Title Page
 Table of Contents
 List of Tables
 List of Figures
 Technical feasibility
 Material compatibility testing
 Liquid metal wetting in angular...
 Liquid bonded fuel rod thermal...
 Fuel rod thermal/mechanical performance...
 Improved designs to enhance LWR...
 Conclusions and recommendation...
 Biographical sketch

Title: Liquid bonded light water reactor fuel
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00097381/00001
 Material Information
Title: Liquid bonded light water reactor fuel enhanced light water reactor safety and performance
Physical Description: xi, 206 leaves : ill. ; 29 cm.
Language: English
Creator: Wright, Richard Frederick, 1958-
Publication Date: 1994
Copyright Date: 1994
Subject: Nuclear Engineering Sciences thesis, Ph. D
Dissertations, Academic -- Nuclear Engineering Sciences -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis (Ph. D.)--University of Florida, 1994.
Bibliography: Includes bibliographical references (leaves 203-205).
Additional Physical Form: Also available on World Wide Web
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by Richard Frederick Wright.
 Record Information
Bibliographic ID: UF00097381
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 002044840
oclc - 33372970
notis - AKN2756


This item has the following downloads:

PDF ( 6 MBs ) ( PDF )

Table of Contents
    Title Page
        Page i
        Page i-a
        Page ii
    Table of Contents
        Page iii
        Page iv
        Page v
    List of Tables
        Page vi
    List of Figures
        Page vii
        Page viii
        Page ix
        Page x
        Page xi
        Page xii
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
    Technical feasibility
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
    Material compatibility testing
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
    Liquid metal wetting in angular gaps
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
    Liquid bonded fuel rod thermal analysis
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
    Fuel rod thermal/mechanical performance analysis
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
    Improved designs to enhance LWR fuel safety
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
    Conclusions and recommendations
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
        Page 185
        Page 186
        Page 187
        Page 188
        Page 189
        Page 190
        Page 191
        Page 192
        Page 193
        Page 194
        Page 195
        Page 196
        Page 197
        Page 198
        Page 199
        Page 200
        Page 201
        Page 202
        Page 203
        Page 204
        Page 205
    Biographical sketch
        Page 206
        Page 207
        Page 208
Full Text






1 T.1


UNI ",'T 01 111MIA
111111111111111111111111111111 I iliI~lll ill 111 11 1 1111
3 1262 08552 3214


The author wishes to express his gratitude to Professor James S. Tulenko for

his vision, patience, guidance, enthusiasm, and breadth of knowledge of nuclear

fuel design. Without his support, this work would not have been possible.

I am also indebted to Dr. Richard G. Connell, Professor Emeritus Glen J.

Schoessow, Thad M. Adams, and Mark Dubecky whose work on this project

parallels my own, and who have provided contributions to this work. I would like

to thank my committee members; Drs. D.E. Hintenlang, E.T. Dugan, G.R. Dalton,

and F.E. Dunnam for their help and guidance. Also Kathy Phillips for all her help.

I would like to acknowledge Dr. Odelli Osser of EPRI and Dennis O'Shay of

Florida Power Corp. for their help in obtaining the ESCORE computer program,

and the U.S. Department of Energy for funding this research.

I would also like to thank Dr. Frederick J. Moody, Dr. J. Edward Schmidt, and

Robert A. Markley for being friends and mentors, and for helping me to put my

education, career and life into the proper perspective.

I would like to give special thanks to my wife, Denise, and my children, Laura

and Rick, and Jim and Betty Hart for their love and support. Most of all I would

like to thank my parents, especially my father, who was my inspiration.


ACKNOWLEDGEMENTS ..................................... ii

LIST OF TABLES .......................................... vi

LIST OF FIGURES .................................. ....... vii

ABSTRACT .................................... ........... x


1 BACKGROUND ....................................... 1

Light Water Reactor Fuel Temperatures and Limits ........... 1
Sodium Bonded Metal Fuel Technology ................... 3
Liquid Bonding in LWR Fuel ................... ......... 4
Potential Benefits ................................. 4
Disadvantages ................................... 5

2 TECHNICAL FEASIBILITY ............................... 8

Choice of Bonding Liquid .............................. 8
Temperature Range Criteria ......................... 9
Nuclear Interaction .............. .. ... .......... 10
M material Com patibility ............................ 16
Fuel Rod Characteristics ........................... 21
Best Candidates ................................. 25
Thermal Considerations .............................. 26
Steady-state Fuel Temperatures ..................... 27
Transient Performance ................ .......... 28
Thermal/Mechanical Limits and Design Criteria ............. 29
Fuel Rod Failure ................. .... ... ... ..... 31
Severe Accident Analysis ............................. 33
M manufacturing ..................................... 35
Results of the LBLWR Feasibility Study ................... 37

3 MATERIAL COMPATIBILITY TESTING ..................... 39

Discussion of Liquid Metal Corrosion ..................... 40


Experimental Assessment of
Bonding Liquid/Cladding Compatibility .................... 45
Materials Used in Experimental Samples ............... 45
Sample Preparation .............................. 47
Test M atrix ..................................... 49
Metallographic Preparation of the Test Specimens ........ 51
Measurement of the Loss of Tube Wall Thickness ........ 52
Transition Layers at the Liquid Metal/Solid
Interface ..................................... 52
Liquid Metal Attack .............................. 53
Results of Material Compatibility Experiments .............. 53
Tube W all Loss ................................. 54
Evaluation of Reaction Layers ....................... 64
Liquid Metal Compatibility with UO2 ...................... 69
Additional Experimental Studies ........................ 72
Fission Gas Flow Through Liquid Metal ................ 72
Liquid Metal-Coolant Interaction ...................... 72
Summary of Experimental Studies ...................... 72


Experimental Studies ................................ 75
Analytical Predictions of Gas Blanketing due to Eccentricity .... 75
Results and Conclusions ............................. 76


Steady-state Fuel Temperatures ........................ 80
Transient Performance ............................... 85
Loss of Coolant Accident ........................... 85
Transient Overpower .............................. 89
Detailed Two-dimensional Fuel Rod Model ................ 93


Background ....................................... 96
ESCORE: Fuel Rod Thermal/Mechanical Performance Code ... 98
ESBOND: LBLWR Fuel Rod Analysis Code ................ 99
Installation on the Unix Platform ..................... 100
ESBOND Gap Conductance Model ................... 102
Liquid Bond Displacement .......................... 104
ESBOND LBLWR Fuel Performance Calculations ........... 107
ESBOND Analysis of the PWR Fuel Rod ............... 108
ESBOND Analysis of the BWR Fuel Rod ............... 120
ESBOND LBLWR Fuel Analysis Conclusions ........... 130


Three--Dimensional Heat Transfer Model ................. 132
Optimized LBLWR Fuel Design Conclusions .............. 140

Experim ental Results ................................ 146
Analytical Fuel Performance Results ................... 147
PW R Fuel Rod ................. ... ............ 149
BWR Fuel Rod ................. ... ............ 150
Recom m endations .................................. 151

SUBROUTINES ................................. 153

REFERENCES ............................................203

BIOGRAPHICAL SKETCH .................................... 206


Table page

2-1 Melting and boiling temperatures for candidate liquids ........ 11

2-2 Thermal neutron absorption cross sections for
liquid m etals ...... ..... .... ........ ............... 13

2-3 Fuel rod parameters for PWR and BWR fuel ............... 22

3-1 ASTM B350 Chemical Composition for
Reactor Grade Zircaloy-4 ............................. 46

3-2 Chemical Composition of Lead-Bismuth Eutectic ............ 46

3-3 Chemical Composition of the Tin Stock ................... 48

3-4 Chemical Composition of the Lead Stock ................. 48

7-1 Maximum fuel temperatures for LBLWR pellet designs ........ 135


Figure page

2-1 Doppler coefficient vs. effective fuel temperature at BOL ...... 15

2-2 Binary alloy phase diagrams for zirconium-bismuth .......... 18

2-3 Binary alloy phase diagrams for zirconium-lead ............. 19

2-4 Binary alloy phase diagrams for zirconium-tin .............. 20

2-5 Uranium dioxide thermal conductivity vs. temperature ........ 24

3-1 Barnstead-Thermolyne furnaces for testing samples .......... 50

3-2 Loss of wall thickness, lead-bismuth samples tested at
1215F for 24 hours .................... ........... 55

3-3 Loss of wall thickness, lead-bismuth samples tested at
13820F for 24 hours .................... .......... 56

3-4 Loss of wall thickness, lead-bismuth samples tested at
1517F for 24 hours ................................. 57

3-5 Loss of wall thickness, lead-bismuth-tin samples tested at
12150F for 24 hours ................. ................ 58

3-6 Loss of wall thickness, lead-bismuth-tin samples tested at
1382F for 24 hours ................. ................ 59

3-7 Loss of wall thickness, lead-bismuth-tin samples tested at
1517F for 24 hours ................................. 60

3-8 Loss of wall thickness, lead-bismuth samples tested at
7500F for 1000 hours ................................ 61

3-9 Loss of wall thickness, lead-bismuth-tin samples tested at
750F for 3500 hours ................................ 62

3-10 Photomicrograph of reaction layer, lead-bismuth sample,
750F for 1000 hours ................................ 65

3-11 Photomicrograph of reaction layer, lead-bismuth-tin sample,
7500F for 1000 hours ............. ................... 66

3-12 Electron beam microprobe reaction layer analysis
lead-bism uth sam ple ................................ 67

3-13 Electron beam microprobe reaction layer analysis
lead-bism uth-tin sam ple .............................. 68

3-14 Optical photomicrograph and electron beam microprobe
results of lead-bismuth-tin sample with UO2 pellets at
1500F for 24 hours ................................. 70

4-1 TRUMP computer model for eccentric rod study ............ 77

4-2 Eccentric rod study with and without gas blanketing ........ 78

5-1 Fuel temperature profile vs. gap conductance (6 kW/ft) ....... 82

5-2 Fuel temperature profile vs. gap conductance (13 kW/ft) ...... 83

5-3 Transient response to a simulated LOCA ................. 88

5-4 Zirconium-water reaction rate constant vs. clad temperature .... 90

5-5 LBLWR fuel response to 15% transient overpower ........... 92

5-6 Cosine axial power shape ............................ 94

6-1 Program logic flow diagram for ESCORE ................. 101

6-2 Rod internal pressure vs. time for unmodified PWR
fuel rod ................... ..................... 109

6-3 Westinghouse 15x15 LBLWR fuel rod dimensions .......... 111

6-4 Westinghouse 15x15 LBLWR fuel average gap conductance . 112

6-5 Westinghouse 15x15 LBLWR fuel fuel temperatures ........ 114

6-6 Westinghouse 15x15 LBLWR fuel rod internal pressure ...... 115

6-7 Westinghouse 15x15 LBLWR fuel fission gas release ....... 117

6-8 Westinghouse 15x15 LBLWR fuel clad strain at EOL ........ 119

6-9 BWR 8x8 LBLWR fuel rod dimensions .................. 121

6-10 BWR 8x8 LBLWR fuel average gap conductance .......... 123

6-11 BWR 8x8 LBLWR fuel fuel temperatures ................. 124

6-12 BWR 8x8 LBLWR fuel rod internal pressure .............. 126

6-13 BWR 8x8 LBLWR fuel fission gas release ................ 127

6-14 BWR 8x8 LBLWR fuel clad strain at EOL ................ 128

7-1 Proposed LBLWR fuel pellet design for optimized performance . 133

7-2 Temperature contours, solid pellet, gas bonded ............. 136

7-3 Temperature contours, solid pellet, liquid bonded ............ 137

7-4 Temperature contours, annular pellet, gas bonded ........... 138

7-5 Temperature contours, annular pellet, liquid bonded .......... 139

7-6 Temperature contours, annular, grooved pellet, gas bonded .... 141

7-7 Temperature contours, annular, grooved pellet,
liquid bonded ................... ....... ........... 142

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Richard Frederick Wright

December 1994

Chairman: James S. Tulenko
Major Department: Nuclear Engineering Sciences

Light water reactor (LWR) fuel performance is limited by thermal and

mechanical constraints associated with the design, fabrication, and operation of

fuel in a nuclear reactor. These limits define the lifetime of the fuel, the maximum

power at which the fuel can be operated, the probability of fuel structural failure

during the fuel lifetime, and the transient performance of the fuel during an

accident. The purpose of this study is to explore one technique for extending

these limits; liquid metal bonding of LWR fuel. Current LWR fuel rod designs

consist of enriched uranium oxide (UO,) fuel pellets enclosed in a zirconium alloy

cladding. The space between the pellets and the cladding is filled at beginning-of-

life by an inert gas (typically helium). This gas space allows for the thermal

expansion and swelling of the fuel, fission gas release, as well as the creepdown

of the clad; additionally, the gap allows the fuel pellets to be inserted into the fuel

rod during the fabrication process. Due to the low thermal conductivity of the gas,

the gas space thermally insulates the fuel pellets from the reactor coolant outside

the fuel rod, elevating the fuel temperatures.

Filling the gas space with a high conductivity liquid thermally "bonds" the fuel

to the cladding and eliminates the large temperature change across the gap. The

resultant lower fuel temperatures directly impact fuel performance limits and

transient performance. Liquid bonding of liquid metal reactor (LMR) fuel has been

used in several research reactors, as liquid sodium is used as the bonding liquid

to limit the peak temperatures during normal operation, and to reduce the stored

energy in the fuel pellets.

The application of liquid metal bonding techniques developed for the LMR

metal fuel program to LWR fuel are explored for the purposes of increasing LWR

fuel performance and safety. An assessment of the technical feasibility of this

concept is presented, including the results of research into materials compatibility

testing and the predicted lifetime performance of Liquid Bonded LWR fuel. A fuel

performance analysis computer program, based on the ESCORE light water

reactor fuel performance code, has been developed and is used to determine the

benefits of liquid metal bonding for light water fuel.

The results of these studies show that the liquid metal bond is compatible with

the cladding and fuel pellets, and could decrease the likelihood of clad failure over

the lifetime of the fuel when compared to conventional gas-bonded fuel rods.

Further studies with the fuel performance code show that the benefits of lower fuel

temperatures over the lifetime of the fuel indicate that this fuel design is safer than

conventional fuel designs, and enhances light water reactor safety and



In an effort to enhance the safety and performance of water reactors, the

development of various innovative fuel designs was explored. Since many of the

safety concerns associated with nuclear reactor fuel deal with high fuel

temperatures, a new fuel rod design that operates at lower temperatures, for a

given power level, would be inherently safer.

Light Water Reactor Fuel Temperatures and Limits

Current light water reactor (LWR) fuel rod operational limitations include

thermal/mechanical limits such as cladding stress and strain, fuel rod internal

pressure, and maximum fuel temperature. These limits result largely from thermal

characteristics of the fuel operated at high linear power levels (kW/ft), large

temperature differences resulting from the poor thermal conductivity of oxide fuel,

and the large temperature drop across the pellet/clad gas gap. The limits define

such factors as the maximum permitted power at normal operation and fuel

temperature margin to melting during anticipated reactor transients.


Due to these high operating temperatures, high energy stored in conventionally

designed light water reactor fuel rods significantly increases the likelihood of fuel

damage during loss of coolant events.

The thermal resistance for heat transfer from the fuel pellet to the coolant for

a typical LWR fuel rod at the beginning-of-life is made up of 1) thermal conductivity

through the fuel pellet (53%), 2) thermal conductivity through the gas gap (35%),

3) thermal conductivity through the cladding (4.7%), and 4) the film drop between

the clad surface and the coolant (7.3%). The ability to transfer heat out of the fuel

rod can be influenced most by modifying the fuel pellet design, or reducing the

thermal resistance across the gas gap. The heat transfer through the fuel pellet

can be enhanced by either increasing the fuel thermal conductivity (i.e. changing

from UO2 to another fuel type), or by decreasing the pellet diameter. Since neither

of these alternatives were deemed acceptable without fundamentally changing the

fuel design, reducing the large thermal resistance associated with the

pellet/cladding gap by replacing the gas between the pellets and cladding with a

liquid metal was explored.

In addition, the benefits of lower fuel temperatures, integrated over the life of

the fuel rod, result in significantly lower fission gas release and fuel swelling. Both

of these factors positively impact the fuel performance at the end-of-life, and could

be useful for extending the fuel to higher burnup levels than those achieved by

conventional fuel.


Sodium Bonded Metal Fuel Technology

Fuel for liquid metal reactors (LMRs) is similar to LWR fuel for most operating

plants. The rods contain fuel pellets made of uranium or plutonium oxide encased

in a stainless steel cladding. An inert gas, typically helium, fills the interstitial

spaces between the pellets and the cladding. In order to develop an inherently

safer reactor design, some LMRs such as the Experimental Breeder Reactor

(EBR-lI) in Idaho Falls [1], use a metallic fuel consisting of

uranium-plutonium-zirconium alloy which is formed into pellets and encased in

stainless steel cladding. The metallic fuel has a much lower melting temperature

(2000F) compared to the oxide fuel (4500F), and a much higher thermal

conductivity (20 Btu/hr-ft-F) compared to the oxide fuel (4 Btu/hr-ft-F). Thus, the

centerline temperature of the metallic fuel is far lower than the oxide fuel for

comparable power and fuel dimensions.

To operate the fuel at acceptable power levels while maintaining margin to the

fuel melting temperature, it was determined that the high thermal resistance

between the pellet and the cladding across the gas gap must be significantly

reduced. To accomplish this, liquid sodium was introduced into the gap, effectively

eliminating the gap resistance [1]. The resulting fuel design was found to operate

safely at high power levels, and to maintain fuel temperature safety margins. In

the event of a fuel rod failure, the liquid sodium inside the rod would mix with the

liquid sodium coolant, and with the exception of the loss of fission product

retention, the rod would maintain its operational integrity.

Disadvantages to using this fuel design are mainly due to fuel manufacturing

and handling, and the lower fission gas retention capability of the metallic fuel.

Extreme care must be taken to isolate the liquid metal from the environment

outside the reactor as sodium reacts violently with air or water.

Liquid Bonding in LWR Fuel

The use of a liquid metal bond in a light water reactor fuel rod would enhance

the heat transfer between the fuel and the reactor coolant, resulting in significantly

lower operating temperature, and a safer fuel design. For this reason, it is

proposed that liquid bonding techniques be investigated for possible use in LWR

fuel design. Several advantages and disadvantages for the proposed design can

be cited.

Potential Benefits

The safety benefits resulting from lower fuel operating temperatures that

influenced the development of liquid bonded LMR fuel can be applied to LWR fuel

design. In order to achieve the high power levels and long fuel life needed in

power reactors, fuel temperature considerations become the principal design


limitation. As was previously stated, a large percentage of the thermal resistance

to removing heat from the fuel occurs in the gas gap between the fuel pellets and

the cladding. By replacing the gas with a liquid, the resistance is dramatically

reduced, and the fuel temperature is significantly lower for the same power level.

The lower radial temperature profile leads to significantly lower stored energy in

the fuel pellet, which is of primary concern during reactor transients. Additionally

the lower temperature reduces the thermal expansion of the pellet and reduces the

fission gas release both of which enhance fuel performance. Both the steady-state

and transient thermal performance of liquid bonded LWR fuel is discussed in

greater detail in Chapter 5.


Most of the disadvantages associated with using liquid bonding techniques in

LWR fuel design stem from lack of developed technology, especially in the field of

materials research. As will be shown, the bonding liquid (liquid metal) must be

chemically compatible with reactor materials including fuel (UO2), cladding

(Zircaloy), coolant (water), as well as fission products, shims, etc.

In addition, for the commercial viability of any new fuel design, it must be able

to replace and coexist with existing LWR fuel. Factors such as nuclear

interactions, performance during reactor transients, propensity to fuel rod failure,

behavior after failure, fission gas release and resultant rod pressure,


manufacturability, and the effect of the liquid bond material on fuel assembly

parameters must be assessed.

The purpose of this study is to

1. Determine the technical feasibility of liquid bonded light water reactor fuel

through a qualitative discussion of candidate bonding liquids, fuel thermal

mechanical limits, fuel reliability, fuel response to postulated severe

accidents, and manufacturing techniques. Through this assessment,

candidate liquid metal(s) will be identified, and a plan implemented for

laboratory testing of the constituent materials, and the development of

analytical tools to determine the performance characteristics of the

proposed fuel design.

2. Demonstrate the material compatibility between the liquid metal bond

material and the Zircaloy-4 cladding through comprehensive testing of

candidate liquid metals at typical reactor operating temperatures, and

anticipated transient temperatures.

3. Demonstrate the performance of liquid bonded LWR fuel by developing an

analytical tool to determine the thermal-mechanical performance under

irradiation conditions.

4. Identify new fuel designs which take full advantage of the temperature

reduction benefits of the liquid metal bond.


The following chapters provide the results of research into each of these

aspects of liquid bonded light water reactor fuel. Based on this research,

recommendations are made concerning the applicability of this advanced fuel

design in enhancing the safety and performance of commercial light water



The potential benefits of liquid bonded LWR (LBLWR) fuel can be realized if

the fuel can be shown to be technically and operationally feasible. This technical

feasibility depends on a number of factors. First, the choice of liquid must be

compatible with materials found in a light water reactor environment. Secondly,

the liquid must remain thermally stable; without experiencing chemical breakdown,

or changing phase over the anticipated temperature range and radiation

environment. In addition, interaction between the liquid and the neutron population

must be minimal. Also, the fuel must demonstrate a clear advantage over current

LWR fuel design, especially in the areas of fuel lifetime extension and safety.

Finally, the liquid bonded LWR fuel must be easy to manufacture and must be able

to replace and coexist with current LWR fuel in a reactor environment.

Choice of Bonding Liquid

Several criteria define the choice of a liquid for use in a LWR liquid bonded fuel

design. The most important of these is the ability of the liquid to maintain its heat

transfer characteristics over the anticipated steady-state operating temperature

range, as well as the expected transient temperature range. The liquid must not

experience chemical breakdown and must remain in the liquid phase over wide

ranges of temperature. In addition, the liquid must expand in a minimal fashion

upon freezing to prevent clad failure when the fuel is at low temperatures.

The selected liquid must coexist with other fuel materials, as well as the reactor

coolant, water. Chemical reaction with these materials over the defined

temperature range must be minimal. In addition, the liquid must have a minimal

impact on the nuclear environment in the reactor. Finally, the fission gases

released from the liquid bonded fuel must be accommodated so that the fuel rod

internal pressure is less than the reactor coolant system pressure at operating


Temperature Range Criteria

Typical LWR fuel operating temperatures range from the coolant temperature

(typically 600F) to the fuel centerline temperature (typically 2500F to 3000F).

This large temperature range is due to the low thermal conductivity of the UO2 fuel

and the significant temperature gradient across the gas gap. The fuel power rating

is limited, in part, by the fuel centerline temperature, which cannot exceed the

melting point of UO2, 47000F [2]. A major consideration for choosing a bonding

liquid is the temperature range over which the liquid remains in the liquid phase.

The material must be liquid at reactor operating conditions, before power is


produced in the fuel; i.e. the material, if solid at low temperatures, must liquefy in

the hot, no power expected temperature conditions of the reactor coolant.

The candidate liquid must have a relatively high boiling point and a low vapor

pressure to assure that it remains in the liquid phase at the highest expected

steady-state or transient fuel centerline temperature. This temperature range

extends from approximately 500F to 3500F. Many candidate liquids such as

water, organic, and molten salts cannot operate effectively over this large

temperature range. It is clear upon considering these limitations that the only

acceptable choices available are the liquid metals. Table 2-1 summarizes the best

choices from among the low melting point liquid metals [2]. Each of these metals

is adequate at the low temperature limit, as the melting points are near or below

600F. However, mercury, sodium, and potassium all boil at temperatures below

the high end of the temperature range and are therefore considered unacceptable.

Gallium, lead, bismuth, tin, lead-bismuth eutectic, lead-bismuth-tin ternary alloy,

and lithium are all acceptable choices based on the temperature range criteria.

Nuclear Interaction

Aside from the temperature criteria, it is also important that the bonding liquid

interact as little as possible with the neutron flux. In addition, the activation

products resulting from interaction between the neutron flux and the bonding liquid

must be evaluated to determine the extent to which they affect the fuel

performance and spent fuel handling. It is expected that the increased radiation

Table 2-1: Melting and Boiling Temperatures for Candidate Liquids

Liquid Melting Point (F) Boiling Point (OF)

Mercury -38 674

Gallium 85 4171

Lead 621 3181

Bismuth 520 3020

Tin 449 4711

Lead-Bismuth 257 3038

Lead-Bismuth-Tin 355 3311

Sodium 207 1623

Potassium 145 1399

Lithium 335 2447

from radioactive isotopes resulting from this interaction would be small when

compared to the activity of the fission products in the spent fuel.

Organic liquids such as Dowtherm are composed of complex molecules which

chemically disassociate in a radiation environment. Liquid metals have been

shown to remain chemically stable in high neutron fluxes which is witnessed by

their use both as reactor coolants and thermal bonding for fuel rods in LMRs.

Table 2-2 shows the thermal neutron cross section for each of the liquid metals

chosen on the basis of temperature [2]. Except for lithium, all of the remaining

candidate liquid metals exhibit very low absorption cross sections and would not

significantly affect the reactor neutron economy. Lithium's relatively high cross

section is due in large part to the presence of 6Li in natural lithium. If this relatively

scarce isotope (7.4 percent by weight) is removed, the lithium absorption cross

section is reduced to 0.370 barn. If desired, the 6Li in natural lithium could be

used as a burnable poison control mechanism to provide a more uniform power

distribution over the life of the fuel. For example, for a fuel rod containing 0.012

Ibm of natural lithium at beginning of life, the inventory of 6Li is reduced by

approximately one-third after one cycle (15,000 MWd/MT). Lithium-6 undergoes

an n-a reaction producing helium and tritium, both of which contribute to the

internal gas pressure.

The neutron absorption effects of lead, bismuth, tin and lithium-7 have been

found to be negligible, and account for fewer absorptions than the Zircaloy-4

Table 2-2

Average Thermal Neutron Absorption Cross Sections for Liquid Metals

Liquid Metal Number Density Micro Cross Macro Cross

(x10-24) Section (barn) Section (cm1)

Mercury .0407 380.00 15.500

Gallium .0511 2.80 0.143

Lead .0330 0.170 0.006

Bismuth .0281 0.034 0.001

Tin .0330 0.625 0.021

Lead-Bismuth .0303 0.094 0.003

Pb-Bi-Sn .0312 0.271 0.008

Sodium .0254 0.525 0.013

Potassium .0134 2.07 0.028

Lithium (natural) .0463 71.0 3.29

Lithium-7 .0460 8.04 0.370


cladding. After absorbing a neutron, bismuth transforms to polonium-210, a highly

radioactive isotope, but the increase in the activity due to the small amount of

polonium is not expected to be significant compared to the fission product activity.

The reduced fuel temperature effect of the liquid bonded fuel has a significant

positive safety effect from a reactivity standpoint. Lower temperature operation

causes a reduction in neutron absorption by uranium-238, as Doppler broadening

of the resonance absorption peaks is lessened. Also, as is shown in Figure 2-1,

the Doppler coefficient for a typical PWR rod in a thermal reactor is a strong

function of temperature [3]. Thus, for a PWR liquid bonded fuel at 9 kW/ft, the

Doppler coefficient is -1.2 x 10-5 Ap/oF compared to a value of -0.95 x 105 Ap/F

for gas bonded fuel. This difference (26% increase in the absolute value) means

that from a safety standpoint, the Doppler coefficient of the liquid bonded fuel will

respond more negatively per degree temperature increase. Since the Doppler

coefficient is one of the fuel's fastest acting safety mechanisms, an accidental

insertion of reactivity will be mitigated faster and more safely by the liquid bonded

fuel. In addition, the lower absorption of thermal neutrons by uranium-238 at

normal operating conditions improves the overall neutron economy.

The power defect contribution from the Doppler coefficient (hot/no power to hot/

power, 9 kW/ft) is much smaller for the liquid bonded fuel (.0005 Ap) than for the

gas bonded fuel (.0013 Ap), because of the smaller fuel temperature change.

Over fuel lifetime, the helium gas bonded fuel has greater production of fissile

plutonium due to the U-238 absorptions resulting from the temperature Doppler


---1.3-- -------
-1.3 --------- --------- f -_ __ _____-__ ________-------------

-1 .1 . .. . . .. .. . .-.. .. .. .. . .. - -.. .. . .. . ... -- . .. .. .. .... -

u 15 -- --- . . ----.--.. ------ --- --------.............
x -1 .6 ..-- .-- -------------.----.. -------------------------............ T -----------------......... ------------------ ------------------
-o -1.2


-1.7 1
400 600 800 1000 1200 1400 1600 1800 2000
Resonance Effective Temperature (F)

Figure 2-1: Doppler Coefficient vs. Effective Fuel Temperature at BOL [3]

broadening effect. This results in the reactivity curve being slightly less negative

for burnup of conventional, gas-bonded fuel than for the lower operating

temperature liquid bonded fuel. Thus, it is expected that a slightly higher

beginning of life boron concentration is required for the liquid bonded fuel to

achieve the same discharge burnup as the gas bonded fuel. Detailed calculations

have been performed to determine the exact neutronic (fuel management) effect

of the reactivity curve and indicate that the LBLWR fuel can operate within the

current control rod patterns and boron concentration ranges.

Material Compatibility

Of critical concern for the use of liquid metal in light water reactor fuel is the

compatibility of the bonding liquid with other LWR core materials. The materials

of importance include the coolant/moderator (light water), zirconium cladding, UO2

fuel, and fission products. The chemical reaction between the liquid and each of

these materials must be minimal over the expected temperature range.

Compatibility with the fuel and cladding is an obvious concern since the bonding

liquid wets both materials. Also important is the liquid bond reaction with the light

water coolant. In the event of a breach of the fuel cladding, violent reaction

between the bonding liquid and water would be unacceptable from a safety

standpoint. Further, the bonding liquid's final disposition and potential effect on the

primary coolant loops is important.


Some of the candidate liquid metals such as potassium and sodium react

violently with water at any temperature. The available literature [4,5] suggests that

lead, bismuth, lead-bismuth eutectic, and tin do not react vigorously with water in

the expected temperature range, and that lithium reacts moderately compared with

other alkaline metals. Gallium is expected to behave corrosively with zirconium at

the expected temperature range and it has been shown that zirconium is soluble

to some degree in lead, bismuth, and tin [6]. Figures 2-2, 2-3 and 2-4 show the

phase diagrams for zirconium/bismuth, zirconium/lead, and zirconium/tin,

respectively. Tests run at 932oF by Hodge, Turner and Platten [6], showed little

zirconium dissolution in the lead-bismuth eutectic under conditions similar to those

expected in an operating fuel rod. There is no indication that significant chemical

reactions occur between the other liquid metals (lead, bismuth, tin, and lithium) and

UO,, zirconium, and fission products.

Material compatibility tests were conducted by Thad M. Adams and Mark

Dubecky under the direction of Dr. Richard G. Connell, Jr., of the Materials

Sciences and Engineering Department and Dr. Glen J.Schoessow of the Nuclear

Engineering Sciences Department of the University of Florida to determine the high

temperature reaction characteristics of the liquid metal(s) chosen with the fuel rod

materials [7,8]. The tests were conducted with donated UO2 fuel and Zircaloy

cladding from the Babcox & Wilcox Fuel Company and SIMFUEL, a simulated

spent fuel, from Atomic Energy of Canada, Ltd. These tests are described in

greater detail in Chapter 3.

Atomic Percent Zirconium


Weight Percent Zirconium ZrZr)

Figure 2-2 Binary Alloy Phase iagram for Zirconium-Bismuth [6
10 20 30 40 50 60 70 80 00 100
Weight Percent Zirconium Zr

Figure 2-2: Binary Alloy Phase Diagram for Zirconium-Bismuth [6]


75 80 85 90 95 100
Weight Percent Zirconium

Figure 2-3: Binary Alloy Phase Diagram for Zirconium-Lead [6]

0 10 20

Atomic Percent Tin
30 40 50

60 70 80 90





M 1200.

E 1000




0 10 20 30 40 50 60
Zr Weight Percent Tin

... .... . ......... 1592C


/ 159820C
21.0 23.5

.18 13270C

1142 C

63 9823 C



Figure 2-4: Binary Alloy Phase Diagram for Zirconium-Tin [6]





-232*C ( 231)888 C

80 90 100

Fuel Rod Characteristics

It is important that any new LWR fuel design be able to replace and coexist

with existing LWR fuel. For this reason, liquid bonded LWR fuel rods for existing

reactors must be the same size as current LWR fuel, perform at or above current

fuel burnup limits, and, when formed into fuel assemblies, exhibit acceptable

handling characteristics during transportation and refueling.

For comparison purposes, PWR 15 x 15, PWR 17x17, and BWR 8x8 fuel rods

are evaluated as liquid bonded fuel rod reference designs. The pellet and cladding

dimensions, shown in Table 2-3, were chosen to be the same as the current

helium filled fuel rod. For this evaluation, the gas gap volume was assumed to be

filled with liquid metal. The size of this gap currently used in conventional gas-

bonded fuel designs is selected based on manufacturing considerations that allow

for pellet insertion, themal considerations which minimize the temperature rise

across the gap during reactor operation, and other considerations including the

accommodation of fission gas. Liquid bonded fuel pellets operate with no

significant temperature drop across the gap; thus, the gap size is not an important

consideration. The use of a liquid metal bond allows the gap size to be increased

to any value desired for ease of manufacturing without concern for the effect on

fuel temperature. As is shown in Chapter 6, however, other factors such as rod

weight and fission gas plenum requirements will influence the gap size and liquid


Table 2-3: Fuel Rod Parameters for PWR and BWR Fuel

Parameter Westinghouse Westinghouse General Electric
S 15x15 17x17 8x8

Pellet OD .3659 in .3225 in 0.41 in
Pellet Length 0.6 in 0.6 in 0.41 in
Cladding ID .3734 in .3290 in .419 in
Cladding OD .422 in .374 in .483 in
Fuel Length 144 in 144 in 150 in
Rod Length 150.45 in 150.45 in 159.6 in
Pre-pressure 545 psia 545 psia 45 psia
Theo. Density 94% 94% 96%
Pellet Volume .00876 ft3 .00681 ft3 .0115 ft3
Pellet Mass 5.959 Ibm 4.614 Ibm 7.868 Ibm
Clad Volume .00253 ft3 .00207 ft3 .00419 ft3
Clad Mass 1.03 Ibm 0.84 Ibm 1.706 Ibm
Gap Volume .000412 ft3 .000338 ft3 .000509 ft3
Liquid Bond Mass:
Lithium .0123 Ibm .010 Ibm .0283 Ibm
Pb-Bi .263 Ibm .216 Ibm .585 Ibm
Pb-Bi-Sn .235 Ibm .192 Ibm .521 Ibm
Total Assembly Weight
Helium 1429 Ibm 1367 Ibm 657.8 Ibm
Lithium 1432 Ibm 1370 Ibm 659.5 Ibm
Pb-Bi 1484 Ibm 1424 Ibm 694.1 Ibm
Pb-Bi-Sn 1478 Ibm 1418 Ibm 690.1 Ibm


Reduced fuel temperatures affect the fuel nuclear performance and material

properties. For example, as shown in Figure 2-5, the thermal conductivity of UO2

decreases with increasing temperatures over the temperature range of interest [9].

Thus, the liquid metal bonded fuel will operate in a more favorable UO2 thermal

conductivity range and will experience lower radial thermal gradients. The

combination of lower fuel temperatures and lower thermal gradients results in lower

thermal expansion and significantly lower thermal swelling. Cracking of the UO2

pellets is also expected to be significantly lower due to the lower operating

temperatures and temperature gradients thus improving fuel performance.

Irradiation testing illustrating these performance enhancements has been

performed on liquid sodium bonded fast reactor fuels using uranium metal, uranium

nitride, and uranium carbide [10]. Chemical reactions between UO2 and sodium

has precluded irradiation testing of liquid sodium bonded oxide fuel.

The total fission gas release rate is a function of the fuel burnup and the

temperature history of the fuel. Fuel maintained at a lower temperature has a

reduced gas release rate, and thus less gas is released over an equivalent fuel

lifetime. Since liquid bonding dramatically reduces the fuel temperature, less gas

is released, requiring less gas plenum volume to accommodate the fission gas.

However, the displacement of the liquid bond material as the cladding creeps down

during operation decreases the available fission gas volume. The gas volume

behavior is discussed in greater detail in Chapter 6.

- \
S1.8 ------ ------ -- --------- ------------- ----------- -----.-- .-..... -. .----------.--- ----------

1.6 -

1.4 ----.---
-j 4 - - .. ..... ... .. ..... ..... ... -. ------------^ ^

1.2--------- -I
1000 1500 2000 2500 3000 3500 4000
Temperature (F)

Figure 2-5: Uranium Dioxide Thermal Conductivity vs. Temperature [9]

It is important to maintain a path for the fission gas from the surface of the fuel

pellet to the gas plenum to avoid local "hot spots" caused by gas blanketing.

Experimental and analytical studies into gas blanketing are shown in Chapter 4.

A liquid bonded fuel rod will be heavier than a conventional LWR fuel rod, by

an amount equal to the weight of the bonding liquid. As is shown in Table 2-3, the

liquid bond increases the fuel rod weight by 0.012 Ibm for lithium, 0.235 Ibm for

lead-bismuth-tin, and 0.263 Ibm for lead-bismuth. The weight of the fuel assembly

is increased by 0.18% for lithium, 3.4% for lead-bismuth-tin, and 3.8% for lead-

bismuth. This increase would not significantly affect fuel handling capabilities.

As is shown in Table 2-3, assembly weight increases in a similar fashion for the

17x17 PWR fuel, and for the 8x8 BWR fuel.

Best Candidates

Based on expected operating temperature range, nuclear interaction, material

compatibility, and fuel rod characteristics, the choice of the bonding liquid comes

down to three candidate liquid metals; lead-bismuth eutectic, lead-bismuth-tin, and

lithium. Each candidate exhibits a low melting temperature to assure liquid

behavior during reactor start-up and operating conditions. Each exhibits a high

boiling point temperature and low vapor pressure to assure liquid behavior at

operating conditions and expected transients. Each does not significantly affect

the core neutron economy, and each has been singled out as having a high

degree of chemical compatibility with other reactor materials at elevated


temperatures. Lithium exhibits a moderate reaction with the light water coolant.

Based on these criteria, the lead-bismuth-tin ternary alloy and the lead-bismuth

eutectic are considered the best candidates, and lithium is considered a backup


Thermal Considerations

As has been shown, the key advantages of liquid bonded fuel is the lower fuel

temperatures associated with the reduced thermal resistance across the gap

between the fuel pellet and the cladding. The reduced thermal resistance has

three important consequences:

1. Steady-state operating temperatures are significantly lower when compared

with conventional fuel with low gap conductance.

2. Stored energy in the fuel is significantly lower. This condition leads to a far

lower peak cladding temperature in the event of a loss of coolant accident.

3. Lower fuel temperatures over the fuel lifetime result in lower fuel pellet

cracking due to lower thermal stress, reduced fission gas release, reduced

thermal expansion, and safer nuclear characteristics. In addition, the fuel

swelling is reduced and the plastic strain in the cladding is lower allowing

for higher fuel burnup.


An initial scoping thermal/hydraulic analysis of the liquid bonded LWR fuel was

performed, prior to detailed fuel performance calculations, to determine the

steady-state and transient characteristics of the fuel, and the advantages

compared with conventional LWR fuel. A qualitative discussion of the LBLWR fuel

temperature characteristics is presented below, while the details of this analysis

are presented in Chapters 5 and 6.

Steady-State Fuel Temperatures

The gas gap thermal resistance results in the high average fuel temperatures

and high thermal gradients observed in conventional LWR fuel. The thermal

gradients are due to the low thermal conductivity of the UO2 resulting from the high

operating temperatures as is shown in Figure 2-5. The temperature drop across

the gap is a strong function of the gap conductance which comprises three

components; conduction through gas, conduction at contact points between the

cladding and the fuel pellet, and radiation heat transfer from the fuel surface to the

inside surface of the cladding. A significant amount of research has been

performed to characterize the gap conductance [11,12,13], which varies as a

function of the fuel burnup. Typically, for PWR fuel, the gap conductance ranges

from 500 Btu/hr-ft2'-F to 3000 Btu/hr-ft2-oF [9] and is a strong function of the gap


For liquid bonded fuel, the sole path for radial heat transfer in the fuel gap

region is conduction through the liquid metal bond. For lead-bismuth with a


thermal conductivity of 8 Btu/hr-ft-F, the gap conductance is 30,000 Btu/hr-ft2-oF.

Thus the temperature drop across the gap, which ranges from 300F to 10000F for

a peak power conventional fuel rod over a typical range of gap conductance, is

negligible for the liquid bonded case.

The advantages of lower fuel temperatures during steady-state operation are

many. For a given power level, the margin to fuel centerline melting is

substantially increased. In addition, parameters such as fuel thermal expansion,

pellet stress and strain, and fission gas release, which are functions of the fuel

temperature, are all reduced. Most importantly, the fuel rod stored energy is

significantly lower, which greatly enhances the fuel transient performance.

Details of steady-state thermal analysis of the fuel rod are presented in

Chapters 5 and 6.

Transient Performance

The lower temperatures expected for LBLWR fuel are important for mitigating

the effects of reactor transients. Specifically, the lower stored energy in the

LBLWR fuel rod reduces the rod heatup associated with loss of coolant and loss

of flow transients. In addition, the large margin between operating fuel

temperatures and the fuel melting point makes it more likely that the rod will

survive power excursions resulting from local reactivity insertion accidents without

undergoing a center melt condition.

The transient performance of LBLWR fuel is assessed in Chapter 5.


Thermal/Mechanical Limits and Design Criteria

Several criteria have been identified [9] with regard to fuel rod

thermal/mechanical design, and fuel performance limits are determined by these

criteria. A qualitative discussion of these criteria, and how they are affected by

the use of liquid bonding follows:

1. Rod Internal Pressure Criterion. "The internal pressure of the highest

power rod in the reactor will be limited to a value below that which could

cause the diametral gap to increase due to outward cladding creep during

steady-state operation and extensive departure from nucleate boiling

(DNB) propagation to occur" [9:66]. The basis for this criterion is to assure

that the diametral gap will not increase causing a decrease in the cooling

water flow area between adjacent rods which will decrease the local heat

transfer coefficient causing an approach to DNB conditions. Though the

DNB condition resulting from flow blockage is still a concern for LBLWR

fuel, the gas pressure in the fuel rod can be maintained at levels similar

to current fuel designs. It is important to note that clad creep-down during

the beginning of life can displace the liquid metal into the gas plenum,

decreasing the volume needed for fission gas accumulation. Design

features such as larger plena, optimized liquid bond loading, and lower rod

pre-pressurization are necessary to accommodate the gas pressure. In

addition, the effects of the liquid metal on the cladding will either decrease


or increase the resistance to creep. This effects are evaluated in Chapter


2. Clad Strain Criterion. "For steady-state operation the total tensile creep

strain is less than 1 percent from the unirradiated condition. For each

transient event the circumferrential, elastic plus plastic strain shall not

exceed a tensile strain range of 1 percent from the existing steady-state

condition" [9:66]. Provided that liquid metal does not react with and alter

the properties of the cladding, there will be no difference between the

performance limits due to clad strain for conventional and liquid bonded

fuel. In addition, lower fuel temperatures will reduce fuel thermal

expansion which, in turn, reduces the pellet-cladding interaction. These

effects are evaluated in Chapter 3.

3. Clad Stress Criterion. "The volume average effective clad stress shall not

exceed the tensile yield strength of the clad material. This criterion arises

from local pellet-cladding interaction due to thermal expansion of the fuel"

[9:67]. Because liquid bonded fuel operates at significantly lower

temperatures, the thermal expansion is reduced, and fuel performance

limits due to clad stress are significantly improved.

4. Clad Temperature Criterion. "The clad surface temperature (oxide-to-metal

interface) shall not exceed 750F for steady-state operation, and 8000F for

short-term transient operation" [9:67] As clad surface temperatures are a

function of the rod power, clad surface area, and the bulk fluid conditions,

fuel performance limits due to clad temperature are the same for liquid

bonded and conventional fuel.

5. Fuel Temperature Criterion. "The maximum fuel temperature shall be less

than the melting temperature of the fuel" [9:68]. As is shown by the

thermal analysis in Chapter 5, the peak operating temperature of the liquid

bonded fuel is significantly lower than conventional fuel, and provides

much more margin to fuel melting. Fuel performance parameters which

are strong functions of fuel temperature such as fission gas release, pellet

thermal expansion, fuel cracking, and pellet/cladding interaction are also

dramatically improved.

Other criteria deal with clad fatigue, plenum spring support, clad flattening due

to axial gaps, and axial rod growth. Fuel performance limits associated with these

criteria are not significantly affected by the presence of a liquid metal bond.

The results of this qualitative discussion of thermal/mechanical fuel

performance limits indicate that the potential for significant improvement exists

from liquid bonded fuel, especially in the areas of maximum fuel temperature and

pellet-clad interaction.

Fuel Rod Failure

To be considered a viable design option, the failure probability of liquid bonded

fuel must be less than or equal to that of conventional fuel. In addition, the


consequences of failure and the effect a liquid bonded fuel rod failure on the core

integrity and the reactor primary coolant system must be shown to be minimal.

As was discussed previously, cladding failure due to thermal/mechanical

considerations is mitigated by the lower operating temperatures in the liquid

bonded fuel rod, which reduce the onset of hard pellet-clad interaction. The

interaction between the liquid metal bond and the cladding must be assessed to

determine whether cladding integrity can be reliably maintained over the fuel

lifetime. Chapter 3 discusses the results of material compatibility tests performed

to characterize the liquid metal interaction with the clad material.

In the event of a failure of a liquid bonded fuel rod during reactor operation,

high temperature water at 2200 psia for PWRs, and 1050 psia for BWRs will

contact the liquid metal. The metal water reaction is of the form

M + xH20 MO, + xH, + Qe (2-1)

The heat of reaction, Qr,,a, varies for each liquid metal. Alkaline metals such

as lithium, sodium, potassium, and cesium, exhibit the most vigorous reaction,

ranging from moderate for lithium, to explosive for cesium. By comparison, lead

bismuth, and tin react in a relatively benign manner, and have been singled out for

use as reactor coolants [14].

Severe Accident Analysis

As discussed previously, liquid bonded fuel is less likely to experience a loss

of fuel rod integrity in the event of a severe accident. This better performance is

primarily due to the lower fuel operating temperatures and the correspondingly

lower stored energy in the fuel rod. However, a severe accident in a reactor core

consisting of both liquid bonded fuel and conventional fuel could expose the liquid

bonded fuel to temperatures above the zirconium metal/water reaction threshold

(1700F), and cause a loss of fuel integrity. What effect the presence of the liquid

bonded fuel has on the accident progression and overall severity must be

determined. A qualitative discussion of the behavior of liquid bonded fuel in a

severe accident is presented.

A Class IX accident in a LWR is defined as an event which falls beyond the

plant design basis. This event involves, in general terms, loss of core cooling, and

loss of active accident mitigation systems (emergency feedwater, core sprays,

etc.). This results in core uncovery, and subsequent core degradation. Factors

such as the speed at which the core uncovers (size of a primary system break),

total stored energy in the fuel assemblies at the time of uncovery, decay heat

levels, and the availability of engineering safeguard features (emergency core

cooling system and residual heat removal), determine the severity of the accident.

For a large break loss of coolant accident (LOCA), core uncovery occurs while

the fuel is essentially at operating temperature. During this phase, the liquid


bonded fuel, which operates at a lower temperature, would retard the overall heat

up of the core. Conventional fuel rods with high stored energy would experience

zirconium-water reaction due to cladding surface temperatures in excess of

17000F. As a result of the exothermic nature of this reaction, large amounts of

heat are generated and concentrated in the vicinity of the fuel pellets, causing fuel

melting and relocation, and the evolution of hydrogen gas.

In a mixed core, as the conventional fuel rods heat up, the adjacent liquid

bonded assemblies will experience an increase in cladding surface temperature

due to radiation heat transfer from the hot neighboring assemblies. In addition,

molten cladding and fuel relocating from disassociated conventional fuel rods could

contact adjacent liquid bonded fuel rods and induce failure. After the LBLWR fuel

cladding is breached, the bonding liquid metal is expelled and is added to the

disassociated core material. The bonding metal could experience a metal-water

reaction, as discussed previously, generating heat and additional hydrogen gas.

Compared to the energetic reaction between steam and zirconium, the oxidation

of the liquid metal bond is not expected to add significantly to the heat and

hydrogen generated.

After the core material is relocated to the bottom of the reactor vessel, the

reactor pressure vessel wall is thermally attacked and fails. The core material is

deposited in the reactor containment, along with the non-condensible gases

(hydrogen) generated during the accident. The hydrogen gas could ignite (in

non-inerted containments) or explode causing pressure spikes inside the

containment. The buildup of hydrogen along with other non-condensible gases

generated by the core material attacking containment structures could

overpressurize and fail the containment, releasing radioactivity to the environment.

From this discussion, there are two major conclusions that can be drawn on the

effects of liquid bonded fuel on Class IX accidents:

1. Liquid bonded fuel will lower the stored energy in the core. This causes the

core to heat up less rapidly, and allows more time for operator mitigation.

The higher the percentage of liquid bonded fuel in the core, the less severe

the heat up of the core is likely to be following a large break LOCA.

2. After all the fuel is failed, the liquid bonded fuel will contribute additional

heat and hydrogen generation due to the liquid metal reaction with water.

Since the volume of liquid metal is much less than the zirconium, and the

reaction is less vigorous, it is expected that this effect is not significant.


Nuclear reactor fuel manufacturing has advanced to a highly automated state.

To be considered a viable commercial option, the liquid metal bonded fuel must

also lend itself to ease of manufacturing. After the fuel rods are sealed, the

differences between the conventional and liquid bonded fuel must be minimal.

Inherent differences such as somewhat higher weight per fuel rod and assembly

must be evaluated to determine whether fuel transportation equipment, refueling

equipment, and in-reactor support structures are impacted.

A brief description of current LMR metal fuel manufacturing techniques is

discussed, and possible application to current LWR fuel manufacturing is


The manufacture of liquid sodium bonded metal fuel for liquid metal reactors

is a complex procedure [1]. The pellets are stacked into the cladding tubes at

room temperature and under "clean room" conditions. These tubes are sealed at

one end, and the open end is attached to a vacuum pump through a tee fitting.

After evacuating all gas from the tube, the fuel rod is heated to a temperature

above the melting temperature of sodium (208F), and the other end of the tee

fitting is connected to a liquid sodium fill tank. The filling valve is opened and the

tube is back-filled with liquid sodium. The tube is cooled and the end cap is

welded in place to seal the rod. The liquid metal freezes upon cooling, but

completely fills the interstitial spaces between the fuel pellets, and between the

pellet stack and the cladding. During reactor start-up, the fuel temperature

increases, and the liquid metal melts. When the coolant temperature reaches the

melting temperature of the liquid metal, the liquid metal is completely melted, and

the fuel rod operates as designed.

The manufacture of liquid bonded LWR fuel could be handled in much the

same way as the liquid sodium bonded LMR fuel described above. Some

reworking of existing LWR fuel manufacturing equipment and methods would be


required to handle liquid metals. Even so, the manufacturing of liquid metal

bonded light water reactor fuel is technically feasible, and is capable of being

automated to a high degree.

It is proposed that a simpler technique be considered which involves placing all

or a portion of the required bonding material in the form of a solid cylinder below

the fuel pellet stack. The pellet hold-down spring is held in compression, and the

fuel rod is evacuated. Upon heating, bond material liquifies and is forced into the

diametral gap between the pellets and the cladding. The rod is then back-filled

with helium and sealed.

Machining tolerances associated with the fuel pellets and cladding dimensions

are extremely important for maintaining a predictable gap dimension and resulting

thermal characteristics. No such constraint is placed on the liquid metal bonded

LWR fuel due to the uniformly low thermal resistance across the gap.

Results of the LBLWR Feasibility Study

The results of this preliminary feasibility study indicated that LBLWR fuel

exhibits sufficient merit to warrant further research. To accomplish this, a study

was funded by the Department of Energy to conduct research in the following


1. Laboratory testing of candidate liquid metals through material compatibility

testing. This work is summarized in Chapters 3 and 4.

2. Detailed calculation of fuel rod steady-state and transient behavior.

Calculations which integrate the effects of burnup dependent parameters

such as fission gas release and fuel dimensional changes, to determine the

fuel rod performance over a typical lifetime in a light water reactor. This

work is summarized in Chapters 5, 6, and 7.


As a result of the feasibility study, two candidate liquid metals, lead-bismuth

eutectic (55.2w/oBi-44.8w/oPb) and a lead-bismuth-tin alloy (33w/oPb-33w/oBi-

33w/oSn), were chosen to experimentally determine material compatibility between

the liquid metals, Zircaloy-4 cladding, and the UO2 pellets. For the purpose of light

water reactor fuel, the compatibility between these materials is determined by the

degree of reaction between the liquid metal, cladding, and pellets, synergistic

effects of the three materials, and changes in the properties of the materials which

would affect its function and the operation of the fuel rod.

This work was performed by Thad M. Adams and Mark Dubecky under the

direction of Dr. Richard G. Connell, Jr., of the Materials Sciences and Engineering

Department and Dr. Glen J.Schoessow of the Nuclear Engineering Sciences

Department of the University of Florida [7, 8]. This work was accomplished by

exposing the cladding material to the liquid metals at elevated temperatures for

extended periods of time. The loss of wall thickness occurring in the cladding, and

the degree of chemical reaction between the cladding and the liquid metal were

determined. The results showed that lead-bismuth-tin alloy gave the best

compatibility performance. A synopsis of this work is presented.

Discussion of Liquid Metal Attack

The phenomenon of liquid metal attack needs to be addressed differently from

the standard idea of corrosion. Typically, corrosion is referred to as the chemical

or electrochemical (galvanic) deterioration of a metal. However, when discussing

liquid metal attack, this concept must be expanded to include solution of the solid

metal in the liquid metal, the degree of attack being dependent upon the solubility

in the liquid metal [15].

Solubility is an important factor in determining the extent of liquid metal attack

of solid metals. Although simple solution of solid metals in liquid metals does

occur, the majority of the attack associated with liquid metal corrosion involves

more complicated concepts of solution and solubility. Solubility does seem to

govern the rate of liquid metal attack on solid metals. Through early experimental

work, it was determined that there are six basic types of liquid metal attack:

simple solution, alloying/intermetallic compound formation, intergranular

penetration, impurity reactions, temperature gradient mass transfer, and

concentration gradient mass transfer [15, 16].

Simple solution of a solid metal by a liquid metal consists of the removal of

surface metal from the solid until the solubility limit for the solid-liquid metal system

is reached [16]. From the phase diagram for the solid-liquid metal system, one

can predict the amount of solid metal that can be dissolved which, in turn, can be

related to the amount of damage to the solid material. As expected, there is a

strong coupling between the surface area of solid metal and the volume of liquid


metal [16]. In general, the smaller the volume of liquid metal, the less the depth

to which attack can occur in the solid metal. In the simplest case, the attack will

continue until the liquid metal becomes saturated with solute from the solid metal.

Although solubility curves for the solid-liquid metal systems can give an accurate

account of the amount of damage, they cannot supply any information as to the

kinetics or rate of the solution process taking place. The kinetics of the process

are of great concern since the liquid metal will be exposed to the cladding in liquid

form for 4-5 years at temperatures over 6000F.

The formation of an intermetallic compound at the solid-liquid metal interface

may be either beneficial or deleterious. Intermetallic layers forming at the solid-

liquid metal interface can act as diffusion barriers that retard further deterioration

of the solid by the liquid [17]. Additionally, metallic barrier layers are added to

LWR fuel in order to improve pellet-clad interaction performance. On the other

hand, since many intermetallic compounds are more brittle than the metal

substrate, they can act to reduce the strength and/or toughness of the substrate


Intergranular penetration/liquid metal embrittlement results from the preferential

attack on the grain boundaries of the solid by liquid metal. Liquid metal

preferential attack of grain boundaries is surface tension driven and causes the

removal of solid metal along the grain boundaries by dissolving solid metal in the

liquid metal [15, 16, 18]. From C. S. Smith's paper on interfaces and grains, it was

shown that for a dihedral or spreading angle of 60O or less, a liquid will wet free


surfaces and penetrate toward the interior along grain boundaries [19, 20]. This

particular manifestation of liquid metal attack is insidious; while no signs of

apparent damage such as material loss or dimension change may be discernible,

liquid metal embrittlement of the solid may reduce strength to such an extent that

catastrophic failure occurs. Electron beam microprobe scanning is one of the

methods used to detect such attack.

Impurities such as oxygen, nitrogen, hydrogen, and carbon have a pronounced

effect on the reactions between solid and liquid metals [16, 18]. The most

pronounced effect that these impurities have on solid-liquid metal systems relates

to the kinetics of reactions, either increasing or decreasing the rate of attack. In

addition, these impurity elements can change surface tension properties and

suppress intermetallic compound formation.

The phenomenon of temperature gradient mass transfer can be related to a

special case of the simple solution process. Temperature gradient mass transfer

is observed in convection loops or heat exchanger tubes where the liquid metal is

in motion through a solid metal channel. In convection loops or heat exchangers,

some sections of the system are at higher temperatures than others. Because

solubility generally increases with temperature, solid metal dissolves in the liquid

in the hot zones, while in colder sections of the loop it may plate out. By such

a mechanism, partial or full blockage of coolant flow can occur. This phenomenon

is characteristic of non-isothermal, dynamic systems, and will not occur in


isothermal systems, or static systems such as a thin layer of liquid metal between

two concentric cylinders.

Concentration gradient mass transfer consists of solid metal dissolving into a

liquid metal, then diffusing through the liquid metal and alloying with another solid

metal [17, 18]. Concentration gradient mass transfer is most commonly seen in

static liquid metal corrosion tests where the solid container material becomes

alloyed with the test specimen or vice versa. The process is driven by a reduction

in Gibb's free energy as the two metals alloy [17, 18].

The testing of liquid metal attack can be done by two methods: static or

isothermal, and dynamic or non-isothermal. Static corrosion testing consists of

placing a solid metal specimen into a liquid metal bath at a specified temperature.

The specimen is exposed to the hot liquid metal for a prescribed period of time

and then evaluated to ascertain the degree of attack. For a dynamic test, a forced

or thermal convection loop is constructed to pump the liquid metal through the

container material in order to simulate a heat exchanger, reactor piping, or similar

component that is expected to experience temperature transients. Hot and cold

sections are purposely built into the loop in order to examine solubility effects such

as temperature gradient mass transfer.

The ability to determine quantitatively the amount of attack is often difficult [19].

The standard procedure used is to measure the weight loss or gain by the sample

after exposure to the corrosive environment. For the case of liquid metal attack,

simple weight loss/gain measurements may be misleading [19]. In order to


investigate the amount of attack when a solid metal is in contact with a liquid

metal, four measurements are considered:

1. Dimensional Changes

2. Compositional changes

3. Weight changes

4. Depth of attack

One or more of these measurements may be used to quantify liquid metal attack.

For this study, the dimensional changes and compositional changes at the liquid

metal-solid interface was used to quantify the degree of attack.

The purpose of this study was to experimentally determine through the use of

static, isothermal testing, the degree of attack experienced by the Zircaloy-4

cladding material when exposed to the candidate liquid metals at temperatures

indicative of

1. Standard operating program (SOP) conditions. Temperatures expected

during hot, full power operation of the fuel (750F) for extended periods of


2. Limiting accident conditions. For reactor fuel, the highest temperatures

expected during a design basis event are associated with loss of coolant

accidents (LOCAs), where heat transfer to the coolant is significantly

decreased leading to a rapid increase in the cladding temperature due to

the stored energy in the fuel. For these tests the temperatures ranged from

12000F to 15000F for short periods of time. High temperature exposure for


short intervals was also viewed as a way to study accelerated attack since

testing over a typical fuel lifetime (35,000 hours) was impractical.

Experimental Assessment of Bonding Liquid/Claddinq Compatibility

The experimental studies at the University of Florida were performed over the

course of two years in a joint effort between the Materials Science and Engineering

Department, and the Nuclear Engineering Sciences Department. A discussion of

the experimental procedures, including the materials, sample preparation, test

matrix, determination and characterization of liquid metal attack is presented, as

well as the interpolation of these results to determine the feasibility of the proposed

liquid bonded LWR fuel design.

Materials Used in Experimental Samples

The Zircaloy-4 cladding used for this investigation was supplied by Babcock &

Wilcox Nuclear Technology of Lynchburg, Virginia. The composition of this

material is in accordance with the ASTM specification B350. Table 3-1 shows the

specifications for reactor grade Zircaloy-4. The cladding was provided in 12 inch

tube sections which were subsequently cut into 6 inch sections and sealed by

welded stainless steel (type 304) end caps. The material provided consisted of

Table 3-1 ASTM B350 Chemical Composition Specification for Reactor Grade


Sn 1.20-1.50
Fe .18-.24
Cr .07-.30
Fe+Cr .28
O .10-.15
C .010-.018
Si .007-.012
Zr balance

Table 3-2 Chemical Composition of Lead-Bismuth Eutectic Supplied by Cerro
Metal Products, Inc.


Cu 0.0001
Sb+Sn 0.005
Ni 0.0001
Pb 55.2
Bi 44.8


B&W 15x15 cladding (0.430 in. OD, 0.030 in. wall thickness), and B&W 17x17

cladding (0.375 in. OD and 0.025 in. wall thickness).

Eutectic lead-bismuth was purchased from Cerro Metal Products of Bellafonte,

Pennsylvania, and was supplied as 0.25 in. diameter rod. The chemical

composition of the lead-bismuth is shown in Table 3-2.

Alumina pellets were used to simulate fuel pellets in the cladding compatibility

tests. Additional tests using depleted UO2 pellets donated by Babcox & Wilcox

Fuel Company to demonstrate the compatibility of the liquid metals with UO2, and

SIMFUEL simulated spent fuel donated by Atomic Energy of Canada, Ltd. were

also used to determine compatibility with UO2 and fission products. The pellet

diameter is 0.366 in. for 15x15 fuel, and 0.322 in. for 17x17 fuel.

The ternary lead-bismuth-tin alloy used in this research was prepared from the

eutectic lead-bismuth with additions of tin and lead stock. Tables 3-3 and 3-4

show the chemical compositions of the tin and lead stock, respectively. The lead-

bismuth-tin alloy was produced by melting on a hot plate under flowing helium

cover gas the proper amounts of lead-bismuth, tin, and lead in order to provide a

ternary alloy composed of 33wt% Pb, 33wt%Bi, and 33wt%Sn. The approximate

melting temperature of the alloy is 2430F [20, 21].

Sample Preparation

Zircaloy-4 cladding sections approximately 6 inches in length were sealed by

welding a stainless steel end cap at one end. These sections were then filled with

Table 3-3 Chemical Composition of the Tin Stock Supplied by Ames Metal
Products, Inc.


Pb 0.0106
Sb 0.0036
Fe 0.055
Bi 0.0033
Al 0.001
S 0.001
Sn Balance

Table 3-4 Chemical Composition of the Lead Stock Supplied by Fisher Scientific


Cu 0.001
Sb+Sn 0.005
Bi 0.0001
Pb Balance


one of the candidate liquid metal alloys. The tube sections were not filled

completely so as to allow for the volumetric expansion of the liquid metal alloy

when heated. Furthermore, from the study of past research in liquid metal attack,

it was determined that there is a strong surface area-to-volume effect for liquid

metals contacting solid metal surfaces [16]. In order to account for this effect, two

studies were made, namely: tubes filled completely with liquid metal to represent

a worst case scenario, and tubes containing simulated fuel pellets made of alumina

(AI20) with liquid metal filling the gaps between the pellets and the cladding. A

second stainless steel endcap was fitted into place to seal the tubes. Additional

tests were run on tubes filled with UO2 pellets to determine the compatibility

between UO2 and the liquid metal.

To minimize oxidation of the exterior of the tubes, the experiments were

conducted in a helium atmosphere. The samples were loaded into the Barnstead-

Thermolyne resistance wound tube furnace (Figure 3-1), and heated to the desired

temperature. The samples remained at the target temperature for a prescribed

period of time and were then allowed to cool to room temperature under a positive

pressure of helium. The samples were then removed for analysis.

Test Matrix

Samples were subjected to two different temperature-time histories;

representing standard operating program conditions (SOP), and typical loss of

coolant accident (LOCA) conditions. SOP involves the day-to-day operation of the

- .e

Figure 3-1: Barnstead-Thermolyne Furnaces for Testing Samples


reactor during which liquid bond temperatures are expected to remain at

approximately 7500F for the life of the fuel (30,000-40,000 hours with shut downs

for scheduled refueling). To simulate the SOP conditions, samples were tested at

7500F for 100-3,500 hours.

During a LOCA, the temperature of conventional fuel cladding can reach

2200F. Cladding remaining at these temperatures for even short periods of time

experiences an energetic oxidation reaction with the steam which is present after

the liquid coolant is lost. Calculations shown in Chapter 5 indicate that liquid

bonded LWR fuel, due to the lower fuel operating temperatures, exhibits peak

cladding temperatures in the range of 1200-15000F for a LOCA. It was decided

to test the samples at this temperature range for times between 6-24 hours to

simulate the cladding response to a LOCA.

High temperature testing at short time intervals was also viewed as a way to

study accelerated liquid metal attack, since testing over a typical fuel lifetime

(35,000 hours) was impractical.

Metallooraphic Preparation of the Test Specimens

After the specimens had cooled to room temperature, they were removed from

the furnace and sectioned using a diamond cut-off saw to produce 0.25 in. long

cross-sections with flat surfaces. These sections were mounted in a 1 in. diameter

mold using a quick setting resin. The mounted samples were polished prior to


Measurement of the Loss of Tube Wall Thickness

Determination of the change in the tube wall dimensions as a result of

exposure of the cladding to the liquid metal was measured directly from a series

of photographs at a magnification of 100x. These photographs were measured

using a dial caliper to 0.001 inches. Measurements from the tested samples were

compared to as-manufactured standard tube wall thicknesses which were also

measured from photographs. From this comparison, an average percent loss of

wall thickness was determined as follows:

( Standard wall Tested wall / Standard wall ) X 100 (3-1)

These average loss values were plotted versus testing time in order to

generate plots that can be used to make predictions over the cladding lifetime.

Transition Layers at the Liquid Metal/Solid Interface

Early in the investigation, transition layers were found to form at the solid-liquid

metal interface. As was discussed, the presence of these layers may have

beneficial or harmful effects relative to the cladding performance. A technique that

was used to characterize the transition layers is described below.

Electron beam microprobe analysis was performed using the JEOL

SUPERPROBE 733 on the transition layers which formed at the liquid metal-solid

interface. The electron microprobe focuses an electron beam than impinges on

the polished surface of the specimen producing characteristic x-rays, whose

wavelengths report the quantitative chemical composition. The microprobe was

used in the line scan mode by setting two end points and allowing the beam to

move in a straight line in small periodic steps. The composition readings are

plotted versus distance traveled in orderto produce composition profiles necessary

to study transition regions.

Liquid Metal Attack

Optical microscopy was used to determine the nature of the liquid metal attack.

Polished specimens were anodized, then examined using a metallograph with a

polarizer and full wavelength interference plate to view the microstructure of the

cladding. Photomicrographs of the internal edge and the main tube wall of the

cladding were made using magnifications of 200x to 500x.

Optical microscopy was also employed on the transition layer formed at the

solid-liquid metal interface in an attempt to correlate thickness of the layer with

length of time exposed, and to examine the integrity of the layer.

Results of the Material Compatibility Experiments

A total of 170 specimens from 79 tests were used to evaluate the liquid metal

attack over a range of temperature-time histories. The specimens were tested


using both the lead-bismuth eutectic and the lead-bismuth-tin alloy. These

specimens were evaluated for the amount of tube wall loss, and the interaction at

the clad-liquid metal interface including the occurrence of liquid metal penetration

of the grain boundaries.

Tube Wall Loss

The average loss of tube wall thickness measurements were made for both the

SOP and LOCA specimens. Results for the LOCA samples are shown in Figures

3-2 to 3-7 for both candidate metals at three different temperatures. Results for

the SOP samples are shown in Figures 3-8 and 3-9.

Figures 3-2 to 3-4 show the results for the eutectic lead-bismuth LOCA

specimens. These specimens show a linear increase in tube wall loss with time

for a specified temperature. Average loss data for specimens tested at 12150F is

shown in Figure 3-2. The two curves represent a specimen filled with liquid metal,

and a specimen that contains alumina pellets and liquid metal. The specimen

containing the large volume of liquid metal experienced a 14% decrease in the

wall thickness after 24 hours. Nuclear Regulatory Commission (NRC) standards

permit a 17% loss of tube wall thickness in one hour (due to clad oxidation). The

second curve in Figure 3-2 representing the cladding tube containing pellets and

a reduced liquid metal inventory, which is more indicative of an actual fuel rod,

exhibits a 9.5% loss in 24 hours. These results demonstrate the surface area to

liquid metal volume effect.

2 2 2 ,o _J oo -J. T .
20- O With Aluimina Pellets
S18- %Loss=0.299*t+0.038
c 16-
- 12-
CL 4

0 5 10 15 20 25
Time (hours)

Figure 3-2: Loss of Wall Thickness, Lead-Bismuth Samples Tested at 12150F for 24 hours [7]

S22- '/oLOSS=U. tT+U.UbD

20- 0 With Aluimina Pellets
S18- %Loss=0.448*t+0.053
- 12-
C) 10-
a0 4-


S5 10 15 20 25
Time (hours)

Figure 3-3: Loss of Wall Thickness, Lead-Bismuth Samples Tested at 13820F for 24 hours [7]

U 22- %LOSS=1.U32+t+U.05~
.20 0 With Aluimina Pellets
S18 %Loss=0.604*t+0.038
e 16
g 14
- 12-
n 10-

a. 4

0 5 10 15 20 25
Time (hours)

Figure 3-4: Loss of Wall Thickness, Lead-Bismuth Samples Tested at 15170F for 24 hours [7]

24- O Liquid Metal Only
0o 22- %Loss=0.433*t+0.073
S20- With Aluimina Pellets

FE 18- %Loss=0.153*t+0.034
I- 12-
) 10-

0- 4

0 5 10 15 20 25
Time (hours)

Figure 3-5: Loss of Wall Thickness, Lead-Bismuth-Tin Samples Tested at 1215F for 24 hours [7]

" 22- /oLUSS=U.O3L L+U.UDO

S20- With Aluimina Pellets
- 18- %Loss=0.273*t+0.028
( 14-
- 12-
On 0 CD
8 6- C

S5 10 15 20 25
Time (hours)

Figure 3-6: Loss of Wall Thickness, Lead-Bismuth-Tin Samples Tested at 13820F for 24 hours [7]

C 22- -/oL.Ub U.U -- L-t-.U/ t
S20 0 With Aluimina Pellets
S18- %Loss=0.342*t+0.010
1 12-

0 0

C- 4-


0 5 10 15 20 25
Time (hours)

Figure 3-7: Loss of Wall Thickness, Lead-Bismuth-Tin Samples Tested at 15170F for 24 hours [7]

) y


SO Liquid Metal Only

0 .1
S0 With Aluimina Pellets
a- %Loss=0.080*tO.611

0 250 500 750 1000 1250
Time (hours)

Figure 3-8: Loss of Wall Thickness, Lead-Bismuth Samples Tested at 7500F for 1000 hours [7]


0 With Aluimina Pellets
S %Loss=0.050*tAO.197



0 1000 2000 3000 400
Time (hours)

Figure 3-9: Loss of Wall Thickness, Lead-Bismuth-Tin Samples Tested at 7500F for 3500 hours [7]


Figures 3-3 and 3-4 show curves of tube wall thickness loss at 1382'F and

15170F respectively. The average loss in tube wall thickness was found to be 19%

in 24 hours for the tubes containing pellets for 13820F, and 25% at 15170F.

Figures 3-5, 3-6 and 3-7 shows curves of tube wall thickness loss at 12150F,

13820F, and 15170F, respectively, for lead-bismuth-tin alloy. These specimens

show a marked reduction in tube wall thickness loss over the lead-bismuth

specimens, ranging from 4% for the 12150F test with simulated fuel pellets at 24

hours to 10-15% for the 15170F test. This may be explained by the transition layer

formed at the solid-liquid metal interface.

For the SOP tests, specimens were tested at 7500F for longer period of time

in order to simulate standard reactor operation. Figure 3-8 shows the results for

the specimens containing lead-bismuth, while Figure 3-9 shows the results for the

specimens containing lead-bismuth-tin. The lead-bismuth specimens were found

to lose 7.5% average tube wall thickness in 1000 hours of operation. The lead-

bismuth-tin specimens exhibit far lower wall thickness loss, with 0.2% measured

for the simulated fuel rod after 3500 hours.

It can be concluded that for fuel lifetimes of 30,000-40,000, a LBLWR fuel rod

using lead-bismuth-tin as the bonding liquid metal will exhibit favorable tube wall

thickness loss characteristics. This is thought to be due to the formation of a

zirconium-tin intermettalic reaction layer.

Evaluation of Reaction Layers

Photomicrographs of both the lead-bismuth and lead-bismuth-tin specimens

tested for 1000 hours at 7500F are shown in Figures 3-10 and 3-11, respectively.

Electron beam microprobe analysis of the reaction layer for both specimens is

shown in Figures 3-12 and 3-13.

The reaction layer for the lead-bismuth specimen shown in Figure 3.10 appears

black in color, and shows a lack of intimate contact with the cladding. The reaction

layer is 3-4 mils thick. Compositional analysis of this layer shown in Figure 3-12

indicates that the reaction layer has an approximate composition of 70 weight

percent bismuth and 30 weight percent zirconium, and formed a BiZr intermetallic


The reaction layer for the lead-bismuth-tin specimen shown in Figure 3.11

appears lighter in color, and remains in contact with the cladding. The reaction

layer is approximately 1 mil thick. Compositional analysis of this layer shown in

Figure 3-13 indicates that the reaction layer has an approximate composition of

72.8 weight percent tin and 27.2 weight percent zirconium, and formed a ZrSn2

intermetallic compound.

From the average loss of tube wall data, it has been shown that the eutectic

lead-bismuth alloy exhibits much poorer compatibility with the Zircaloy-4 cladding,

with losses 10-15 times that observed for the lead-bismuth-tin alloy. It is apparent

that the ZrSn2 intermetallic layer acts as a diffusion barrier, which, once formed,

effectively stops the attack of the liquid metal on the cladding wall.

Figure 3-10: Photomicrograph of Reaction Layer, Lead-Bismuth Sample, 750F for 1000 hours [7]

2,4 Pellet r

Reaction \

r. -'.- ,., Bismuth-Tin 4

Figure 3-11: Photomicrograph of Reaction Layer, Lead-Bismuth-Tin Sample, 750F for 3500 hours [7]

Figure 3-11: Photomicrograph of Reaction Layer, Lead-Bismuth-Tin Sample, 750F for 3500 hours [7]

2 60-





0 50 100 150 20C
Width of Analysis (microns)
-- Zr -- Bi

Figure 3-12: Electron Beam Microprobe Reaction Layer Analysis Lead-Bismuth Sample [7]







0 5 10 15 20
Width of Analysis (microns)
Zr -A- Sn

Figure 3-13: Electron Beam Microprobe Reaction Layer Analysis Lead-Bismuth-Tin Sample [7]

Liquid Metal Compatibility with UO,

As discussed previously, compatibility tests were performed on Zircaloy tubes

filled with liquid metal. Some of these tests included alumina pellets to simulate

the effects of liquid metal volume reduction due to fuel pellets. A second set of

tests were conducted to determine the effects of UO2 pellets on the compatibility

of liquid metal bonding material with other fuel materials. Two liquid metal alloys

were tested; lead-bismuth-tin, and bismuth-tin-gallium. Tests were run with

Zircaloy tubes containing depleted UO2 pellets and filled with the liquid metal

bonding alloy. The test specimens were tested at 7500F for 500 hours,

representing standard operation procedure (SOP) conditions, and 1500'F for 24

hours representing loss of coolant accident (LOCA) conditions.

Slight differences were observed for the lead-bismuth-tin samples compared

to the previous tests containing alumina pellets, namely, the formation of

discernable ZrSn2 crystals in the intermetallic layer as is shown in Figure 3-14.

Previous tests, without UO,, exhibited a uniform intermetallic layer with no

observed crystal formation. Also shown in Figure 3-14 is an electron microprobe

analysis of the intermetallic layer which shows a region of zirconium-uranium oxide

which shows up as black. This oxide contains a large percentage of zirconium,

and may be the precursor to the formation of the ZrSn2 crystals.

Width of Analysis (microns)
-*- Sn- U Zr

Figure 3-14: Optical Photomicrograph and Electron Microprobe results of
Lead-bismuth-tin Sample with UO, pellets at 1500F for 24 hours

A hardness test was performed to determine the hardness of the ZrSn2 layer

relative to the substrate Zircaloy. These tests indicate that the intermetallic ZrSn2

layer is significantly harder than Zircaloy (237 hV vs. 167 hV), and may add to the

mechanical stability of the fuel rod and protect against pellet-clad interaction.

These tests shown that the properties of the UO2 pellet are largely unaffected

by the liquid bond, as no penetration into the pellet was observed, and no uranium

was detected in the bulk liquid metal outside of the intermetallic layer.

Similar tests conducted using a bismuth-tin-gallium alloy indicate that the fuel

matrix is soluble to some degree in the liquid metal, making this alloy unacceptable

for long term compatibility with UO2.

In summary, the presence of UO2 pellets was found to have a definite effect

on the morphology and abundance of intermetallic compounds. The lead-bismuth-

tin alloy shows the formation of a zirconium-uranium oxide layer at the surface of

the pellet, and a thin intermetallic layer made up of small crystals containing ZrSn2.

The bismuth-tin-gallium alloy produced a larger amount of intermetallic

compounds, as well as dissolving some of the UO2. It is therefore deemed

unacceptable for use as a bonding agent.

The results of these tests show that the lead-bismuth-tin alloy in the presence

of UO2 pellets does not affect the performance of the LBLWR fuel under standard

operating and LOCA conditions.


Additional Experimental Studies

Additional experimental studies were conducted to determine the flow of fission

gas through a small liquid metal-filled gap, and liquid metal-coolant at elevated

temperatures. The results of these studies are summarized below.

Fission Gas Flow Through Liquid Metal

Experiments were conducted to determine the flow of fission gas through the

liquid metal. These tests showed that helium and nitrogen gas readily rose

through the liquid metal and did not blanket either the pellet or clad. These results

confirm the data reported for fission gas release in sodium bonded fuel [22].

Liquid Metal Coolant Interaction

Additional tests were run to confirm the non-reactive nature of the liquid metal

with coolant water in case of a rod defected by a fretting mechanism. The tests

with liquid lead-bismuth-tin at 6000F and water (<212F) showed no reaction.

Summary of Experimental Studies

The following conclusions can be drawn from these experiments:

1. The lead-bismuth-tin alloy demonstrates better compatibility with Zircaloy-4

than lead-bismuth eutectic.

2. Extrapolation of the average loss of tube wall thickness data predicts 0.3%

loss in 5,000 hours under standard reactor operating conditions using the

lead-bismuth-tin alloy. This corresponds to less than 2% loss over the fuel

lifetime assuming the correlation is valid over longer exposure times.

3. The lead-bismuth-tin alloy exhibits no significant reactions when exposed

to UO2 pellets at prototypic temperatures. Bismuth-tin-gallium, however,

reacts with both the cladding and fuel and is deemed unacceptable for use

in LBLWR fuel.

4. On the basis of the tests conducted to date, the lead-bismuth-tin alloy

meets all of the material compatibility requirements for a candidate liquid

metal to be used in a light water reactor fuel design.

5. Additional studies which examined gas bubble transport through small liquid

filled gaps, and liquid metal-coolant interaction failed to produce any "show-

stoppers" which would preclude the use of liquid metal in light water reactor


6. Experiments are currently underway to determine the compatibility of the

lead-bismuth-tin with fission products. SIMFUEL, a simulated spent fuel

obtained from Atomic Energy of Canada, Ltd. is being used in this study.


Experimental studies were conducted to determine the wetting characteristics

of the liquid metal bond material, especially in small gaps. Such concerns arise

due to the small diametral gaps, and eccentricities associated with the fuel

manufacturing process. These eccentricities are a problem for conventional fuel

rods as the unequal gas gap can result in local hot spots on the cladding.

Fabrication of LBLWR fuel rods involves the insertion of pellets into the

cladding tube, and the introduction of the liquid metal into the cladding so that it

fills the spaces between the pellets, and the annulus between the pellets and the

cladding. As was discussed in Chapter 2, there are several methods for filling the

tubes. One technique is to apply a vacuum to a loaded fuel rod at an elevated

temperature, and back filling with liquid metal. A second method is to load solid

metal slugs into cold tubes tube before loading pellets, using the spring to supply

compression. As the rod is heated, the metal melts and is forced into the gaps by

the force of the spring. In either case, the rods must be inspected to be sure that

the liquid metal fills the gaps.

An experimental study was performed to characterize the wetting behavior of

liquid metal in small gaps. The results of this experiment were used to determine

the effect of gas blanketing due to inhomogeneous distribution of the liquid metal

on the fuel rod temperature profile.

Experimental Studies

Experiments were carried out under the direction of Dr. Glen J. Schoessow of

the Nuclear Engineering Sciences Department of the University of Florida to

confirm the fabrication and wetting behavior of the liquid metal/ UO/Zr bond. For

these tests, UO2 or A1203 pellets were loaded into quartz tubes approximating the

cladding with solid lead-bismuth alloy on top of the pellets. The tubes were

evacuated, the rods were heated to 400oF, and the liquid alloy was allowed to

flow by gravity around the pellets. These tests showed that due to surface

tension, the lead-bismuth alloy will not wet dimensions of one mil or less. An

analytical study was performed to determine the effects of gas blanketing on the

radial temperature profile, and the fuel centerline temperature.

Other liquid metals were tested to determine wettability, including non-lead

alloys such as tin-bismuth-gallium. These studies showed similar results.

Analytical Predictions of Gas Blanketing due to Eccentricity

A two-dimensional (radial and circumferential) model of 1800 section of a

Westinghouse 15 x 15 fuel rod was modeled using the TRUMP generalized heat

transfer computer code [23]. The rod, cladding, and gap material are all modeled,


and the pellet and cladding are assumed to be misaligned with a .001 inch gap on

one side. One-tenth of the rod is assumed to be gas blanketed (i.e. the gap is

assumed to be .001 inch wide and filled with helium), and the rest of the gap

contains liquid metal bond. The rod is assumed to operate at an average power

of 6 kW/ft, and the peak axial location has been modeled with a local peaking

factor of 1.2. The fuel pellet and cladding dimension, as well as the reactor

coolant thermal/hydraulic conditions are assumed to be at hot, full power

conditions. The computer model used in this analysis is shown in Figure 4-1.

Results and Conclusions

Two separate runs were made

1. Symmetric gap is completely filled with liquid metal

2. Eccentric gap with gas blanketing the rod for < .001 inch gap

A comparison of the radial temperature profile for the symmetrical (completely

liquid bonded) case, and the minimum and maximum gap eccentricity for the

liquid/gas bonded case are shown in Figure 4-2. Also shown is the radial

temperature profile for a conventional gas bonded fuel rod.

As shown in Figure 4-2, there is a local "hot spot" associated with gas

blanketing in the minimum clearance area. Circumferential heat conduction

mitigates this effect, however, and the net result is a radial temperature profile

which is less than 100F higher than the maximum clearance (liquid bonded)




Figure 4-1: TRUMP Computer Model for Eccentric Rod Study

Effect of Gas Blanketing
due to Pellet Eccentricity

I 0.15
Radius (in)

-- Min clearance -+- Max clearance -X- Symmetric --- Gas Only

Figure 4-2: Eccentric Rod Study With and Without Gas Blanketing

G 1'



profile. In addition, the overall fuel temperatures are slightly higher (approximately

250F at the centerline) for the eccentric rod than for the symmetrical rod, owing to

the 10% reduction in high thermal conductance area.

Both cases exhibit far lower centerline temperatures and stored energy than

the conventional fuel rod which is, in a sense, completely gas blanketed.

The results of this study indicate that

1. Liquid metal incorporated into the LBLWR fuel rod will fill the gaps between

the pellets and cladding completely for any gap greater than approximately

0.001 inch.

2. In the event that small areas of the rod are gas blanketed due to rod

eccentricity or other causes, the resultant increase in peak fuel

temperatures is small due to circumferential heat conduction from regions

of poor conductance (gas blanketed), to regions of high conductance (liquid



The potential benefit of liquid bonded LWR fuel lies in the reduction of the fuel

centerline temperatures and the corresponding reduction in fuel stored energy. In

order to evaluate the performance of the liquid metal bonded LWR fuel, detailed

thermal analyses were performed using the TRUMP [23] generalized heat transfer

computer code. In addition, a fuel thermal/mechanical computer code, ESBOND,

was developed to assess the LBLWR fuel performance over a typical fuel cycle.

The results of these calculations are discussed in Chapter 6.

Steady-State Fuel Temperatures

The steady-state operating characteristics of the liquid bonded LWR fuel were

determined by constructing a simple one-dimensional radial heat conduction

model. This analysis considers a typical PWR fuel rod design, (17 x 17 PWR rod)

operating at average (6 kW/ft) and peak (13 kW/ft) linear power, with forced

convection flow along the outside of the cladding. The LWR fuel dimensions are

used with the gas gap replaced by liquid metal. Fuel rod dimensions are taken

from Table 2-3. The forced convective heat transfer coefficient between the


cladding and the light water coolant is assumed to be 4000 Btu/hr-ft2-oF, and

uniform heat generation is assumed to occur throughout the fuel volume.

The TRUMP generalized heat transfer computer program was used to

determine the steady-state radial temperature profiles for the liquid bonded fuel,

and for conventional LWR fuel.

Over the life of a conventional LWR fuel rod, the thermal conductance that

occurs in the gas gap varies from 500 to 3000 Btu/hr-ft2-'F [9]. Liquid metal

reactor fuel, on the other hand, which employs liquid metal bond material in the

fuel-cladding gap, exhibits virtually zero thermal resistance across the gap

throughout the fuel lifetime.

A radial heat conduction model consisting of 20 fuel nodes, and 6 cladding

nodes was constructed using TRUMP. The effect of varying the thermal resistance

between the fuel and cladding was studied by performing several steady-state

calculations. The effect of varying the gap conductance for a typical fuel rod

(Westinghouse 17x17), is shown in Figure 5-1 for a average linear power of 6

kW/ft. Figure 5-2 shows the effect of varying gap conductance for a peak linear

power of 13 kW/ft. These results show that the centerline fuel temperature for a

LBLWR rod can be reduced from 1500F-6500F for an average power rod, and

450F-16000F for a peak power rod by eliminating the gap resistance using the

liquid metal bond.





U.U2 U

.U4 0.06 0.08 0U. U.1Z U.14 U.b6 U.its U.z
Radius (in)
h=500 -- h=1000 h=1500
-8- h=2000 X- Liquid Btu/hr-ft2-F

Figure 5-1: Fuel Temperature Profile vs. Gap Conductance (6 kW/ft)

Westinghouse 17x17 Liquid Bonded Fuel
Effect of Gap Conductance (6 kW/ft)
I-------------- -------------4 ---- - ------j- - ---- -- --- - -
8 00 - 1

i ~ Cli i Clad

----- ----- -----

00- Pellet -- ---------- ----..- ----.... '--
600- ------- ---------------i

600 1------------i

Westinghouse 17x17 Liquid Bonded Fuel
Effect of Gap Conductance (13 kW/ft)

0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
Radius (in)
Sh=500 -+- h=1000 -K- h=1500
-- h=2000 -X- Liquid Btu/hr-ft2-F

Figure 5-2: Fuel Temperature Profile vs. Gap Conductance (13 kW/ft)


Thus, a reduction of the fuel centerline temperatures translates directly into

increased operating margins for LBLWR fuel, as compared to conventional LWR


Lower operating fuel temperatures also increase the fuel thermal conductivity,

as is shown in Figure 2-3. Higher fuel thermal conductivity decreases the radial

temperature gradient in the fuel pellet. The integrated effects of lower fuel

temperatures and radial temperature gradient on operating fuel characteristics such

as fission gas release, fuel cracking and swelling, fuel-cladding interaction, and

clad integrity will be assessed using the fuel lifetime calculation code that is

described in Chapter 6.

It can be concluded from this simple steady-state analysis that the liquid

bonded fuel operates at far lower temperatures than the conventional LWR fuel,

especially at the beginning of life when the maximum thermal resistance occurs

for conventional fuel. This is primarily due to the lack of thermal resistance across

the gas gap, and to a lesser extent, the improved thermal conductivity of UO, at

these lower temperatures. For the 6 kW/ft case, the contribution of the reduced

thermal impedance over the liquid metal gap to the reduction in centerline

temperature at beginning of life is 66%, compared to 34% from the increased

thermal conductivity.

Transient Performance

The TRUMP model used to determine the fuel steady-state operating

temperatures can also be used to evaluate the LBLWR fuel performance in the

event of a postulated accident. Two accident scenarios are examined:

1. Loss of coolant accident -- Instantaneous transition from forced convection

heat transfer at the start of the event to steam cooling, coupled with an

instantaneous reduction to zero power.

2. Transient overpower -- Step increase in fuel pin linear power resulting from

a local reactivity excursion.

Loss of Coolant Accident

In a large number of design basis accidents, heat transfer to the coolant is

sharply curtailed due either to loss of flow or loss of coolant. In these cases voids

appear in the core, shutting down the nuclear reaction, but causing the cladding

temperature to rise sharply due to the loss of heat transfer from the cladding

surface. This rapid increase in clad temperature is due to the stored energy in the

fuel which is given by

Q = pp v [ T(r) T,] dV

where p is the fuel density

c, is the fuel specific heat

V is the fuel volume

T(r) is the radial temperature distribution

and T, is the fuel rod outer surface temperature

For a cylinder with internal heat generation, the radial temperature profile is

given by

T(r) = T, + (To TJ ) [ 1 (r/R)2 ] (5-2)

where To is the fuel centerline temperature

and R is the outer radius of the fuel rod

Substituting the temperature profile into equation 5-1 and integrating over the

volume yields

Q/L = npc, ( T T ) R2/2 (5-3)

where Q/L is the stored energy per unit length of fuel rod

A transient heat conduction calculation was performed to determine the effect

of the stored energy on the cladding temperature after a loss of coolant event.

The steady-state temperature profiles for both liquid bonded and conventional fuel

types were used as initial conditions for the transient. To simulate the power

shutdown associated with the sudden loss of moderator, a step change in the fuel

volumetric heat generation rate from 13 kW/ft to decay heat levels is assumed at

the beginning of the transient. For simplicity, the decay heat is conservatively

assumed to remain at 6% of operating power throughout the transient. To

simulate the loss of coolant, a step change in the cladding surface heat transfer

coefficient from 4000 Btu/hr-ft2'-F (forced liquid convection) to 10 Btu/hr-ft2-F

(steam cooling) is assumed at the beginning of the transient.

The steady state temperature profile in the fuel pellet is highly peaked due to

the large heat generation rate, and the low thermal conductivity of the fuel. With

the drastic reduction in the cladding surface heat transfer coefficient, the

temperature profile is forced to assume a much flatter shape, which causes a large

increase in the cladding surface temperature. The fuel and cladding quickly reach

a quasi-equilibrium temperature which results in high cladding temperatures, as is

shown for both liquid bonded and conventional 17 x 17 fuel in Figure 5-3. These

temperatures are somewhat conservative due to the constant decay power level.

For a more realistic decay heat curve, the rate of temperature increase would be

continually less for both fuel types. The assumption of constant decay power is

valid for comparisons over the first 40 seconds as is shown in Figure 5-3. Due to

the lower average fuel temperature for the liquid bonded fuel, the maximum

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs