Examination of the Conversion of the ... Submarine Fleet from Highly Enriched ... to Low Enriched Uranium

advertisement
Examination of the Conversion of the U.S.
Submarine Fleet from Highly Enriched Uranium
to Low Enriched Uranium
by
Cameron Liam McCord
S.B., Massachusetts Institute of Technology (2013)
Submitted to the Department of Nuclear Science and Engineering
in partial fulfillment of the requirements for the degree of
Master of Science in Nuclear Science and Engineering
MASSACHUSETTS INSTITUTE
OF TEOI.-N'OLOGY
at the
JUL 2 9 2014
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
LIBRARIES
June 2014
@ Massachusetts Institute of Technology 2014. All rights reserved.
Signature redacted
Author .............................................
Department of Nuclear Science and Engineering
May 19, 2014
........................
Certified by...
R. Scott Kemp
Assistant Professor, Department of Nuclear Science and Engineering
Signature redacted
Signature redacted Thesis Supervisor
Certified by..................
4
Owen R. Cote
Associate Director, MIT Security Studies Program
Signature redacted Thesis Reader
Accepted by.................
( jz,/Mujid S. Kazimi
TEPCO Professor of Nuclear Engineering
Chair, Department Committee on Graduate Students
jkilla"ftm
Examination of the Conversion of the U.S. Submarine Fleet
from Highly Enriched Uranium to Low Enriched Uranium
by
Cameron Liam McCord
Submitted to the Department of Nuclear Science and Engineering
on May 19, 2014, in partial fulfillment of the
requirements for the degree of
Master of Science in Nuclear Science and Engineering
Abstract
The nuclear reactors used by the U.S. Navy for submarine propulsion are currently fueled by highly enriched uranium (HEU), but HEU brings administrative and political
challenges. This issue has been studied by the Navy before, and promising new fuel
technology developments, combined with new motivations, warrant a fresh look at low
enriched uranium (LEU). Evidence suggests that the mission space of the submarine
force might be changing in such a way to relax previous constraints that made LEU
cores more challenging. Estimates show that high density fuels can produce LEU
core sizes that are comparable to HEU core sizes. One of the difficult constraints will
be meeting the reactivity requirements to overcome xenon poisoning throughout the
lifetime of the core, especially at higher burnups, where LEU cores exhibit different
reactivity behavior. A new LEU design should not prove prohibitive from a thermal
hydraulic perspective, unless increases in core volume or fuel-to-moderator ratio require a significant increase in pumping power. LEU designs that require increases in
weight or size of the reactor compartment are not outright infeasible, and alterations
to the reactor compartment might provide a necessary recovery of weight and volume.
The concept of LEU redesigns resulting in unattractively large new cores that fail to
meet performance standards is a thing of the past. A major investigation of advanced
naval reactor technology and LEU designs should be made, and a thorough analysis
should be conducted soon.
Thesis Supervisor: R. Scott Kemp
Title: Assistant Professor, Department of Nuclear Science and Engineering
Thesis Reader: Owen R. Cote
Title: Associate Director, MIT Security Studies Program
2
Disclaimer
The views expressed in this thesis are those of the author and do not reflect the
official policy or position of the Department of Defense or the U.S. Government.
This thesis does not make use of any classified material; rather, it is derived from
material in open source literature.
3
Acknowledgments
First and foremost, I would like to thank R. Scott Kemp for first introducing me
to this topic and guiding me, from start to finish, in producing this thesis. I'd like to
thank him for his constant enthusiasm, his acceptance of nothing shy of perfection,
and for always taking the time to point out "teaching moments."
I'd like to thank my thesis reader, Owen R. Cote, for his guidance, interest, and
calming influence.
I'd like to thank Richard K. Lester for helping me get involved in the Nuclear
Science and Engineering Department and for being a constant source of advice and
mentorship.
I'd like to thank the faculty of the Nuclear Science and Engineering Department
for their collective help over the past four years. Specifically, I would like to thank
Dr. Alan Hanson, Dr. Michael Short, Dr. Tom Newton, and Dr. Arthur Baggeroer.
I'd like to thank CAPT Curtis Stevens, CAPT Steven Benke, and LT Matthew
Mink for helping me reach my goals: commissioning in the Navy, becoming a future
submarine officer, and being accepted into the Master of Science program here at
MIT.
Lastly, I would like to thank my family. I'd like to thank my brother, Brendan, for
spurring my interest in the submarine force and for trail blazing the MIT experience.
I'd like to thank my sister, Caitlin, for her perpetual optimism and kind words during
late nights. I'd like to thank my parents, Michael and Tara, for their love and support
and for all the sacrifices they've made over the past 23 years to help me get to where
I am today.
4
Contents
1
2
1.1
The Submarine's Mission . . . . . . . . . . . . . . . . . . . . . . . . .
10
1.2
Overview of Submarine Reactors
. . . . . . . . . . . . . . . . . . . .
11
1.3
Thesis Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14
1.4
Thesis Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15
17
Motivations for Investigation
. . . . . . . . . . . . . .
17
. . . . . . . . . . . . . . . . . . . .
19
2.2.1
Funding and Innovation for Naval Reactors . . . . . . . . . . .
19
2.2.2
Timing of Research . . . . . . . . . . . . . . . . . . . . . . . .
21
. . . .
21
2.3.1
The Non-Proliferation Treaty Loophole . . . . . . . . . . . . .
21
2.3.2
The Fissile Material Cut-Off Treaty and Implications for In-
2.1
Changing Role of Submarines in the Future
2.2
Bolstering Submarine Innovation
2.3
2.4
3
9
Introduction
Addressing Problems with International Treaty Requirements
spection Security . . . . . . . . . . . . . . . . . . . . . . . . .
24
Reducing the Need for HEU Stockpiles . . . . . . . . . . . . . . . . .
25
Previous Work on Conversion to LEU
27
3.1
The Navy's 1995 Report on LEU Conversion . . . . . . . . . . . . . .
28
3.2
The Ippolito Study . . . . . . . . . . . . . . . . . . . . . . . . . . . .
30
3.3
The Ma and von Hippel Study . . . . . . . . . . . . . . . . . . . . . .
31
3.4
The Glaser and von Hippel Study . . . . . . . . . . . . . . . . . . . .
32
3.5
The Miller Speech . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
5
The Ward Study .............................
33
3.7
The Navy's 2014 Report on LEU Conversion . . . . . . . . . . . . .
35
.
3.6
4 Fuel and Neutronics Considerations
37
General Fuel Criteria ...........................
. . . .
38
4.2
LEU Fuel Design . . . . . . . . . . . . . . . . . . . . . . . .
. . . .
40
4.3
Fuel Weight and Volume Estimates . . . . . . . . . . . . . .
. . . .
43
.
.
4.1
Current HEU-Core Estimates
. . . . . . . . . . . . .
. . . .
44
4.3.2
New LEU-Core Estimates . . . . . . . . . . . . . . .
. . . .
46
4.3.3
Comparing HEU and LEU Core Estimates . . . . . .
. . . .
47
4.4
Reactivity and Poisoning . . . . . . . . . . . . . . . . . . . .
. . . .
48
4.5
Reactivity Behavior Differences in LEU Cores . . . . . . . .
. . . .
53
4.6
Considerations on Safety, Fuel Disposal, and the Environment . . . .
60
4.7
Sum mary
. . . .
62
.
.
.
.
.
4.3.1
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
5 Chemistry and Thermal Hydraulics Considerations
Chemistry and Corrosion Constraints . . . . . . . . . . . . .
64
5.2
Selection of Core Geometry and Modeling
. . . . . . . . . .
65
5.3
Thermal Hydraulic Constraints
. . . . . . . . . . . . . . . .
69
5.3.1
Minimal Flow Rates and Noise Production . . . . . .
71
5.3.2
"Walk Away" Ready Heat Decay Requirement . . . .
73
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
74
Summary
.
.
.
.
.
.
5.1
5.4
75
The Speed Requirement
. . . . . . . . . . . . .
75
6.2
General Constraints on Reactor Weight and Size
.
77
6.3
Specific Constraints on Reactor Weight and Size
78
6.4
Shielding as a Source of Size and Weight . . . .
80
6.4.1
Occupational Radiation Exposure . . . .
80
6.4.2
Design Considerations with Shielding and Safety for an LEU
.
.
.
6.1
.
Physical Constraints
C ore . . . . . . . . . . . . . . . . . . . .
.
6
64
6
81
6.5
Strategic Capability, Submarine Readiness, and Deployment Schedules
82
6.6
Minimizing Time in Port . . . . . . . . . . . . . . . . . . . . . . . . .
85
6.6.1
Refueling and Hulls . . . . . . . . . . . . . . . . . . . . . . . .
85
6.6.2
Effects of Refueling on Depth Performance . . . . . . . . . . .
86
. . . . . . . . . . . . . . . . . . . . . . .
87
6.7.1
Reactor Compartment Alterations . . . . . . . . . . . . . . . .
88
6.7.2
The French Integrated Design Revisited
. . . . . . . . . . . .
89
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
91
6.7
6.8
7
Maximizing Time Deployed
Sum mary
Conclusion and Recommendations
92
7.1
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
92
7.2
Recommendations for Future Work . . . . . . . . . . . . . . . . . . .
94
7
List of Figures
4-1
High Density Fuels.[2]
.......................
43
4-2
Fuel Weight and Volume Estimates. . . . . . . . . . . . . . . .
48
4-3
Reactivity Contribution from Xenon.[3] . . . . . . . . . . . . .
50
4-4
keff Required vs. Burnup. . . . . . . . . . . . . . . . . . . . .
52
4-5
keff Actual vs. Burnup.
. . . . . . . . . . . . . . . . . . . . .
53
4-6
keff vs. Burnup for 2%, 5%, 20%, and 97.3% Enrichment.
56
4-7
Reactivity vs. Burnup for 2%, 5%, 20%, and 97 .3% Enrichment.
56
4-8
CASMO Burnup Parameters.
4-9
CASMO Data .
.
.
.
.
PW R Plant.[1. ...........................
.
13
1-1
. .
57
. . . . . . . . .
58
4-10 Pin-to-Pitch Ratio. [4] ........
. . . . . . . . .
59
4-11 K-eff vs. Pin-to-Pitch Ratio.[4]
. . . . . . . . .
59
.
.
.
.
...........
.
. . . . . . . . .
.
Critical Heat Flux.[4] ........
66
5-2
Maximum Fuel Temperature.[4]
67
5-3
MITR Plate Fuel Design.[2]
6-1
Submarine Propulsion System.[1]J
79
6-2
Naval Reactor Shielding.[1] .
81
6-3
French Integrated Design.[5] ...
.
5-1
...
68
...
90
8
Chapter 1
Introduction
The nuclear reactors used by the U.S. Navy for submarine power and propulsion are
currently fueled by highly enriched uranium (HEU).'[6] That fuel provides a high
power density, compact size, and a near-lifetime reactor core without the need for
refueling. It also maximizes submarine availability. Despite these technical advantages, HEU brings administrative and political challenges.
For these reasons, the
possibility of the Navy switching to a low enriched uranium (LEU) fuel has been
discussed and studied for many years. 2 [6] In recent years, promising new fuel technology developments, combined with new motivations, warrant a fresh look at LEU
fuel. These motivations include: (1) a changing role for submarines in the future
that reduces operational constraints; (2) supporting submarine reactor innovations;
(3) reducing proliferation risks that arise from the Non-Proliferation Treaty loophole,
and enabling a more comprehensive Fissile Material Cutoff Treaty; and (4) reducing
problems arising from the need to maintain HEU stockpiles.
The history of submarine warfare and naval reactor design is complex. In order to
meaningfully consider the creation of new designs, it is important to understand the
role of the submarine and the motivations behind the technical constraints affecting
naval reactors.
1
2
HEU is uranium containing 20% or more of the isotope uranium-235.
LEU is uranium containing less than 20% of the isotope uranium-235.
9
1.1
The Submarine's Mission
The missions and purposes of military submarines have been changing constantly
since their first primary military impact during World War I. The use of German
U-boats to impinge upon civilian open-sea traffic in World War I helped to set a
precedent for the importance of undersea warfare in World War II. The Cold War era
brought about important new submarine technologies. Most notable was the creation
of the first nuclear-powered submarine in 1955, the USS Nautilus. The combination of
nuclear power and a new teardrop-shaped hull increased submerged operating speeds
and removed propulsion constraints on operational endurance. Nuclear power allowed
submarines to stay submerged longer, with no requirements to snorkel for oxygen.[7]
Only food and supplies limit mission duration.
These improvements enabled sub-
marines to undertake new missions. In the 1960s, the U.S. Navy's submarine force
was split into the two major components that exist today: ballistic missile submarines
(SSBNs) and fast attack submarines (SSNs).[7] These submarines serve the following
roles:
* Peacetime Operational Support: The secrecy and stealth of submarines allow
them to operate undetected in potentially hostile territory in a specific region
until they are needed.[7]
" Intelligence, Surveillance, Reconnaissance (ISR): ISR missions are perhaps the
main use of fast attack submarines today.
A submarine can easily infiltrate
an area and watch and listen without being detected, allowing it to intercept
communications and gather valuable intelligence. Although the overall number
of missions has dropped because the number of fast attack submarines has
decreased, the proportion of missions that are ISR missions has more than
doubled in the past two decades, as information-gathering technologies and
experience have improved.[7]
" Special Operations: Submarines provide an option for working with and deploying special operations forces. They can stealthily allow troops to be inserted
10
and removed from operating areas without detection.[7
" Precision Strike: Submarines have the ability to surface and fire precision strike
missiles, such as Tomahawk Land Attack Missiles (TLAMs) from distances of
over 650 miles. With the creation of the newer Guided Missile Submarines
(SSGNs), which can carry up to 154 Tomahawk missiles, the strike capability
of the submarine force has increased. [7]
" Sea Denial: Perhaps one of the most basic and important missions of submarines
is sea denial, whether it is blocking enemy ports, pursuing suspicious or dangerous sea vessels, or escorting and enabling peaceful civilian ocean trade. [7]
These missions form the backbone of the U.S. Navy's submarine force, and the
overall demand for submarines to carry out such missions can be expected to increase
in coming decades even though the number of available submarines may sometimes
decrease.
In particular, the need for cost-effective methods for gathering unique
intelligence and the need for covert Special Forces as a preferred option for conflict
resolution can be expected to grow in the future. [8] Similarly, the demand for capable
fast attack submarines will likely continue to expand as more sophisticated weapons
and detection systems proliferate in the coming decades. [8] With a greater focus
on Asia, the Pacific area, and China's rapid naval modernization, more submarines
are being switched to west coast and Pacific ports. President Obama has openly
announced a "rebalancing" in Asia, and submarines can be expected to play a key
role in this effort. [9]
1.2
Overview of Submarine Reactors
This section describes the relevant history of naval reactors and the origins of some
of the main technical constraints.
The Navy's expertise in developing and operating nuclear reactors is longer running than that of any other organization in the world. Naval reactors have undergone complex experimentation and optimization since their introduction in the 1950s.
11
Currently, the naval reactors used in U.S. submarines are pressurized water reactors
(PWRs) that run on HEU enriched to above 93% uranium-235.[3] This higher enrichment level has been standard for the past several decades as it gives submarines a
longer core lifetime, which allows them to operate for longer periods of time without
having to refuel. 3 [3] Currently, the new Virginia Class submarine has a core lifetime of 33 years, which is the same amount of time that the boat is expected to be
operational. [101
Several important factors distinguish naval reactors from terrestrial power reactors. The thermal efficiency of naval reactors is lower than that of terrestrial reactors
because variable power demands are incompatible with optimization of the thermal
cycle.[3] In addition, naval reactors must be extremely compact to fit within the dimensions of the hull of the submarine and to limit the weight of the core.[5] Much of
this space and weight comes from the large amounts of shielding that are necessary
to block neutron and gamma radiation and allow the submarine crew to work and
operate safely near the reactor compartment.[3] Naval reactors also must be able to
restart rapidly after a shutdown in case of an emergency, while terrestrial reactors
can use a more gradual restarting approach.4 Finally, naval reactors must be able
to withstand battle conditions and changes in orientation as the submarine dives
and rises.5 In general, naval reactors are designed with an emphasis on performance
and power demand as opposed to economic optimization. Figure 1-1 shows a typical
schematic of a PWR plant.
On a submarine, the plant consists of two main compartments: a shielded primary
compartment that contains the reactor, steam generator, and primary coolant loop;
and an unshielded secondary compartment that contains the steam turbines, drive
train, and condensers.[5] A modern pressurized water submarine reactor generally
'HEU has been used as the standard for naval reactor fuel since 1953.
4
Naval reactors need to be able to restart rapidly in the case of a reactor malfunction or accidental
SCRAM (complete shutdown) of the reactor. The submarine needs propulsive power at all times.
Terrestrial research and commercial reactors can restart over hours or days.
5
Battle conditions mean that the submarine is expected to remain operational in the event of a
depth charge or torpedo strike (both of which would result in massive vibrations and a shock wave).
Submarines typically dive at angles of as much as 30 degrees and naval reactors have to be able to
handle this change in pitch and any effects it might have on the control rods or coolant flow.
12
STEAM
ITO TURBINE)
SECONDAR Y
LOOP
FEEDwATER
-
CONTROLS
CONDENSERI
--
CORE
DOWNCOMER
---
SHROUD-
REACTOR
VESSEL
STEAM GENERATOR
PRtMARY LOOP
PUMP
WATER
Figure 1-1: PWR Plant.[1]
contains a primary loop that is pressurized to 15 MPa with an inlet temperature of
about 290 degrees Celsius and an outlet temperature of about 320 degrees Celsius. [3]
The outlet water enters a steam generator where it transfers its heat to cooler water
in the secondary loop that is pressurized to 7 MPa and has an inlet temperature of
about 225 degrees Celsius and an outlet temperature of about 285 degrees Celsius. [3]
This secondary loop produces steam that drives a turbine to produce shaft rotation
and propel the submarine.
The Navy has experimented with many reactor designs over its history and has
a wealth of experimental and operational experience. The first successful prototype
design was the STR (Submarine Thermal Reactor) in 1953, which sustained full power
operation for 96 hours.[3] This was followed by the S1W in 1955, which sustained full
power operation for 66 days. This SiW reactor became the design of choice for the
first nuclear submarine, the USS Nautilus.[3] This reactor design included uraniumzirconium fuel plates with a zirconium cladding. [3]
The SiW design accomplished the original goals of a submarine reactor at that
13
time: it was safe and provided the necessary performance in testing.6 [3 This specific design had a maximum fuel temperature of 645 degrees Celsius, and the reactor
temperature was ultimately limited by the ability to maintain high pressures in the
pressure vessel as well as physical limitations that accompanied such a demand.[3]
Because of these high pressures and the accompanying pressure drop resulting from
the temperature drop in the coolant, a large amount of pumping power was initially
needed to circulate coolant.[3] However, the Navy made improvements over many iterations in the next several decades and arrived at the current state-of-the-art reactor
in the Virginia Class submarine, the S9G. This design has advanced steam generators
and improved corrosion resistance, thereby reducing the lifecycle cost. It also has a
very high energy density, which eliminates the need for large pumping capability.[3]
1.3
Thesis Objectives
The first objective of this thesis is to assess the validity and strength of the reasons
set forth in the introduction for an investigation of the feasibility of converting naval
reactors to LEU fuel.
The second objective of this thesis is to understand and analyze the naval reactor design constraints that would be imposed on a new LEU reactor design. These
include the standard naval reactor submarine requirements of compactness, crew protection/safety, reliability, ruggedness, maneuverability, endurance, and quietness.[11]
The selection of a fuel type, density, and geometry is constrained by the need
for the submarine to withstand "battleshock" and severe vibrations. 7 [11] In addition,
the fuel will need to be designed in such a way that power can be transferred to the
coolant with very low flow rates to ensure that the submarine is quiet.[11}
The new LEU core must also meet all of the current requirements of naval reactors for power and reactivity, because of the need for submarine naval reactors to
stop/start quickly. A reasonable estimate for required reactor power is between 135
6
The S1W design was tested extensively from 1951 to 1953 and achieved the desired operating
conditions with few or no accidents.
7
Naval reactors could expect up to 5OG's of force.
14
and 140 MWt.
The reactor is expected to operate at an average of 15-20% of full
power based on a typical combination of a fast attack submarine rapidly transiting
to a location, followed by reduced power use during operations. 9 The new core must
meet these varying power demands within the tolerated limits for thermal stresses on
the fuel cladding, and be able to overcome (with excess reactivity) the effects of poisons that are produced by normal operation. 10 In general, naval reactors are expected
to operate at maximum power for extended periods of time and be able to quickly cut
power to lower levels and restart even after short-term or long-term poisoning from
xenon and samarium.
Constraints on reactor size also exist, and the most successful redesigns will not
increase core size. However, if the core size must be increased to reach the same (or
other acceptable) performance characteristics, the size increase should be minimized
or compensated for by improvements in other areas. The current core lifetime for naval
reactors is approximately 33 years. Any new core design project, while investigating
refueling options, should strive to produce a meaningful core lifetime, capable of
satisfying the Navy's requirements for fleet size and fleet availability."
Finally, in combination with complementary changes, any new reactor will need
to meet existing operational, logistical, and strategic requirements within reasonable
budgetary and resource constraints.
1.4
Thesis Overview
The above objectives will be revisited throughout this thesis. Chapter 2 examines the
motivations for conversion in greater detail. Chapter 3 reviews previous studies on
the conversion of naval reactors to LEU and the Navy's past and emerging positions
regarding reactor conversion.
Chapter 4 defines fuel and minimum neutronic and
8
From a conversation with former Naval Reactor Engineer.
From a conversation with former Naval Reactor Engineer.
10 "Poisons" refer to fission products that result from normal operation and act as neutron absorbers, thus "poisoning" the normal expected neutron performance of the core.
"The Seawolf Class submarine has a core lifetime of 30 years, and the Virginia Class submarine
has a core lifetime of 33 years.
9
15
lifetime requirements for an effective naval reactor that meets operational needs.
Chapter 5 examines the chemistry and thermal hydraulic properties needed for safe
and reliable operation. Chapter 6 examines physical, geometric, weight, and shielding
constraints.
Finally, Chapter 7 provides conclusions and suggested directions for
future studies on naval reactor conversion.
16
Chapter 2
Motivations for Investigation
2.1
Changing Role of Submarines in the Future
The "traditional" mission of the submarine, especially the smaller fast attack submarine, has significantly changed over the past several decades. Submarines are not as
often performing their World War II or Cold War missions of chasing, tracking, or
destroying enemy submarines. The mission space of the submarine force has shifted
more toward stealth missions; Intelligence, Surveillance, and Reconnaissance (ISR)
missions; and Special Forces-related deployments, all of which, in general, do not require the same amount of average propulsive power by the submarine.'[8] As a result,
the mission requirements of increased underwater speed and rapid transit are less
stringent than in the past.2 If the mission space during the Cold War era resulted in
a typical submarine operating at an average level of 25% of maximum power over its
lifetime, the mission space today is closer to 15%.3 This trend toward reduced power
demands is also likely to continue in the future.
A lower average power profile means that the reactor cores of submarines require
'Stealth missions, ISR missions, and Special Forces operations are typically characterized by a
period of transit (high power) followed by long periods of quiet submerged activity (little or no
power).
2
These mission requirements will still remain, but in smaller proportions than previously was the
case.
3The actual decrease
is unknown. The important point is that the average power demand over
the lifetime of the submarine is decreasing.
17
less total energy (in the form of uranium-235 that can be fissioned) at the beginning
of their lives. This reduction in energy demand means that less total energy needs to
be extracted from the core during the lifetime of operation of the submarine. This
energy extraction quantity of a core is known as core burnup.4
The reduction in
average power demand outlined above translates to an average of 40% lower average
core burnup per submarine. Finally, a reduction in average core burnup per submarine
means that, for a given level of enrichment, cores can be smaller and more compact
because they do not initially need to contain as much fuel.5 As long as the average
burnup per submarine stays below a design threshold across the fleet, reductions in
average burnup per submarine appear possible. 6
This reduction in average burnup per submarine makes the use of LEU cores more
promising. One of the initial problems with LEU core designs was the challenge of fitting the required amount of uranium-235 needed for lifetime energy extraction into a
sufficiently compact space. However, with a reduction in the necessary energy extraction, this problem becomes simpler to solve. Accordingly, any significant reduction in
the average burnup per submarine helps to improve the potential feasibility of LEU
cores.
Because of the changing nature of the modern submarine's mission space and
the structure of the submarine force over the next several decades, the performance
requirements to be met by individual submarines should be studied closely. Although
naval reactors are moving toward missions that require less fuel, the Navy expects
that the number of submarines in its fleet will drop below what it currently predicts
is the requisite number of submarines to accomplish its mission.'[10] As a result, the
4The "core burnup" is defined as the energy extracted from the fuel per total mass of the fuel.
Thus, a lower average core burnup for a submarine means less energy is ultimately extracted from
the core in that submarine.
5
The amount of fuel in the core at the beginning of the submarine is proportional to how much
energy the core contains. If the energy requirements decrease, the size can decrease as well.
6
1t is also important to recognize that a reduction in burnup will not affect the submarine's overall
top speed. Submarines with a reduced core burnup will still meet the same peak power demands
that submarines can currently meet because peak power is not a function of core burnup, but rather
a function of neutronics, fuel temperature, and thermal hydraulics.
7The number is expected to drop from the current 54 submarines below the stated requisite of
48 submarines in FY2025.
18
Navy is facing difficult budgetary tradeoffs and is striving to get the longest possible
lifetime out of each submarine and to minimize maintenance, refueling, and other
associated costs. These forces suggest preserving long-lived HEU cores, unless LEU
cores can offer compensating savings or advantages. The Navy's analysis, therefore,
will need to take account of all relevant factors. These factors include the question
of how conversion to LEU fuel will affect future requirements and, indeed, whether
LEU cores might offer some non-technical advantages over traditional HEU fuels.
2.2
2.2.1
Bolstering Submarine Innovation
Funding and Innovation for Naval Reactors
Beyond mission considerations, an LEU conversion program may bring certain organizational advantages to the Navy. The Office of Naval Reactors stands to benefit
greatly from a thorough investigation of new and advanced naval reactor fuel technologies and high density fuels. This investigation could bring an increase in funding,
promote technology innovations, and result in knowledge that would lead to new LEU
reactor designs or new optimizations to existing HEU reactor designs. Currently, most
of Naval Reactors' resources (experts, laboratory resources, and program funding) are
being used on the Land-Based Prototype Reactor Program and the Ohio Class SSBN
Replacement Program.[12] As the Ohio Replacement program begins to wind down,
a new project investigating advanced naval reactor fuel technology and high density
fuels, both enabling technologies for LEU conversion, would bring increased funding,
innovation, and expertise that will benefit Naval Reactors.
A thorough re-examination of the pros and cons of converting submarines to LEU
fuel might take five years and one to two million dollars, a relatively small effort
for the Navy.8
The design of a functioning LEU core, however, would be a much
larger project for Naval Reactors. In 1995, the Navy estimated that this might cost
around $800 million, supporting Naval Reactors research for at least 20 years.[11] A
8
A reasonable estimate would be a five-year study costing approximately $1.25 million.
19
compelling case to investigate this issue would help to ensure a steady flow of funding
to Naval Reactors over the coming decades. This funding could pay dividends in the
future, no matter what the final outcome of the investigation. For example, a steady
source of research funding would enable research to be accomplished much more
efficiently and quickly. By contrast, fluctuating levels of annual funding for a project
generally result in research personnel being switched to and from other assignments
and government contractors being alternately hired and retained, thereby wasting
significant portions of the funding because of the resulting inefficiencies. A steady
source of adequate funding is also more likely to enable researchers to explore very
novel designs, which could lead to discoveries that will prove to be beneficial in the
future regardless of whether the reactor cores use LEU or HEU as fuel.
Second, in order to prevent a very possible "innovation plateau" in naval reactor
technology and to avoid allowing the United States' competitors to narrow the submarine technology gap, it is important that the United States keep investing in naval
propulsion research. A steady source of funding could help ensure that the United
States will continue to own the world's best submarines in the future and could be
used to create broader innovations in submarine technology. 9 [13]
A reliable source of funding and an environment that promotes innovation would
foster creativity in the work force and enable researchers at Naval Reactors to participate in more cutting-edge, diverse, and interesting projects than would otherwise be
the case.
0
This increase in attractive work projects may help recruit new engineers
and retain current talent.
Finally, at the conclusion of the investigation, Naval Reactors will have gained a
substantial amount of knowledge regarding LEU conversion and, even more broadly, a
vast amount of knowledge regarding advanced naval reactor fuel technology and high
9
This is derived from conversations with submarine expert Owen Cote, in which he suggests that
innovation in the submarine community might have leveled off in the past decade or so after marked
increases in innovation driven by Cold War era budgets and technology increases. He suggests that
Naval Reactors might be looking for a new area in which to devote more time and energy and that
this LEU fuel source could be that area.
0 These projects refer to private sector opportunities to work on
advanced reactors (Generation
IV, Generation IV+), other new reactor technologies, or other projects requiring similar engineering
expertise.
20
energy density fuels. This knowledge will allow Naval Reactors either to implement
a conversion to LEU reactors or, if the conversion to LEU proves to be unworkable,
to explain exactly why the proposal is not credible, putting to rest the long-standing
push for conversion.
2.2.2
Timing of Research
Large-scale naval construction programs generally last for several decades, from conception to initial planning to completion. For example, the origins of the Ohio Class
SSBN Replacement Program can be traced back to 2006.[14] The new SSBNs are
scheduled to become active in 2021. This time frame lasts 15 years and likely entails
a less dramatic re-design than the switch from HEU to LEU. The need for a program
to design and construct new fast attack submarines to ultimately replace the current
Virginia Class SSNs will be required to begin as soon as 2015 in order to deliver
submarines starting in 2030.11 The conversion to LEU fuel is a complex issue that
needs to be investigated in conjunction with new submarine designs and should begin
in earnest as soon as possible to most effectively inform the new design processes.
2.3
Addressing Problems with International Treaty
Requirements
2.3.1
The Non-Proliferation Treaty Loophole
The Treaty on the Non-Proliferation of Nuclear Weapons (NPT) is the primary international agreement controlling nuclear materials and restricting their use in producing
nuclear weapons. Among its multiple aims, the NPT prohibits non-nuclear-weapons
countries from using nuclear materials to produce nuclear weapons and requires that
stockpiles of such materials be subject to safeguards provided by the International
"This assumes both (1) that the U.S. Navy will want new fast attack submarines online in 2030
and (2) that a new fast attack submarine construction program will follow the same general timeline
as the Ohio Class Replacement Program.
21
Atomic Energy Agency (IAEA). However, the NPT contains a "loophole" that can be
used by non-nuclear-weapons states to exclude quantities of directly weapon-usable
HEU from IAEA safeguards. In particular, the NPT implicitly exempts HEU for nuclear propulsion from IAEA safeguards. Thus, any non-nuclear country could publicly
announce that it will produce or stockpile HEU for a reactor in a nuclear submarine
to be constructed in the future but covertly intend to use the HEU for the immediate
or future production of nuclear weapons.[15]
Although this possibility has existed for many years, the need to address the NPT
loophole has become more urgent in recent years. At the time the NPT was drafted,
it seemed implausible that countries of proliferation concern could produce large
quantities of HEU. Today, that has changed, as technology to enrich uranium has
become simpler and more countries view uranium enrichment as providing a nuclearweapon option.[16] Furthermore, potential proliferators have already indicated that
they may utilize the naval propulsion loophole. Iran announced in July 2012 that it is
planning to undertake steps to use nuclear submarines in its navy. [17] Moreover, Iran's
atomic energy director specifically stated that if HEU were needed for its submarines,
Iran would inform the IAEA so that the agency could take steps to ensure that the
HEU would be available. 12 [17] Given international concerns about Iran's past interest
in nuclear weapons, as well as Iran's extremely difficult security situation, it is not
hard to imagine that the HEU could be diverted for use in producing nuclear weapons.
In addition, India, Brazil, Argentina, Pakistan, and Venezuela have also expressed
their intentions to lease or build nuclear-powered submarines sometime in the near
future. [18] Brazil will apparently be the first country to test the NPT "loophole"
allowing non-nuclear countries to remove nuclear materials from IAEA safeguards for
non-explosive military activities. Brazil is believed to be negotiating with the IAEA
for an exemption from safeguards when its nuclear material is being used for naval
propulsion purposes in a future nuclear submarine. 13 [19]
12
In addition, given more recent discussions between the P5+1 countries and Iran regarding their
nuclear program and enrichment capabilities, it is unclear to what extent demands like this will be
recognized.
1 3 Although Brazil currently plans to use LEU as fuel for its nuclear submarines, it would retain
the option to use HEU at some point in the future.
22
Another concern is the Fissile Material Cutoff Treaty (FMCT), which is a proposed
international agreement that would limit the production of fissile material for nuclear
weapons. The United States has endorsed the FMCT negotiations and has proposed
that the treaty cover all levels of HEU, not just weapons-grade HEU (greater than
90% uranium-235) as proposed by Russia. The United States has also recognized
that there are legitimate civilian and military uses for fissile material and that the
treaty must take into account those uses while preventing the use of such material
for weapons production.[20}
It is in the interest of the United States and international security community
broadly to close the NPT loophole and to prevent the creation of a similar loophole
in the FMCT. Otherwise, such loopholes provide opportunities for renegade countries
to avoid IAEA safeguards and use HEU for nuclear weapons.
In addition, the United States has worked to reduce civilian uses of HEU and to
convert those facilities to the use of LEU in order to reduce the risk that HEU might
be stolen. It has supported United Nations Security Council Resolution 1887, which
in 2009 called upon all countries to minimize to the greatest extent that is technically
and economically feasible the use of HEU for civilian purposes, including by working to
convert research reactors and those reactors involved in radioisotope production. [17
However, the risk of theft is not limited to civilian fuel cycles. Conversion of naval
propulsion to LEU would further reduce the risk of nuclear terrorism.
Current attempts by the United States to close loopholes and reduce the availability of HEU are weakened because HEU is still used in its nuclear submarines and
ships.14[17} Of the five NPT nuclear-weapons states, three of them (the United States,
the United Kingdom, and Russia) use HEU in their nuclear submarines. The United
States and the United Kingdom use weapons-grade HEU (uranium-235 levels of more
than 90%) in their submarines, while Russia's submarines are believed to be powered
by HEU at 21% to 45% enrichment levels. [21] Both France and China now use LEU
in their nuclear fleets. 5 [18]
14According to the U.S. Naval Registrar, there have been 209 nuclear ships and are currently 72
commissioned nuclear vessels (10 CVNs, 14 SSBNs, 54 SSNs, 4 SSGNs).
"Referring primarily to the French Rubis-class submarine (7.5% enrichment) and the new Chinese
23
One of the primary obstacles to improving the United States' negotiating position is that the U.S. Navy has maintained that switching its nuclear fleet to LEU
could create serious performance problems, but it has not yet publicly comprehensively studied the potential advantages of LEU in detail. [111 The United States would
strengthen its position if the Navy had conducted such a study so that it could speak
with more authority about the feasibility of using LEU.
The United States would occupy a central role in any future effort to limit, by
treaty, the production of HEU by other nations, and to close current existing loopholes that allow the use of HEU for nuclear propulsion. The United Kingdom relies
heavily on the United States' nuclear propulsion technology and as a result will almost
certainly follow the United States' lead in addressing HEU.[22] Russia has already
lowered the enrichment level of the HEU used for its naval reactor fuel to about
21% to 45% and therefore would presumably be more willing to take steps to control weapons-grade HEU.[21] If the United States' submarines could ultimately be
converted to LEU fuel, it might then be possible to ban the use of HEU by all countries, thereby eliminating a current proliferation risk and increased threat of nuclear
terrorism that persists because of the use and potential use of HEU by others.
2.3.2
The Fissile Material Cut-Off Treaty and Implications
for Inspection Security
Although an unofficial draft, a proposed version of the FMCT treaty now in wide
circulation suggests in Article I that all signatories would be required to declare to
the IAEA all fissile materials in both the civilian sector and those in use in military
reactors, including fuel used for current naval reactors.[23] In addition, states would
agree to disable and dismantle all fissile material production capabilities.
Lastly,
states would be required to submit to IAEA inspection to ensure that all fissile material was being disposed of and handled properly. Such drafts of the FMCT often also
propose a comprehensive verification regime to inspect and verify quantities of HEU
Xia-class submarine (5% enrichment).
24
used in naval reactors. All of these terms could have deleterious and unpredictable
strategic effects on the U.S. Navy by exposing sensitive and classified naval reactor
information as well as general confidential data about U.S. Navy submarines, their
operations, and their capabilities.1 6 Switching to LEU would allow the Navy to avoid
fully these potential intrusions that might arise from an FMCT.
2.4
Reducing the Need for HEU Stockpiles
The United States no longer enriches HEU, but as of January 2013 it was estimated
that its total HEU inventory was about 604 metric tons. Of this total, 100 metric
tons had already been fabricated into fuel for naval reactors, and another 159 metric
tons had been designated for the naval-fuel reserve. [24] All future HEU used for naval
reactors is to come from a combination of this stockpile and from the reprocessing of
pre-existing uranium from weapons.
Because the large-scale reprocessing of weapons material is unpredictable and
limited" and the likelihood of the United States producing more HEU is in the near
term seems unlikely,1 8 the Navy may find itself unable to ensure the continued supply
of HEU in a timely manner.[25}[261 Although current HEU stockpiles would not be
depleted for several more decades, the problem of replenishing those stocks will need
to be addressed far (probably decades) in advance.' 9 One way to ensure the continued
6
This assumes that inspectors would routinely examine naval reactor fuel and potentially even
fuel in cores in active submarines. This would require bringing submarines into port (potentially
dry-dock) for the inspectors to examine. Also, inspections could be random and would then require
submarines to pull in at inconvenient times and for unknown amounts of time. Finally, inspectors
would be given access to the reactor compartment for inspections.
' 7 While current political rhetoric suggests that disarmament and reprocessing trends will continue,
there are also key examples of reprocessing programs that presently are drastically over budget
and/or temporarily halted, such as the project to construct the Mixed-Oxide Fuel Fabrication Facility
(MFFF) at Savannah River.
8
Currently, the United States has no American-owned uranium enrichment capacity. USEC,
the domestic company that was performing enrichment, closed its last facility in June 2013 and
has declared bankruptcy after years of financial struggles. The enrichment that is occurring in the
United States is being performed in New Mexico by a European consortium using European and
Russian technology.
9
Although it is not possible to predict precisely when the HEU inventory will be depleted, it is
possible to make a rough estimate of when this might occur. It was estimated that the United States'
total HEU inventory in 1996 was almost 750 metric tons. "Highly Enriched Uranium: Striking a
Balance," Report of the National Nuclear Security Administration, January 2001. As indicated
25
success of the U.S. submarine fleet far into the future might be to reduce the need for
HEU by considering LEU cores prior to the expiration of current HEU stocks. HEU
stockpiles would continue to decrease for several more decades as HEU naval reactor
submarines are slowly phased out and LEU naval reactor submarines are phased in. In
addition, if a transition to LEU fuel is eventually successful, the United States could
continue to reduce its HEU stockpiles as HEU fuel is blended down or fabricated into
LEU fuel. Trends already indicate that HEU stockpiles are dropping from this downblending process.[?] This means that HEU quantities would continue to decrease and
the long-term goal of eliminating all HEU quantities and sources would remain intact.
above, the United States' total HEU inventory in 2013 was approximately 600 metric tons. Thus,
one could estimate that the HEU inventory decreases by almost ten metric tons per year (150/17).
At this rate, the HEU inventory would be gone in roughly 60 years.
26
Chapter 3
Previous Work on Conversion to
LEU
There has been a longstanding interest in the conversion of naval reactors from HEU
to LEU; this is not a new area of concern. MIT Professor Emeritus Marvin Miller
first expressed interest in the area in 1974.1 In 1990, an important thesis on the topic
by MIT graduate student Thomas Ippolito contained analyses of some possible LEU
reactor core designs for submarines. Following that, in 1995 the Office of Naval Nuclear Propulsion issued a major report assessing the feasibility of converting the U.S.
Navy's submarine fleet from HEU to LEU.[11] Over the past two decades, countries
with major nuclear submarine programs have considered the feasibility of using LEU
as fuel for submarines, and scholars have intermittently discussed this possibility. At
this time, both France and China power their nuclear submarines with LEU rather
than HEU.2 [18] The other three countries with large nuclear submarine fleets: the
United States, the United Kingdom, and Russia, continue to utilize HEU in their
submarines.
'From a private conversation with Professor Marvin Miller.
Referring primarily to the French Rubis-class submarine (7.5% enrichment) and the new Chinesexia Class submarine (5% enrichment).
2
27
3.1
The Navy's 1995 Report on LEU Conversion
In the 1995 Report on the potential conversion to LEU from HEU, the Navy's Office
of Naval Nuclear Propulsion outlined several problems that it believed would result
from converting its submarines and carriers to LEU. [11] Those problems fell into two
main categories: (1) the use of LEU in existing submarines would require that nuclear
cores be replaced before the lifetime of the submarine had otherwise ended; and (2)
redesigning naval reactors and submarines to use LEU would be extremely expensive
and disruptive. [11]
The 1995 Report stated that conversion to LEU based on current reactor designs
would significantly affect the endurance of the nuclear cores. For example, attack
submarines that do not need to be refueled during their roughly 33-year lifetimes
would need to be taken out of service two or three times for refueling with LEU.
According to the 1995 Report, the refueling process might involve cutting through
a submarine's hull in order to replace the nuclear core. In any event, submarines
undergoing refueling would be out of service for extended periods of time, thereby
effectively reducing the number of submarines available for sea duty at any given time.
The report stated that each attack submarine (an SSN) would spend an additional
2.5 years (8% of its life) in shipyards, and each strategic deterrent submarine (an
SSBN) would spend an additional two years (4% of its life) in shipyards not available
for sea duty. In order to maintain fleet readiness, the Navy stated that it would need
to construct four additional SSNs and one additional SSBN. According to the 1995
Report, the annualized cost of these additional submarines would be an additional
$500 million.[11]
The Navy acknowledged that it could redesign its submarines and ships to accommodate a larger core based on LEU that would not need refueling during the vessels'
lifetimes. However, the report indicated that, because LEU has a lower concentration
of uranium-235, the nuclear core must be larger in volume in order to provide the
same endurance. All else being equal, a larger core volume will require that the space
provided for the reactor be increased. As a result, the Navy believed that it would
28
need to redesign its submarines to provide for this additional space demand. The
report did not explain how the calculations of increased core volume were made and
what assumptions were used to estimate what additional space would be needed in
the reactor compartments. [11]
With regard to designing new submarines, the 1995 Report maintained that the
same performance with a LEU core would require that the sizes and weights of the
reactor vessel, pressurizer, and related components be increased to accommodate the
larger core. This in turn would increase the size and weight of the reactor compartment and the shielding needed to protect the crew. As a result, the submarine's
volume would need to be increased to add buoyancy to compensate for the additional weight. According to the Navy, if attack submarines (SSNs) were redesigned
to accommodate lifetime LEU cores (20% enriched), the machinery weight would increase by 18% and the hull diameter would increase by three feet. The Navy believed
that such a heavier, larger submarine would be detrimental to tactical characteristics,
would have a longer stopping distance, and might be less maneuverable. The Navy
acknowledged that it could probably fit a lifetime LEU core into an SSBN's larger hull
without redesigning the hull, although it would need to address the greater weight
during the design phase.[11]
The 1995 Report further stated that the costs of converting the nuclear fleet
to LEU would be substantial. It estimated that, if current reactor designs are not
changed, maintenance costs would increase by about $1.8 billion per year (in 1995
dollars and assuming a 1995 baseline) and that additional submarine and ship construction would be necessary to provide the same at-sea force levels. Alternatively,
if submarines and ships are instead redesigned to accommodate larger LEU cores,
the report indicates that the one-time redesign costs would be about $4.7 billion,
and increased construction costs for new submarines and carriers based on natural
replacement rates would be about $1.1 billion per year more for LEU-fueled vessels
as compared to HEU-fueled vessels. The report additionally stated that conversion
to LEU would provide no technical advantage to the Navy and would actually be
detrimental from an environmental perspective. [11]
29
3.2
The Ippolito Study
Five years before the Navy issued the 1995 Report, Thomas Ippolito completed a
study examining the tradeoffs that exist if LEU were used as a nuclear submarine
reactor fuel rather than HEU. His study addressed such factors as core life, core
size, total power, and reactor safety. To evaluate these tradeoffs, Ippolito simulated
three 50 MWt reactor designs using uranium fuel enriched to 7%, 20%, and 97.3%
respectively.
He assumed that the 7% and 20% designs (LEU) were fueled with
uranium dioxide (UO 2 ) in a caramel configuration, a design developed by France and
currently used in all of its submarines. He also assumed that the design using the
97.3% fuel (HEU) was a dispersion type.[1]
Based on his data, Ippolito concluded that the 20% enriched core could have a
lifetime of 20 years, the same as the HEU core at that time, if the core volume were
increased to 1.7 to 2.5 times the volume of the HEU core. He also suggested that the
U.S. Navy's submarines could accommodate the increased core volume by switching
to the French integral core design.
Under this design, the reactor and the steam
generator are integrated into one unit rather than being separated as under the U.S.
Navy design. As a result, the space needed to house the reactor and steam generator
under the French design would presumably be significantly less as compared to the
U.S. Navy design.[1]
Although the Ippolito study predated the 1995 Report, the 1995 Report did not
refer to the Ippolito study nor did it expressly address any of the results and conclusions of that study. Because the 1995 Report was silent in this regard, the extent to
which the Navy actually considered the elements of the Ippolito study is unclear.[1]
The Ippolito study was one of the first technical attempts made at understanding
the constraints associated with a new LEU core in submarines and forecasting actual
new designs. The study resulted in an increase in attention to the issue and most likely
resulted in the 1995 Report from the Navy in response. 3 However, the Ippolito study
3
A private discussion with Professor Marvin H. Miller of MIT suggested that the 1995 Report
issued by Naval Reactors was a response to increased attention that arose after the Ippolito thesis
was published.
30
looked at a small part of the design space and fell short in several key areas: it modeled
naval reactor fuel as pins instead of plates, and it considered only lifetime cores as
the end goal. 4 This approach resulted in unattractively high core volumes for LEU
reactors.5 [1] These high core volumes ultimately led to a perception among the public
and even the naval reactor community that larger cores were the inevitable result of
using LEU. In fact, this is not necessarily true, and increases in fuel technology since
the Ippolito study, in combination with a relaxation of the lifetime criteria, might
actually yield far more attractive results. 6
3.3
The Ma and von Hippel Study
In 2001, Ma and von Hippel published a study which addressed the 1995 Report
and performed additional calculations based on alternative designs. Their alternative
designs were also based on the caramel fuel designs used by France in its Rubis-class
submarines, which utilize LEU, and on data involving LEU-fueled Russian icebreaker
ships. Starting with the assumption and results of Ippolito's study, Ma and von Hippel modified the core design modeled by Ippolito for the Rubis-class submarines to
achieve power of 130 MWt, approximately the power of a Los Angeles class attack
submarine.[5] They increased the size of the reactor core by 2.6 times to achieve the
130 MWt power. They then scaled up the core volume by an additional factor to
increase the core life from 20 years (the result of Ippolito's study) to 33 years (the
Navy's goal for the newer, larger Virginia class attack submarines). The resulting dimensions of the reactor core for LEU fuel were 50% larger than Ippolito's calculation
but, according to the authors, were much smaller than the reactor compartment of a
Los Angeles class submarine. (However, it is still speculative as to whether the hypothetical LEU core would fit into the reactor compartment without modifying the size
4Naval reactor fuel is plate fuel, not pins. The study seriously considered only the design results
that come from a lifetime core design, without analyzing other intermediate design options such as
periodic refueling.
5
Some of the redesign estimates resulted in as high as a 2.5 increase in core volume.
6
"Increases in fuel technology" refers to advancements made in uranium loading capability of
certain high density fuels.
31
of the compartment.) 7 Their conclusion was that the Navy could successfully pursue
new, innovative reactor designs using LEU that would meet its size and performance
requirements. In short, Ma and von Hippel concluded that the 1995 Report had been
unduly pessimistic in assessing the possibility of converting the nuclear fleet from
HEU to LEU.
3.4
The Glaser and von Hippel Study
In a 2002 paper, Glaser and von Hippel maintained that the conversion of naval
propulsion reactors to LEU has thus far attracted relatively little attention even
though the amount of HEU required for naval reactors is greater than the amount
currently required for research reactors.' The authors argue that conversion of the
world's nuclear navies to LEU is important for nuclear nonproliferation, preventing
terrorists from acquiring HEU, and verification under the proposed FMCT. The paper states that the 1995 Report did not adequately address the conversion issue and
that the 1995 Report simply emphasized that conversion to LEU would not give
the U.S. Navy any technical advantage.[21] In addition, the authors noted that any
independent review of the impact of propulsion reactor conversion to LEU on submarine performance is hampered by the fact that virtually all fuel and reactor design
information is classified.[21]
3.5
The Miller Speech
In 2003, Professor Marvin Miller presented a speech in which he stated that the
challenge of converting existing HEU-fueled naval reactors, particularly submarine
reactors, to the use of LEU fuel is more daunting than conversion of land-based
7
While the core itself might be small enough to fit inside the reactor compartment, it is important
to note that the core takes up a small fraction of the total volume in the compartment. The majority
of space is comprised of the pressure vessel, piping, heat exchangers, other components, shielding,
and finally empty space to allow the crew to move and work.
8
This assertion is in reference to the IAEA and international attempts to convert all research
reactors from HEU to LEU.
32
research reactors. He emphasized the lack of space on submarines and the difficulty
of increasing reactor core volumes to maintain the same power and fuel lifetimes. He
also maintained that the hostile and hazardous environments in which naval reactors
operate might rule out the use of higher density fuels that can be used to convert
research reactors to LEU.9
However, Miller also stated that it may be possible to design new nuclear-powered
ships from the ground up to use LEU. In this regard, he referred to the French Rubisclass submarines that use 7% enriched uranium and Ippolito's paper calculating that
LEU enriched at 20% could extend the core lifetime to 20 years with an increase of
core volume of about 2.5. He explained that the 1995 Report concluded that any
significant increase in core volume would be unacceptable to the U.S. Navy and that
the Navy's goal is to build the best nuclear submarines in the world. According to
Miller, the Navy's contention that attempting to increase the uranium density of the
fuel sufficiently to compensate for going to LEU without an increase in core volume
would simply compromise its performance is strongly held but impossible to verify
without access to classified documents.{27
Miller's suggestion is that the U.S. Navy provide leadership by example by investigating the potential for using new LEU fuels. The investigation might focus on
the type of fuel being developed under the Reduced Enrichment for Research and
Test Reactors (RERTR) program to convert the remaining HEU research reactors
and also the possibility of non-intrusive but credible monitoring of the naval reactor
fuel cycle. [27]
3.6
The Ward Study
In a 2011 study, Rebecca Ward examined the U.S. Navy's assertions made in the
1995 Report and focused on its primary argument against conversion to LEU in what
9
An example of a hostile or hazardous environment that a submarine naval reactor might have to
endure that a land-based research reactor would not have to endure is an "all ahead full" maneuver,
in which the submarine goes from forward movement to backward movement with full power. This
maneuver causes considerable vibrations that will potentially impact the reactor.
33
she terms the economic penalty. Ward first discussed the Navy's stated reasons for
resisting the conversion of its submarine reactors to LEU. She noted that the Navy
cited a number of factors as being essential requirements and indicated that failure to
meet any one of them would eliminate the use of a different reactor core and/or fuel
system. She stated that compactness (core size) and endurance (core life) are most
often cited as the major obstacles to conversion. However, one expert also maintained
that maneuverability could also be compromised by a conversion to LEU.[28
Ward also discussed France's experience with conversion to LEU and its relevance
to the issues raised in the 1995 Report. She explained that all French submarines
now contain reactor cores that use LEU designed as caramel fuel with the average
enrichment being about 7%. She stated that a new class of French submarines,
the Barracuda-class, is scheduled to begin operation in 2015. The Barracuda-class
submarine will be almost twice as large as the Rubis-class submarine and will have an
increased core life. The Barracuda-class submarines will be fueled by LEU caramel
fuel with some improvements.' 0 According to Ward, France's new submarines show
that improvements to core design have enabled increased power and/or longer core
life, again suggesting that efforts by the U.S. Navy to optimize reactor design for an
LEU core could yield more impressive results than presented in the 1995 report.[28]
With regard to economic considerations, Ward asserted that while fuel conversion
would no doubt prove costly for the Navy, it seems unlikely that cost concerns are
truly driving the Navys opposition to conversion. She explained that the 1995 Report set forth two main options for conversion: (1) maintaining current core volume
and thereby reducing core life, or (2) increasing core volume and maintaining lifetime
cores. If the Navy were to pursue the second option and completely redesign submarine reactor cores from the ground up, the Navy would face substantial upfront costs
of $5.5 billion. However, annual construction costs and operating expenses would
be lower because fleet availability would not be adversely affected.
Ward further
contended that conversion to LEU would likely reduce the 1995 Report's projected
annual operating costs because fuel fabrication costs for LEU should be lower, in part,
10
The fuel consists of flat squares of uranium dioxide ceramic embedded in a zirconium alloy grid.
34
because of decreased security requirements for LEU as opposed to HEU. " [28]
3.7
The Navy's 2014 Report on LEU Conversion
Recently, at the request of Congress, the Navy issued a report in January 2014 that
re-assessed the possibility of employing an LEU naval nuclear propulsion system in
its submarine fleet. The 2014 Report concludes that there are two main options for
developing a new LEU system: (1) substituting LEU fuel for HEU fuel in current
reactors; and (2) developing a new fuel system that can increase uranium loading in
order to compensate for some of the expected impacts of LEU fuel. More specifically,
the report maintains that converting existing submarine reactors to LEU is impractical and extremely costly. [12] Furthermore, the report states that using LEU in today's
submarines would require two or three refuelings, which would adversely impact the
strategic ability of the submarine force. The report maintains that, although it is
possible to design a larger LEU core with performance characteristics similar to those
for the existing HEU core, this would require a significant increase in volume, which
would have deleterious effects on an individual submarine's performance. [12]
At the same time, it is significant that the 2014 Report suggests that development
of a more advanced fuel system is feasible.
The report states that "recent work
has shown that the potential exists to develop an advanced fuel system that could
increase uranium loading beyond what is practical today while meeting the rigorous
performance requirements for naval reactors. Success is not assured, but an advanced
fuel system might enable either a higher energy naval core using HEU fuel, or allow
using LEU fuel with less impact on reactor lifetime, size, and ship costs." [12] Although
the report says that this advanced fuel system option is possible, it also states that
such a development would take years of focused research and that its researchers,
physical resources, and funding are currently dedicated to supporting the existing
fleet and developing technologies for other new reactor and submarine designs. The
"Security requirements for storing and transporting HEU require much larger and more extensive
security details that would not be necessary if the fuel was LEU.
35
report concludes that, without additional funding, the Navy can do no more than
conduct research on advanced fuel systems at the concept stage.[12]
36
Chapter 4
Fuel and Neutronics
Considerations
Naval reactors have certain specific design constraints that are more stringent than
terrestrial power reactors. Naval reactor core designs must meet several general requirements: they must be compact, produce an appropriate fuel burnup with a core
lifetime that meets submarine fleet requirements, perform safely under varying chemistry conditions while resisting damage from corrosion, and be able to withstand
possible accidents and stresses from war-time conditions. 1 [4]
This chapter will focus on the requirements of naval reactor design that pertain to
the selection of fuel, the neutronic performance of the core, the geometry of the core,
and the safety concerns that result from a naval reactor during refueling and disposal
of waste. The chapter will describe requirements of traditional naval reactors, namely
those requirements that have evolved over time to dictate the design constraints of
the current HEU-fueled, PWR reactors, as well as describe design challenges that
arise when considering a new LEU core.
'Burnup is defined as energy extracted from fuel per mass. A typical burnup for a commercial
PWR is 30 MWD/kgF. Submarine fleet requirements generally call for a long core burnup, but as
we will see later in the thesis, this does not necessarily translate into a requirement for a lifetime
core.
37
4.1
General Fuel Criteria
The first design area to examine is the naval reactor fuel itself. The fuel should adhere
to the following:
" Naval reactors are traditionally designed to produce a reactor coolant outlet
temperature of 320 degrees Celsius in order to produce steam that is required
for power generation.[3] This high outlet temperature mandates a high fuel
temperature, which in turn requires a high fuel melting temperature to provide
an appropriate safety margin in case of an accident that increases the fuel
temperature.
" The fuel should be volumetrically stable because volume changes can damage
the core, both in terms of physical destruction of the core and also by altering
the expected neutronic performance. 2
"
Naval reactor fuel needs to be more mechanically stable than terrestrial fuel.
Unlike fuel pins used in most land-based reactors, which provide limited additional structural integrity to lateral forces, naval reactor plate fuel must be able
to withstand battle conditions. These battle conditions can manifest themselves
in many different forms, but one of the most obvious is the need for the reactor
to withstand a nearby explosion by torpedo, mine, or depth charge. This explosion would result in a sudden shockwave that would put an immense amount of
strain on the reactor components and fuel. Therefore, the fuel and other components need to be designed in such a way that they can handle in excess of 50G's
of force without fracturing or displacing.[11] This requirement has most likely
resulted in naval reactor fuel designs having a greater emphasis on structural
performance and mechanical stability than terrestrial reactors. Good characteristics of a strong and durable reactor include matrix or monobloc fuel with
as few separate parts as possible, and a cladding that maintains adequate ten-
2
The fuel could expand, causing it to crack or break and result in physical damage to the fuel
matrix. This expansion could also alter the neutronic performance of the core by creating a local
area with a different fuel temperature or density, or by creating an area to hold fission gases.
38
sile strength even after high neutron exposure and at high temperatures, while
being resistant to chemical corrosion. Pins do not provide adequate support,
and standard plate fuel can bow under increased pressure or radiation damage,
which can result in localized occluded regions in the core where there would
be inadequate coolant flow. This situation could result in local hotspots that
would damage the core.
" The fuel cladding should interact chemically with the coolant in an acceptable
way. As a general philosophy, naval reactor designs generally favor slow and predictable corrosion of components over potentially more rapid and catastrophic
corrosion accidents. This means that the plant will probably accept surface
oxidation as a tradeoff for a lower probability of a more aggressive stress corrosion cracking. Unlike a larger terrestrial plant environment with more limits on
coolant chemistry, naval reactor operators can more readily control the chemistry of the coolant.3
" The fuel needs to tolerate internal pressure arising from fission gas production
over the long life of service. This is especially true because naval reactors are
expected to undergo a significantly higher burnup than a terrestrial reactor and
to produce a large amount of fission gases. This can be compensated for either
by having void space built into the fuel to trap the fission gases in the fuel, or
by allowing them to escape the fuel but have some type of larger plenum to
trap the gas and volatile fission products as they accumulate. 4
"
The fuel should allow high uranium loading (high content of uranium). Uranium
loading is the density of uranium metal in the fuel, usually given in grams
per cubic centimeter. High uranium loading allows naval reactors to produce
3
They can opt to create a slightly more oxidizing coolant chemistry in the initial phases of
operation, resulting in the creation of a small oxidation layer over the fuel cladding. Later in
operation, after a small layer of oxide has formed, the chemistry can be returned to a state that
produces much lower oxidation. This combination, initial rapid oxidation followed by slow oxidation,
results in an oxidation scheme that is easily monitored and is robust over longer periods of time.
4
A large plenum would be installed on top of the core to catch fission gas as it is produced and
contain it in a controlled manner.
39
the necessary energy density that the compact size requirement of submarines
dictates.
These above requirements will inform the fuel selection process. With them outlined, we can move on to look at specific LEU fuel options.
4.2
LEU Fuel Design
In designing a new naval reactor core, the goal is to choose an appropriate LEU fuel
type with desirable characteristics that satisfy the design constraints above.
The
focus of this section will be to enumerate a design strategy for selecting a fuel and
testing its performance.
Before moving too far along in the design process, it is important to perform a
"proof of concept" to examine if LEU naval reactor cores can indeed meet endurance
requirements without an unacceptable increase in core volume. Core volume increases
have historically been a basis for rejecting LEU options for the submarine force.
Therefore, one of the main focuses of this section will be on new LEU fuels with
high uranium loading, with the goal of showing that new fuel technology and the
possibility of high uranium loading have combined to make LEU cores more attainable
and attractive in size.
Uranium loading refers to the density of uranium in a fuel. Most fuels consist of a
uranium compound in some matrix element, and therefore uranium only constitutes
a portion of the actual fuel. This uranium is then in turn divided into its constituents
of uranium-235, uranium-238, and other minor uranium isotopes according to its
enrichment level. Different fuel types have different densities of uranium, and this
density is a key factor when considering lower enrichments in LEU fuels. An LEU
fuel can compensate for any losses from lower enrichments of uranium-235 with higher
uranium loading to reach comparable levels of uranium-235 that would be present in
an HEU fuel.
One high uranium loading fuel type of interest is Uranium-Molybdenum fuel, or
40
UMo fuel. 5
Researchers studied UMo fuels extensively in the 1960s for potential
use in fast reactors and have closely studied them again more recently. Efforts to
convert research and test reactors from HEU to LEU have sparked an increase in
research and funding into high uranium loading fuels.
Many of these new types
of fuels are possible candidates for a new naval reactor LEU fuel.
One particular
variant of UMo fuel is a monolithic foil that has a uranium density of 17.4 grams
of uranium per cubic centimeter (gU/cc), which is much higher than most fuels,
and therefore higher than what one could consider the uranium loading value is
for a current naval reactor.[2] This UMo fuel has achieved high burnup levels of
65 megawatt-days per kilogram of fuel (MWD/kgF), which is another plus for a
naval reactor fuel.[2] Although small-scale plates of monolithic UMo fuel have been
produced and successfully tested, monolithic UMo plate fuel currently cannot be
manufactured on large enough scales for use in a naval reactor design.6
In addition to UMo monolithic fuel, one can also consider UMo dispersion fuel,
which consists of UMo fuel dispersed in an aluminum matrix. This fuel is easier to
manufacture than UMo monolithic fuel but has been less commonly used in the United
States in reactor replacement programs because of a lack of a domestic interest in
this fuel as a research reactor replacement. [2] These dispersion fuels contain uranium
densities of less than 10 gU/cc, but still contain higher uranium loading values than
most current fuels. [2]
One of the major challenges for UMo dispersion fuel has been the interaction between the UMo and the matrix aluminum that houses it. Experience shows that under
certain irradiation scenarios, namely higher burnups, unpredictable fuel swelling has
been seen at the interface between the UMo and the aluminum.[29] Researchers have
also observed this fuel swelling in the UMo monolithic fuel to a lesser degree in the
interface between the solid UMo and the aluminum cladding. However, taken inde5Most notably, this new fuel type has been experimented with by the RERTR program (reduced
enrichment for research and test reactors) for use in converting HEU research reactors to LEU. For
example, U-10Mo is 90% uranium and 10% molybdenum.
'The fuel manufacturing center, which is located in Lynchburg, Virginia, is continually working
on perfecting the manufacturing process. Not much is known about this process other than personal
conversations with Professor T.H. Newton that suggest the fabrication process could be perfected
by 2016.
41
pendently, both components, the UMo and the aluminum, have performed in a stable
manner under various irradiation conditions, meaning that there is either some fundamental incompatibility between the UMo and aluminum fuel or there exists room
for further analysis to solve this problem.729][30]
Studies have yielded a large amount of experimental data on specific fuel characteristics. These characteristics are important when examining a new fuel for potential
use in naval reactors. The following is a list of characteristics that are experimentally
known for UMo fuels and would need to be determined (if not known already) for
other high density fuels to be analyzed: (1) heat capacity, (2) thermal expansion, (3)
melting temperature, (4) thermal conductivity, (5) tensile and compressive properties
(6) swelling, and (7) corrosion properties. [30] These are properties that are important
to know in order to assess how well a fuel type meets the criteria outlined at the beginning of this chapter. Analysis of these specific fuel characteristics lies outside the
scope of this thesis, but they are presented to help describe the process for selecting
a new high uranium loading, LEU fuel.
Figure 4-1 shows other high uranium loading fuels and lists uranium density and
the weight fraction of uranium for each fuel type. This topic of fuel performance
and selection has been studied more extensively in the Ippolito thesis, "Effects of
Variation of Uranium Enrichment on Nuclear Submarine Reactor Design." [1] Ideally,
many high uranium loading fuel options from Figure 4-1 should be examined and
tested in a similar manner to arrive at the best possible option. Once a viable fuel
type has been selected and results have shown that it meets the requirements for fuel
performance, one can determine the amount of fuel that will be required to meet the
current power requirements of a submarine.
7This means that when irradiated by itself, UMo has remained
stable and exhibited favorable
performance characteristics. Similarly, aluminum has performed well under irradiation conditions
and is a popular choice for many matrix elements.
42
FUEL CHARACTERISTICS
W#-RIK4
Makix
Matei"l
Rtw%
Ahy
Density
(eMI)
2-7
3-94
UA:
Al
Al
Al
Al
5.7
6.8
8.14
6-42
0.653
0-748
0.815
0-717
3.72
5.07
6.a3
4.0
10.96
8.30
11.19
9.50
0.882
0-848
0.809
0.A8
9.07
7.04
9.72
Russia Oxide
Al
Al
At
Al
Usis
Al
Al
Al
1096
12.20
15.30
0.895
0-927
090AM
9.81
11.31
14.60
U-1OMo
At
Al
Al
Al
Al
Al
Al
17.82
17-20
17.36
17.55
17.72
18.09
"6An
0.90
091
15.32
15.5
15.97
Al
Al
Al
At
17-74
17-84
17.59
17.01
16.5
16-48
Al
Al
At
Al
16.99
1.41
10.86
17.94
14.15
14.77
15.51
17.22
Al
Al
17.40
14-30
19.05
FuelAloy
Al
At^~
UAIs
Uh
UA16
U-Qlo
U-BUD
U-7Mo
U4-*
U-4Woa
Canspd t 1Mao
U-Mo-lPt
U-6Mo-0.6Ru
U--QMo-0.lSi
U-Sopb-asir
U-44
U-5Nb-3Zr
U-2Mo-lNb-lZr
UfFe
UN
U
Mayr
WO"
-MO-
0.92
0.93
0-94
0.96
(9k"6
8.26
10.32
16.0
17.37
13.81
16.52
15.30
0.962
0.044
1.00
16.74
13.50
19.O5
Rho,# = Density of Dispersed Phase
Fadrian of Uriniuvn Dispersed Phase
WO"RIh= Density of UnI
in Dispersed Phase
(a)As=mned toxmnsist arof wt% UA16 and 31 wt-% UA14 after bfinkcuion.
(b) Russimo aide pawder is Imudy refend to as U02. but is acualy U#O.AAhal density oftias
oxide paideris9- 109tOm. We us a densty of 9.5 gkun.
W# = Woi
Figure 4-1: High Density Fuels.[2]
4.3
Fuel Weight and Volume Estimates
As a first level of analysis, we can estimate the fuel weight and volume requirements
from current HEU cores in submarines and use this result as a baseline for comparison
43
with new LEU designs.
4.3.1
Current HEU-Core Estimates
The first step in estimating core parameters for a current HEU naval reactor is to
analyze the typical performance of a submarine over its lifetime and understand the
total energy demand during this time. To do this, we calculate the number of fullpower days (FPD) that a typical submarine will require. The number of FPDs is the
number of 24-hour periods a reactor is scheduled for operation at full power output.
number of FPD=
[
time deployed
year
]o[average power profile] [core lifetime]
(4.1)
The Los Angeles Class fast attack submarine has a 130 MWt reactor, and we assume that the typical power profile of the reactor is an average of 25% of maximum
power. 8 [5] In addition, a typical submarine is active for an average of 6 months out
of a year.' We can then relate the number of FPDs to the number of megawatt days
(MWD), also known as the total energy our reactor produces over its lifetime.
number of MWD = [number of FPD][core power]
(4.2)
This calculation yields 195,000 MWD of energy required over the lifetime of the
submarine. This is the total energy demand that the core has to provide. The next
step in the analysis is to incorporate an estimate for the total amount of energy
that can be extracted from the fuel on a per mass basis, known again, as burnup.
An estimate of a current naval reactor burnup is 50 MWD/kgF. 10 [4] This is a high
estimate.
It is advantageous for the burnup to be as high as possible in a size-
constrained core because it means less fuel is ultimately necessary. Picking a high
8This 25% average is a combination of rapid transit at high power and a majority of time spent
at low or no power while submerged performing mission essential activities like ISR, Special Forces
operations, or other stealth missions.
9
Submarine deployments have increased in duration to about 7 months out of the year, but when
we account for periodic maintenance on the boat for periods of up to 6-24 months approximately
every 7 years, this results in 6 months per year as a good estimate.
10Typical
PWRs run for burnups of 30-50 MWD/kgF.
44
estimate for this calculation is important for this proof of concept. If we can show
that we can produce an LEU core fuel volume that is comparable to an HEU fuel core
volume assuming a very high current burnup in the HEU core, it shows that when
a more accurate, and most likely lower, burnup value is used our estimates will still
hold. We can divide the total energy demand (number of MWD) by the burnup to
get the weight of fuel that is required.
kg fuel required = number of MWD
burnup
(4.3)
Using the total energy demand and burnup estimate results in an estimate of 3,900
kg of fuel in the core. We can then determine the weight of uranium required by
multiplying the weight of fuel required by the weight percent of uranium in the fuel.
From this, we can estimate the weight of uranium-235 by multiplying the weight of
uranium by an enrichment value.
kg uranium required= [kg fuel required][weight percent uranium]
(4.4)
kg uranium 235 required = [kg uranium required][enrichment percent]
(4.5)
Assuming Russian icebreaker fuel is similar in characteristics to U.S. naval reactor
fuel, which is a claim made by Ma and von Hippel, the fuel contains about 48.2wt%
uranium and 51.8wt% zirconium.[5] This yields approximately 1,900 kg of uranium
and approximately 1,850 kg of uranium-235."
Once the weight of fuel is known, the uranium density of the fuel can be used
to determine the total volume of the fuel. We estimate the uranium density for a
naval reactor as 4.5 gU/cc based on an analysis referenced in a report by Ma and von
Hippel that lists this density for a uranium zirconium fuel. [5]
fuel volume =
kg fuel required
uranium uload
uranium loading
"This is assuming an enrichment estimate of 97.3%.
45
(4.6)
This gives a fuel volume of .420 m. This number corresponds to the total volume of
fuel meat (consisting of both uranium and zirconium) that will be contained in the
fuel. The displaced volume in the core will include this fuel volume as well as the
volume of cladding and spacers that make up the total fuel assembly. The process for
determining these values is described later in the thesis, but since these additional
displacements are assumed to be small compared to the fuel meat, it can be assumed
that there is some relatively fixed ratio between fuel volume and total core volume,
assuming an unchanged fuel-to-moderator ratio.' 2
4.3.2
New LEU-Core Estimates
The process for estimating the fuel weight and size of a new LEU core is similar
to the above process for the current HEU naval reactor core. The following process
can be used for a range of prospective LEU fuels.
Different fuels have different
burnup limits, uranium densities, and other fuel performance characteristics that will
make some more attractive options than others. For this analysis, we use UraniumlOMolybdenum (U-1OMo) fuel enriched to 20%.
In this notation, the number 10
refers to the weight percent of molybdenum in the fuel. This fuel is selected because
it has been studied more extensively than others due to its potential use in research
reactor replacements.
The first step is to assume the same performance requirements of 1,500 FPD and
195,000 MWD of energy over the lifetime." The next step is to select a burnup value
for the new fuel. The maximum burnup limit of U-1OMo is 65 MWD/kgF, but a
more conservative estimate of 50 MWD/kgF is used.[29] We select 50 MWD/kgF to
include a safety margin and to account for the fact that material properties tend to
suffer at higher burnups, as well as to provide a margin on available reactivity toward
the end of the core lifetime.' 4 The achievable burnup of the fuel will be one of the
12As will be discussed in Chapter 5, LEU cores might be optimized at different fuel-to-moderator
ratios.
13 It is important to note that these performance requirements might actually be changing as
referenced in Chapter 2.
14Naval reactors have specific constraints on available reactivity throughout their lifetime, and
this is most jeopardized at the end of the core lifetime. With a very high burnup, the available
46
main parameters to experiment with for future work."'
Low burnups will result in
larger core volumes, and high burnups will result in the boundaries of safety limits
and reactivity limits being tested.
Once the burnup limit of the fuel is known, in this case 50 MWD/kgF, we can
perform a similar process to determine the weight of fuel. This results in the same
estimate of 3,900 kg of fuel. The U-10Mo fuel is 90wt% uranium and 10wt% molybdenum. Using the U-10Mo value, this yields a weight of 3,500 kg of uranium and a
weight of about 700 kg of uranium-235.1 6
This uranium loading for U-10Mo fuel is 15.32 gU/cc. Different fuels also have
different uranium densities so this value is expected to change based on the fuel. This
value results in a total volume of fuel of .230 m3
,
which is 54% of the HEU core
estimate volume.
The other high density fuel options shown in Figure 4-1 all have different characteristics of burnup limit and uranium loading. Improvements in fuels, as a result of
a Naval Reactors research project, might result in significant improvements for core
fuel weight and volume.
4.3.3
Comparing HEU and LEU Core Estimates
This estimate shows that, using 20% enriched U-10Mo fuel, it is possible to achieve
the same uranium loading in an LEU core with a fuel volume of .230 m3 as compared
to an HEU core with a fuel volume of .420 m3 . In this estimate, the LEU core volume
is primarily a function of burnup, weight percent uranium in the fuel, and uranium
loading of the fuel. These are parameters that will vary significantly based on the
LEU fuel type that is being examined. The ability to estimate a smaller LEU core
is important and shows the value of high uranium loading fuels in solving the size
issue that has been a problem with LEU redesigns in the past. Figure 4-2 shows the
results of the same analysis with some other LEU fuel types.
reactivity could fall below a requisite threshold.
1 5 The ultimate burnup value will have to be selected as the result of a modeling and simulation
using a full core burnup modeling program.
16
This is assuming an LEU enrichment of 20%.
47
FUEL WEIGHT AND VOLUME
uranium
wt%
loading (gU/cc) uranium
Fuel
97.3% UZr
4.5
48.20%
20% U-10Mo
15.32
90%
20% U-7Mo
16.32
93%
20% USi
9.81
89.50%
fuel
(kg)
3,900
3,900
3,900
3,900
uranium
(kg)
1,900
3,500
3,600
3,500
uranium 235
(kg)
1,850
700
720
700
volume
(mA3)
0.42
0.23
0.22
0.36
Figure 4-2: Fuel Weight and Volume Estimates.
4.4
Reactivity and Poisoning
Unlike terrestrial reactors, naval reactors need to shutdown and start back up in
very short periods of time, which imposes a unique requirement on reactivity. The
reactor must maintain enough reactivity in the core so that the submarine can rapidly
change power levels and the operator can restart the reactor in case of an emergency
or accidental SCRAM without waiting.' 7 This requirement results from the need to
be mobile at all times, especially to escape a dangerous situation in combat. The
ability to restart the reactor is inhibited by the buildup of fission products, which act
as neutron poisons, absorbing neutrons that were previously being used to sustain
fission reactions, and reducing the overall reactivity of the core. 18 [31]
One of the main neutron-absorbing fission products resulting from steady-state
operation is xenon-135. Xenon-135 continues to accumulate after reactor shutdown
because it will form from the decay of its parent isotope, iodine-135. In significant
enough quantities after shutdown, this negative reactivity contribution can inhibit
the ability to restart the reactor. This phenomenon, known as reactor dead time, is
acceptable in terrestrial reactors but must be avoided in naval applications.[3]
Currently, naval reactors are designed to operate at normal power producing levels
with .the control rods partially inserted. These control rods absorb neutrons and
provide negative reactivity. If the reactor shuts down and needs to be restarted, the
7 A SCRAM refers
to the emergency shutdown of a nuclear reactor.
' 8 Reactivity is defined as the departure from criticality. The k-effective is the effective neutron
multiplication factor: the average number of neutrons from one fission that cause another fission.
Reactors operating at steady-state are critical and have k = 1, reactors that are increasing power
levels are supercritical and have k > 1, and reactors that are decreasing power levels are subcritical
and have k < 1.
48
control rods can be removed beyond their normal position to overcome xenon-135
poisoning in the core. This technique is known as burning through the xenon-135
and ensures that the submarine can maintain power and mobility at all times.
The specific behavior of xenon-135 concentrations after shutdown can be obtained
from decay equations, and the contribution of xenon-135 to negative reactivity can
also be determined. The following equations help us to do that, where p(t) is the
reactivity contribution from xenon-135 as a function of time.
1 L-,pX(t)
[3]
'
P(t) = ----
vep
p(t)
=
25
7'AX (e-Axt - e-AIt)][3]
7'-I' e- xt + A," '[ AX'^x+ +Oa,X
Ve-p
T
-x=
(4.7)
(4.8)
= neutron f lux
fission yield of xenon
1= fission yield of iodine
Ef = fission cross section of iodine
Ax
=
decay constant of xenon
A1
=
decay constant of iodine
oa,x= absorption cross section of xenon
=a,=
absorption cross section of poison
e = fast fission f actor
p = resonance escape probability
v = average neutron yield per fission
This analysis is from "Nuclear Naval Propulsion" by M. Ragheb.[3] Plugging in
known values for the above parameters, we can solve for the xenon-135 maximum
49
reactivity contribution.
PXe,max = -0.525 at t = 10 hours
(4.9)
Figure 4-2 shows the negative reactivity contribution from xenon-135 for a given
time period after reactor shutdown.
0
Reactor Dead Time
0.2
E
-0.6
0
-0.7
10
20
30
40
50
60
70
80
....
Time After Shutdown [hours]
Figure 4-3: Reactivity Contribution from Xenon. [3]
The negative reactivity contribution from xenon-135 reaches a minimum value of
about -0.525 at about 10 hours after reactor shutdown.[3] This is the reactivity value
that a naval reactor will have to be capable of overcoming at the end of the core
lifetime.
We can use the following equations to develop scaling factors for k-effective at
the beginning and end-of-life, reactivity at beginning and end-of-life, and how they
relate to core burnup. The change in reactivity as a function of burnup can be closely
50
approximated as a linear relationship.[31]
p = keff
1
-
keff
(4.10)
pf = po - aB[31]
(4.11)
final reactivity
p=
po = initial reactivity
a = constant relating burnup to reactivity
B = burnup
By plugging in our reactivity contribution from xenon-135 value, we can solve for the
k-effective needed to overcome this.
.525
kef
kef
=
-
1
;> 2.105
(4.12)
(4.13)
If the core is required to overcome peak xenon-135 poisoning, this means that a keffective of 2.105 is necessary at a minimum. (Naval reactors would most likely have
some safety margin on this value.) We can also derive equations for the k-effective
that is required at the beginning-of-life as a function of burnup to see that as burnup
increases, the required k-effective increases quickly.
.525 = po - aB
po = .525+ aB = kef
kef
keff required =
1
(1 - .525) - aB
(4.14)
(4.15)
(4.16)
Figure 4-4 shows the k-effective required at the beginning-of-life as a function of
maximum burnup. Finally, we can develop an equation for the actual k-effective that
51
3.0
2.5
71
1.5
0.00
002
0.04
0.06
0.A
0.10
0.12
0.14
aB
Figure 4-4: keff Required vs. Burnup.
would be present in the core as a function of burnup if we impose a limit of at least
2.105 as burnup increases towards infinity.
- 1
p =po-aB=keff
keff
keff actual =
1
(1 - po) + aB
(4.17)
(4.18)
Figure 4-5 shows the actual k-effective of the reactor as a function of burnup, assuming
an initial k-effective of close to 2.5. Figure 4-5 suggests that a naval reactor will
require a large beginning-of-life k-effective, as high as the maximum theoretical keffective value of approximately 2.5. This maximum value is a function of the numbers
of neutrons generated per fission, commonly known as v. For uranium-235, v =
2.418.[31]
52
2.5-
2.0
C.)
0.00
0.02
0.0
0.06
0.0
0.10
0.12
0.14
aB
Figure 4-5: keff Actual vs. Burnup.
4.5
Reactivity Behavior Differences in LEU Cores
It is important to note that LEU cores could experience different burnup reactivity correlations when compared to HEU cores. Lower fuel enrichment means less
uranium-235, which means more uranium-238. A higher concentration of uranium238 will result in more actinide formation from neutron capture, lowering the reactivity level, and resulting in a harder spectrum.[21 However, because both an HEU
core and LEU core are expected to produce the same total quantity of fission products per unit burnup, this means that fission products will absorb proportionally
fewer neutrons in an LEU core, which ultimately means that fission products will
contribute proportionally less to the negative reactivity contribution. This reduction
in the negative reactivity contribution of fission products might mean that an LEU
core is ultimately easier to manage and control because the variations in reactivity
that result from rapidly decaying fission products are fewer.
The odd-numbered plutonium isotopes produced by uranium-238 neutron capture
53
are fissile and give a higher neutron yield per absorption than uranium-235, 2.10
versus 2.06 for thermal neutrons.[31] However, because plutonium fission requires two
neutrons, one to produce the plutonium-239 from absorption by uranium-238 and one
to fission the plutonium-239, this still results in a decrease in neutron economy.
Naval reactors tend to operate at high temperatures (compared to research reactors), exacerbating the negative reactivity of uranium-238. At high temperatures,
Doppler broadening increases the absorption cross-sections of the uranium-238.19
An intriguing aspect of switching to LEU fuel is the possibility of using uranium238 concentration as a safety feedback mechanism by exploiting the resonance Doppler
broadening characteristics of uranium-238. If the core experiences a sudden reactivity
insertion, by either a battle transient or a normal steady-state operation accident, the
reactor power will initially increase. This power increase will result in an immediate
fuel temperature increase on the timescale of 10-1
seconds.
The average neutron
lifetime is on the timescale of 10-' seconds. Therefore, the increase in fuel temperature happens quickly enough that the fuel will experience an additional generation of
neutrons. With this temperature increase in the fuel, there will also be a corresponding broadening of the uranium-238 resonances, so more of these second generation
neutrons will be absorbed during the thermalization process. As a result, the power
increase will be arrested.
This is not exclusively an LEU-specific mechanism, but
because of the concentration of uranium-238 in LEU, the mechanism is more pronounced. This stabilizing effect of uranium-238 can be easily overcome (in instances
when the operator wants a sudden power increase for rapid propulsion changes) as
long as there is requisite excess reactivity in the core. If the design exists in such a
way that there is excess reactivity present at full power operation, it should be relatively easy to jump from normal submarine operating power, 20-25%, to full power
by simply overpowering the resonance broadening phenomena.2 0
The end-of-life k-effective value will have to be the necessary 2.105 to overcome
1 9 Doppler broadening refers to the broadening of cross-section resonances due to motion of nuclei.
This 25% average is a combination of rapid transit at high power and a majority of time spent
at low or no power while submerged performing mission essential activities like ISR, Special Forces
operations, or other stealth missions.
20
54
xenon-135 poisoning. All of the above reactivity information and the factors described
above must be taken into consideration and will make it more difficult when computing the beginning-of-life k-value for an LEU naval reactor core. This value will be
selected so that, at the end of its lifetime, the core has enough reactivity to overcome
any negative reactivity insertions that might arise. An LEU core will have a different
reactivity requirement than an HEU core and will require in-depth analysis, beyond
the scope of this thesis, to pick an appropriate exact k-effective value. However, as
we can see in Figure 4-6, as burnup increases, the difference in reactivity loss between
higher enriched uranium and lower enriched uranium fuels becomes more pronounced.
This means that at the high burnups that naval reactors require, they could be expected to lose a significant amount of the reactivity that they need for quick power
fluctuations and restarts.
Using a modeling program called CASMO to produce core burnup simulations, we
can construct a graph of k-effective in a core as a function of enrichment and burnup
by using generic core parameters from a typical PWR. This graph is produced using
uranium dioxide fuel that is typical for a commercial PWR. This fuel type is different
from naval reactor fuel and will therefore have a lower k-effective, but the important
information is the relationship between enrichment and k-effective and reactivity at
high burnup. Figure 4-6 shows several enrichment levels and how k-effective changes
as a function of burnup. Figure 4-7 shows several enrichments and how reactivity
changes as a function of burnup. The specific parameters used in the simulation are
shown in Figure 4-8. Figure 4-9 shows the specific values for the analysis.
The high enrichment case of 97.3% decreases from k-effective = 1.805 at zero
burnup to k-effective = 1.718 at a burnup of 40 MWD/kgF, for Ak = .087. The
lower enrichment of 20% decreases from k-effective = 1.524 at zero burnup to keffective = 1.212 at a burnup of 40 MWD/kgF, for Ak = .312, a much more rapid
loss of k-effective. The high enrichment case of 97.3% decreases from a reactivity of
p = .446 at zero burnup to a reactivity of p = .418 at a burnup of 40 MWD/kgF, for
Ap = .028. The lower enrichment of 20% decreases from a reactivity of p = .344 at
zero burnup to p = .175, for Ap = .169, a much more rapid loss of reactivity. While
55
I
1.9
1.6
1.57
223
q)1.3
1
0.7
1.21%I
2.0%
0.7
1
3
S
7
9
11
13
25
17
19
21
23
2S
27
29
31
33
3S
37 39
Burnup [MWD/kg]
vs. Burnup for 2%, 5%, 20%, and 97.3% Enrichment.
Figure 4-6: ke
O.5
OA
97.3%
0.3
0.2-
20.0%
S0.1
M
C
5.0%;
-
-0.1--0.2
20
3
S
7
9
11 10
B
17
1921
23
25
27
29
31 33 3S 37 39
Burnup [MWD/kg]
Figure 4-7: Reactivity vs. Burnup for 2%, 5%, 20%, and 97.3% Enrichment.
56
CASMO BURNUP PARAMETERS
PWR infinite Lattice K-eff Calculation
Pressure: 15.51 MPa
Power Density in the Fuel: 104.5 kW/L
Coolant / Moderator Temperature: 583K
Average Fuel Temperature: 873K
Coolant is Approximated as 88.81% Natural Oxygen
and 11.19% Hydrogen-1 by Weight
Fuel is U02 at 10.42 g/cc
Pin: 0.4096cm Fuel Radius, 0.41786cm Outer Gap
Radius, 0.47506cm Outer Clad Radius
Gap is Modeled as Air
Clad is Zircaloy
Pins are in a Standard 17x17 Layout
Pitch: 1.26cm
Figure 4-8: CASMO Burnup Parameters.
we cannot directly compare PWR fuel geometry to existing or potentially novel high
density naval reactor fuels, we can expect that the same trend of high enrichment
versus lower enrichment will hold true.
One of the main areas for future work is to determine if a new high density LEU
fuel will primarily be burnup-limited or reactivity-limited. Burnup-limited refers to
performance limits that arise from the actual physical damage that will result to the
fuel assembly at high burnup. This is a function of the fuel design and is especially a
concern for novel high density fuels. Reactivity-limited refers to performance limits
that will arise from the eventual loss in reactivity levels at the end of the core lifetime.
Ideally, advanced fuel design might develop to the point where only reactivity limits
burn up.
'
The next step in the design process will be to select a fuel-to-moderator ratio.2
Figure 4-10 shows an example of the pin-to-pitch ratio, or the ratio between fuel pin
diameter and coolant flow in a typical pin-fuel PWR. This ratio is also known as the
fuel-to-moderator ratio in non-pin-type reactors.
This ratio is an integral part of reactor design and will play a role in determining
21
The fuel-to-moderator ratio is defined as the ratio of the number of fuel atoms in the core to
the number of moderator (or water) atoms in the core. As temperature increases, fuel volume and
density remain relatively constant, and moderator volume remains constant, but moderator density
decreases. This mechanism relates the fuel to moderator ratio to the k-effective.
57
I:EIA4T1VITCA
K-INF
Burnup
[MWDhIrgFj
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Enrichment
20%
5%
1.38715 1.52438
L34542 1.51149
1.33575 1.50837
1.3282 150467
1.31688 1A9788
130563 1A9178
L29418 1A8617
1.28267 1A8088
1.2713 1A7581
1.26015 1A7091
1.2493 1A6613
179014
1.78821
178506
1.782
Bunup
jLM/kIgFn
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
1.77897
18
1.77597
177299
L77004
176716
176432
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
97.30%
180532
1.80453
180401
180334
1.80199
180064
179931
179799
1.79666
179535
1.79403
1.23875
1A6143
179273
1.22853
1A5681
1A5223
1A4551
1.43454
1A2377
1.41315
1.79143
1.21862
L20437
L18203
1.16107
1.14134
1.1225 1A0269
1.10444 139244
1.08708
138241
L07032 1.37254
L0541
136281
103837 L3532
1.02309 1.34373
100825 1.33437
0.99384
0.97985
0.96636
0.95329
0.9406
0.92829
0.9164
0.90493
0.89389
0.88326
0.87306
0.8633
0.85396
0.84505
L76149
1.75868
1.75589
1.75311
1.32511
1.31594
1.75034
1.30687
L74759
1.29789
128902
1.26024
1.27152
1.26286
1.25426
1.74485
12457
1.23719
122871
1.22027
121187
L74212
1.73941
L73673
173407
173141
1.72877
172612
172348
172084
171821
Enrichment
20% 97.30%
5%
0.279097 0.343996 OA46082
0.256738 0.338401 OA45839
0.251357 0.337033 OA45679
0.247101 0.335402 0A45473
0.240629 0.33239 OA45058
0.234086 0.32966 OA44642
0.22731 0.327129 OA44231
0.220376 0.324726 OA43823
0.213404 0.322406 0443412
0.206444 0.320149 OA43006
0.199552 0.317932 0.442596
0.192735 0.315739 OA42192
0.186019 0.313569 0.441787
0.1794 0.311404 OA41384
0.16969 0.308203 0.440782
0.153998 0.302912 0A39795
0.138725 0.297639 .438833
0.123837 0.292361 OA37877
0.109131 0.287084 0.436927
0.094564 0.281836 0A35981
0.060105 0.276626 0435041
0.0657 0.271424 0A3412
0.051323 0.266222 .433209
0.036952 0.261011 0A32299
0.022569 0.255803 0A31392
0.006182 0.250583 0A30488
-0.006198
-0.020564
-0.034811
-0.048999
-0.063151
-0.07725
0.245346 0A29585
0.240087 0A28682
0.234813 0427783
0.229519 0A26885
0.224217 0A25987
0.218896 0A25092
0.21354 0424205
-0.105058 0.208147 0A23322
-0.118706 0.202717 0A22436
-0.132169 0.197239 0421554
-0.145397 0.191717 0A20666
-0.158346 0.186138 0419779
-0.171015 0.180509 0A18888
-0.183362 0.174829 0.417999
0.091227
Figure 4-9: CASMO Data.
the neutronic feedback mechanisms as well as setting thresholds for thermal hydraulic
performance. As we have shown earlier, naval reactors have special requirements for
k-effective. Once a designer has taken the steps to maximize burnup and reached an
58
Pitch
Coolant Flow
Pin
Coolant Flow
Figure 4-10: Pin-to-Pitch Ratio. [4]
appropriate k-effective for the core, he or she can use a simple modeling code like
CASMO or Pinspec to find the optimal fuel-to-moderator ratio. With a predetermined fuel type and weight, one can produce a graph like Figure 4-11, which shows
the k-effective as a function of fuel-to-moderator ratio.
1.5
-
1.4
1.3
1.2
0.9
0.8
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
Pin-to-Pitch Ratio
Figure 4-11: K-eff vs. Pin-to-Pitch Ratio.[4]
This process will yield an optimal fuel-to-moderator ratio that can be used to
59
draw inferences regarding the geometry and design of an individual plate element.
One of the final steps will be accounting for geometric and material buckling.
These terms capture the effects of neutron leakage from both the geometry of the core
and from the material properties of the core. Buckling will result in a larger necessary
core volume than would be true if neutron leakage was neglected completely.
4.6
Considerations on Safety, Fuel Disposal, and
the Environment
The final consideration for naval reactor fuel that we discuss involves possible safety
effects on shipyard workers and the public during fuel disposal once the core's lifetime
is up and it has been depleted. Historically, the Navy has been concerned with the
safety effects that a new LEU core might have on the radiation exposure of workers
on the submarine or in the shipyard during refueling. 22 [11} This section discusses
these concerns, their origins, and what effects they might have. The Navy has a long
history of responsible waste disposal practices for spent fuel from naval reactors and
has made significant improvements in waste disposal radiation exposure levels since
1966. [11] These exposure level reductions have resulted from better waste disposal
practices, advancements in temporary shielding, and reducing the need for handling
waste by moving to longer lifetime cores. As a result, radiation exposure to shipyard
workers has been steadily decreasing, a trend that the Navy seeks to continue.
The current waste disposal practice for submarines takes place after the boat's
lifetime has been reached and the vessel has been decommissioned. The submarine
is then typically moved to a shipyard that specializes in waste disposal.23 Workers
remove the spent fuel from the reactor pressure vessel in two pieces and place each
piece in a waste storage canister, which is then shipped to the Navy's Expended Core
22
The report states that the occupational radiation exposure to shipyard workers would increase
both as a result of an increase in fuel waste canisters being moved and also that LEU waste produces
more hazardous waste in terms of radiation behavior.
23
Currently, this shipyard is the Puget Sound Naval Shipyard and Intermediate Maintenance
Facility in Bremerton, Washington.
60
Facility (ECF). Historically, this facility would analyze spent naval reactor fuel to
determine how new designs could be better optimized. After analysis at the ECF,
the spent fuel canister will ultimately await storage in a geological depository, but it
is first shipped to a temporary spent fuel facility at the Idaho National Laboratory.
The remaining reactor compartment is still radioactive from years of contact with
the core and will be cut from the submarine, drained, and sealed. The rest of the
reactor compartment can be disposed of in a low-level waste disposal area because
the radiation is minimal and contained." [321
Currently, a fast attack submarine produces two containers of waste at the end
of its lifetime. The Navy has estimated that shipyard workers receive a total annual
average of 60 man-rem exposure when refueling or defueling submarines.[11] The
1995 Report on LEU conversion in naval reactors estimated that switching to LEU
fuel could result in as many as seven containers of waste per submarine acquired
at multiple refuelings over the lifetime of the submarine, and as much as a total
annual average of 249 man-rem exposure.2"
These estimates regarding increases in
radiation exposure are the result of an assumed increase frequency of refueling from
fission products that are in equilibrium concentrations as a function of core power
(not any increase in uranium loading or core volume).
By using a new fuel with
increased uranium loading, the Navy should be able to maintain the status quo of
two containers of waste per submarine.
One important consideration is the different radioactive behavior of LEU fuel as
compared to HEU fuel and how this might affect radiation exposure levels. Spent,
high-burnup LEU fuel will have higher concentrations of minor uranium elements
like uranium-232, uranium-234, and uranium-236. These elements can exhibit more
hazardous radiation behavior. Specifically, uranium-232 can decay into thalium-208,
which releases an extremely energetic gamma particle that can be dangerous to nearby
workers if not handled correctly. [331
More generally, the higher concentration of uranium-238 in LEU fuel will cause
24These areas are located at a Department of Energy site in Hanford, Washington. The Navy has
disposed
of over 100 naval reactor compartments at the Hanford site.
25
It is important to note that this number can be reduced by having more workers.
61
a higher production of transuranics (namely, the production of plutonium).
These
transuranics tend to decay by spontaneous fission and neutron emission to a higher
degree than the non-transuranic waste produced by HEU. This increase in neutron
emission means that the cask storage system used to shield and protect against radiation exposure will need to consist of thick metal shielding as it currently does (to
block gamma decay) as well as hydrogenous material to slow down neutrons.[33]
The Navy would need to resolve this concern about the different treatment of
LEU waste as compared to HEU waste. The solution could be a process to ensure
that the quality of the LEU is high by working to ensure the LEU fuel has as small
of a concentration of minor uranium elements as possible.
Another solution is a
"work-around" new design for an intermediate storage container that can handle the
different LEU waste radiation profile.
In addition to the above technical obstacles, there exists a strong political barrier
to new waste disposal options that might be required for a switch to LEU fuel. In
1995, the State of Idaho, the U.S. Navy, and the Department of Energy all reached
an agreement, which sets limitations and timelines on the interim storage of nuclear
waste at the Idaho facility.[34] Some key provisions of this agreement are: (1) the
Navy will send 575 shipments of spent nuclear fuel to Idaho over a 40-year period
(1995-2035); (2) a target date of 2035 will be set for removal of all spent fuel from
Idaho for permanent disposal; and (3) transuranic waste will be removed by 2016.[34]
Waste disposal of nuclear spent fuel presents complex technical and political problems. Ideally, a new LEU core design would produce no significant additional quantities of waste, and this waste would need no additional handling requirements for
safety purposes. However, these two goals seem difficult for the Navy to meet and
raise concerns about disposing of the new LEU spent fuel.
4.7
Summary
Submarines present unique design requirements and constraints on naval reactors.
These constraints make using LEU as opposed to HEU challenging, but not impossi-
62
ble. New fuel technology and increases in uranium loading of fuels make it possible
to reduce LEU core sizes significantly when compared to previous estimates.2" This
is an important proof of concept to help remove the preconception that LEU cores
will require significant increases in size. A systematic design process would include
selecting a fuel type and testing it to determine maximum allowable burnup. Next, it
is necessary to determine reactivity constraints for the fuel design and the maximum
achievable burnup. Finally, one must refine the fuel geometry and fuel-to-moderator
ratio that will meet both reactivity demands and thermal hydraulic constraints that
will be discussed in later chapters.
26The
estimate using high uranium loading LEU fuel results in a core volume decrease to .7 times
the size of a current naval reactor core. The Navy's 1995 Report estimated that a 20% LEU core
would require a hull increase in the submarine of 3 feet. The previous thesis by Ippolito estimated
an increase in core volume of 1.7 to 2.5 times the current core volume.
63
Chapter 5
Chemistry and Thermal
Hydraulics Considerations
This chapter defines constraints in naval reactors that arise from chemistry-related
concerns and requirements for thermal hydraulic performance. It builds on the ideas
of Chapter 4, further defining requirements for naval reactor fuel, cladding, and components selections that pertain to corrosion. This chapter also discusses how a new
LEU design might impact thermal hydraulic performance.
5.1
Chemistry and Corrosion Constraints
In addition to major stresses that might result from battle conditions, the reactor
needs to be able to withstand corrosion damage from normal operations. These are
of two main types: general corrosion and pitting corrosion.[35]
General corrosion is the result of the water coolant interacting with the cladding,
causing oxidation, and forming hydrides. [35] This process is dependent on the chemistry of the water (pH and other ions) as well as the temperature gradient of the
cladding and coolant. If the pH of the water is too high, a large oxidation layer can
form. These layers typically have much worse heat transfer properties than the original cladding. This situation results in operational problems: the fuel temperature
can spike; and the fuel and the cladding can ultimately separate during start-up or
64
shut-down because of the large temperature difference.[35] Both of these problems
result in significant core damage and are unacceptable for a naval reactor.
Pitting corrosion occurs when the cladding is subjected to certain ions in the
coolant water. This results in localized galvanic corrosion, which causes small holes
to form on the surface of the cladding, weakening the material over time. [35] This
problem can be prevented if a thin oxide layer is formed on the cladding by raising the
pH of the coolant in a predictable and controlled fashion for a period of time. [35] [36]
Naval reactors tend to undergo slower, more predictable corrosion and build-up of
crud.1 This crud can be removed by a "crud-burst," where the pump power is rapidly
changed and the flow speed of the coolant changes dramatically. This will accelerate
fragile crud deposits so they detach from the fuel, entering the coolant where they
can later be filtered out. [36] If the material is not filtered out, it will start to damage
components and alter neutronic performance of the core. This rapid change in pump
power can also occur unexpectedly if the submarine is forced to change speed rapidly.
Crud-bursts, therefore, can be dangerous for the submarine if they occur unexpectedly
during operation. For this reason, rapid build-up of crud in naval reactors is avoided
when possible.
5.2
Selection of Core Geometry and Modeling
The geometry depends on the design of the fuel elements, which in turn depends
on both temperature and neutronic considerations. After the fuel type and cladding
are selected, one can experiment with several design options and analyze the options
using appropriate nuclear modeling codes.
The procedure for designing an individual plate element is to determine the dimensions of the plate by setting safety limits and solving for the size. Two of the
main limits are listed below.
* Critical heat flux: The plate element must be thin enough to avoid a departure
'Crud in this context simply refers to the material that builds up and forms on the surface of
interest. It is not to be confused with Chalk River Unidentified Deposits (CRUD).
65
from nucleate boiling (DNB) in the coolant on the surface of the plate.2 [37]
Figure 5-1 shows a coolant channel where fuel plates on either side have reached
the critical heat flux. Past this region, the DNB region, a steam layer forms
on the cladding and heat transfer from the cladding to coolant is drastically
reduced. If the plate is too wide, the temperature profile will be such that the
surface of the plate that touches the coolant will be too hot, resulting in damage
to the cladding and fuel. [4]
Liquid
core
DNB
Bubble layer
)I
Figure 5-1: Critical Heat Flux. [4]
o Maximum fuel temperature: The plate element must be thin enough to avoid
2
Departure from nucleate boiling is defined as the point at which the heat transfer from a fuel
rod or plate rapidly decreases due to the insulating effect of a steam blanket that forms on the rod
surface when the temperature continues to increase.
66
dangerous temperature peaking in the center of the fuel. Figure 5-2 shows a
cross-section of a plate element of plate width 2a, cladding width 6c, and coolant
flow on both sides. If the plate element is too wide (a is too large), the fuel in
the center of the plate will reach temperatures that are too high, which could
damage the fuel or cause melting.[4]
Fuel
q
C ad
Tei
Tei
z
x
Figure 5-2: Maximum Fuel Temperature. [4]
Once we know these limits, we can factor in a safety margin to avoid potential accident scenarios. These calculations should yield the desired dimensions of a complete
plate element.
67
After we know the dimensions of the plate element, we can organize individual
plate elements in a cylindrical fashion for optimal buckling until the requisite total
fuel weight of the core has been reached by the individual plate elements. The fuelto-moderator ratio, which was previously solved for, will determine the spacing of
these elements in the core. An example of a plate fuel core design for the MITR is
shown in Figure 5-3.
Inner
Reflector
Element
A-ang sample
area
Figure 5-3: MITR Plate Fuel Design. [2]
With a candidate core geometry, we can use a full core burnup simulation code
such as SERPENT to analyze the core over a lifetime and assess the viability of the
design. Many iterations of this process will be necessary to achieve a core design that
meets the necessary power profile, end-of-life k-effective, and produces meaningful
burnup.
68
5.3
Thermal Hydraulic Constraints
Naval reactors have both general and Navy-specific requirements that govern thermal
hydraulic performance. The general requirements do not differ greatly from terrestrial
reactors. Naval reactors usually produce propulsive power by passing coolant through
the core, converting the coolant to steam to turn turbines, and converting the turbine
rotation into propeller rotation. [3] The achievable propulsive power is therefore a
function of steam production, which is in turn a function of core outlet temperature.
Unlike a terrestrial reactor, however, which is usually at either zero power or near
full power, and can therefore be optimized to perform at those levels, a naval reactor
is expected to operate at many different intermediate power levels, depending on the
steam demand for a given speed. As a result, naval reactors will tend to perform
at lower efficiencies, but this is tolerable because the Navy is not as concerned with
economic constraints.
Some specific requirements for the thermal hydraulic performance of naval reactors are the ability to withstand power transients from battle conditions and passively
remove heat from the core in an emergency scenario. Naval reactors also have the requirement to remove heat from the core during low-power operation with a constraint
on pump power levels due to the noise levels they can produce. [11] These requirements
will be discussed in more detail in Chapter 6. This section will focus primarily on
developing relationships between fundamental reactor thermal hydraulic properties
to help make inferences on how different LEU core designs could alter requirements
or possibly prove prohibitive.
Starting with the following equation, we can find the heat removal rate from the
core by the coolant.
Q=
cAT
Q = heat removal rate
m = mass flow rate of coolant
C = specific heat of water
69
(5.1)
AT = temperature change across core
For the sake of this analysis, we assume that this Q value is the same for both an HEU
and LEU core because both cores are intended to produce the same power levels. If
we want to meet the same steam requirements as well, then the AT coolant value
across the core should be the same. The specific heat of water can be approximated
as a constant over small temperature changes.[4] This means that 7h will remain the
same as well. With these assumptions, we can relate rh to pumping power.
pump power requiredoc 7
(5.2)
p
AP = pressure change in core
p = density of coolant
This equation relates the required pumping power for the plant to the overall pressure
change throughout the plant. We can further dissect the pressure change term into
its subcomponents.
AP = APinertia+ APacc + APravity + APfriction + APform
(5.3)
We now have a complete equation for the pressure change throughout the plant. We
can neglect the APinertia term if we assume we are looking at a steady state scenario
in the plant. We can also neglect the APace term, as it will be negligible in most
cases.[4] This leaves us with three terms contributing to the overall pressure change:
APgravity, APfriction, and APform.
APgravity oc pgL
(5.4)
L = length of core
APriction Oc f
L P(2)
p(-)
f = friction factor
70
(5.5)
D = flow diameter
v = f low velocity
APfom oc rp(
)
(5.6)
r = form loss coefficient
If we assume that a new LEU core design will be the same size as a current HEU
core, this means that the height of the core L is constant. The pressure will therefore
be a function of the fuel-to-moderator ratio (or the amount of fuel compared to the
amount of coolant). [41 The fuel-to-moderator ratio will determine both the change in
the coolant piping diameter D as well as the coolant velocity v.[4] How the fuel-tomoderator ratio changes when compared to a current HEU core will dictate increases
or decreases in the APfj
7 jd and APfom terms. This means that a new LEU could
have a larger or smaller pressure change and therefore a larger or smaller required
pump power.
If we assume that a new LEU core design will require a larger core, the height of the
core L will increase to maintain optimal buckling. This will cause the AP,.jtV term
to increase. However, an increase in core size with a constant rh will mean a decreased
coolant velocity v throughout the core (the total mass of coolant is distributed over
more channels, each with a lower velocity, resulting in the same overall 7h). This
decrease in coolant velocity will cause APfrii,, and APfarm to decrease, if all else is
held equal. Therefore, core-volume increases tend to result in an overall decrease in
required pump power. However, APfrictio
and APf m terms arising from different
fuel-moderator ratios could overcome this benefit.
5.3.1
Minimal Flow Rates and Noise Production
Submarines have a unique requirement to remain as quiet as possible, another difference from terrestrial reactors. The success of a submarine often depends on its
71
ability to remain undetected and to avoid classification by an adversary.3 There are
many sources of noise in a submarine and many methods of detection used by other
submarines, surface vessels, and stationary arrays. Submarines produce two main
categories of noise: broadband noise and narrowband noise. [13] Broadband noise generally results from flow over the hull or cavitation on the propulsor.[13] Narrowband
noise consists of specific tonals that result from components and rotating machinery,
each of which has a specific frequency signature. [13] Narrowband noise can be filtered
out from the background noise to both locate and classify a submarine.[13] As such,
these components are designed to be as quiet as possible by minimizing resonance
modes, and reducing the frequency of any residual noise.
Submarines are designed with several mechanisms to limit noise: (1) a surface
rubber coating that dampens noise; (2) advanced propellers that provide high-thrust
while minimizing cavitation; and (3) springs and rubber grommets placed at certain
locations in the submarine to isolate noises from the hull.4 [3]
One unique source of noise in a submarine is the reactor, and more specifically,
the noise signature of the coolant flow through the piping and the noise signature
of the primary coolant pumps during operation (especially turning the pumps on or
off).[38] In particular, pumps tend to produce higher frequency narrowband tonals,
the most revealing kind of noise for identification and classification. [38] A new LEU
naval reactor will almost certainly have a different noise profile and must be designed
in such a way that limits noise. [38] If a new design does require significantly redesigned
pumps, either due to a new layout or a new power requirement, these pumps will have
to be designed to remain as quiet as possible.
The magnitude of the steady-state noise that a pump produces is generally proportional to the pump power required to overcome the pressure drop it must pump
across.[38] If an LEU core required an increase in pumping power, this could mean
noisier pumps and more difficulty remaining quiet in tactical scenarios. Conversely,
3
Classification in this context refers to an adversary not only detecting a submarine, but recording
it, processing the signals, and developing a methodology for identifying it more easily in the future.
4
This rubber coating surrounds the hull of the submarine and absorbs vibrations. Cavitation is
caused by the collapse of gas bubbles formed underwater by mechanical sheer.
72
if an LEU redesign resulted in a lower required pump power, as seems more likely to
be the case for a larger LEU core, this would mean that pumps might be quieter and
the submarine's stealthiness could be improved.
5.3.2
"Walk Away" Ready Heat Decay Requirement
One of the central philosophies in nuclear reactor safety is designing the system to
be able to successfully dissipate decay heat. Heat production does not stop immediately after the reactor is shutdown because fission products that have been produced
during normal operation continue to decay and produce heat. This heat decays after
shutdown according to the following equation.
P(t) =6.48 x 10-P[(t
Po
=
-
- t~0 .2][3]
(5.7)
reactor power before shutdown
To = time since reactor shutdown (measured from startup time)
t = time since reactor startup
This means that the power of the reactor is a positive function of both the time that
the reactor was operating before shutdown as well as the operating power and the
time after shutdown. Naval reactors have developed around a philosophy of "walk
away" ready heat decay.5 This means that in the event of an accident and a need
to SCRAM (suddenly shut down) the reactor, heat is still dissipated through passive
circulation in the system. It is also desirable that the passive cooling potential be
significant enough to allow the submarine to be operated at very low power without
using noisy pumps.' This capability allows the submarine to maintain hotel power,
and move submerged at very low speeds, while conducting stealth operations without
noise associated with pumps.
The above analysis of pressure changes in the reactor suggests that an LEU core
'This refers to the ability to passively cool the core.
6 rom a private conversation with a Naval Reactor Engineer.
73
design with a significant increase in the pressure change could limit the ability of the
reactor to be passively cooled; natural circulation would be inhibited by having to
overcome a larger AP.
5.4
Summary
The performance environment of a naval reactor imposes requirements on the fuel,
cladding, and other components to ensure they do not suffer damage under potentially
corrosive conditions and that they behave in a predictable fashion.. The underlying
thermal hydraulic performance of a naval reactor is much like terrestrial reactors,
which leads to similar constraints on thermal hydraulic performance. Differences in
constraints arise when we consider new LEU designs that might alter the required
pump power of the plant, either due to an increase in core size or an alteration of the
fuel-to-moderator ratio. This need for an increase in pump power could have negative
effects on passive cooling ability and noise production. However, further analysis is
necessary to more accurately predict how an LEU core will alter the pressure change
in the plant.
Preliminary analysis shows that LEU redesigns might even result in
decreases in required pump power.
74
Chapter 6
Physical Constraints
This chapter focuses on physical constraints that arise as a result of performance and
strategic requirements for submarines. The first performance requirement considered
is maximum speed, which relates to the required power to move the boat through
water based on its size and weight. The next requirement is the strategic force level
requirement of the submarine fleet. Both of these requirements, speed and strategic
force level, are analyzed in light of potential LEU redesigns. Specifically, we assess in
more detail LEU redesigns that result in larger cores in order to make inferences on
feasibility.
6.1
The Speed Requirement
In Chapter 2, it was argued that the mission space of the submarine force requiring
high burnup cores might be declining in the future. One requirement that is unlikely
to change is the requirement for a submarine to transit quickly at maximum power in
the case of an emergency or a battle scenario. If one assumes that the current Virginia
class submarine speed must be met by future submarines, an LEU core could present
a design challenge. An empirical relationship between shaft power and speed is:
propulsive power required = 0.06977CdV 21 3v 3 [3
75
(6.1)
Cd = drag coef ficient
V = displacement [im 3 ]
v = speed [knots]
A conservatively high estimate for the drag coefficient value, Cd, for a Virginia class
submarine is 0.035.[1] An estimate for the total displacement of a Virginia class is
7,700 m .[5] The maximum submerged speed of a Virginia class submarine, v, is
estimated at 35 knots.[3] Using the above equation and estimates for drag coefficient,
displacement, and maximum speed, we get a requirement for propulsive power of
40.8 MWe, and assuming a plant efficiency of .33%, this yields a propulsive power
requirement of 122.4 MWt for the plant. 1 [4]
This equation can also be used to derive an important relationship between propulsive power required and reactor power available according to the scaling laws below.
hull diameter D oc r
(6.2)
displacement V oc r2
(6.3)
3
propulsive power P oc Cdr4 /
(6.4)
reactor power P, oc r3
(6.5)
r > Cdr4/3 and P, > P as long as Cd < r/ 3
(6.6)
Because reactor power P, scales faster with radius than the propulsive power P,
reactor size should not be a hard constraint in terms of allowing the submarine to
achieve the necessary speed. This is assuming that the drag coefficient Cd does not
scale faster than r5 / 3 , which is true.[39] Although the drag coefficient changes with
the radius of the submarine, the maximum drag coefficient is typically about 2.
In addition, the reactor compartment is just a small portion of the entire submarine. The weight of a reactor compartment for a Virginia class submarine is estimated
'Plant efficiency is simply the rate at which a plant can convert thermal energy into useful electric
energy.
76
at between 1,130 tons and 1,680 tons.[32 If we take the conservatively high estimate
for the reactor compartment of the submarine to be 1,680 m 3 , we find that the reactor compartment of a Virginia class submarine will comprise about 22% of the total
weight of the submarine. 2 [5] This means that the reactor compartment itself only
comprises about a fifth of the total weight of the submarine, making it unlikely that
changes to the size of the reactor compartment (in the axial direction of the boat)
will result in a significant alteration of the submarine's maximum speed.
We have shown that an increase in core volume does not prove prohibitive for a
submarine from a speed perspective and we have also shown that an increase in reactor
compartment weight is unlikely to prove prohibitive because the reactor compartment
constitutes a small portion of the total submarine weight and volume. However, size
increases might have detrimental effects on the performance of the submarine in terms
of maneuverability and tactical ability.
6.2
General Constraints on Reactor Weight and
Size
In addition to speed requirements, submarines must also be able to dive and turn
quickly underwater, and to transit underwater in shallow water. These are capabilities that are likely to be affected by weight and size increases of the boat. Weight
increases will likely make a submarine less maneuverable in its turning ability. Similarly, a larger submarine is more difficult to turn and is more restricted in its ability
to submerge in shallow water. Weight increases or fluctuations can also alter the
submarine's buoyancy and center of mass, changing the performance ability of the
boat.[11]
A significant increase in both size and weight could result for two main reasons: (1)
the increase in weight from additional fuel, components, or shielding is so much that
it requires additional buoyancy for the submarine, meaning larger ballast tanks and
2
The 1,680 m3 of the reactor compartment, divided by the 7,700 m 3 of the entire submarine
yields 22%.
77
a larger submarine; or (2) the actual addition of the fuel, components, and shielding
takes up so much space that it requires the hull diameter of the submarine to increase,
which increases the total displacement of the boat.
These more specific constraints on size and weight are discussed in the following
sections. It is interesting to note, however, that the Navy is already proposing largescale changes to current submarine size and weight characteristics. New procurements
of the Virginia class submarine starting in 2019 will potentially have the new Virginia
Payload Module (VPM). This module is an additional 70-foot long mid-body extension to the submarine, which will carry four additional large-diameter vertical launch
tubes. These launch tubes could be made to hold up to 28 Tomahawk cruise missiles
among other types of payloads.[10] This new section will no doubt add a massive
amount of weight to the boat (a new 70-foot section is an additional 18.6% of the
entire length of the submarine). This would impact maximum submerged speed characteristics as well as ballast and buoyancy needs. If the Navy is willing potentially to
install such a module, it seems reasonable that a slightly larger reactor compartment
in terms of length and weight would also be achievable.
6.3
Specific Constraints on Reactor Weight and
Size
An important constraint is the reactor core size and the size of the accompanying
pressure vessel. The most successful redesigns will not increase core size, but if the
core size must be increased to reach adequate performance characteristics, the size
increase might be compensated for by size reductions in other areas.
A reactor for the Virginia class fast attack submarine would need to fit within a
hull diameter of around 33-34 feet.[32] The existing reactor pressure vessel for that
submarine is about 66 inches in diameter, and has a total reactor pressure vessel
volume of about 1-2 m 3 .[401
The reactor design is constructed in such a way to be as compact as possible.
78
Telescoping control rods help limit the size of the overall reactor to little more than
the size of the pressure vessel.[3] The entire reactor compartment of a current Virginia class submarine can be estimated at about 40 feet in length and a diameter of
33 feet.[32] The reactor compartment consists of primary and secondary shielding.
The primary shielding covers only the core and pressure vessel. The larger secondary
shielding covers the primary shielding compartment as well as the pressurizer, the
main coolant pump, the steam generator, and all of the piping connecting the various
components. Piping from the steam generator then leads out of the shielded compartment to the main turbine and condenser. Figure 6-1 shows the typical location
of a reactor compartment and a typical submarine propulsion system.
A "A C.amq ftu.0
MAIN ENGINE THROTTLE
wqn.M "W-
PRESSURISER
STEAM
CONTROL
GENERATOR
TURBO-GENERATOR
MOTORS
GEARING
\REDUCTION
MAIN TURBINLECEA
\\
CLT
PROPULSION
\
MOTOR
THRUST BLOCK
BATTERY
REACTOR
MI
MAIN COOLANT PUMPONENE
ED
BLED
CONDENSER
ODNE
MOTOR GENERATOR
Figure 6-1: Submarine Propulsion System.[1]
In addition to weight increases that might result from alterations to the core, an
increase in core volume would also increase the shielding required around the reactor
compartment, which when compared to weight increases from the core itself, will be
much larger.
79
6.4
Shielding as a Source of Size and Weight
Historically, the Navy has been concerned about the safety effects that a new LEU
core might have on the radiation exposure of workers on the submarine or in the
shipyard during refueling.[11] This section will discuss these concerns, their origins,
and what effects they might have.
6.4.1
Occupational Radiation Exposure
Because of the crew's proximity to the core, naval reactors must be designed to
reduce the crew's radiation exposure. Traditionally naval reactors have five barriers
that protect the crew and the environment from dangerous releases of fission products:
(1) the fuel is developed to contain fission products that are produced over the lifetime
of the reactor and to expand as fission product gas is formed; (2) fuel elements are
contained in cladding and sealed; (3) the core is contained in a reactor pressure vessel;
(4) the entire nuclear reactor system is contained in its own, independently sealed
reactor compartment on the submarine; and (5) the physical hull of the submarine
is a last resort for keeping in radiation and acting as a barrier to the environment.[1]
Figure 6-2 shows traditional shielding in a naval reactor design.
The new reactor must perform just as well, if not better than, current HEU reactors in terms of radiation exposure risk. It must meet all of the Navy's current
safety requirements and must be constructed in a way that does not result in greater
occupational radiation exposure for workers who are operating within the primary
containment area when compared to the current radiation exposure levels on submarines from HEU cores.
There are two main sources of radiation that have to be shielded: gammas and
neutrons. These radiation sources require different shielding approaches that do not
necessarily complement one another.
In addition, radiation is produced not only
directly in the core but also in the primary coolant and in all components that interact
with it.
One of the first steps in designing a successful shielding system is determining
80
SCHEMATIC
(SHOWI$G
OF A REACTOR COMPARTMENT
SIELD
COMPONENTS AND TERMIWOLOGY OF SHELDIG)
OVER S-1rV
FROM SHIELD-
HO AIAT
ALL TIMES
L
POWER.
HURSEM
PLANET
TAO
-
ER
PEE NES AT
AEERAL
PER DAY
ACCESS DURING
LOW P4OWER, SEERL
P- - - - MLER.TE
PV
VEIMRTSELS
POUS
-
_____
\
REACTOR COOLANT
SYSTEM (PRIMARY). CON4TAINS
RtADOACTIVE COOLANT
MUST BE SHIELDED
REFLECTOR
THERMAL SWILD
Figure 6-2: Naval Reactor Shielding.[1]
an appropriate level of radiation exposure for personnel who will be operating and
working in proximity to the reactor.
This acceptable dose rate will influence how
thick the shielding is, what material the shielding is made of, and the physical layout
of the shielding. Ultimately, these will determine how expensive the shielding will be
and will also set limits on how long personnel can work in proximity to the reactor.
The next section will focus on specific scenarios that might require different shielding procedures for new LEU core redesigns.
6.4.2
Design Considerations with Shielding and Safety for an
LEU Core
Based on the Navy's safety philosophy, any redesign of the reactor must meet, or
exceed, current safety goals. With this as an assumption, a new LEU reactor design
should in no way increase radiation exposure to the crew.
If LEU redesigns result in larger cores, additional requirements for more shielding
must be met in order to keep the crew safe. A larger core will have more exposed
81
surface area and therefore the potential for more neutron leakage.3 Designs with
increased neutron leakage will result in the need for more shielding weight and volume.
LEU cores could have the potential to reduce radiation exposure (or the need for
additional shielding) if an integrated design is selected. Such designs reduce the need
for larger amounts of coolant piping leading from the core to the heat exchanger that
exists in the current naval reactor design. Because integrated heat exchangers would
be connected directly to the pressure vessel, there is no need for primary coolant
piping that carries radioactive primary coolant in the loop. 4 However, it should
be noted that any improvements to radiation exposure that might be gained from
this new integrated design might be lost due to the potentially more complicated
cleaning process posed by the design. Also, because the heat exchangers, which
require periodic maintenance are so close to the pressure vessel and core, the risk for
occupational radiation exposure from normal maintenance increases. 5
6.5
Strategic Capability, Submarine Readiness, and
Deployment Schedules
Any significant physical change or alteration in performance made to the submarine
as a result of a new reactor design could affect the submarine's overall availability.
The Navy has stated that it requires 48 SSNs to carry out its mission successfully
and meet force protection goals.[10] The submarine force structure currently contains
54 SSNs, with two additional Virginia class submarines being added each year and
larger numbers of aging Los Angeles class submarines being decommissioned.[10] The
number of submarines in the force can be expected to peak in FY14 and FY15 at 55
boats. [10 However, because of the decommissioning of large numbers of Los Angeles
3
Neutron leakage depends on the geometric buckling, reflector size, and neutron flux of the core.
In general, larger cores have less fractional leakage for the same flux. However, moving to LEU will
require increased flux for a given power level. This is partially offset by an increase in uranium
density. In order to determine the neutron leakage, a one-group diffusion calculation is necessary.
4
Because the primary coolant passes through the core, it gets irradiated, and therefore needs to
be shielded throughout the primary loop.
5
From a private conversation with a Naval Reactor Engineer.
82
class submarines in the coming years, this number will start to drop and could bottom
out at just 42 boats in FY29 if the procurement trend continues.[10] This number is
then expected to rise back up to the requisite 48 in FY35.7 [101
In addition to the overall force level, the Navy also requires that 10 fast attack
submarines be out on a day-to-day basis as well as having 35 submarines able to
respond in a certain period of time should a war-time scenario arise.[10] If the force
does drop to just 42 boats, the war-time demand submarine quota would only be 32,
three short of the goal. The Navy is trying to mitigate this shortfall in submarine
numbers and has been doing so for the past decade. Some of the solutions for this
impending problem are extending the service lifetime of current boats by increments
of 3 to 24 months where possible and fighting for the additional procurement of four
Virginia class submarines not currently planned for. 8 [101
If LEU submarines have a lower availability factor because of the need for occasional refueling, the total number of boats may need to increase in order to meet
minimum force requirements. The following equations relate the total force level to
the fleet requirements defined by a war-time surge.
f leet requirements = f leet size [A]
(6.7)
A = 1- U
(6.8)
A = availability
U = unavailability
U =
n
m+n
6
(6.9)
The battle for procurement of two Virginia class boats per year was a huge victory for the Navy.
Although the numbers of boats are going to eventually drop below the published force level goals,
the submarine force is well positioned to thrive for the next several decades.
'This alone shows that there is some flexibility in the size of the strategic force.
8J[ is important to note that these numbers: a total force level of 48, ten fast attack submarines
out on a day-to-day basis, and a war-time surge of 35 boats in a given period of time, are all
numbers that are based on current calculations and predictions and could change in the coming
decades as technology advances, the role of submarines changes, or the geopolitical context that
calls for submarine operations shifts.
83
The quantity
n
is called the unavailability factor U. This represents the percentage
that any given submarine is unavailable for deployment. Naturally, the quantity 1-U
is defined as the availability, A, of a submarine.
n = time in overhaul or refueling
m = time deployed = dL
L = lif etime of core
d = deployment constant
In this equation, n refers to total time in overhaul that a submarine experiences
during its lifetime. The value m refers to the time that an individual submarine is
deployed. The value m is further divided into a ratio of the lifetime of the core of
the submarine, L, and its deployment factor, d. The deployment factor d is a ratio
of time a normal submarine spends out on deployment versus the time it spends in
port for typical, short-term maintenance. 9 A good estimate of this d value is 0.5,
corresponding to 6 months out per 12 month cycle.
(6.10)
d= 6 months
12 months
This equation will help to understand the relationship between refueling, longer lifetime cores, and meeting Navy force requirements.
Currently, the Navy defines its
necessary fleet size at 48 boats, with a war-time surge requirement of 35 boats. This
yields a required A of .73, and a related U of .27.
wartime surge requirement
fleet requirement
_
35
48
= - < =.3
A(6.11)
In order to minimize the fleet size while still meeting fleet requirements, we want to
9
This refers to normal maintenance between deployments, work-ups, etc. and not longer maintenance that requires the submarine to be in dry dock for extended periods of time. This long-term
maintenance is captured by the n quantity.
84
minimize A and maximize U. This can be done in two ways:
0 make n small (decrease the time a submarine is in port for maintenance or
refueling)
* make m large (increase the time core lifetime of a submarine)
6.6
6.6.1
Minimizing Time in Port
Refueling and Hulls
One obvious effect of any new LEU core design that requires additional refueling is
that submarines must be in port for longer periods of time, thereby detracting from
their ability to be mission-ready and deployed.
One possible strategy for decreasing the time in port for refueling of a new LEU
reactor design would be to install a refueling hatch to streamline the refueling process.
If an LEU core could be designed that met the performance and size requirements of
a current fast attack submarine, it might nevertheless require multiple refuelings over
the lifetime of the submarine. Currently, the refueling process for Los Angeles class
boats is slow and costly, requiring the submarine to be in dry dock for long periods of
time, and involving an extensive process by which the hull is cut and the entire core
is removed.[11] For many reasons, this process is problematic. One solution to this
refueling problem from an economic and duration standpoint is to install hatches on
the top of the reactor that can be opened more easily when the submarine is being
refueled. France is currently using this refueling strategy in its Rubis-class submarine.
Those submarines have achieved a refueling time of as little as five months, and the
submarines refuel every seven to ten years.10[5]
A= 1-
n
= n
= .416
= 1= 1= .92 > .73
m+&
dL + n
5+.416
_
(6.12)
10This can be compared to the current estimates of U.S. submarine refueling time, which is 24
months (in Naval Shipyard Portsmouth).
85
A refueling time of 5 months and a core lifetime of 10 years yields an availability
factor of A = .92.
This suggests that the current refueling strategy used by the
French would be a potential solution for meeting force requirement levels.
Moreover, another option for shortening refueling duration is a modular reactor
compartment. This compartment would contain a complete and fully fueled core and
would be inserted in a manner similar to the Virginia Payload Module. Estimates
show that it is projected to take roughly 18 months from delivery of a payload module
until it is fully integrated into the submarine." The Virginia class retrofitting would
be analogous to a new refueling process with a refueling time of 18-24 months.' 2
A=1-
+ =1-=1- 2 =.71~.73
m+n
dL+n
5+2
(6.13)
A refueling time of 24 months and a core lifetime of 10 years yields an availability
factor of A = .71. This suggests that a refueling strategy based on modular compartments would be a potential solution for meeting force requirement levels.
6.6.2
Effects of Refueling on Depth Performance
One result of a hatch system would be its effect on the submarine's diving ability
and crush depth. The crush depth is the depth at which a submarine can safely dive
without risking serious damage to its hull and is primarily a function of the material
strength of the hull. Installing a hatch system would potentially weaken the integrity
of the hull. This would most likely result in a decrease of the crush depth, which
could have significant strategic impacts on the submarine.
Submarines spend the majority of their time relatively close to the surface (either
at periscope depth or in the range of about 100 feet submerged)." However, depending on the mission requirements, submarines will sometimes need to dive to greater
"From NAVSEA briefing "Virginia Class Submarine Program Status, Sea Air Space Symposium
2013."
12 An additional 6 months are factored in to account for the additional time it is estimated to
take to install a more complex core module that has to be more fully integrated into the submarine
propulsion system when compared to a stand-alone Virginia Payload Module.
13 From conversations with submarine officers.
86
depths (anywhere from 800 feet to 1200 feet). This maximum depth, or crush depth,
varies based on the submarine in question, but we can estimate it at around 1200
feet.[3] This ability of the submarine to submerge safely to these depths is a result
of extremely strong steel composite hulls and sophisticated welding between joints
and compartments. [3] As discussed above, the addition of a new hatch system to help
expedite the refueling process, if necessary, would potentially reduce the crush depth.
If the hatch were designed so that it did not require extensive maintenance to open
(i.e., physically cutting the hull of the submarine), it would most likely not be as
strong as the surrounding hull. If the hatch were designed to be as strong as the hull
of the submarine and not detract from the diving performance of the boat, it would
most likely be complicated and more costly than the hatches that are currently on
submarines for personnel to enter and exit the boat, due to the required increase in
the size of the hatch to fit the entire pressure vessel. The development of a hatch
system that is both large enough to service the core and also strong enough to not
limit the diving ability of the submarine is challenging, but very feasible.
A reduction in the diving ability of submarines because of an LEU core is a
potential reason why a non-lifetime LEU core redesign might be an unattractive
option, and would require force structures to increase in size to account for the longer
refueling times that would result.
6.7
Maximizing Time Deployed
The second solution to maximizing a submarine's availability is maximizing the time
that the submarine can be deployed. One response to this problem is to consider the
changing mission space of the submarine under the assumption that a lower average
burnup per submarine is possible, as discussed in Chapter 2. If submarines can indeed
perform in coming decades on a lower average burnup per submarine, an LEU core
with such characteristics becomes more attractive. The new LEU core could have a
lower burnup than the current HEU cores, relaxing the stringent constraint on core
87
size by making it easier to reach the necessary energy density.' 4
However, assuming that this mission space remains the same as it is currently, and
the burnup requirement does not change, the incentive to have a longer core lifetime
is clear in terms of maximizing deployable time.
Naval reactors have been developed to minimize life-cycle and operation costs
and as such have been designed to have cores that will last the projected lifetime
of the boat, or at most require one refueling.[11] The current lifetime core for the
Virginia class submarine is estimated to last 33 years.[5] In the Navy's criticism of
LEU cores, the Navy estimates that it would require about $1 billion per submarine
to refuel if the core did not last the entire life of the boat. [11] It will be imperative
from both a strategic force level perspective and a budgetary perspective to ensure
that an LEU redesign has the potential for a long core lifetime or affordable refueling.
Longer lifetime cores tend to require heavier and larger cores, which as discussed
above, can present other design challenges. While analysis suggests that perhaps the
longest lifetime cores possible are not imperative, minimizing the size and weight
effects that accompany efforts to increase core lifetime are important. Reductions in
size and weight can be obtained from a combination of general reactor compartment
alterations and a more compact integrated reactor design.
6.7.1
Reactor Compartment Alterations
As described in the previous section on design strategy, a new LEU core could require
either less shielding or more shielding, depending on the design. If a new core design
required less shielding, this would make the reactor compartment lighter and would
allow the recovery of valuable space around the reactor area and working space in
the reactor compartment.
Any additional space that can be gained is a positive
development.
With regard to altering the reactor compartment, the introduction of the Virginia
"If a new LEU core must only meet a typical average power profile of 15-20% of a 140 MWt core
as opposed to the current 20-25%, it can have a lower burnup, or lower energy extraction of the
fuel. This means that the new LEU core can be smaller because it can contain a lower amount of
uranium that will ultimately need to be burned during the submarine's operation.
88
Payload Module (VPM) suggests a substantial amount of leeway in terms of weight
and lateral size restraints. (The VPM could add as much as 70 feet along the axis
of the boat.) [10] The reactor compartment likely could be elongated in the lateral
direction of the submarine if necessary to accommodate additional shielding or components. The most restrictive dimension of the reactor will be its height, which will
affect the hull diameter but not the lateral dimension of the reactor compartment.
6.7.2
The French Integrated Design Revisited
Another solution for saving valuable space is a more integrated reactor design. The
French company Technicatome produces an integrated steam generator design, which
consists of a steam generator that is directly integrated into the reactor pressure
vessel. [5] This design reduces the need for additional piping for the primary coolant.
The design is shown in Figure 6-3.
The design resembles a traditional pressure vessel structure, but the vessel encapsulates not only the core but also small vertical steam generators, that ultimately
produce steam at the top-most portion of the pressure vessel.[5] Pipes then extract
the steam and carry it away to the turbine. The feed water enters the pressure vessel
just above the steam generators. Not only is this design more compact in terms of
components and the piping between them, but also perhaps an even more significant
result is the decrease in the need for extra shielding. With a combined core and steam
generator contained in one pressure vessel, one large shielding could conceivably be
used around this component, eliminating the need for two layers of shielding and
presumably reducing the overall shielding volume and weight.
Moreover, a reduction in piping might also reduce radiation levels produced from
radioactive coolant and lower noise levels that result from more coolant flows in
pipes. 15
France has successfully used this integrated design to construct a 48 MWt reactor
'5 At the same time, the integrated system, in general, involves a more complicated piece of
equipment that could experience more failure modes than the traditional separate design. The
integrated design could also be more difficult to perform maintenance on, and having high-pressure
steam so close to the core and a major radiation source may present a design and safety challenge.
89
-
Pump
GeCiranit
Primard
Auxffiary
cfrcuit
Figure 6-3: French Integrated Design.[5]
with a 10-year core life inside of its Rubis-class attack submarine. [5] The hull diameter
of the Rubis-class submarine is 25 feet, which is smaller than that of the Virginia class
submarine by about 8 feet.[5] Thus, France was able to design a successful integrated
system that had a maximum dimension (presumably the vertical direction) that did
not impede it from fitting the system into the hull. Even though the Rubis-class
produces almost 100 MWt less power than the Virginia class, this buffer of 8 feet,
combined with other innovative design strategies, suggests that an integrated design
in a Virginia class submarine might not only be possible but might be a key design
feature in a proposed LEU reactor.
90
6.8
Summary
The performance requirements of a submarine, mainly speed and maneuverability,
mandate certain requirements for its size and weight. We have shown, through scaling factors that dictate submarine speed, size, and propulsive power required, that
reactor size should not be a hard design constraint from the perspective of maximum
speed. Instead, constraints on reactor size will result from requirements on submarine maneuverability. These size increases can affect a submarine's center of mass and
buoyancy, and make stopping abruptly and turning in tight directions more difficult.
Even this constraint, however, does not seem difficult to satisfy. If the Navy is willing
to include large alterations to the submarine in terms of the new Virginia Payload
Module, perhaps increases in size and weight are indeed tolerable. In any event, we
have also outlined strategies in terms of reactor compartment layout and more integrated reactor designs that should hopefully be able to mitigate any increase in size
and weight that might result from a new LEU core design.
91
Chapter 7
Conclusion and Recommendations
7.1
Conclusion
Current reactors in the U.S. Navy's submarine fleet are fueled with HEU cores with
lifetimes of up to 33 years. The Navy developed these specifications over time by
balancing reactor endurance with compactness, and utilizing the best available fuel.
The proposed conversion of naval reactors from HEU fuel to LEU fuel has already
been studied by the Navy twice. The first of study in 1995 concluded that: (1) the
use of LEU in existing submarines would require that nuclear cores be replaced before
the lifetime of the submarine had otherwise ended; and (2) redesigning naval reactors
and submarines to use LEU would be extremely expensive and disruptive. The later
study in 2014 concluded that there are two main options for using LEU fuel in the
future: (1) substituting LEU fuel for HEU fuel in current reactors; and (2) developing
a new fuel system that can increase uranium loading in order to compensate for some
of the expected impacts of LEU fuel.
Compelling motivations for re-assessing the feasibility of conversion to LEU, combined with new fuel technology and reactor technology advancements, justify a new,
thorough investigation of the issue. Evidence suggests that the mission space of the
submarine force is changing in such a way that would allow decreases in average
burnup per submarine. This would warrant a relaxation of previous constraints that
mandated higher burnups and thus made LEU cores more challenging to design. Also,
92
increases in high density fuel technology and improvements in reactor components like
steam generators create design space that make LEU options more attractive. In addition, proliferation concerns have arisen, for example, from a potential loophole in the
NPT that might allow other countries to produce and store HEU without adequate
safeguards provide a strong incentive to work toward reducing all HEU stockpiles. Finally, an investment in new reactor technology and a thorough re-assessment of new
LEU cores should produce significant benefits for Naval Reactors as an organization,
and possibly also for the submarine force by fostering technology innovation.
With these motivations in mind, this thesis analyzed naval reactors to establish a
set of basic constraints on fuel types, fuel performance and reactivity, core geometry,
chemistry and corrosion performance, thermal hydraulics performance, and physical
plant performance. The thesis then revisited each constraint with a postulated LEU
core design to estimate how this design might augment existing performance.
One of the most difficult constraints for an LEU design will be meeting the reactivity requirements to overcome xenon-135 poisoning throughout the lifetime of the
core, especially at higher burnups. This is because LEU cores exhibit more rapid
declines in reactivity at high burnups when compared to HEU. This loss of reactivity
over time, combined with the physical degradation of the fuel at high burnups, will
determine the maximum lifetime of a novel LEU core design.
From a thermal hydraulics standpoint, a new LEU design should not prove to be
prohibitive.
LEU cores should perform acceptably unless increases in core volume
or the fuel-to-moderator ratio require a significant increase in pumping power that
either results in large, noisy pumps or eliminates the reactor's ability to be passively
cooled in an emergency scenario.
Lastly, LEU designs that require increases in weight or size of the reactor compartment are not necessarily infeasible. Analysis shows that reactor size will produce
enough power to meet top speed requirements faster than resulting volume increases
will detract from the ability to meet those requirements. Instead, increases in weight
and size of the reactor compartment potentially will negatively impact submarine
maneuverability. The Navy appears willing to include large alterations to the sub-
93
marine in terms of the Virginia Payload Module, suggesting maneuverability is not
ultimately a hard constraint. Because space in submarines is limited, every effort
should be made to keep a new core design small, while still producing a refueling
schedule that meets Navy requirements for fleet size and war-time surge capability.
Fleet size and availability provide a strong incentive for lifetime cores, but analyses
show that current French refueling schedules and core lifetimes of ten years could
also work. Alterations to the reactor compartment, either by slightly elongating the
compartment or switching to an integrated reactor design, might compensate for any
increases of weight and volume expected with redesigns.
The assumption that LEU core redesigns will result in impermissibly large new
cores that fail to meet performance standards can no longer be made. Advancements
in fuel technology and reactor technology suggest that a workable LEU core design
is increasingly possible. In conclusion, the Navy has compelling reasons to make a
significant investigation of advanced naval reactor technology and LEU designs, and
to conduct a thorough analysis of the feasibility of LEU conversion.
7.2
Recommendations for Future Work
For each of the constraints that axe developed on fuel and neutronics performance,
chemistry and thermal hydraulics performance, and physical alterations, more thorough testing and modeling needs to be done.
Specifically, the following are two
important areas for early future work:
* Various high density fuels outlined in Chapter 4 should be analyzed under conditions similar to those in naval reactors to determine if they truly meet the
requirements for use in the demanding environment.
This testing should in-
clude high temperatures, high burnups, potentially corrosive chemistry, and
simulation of the required battleshock conditions.
" Full core burnup models should be run for several different high uranium loading fuels and core geometries to determine if these cores can meet reactivity
94
requirements at high burnups and accordingly to ascertain which core lifetimes
and volumes axe truly attainable.
95
Bibliography
[1] T. Ippolito, "Effect of Variation of Uranium Enrichment on Nuclear Submarine
Reactor Design," Master's thesis, MIT, 1990.
[2] T.H. Newton Jr. M.S. Kazimi and E. E. Pilat, "Development of a Low-Enriched
Uranium Core for the MIT Reactor," tech. rep., MIT, 2007.
[3] M. Ragheb, "Nuclear Naval Propulsion," tech. rep., Nuclear Naval Propulsion,
Nuclear Power: Deployment, Operation and Sustainability, 2011.
[4] N. E. Todreas and M. S. Kazimi, Nuclear Systems: Thermal Hydraulic Fundamentals. CRC Press, Taylor and Francis Group, 2012.
[5] C. Ma and F. von Hippel, "Ending the Production of Highly Enriched Uranium
for Naval Reactors," The NonproliferationReview, 2001.
[6] "IAEA Safety Glossary: Terminology Used in Nuclear, Radiation, Radioactive
Waste, and Transport Safety," tech. rep., Department of Nuclear Safety and
Security, 2006.
[7] "United States Naval Academy Submarine Warfare Education Guide."
[8] M. Gorenflo and M. Poirier, "The Case for More Submarines," tech. rep.
[9] P. B. Obama, "Remarks by President Obama to the Australian Parliament,"
tech. rep., The White House Office of the Press Secretary, 2011.
[10] R. O'Rourke, "Navy Virginia (SSN-774) Class Attack Submarine Procurement:
Background and Issues for Congress," tech. rep., Congressional Research Service,
2014.
[11] N. N. P. Director, "Report on Use of Low Enrichment Uranium in Naval Nuclear
Propulsion," tech. rep., Naval Reactors, 1995.
[12] N. N. P. Director, "Report on Low Enriched Uranium for Naval Reactor Cores,"
tech. rep., Naval Reactors, 2014.
[13] 0. R. Cote, "Associate Director, MIT Security Studies Program.".
96
[14] R. O'Rourke, "Navy Ohio Replacement (SSNBNX) Ballistic Missile Submarine
Program: Background and Issues for Congress," tech. rep., Congressional Research Service Report, 2013.
[15] C. J. Harvey, "At Sea Over Naval HEU: Expanding Interest in Nuclear Propulsion Poses Proliferation Challenges," in Nuclear Threat Initiative, 2010.
[16] R. S. Kemp, "Technology Matters: The Gas Centrifuge, Supply-Side Controls,
and the Future of Nuclear Proliferation," InternationalSecurity, vol. 38, 2014.
[17] G. Theilmann, "Submarine Nuclear Reactors: A Worsening Proliferation Challenge," tech. rep., Arms Control Association, 2012.
[18] J. C. Moltz, "Closing the NPT Loophole on Exports of Naval Propulsion Reactors," The NonproliferationReview, 1998.
[19] A. de Sa, "Brazil's Nuclear Submarine Program: A Historical Perspective," Master's thesis, Princeton University, 2013.
[20] A. L. Kennedy, "Remarks on FMCT," tech. rep., U.S. Permanent Representative
to the Conference on Disarmament, 2012.
[21] A. Glaser and F. von Hippel, "On the Importance of Ending the Use of HEU in
the Nuclear Fuel Cycle: An Updated Assessment," in 2002 InternationalMeeting
on Reduced Enrichment for Research and Test Reactors, 2002.
[22] F. von Hippel, "IPFM," tech. rep., International Panel on Fissile Materials Blog,
2014.
[23] "September 2009 Draft of the FMCT," tech. rep., International Panel on Fissile
Materials, 2009.
[24] "2013 Report," tech. rep., International Panel on Fissile Materials, 2013.
[25] M. Wald, "Company Struggles to Keep U.S. in the Uranium Enrichment Game,"
tech. rep., New York Times, 2014.
[26] M. Wald, "Kentucky Operator to Cease Enrichment of Uranium," tech. rep.,
New York Times, 2013.
[27] M. Miller, "The Use of HEU in Naval Nuclear Reactors and Its Implications for
a Fissile Material Cutoff Treaty (FMCT)," 2003.
[28] R. Ward, "Prospects for Conversion of U.S. Naval Propulsion Reactors," tech.
rep., 2011.
[29] Y. S. K. G. L. H. M. K. M. T.H. Newton. J. Rest and S. L. Hayes, "U-Mo Fuels
Handbook, RERTR Program," tech. rep., Argonne National Lab, 2006.
97
[30] "Good Practices for Qualification of High Density Low Enriched Uranium Research Reactor Fuels," tech. rep., IAEA Nuclear Energy Series, 2009.
[31] M. Benedict T.H. Pigford and H. W. Levi, Nuclear Chemical Engineering.
McGraw-Hill Book Company, 1981.
[32] "Naval Nuclear Reactor Compartment Shipments on the Columbia River," tech.
rep., Oregon Department of Energy, 2013.
[33] A. Hanson, "Executive Director, MIT International Nuclear Leadership Education Program.".
[34] "1995 Settlement Agreement: Overview," tech. rep., Idaho Department of Environmental Quaility, 1995.
[35] D. A. Jones, Principles and Prevention of Corrosion. 1995.
[36] M. Short, "MIT Assistant Professor of Nuclear Science and Engineering.".
[37] "Glossary," tech. rep., U.S. Nuclear Regulatory Commission, 2014.
[38] A. B. Baggeroer, "MIT Professor of Mechanical & Ocean Engineering.".
[39] R. Serwat, Physics for Scientists and Engineers. Saunders College Publishing,
1990.
[40] "Over 112 Million Miles Safely Steamed on Nuclear Power," tech. rep., United
States Naval Nuclear Propulsion Program, 1997.
98
Download