORNL,
and is derlved from an estimate for the ART,

Therefore estimated heat rate in expan31on chamber 1s 150 + 120 +
130 2’ 400 KW,

Assuming that fuel is bought ihto chanmber at 12500F and experiences
8 100°F temperature rise before being expelled,

400 KW x 3415 BTU/hr kw = 1,366,000 BTU/hr,

1,366,000 BTU/hr = (.27 BTU/1b °F x 100°F) = 50,600 lb/hr, fuel flow
required to remove heat, |

If a temperature rise of T5°F is experienced, 67,500 1b/hr will be
required.

Assuming that inlet tempersture is llTBOF, expansion tank exit tem-
perature would be 12750F and average temperature 1225°F with 100°F rise.

Use of a helium purge for the system would result in reduction of

ans5i0R, chamber by no more than 30%-

 
A,

w255 -

APPENDIX 13.1

BREAKDOWN  OF BASIC REACTOR POWERED SYSTEM COMPONENT WEIGHTS

Category A and B (Steam Propulsion Machinery)

L.

2.

= W

oo ~I &N \Wn

10.
11.
12,
13.
1k,

- 15.

16,

Main propelling units

Main shafting

Main shaft bearing

Iubricating oil system

Main condenser and ailr ejector
Circulating, condenser, and booster pump
Propellers

Steam and exhaust piping
Water and service piping
Insulation and logging
Floors, gratings, and adjuncts
Auxiliaries

Fittings and gears

Ligquids

Total weight

Specific weight

Category C and D (Reactor Plant Machinery)

1,

Reactor Proper

Pressure shell
Thermal shield

Fuel

130,150 1b
86,480
14,810
19,650
36,040
13,435
18,280

69,580

72,110

21,530
22,400
2,200
12,500
42,510

601,675

17.19 1b/SHP

13,172
4,100

11,365
~256~

Coolant

Heat exchanger Structure and headers
Moderator rods and cladding

Moderator support structure

Control rod

-Poison rods and cladding

| Nickel shield

Miscellaneous (5%.total reactor weight)

Total reactor weight

Steam Generating System

Dry boiler, 2 at 55,000 15 each

Salt holdup in boiler, hOOO 1b each

Water in boiler, 4000 1b each

- Dry superheater, 2 at 9000 1b each

 Salt in superheater, 2300 1b each

Secondary salt plumbing, total

Salt in secondary plumbifig, total

Steam and salt in lines

Salt pfimps, L at 4000 1b each

Boiler recirculating pumps, 2 at 6000 1b each
Additibnal feed water heating

Thermal insulation

Total |

Additional structural support at 25% of total

Total

3,430
12,550
2,185
5,330

80
L, h52
9,510

3,482

69,656

110,000
8,000
8,000

18,000
4,600
3,000
8,000
4,000

16,000

12,000
8,000

4,000

199,600

0,000

2kg, 600
T

L.

2,

3.
b,

-257-

Dump Tanks (Primary and Secondary) .
Fuel Pumps, 3 at 4000 1b each

Miscellaneous (Instruments, additional lines, etc)

Total Weight

Specific Weight, 431,260/35,000

Category E (Radiation Shielding)

Primary Shield

Tank inner wall

Lead

Tank outer wall

Water

Shield plug

Total weight of primary shield
Secondary Shield

Aft-face

Top face

Top hat

Side faces

Forward face

Superheater shadow shields

Total weight of secondary shield
Total Shielding Weight

Specific Weight of Shield

30,000 1b
12,000
70,000
431,260

12.32 1b/SHP

13,800 1b
51,000
14,200
117,000
28,140

22k ,1h0

142,620
248,320
13,380
78,900
10,000
10,000
503,320
727,460

20.78 1b/sHP
 

 

~258-

Category F and G (Blectric Plant)

Total weight 210,000 1b

Specific weight - - 6.00 1v/sHP

 

Category H and J {Independent Systems)
Total weight | | 182,000 1v

Specific weight 5.2 1b/SHP

 

Category L (Tools, Equipment, and Spare Parts)

Total weight 70,000 1b
~Specific weight o 2,00 1b/SHP
Fuel 0il "
Total weight [ 0 b
Specific weight | | 0 1b/SHP <
Total Systefi Weight 2,281,000 1b

 

Specific Weight of Entire Plant ' 63.5 1b/SHP
R

Ref, No.

10

11

12

13

=259~

BIBLIOGRAPHY

Poppendiek, H, F., and Palwer, L, D,, "Forced Convection
Heat Transfer Between Parallel Plates and in Annuli with
Volume Heat Bources Within the Fluids", ORNL-1701, May 11,
1954, Unclassified.

Poppendiek, H. F. and Palmer, L. D., "Forced Convection
Heat Transfer in Plpes with Volume Heat Sources Within
the Fluids", ORNL 1395, December 17, 1954, Unclassified.

Field, F, A., "Temperature Gradient and Thermal Stresses in
Heat Generating Bodies", ORNL-CF-54-5-196, May 24, 1954,
Confidential.

Fraas, A, P., "In Pile Testing of High Temperature Moderating
Materials", ORNL CF-56-9-99, September 2k, 1956, Secret,

MacPherson, H, G., et al, "A Preliminary Study of Molten
Salt Power Reactors", ORNL CF-57-4-27, April 29, 1957, Secret.

Weinberg, A. M., et al, "Molten Fluoride Reactors", ORNL CF-
57-6-69, June 1957, Unclassified,

Davidson, J. K., Aebert, R. J., Morecroft, B. T., "A Fused-
Fluoride Homogeneous Reasctor System for Submarine Propulsion
(SAR Phase III Study)", KAPL-992, September 15, 1953, Secret,

"DIG Project Progress Report', KAPL-1800~1, February-March,
1957, Secret,

Smith, C, O0,, "Notes on Materials Engineering for ORSORT
Students", First BEdition, ORNL, 1956-57, Unclassified,

Smith, C. 0., "Stress Analysis for ORSORT Students", Second

Edition, ORNL, October, 1956, Unclassified.

Cooper, W, E., "Modified Structural Design Basis, SAR Reactor
Components-I11", KAPL-N1-WEC-9, Rev. 1, May 1, 1957, Unclassified,

Fraas, A, P., and Laverne, M, E,, "Heat Exchanger Design
Charts", ORNL-1330, December 7, 1952, Secret,

Cohen, S, I.,, and Jones, T, N., "Measurement of the Friction
Characteristics for Flow in the ART", ORNL-57-3-95, March 19,
1957, Secret.

 
Ref, No,

14

15
16
17
18

19

20
21

22

.23

2k,

25

26

-260-

Platus, D. L., and Greenstreet, B, L., "Deflection Equations
for Various Loadings of Circular -Arc Curved Beams",
ORNL-57,4~96, April 22, 1957, Unclassified,

pPotter, P. J., "Steam Power Plants", The Ronald Press Co.,
New York, 1949,

Wllner, B. M., and Stumpf, H., J., "Intermediate Heat
Exchanger Test Results", ORNL-CF- 5h 1~155, January 29, 1954,
Secret.

McAdems, W, H,, "Heat Transmission", 3rd Edition, McGraw-
Hill, New York, 1954, Unclassified,

ASME "Boiler and Pressure Vessel Code", New York, 1952,
Unclassified.

Lewié, W. Y., and Robertson, S. A., "The Circulation of
Water and Steam in Water Tube Boilers, and the Rational
Simplification of Boiler Design', Institution of Mechanical

Englneers, Proceedings of March 19&0, London, Unclassified,

Glasstone, 8., "Principles of Nuclear Reactor Engineering",
Princeton, 1955, Unclassified.

The Babcock and Wilcox Co., "Steam, Its Generation and Use",
New York, 1955, Unclassified,

Davies, R, W,, et al., "ORSORT Reactor Design and Feasibility
Study 600 ww Fused Salt Homogeneous Reactor Power Plant",
ORNL CF-56-8-208, Oak Ridge, 1956, Secret.

Kays, W. M, and Londom, A, L,, "Compact Heat Exchangers",
National Press, 1955, Unclassified.

Peak, R, D, and Cooper, M. H., Report of Experiment No,
7405-1-1, "Heat Exchanger Evaluation, Type IHE-2, ORNL-1l
and 2", 12-15-55, Becret.

Peak, R. D., and Cooper, M. H., Report of Experiment No,
7405~1-2, "Heat Transfer and Pressure Drop Correlations,
Internediate Heat Exchanger Type IHE-3", ORNL-l and 2,
2-1-56, Secret,

Enstice, L. R,, and Hopkins, H. C,, Report of Experiments
No, 7405-1-3, "Heat Transfer and Pressure Drop, Intermediate
Heat Exchanger Type IHE-3", ORNL-1 and 2, 6~1-56, Secret.

3
Ref, No.

e

28

29

30

31

32

33
34

35

36

37

38

39

40

~261.-

Cohen, 8, I., and Jones, T:-N., "Measurement of the Friction
Characteristics for Flow in the ART Fuel-to-NsK Heat
Exchanger"”, ORNL CF-57-3-95, 3-19-57, Secret,

Wantland, J. L., "Thermal Characteristics of the ART Fuel-
to-NaK Heat Exchanger", ORNL CF-55-12-120, 12-22-55, Secret.

Wantland, J, L,, "Transverse Pressure Difference Across
Staggered and Inclined Spacers in the ART Fuel-to-NasK Heat
Exchanger", ORNL CF-56-6-143, 6-29-56, Secret.

Stone, J. J., and Mann, E. R., "Oak Ridge National Laboratory
Reactor Controls Computer”, ORNL 1632, March 1, 1956,
Unclasgified,

Schultz, M. A., "Control of Nuclear Reactors and Power Plants",

- McGraw-Hill, 1955, Unclassified.

Nuclear Power Branch of Central Technical Department, "Shielding
Notes", Shipbuilding Division, Bethlehem Steel Company,
Unclasgified,

Rockwell, Theodore III, "Reactor Shielding Manual', TID-T00L,
March 1956, Unclassified.

Hughes, D, J., and Harvey, J. A,, "Neutron Cross Sections”,
BNL-325, July 1, 1955, Unclassified.

Blizard, E. P., "Nuclear Radiation Shielding", ORNL, September
17, 1956, Unclassified.

Quarterly Progress Report for Period Ending December 31, 1956,
"Aircraft Nuclear Propulsion Project", ORNL-2221, March 12,
1957, Secret,

Stevenson, R,, "Introduction to Nuclear Engineering", McGraw-
Hill, 1954,

"Report of the 1953 Summer Shielding Sessions", ORNL-1575,
June 1, 195k, Secret.

Weir, J, R, Jr., et al, "Inconel as a Structural Material for
a High-Temperature Fused Salt Reactor", ORNL 2264, June k4,
1957, Secret,’

Cohen, 8. I,, et al, "A Physical Property Summary for ANP
Fluoride Mixes", ORNL-2150, August 23, 1957, Secret,
 

Ref, No.

41
2
43
4y
45
k6
b7
48
k9
50
51
52
53
54

55

56

57
58

29

60

~262-

Cohen, g8, L., ORNL, Personal Communication, July, 1957,
DeVan, J. H., ORNL, Personal'Communication, July, 1957,
Doney, 1. M., ORNL, Personal Communication, July, 1857,
Douglas, D, A., ORNI, Personal Communication, July, 1957,
Grimes, W, R., ORNL, Personal Communication, July, 1957,
Hikido, 7., ORNL, Personal Communication, July, 1957,
Inouye,.H., ORNL, Personsl Communication, July, 1957,
Hoffman, E, E., ORNL, Personsl Communication, July, 1957,

Lackey, M, E,, ORNL, Personal Communication, July, 1957,

Samuels, G., ORNL, Personal Communication, July, 1957,

Thowa, R, E., ORNIL, Personal Communication, July, 1957,
"Liquid Metals Handbook", Navexos P-733 Rev,, June, 1952,
"Metals Reference Book", Interscience Publishers, 1952,
Manthos, E, J,, e4 al, "Disassembly ang Examination of

the BeO Moderatop Capsule Test, ORNL-lO-h", ORNI, CF-57-3-53,
March 13, 1957, Secret,

Bolta, C, ¢,, "In.pile Test of Moderstor (BeO) Material
ORNL-10.4 " ORNL~CF~57-2-123, February 15, 1957, Secret.

Thoma, R, E,, "Reactor Coolant Mixtures", ORNL Intra-
Labora tory Memorandum to A, p, Fraas, Dated June T, 1957,

Nessle, g, Joy ORNL, Persongl Communication, July 1957,
Quarterly Progress Report for Perieg Ending September 10, 1956,
"Aircraft Nucleap Propulsion Project”, ORyL 2157, November
"Zebra Project Quarterly Project Report", Curtisg Wright
Research Department, CWR-470, March 31, 1957, Secret,

Bote, R, R., Einstein, I, T., Kinney, w, E., "Description

and Operation Manual for the Three Group Three Region Reactor
Code for ORACLE", ORNL CF-55-1-76, January 13, 1955, Unclassifie&,

 
Ref, No,

61

62

63

6L

65

66

67

68
69

70

71

72
13

s

-263~

"Neutron Cross Sections for Multigroup Reactor Calculations”,
Curtiss Wright Research Department, CWR-413, September, 1955,
Secret,

Deutsch, R, W., "Calculation of the Neutron Age in Hydrogeneous
Mixtures", KAPL Memo RWD-13,

R. Ramanna, et al, "On the Determination of Diffusion ang
Slowing Down Constants of Ordinary Water and Beryllium
Oxide Using a Fused Neutron Source", P/872.

Dunning, F. 5. and LeDoux, J, C., "Bazard Consideration
for a 100 MW Fused Salt Reactor", ORNL CF-57-8-8, to be
published, Secret,

Sense, Karl A., et al, "Vapor Pressures of the Sodium
Fluoride-Zirconium Fluoride System and Derived Information"”,
BMI-106k, Unclassified.

Palmer, L. D., "A Preliminary Analysis of the Temperature
Structure Within a Solid Moderator Rod, Cylindrical Reactor",
ORNL CF-57-4-138, April, 1957, Confidential. '

Quarterly Progress Report for Period Ending June 30, 1957,
"Aircraft Nuclear Propulsion Project", ORNL~2340, To be
published, Secret, '

Stehn, R. J., and Clancy, E. F., "General Electric Chart of
the Nuclides", April, 1956. '

Walker, C. 8., "Reactor Control", ORNL CF-57-1-1, January 5,
1957.

Holmes, D, K., "Calculation of Average Lifetime of Neutrons
Using the Results of Multigroup Calculation”, Y-F10-15,
October 2, 1950,

Geortzel, G., "An Estimation of Doppler Effect in Intermediate
and Fast Nuclear Reactors”, P/613.

Kinyon, B, W., ORNL, Personal Communication, July, 1957.

Schroeder, J, H., et al., "A Preliminary Design of a Unit
Shield", ORSORT Study Group, ORNL CF-58-813, To be published,
Secret,

Dee, J. R. and Woodsun, H, C., "An Analysis of F, P. -Ray
Spectrum”, Volume 4 and Document No, NARF-56-41T, FZK-9-109,
1956,