3

Goals for a Fusion Pilot Plant

A fusion pilot plant is required to demonstrate key performance and cost metrics to directly enable first-of-a-kind (FOAK) commercial fusion power plants. This creates a set of interconnected requirements in the scientific, technical, and economic missions of a fusion pilot.

INTEGRATED FUSION AND ELECTRIC POWER PERFORMANCE

The primary goal of the pilot is to produce electricity from fusion energy. Therefore, there are basic requirements for the fusion core and the plant’s electric performance.

The first requirement is that the concept must demonstrate reliable net energy gain from fusion reactions. The first hurdle is net plasma gain, Qp—that is, the ratio of fusion power to input power to the plasma, must exceed unity. It is self-evident that this objective is the minimum required for a power plant, but it is also an important threshold in fusion science since this also represents the minimum where the fusion plasma can start to heat itself in order to sustain the temperatures necessary for fusion. No fusion concept has yet met this criterion, although deuterium-tritium (D-T) tokamaks have achieved up to Qp = 0.67. The second hurdle is net plant electrical energy gain Qe, such that the amount of time-averaged electrical power generated by the fusion core exceeds the electrical power required for operation. This step requires the conversion of fusion energy into electricity taking into account the total electrical power for plasma heating, and any other forms of power required (e.g., pumps for heat removal or refrigerators), and there-

fore requires plasma gain well above one. Furthermore, this step requires the use of appropriate technologies to efficiently capture and convert the fusion power to electricity.

The second requirement of the pilot is that it provides cost certainty to the marketplace in terms of capital cost, construction time, control of radioactive effluents including tritium, the cost of electricity, and the maintenance/operating schedule and cost. The electrical market and fusion science guide the range of power generation in a pilot. The upper range of mature technology electrical generating units, ~400-1500 MWe, should be avoided in the pilot because these will inherently not meet the minimum cost requirement. The lower limit for a pilot is set by two considerations. First, cost certainty requires a minimum power generation scale such that the extrapolation to FOAK fusion power plant is reliable and reasonable. The FOAK power is not well defined, in part because its power level will be set by the pilot performance and marketplace considerations, but a range of 100-500 MWe appears to be a reasonable entry point for a dispatchable, centralized electricity source such as fusion. Secondly, the science of achieving sufficient Qp and Qe requires a minimum fusion power because the fusion power is also the plasma’s primary heating source. While configuration dependent, most public¹ and private² designs for fusion plants are in the 100-500 megawatt thermal (MWth) range, which would approximately translate to a gross electric power of 40-200 megawatt electrical (MWe) assuming a conversion efficiency of 40 percent, while net electric power depends on multiple design factors including Qp, cooling power, and wall plug efficiency of plasma heating.

The third requirement of a pilot plant is the demonstration of sufficient fusion power duration that the plant is providing certainty as to the cost and viability of FOAK operation in the marketplace. The nature of fusion technology and science leads to several distinct phases of demonstration and learning for a pilot plant toward the ultimate goal of operating the pilot in a manner closely representative of FOAK operations. These phases are set by the characteristic timescales involved in the fusion system.

MATERIALS AND MANUFACTURED COMPONENTS

It is evident from input received from electrical utilities that in order for the fusion pilot plant to provide the foundation necessary to enable a FOAK fusion power plant, the pilot plant in Phase 2 operation will need to produce continuous energy, or perhaps with a representative availability of at least 50 percent, for at least one, if not two, environmental cycle(s). An environmental cycle is characterized as a period that includes installation of integrated core components, fusion plant operation in which the fusion environment degrades the component performance to the point where it must be replaced or repaired, and that such maintenance actions are taken to allow further operation of the plant. While an environmental cycle is not strictly defined numerically, because it can vary across design approaches, based on our present knowledge base it likely represents a significant (on the order of one full power year) operational time before maintenance/repair would be required in a FOAK fusion power plant. It is important to note that this range of times would be familiar to and expected for operators of FOAK power plants in the marketplace. The requirement for operation for an environmental cycle imposes clear goals for the performance of key components. PFCs or armor, internal structural and blanket materials, power extraction systems, and breeding materials (if required) need to successfully demonstrate operation up to a global neutron or ion surface/ wall loading equivalent to approximately 1 to 3 MW-year m⁻² on the PFC armor, and approximately 2 to 3 MW-year m⁻² on the internal structural and blanket materials. This will enable these components to experience the coupled environmental degradations arising from combinations of plasma particles, thermal loading, ionizing radiation, and high-energy neutrons. While this combination may vary across approaches, all pilot designs must account for this degradation. Table 3.1 provides goals for the PFCs and structural materials in Phase 2. Phase 2

TABLE 3.1 Goals for Material Operation During Phase 1 and 2 of a Fusion Pilot Plant

NOTE: Phase 3 has the same goals as Phase 2 with the addition of more environmental cycles and/or modified materials and components.

will also need to demonstrate effective remote maintenance and component replacement in order to provide greater certainty for the maintenance time and operation and maintenance costs of the FOAK fusion power plant. The combination of neutron flux, remote maintenance strategy, material properties, and facility availability will determine the duration of Phase 2 to test material degradation discussed in Table 3.1 and can vary depending on fusion concept.

The Phase 1 operation will focus on power generation for a relatively short duration in which the challenges associated with the power rate (global power, particle flux rate) are overcome but without the degradation that will occur after days of operation. Phase 1 itself is expected to have two operating sequences. The large engineering components such as magnets, vacuum vessel, and plasma heating and cooling, will need to be commissioned in a pilot. Phase 1a (Table 3.1) would focus

on establishing the equilibrated fusion power performance and immediate thermal response of the PFC actively cooled components, which is their first integrated test at representative thermal conditions. The duration of plasma operation and heat exhaust would be increased within Phase 1b in order to demonstrate the ability to reach a steady rate of fusion power, capture the fusion energy, and demonstrate the capability to generate electricity. From a material and components viewpoint Phase 1 demonstrates solutions to the so-called “zero dpa (displacements per atom)” integration challenges—that is, issues centered around power generation and extraction but at sufficiently low energy/particle fluence such that the components should experience minimal degradation by the fusion environment. A list of goals for the Phase 1 operation are also provided in Table 3.1.

The Phase 2 operation focuses on achieving a full environmental cycle for the materials and components. The structural materials of the first wall and blanket will degrade under neutron bombardment. One can use a well characterized material such as low-activation ferritic martensitic alloys to estimate the requirement of this phase. Confidence exists that these materials can adequately perform up to doses of 50 dpa and 500 ppm He accumulation within the temperature range from 400 to 550°C arising from fast neutron-induced transmutation. For 14 MeV neutrons from D-T fusion, this translates to an incident neutron energy fluence of ~5 MW-year m⁻² at the first wall surrounding the fusion plasma. The optimal materials/ technology choice of other key components such as the plasma-facing components for the divertor, first wall armor, and associated active cooling methods that can meet the power exhaust, material erosion, and tritium retention requirements are less certain at this time. Therefore, the Phase 2 requirement is to meet a neutron wall loading of 1 to 3 MW-year m⁻² on the PFC armor and 2 to 3 MW-year m⁻² on the structural materials of the vacuum vessel and breeding blanket such that modifications to the structural materials and first wall components will be significant and measurable, but are not expected to result in failure of the components and entail their immediate replacement.

Meeting the phase 2 requirements leads to the possibility of introducing a Phase 3 of pilot plant operation that would have the goal of further defining the mean time to failure as well as providing the option to serve as a component test facility to evaluate the performance of advanced materials and technology in the complex fusion neutron environment, which includes significant cyclic stress, chemical compatibility, and neutron-induced materials degradation challenges. To enable Phase 3 operations with the mission of testing advanced technology, the pilot plant could be designed with the flexibility to replace sectors and more extensive remote maintenance capability to change out PFC/divertor, structural, and blanket components to accommodate different technologies.

The goals for material performance in pilot phases are summarized in Table 3.1.

FUEL AND ASH

Fusion produces net energy through nuclear reactions of various light isotopes that eventually result in helium. For example, our sun primarily produces energy due to the fusion of four hydrogen nuclei into a single helium nuclei, through a chain of reactions that involves the intermediate production of helium-3 and deuterium, the stable heavier isotope of hydrogen. Alpha particles are the most energetically favorable end state of fusion due to fundamental nuclear physics. Therefore, subsequent fusion of alpha particles is negligible and is considered the “ash” of fusion energy. The fact that the primary end product of fusion is a stable inert gas with no safety or disposal concerns is one of fusion energy’s most attractive features.

Terrestrial fusion sources do not use stellar fuel cycles based solely on hydrogen since they are too slow and generate low power density. Other fusion fuel combinations are considered that have higher (but variable) reaction rates and immediate products. These include D-T fusion resulting in a helium and neutron, deuterium-deuterium (D-D) fusion producing helium-3 (also stable), tritium, neutrons, and protons, deuterium-helium-3 (D-³He) fusion producing alpha particles, and a proton and proton-boron (p-¹¹B) producing three alphas. The various combinations of these reactions, their immediate products, and the helium ash make up the fusion fuel cycle. The production of energetic neutrons and gamma rays results in machine activation, leading to concerns for radioactive material disposal, and material radiation damage induced degradation as discussed in the previous section.

Ash Removal

Helium ash removal is required for all fusion fuel cycles. The helium is born as a nucleus (alpha) at extremely high kinetic energies in the fusion plasma, constituting some fraction of the fusion energy. The helium ash must be removed at a steady state rate from the fusion plasma system and replaced with the fusion fuels, to maintain a constant fusion power. The helium is eventually removed as a neutral gas particle at near room temperature in a region outside of the fusion plasma, typically a pump adjacent to plasma-facing components where the helium ions are neutralized into helium atoms. The helium ash undergoes varying degrees of energy loss before impinging on the PFCs. The impingement generally poses a significant challenge to materials because helium, as an inert element, is insoluble

in materials. If the helium is promptly lost to PFCs near its birth energy (3.5 MeV) it will deeply embed in materials and cause blistering and spallation that results in material loss, while also causing high local heat flux. In some configurations the helium ash is at intermediate energies (>1 keV) and this can produce helium bubbles and cavities in the material, which can degrade thermal conductivity and drive swelling of the material. In a common configuration, called the divertor, the helium is at low energies (10-100 eV) but high flux density, which causes major surface modifications such as the growth of tendrils in PFC refractory metal surfaces.

The coupled issues of ash removal and the evolving viability of PFCs due to helium bombardment will be a critical requirement to demonstrate in all fusion pilots, regardless of configuration or fuel cycle.

D-T Fuel Cycle

D-T is the most common fusion fuel cycle presently under consideration due to its relatively high reactivity (10 to 100 times higher than the other fusion reactions at 100 million degrees Kelvin) and high energy gain per reaction. Tritium has a short half-life of 12.3 years, decaying into helium-3, and therefore no natural source of tritium exists. In D-T fusion the resulting 14.1 MeV neutron, which carries 80 percent of the fusion energy, is slowed down in a surrounding blanket and forced to undergo another reaction, mostly with lithium-6, which produces a helium atom and a tritium atom. The ratio of tritium produced in the blanket to the D-T neutrons is called the tritium breeding ratio (TBR). TBR must be at or above unity in order to have the D-T fusion power plant sustain itself, given that its consumables are deuterium and lithium. Detailed calculations of the neutron interactions in the blanket show that TBR can exceed unity due to the presence of other neutron reactions, which also increase the thermal power by an energy multiplier M compared to fusion power (Figure 3.1). This excess is important since it is envisioned to be the source of “starting” tritium for subsequent fusion power plants.

The present world inventory of commercially available tritium is highly limited (~40 kg) and largely produced as a by-product of heavy water fission plants (Figure 3.2). While Ontario has committed to, and is in the process of refurbishing 10 of its 18 CANDU reactors, the tritium supply from CANDUs beyond the year 2040 has large uncertainty. In addition, the ITER fusion project will store ~4 kg of tritium and consume ~12 kg but not breed tritium in any significant quantity. The deployment of advanced fission reactors that could produce tritium is also unclear. The uncertain tritium production/recovery of the next 20 years, coupled to 12.3 year half-life of the tritium, motivates proceeding with the construction and operation of a pilot plant as soon as practical.

Some advanced fission reactors, specifically fluoride salt-cooled high-temperature reactors (FHRs) and some molten salt reactors use fluoride salt coolants that contain lithium and produce tritium at rates comparable to heavy water reactors. Because these advanced fission reactors require systems for tritium control and recovery, they may have capability to extend tritium supplies and also demonstrate tritium control and recovery technologies that can be used in fusion power plants.

The rate of tritium production in a D-T pilot plant breeding blanket represents a significant step up from present commercial experiences. For example, a 500 MWth fusion system burns ~75 g of tritium per day or 28 kg/full power year. With a TBR~1 this implies the production of ~75 g tritium/day in the blanket, the recovery and processing of this tritium, and its recycling back into the D-T fusion plasma (Figure 3.1). This can be compared to the CANDU 650 MWe fission reactor, which produces ~150 g-T/year in the heavy water moderator. This tritium is periodically (~annually) removed and stored. The entire 12 GWe CANDU Ontario fleet produces ~2.5 kg/year, which due to the tritium decay results in a steady-state ~30 kg inventory, representing the majority of the world’s supply. From this comparison several insights emerge.

There are presently many blanket concepts being considered to meet the simultaneous demands of fusion power removal, neutron shielding, and tritium breeding. Concepts include a variety of solid and liquid blankets in various configurations to achieve these goals (Figure 3.1); however, all concepts are at low technology readiness level (TRL). Furthermore, this low readiness has to be addressed since among other reasons it has significant design implications regarding the maintenance scheme for the pilot plant. The choice of coolant and the range of operating temperature of the blanket are important for overall efficiency of the pilot plant, and this has far-reaching consequences on the design performance of the blanket. The blanket also serves as a primary component in shielding sensitive components, such as magnets, from the deleterious effects of neutrons and high-energy photons. The radial thickness of the blanket is a key consideration in setting the overall size of the fusion pilot plant.

The required inventory of tritium at a fusion pilot site depends on several design features. The first feature is the burn fraction—that is, the fraction of tritium injected into the fusion plasma that undergoes fusion. This fraction will be small because the helium ash must be kept at a low concentration (<10 percent) in the plasma and therefore most of the exhausted particles (>90 percent) are un-burned deuterium and tritium atoms, which must be processed and put back into the plasma. Burn fractions can be as low as 1 percent, which would necessitate 7,500 g of tritium to be processed per day at 500 MW fusion power. Developing a direct purification and recycle loop (i.e., direct internal recycle) as closely coupled as possible to the fusion device would be the most effective way to reduce fuel cycle building inventory. The removal of ash and non-fuel impurities in this process stream is difficult and has to be addressed.

The second important feature is the characteristic timescale to process the exhausted D-T mix and the tritium recovered from the blanket. The third design feature is the amount of tritium that is retained in materials inside the fusion reactor such as PFCs and blanket materials. Tritium permeates all materials to various degrees, so tritium confinement barriers are a necessity as the large components of the core could accumulate a significant quantity of tritium. Together these features determine the tritium residing in the plant because the tritium fuel in the plasma core is negligible (<1 g). Furthermore, they define the required minimum “starting” inventory of tritium fuel, which is important for the plant’s power availability in case of a shutdown or for the start of a new fusion power plant. There is a strong motivation to decrease tritium site inventories at a pilot to decrease the licensing burden, improve safety, and enable easier start and restart of a pilot and fusion power plants.

Tritium purification, handling, isotope separation, and storage have historically been developed as batch or semi-continuous processes since these are cost-effective approaches for the relatively low processing rates required for non-fusion applications. The need to demonstrate continuous operation of these systems in a D-T fusion pilot plant will also likely require continual operation of the tritium processing system, which has not yet been demonstrated. Effective recovery of

tritium from the blanket is an important part of tritium processing. Permeation, solubility, and materials handling issues with lithium-containing metals and salts will need to be resolved since all blanket concepts face challenges and presently have low TRL.

Strict tritium management will be required for a D-T fusion pilot plant to meet regulatory controls and site limits, ensure compliance with safety bases, and maintain continuous facility operation. Previous operational experience with tritium in fusion facilities has been limited to Tokamak Fusion Test Reactor (TFTR) and Joint European Torus (JET), where the lifetime inventories were ~100 g. Tritium accountability for these machines was burdensome. A D-T pilot plant will have a tritium throughput of several kilograms a day and will have tritium breeding in the blanket. ITER’s tritium accountancy process requires collection of all the tritium into hydride beds for calorimetry, which is not directly feasible for a continuously operating pilot plant.

Alternative Fuel Cycles to D-T

The technology challenges that face alternative fuel cycles are less defined because most research efforts to date have concentrated on the D-T cycle. Nonetheless, alternate fuel cycles have been proposed and studied such as D-D and p-¹¹B that have abundant terrestrial fuel sources, and D-³He, which would likely require a lunar source of He-3 for a FOAK fusion power plant. The U.S. He-3 inventory is ≈20 kg and would be capable of producing ≈400 MWth-y. Additional inventory would be required for the processing system. This quantity of He-3 appears to be sufficient to complete Phase 2 of the pilot operation for the minimum electrical power but may limit Phase 3. There are additional resources available of He-3 from the decay of tritium produced in the CANDU reactors, but these have not been assessed.

The most significant design advantage of these alternate fuel cycles is removing the requirement for tritium production/breeding and recovery from the reactor

blanket. Conversely these fuels have hundred-fold lower reactivity rates at 10 keV (100 million degrees Kelvin) where D-T achieves net energy gain, and therefore alternate fuels require both higher plasma operating temperature and much higher triple product (Figure 3.3).

As well, Figure 3.3 shows that alternate fuels have a lower fraction of fusion power released as neutrons. In some fuel cycles varied amounts of tritium are produced, and in all the fuel cycles some neutrons are produced either directly (D-D) or as side reactions that occur in the plasma, and therefore issues associated with tritium (radiological safety) and neutrons (activation, volumetric damage) are not eliminated but reduced to a variety of degrees. These alternate fuel neutrons are also of lower energy than those produced by D-T, decreasing transmutation rates, especially ⁴He production. Therefore, alternate fuels feature lower demands on neutron tolerant materials, activation, and tritium safety than D-T. Conversely the alternate fuels face more severe global design challenges for PFC surface heat removal since a larger fraction of the fusion energy is released as charged particles or photons (neutrons volumetrically heat a D-T blanket). Alternate fuel cycles often feature concepts that use open magnetic topologies and linear systems in order to meet this challenge, yet these have less experimental vetting than closed magnetic topologies. Direct energy conversion that converts the charged particle and plasma radiation directly to electricity is an example of an innovative technology that

could be applied to this challenge. Technology and science innovations must also be compatible with helium ash control and removal.

RELIABILITY AND AVAILABILITY

The pilot plant should provide operational and test data needed to assure reliability for the subsequent FOAK fusion power plant, which must be capable of operating with high availability (eventually greater than 85 percent). High reliability and availability will be an expectation for commercial customers based upon performance now achieved by existing fission power plants. A major focus for reliability will be to achieve very low forced outage rates. Scheduled maintenance outages for repairs, refurbishment, and component replacements are expected and accepted, because they can be scheduled during periods of seasonal low energy demand. But the ability to complete scheduled maintenance outages expeditiously, in time periods of several weeks will be preferred.

Reliability involves both short-term and long-term phenomena. Operations in the pilot plant should be expected to occur in phases, including commissioning, initial low power, and transient testing (Phases 1a/1b). An important point is that full tritium self-sufficiency is not required during this phase. However, subsequently the pilot plant must operate at full power in Phases 2 and 3 with availability greater than 50 percent to demonstrate reliability of those components that require periodic replacement through at least one full environmental cycle. Operating durations beyond one full environmental cycle will enable assessment of component lifetimes and assessment of the replacement/repair times and cost. Tritium consumption for D-T systems during this phase will be sufficiently large that complete, or nearly complete, tritium self-sufficiency is required. Alternate fuels will not be subject to this tritium breeding requirement. Significant upgrades and plant modifications may occur between these phases.

Modern approaches to reliability engineering, including computational modeling and engineering analysis to simulate plant assembly, operations, and maintenance, that incorporate best practices and lessons learned from large-scale project construction and systems integration to perform engineering computer aided design, structural analysis, and process and control modeling should be applied and

demonstrated. Fusion components such as plasma-facing components, the structural materials for the vacuum vessel, shielding and blanket modules operate under severe thermal and radiation environments. Substantial research and development (R&D) has been devoted to structural materials performance under neutron irradiation for these components, but design and integration into a functioning fusion power plant is highly complex and is one of the primary objectives of a pilot plant.

ENVIRONMENTAL AND SAFETY CONSIDERATIONS

Fusion provides the promise of a safe and environmentally acceptable energy source. These attributes are key to the public acceptance and overall economic competitiveness of fusion energy.

Demonstration of safe operation of the fusion pilot plant is one of its most important goals. In D-T systems, tritium dominates the plant’s source term, and mitigation of tritium release is key to the safety case. There is significant experience with the TFTR³ and JET operations, and with fission power plants (particularly CANDU reactors⁴), as well as with the design of ITER,⁵ that shows that tritium releases may be kept within allowable limits via design and choice of materials. All fusion concepts produce intense ionizing radiation. Neutron activation of structural materials will contribute to the source term in all fusion concepts, but again, through design, and primarily through material choice, this portion of the source term can be mitigated. The combination of source term plus pathway to release (e.g., via a large release of energy) can be minimized via design and materials choice.⁶

From an environmental perspective, minimizing waste volume and hazard overall, and avoiding greater than Class C waste as much as feasible (and completely, if possible) in the pilot plant is key. In a fusion pilot plant, greater than Class C waste can be avoided through the use of low-activation materials⁷ and/or the use of alternate fuel cycles, which lowers the material activation. The pilot plant may generate some greater than Class C waste if some low-activation materials will not be fully qualified, but it should demonstrate that the FOAK fusion power

plant could use these low activation materials.⁸ The fusion pilot plant will generate radioactive products requiring disposal in a near-surface disposal facility, and its design should seek to generate the minimum volume of waste.

REGULATORY PROCESS

The U.S. Nuclear Regulatory Commission (NRC), an independent federal agency, regulates the Nation’s civilian use of nuclear materials. This authority was granted to the NRC by Congress through the Atomic Energy Act of 1954, as amended, hereinafter referred to as the AEA.⁹ In 2009, the NRC determined, “as a general matter, that the NRC has regulatory jurisdiction over commercial fusion energy devices whenever such devices are of significance to the common defense and security, or could affect the health and safety of the public.”¹⁰

The NRC has not yet established a framework for fusion power reactors but is required to do so by December 31, 2027, per the Nuclear Energy Innovation and Modernization Act (NEIMA), which was signed into law January 16, 2019.¹¹ Section 103 of NEIMA requires the NRC to “complete a rulemaking to establish a technology-inclusive, regulatory framework for optional use by commercial advanced nuclear reactor applicants for new reactor license applications” by December 31, 2027. NEIMA defines an “advanced nuclear reactor” to include fusion reactors.

The NRC has begun to consider how best to regulate fusion power reactors, and on October 2, 2020, the NRC directed the staff “to consider the appropriate treatment of fusion reactor designs in our regulatory structure by developing options for Commission consideration on licensing and regulating fusion energy systems.”¹² To this end, the NRC staff is evaluating three main approaches: (1) treat a fusion power reactor as a “utilization facility”; (2) regulate the use of byproduct material at a fusion facility; or (3) use a hybrid approach.¹³ As a part of this process, the NRC held a public meeting October 6, 2020, jointly with DOE to discuss the regulatory framework for fusion.¹⁴

Utilization Facility

The term “utilization facility” is defined in the AEA to mean:

The NRC has, in turn, defined “utilization facility” to specifically include nuclear fission power reactors.¹⁵ Thus, the existing technical regulatory requirements under 10 CFR Part 50, “Domestic Licensing of Production and Utilization Facilities,” have been tailored to nuclear fission power plants. 10 CFR Part 52, “Licenses, Certifications, and Approvals for Nuclear Power Plants,” provides an alternative licensing process to Part 50 but references the detailed technical requirements of Part 50.

Although the NRC’s current regulatory definition of “utilization facility” does not include fusion reactors, the broad statutory definition provided in the AEA could permit a future determination by the NRC to include fusion plants within the definition of “utilization facility.” For example, in 2014, the NRC used a direct final rulemaking¹⁶ to modify the definition of “utilization facility” to include the SHINE facility, which is an accelerator-driven subcritical device used to irradiate uranium to produce molybdenum-99 for medical purposes.¹⁷ The decision to modify the definition of “utilization facility” and also use the 10 CFR Part 50 framework primarily stemmed from the fact that the process involved the fission of uranium atoms (a special nuclear material¹⁸) and the associated safety requirements such as criticality control, heat removal while operating and shutdown, and containment of radioactive fission products were determined to be applicable.¹⁹

Fusion power reactor designs under development do not use isotopes of uranium or fissile materials in their processes. The fuel types are isotopes of hydrogen, with lithium-containing blankets. Fusion power plants cannot have a chain reaction. As a result, safety issues associated with fusion are different from those associated with fission reactors and stem from control of relatively short-lived radioactive material, such as tritium and longer-lived radioisotopes generated by neutron activation of metallic materials in the structures.²⁰ If the NRC were to treat fusion reactors as “utilization facilities,” it would not need to use 10 CFR Part 50. Instead, the regulatory framework could be tailored to the hazards specific to fusion through rulemaking, as further discussed in the section “Hybrid Approach,” below.

Byproduct Material

In the AEA and in NRC’s regulations, radioisotopes used as fuel in a fusion facility, such as tritium, or generated during the fusion process, such as tritium and activation products, are byproduct material.²¹ This determination assumes the NRC would consider radioisotopes produced in a fusion reactor as being produced in an accelerator. The regulatory framework specified in 10 CFR Part 30, “Rules of General Applicability to Domestic Licensing of Byproduct Material,” and 10 CFR Part 20, “Standard for Protection Against Radiation,” contains the requirements that must be met when handling byproduct material as would be neces-

sary in a fusion facility. The regulations specify requirements for environmental protection, radiation protection of workers and the public, control of liquid and gas effluents, training of staff, operating procedures and processes, decommissioning, and depending on the activity levels present, requirements for hazards analyses and emergency planning and response.

The NRC uses a risk-informed regulatory process and as such scales its regulatory requirements according to the risk posed to public health and safety from the facility and/or activity. The regulatory framework in 10 CFR Parts 20 and 30 addresses the majority, if not all, of the radiological safety issues associated with a fusion facility without the unnecessary requirements specified in 10 CFR Part 50. The NRC is able to add any requirements necessary to provide reasonable assurance of public health and safety through the rulemaking process. Use of 10 CFR Parts 20 and 30 allows Agreement States²² to regulate the fusion devices unless the NRC determines that its authority should not be delegated to the states. Fusion devices not located in an Agreement State would be regulated by the NRC.

Right-sized regulation that ensures safety but is free from unnecessary burden is needed to enable rapid innovation in fusion technology and for fusion to be an economically viable non-carbon emission energy generation source. Provided the NRC determines that the material generated in a fusion facility can be categorized as byproduct material, the existing regulatory framework provided under 10 CFR Parts 20 and 30 is well suited to fusion technology.

Hybrid Approach

In a hybrid approach, the NRC could develop a new regulatory framework through rulemaking that uses aspects of 10 CFR Parts 20 and 30, classifies fusion facilities as “utilization facilities,” and uses some aspects of 10 CFR Part 50, such as licensing of operators, if necessary. As such, the regulatory framework would be tailored to the hazards posed by fusion and would only impose regulatory requirements that the Commission deems as necessary to provide reasonable assurance of adequate protection of public health and safety and to promote the common defense and security and to protect the environment.

Decommissioning

Upon cessation of operations of a nuclear facility, the decommissioning process is initiated whereby the facility is radiologically decontaminated and demolished. In the process, radiologically contaminated or activated materials are packaged and shipped to a licensed disposal facility as waste. Structures and the site grounds are then remediated to ensure that any residual radioactivity remaining on the site is reduced to a point of either unrestricted or, under some circumstances, restricted use. The facility license is then terminated. Decommissioning of nuclear fission reactors has been occurring successfully and efficiently at many facilities. The NRC indicates on its website that approximately 100 material licenses are terminated each year.²³

The NRC has extensive regulations that govern decommissioning of all nuclear facilities to protect the workers and the public during the entire process and until the site is released for use. To plan for decommissioning, the NRC requires that license applicants provide assurance that they will have sufficient financial resources available to complete the process.²⁴ In such a case, a decommissioning funding plan or a certification of financial assurance for decommissioning is required to be submitted before a license is issued. The cost of the decommissioning depends on the complexity of the process and the amount and type of waste requiring disposal and is accounted for in the decommissioning plan and in the regulatory requirements. The classification of the waste depends on the concentration and form of the materials.²⁵ The NRC requires that operations of a facility be conducted in a manner that minimizes contamination as a means of reducing the complexity of the decommissioning process.²⁶ The TFTR tokamak D-T fusion experiment was successfully decommissioned in 2002.²⁷ Advances in radiological decommissioning techniques are being made every year in the areas of remote tooling for segmentation and packaging of large, highly activated reactor components and likely can be applied to the future decommissioning of a fusion power reactor.²⁸

Most fusion designs will need to manage and dispose of tritium safely during decommissioning, which may be the most difficult issue that will need to be addressed if the tritium is mixed with long-lived isotopes in such concentrations that it is classified as greater than Class C waste.²⁹ These isotopes include carbon-14 and nickel-59 and niobium-94 in activated metal. High temperature treatment of the fusion components can be used to recover tritium, since it is desirable to recover the tritium for other fusion plants rather than dispose of it.

For all fusion designs, low activation metals can be used to reduce the radiological hazard to workers and decommissioning costs.³⁰ Currently, greater than Class C waste can only be disposed of in a high-level radioactive waste repository, which does not yet exist within the United States. The NRC is conducting rulemaking that is considering whether some concentrations of isotopes exceeding Class C limits can be safely disposed of in a low-level waste near surface disposal facility, and this rulemaking could reduce the complexity of disposal of waste containing greater than Class C waste.³¹ If no repository is available and the NRC determines that the greater than Class C waste generated at a fusion facility cannot be disposed of in a low-level near surface waste facility, then prolonged storage may be necessary until a repository is available.

If the waste contains only short-lived isotopes other than Ni-63, Sr-90, and Cs-137, then it can generally be disposed of in a low-level waste near surface disposal facility, currently licensed by an Agreement State.^32,33 However, if the waste contains only short-lived isotopes including certain concentrations of Ni-63, Sr-90, or Cs-137 then it would be classified as greater than Class C waste and it may need to be stored on site for several years to allow for radioactive decay and/or blended to acceptable limits and then disposed of in a low-level waste near surface disposal facility.³⁴

ECONOMIC CONSIDERATIONS

As one considers new generation pilot electrical plants and their economic value, we can look at how pilot plants have been funded in the past, their generating scale, and what has driven the economic value seen by the participants in developing these pilot plants. Table 3.2 compiles these results.

TABLE 3.2 Historic and Recent Examples of Various Pilot Generation Plants

^a Not completed.

^b Switched to natural gas in 2017.

^c Average price.

^d Capacity factor of these pilots are lower than those for other baseload pilot plants noted. Solar average for early plants were between 13-19 percent, Ivanpah capacity factor 24 percent, Dominion offshore wind capacity factor 42 percent.

^e Based on economic development for future off-shore wind plus transition to Virginia’s zero carbon goal by 2045. $210 million in economic output, 900 jobs created for construction, 1,100 jobs created for operation and maintenance, $21 million in local government revenues.

^f Based on publicly announced DOE total support of 3.2 B$ combined for the Natrium and X-energy projects with minimum 50 percent cost share from developers, and the cost split evenly between projects.

SOURCE: Shipppingport Atomic Power Station data from Wikipedia, https://en.wikipedia.org/w/index.php?title=Shippingport_Atomic_Power_Station&oldid=964406210;

Texas Clean Energy, Kemper County Energy, and Callide Oxyfuel data from World Nuclear Organization, https://www.worldnuclear.org/information-library/energy-and-the-environment/clean-coal-technologies.aspx;

Solar PV Grid Scale data from Lawrence Berkeley Laboratory; Ivanpah Solar Power Facility data from Wikipedia, https://en.wikipedia.org/w/index.php?title=Ivanpah_Solar_Power_Facility&oldid=977491449;

Vogtle Electric Generating Plant data from Wikipedia, https://en.wikipedia.org/w/index.php?title=Vogtle_Electric_Generating_Plant&oldid=964406778;

Dominion Energy Offshore wind data from Hampton Roads Alliance, https://hamptonroadsalliance.com/wpcontent/uploads/2020/09/Offshore-Wind-Economic-Impact-Report-092820.pdf and VA SCC ruling: Case # PUR-2018-00121, November 2, 2018;

Va. Transformation & Security Act of 2018, https://lis.virginia.gov/cgi-bin/legp604.exe?181+ful+CHAP0296+pdf;

NuScale data from https://www.sciencemag.org/news/2020/11/several-us-utilities-back-out-deal-build-novelnuclear-power-plant; NuScale is planning to increase the power by 25 percent to 924 MWe with minor design changes without any major changes to the design. The cost per MWe and the total cost listed in this table are based on the 720 MWe design from https://newsroom.nuscalepower.com/press-releases/news-details/2020/NuScale-Power-Announces-an-Additional-25-Percent-Increase-in-NuScale-Power-Module-Output-Additional-Power-PlantSolutions/default.aspx?utm_source=nuscalepower&utm_medium=web&utm_campaign=default-hero-1.

State, Local, and Engineering, Procurement, and Construction Support

Vertically integrated electric utilities are physically tied to the regions they serve by the large amount of infrastructure invested in their transmission and distribution network plus generation assets. These utilities have a vested interest in ensuring they respond to the desires of the populace they serve by being engaged in their communities, providing service at the level their customers require, and providing the type of generation their customers demand while they ensure reliability and resiliency. By responding to their customer base with the lowest cost of service possible while providing the services required, they obtain public support, political support and regulatory support. Unlike other industries, if they are to remain in this business sector, they cannot simply move to another state if they do not like the direction being taken by their customers, politicians, or regulators but rather must either work with them or sell their franchise to someone who does.

As a result of this relationship, state and local governments tend to work with energy providers as they know how to attract industry to increase employment and improve the quality of life for the populace that they serve with the energy they provide. Given that electricity is fundamental to modern society, the financial health of this service provider should be in the state’s interest.

Looking at past pilot plants in the United States from Texas, Mississippi, California, and Virginia (Table 3.2), one can see where state plus local engagement has aided the development of these installations. The support for these facilities has come in various forms such as investment tax credits, deferment of taxes, legislative support for cost recovery up to a defined amount, and training assistance for personnel.

Engineering, procurement, and construction (EPC) industries may participate as strategic partners for the building of the pilot plant if the industry participants have a relationship and there is belief that they will partner on future generation facilities.

Federal Support

Federal support of pilot plants of any new advanced electricity generation technology is important. A recent example is the October 2020 announcement of federal support for advanced fission concepts (Table 3.2). A fusion pilot plant provides a number of potential national security benefits, including increased scientific knowledge that can support other advanced applications for the nation; sustainability of fuel supply; enhancing the diversity of firm energy sources for the future with the potential to assist with the restoration of the transmission system; and development of a new industry to provide economic value to the nation as a provider of materials, engineering, and facilities to other countries.

As has been seen in other projects, the federal government has been actively engaged in cost sharing.³⁵ Fusion should be considered as a payment-for-milestones as described in the recently enacted bill HR-133 Sec. 2008. Other types of cost sharing include tax incentives such as federal tax credits, production tax credits, grants to specific companies, enhanced rates of return for utility investments, and R&D investment through universities and National Laboratories.

Generation Size and Cost

The cost of new technology pilot plants has varied over the years and with the generating technology. Not surprisingly generating costs of pilot plants are typically above those for mature plants. Table 3.2 gives a snapshot of pilot generation projects, along with some of their costs to construct reported in 2020 dollars per MW. These examples help provide the proper context for a fusion pilot plant whose goal is to accelerate the development of a FOAK fusion power plant. Indeed, it is this stated goal of a pilot that differentiates it from fusion devices designed to address purely technical or scientific challenges. Figure 3.4 shows, for recent examples

from fission, that as the maturity and completeness of the design concept increases the absolute costs generally tend to increase. This example clearly illustrates the importance of developing a mature design to develop a reliable cost and schedule basis, and obviously motivates one to minimize the pilot’s scale and cost while still meeting its technical requirements.

The first consideration is total cost. The ability of individual investors and utilities to invest in a pilot plant and/or FOAK power plant depends on the overall cost estimate for the pilot facility. Their investment will likely be based on the overall capital cost of the pilot facility and the firm’s capitalization capacity or risk investment capability. Input to the committee indicated that only the largest U.S. utilities have capitalization capacity that would allow ~$5 billion investments in a known technology, and many utilities would be more limited than this in their total cost. An exception to this in Table 3.2 is Vogtle at more than $25 billion cost. Vogtle uses well-established fission technology so is not a pilot in that sense but is listed because it was the first major fission build in a generation in the United States with a depreciated life expectancy of 40 years. It continues to face challenges even with substantial government backing, and its cost escalations provide a cautionary

tale about large size and total costs. Conversely, smaller investments will motivate utilities to purchase the technology, while larger costs can risk the viability of the entire utility. This immediately suggests a similar limit, effectively set by the U.S. electrical marketplace, for a FOAK fusion power plant that even following a fusion pilot will be a relatively immature technology. The maximum FOAK cost should be evaluated periodically based on input from the energy marketplace, consistent with the final recommendation of Chapter 2.

Given this reality of the U.S. market, the next consideration is the total cost for the fusion pilot plant. The pilot plant has linked technical and economic goals with significant private-sector contributions. The private-sector developers must in turn follow the constraints of the marketplace in which they will deploy their technologies in order to attract investment. If the private sector, even with government backing (Table 3.2), will not accept a total price past $5 billion to $6 billion for generating technology already demonstrated at scale, then it will certainly not accept this for the pilot plant. In addition, a pilot plant should have long term benefit to the industry’s development so the capital cost can be depreciated over a long period of time.

A pilot plant does not need to generate power at the same level of a FOAK fusion power plant or NOAK power plant, which would be determined by economies of scale and marketplace demands for a mature energy technology. The pilot plant will not be designed to demonstrate the normalized generation construction costs, $/MW, of a FOAK fusion power plant for several reasons. The first is that the technology readiness is less mature, which is the motivation for building a pilot plant. Compared with other technologies such as fission, solar power, and off-shore wind generation, fusion is at a lower level of technological readiness. The fusion pilot plant will improve technological readiness. The second is that there are economies of scale that can be realized by designing and operating the plant at higher electrical powers. These can be realized by making the plant larger on the basis of the same scientific and technological basis but will increase the cost. This is in conflict with the goal of seeking to build the pilot plant at the lowest possible cost. An example of a relatively low power pilot plant is the Dominion Energy Off-shore Wind pilot at 12 MWe. The normalized cost was relatively high at $25 million per MWe. The higher normalized cost should be expected at low power because some costs (permitting, design) are relatively insensitive to scale and represent a minimum cost. Nonetheless, there is a strong motivation to reduce the normalized generation costs,

$/MW, even for the fusion pilot plant. This is because it reduces the extrapolation to the marketplace. It will provide both the developers and utility operators greater assurance in committing to a FOAK fusion power plant. Projections of the normalized cost for a FOAK commercial plant will be a significant consideration at the decision point to construct the pilot plant, discussed in Chapter 5. Hence, this metric should be considered during the design phase of the pilot plant. Due to the changing marketplace, engaging experts in the electrical industry at the time of this decision would be valuable.

The levelized cost of energy (LCOE) for the FOAK fusion power plant will be defined in large part by the normalized generation construction costs, facility availability, and the operation and maintenance costs. Operating the pilot plant into a third phase represents an opportunity to improve our knowledge of mean time to failure for components, better defining the operating lifetime in the complex fusion environment and potentially testing advanced materials. It is important to note this benefit gained from use of the pilot plant must be balanced against the additional costs associated with Phase 3 operations. The operating and maintenance phase of the pilot plant will provide valuable data to estimate the operating and maintenance costs of the FOAK fusion power plant. While increased normalized costs are to be expected in the pilot plant due to the low technological maturity, developing an integrated solution that reduces the operating and maintenance costs and demonstrating it on the pilot plant would be advantageous. This has implications for the design of the in-vessel components, including the blanket, and remote maintenance.

This conclusion indicates a maximum normalized cost 100M$ per MW, which is more than four times larger than the examples cited in Table 3.2. Such a high normalized cost for the pilot plant would unlikely be sufficient to motivate funding for a FOAK fusion power plant. This motivates design efforts to decrease the normalized cost of the pilot plant and/or provide a clear pathway to reduce normalized cost through scale or parallel technology improvements.

Siting is another important factor in cost. One would have to assume that due to the technical challenges associated with fusion technology, the cost to build a pilot plant would be similar to those of advanced firm generating plant pilots or the offshore wind pilot. If possible, as much balance of plant that could be repurposed from an existing generator facility that is being retired, but was properly maintained, could reduce the balance of plant expenses. Otherwise, the pilot plant should use a standardized balance of plant design that fits the overall balance of plant parameters and enables operation of the smallest plant capable of meeting

the performance criteria. The overall cost requirement on a pilot facility could also be reduced if the plant is sited near a utility location with synchronous condensers, storage, or standby generators, since these features are required to support a fusion pilot plant. The present cost risk for fusion is the thermal side of the plant, not the balance of plant, so the value should be considered based on the cost of the thermal side of the plant and using the minimum cost balance of plant design to provide electrical energy. The balance of plant could have a cost profile as large as the thermal fusion core, and therefore this is an important consideration for a fusion pilot to meet its cost goals.

NOTES

1. See, for example, J.E. Menard, T. Brown, L. El-Guebaly, M. Boyer, J. Canik, B. Colling, R. Raman, et al., 2016, Fusion nuclear science facilities and pilot plants based on the spherical tokamak, Nuclear Fusion 56:106023, https://doi.org/10.1088/0029-5515/56/10/106023.

2. See, for example, B. N. Sorbom, J. Ball, T. R. Palmer, F. J. Mangiarotti, J. M. Sierchio, P. Bonoli, C. Kasten, et al., 2015, ARC: A compact high-field fusion nuclear science facility and demonstration power plant with demountable magnets. Fusion Engineering and Design 100:378, https://doi.org/10.1016/j.fusengdes.2015.07.008.; M. LaBerge, 2007, An acoustically driven magnetized target fusion reactor, Journal of Fusion Energy 27:65, https://doi.org/10.1007/s10894-007-9091-4.

3. C. A. Gentile, S. Raftopoulous, P. LaMarche, M. Viola, T. Walters, M. Kalish, T. Kozub, et al., 1996, TFTR tritium operations lessons learned, Fusion Technology 30:1564-1566, https://doi.org/10.13182/FST96-A11963173.

4. Canadian Nuclear Safety Commission, December 2009, “Tritium Releases and Dose Consequences in Canada in 2006”, INFO-0793, https://nuclearsafety.gc.ca/pubs_catalogue/uploads/CNSC_Release_and_Dose_eng_rev2.pdf.

5. I. R. Cristescu, I. Cristescu, L. Doerr, M. Gulga, and D. Murdoch, 2007, Tritium inventories and tritium safety design principles for the fuel cycle of ITER, Nuclear Fusion 47(7):S458, https://doi.org/10.1088/0029-5515/47/7/S08.

6. M. Victoria, N. Baluc, and P. Spätig, 2001, Structural materials for fusion reactors, Nuclear Fusion, 41(8):1047, https://doi.org/10.1088/0029-5515/41/8/308.

7. D. A. Petti, K. A. McCarthy, N. P. Taylor, C. B. A. Forthy, and R. A. Forrest, 2000, Re-evaluation of the use of low activation materials in waste management strategies for fusion, Fusion Engineering and Design 51-52:435-444, https://doi.org/10.1016/S0920-3796(00)00229-5.

8. L. El-Guebaly, L. Mynsberge, C. D’Angelo, A. Rowcliffe, B. Pint, and ARIES-ACT TEAM, 2017, Design and evaluation of nuclear system for ARIES-ACT2 power plant with DCLL blanket, Fusion Science and Technology 72:17-40, https://doi.org/10.1080/15361055.2016.1273669.

9. Atomic Energy Act of 1954 (P.L. 83-703).

10. SRM-SECY-09-0064, “Regulation of Fusion-Based Generation Devices” (Apr. 20, 2009) (SECY-09-0064).

11. U.S. Senate Committee on Environment and Public Works, “Nuclear Energy Innovation and Modernization Act (NEIMA),” https://www.epw.senate.gov/public/index.cfm/neima, updated January 28, 2019.

12. SRM-SECY-20-0032, “Rulemaking Plan on Risk Informed, Technology-Inclusive Regulatory Framework for Advanced Reactors (October 2, 20202) (SECY-20-0032).

13. W. Reckley, 2020, “Thoughts on NRC Regulatory Approach for Fusion,” presented at the DOE/ NRC/FIA Public Forum on a Regulatory Framework for Fusion, updated October 6, 2020. https://science.osti.gov/fes/Community-Resources/Workshop-Reports/DOE-NRC-FIA-PublicForum.

14. W. Reckley, 2020, “Thoughts on NRC Regulatory Approach for Fusion,” presented at the DOE/ NRC/FIA Public Forum on a Regulatory Framework for Fusion, updated October 6, 2020. https://science.osti.gov/fes/Community-Resources/Workshop-Reports/DOE-NRC-FIA-PublicForum.

15. C.F.R. 50.2 (definitions of “utilization facility” and “nuclear reactor”).

16. Definition of Utilization Facility, Direct Final Rule, 79 Fed. Reg. 62,329 (Oct. 17, 2014).

17. SHINE Preliminary Safety Analysis Report, Chapter 4—Irradiation Unit and Radioisotope Production Facility Description dated May 31, 2013 (ADAMS Accession No. ML13172A265).

18. Title I of the Atomic Energy Act of 1954 defines “special nuclear material” as plutonium, uranium-233, or uranium enriched in the isotopes uranium-233 or uranium-235.

19. SECY-14-0061, “Direct Final Rule: Adding Shine Medical Technologies, Inc., Accelerator-Driven Subcritical Operating Assembly to the Definition of Utilization Facility”(June 16, 20).

20. S. Sandri, G. M. Contessa, M. D’Arienzo, M. Guardati, M. Guarracino, C. Poggi, and R. Villari, 2019, A review of radioactive wastes production and potential environmental releases at experimental nuclear fusion facilities, Environments 7(1):1-12, https://doi.org/10.3390/environments7010006.

21. See 10 C.F.R. § 30.4 (definition of “byproduct material”).

22. United States Nuclear Regulatory Commission, “Agreement State Program,” https://www.nrc.gov/about-nrc/state-tribal/agreement-states.html, updated January 13, 2021.

23. United States Nuclear Regulatory Commission, “Decommissioning of Nuclear Facilities,” https://www.nrc.gov/waste/decommissioning.html, updated August 26, 2020.

24. See 10 C.F.R. 50.75, “Reporting and recordkeeping for decommissioning planning,” and 10 C.F.R. 30.35, “Financial assurance and recordkeeping for decommissioning.”

25. See 10 CFR 61.55, “Waste classification.”

26. 10 CFR 20.1406(c), the NRC requires all licensees to minimize the introduction of radiological contamination into the site environment to keep doses to as low as reasonably achievable and to facilitate decommissioning.

27. E. Perry, J. Chrzanowski, C. Gentile, R. Parsells, K. Rule, R. Strykowsky, and M. Viola, 2003, “Decommissioning of the Tokamak Fusion Test Reactor,” presented at the 20th IEEE/NPSS Symposium on Fusion Engineering, https://doi.org/10.1109/FUSION.2003.1426633.

28. D. Stenger, A. C. Roma, and S. S. Desai, July 2019, “Innovation in decommissioning and their application abroad,” Nuclear News,https://www.hoganlovells.com/~/media/hogan-lovells/pdf/2019/2019_08_16_jul19nn-innovation-reprint.pdf?la=en. Pp. 45-52.

29. See 10 CFR 61.55, “Waste classification.”

30. S. Sandri, G. M. Contessa, M. D’Arienzo, M. Guardati, M. Guarracino, C. Poggi, and R Villari, 2019, A review of radioactive wastes production and potential environmental releases at experimental nuclear fusion facilities, Environments 7(1):1-12, https://doi.org/10.3390/environments7010006.

31. D. Stenger, A. C. Roma, amd S. S. Desai, July 2019, Innovation in decommissioning and their application abroad, Nuclear News,https://www.hoganlovells.com/~/media/hogan-lovells/pdf/2019/2019_08_16_jul19nn-innovation-reprint.pdf?la=en. Pp. 45-52.

32. See 10 C.F.R. 30.4 (definition of “byproduct material”).

33. United States Nuclear Regulatory Commission, “Agreement State Program,” https://www.nrc.gov/about-nrc/state-tribal/agreement-states.html, updated January 13, 2021. United States Nuclear Regulatory Commission, “Decommissioning of Nuclear Facilities,” https://www.nrc.gov/waste/decommissioning.html, updated August 26, 2020.

34. See 10 CFR 61.55, “Waste classification.”

35. Office of Nuclear Energy, 2020, “It’s Time for the United States to Demonstrated Advanced Reactors,” https://www.energy.gov/ne/articles/it-s-time-united-states-demonstrate-advancedreactors-0.