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