Content uploaded by Michael V. Glazoff
Author content
All content in this area was uploaded by Michael V. Glazoff on Sep 12, 2016
Content may be subject to copyright.
The INL is a U.S. Department of Energy National Laboratory
operated by Battelle Energy Alliance
INL/EXT-16-37972
Revision
0
Comparison of Nuclear
Fuels for TREAT:
UO2 versus U3O8
Michael V. Glazoff,
Isabella
J. van Rooyen,
Benjamin
D. Coryell, and
Clemente
J. Parga
May 2016
DISCLAIMER
This information was prepared as an account of work sponsored by an
agency of the U.S. Government. Neither the U.S. Government nor any
agency thereof, nor any of their employees, makes any warranty, expressed
or implied, or assumes any legal liability or responsibility for the accuracy,
completeness, or usefulness, of any information, apparatus, product, or
process disclosed, or represents that its use would not infringe privately
owned rights. References herein to any specific commercial product,
process, or service by trade name, trade mark, manufacturer, or otherwise,
does not necessarily constitute or imply its endorsement, recommendation,
or favoring by the U.S. Government or any agency thereof. The views and
opinions of authors expressed herein do not necessarily state or reflect
those of the U.S. Government or any agency thereof.
INL/EXT-16-37972
Revision
0
Comparison of Nuclear Fuels for TREAT:
UO2 versus U3O8
Michael V. Glazoff, Isabella J. van Rooyen, Benjamin D. Coryell, and
Clemente J. Parga
May 2016
Idaho National Laboratory
Idaho Falls, Idaho 83415
http://www.inl.gov
Prepared for the
U.S. Department of Energy
Office of Nuclear Energy
Under DOE Idaho Operations Office
Contract DE-AC07-05ID14517
INTENTIONALLY BLANK
iii
ABSTRACT
The Transient Reactor Test (TREAT) facility that resides in the Materials
and Fuels Complex at Idaho National Laboratory first achieved criticality in 1959
and successfully performed many transient tests on nuclear fuel until 1994, when
its operations were suspended.1 Resumption of operations at TREAT was
approved in February 2014 to meet the U.S. Department of Energy Office of
Nuclear Energy objectives in transient testing of nuclear fuels. The National
Nuclear Security Administration’s Office of Material Management and
Minimization is converting TREAT from its existing highly enriched uranium
core to a new core containing low-enriched uranium (i.e., U-235°less than 20%
by weight).1
The TREAT Conversion Project is currently progressing with conceptual
design phase activities. It is very important to make the right decision on what
type of nuclear fuel will be used at TREAT. In particular, one has to consider
different oxides of uranium and, most importantly, UO2 versus U3O8.1 The
objective of this report is to evaluate and compare the differences and similarities
between these two uranium oxides, including thermodynamic behavior at
atmospheric and subatmospheric pressure (i.e., crystal structure, phase stability,
and red-ox reactions), some thermo-physical properties (i.e., heat conductivity
and volumetric expansion and densification during fabrication), chemical
reactivity with fuel constituents during fabrication and in contact with assembly
materials (i.e., zirconium-based cladding) at operational environments
(i.e., steady-state and planned transients), and design basis accident scenario
(i.e., reactivity accident transient).
In this report, the results are documented pertaining to the choice mentioned
above (UO2 versus U3O8). The conclusion in favor of using UO2 was made based
on analysis of historical data, up-to-date literature, and self-consistent
calculations of phase equilibria and thermodynamic properties in the U-O and
U-O-C systems. The report is organized as follows: first, the criteria used to
make the choice were analyzed and, second, existing historical data and current
literature were reviewed. This analysis was supplemented by construction and
examination of the U-O and U-O-C phase diagrams in design operation
conditions. Finally, the conclusion in favor of the UO2 down selection is
summarized and explained.
iv
INTENTIONALLY BLANK
v
ACKNOWLEDGEMENTS
One of the authors (MVG) would like to express his sincere gratitude to
Professor Indrajit Charit (University of Idaho, Moscow, Idaho) for his valuable
thoughts and ideas shared during his presentation of the two classes “Nuclear
Materials Science” and “Interaction of Radiation with Substance” in 2011
through 2012.
vi
INTENTIONALLY BLANK
vii
CONTENTS
ABSTRACT ................................................................................................................................................. iii
ACKNOWLEDGEMENTS .......................................................................................................................... v
1. INTRODUCTION AND CRITERIA USED FOR UO2 VERSUS U3O8 SELECTION .................. 1
2. THERMODYNAMICS AND KINETICS OF URANIUM OXIDES ............................................... 5
3. POTENTIAL INTERACTIONS INSIDE NUCLEAR FUEL (UO2 AND U3O8
REACTING WITH GRAPHITE) .................................................................................................... 11
3.1 Uranium Dioxide UO2 versus U3O8 – General Considerations for Graphite
Reactions ................................................................................................................................ 11
4. AIR INGRESS AND ITS INFLUENCE ON UO2 FUEL STOICHIOMETRY AND
ZIRCONIUM-BASED CLADDING ............................................................................................... 23
5. VOLUMETRIC CHANGES DURING FUEL BLOCK MANUFACTURING (950°C)
AND DURING SUBSEQUENT OPERATION UP TO 820°C ....................................................... 24
6. OPTIMAL MANUFACTURING CONDITIONS ........................................................................... 26
7. THERMAL CONDUCTIVITY OF UO2 AND U3O8 ....................................................................... 27
8. INTERACTIONS OF NUCLEAR FUEL WITH CLADDING MATERIAL ................................. 28
9. CONCLUSIONS .............................................................................................................................. 31
10. REFERENCES ................................................................................................................................. 32
FIGURES
Figure 1. Crystalline lattices of UO2 (left) and U3O8 (right).4 ...................................................................... 3
Figure 2. Bulk fuel density versus oxygen content for uranium oxides sintered at 1450°C for
2 hours in argon.6,5 ....................................................................................................................... 3
Figure 3. Variation of total thermal conductivity as a function of temperature and linear thermal
expansion coefficient of UO2 with temperature.7 ........................................................................ 4
Figure 4. Variation of the UO2 creep rate with and without irradiation.6,2 ................................................... 5
Figure 5. The calculated oxygen/uranium phase diagram.13 ......................................................................... 8
Figure 6. The oxygen/uranium phase diagram according to Reference 14, calculated................................. 9
Figure 7. The U-O phase diagram constructed experimentally.15 ............................................................... 10
Figure 8. Isothermal cross section of the U-C-O phase diagram.16 ............................................................. 10
viii
Figure 9. Values of Gibbs free energy for urania-graphite system as a function of temperature and
pressure.18 .................................................................................................................................. 23
Figure 10. Variation of total thermal expansion and linear thermal expansion coefficient of UO2
with temperature.6,5 .................................................................................................................... 24
Figure 11 Thermal conductivity of unirradiated polycrystalline UO2.00.25 .................................................. 27
Figure 12. Thermal conductivity of U3O8 corrected to 100% theoretical density as a function of
temperature according to different data.27 ................................................................................. 27
Figure 13. Temperature dependence of the phase composition for the alloy Zy-2. .................................... 29
Figure 14. Temperature dependence of the phase composition for zircaloy-4. .......................................... 29
Figure 15. Temperature dependence of the phase composition for ZIRLOTM. ........................................... 29
Figure 16. Temperature dependence of the phase composition for the alloy M5TM. .................................. 30
Figure 17. Temperature dependence of the phase composition for the alloy Zircaloy-3 (our
calculations). .............................................................................................................................. 30
TABLES
Table 1. Some nuclear and physical properties of UO2 and U3O8. Macroscopic cross sections
(fission and absorption) are for neutron energy 0.025 eV.2,5 ....................................................... 2
Table 2. Gibbs free energy for the reaction 3UO2+O2 = U3O8 as a function of temperature at
pressure P=0.05 mbar, corresponding to the optimal conditions determined at ANL.9 .............. 6
Table 3. Values of Gibbs free energy for the reaction 3UO2+O2 = U3O8 as a function of temperature
at ambient pressure P = 1,013.25 mbar (or about 1 atm). This table also illustrates
polymorphic transformations that take place at 210°C for s1 s2; at 397°C for s2 s3;
and at 557°C for s3 s4 (our calculations). ................................................................................ 7
Table 4. Gibbs free energy as a function of temperature for the reaction UO2 + C = 2CO + U at
ambient pressure (101.325 kPa = 1,013.25 mbar ~ 1 MPa). ..................................................... 12
Table 5. Values of Gibbs free energy for the reaction UO2 + C = 2CO + U as a function of
temperature at an external pressure of 0.05 mbar. ..................................................................... 13
Table 6. Values of Gibbs free energy for the reaction UO2 + 4C = UC2 + 2CO as a function of
temperature at ambient pressure P = 1,013.25 mbar (or about 1 atm). ...................................... 14
Table 7. Values of Gibbs free energy for the reaction UO2 + 4C = UC2 + 2CO as a function of
temperature at ambient pressure P = 0.05 mbar. ....................................................................... 15
Table 8. Values of Gibbs free energy for the reaction UO2 + 4C = UC2 + 2CO as a function of
temperature at P = 0.01 mbar. ................................................................................................... 16
ix
Table 9. Values of Gibbs free energy for the reaction UO2 + 4C = UC2 + 2CO as a function of
temperature at P = 0. .................................................................................................................. 17
Table 10. Values of Gibbs free energy for the reaction U3O8 + 8C = 3U + 8CO as a function of
temperature at ambient pressure P = 1013.25 mbar (or about 1 atm). ....................................... 18
Table 11. Values of Gibbs free energy for the reaction U3O8 + 8C = 3U + 8CO at P = 0.05 mbar. ........... 19
Table 12. Values of Gibbs free energy for the reaction U3O8 + 8C = 3U + 8CO at P = 1 atm. .................. 20
Table 13. Values of Gibbs free energy for the reaction U3O8 + 8C = 3U + 8CO at P = 0.05 mbar. ........... 21
Table 14. Values of Gibbs free energy for the reaction U3O8 + 8C = 3U + 8CO at P = 0. ......................... 22
Table 15. Coefficients of thermal expansion of the MOX fuel and stoichiometric UO2.00.23 ..................... 25
x
INTENTIONALLY BLANK
1
Comparison of Nuclear Fuels for TREAT: UO2 versus
U3O8
1. INTRODUCTION AND CRITERIA USED FOR
UO2 VERSUS U3O8 SELECTION
There is a major effort underway for limiting or eliminating the international trafficking of highly
enriched uranium. For solving this problem, the Reduced Enrichment for Research and Test Reactors
Program was started in 1978 by the U.S. Department of Energy (DOE) to develop technology needed to
convert highly enriched uranium to low-enriched uranium (LEU) fuels. LEU is supposed to contain less
than 20% of the U-235 isotope and is much more resistant to proliferation.
Among other things, nuclear fuel for a research reactor must possess the following properties:
(1) high value of thermal flux and (2) the density of the LEU fuel must also be high.
This second goal typically can be achieved through one of two approaches: (1) increase the loading
of the fuel in the fuel element or (2) change the fuel composition to one that contains more fissile
uranium isotope(s). To convert all considered reactors to LEU fuels, a density of at least 8 g-U/cm3 will
be generally required.2 From this perspective, consider dispersion fuels and monolithic fuels. A typical
dispersion fuel consists of fuel in powder form that is dispersed, for example, in a graphite matrix that
is clad between cover plates (e.g., made of zircaloy-4 or M-5TM zirconium-based cladding alloys) that
are being considered for the Idaho National Laboratory (INL) Transient Reactor Test (TREAT)
facility.1 The highest possible uranium loading can be achieved for some U-Mo alloys: 17.0 g/cm3 with
U-10Mo 17.5 g/cm3 with U-7Mo.2 However, for many older experimental reactors, ceramic uranium-
based fuel remains the main choice because its properties have been studied extensively and it
possesses a number of well-known advantages over metallic fuels (i.e., higher melting point, chemical
inertness, property isotropy, and reduced probability of re-criticality during core meltdown).
A very detailed analysis of the different types of fuel was conducted by van Rooyen et al. in their
recent trade study.1 The choice was made in favor of the uranium-based (i.e., U-235) ceramic fuel
dispersed in the graphite matrix; therefore, it is necessary at this stage to determine which of the oxides
of uranium (i.e., UO2 versus U3O8) would perform better under the specific conditions of the TREAT
reactor. This determination will allow us to consider only the most suitable ceramic fuel enriched with
the U-235 isotope (rather than U-233), although there should not be any difference in the chemical
behavior of both isotopes and their respective oxides.
In Oak Ridge National Laboratory’s (ORNL’s) earlier work addressing the issue of safe storage of
nuclear fuel, the problem of thermal stabilization of the fuel was considered.3 The authors noted that
the temperature of the process was chosen mostly based on the basis of eliminating residual moisture
and volatile impurities. It was concluded that conversion to U3O8 was sufficient to accomplish all
desired goals. The preferred storage form is U3O8 because it is more stable than UO2 or UO3 in
oxidizing atmospheres. Heating in an oxidizing atmosphere at 750°C for at least 1 hour will achieve the
desired thermal stabilization.3
Also, the U3O8 oxide form had been utilized at TREAT earlier, before the restart efforts were
initiated.1 It is possible that the choice of U3O8 was made to ensure there is not any fuel oxidation (and,
therefore, dimensional instabilities) associated with operations of TREAT.
The prevalent species of uranium oxide are the chemical forms UO2, UO3, and U3O8; however, a
number of other phases such as UO; U4O9; U3O7; and a number of non-stoichiometric phases are also
described.
2
Literature states that UO is the lowest oxide (U+2 oxidation state) that can be identified only in thin
films (i.e., native oxide) when uranium is exposed briefly to open air.2 It is also well established that
uranium dioxide (i.e., UO2) can exist in a wide range of variable compounds depending on temperature,
environment, and the pressure. The U3O7 oxide is formed as a result of the phase transformation
reaction2:
U3O7 2UO2 + UO3 . (1)
It forms at a temperature of approximately 150°C and transforms to U3O8 at around 375°C. During its
turn, the U3O8 oxide is unstable above 450°C and converts back to UO2 at higher temperatures. This
sequence of phase reactions (taking place at different temperatures) can be illustrated as follows:2
UO2 + 2UO3 U3O8 3UO2 + O2 . (2)
A major issue about the UO2 dioxide is that, depending upon temperature and oxygen pressure in the
system, it can deviate from its stoichiometric composition in both directions (i.e., toward deficiency and
excess of oxygen in the crystalline lattice). If the oxygen to uranium atom ratio is 2.0, the UO2 is
stoichiometric. If an oxygen-deficient or excessive uranium exists (i.e., O/U less than 2.0), the fuel is
called hyper-stoichiometric fuel (UO2-x). If an O/U greater than 2.0, UO(2+x) is called hypo-stoichiometric
fuel (x is a small fraction).4
The departures from stoichiometry (i.e., self-diffusion in fuel itself and inter-diffusion between fuel
and cladding materials to form hyper-stoichiometric or hypo-stoichiometric fuel during the reactor
operation) may result in deterioration of the UO2 creep properties. However, for a dispersion-type fuel,
this issue seems to be of very minor importance. Some nuclear and physical properties of UO2 and U3O8
(i.e., the only two uranium oxides that are used for fuel development and/or safe storage) are presented in
Table 1.
Table 1. Some nuclear and physical properties of UO2 and U3O8. Macroscopic cross sections (fission and
absorption) are for neutron energy 0.025 eV.2,5
Nuclear Properties (Natural Uranium)
Crystal Structure and Physical Properties
Compound
Fission
Σf, in
cm
2
/cm
3
Absorption
Σa, in
cm
2
/cm
3
n
Unit
Cell
Type
Lattice
Paramete
r
Atoms
per
Cell
Melting
Point,
°C
X-ray
Densit
y
U
Content,
a/o
U
Content,
w/o
UO2 0.102 0.185 1.34 FCC
(CaF2
type)
a=5.468 4 2,780 10.96 33.3 88.15
U3O8 0.065 0.120 1.34 Ortho-
rhombic
a=6.70
b=11.94
c=4.14
2 2,500 8.39 27.2 73.61
From Table 1, it is clear that the nuclear properties of both oxides are quite similar (fission and
neutron absorption macroscopic cross sections). However, the melting point of UO2 is somewhat higher
(2780 versus 2500°C). Also, the UO2 has a crystalline structure of the CaF2 type (cF12; FCC –face
centered cubic), while U3O8 is orthorhombic with different values for all three lattice parameters a, b, and
c. This causes an undesirable anisotropy of properties for U3O8. It might not represent a significant
impediment for the safe storage of nuclear fuel, but for its use inside a nuclear reactor as a fuel it might be
problematic because of the dimensional instability issues. The crystalline lattices of both oxides are
presented in Figure 1.4
3
Figure 1. Crystalline lattices of UO2 (left) and U3O8 (right).4
Continuing the analysis of Table 1, UO2 has both higher atomic density and actual density compared
to U3O8 (10.96 versus 8.39g/cm3). The nuclear fuel fabricated with UO2 provides a number of advantages
for UO2 compared to U3O8, including (a) higher uranium density, (b) high value of thermal conductivity,
(c) high capability to contain and retain fission product gases in the fuel and (d) high value of linear
power rating (q) of the fuel element:6,2
. (3)
In terms of bulk nuclear fuel density and its dependence on UO2 stoichiometry, the results shown in
Figure 2 were reported.6,1
Figure 2. Bulk fuel density versus oxygen content for uranium oxides sintered at 1450°C for 2 hours in
argon.4,3
∫
=dTTkq ),(
ρ
4
From Figure 2, it follows that the acceptable values of the sintered bulk UO2 fuel density (around
9 g/cm3) can be achieved with very insignificant deviations from the stoichiometric composition
(i.e., around UO2.01). This does not affect its melting temperature and thermo-mechanical properties in a
detrimental way. Consequently, the level of creep rate will remain almost the same as for the
stoichiometric compound UO2 (see Figure 4), while thermal conductivity (i.e., a property that is very
sensitive to deviations from stoichiometry) will also be sufficiently close to that of the UO2 (see Figure 3).
Figure 3. Variation of total thermal conductivity as a function of temperature and linear thermal
expansion coefficient of UO2 with temperature.7
Generally, thermal conductivity varies quite substantially with deviations from UO2 stoichiometry
(i.e., lower at elevated temperatures [above 600°C] for hyper-stoichiometric materials [e.g., UO2.01 ] and
higher for hypo-stoichiometric materials [e.g., UO1.97]).1 Comparison of thermal conductivities of UO2
and U3O8 can be found in Figure 3 in much greater detail.
Irradiation swelling in UO2 will obviously depend on the fuel burn-up and temperature of the process.
Finally, creep rate will be higher in the case of higher irradiation than for normal conditions (see
Figure 4).
5
Figure 4. Variation of the UO2 creep rate with and without irradiation.4,1
2. THERMODYNAMICS AND KINETICS OF URANIUM OXIDES
Susceptibility of UO2 to oxidation depends on the average particle size.8 In very fine particles, UO2
becomes easily oxidized to U3O8 in the presence of oxygen gas. However, when particles of UO2 are at
least 0.3-µm large, UO2 is fairly resistant to oxidation.8 Very dense, sintered UO2 pellets will not get
oxidized for many years because they are protected by the slightly oxidized thin surface film that is
formed on large UO2 grains.8 In any case, the UO2-based fuel must operate at as low of an external
pressure as possible.a The influence of external pressure on the value of the Gibbs free energy changes for
the 3UO2+O2 = U3O8 (see Table 2).
It is clear that this reaction becomes thermodynamically prohibited at temperatures slightly higher
than 1000°C (∆G > 0).
On the other hand, at ambient pressure, this phase reaction is allowed thermodynamically in the entire
range of temperatures considered in Table 3.
This requirement is particularly important for the TREAT reactor because the quality of vacuum inside
the assembly affects heat transfer through the fuel-to-cladding gap in a profound way (see Reference 5).
Consequently, a conclusion was made in the Argonne National Laboratory (ANL) study9 that pressure in the
gap should be no higher than 0.05 mBar, which can be considered a perfect vacuum for thermal
calculations. All calculations were made using the TAB module of ThermoCalc v.2015a and the SSUB4
database for thermodynamic properties of individual substances.10
a See discussion below about pressure-dependent chemical interactions of UO2 and the graphite matrix.
7
Table 3. Values of Gibbs free energy for the reaction 3UO2+O2 = U3O8 as a function of temperature at
ambient pressure P = 1,013.25 mbar (or about 1 atm). This table also illustrates polymorphic
transformations that take place at 210°C for s1 s2; at 397°C for s2 s3; and at 557°C for s3 s4 (our
calculations).
UO2 does not have any polymorphs (i.e., crystalline modifications), while U3O8, according to the
literature data,10,11 has four polymorphs (sometimes denoted as s1, s2, s3, and s4; see Table 2 and Table 38),
of which only two are stable in normal conditions.11 These polymorphic modifications of U3O8 are called
“α” and “β.”12 The most stable α−modification possesses orthorhombic crystalline structure, while the
β−modification is hexagonal.10 They are related but not identical in terms of their respective crystalline
structures. Indeed, in the β−U3O8 structure, the uranium atoms are located in a single three-fold position,
while in the α−U3O8, they occupy in two types of positions: two-fold and four-fold.
The uranium-oxygen phase diagrams described in the literature do not yield a coherent picture.
Indeed, depending on the type of report (i.e., experimental versus computational) and the date of its
publication, one can state that there is still no consensus about the number and stoichiometry of
intermediate phases present in the system.13,14,15,16,17 These phases as UO (in thin films1) (i.e., UO2, U3O8,
8
UO3, U4O9, U3O7, and different non-stoichiometric compositions of the UO2±δ) have been reported. U2O5,
U13O34, and U5O12 were noted in one reference (phase diagram),4 as well as U8O21±x.16 In the latter case, it
was pointed out that the U8O21±x oxide produces metastable states that depend on sample morphology.16
Because these phases (i.e., U2O5; U13O34; U5O12; and U8O21±x) were reported to be stable only at relatively
low temperatures or in certain specific morphologies, we do not consider them here.
Of particular interest is the fact that the melting temperature of stoichiometric UO2 (around 2865°C)
decreases quite substantially when the oxygen/uranium ratio is 1.68 (2425°C), and also when the
oxygen/uranium ratio is 2.25 (2500°C) (i.e., for both hypo and hyper-stoichiometric compositions).12 In
turn, as pointed out, both types of deviations from stoichiometry, quite unexpectedly result in the
undesired increase in diffusion rate and other properties dependent on interatomic mobility.3
The two sources13,14 differ quite substantially in the phase compositions at low temperatures. This is
not surprising given that it is difficult to achieve equilibrium at low temperatures of 200 to 600°C in
experimental studies. However, they are consistent in providing the melting temperature of UO2 at 2850
to 2852°C. In both diagrams, UO2 melts congruently. The existence of several polytypes of U3O8 is also
an established fact, as well as that of UO3 and U4O9 (denoted sometimes as UO2.25). The temperature of
non-variant phase reaction U3O8 UO2 + GAS is also established reliably at about 1875°C (ambient
pressure) (see Figure 5 and Figure 6).
Figure 5. The calculated oxygen/uranium phase diagram.13
9
Figure 6. The oxygen/uranium phase diagram according to Reference 14, calculated.
The terminal points of the maximum and minimum solubility of oxygen in UO2 are unambiguously
determined as about 31 at.% U and about 38 at.% U, respectively, in both studies. The corresponding
temperatures are defined by the non-variant phase transformation at 2425°C (maximum U solubility) and
1875°C (decomposition of U3O8, minimum U solubility).
Therefore, one can say with confidence that all phase transformations related to production and
functioning of ceramic UO2-based nuclear fuel are very well established. Discrepancies are related only to
the relatively low-temperature regions of this phase diagram, which are considerably less important for
manufacturing and operation of nuclear fuel (see phase diagram15 in Figure 7).
A potential chemical interaction between nuclear fuel and the carbon (i.e., graphite) dispersion matrix
is an important issue and is discussed later. Here, only the analysis of the isothermal cross section of the
U-O-C phase diagram was conducted (see Figure 8).16
It is clear from this diagram that there is not any solubility of graphite in UO2; therefore, a chemical
reaction of UO2 reduction with C is hardly feasible in the present conditions. There is some solubility of
UO2 in UC, but this effect is not relevant for the TREAT developed fuel.
11
3. POTENTIAL INTERACTIONS INSIDE NUCLEAR FUEL (UO2 AND
U3O8 REACTING WITH GRAPHITE)
It is known that under certain conditions UO2 or U3O8 could chemically interact with the graphite
matrix (i.e., “carbothermic reduction”).We address this issue in detail in the following subsections and
consider a number of potential interactions and their feasibility, both at normal working conditions and at
air ingress conditions.
3.1 Uranium Dioxide UO2 versus U3O8 – General
Considerations for Graphite Reactions
In UO2, uranium is tetravalent, while in U3O8 its degree of oxidation, at least formally, is 5.33.
Therefore, U3O8 uranium is more oxidized than in UO2. Consequently, one could expect that if any
chemical interactions among uranium oxides and graphite are possible, U3O8 will get reduced by carbon
in somewhat milder conditions that UO2. However, one has to consider the fact that UO2 can easily form
non-stoichiometric compounds of UO2±δ, while ThO2 and PuO2 do not display such deviations from
stoichiometry. These lattice instabilities invite additional scrutiny in terms of potential interactions with
graphite. There is a clear need for calculating or revisiting the electronic structure of both oxides in order
to understand the peculiarities of their oxidation behavior. For now, a number of potential reactions are
considered from a thermodynamic standpoint.
1. UO2 + C = 2CO + U
For this reaction at ambient pressure, the temperature dependence of the Gibbs free energy of reaction
on temperature was considered (TAB module, Thermo-Calc 2015a; see Table 4).10
We see that for this hypothetical reaction to become allowed thermodynamically, the temperature in
the system must be above 2400 K (approximately 2123°C). However, metallic uranium will melt at
1408 K (about 1135°C).
When external pressure is equal to 0.05 mbar (determined as the maximum allowed or fuel elements
at ANL), we get the results shown in Table 5.
At a low pressure of 0.05 mbar, this reaction becomes thermodynamically possible at a temperature
somewhat above the metallic uranium melting point (Tm = 1135°C) (i.e., around about 1450°C). In
other words, this reaction is pressure sensitive but requires very deep reduction of uranium in the UO2
to metallic uranium. From this perspective and considering our general thoughts above, the next
reaction considered is much more energetically favorable as pressure approaches zero and requires
much more attention.
2. UO2 + 4C = UC2 + 2CO
For this reaction we get the following results at ambient pressure (see Table 6).
Consequently, this reaction is thermodynamically prohibited at any temperature of interest at ambient
pressure.
Therefore, when the pressure inside a fuel element becomes 0.05 mbar, this reaction becomes
thermodynamically permissible around about 2000°C (see Table 7). Exploring a hypothetical case
when P = 0.01mbar, we get different results (Table 8).
In this case, the studied reaction already becomes thermodynamically permissible at about 1800°C.
Finally, considering a hypothetical case when P=0, we get the results shown in Table 9.
12
Table 4. Gibbs free energy as a function of temperature for the reaction UO2 + C = 2CO + U at ambient
pressure (101.325 kPa = 1,013.25 mbar ~ 1 MPa).
13
Table 5. Values of Gibbs free energy for the reaction UO2 + C = 2CO + U as a function of temperature at
an external pressure of 0.05 mbar.
14
Table 6. Values of Gibbs free energy for the reaction UO2 + 4C = UC2 + 2CO as a function of
temperature at ambient pressure P = 1,013.25 mbar (or about 1 atm).
15
Table 7. Values of Gibbs free energy for the reaction UO2 + 4C = UC2 + 2CO as a function of
temperature at ambient pressure P = 0.05 mbar.
16
Table 8. Values of Gibbs free energy for the reaction UO2 + 4C = UC2 + 2CO as a function of
temperature at P = 0.01 mbar.
17
Table 9. Values of Gibbs free energy for the reaction UO2 + 4C = UC2 + 2CO as a function of
temperature at P = 0.
Therefore, the studied reaction becomes thermodynamically permissible at the relatively low
temperature of about 300°C.
This implies the need to select the maximum and minimum pressure and gas or vacuum pressure
inside the fuel/cladding gap very carefully and not only considerations of heat transfer, but also taking
into account the facts about chemical interaction as stated above.17
3. U3O8 + 8C = 3U + 8CO
For this reaction, at ambient pressure, we get results shown in Table 10.
At pressure equal to P = 0.05 mbar, results were obtained as shown in Table 11.
18
Table 10. Values of Gibbs free energy for the reaction U3O8 + 8C = 3U + 8CO as a function of
temperature at ambient pressure P = 1013.25 mbar (or about 1 atm).
19
Table 11. Values of Gibbs free energy for the reaction U3O8 + 8C = 3U + 8CO at P = 0.05 mbar.
This reaction becomes possible at temperatures around 1100°C.
4. U3O8 + 14C = 3UC2 + 8CO
For this reaction at ambient pressure, we get the results shown in Table 12.
At P = 1 atm this reaction is impossible in the whole range of studied temperatures.
When P = 0.05 mbar, we get the results in Table 13.
Therefore, at P = 0.05 mbar, U3O8 can be reduced to UC2 beginning at about 1300°C. Finally, for a
hypothetical case when P = 0, we get the results seen in Table 14.
Consequently, U3O8 becomes unstable with respect to reduction to UC2 at a very low temperature of
about 200°C.
These results are conveniently summarized in Figure 9.
20
Table 12. Values of Gibbs free energy for the reaction U3O8 + 8C = 3U + 8CO at P = 1 atm.
21
Table 13. Values of Gibbs free energy for the reaction U3O8 + 8C = 3U + 8CO at P = 0.05 mbar.
22
Table 14. Values of Gibbs free energy for the reaction U3O8 + 8C = 3U + 8CO at P = 0.
23
Figure 9. Values of Gibbs free energy for urania-graphite system as a function of temperature and
pressure.18
When summing up this section of the report in terms of chemical stability with respect to the process
of reduction with carbon, both oxides exhibit unstable behavior at very low pressure. However, as
expected, U3O8 is reduced more easily in the carbon matrix (i.e., at lower temperatures with all other
conditions being equal) than UO2. Consequently, no advantages could be gained by utilizing the U3O8–
based fuel.
4. AIR INGRESS AND ITS INFLUENCE ON UO2 FUEL
STOICHIOMETRY AND ZIRCONIUM-BASED CLADDING
In the unlikely event of air ingress, the fuel rods will be exposed to air-containing atmospheres at high
temperatures. In comparison with steam, the presence of air is expected to result in a more rapid
escalation of the accident. In particular, the presence of air can lead to accelerated oxidation of the
zircaloy cladding compared to that in steam, because of the faster kinetics, while the 85% higher heat of
reaction will introduce the positive feedback loop and drive the process even further.19
Previously, air ingress has been shown to cause poor heat transfer. In technical literature,20 it is noted
that the combined effect of these factors can give rise to an increased rate of fuel assembly degradation. In
oxygen-starved conditions, nitriding of the metal can occur (this requires thermodynamic verification);
the resulting zirconium nitride is highly inflammable.5 It can detonate on re-introduction of oxygen or
steam (e.g., during core re-flood). Furthermore, the exposure of UO2 to air at elevated temperatures can
lead to an increased release of some fission products, notably the highly radiotoxic ruthenium;21,22 air also
is likely to further weaken the damaged cladding as a barrier against fission product release.
24
The mechanisms of oxygen stoichiometry variation in UO2 at different temperature and oxygen
partial pressure were studied using DFT, or Density Functional Theory formalism.23 The authors
emphasized that very limited experimental studies are available to understand the atomic structure of UO2
near the surface and the defect effects of near surface on stoichiometry. By using their computational
approach, the authors concluded that, under the poor oxygen conditions, the stoichiometry is switched
from hyper-stoichiometric at 300 K, with a depth around 3 nm to near-stoichiometric at 1000 K and hypo-
stoichiometric structure at 2000 K. Furthermore, at very poor oxygen concentrations and high
temperatures, the results obtained by the authors also suggest that the bulk of the UO2 is energetically
more favorable to be hypo-stoichiometric, although the surface of OU2 remains near-stoichiometric.23
5. VOLUMETRIC CHANGES DURING FUEL BLOCK
MANUFACTURING (950°C) AND DURING SUBSEQUENT OPERATION
UP TO 820°C
Data on the thermal expansion characteristics of UO2 is presented in Figure 10.24,5
Figure 10. Variation of total thermal expansion and linear thermal expansion coefficient of UO2 with
temperature.6,5
The coefficients of thermal expansion of the stoichiometric UO2.00 are presented in Table 15.
25
Table 15. Coefficients of thermal expansion of the MOX fuel and stoichiometric UO2.00.23
Temperature
(K)
Relative Thermal
Expansion
∆L(Τ)/L(273)
Average Linear
Thermal Expansion
Coefficient
(1/K)
True Linear
Expansion
Coefficient
(1/K)
Density of UO2.00’
kg/m
3
3.0000E+02
2.6811E-04
9.9300E-06
9.7555E-06
1.0961E+04
4.0000E+02
1.2456E-03
9.8081E-06
9.7841E-06
1.0929E+04
5.0000E+02
2.2283E-03
9.8161E-06
9.8388E-06
1.0897E+04
6.0000E+02
3.2187E-03
9.8430E-06
9.9196E-06
1.0865E+04
7.0000E+02
4.2195E-03
9.8817E-06
1.0026E-05
1.0832E+04
8.0000E+02
5.2333E-03
9.9304E-06
1.0159E-05
1.0800E+04
9.0000E+02
6.2628E-03
9.9885E-06
1.0317E-05
1.0766E+04
1.0000E+03
7.3001E-03
1.0041E-05
1.0515E-05
1.0733E+04
1.1000E+03
8.3725E-03
1.0124E-05
1.0782E-05
1.0699E+04
1.2000E+03
9.4768E-03
1.0223E-05
1.1120E-05
1.0664E+04
1.3000E+03
1.0620E-02
1.0341E-05
1.1529E-05
1.0628E+04
1.4000E+03
1.1810E-02
1.0479E-05
1.2008E-05
1.0590E+04
1.5000E+03
1.3054E-02
1.0639E-05
1.2558E-05
1.0551E+04
1.6000E+03
1.4359E-02
1.0821E-05
1.3177E-05
1.0511E+04
1.7000E+03
1.5733E-02
1.1025E-05
1.3865E-05
1.0468E+04
1.8000E-03
1.7182E-02
1.1252E-05
1.4622E-05
1.0423E+04
1.9000E+03
1.8714E-02
1.1502E-05
1.5447E-05
1.0376E+04
2.0000E+03
2.0337E-02
1.1776E-05
1.6341E-05
1.0327E+04
2.1000E+03
2.2057E-02
1.2073E-05
1.7302E-05
1.0275E+04
2.2000E+03
2.3883E-02
1.2394E-05
1.8331E-05
1.0220E+04
2.3000E+03
2.5820E-02
1.2738E-05
1.9427E-05
1.0162E+04
In the range from 0 to 650°C, the expressions recommended by Martin25 should be used for relative
thermal expansion, the true linear expansion coefficient, and density of solid stoichiometric UO2.00 or very
similar MOX fuel26:
In the temperature range from 923K up to the melting point of UO2, the regression equations are
given as follows:7
26
In the case of U3O8, the situation with thermal expansion is somewhat more complex.25 Detailed
thermal expansion data for U3O8 have been obtained by high-temperature x-ray diffractometric data
between 22 and 1100°C.26 It was established that stoichiometric U3O8 changes continuously, reversibly
and anisotropically above room temperature with expansion along the a-axis and contraction along the
b-axis from orthorhombic to hexagonal symmetry at 350°C ± 10°C.26 A small but continuous contraction
along the c-axis occurs up to ll00°C. The loss of oxygen begins around 600°C, but this does not result in
any discontinuity in both parameters of the hexagonal phase (i.e., “a” and “c”) up to 875°C.
However, in the temperature range from 875 to 925°C, the structure undergoes a change to lower
symmetry as a result of contraction along the a-axis and expansion along the b-axis
(i.e., expansion-contraction anomaly). This structural change is accompanied by a more extensive loss of
oxygen and is usually irreversible unless the crystallite size is sufficiently small (i.e., about 0.05 µm). On
the other hand, during fuel block manufacturing, the temperature can reach as high as 950°C, while at
normal fuel operation conditions, it is maintained at about 820°C.
Therefore, because of anisotropic and some irreversible changes in the U3O8 crystalline structure and
tendency to lose oxygen at temperature above 600°C, UO2 shows significantly more thermal stability over
the thermal stability of U3O8. In addition, the thermo-physical properties of UO2 are exceptionally well
understood and display highly predictable behavior without any discontinuities and allotropic
transformations. This should serve as yet another argument in favor of UO2 rather than U3O8, that can
undergo polymorphic transformations.
6. OPTIMAL MANUFACTURING CONDITIONS
In this report, a number of factors were considered that might affect the chemical and dimensional
stability of the TREAT nuclear fuel. Assuming that the UO2–carbon matrix fuel will be selected for the
TREAT restart, one has to consider the optimal gas filling of the gap between the fuel and the Zry-4
cladding. Special attention was paid to the two types of conditions existing at the same time in the fuel.
On the one hand, the presence of even minor amounts of oxygen in the gap is highly undesirable because
of the potential oxidation of carbon to CO or CO2 at 820°C. Also, a possibility of the oxidative process
resulting in the chemical reaction UO2 U3O8 should be excluded. Consequently, the atmosphere in the
gap should be that of an inert gas and the pressure, as recommended in the ANL study, should be
0.05 mbar or lower.
However, this second process is always a distinct possibility if the temperature in the system is
sufficiently high and, at least in parts of the composite fuel, pressure upon the UO2 particles could be
negligible. This second reaction should be explored further, including the first-principles atomistic
calculations under different stress field distribution in the composite.
27
7. THERMAL CONDUCTIVITY OF UO2 AND U3O8
Thermo-physical properties of both oxides were studied quite extensively, especially for UO2.25,27,28
The summary thermal conductivity curves as functions of temperature for UO2 are presented in Figure 11
(corrected for 100% density).22
Figure 11 Thermal conductivity of unirradiated polycrystalline UO2.00.25
The thermal conductivity of orthorhombic α-U3O8 has been measured in air from 300 to 1100K using
an axial heat flow comparative set-up23 (see Figure 12). The results show that the conductivity decreases
monotonically with increasing temperature. The observed conductivity can be explained in terms of the
phonon-defects and phonon–phonon interaction processes.
Figure 12. Thermal conductivity of U3O8 corrected to 100% theoretical density as a function of
temperature according to different data.27
Comparing the data for UO2 and U3O8, the thermal conductivity of UO2 is somewhat higher than that
of U3O8 at all temperatures, although the exact ratio will be temperature-dependent. For example, at
900°C it is ~ 2 times higher for UO2.b For both ceramic materials, as expected, the thermal conductivity
b The conversion formula is: 1 W/(cm*C) = 100 W/(m*K)
28
values were low. Indeed, for comparison, the thermal conductivity of pure aluminum is 240 W (m-1 C-1)
at 100ºC.3 As the neutron fluence (to which both materials were exposed) grows, thermal conductivity
decreases for both oxides.29
In the recent experimental work29 it was established that for UO2, the thermal conductivity was
10.2 W/m K at 323K, and with increasing temperatures its behavior, it was inversely proportional to the
temperature. At 648K the thermal conductivity is 4.9 W/m K. The thermal conductivity peaks between
200 and 250K and decreases for decreasing temperatures.29
U3O8 was studied with different irradiation doses (from 0 Ar+/cm2 to 2 × 1016 Ar+/cm2). Its thermal
conductivity was 1.67 W/m K at 333K and 1.97 W/m K at 648K. The self-annealing effect was found
to be stronger than in UO2; therefore, thermal conductivities at 648K were from 1.3 up to 1.86 W/m K
for the above-mentioned doses. For lower doses, the thermal conductivity decreases with increasing
dose, but then starts increasing again for higher doses. The possible cause of this effect might be related
to re-crystallization and/or formation of a second phase (e.g., UO2+x in U3O8). In general, it was found
that oxidation of UO2 has a stronger influence on the thermal conductivity than irradiation with argon
ions.27 However, one must bear in mind that in the conditions of the TREAT reactor, there is very little
neutron damage to materials.
8. INTERACTIONS OF NUCLEAR FUEL WITH CLADDING MATERIAL
This is a complex topic that requires an in-depth study. Potential chemical interactions include
hydrogen uptake and oxidation (from outside the Zr-based cladding); irradiation-assisted corrosion
phenomena; swelling of fuel with extended burn-up; and many other effects. In any case, it will be
necessary to reliably establish the temperature distribution, neutron flux, fluence, chemical and phase
composition of the selected zirconium-based nuclear cladding alloy, and the external pressure inside and
outside a given fuel element. For all of these parameters, it is highly desirable to have not just
averaged-out values, but the parameter field resolved in time and in space (for a given geometry of the
fuel element).
As one of the preliminary steps in this work, we have calculated the phase composition as a function
of temperature for all zirconium-based alloys considered for the TREAT fuel element cladding:
zircaloy-2; zircaloy-4; ZIRLOTM; and M5TM. These results are presented in Figure 13 through
Figure 16.28,29,30
⋅
⋅
⋅
⋅
⋅
29
Figure 13. Temperature dependence of the phase composition for the alloy Zy-2.
Figure 14. Temperature dependence of the phase composition for zircaloy-4.
Figure 15. Temperature dependence of the phase composition for ZIRLOTM.
30
Figure 16. Temperature dependence of the phase composition for the alloy M5TM.
Zircaloy-3 is also an interesting material.19 It contains about the same amount of iron as Zircaloy-4
(0.25%Fe), but no nickel, no chromium, and only a small amount of tin compared to all other zircaloys
(0.25%). Its temperature dependence of the phase composition is provided in Figure 17.
Figure 17. Temperature dependence of the phase composition for the alloy Zircaloy-3 (our calculations).
It is known from the literature23,25,31,32 that corrosion, mechanical, and thermo-physical properties of
these materials can be changed in a favorable way by optimizing the so-called “temper” of these heat-
treatable materials (via the construction of TTT and CCT diagrams: “Time-Temperature-Transformation”
and Continuous Cooling Transformation”). This work is planned for the future.
31
9. CONCLUSIONS
In this brief report, a preliminary conclusion (based on both the computational thermodynamics
results of the authors and the literature analysis) was made for selecting UO2 versus U3O8 as the nuclear
fuel (to be dispersed in graphite) for the TREAT LEU fuel fabrication and reactor operational and
accident conditions. While the choice was made in favor of UO2 it is important to keep in mind that both
oxides could be used as fuel for TREAT if necessary. Some arguments in favor of the UO2 are given
below:
1. High symmetry of the UO2 crystalline lattice and isotropy of properties in a broad temperature
range ensure dimensional stability of the fuel assemblies.
2. While UO2 may react with the graphite matrix forming UC2, for the U3O8 oxide, the onset of such
chemical interactions takes place at lower temperatures (with all other factors being the same).
3. The issue of air ingress is well understood for UO2 monolithic fuel, but not for the UO2 fuel
dispersed in the graphite matrix.
4. It is important to understand the chemical reactivity mechanisms of UO2 at elevated temperatures.
This issue requires further study. However, using the thermodynamic calculations in this report,
one can state that there will be competition for UO2 in two potentially major chemical processes
corresponding to the above scenario: (1) reduction of UO2 by the graphite matrix accompanied by
the formation of UC2 and CO and (2) oxidation of UO2 to U3O8 by the ingress air. Furthermore,
the onset temperatures for both processes are pressure-dependent. Of the two processes, the
oxidation to U3O8 is probably the worst because it is accompanied by the specific volume change
of about 38% and subsequent nonlinear oxide layer growth.31 Detailed experimentation and first-
principles calculations could shed light on this very complex issue.
5. While thermal conductivity of UO2 and U3O8 is low, it is ~2 times higher for UO2
c, although the
exact ratio will be temperature-dependent.
6. Thermal expansion of UO2 increases monotonically with temperature, changing about 0.75% in
the 0 to 1000°C temperature range. However, in the case of U3O8, the situation is more complex.
Stoichiometric U3O8 changes continuously with increasing temperature, reversibly and
anisotropically, above 25°C with expansion along the a-axis and contraction along the b-axis
from orthorhombic to hexagonal symmetry at 350°C ± 10°C.26 A small but continuous
contraction along the c-axis occurs up to 1100°C. The loss of oxygen begins around 600°C. This
does not result in any discontinuity in both parameters of the hexagonal phase (i.e., “a” and “c”)
up to 875°C. However, in the range from 875 to 925°C, the structure undergoes a change to lower
symmetry as a result of contraction along the a-axis and expansion along the b-axis (expansion-
contraction anomaly). This structural change is accompanied by a more extensive loss of oxygen
and is usually irreversible. For these reasons, UO2 is somewhat more preferable from the thermal
expansion behavior standpoint as well.
7. Based on the considerations given above, from the physico-chemical perspective, the choice in
favor of UO2 versus U3O8 seems to be somewhat more justified for all stages of the fuel
production process, including the composite fuel compaction, manufacturing of the fuel rods, and
their subsequent exploitation in the TREAT reactor. This conclusion is substantiated by the fact
that U3O8 will undergo a partial transformation into UO2 during the developed production
process.
There are a number of things that still need to be explored further. One issue is development of
zirconium alloy tempering that would provide the best possible corrosion resistance for the selected
c At temperature ~900°C
32
Zr-based alloy (e.g., construction of the TTT and CCT-diagrams, analysis of mechanical behavior, and
selection of the optimal heat treatment). Secondly, the generalized Ellingham-Richardson diagram will
need to be constructed for the zirconium alloy of choice in conditions of air ingress and at normal
working conditions. Third, to gain fundamental understanding of the UO2–graphite reactions, it is
desirable to study the electronic density distributions for this reaction taking place under different
conditions (density functional theory). Finally, potential interactions between the fuel and the Zr-based
cladding need to be understood in detail.
10. REFERENCES
1. van Rooyen, I. J., N. E. Woolstenhulme, R. K. Jamison, E. P. Luther, and L. Valenti, 2015, TREAT
Conversion LEU Fuel Design Trade Study, INL External Report INL/LTD-14-31704, Revision 1,
November 2015.
2. Murty, K. L. and I. Charit, 2013, “Introduction to Nuclear Materials: Fundamentals and
Applications,” Wiley-VCH.
3. Thein, S. M. and P. J. Bereolos, 2000, Thermal Stabilization of 233UO2, 233UO3, and 233U3O8,
ORNL/TM-2000/82, July 2000.
4. Zumdahl, S. C., 2009, Chemical Principles, 6th Ed., Houghton Mifflin Company, ISBN 0-618-
94690-X.
5. Charit, I., 2012, “Introduction to Nuclear Materials,” University of Idaho.2
6. Ma, B. M., 1982, Nuclear Reactor Materials and Applications, 1st edition, Springer.
7. Popov, S. G., J. J. Carbajo, V. K. Ivanov, and G. L. Yoder, 2000, Thermophysical properties of MOX
and UO2 fuels including the effects of radiation, ORNL/TM-2000/351.
8. Walters, L. C., 2004, “Nuclear Fuel: Design and Fabrication,” in Encyclopedia of Energy 4: 341-350.
9. Papadias, D., 2014, “Thermal Analysis of a TREAT Fuel Assembly,” ANL/GTRI/TM-DRAFT, ANL
Nuclear Engineering Division, July 9, 2014.
10. Thermo-Calc v.2015a, Stockholm, Sweden, 2015.
11. Uranium Processing and Properties, 2013, J. S. Morrell and M. J. Jackson (eds.), Springer, Berlin.
12. Simnad, M. T., 2002, “Nuclear Reactor Materials and Fuels,” in Encyclopedia of Physical Sciences
and Technology, 3rd edition, Vol. 10, pp.775-815, Academic Press.
13. Caneiro, A. and J. P. Abriata, 1984, “Equilibrium Oxygen Partial Pressure and Phase Diagram of the
Uranium-Oxygen System in the Composition Range 2.61< O/U < 2.67 between 844 and 1371 K,”
Journal of Nuclear Materials 126: 255-267.
14. Thompson, W. T., B. J. Lewis, E. C. Corcoran, M. H. Kaye, S. J. White, F. Akbari, Z. He, R. Verrall,
J. D. Higgs, D. M. Thompson, T. M. Besmann, and S. C. Vogel, 2007, “Thermodynamic treatment of
uranium dioxide based nuclear fuel,” International Journal of Materials Research 98: 1004-1011.
15. Bykov, M. A. and I. A. Uspenskaya, 2003, “Thermodynamic Properties of Low-Stability Phases and
the Phase Diagram of the Uranium-Oxygen System,” Zhurnal Neorganicheskoi Khimii 48: 1715-
1717.
16. Pillai, S. R., S. Anthonysamy, P. K. Prakashan, R. Ranganathan, P. R. Vasudeva Rao, and C. K.
Mathews, 1989, “Carburization of stainless steel clad by uranium-plutonium carbide fuel,” Journal of
Nuclear Materials 167: 105-109.
17. https://en.wikipedia.org/wiki/Uranium_dioxide.
33
18. Parga, C. J., I. J. van Rooyen, M. V. Glazoff, and E. P. Luther, 2016, “TREAT Reactor LEU Fuel-
Clad Chemical Interaction Empirical-Modeling Analysis,” TopFuel 2016 ANS Conference, Boise,
Idaho, September 11 through 16, 2016, submitted.
19. Duriez, C., M. Steinbrück, D. Ohai, T. Meleg, J. Birchley, and T. Haste, 2009, “Separate-effect tests
on zirconium cladding degradation in air ingress situations,” Nuclear Engineering and Design.
20. Okamoto, H., 2015, “C-U (Carbon-Uranium), Binary Alloy Phase Diagrams,” Second Edition, T. B.
Massalski (ed.), Vol. 1, p 892-893, Journal of Chemistry, Vol. 2015, Article ID 142510.
21. Parker, G. W. et al., 1967, “Out of Pile Studies of Fission-Product Release from Overheated Reactor
Fuels at ORNL, 1955-1965,” ORNL-3981, July 1967.
22. Iglesias, F. C. et al., 1990, “Measured Release Kinetics of Ruthenium from Uranium Oxides in Air,”
Proceedings of International Seminar on Fission Product Transport Processes during Reactor
Accidents, J. T. Rogers (ed.), Hemisphere Publishing Corp, Washington DC.
23. Yu, J., B. Valderrama, H. B. Henderson, M. V. Manuel, and T. Allen, 2015, “Near Surface
Stoichiometry in UO2: A Density Functional Theory Study,” Journal of Chemistry 2015: 142510.
24. “Transient Reactor Test Facility (TREAT) FSAR,” 2014, Document ID: S3942-0001-YT, Revision 8,
November 12, 2014.
25. IAEA, 1965, “Thermal Conductivity of Uranium Dioxide,” Technical Report No. 59, Vienna.
26. Martin, D. G., 1988, Journal of Nuclear Materials 152: 24. Also see: Tokar, M. et al., 1973, Nuclear
Technology 17: 147
27. Pillai, C. G. S., A. K. Dua, and P. Raj, 2001, “Thermal conductivity of U3O8 from 300 to 1100 K,”
Journal of Nuclear Materials 288(2-3): 87-91.
28. Weisensee, P. B., 2011, “Thermal Conductivity of UO2 and U3O8 Epitaxial Layers Damaged by Ion
Irradiation,” M.S. Thesis, University of Illinois, Urbana-Champaign.
29. Glazoff, M. V., 2015, “Physical and Mechanical Metallurgy of Zirconium Alloys for Nuclear
Applications: a Multi-Scale Computational Study,” Ph.D. Dissertation in Nuclear Engineering, the
University of Idaho.
30. Glazoff, M. V., A. Tokuhiro, S. N. Rashkeev, and P. Sabharwall, 2014, “Zirconium, Zircaloy-2, and
Zircaloy-4: Computational Thermodynamics and Atomistic Perspective on Hydrogen Uptake and
Oxidation,” Journal of Nuclear Materials 444: 65-75.
31. Nuclear Corrosion Science and Engineering, 2012, D. Feron (ed.), Woodhead Publishing Limited.
Anderson, W. K., C. J. Beck, A. R. Kephart, and J. S. Theilacker, 1962, “Reactor Structural
Materials: Engineering Properties as Affected by Nuclear Reactor Service,” ASTM Special Technical
Publication No. 314
32. Szpunar, B., J. A. Szpunar, V. Milman, and A. Goldberg, 2013, “Implication of volume changes in
uranium oxides: A density functional study,” Solid State Sciences 24: 44-53.
