SPARC as a platform to advance tokamak science

The past ten years, and the past five in particular, have seen a significant shift in the fusion research landscape, with increases in private funding so large that they now match or exceed public funding in many places.¹ These privately funded ventures are aimed toward commercializing fusion power and so are targeted in their research interest: resources are dedicated to gaining information that directly leads to a fusion power plant or components thereof. Publicly funded research, however, in many cases has the larger goal of advancing the frontiers of human knowledge and supporting education in addition to solving the specific challenges of fusion power.

While these missions are different, in many cases they overlap. The research performed in the public sector over the past 70 years has brought fusion energy science and technology to the point where near-term commercialization is possible. Without the past achievements of the worldwide fusion research program, it is very unlikely any private companies would be where they are today. This is certainly true for Commonwealth Fusion Systems (CFS), whose SPARC tokamak design² is based on the extensive physics knowledge gathered over the decades and recently refined for the design of ITER.³ It is also squarely in the interest of CFS and many other fusion companies for the public program to continue advancing the understanding of tokamak science in the future.

This overlap in mission also applies in the other direction. While the primary mission of the SPARC tokamak is to demonstrate Q > 1 (scientific breakeven)⁴ and to set the stage for construction of the ARC power plant (early iterations of which are discussed in Refs. 5 and 6), the SPARC project has and will continue to learn new physics along the way. The SPARC team has from the beginning emphasized the importance of making its scientific results available to the broader physics community, publishing an initial series of seven papers in the Journal of Plasma Physics describing the physics basis for SPARC,^2,7–13 continuing to publish subsequent scientific results in other journals,^14–18 and presenting at conferences and seminars. CFS has also made some physics calculation results and codes publicly available on GitHub for any that wish to pursue their own studies.¹⁹ 

Together, public and private sector research efforts have the potential to mutually benefit one another. Tightly focused private research will push the field in very applied directions, benefitting from and continuing to contribute back to publicly funded work. This overlap in mission and the benefit of public and private fusion programs working together has been clear from the very beginning of the SPARC project, which originated as a joint project between CFS and the Massachusetts Institute of Technology (MIT) Plasma Science and Fusion Center (PSFC).²⁰ The deep expertise and ability to do exploratory physics work at MIT complements the speed and tightly focused nature of a company like CFS. While there are occasionally different priorities that necessitate reconciliation, the partnership is mutually beneficial and has enabled the SPARC project to start at scale and move rapidly.

While all of the scientific work so far on SPARC has been predictive and used in the design of the tokamak and plant, the machine is now under construction and will soon start producing experimental data. As a part of the CFS belief that open science is the most effective way to advance the understanding of fusion, the intent is to publish physics results from SPARC once they are available. CFS will continue to work in close collaboration with academic, US Department of Energy National Laboratory, and international laboratory partners, building on the success of the past five years and contributing to the missions of both types of organizations.

This article focuses primarily on the potential of SPARC to contribute as a part of a joint public–private scientific program. The SPARC tokamak is unique compared to any publicly funded research device today and has the potential to contribute significantly to the worldwide understanding of tokamak physics. The rest of the article is organized as follows. Section II describes the SPARC tokamak and features of the tokamak and plant that make it attractive as a research facility. Section III describes the access SPARC will provide to unique plasma parameter ranges and the research potential in these areas. Section IV describes possible modes of collaboration between the SPARC team and the publicly funded program, including the present intent for SPARC data access policies. Section V summarizes these points as well as discussing possible next steps to further integrate SPARC physics into the publicly funded physics program.

SPARC will be a high field (toroidal field on axis $B 0 = 12.2$ T), compact (major radius $R 0 = 1.85$ m and minor radius a = 0.57 m), standard aspect ratio (⁠ $ε = a / R 0 = 0.31$⁠) tokamak capable of operating with deuterium–tritium fuel and designed to achieve up to P_fus = 140 MW and Q = 11 in nominal H-mode operation (⁠ $H 98 , y 2 = 1.0$⁠).² The high field and compact size are enabled by novel high temperature superconducting (HTS) magnets jointly developed by CFS and MIT.²¹ A rendering is shown in Fig. 1.

At the time of this article's writing, SPARC is under construction in Devens, Massachusetts, USA, as seen in Fig. 2. The primary mission of the machine is to demonstrate Q > 1 in a tokamak for the first time and to gain the physics, engineering, and operational knowledge needed to design the first ARC tokamak power plant, in parallel with an R&D program focused on fusion power plant technology. It is anticipated that the majority of this mission will be achieved within the first few years of SPARC operations, at which point it is possible that SPARC pursues much broader research objectives. The facility has a design lifetime of 10 years and 13 000 pulses.

While the SPARC facility is designed first and foremost to achieve its primary missions, there are a number of capabilities of the facility that make it attractive for use in more general scientific research.

The SPARC tokamak will have 50 ports to the vacuum vessel arranged around the machine between toroidal field coils. All ports are in the poloidal plane of the tokamak (i.e., there are no tangential ports). There will be 18 midplane ports (horizontal and centered at Z = 0) that measure approximately 1.2 × 0.5 m. Of these, ten will have permanently installed equipment: seven with pairs of four-strap ion cyclotron resonance heating (ICRH) antennas, two with leads for vertical stability coils and feedthroughs for ion cyclotron discharge cleaning (ICDC) antennas, and one for vacuum pumping. The other eight midplane ports will have removable “port plugs” that can be installed and removed from the outside of the tokamak without personnel access inside of the vessel.

In addition, there will be 16 upper off-midplane ports and 16 lower off-midplane ports, which are up-down symmetric and enter the vacuum vessel at a poloidal angle. These ports measure approximately 0.3 × 0.2 m. For each of the upper and lower sets, four locations are used for vacuum pumping and 12 will have port plugs.

The ability to remove and replace port plugs in SPARC will be important once the tokamak begins to operate with tritium fuel. Tritium operation leads to both tritium contamination of in-vessel surfaces (though this can be mitigated through baking and other wall conditioning) and to activation of materials by 14 MeV neutrons produced by the DT fusion process. Early in its operation, in-vessel maintenance will be possible in SPARC, but as the machine becomes increasingly activated all maintenance must be performed from outside of the tokamak. Port plugs that can be removed and replaced from outside the machine facilitate upgrades and repairs without in-vessel entry.

SPARC will initially be equipped with the set of plasma diagnostics required to accomplish early milestones, but will also include some other components that must be installed in difficult to maintain regions, such as behind plasma facing components. Initial concepts were presented in Ref. 22, and future publications will provide more specific details on final capabilities. SPARC is presently designing the following diagnostic systems: magnetics, interferometry, neutral gas, spectroscopy [near infrared (NIR) to x-ray], visible/IR imaging, bolometry, Langmuir probes, thermocouples, Thomson scattering, electron cyclotron emission, edge reflectometry, and a suite of neutron measurements (which are critical for measuring fusion power) such as time-resolved neutron counters and spectrometers. Notably, SPARC will not be equipped with any beam-based diagnostics as SPARC will not have any neutral beams. The present evaluation of operational risk has also excluded scanning probes or re-entrant material exposure stands.

Inside of the tokamak hall, the neutron and gamma environment during a pulse will be above levels where most plasma diagnostic equipment is expected to survive. Five diagnostic laboratory spaces, totaling approximately 375 m² of floor space, are on the other side of a 2.5 m thick concrete wall. As shown in Fig. 3, there are 30 general-use penetrations in this wall and a single dedicated aperture for poloidal imaging of neutrons and gammas. Each of the 30 general-use penetrations will be filled with a shielded plug, into which laser, x-ray, microwave, optical imaging, and spectroscopy beamlines can connect the tokamak hall to the diagnostic laboratories. The diagnostic laboratories will be far more easily accessible due to the reduced activation and tritium exposure concerns compared to the tokamak hall, meaning that there will be fewer restrictions on who can enter the diagnostics laboratories and less training will be required than for accessing the tokamak hall. This also facilitates collaborator access.

SPARC's combination of replaceable port plugs and shielded diagnostic laboratories will allow for considerable flexibility to upgrade the SPARC diagnostic suite in the future. As will be discussed in more detail in Sec. IV, one avenue of possible collaboration with publicly funded research would be to field diagnostics on SPARC that can study additional areas of physics. This flexibility will also allow SPARC to test new diagnostic designs that use ARC-relevant technology, materials, and components, as many of the early campaign SPARC diagnostics leverage existing techniques that do not translate to a long-pulse fusion power plant.

In addition to its diagnostic suite, SPARC will also be equipped with a reasonably flexible set of actuators. First, and most centrally, SPARC will have a large number of magnets designed to create varying plasma shapes, including complex time dependence with high current ramp rates. SPARC will have six superconducting central solenoid (CS) modules, eight superconducting poloidal field (PF) coils, four copper poloidal field coils (divertor coils, DIVC) inside of the toroidal field coil bore, a pair of copper vertical stability coils (VSC) inside the vacuum vessel, and 18 copper error field correction coils (EFCCs) arranged at three poloidal and six toroidal positions. All of the CS, PF, and DIVC coils can be independently controlled. The VSC will be connected in anti-series, such that the current in one is always in the opposite direction to the current in the other. All coils will be up-down symmetric.

The EFCCs are designed to robustly achieve both error field correction and to apply n = 3 resonant magnetic perturbations (RMPs),²³ but the initial configuration of the EFCC power supplies focus primarily on error field correction. A copper patch panel outside of the cryostat means that the coils could be reconfigured into a variety of error field correction and RMP combinations.

For plasma heating, SPARC will be equipped with ion cyclotron resonance heating (ICRH) that is expected to deliver up to 25 MW of coupled heating power through seven pairs of four-strap antennas (14 antennas total). The 120 MHz waves can heat helium-4, D, and DT plasmas at 12 T using a helium-3 minority.¹¹ In addition, the same frequency can heat helium-4 and D plasmas at 8T using a hydrogen minority. While the schedule for installing the full heating power is not yet finalized, SPARC anticipates having at least half of this ready for the first campaign.

Note that SPARC will not have electron cyclotron heating (ECH) due to the very high frequencies required to resonate at SPARC's magnetic field (more than 300 GHz) though if appropriate sources were developed, they could be fielded on the facility. SPARC will not have lower hybrid current drive (LHCD) due to its technical feasibility in the short term, but while no engineering studies have been performed to evaluate viability, it has not been precluded from future installation. Finally, SPARC also will not have neutral beams due to their complexity, cost, size, the very high density in SPARC (meaning that high energies would be required for core heating), and lack of tangential port access. It is unlikely that neutral beams would be supported in the future.

SPARC will also have a flexible set of gas injection for both fueling and impurity injection. SPARC will be capable of injecting the following gasses: H, D, T, ³He, ⁴He, Ne, Ar, Kr, and Xe. While often used in present-day tokamaks, nitrogen injection will likely not be possible on SPARC due to incompatibility with the tritium processing system. Impurity and non-tritium fuel species can be seeded into the divertor, and all species can be introduced into the main chamber, though not all injection locations have been finalized. For disruption mitigation, SPARC will be equipped with six massive gas injection (MGI) valves located on off-midplane port plugs.

Finally for actuators, while not a part of the initial build, SPARC will have the ability to install both high-field side and low-field side pellet injectors. A guide tube will be installed to the high-field side such that later installation would only be required ex-vessel. The facility will also be equipped with appropriate utilities and access for pellets should they be added in the future, though currently only deuterium pellets are envisioned, not tritium. Pellets may be used for both fueling and for edge localized mode (ELM) pacing.

Dedicated systems for wall conditioning will include baking, glow-discharge cleaning, and ion cyclotron discharge cleaning (ICDC). Glow discharge cleaning and high-temperature bakes (at approximately 350 °C) are anticipated at the beginning and end of campaigns, while low temperature bakes (at approximately 150 °C) and ICDC will be used throughout. The ICDC can also be used for boronization via deuterated or hydrogenated diborane injection.

The SPARC plasma control system will be based on an in-house control framework known as neutrino. Developed primarily in C++ and Python, this framework has emphasized simplicity and flexibility from the beginning, meaning that it can be readily upgraded to include additional functionality. Data storage will largely be handled by commercially available systems, and there will be an emphasis on facilitating connections with existing fusion software.

To summarize, SPARC will be equipped with a fairly wide range of diagnostics and a number of actuators. This initial set of diagnostics and actuators could be supplemented in the future via upgrades, facilitated by replaceable port plugs and diagnostic laboratories shielded from neutrons. It is acknowledged that some other tokamaks are more fully diagnosed and/or have a wider variety of actuators, but SPARC's flexibility is most significant when considering the unique plasma regimes that will be accessible in SPARC, as described in Sec. III.

The set of machine capabilities described above mean that SPARC will be equipped to take scientific advantage of the unique parameter regimes to which it will have access. In particular, SPARC will have the unique ability to address several of the critical areas of fusion research identified in recent US fusion community,²⁴ Fusion Energy Sciences Advisory Committee (FESAC),²⁵ and National Academies of Sciences, Engineering, and Medicine^26,27 reports. This section is organized into the research areas to which SPARC may contribute, with a focus on areas also identified as high priority by the public fusion research program.

First, SPARC will operate with magnetic fields (up to 12.2 T on axis) and plasma currents (up to 8.7 MA) that are significantly higher than other tokamaks of similar size. For this reason, SPARC will be able to extend empirical scaling relationships for many phenomena (such as energy confinement, access to H-mode, and density peaking), adding data at a similar size (⁠ $R 0 = 1.85$ m) to ASDEX Upgrade (⁠ $R 0 = 1.65$ m),²⁸ DIII-D (⁠ $R 0 = 1.66$ m),²⁹ EAST (⁠ $R 0 = 1.70$ m),³⁰ and KSTAR (⁠ $R 0 = 1.80$ m).³¹ For example, SPARC is quite similar to KSTAR geometrically, with comparable major radius, minor radius, elongation, and triangularity, but will operate at approximately four times the magnetic field and plasma current.

The SPARC plasma current flattop is designed to be 10 s, spanning more than ten energy confinement times (which is predicted to be up to 0.77 s).² In addition, simulations show that the current profile should be well equilibrated after approximately 5 s, such that the SPARC plasma will be stationary in both thermal transport and current relaxation.² 

SPARC will also be able to operate at 8 T, accessing plasmas with higher normalized pressure and providing more data at intermediate field. The only planned external current drive in SPARC is inductive via the central solenoid. SPARC operation will be able to determine the contribution of bootstrap current to the plasma current in these inductive regimes, and so will inform ARC design, which also presently plans on operating with only solenoid driven current. SPARC will, however, likely be able to utilize many of the same time-dependent techniques that are envisioned for “advanced tokamak” operational regimes, such as freezing current profiles with early heating, though such phenomena are not intended to be part of standard pulses on SPARC.³² 

The nominal SPARC DT H-mode, known as the primary reference discharge, is projected to achieve a volume average temperature of more than 7 keV and a volume average density of more than $3 × 10 20$ m⁻³.² Using the same methodology shown in Ref. 2, one can show that it is possible to achieve densities of up to $6 × 10 20$ m⁻³ with maximum heating power, though that would produce more fusion power than the magnet cooling is designed to remove. Similarly, SPARC is predicted to be able to generate volume averaged temperatures of more than 13 keV, though at lower fusion power and gain than the nominal discharge. These high absolute pressures are not accessible on existing tokamaks and so will provide experimental data in new regimes for both empirical scalings and validation of computational models. SPARC intends to contribute such data to publicly organized empirical scaling databases to improve these models.

In terms of non-dimensional parameters, SPARC is predicted to achieve $ρ *$ as low as 0.0027, $ν *$ as low as 0.0073, and β_N up to 1.0 m T MA⁻¹.² 

In addition to L-mode and standard ELMy H-mode, SPARC will have the opportunity to investigate other enhanced confinement regimes. Access to the I-mode regime is predicted to widen with increased magnetic field.³³ While SPARC is not relying on access to I-mode for high gain operation, existing confinement scalings (which have large uncertainty) predict favorable results for I-mode operation in SPARC.³⁴ SPARC operation will likely contribute data to both I-mode access and confinement databases.

In addition to I-mode, SPARC may investigate operation in the Super H-mode³⁵ and Wide Pedestal Quiescent H-Mode (WP QH-mode),³⁶ though the lack of torque injection in power plant-relevant conditions may make access challenging.³⁷ There is also the possibility of encountering unexpected confinement physics when operating at the unique absolute and normalized parameters of SPARC.

The data gathered on confinement, confinement regimes, and empirical scalings will be used to validate the predictive methods used to design ARC and other future devices. In particular, the data that SPARC should be able to contribute to empirical scaling relationships (such as confinement, density peaking, regime access, and others) will likely benefit all future tokamak design and operation efforts, both publicly and privately funded. While the majority of the learning is expected to be most applicable to tokamaks, it is possible that some physics is more generally useful for other fusion concepts. Note that SPARC does not anticipate having turbulence diagnostics as a part of the initial set listed in Sec. II, so detailed transport studies would likely require new or upgraded diagnostics.

Closely linked to its mission of high gain and high fusion power, SPARC is predicted to access regimes with $Q ≈ 11$ and so will be well into the “burning plasma” regime that is typically defined as Q > 5, where alpha heating is larger than externally applied heating. This will be accompanied by a significant population of fast particles and significant fast ion pressure.

The topic of burning plasma physics has been written about extensively in other articles.³⁸ Many consider the physics of burning plasma physics (particularly the non-linear, self-consistent equilibrium between energetic fusion products and the thermal plasma) to be one of the more important topics of study in future tokamaks.^24,25

SPARC operation will likely be able to show whether or not alpha slowing is purely classical. In addition, SPARC operation will provide some insight into the interaction of alpha particles with Alfvén eigenmodes and other MHD physics, similar to what is anticipated in ARC.¹³ Finally, SPARC will also likely see any fast ion effects on turbulence that may be relevant to ARC regimes.

While not a significant portion of SPARC's initial operational mission, SPARC could provide an excellent platform to study these effects in collaboration with the public program. The early campaign diagnostics listed above are not expected to be able to support all of these burning plasma research topics, so measurement enhancements will be required to do so. Every measurement of alpha physics in burning plasmas on SPARC would be novel and could lead to significant advancement in the understanding of new plasma regimes.

In this way, SPARC could be a significant complement to physics studies on ITER. SPARC will provide early looks at Q > 1 and burning plasmas in the 2020s before these are seen on ITER. ITER will likely have significantly more diagnostic coverage in a larger machine and will run with much longer plasma discharges (in terms of absolute duration, though only somewhat longer when normalized to plasma timescales). Data from SPARC and ITER will span significant parameter ranges and by combining data will hopefully lead to better physics understanding of these phenomena.

Another area of interest for SPARC will be disruption prediction, avoidance, and mitigation. This is clearly an active area of research on many tokamaks around the world and will be important for any tokamak fusion power plant.³⁹ SPARC is designed to explore disruption physics in power plant-relevant regimes of high stored energies, while being more robust than a power plant to off-normal conditions.

Early disruption work on SPARC will focus on validating the disruption physics basis¹² that was used in the design of SPARC to confirm the assumptions that went into setting structural load specifications. This will be useful in reducing uncertainty or eliminating conservatism for the design of the ARC vacuum vessel and its in-vessel structures. In addition, the massive gas injection (MGI) system will be commissioned and the gas mixture and trigger algorithms will be refined to mitigate damage to plasma facing components. SPARC has been designed to be resilient to runaway electron losses by having passively cooled plasma facing components. This removes the possibility of a cooling fluid leak, allowing for flexibility in learning how to manage disruption losses without risk of significant down-time.

SPARC also has hardware dedicated to study mitigation of runaway electrons. A passive, non-axisymmetric runaway electron mitigation coil (REMC) will be installed in SPARC¹⁶ to investigate the ability to suppress the generation of runaway electrons before they reach energies that will lead to component damage. A similar passive coil will also be installed in DIII-D.⁴⁰ The efficacy and engineering of these coils will inform the potential to use a similar coil in ARC or other future power plants.

Later work, however, will focus much more on characterizing disruption causes and mapping them to operation of a power plant in which the same pulse is run repeatedly. Other tokamaks around the world could also contribute to these studies, though it may require different types of experiments than are typically performed. In particular, running the same discharge repeatedly (such as a study on JET)⁴¹ is more representative of how a power plant will operate. Combining datasets of this type could lead to improved understanding of disruption causes and ways to reduce disruptions in future machines.

A principal challenge for any practical fusion energy system is the management of extremely high power flows into and out of the device.²⁴ While this is more pronounced in SPARC due to its compact size, the general challenge of power exhaust and particle control will apply to any fusion power plant and many of the physics studied in SPARC will apply reasonably well to any tokamak-based power plant.

SPARC is designed with fully up-down symmetric divertors and will have a fully metallic first wall, consisting of either pure tungsten or tungsten heavy alloy (WHA) plasma facing components throughout the machine. Pure tungsten is used in the areas of the first wall anticipated to receive the highest heat fluxes, while the more easily machined tungsten heavy alloy⁴² is used in lower heat flux regions, as it has a lower temperature limit. The decision to have a tungsten first wall was made primarily to reduce tritium retention and to more closely resemble the anticipated choice of plasma facing material in ARC. Similar logic led to the decision for metallic walls in Alcator C-Mod, ASDEX Upgrade, JET, and ITER.^43–46 

SPARC will be able to explore a range of dissipative divertor regimes. Magnet and power supply systems allow for strike point sweeping of approximately 0.5 m/s, while gas fueling and divertor geometries are designed to enable full detachment at high power. The operational constraints of heat flux and tungsten sputtering at low dissipation and detachment stability at high dissipation remain key uncertainties in the design of ARC or any tokamak-based power plant. This is especially true when also aiming to achieve acceptable energy confinement (i.e., $H 98 , y 2 ≈ 1.0$⁠).

The design of SPARC's plasma facing components (PFCs) assumed single null geometries to avoid requiring near-double null operation to manage power exhaust. Estimates of ELM energy flux from Refs. 9 and 10 indicate that SPARC may tolerate a few Type-I ELMs at full field and current without reaching PFC surface melting, but at lower ELM energies, it may be possible to sustain an ELMy H-mode for a portion of the flattop. This gives SPARC flexibility to develop power exhaust and particle control techniques without having to completely solve core-edge integration before operation.

Ultimately, ARC must handle high duty factor operation with low erosion. SPARC will not be well equipped to investigate material migration but does plan to explore access to and stability of detached divertor regimes in a range of magnetic configurations, including single null, double-null, and X-point target divertor geometries. This advanced divertor mission has received particular focus in diagnostic design, as many instruments need to be integrated with PFCs and cannot be upgraded later via port plugs. It is anticipated that after initial operation and demonstration of Q > 1, a considerable portion of SPARC's operational time will be dedicated to studies of integrated scenarios relevant to ARC, which may also complement work on other tokamaks around the world that are studying boundary physics. SPARC should be able to access absolute edge plasma conditions prototypical of what is expected on ARC.⁴⁷ 

Due to SPARC's high density and moderate size, it will also be able to access regimes of edge opacity, $n e , sep a$⁠, that are not accessible in other machines. In particular, having the ability to study regimes with high edge opacity but low Greenwald fraction will provide an opportunity to study the physics of edge gas fueling in new regimes and will likely inform the need for pellet fueling in future machines.

In addition to the edge opacity, SPARC will access regimes predicted to have very narrow scrape-off layer heat flux widths due to the large poloidal field.^48,49 Experimental data from SPARC will, thus, serve as an excellent validation of existing empirical predictions^48,49 and could validate computational models, such as XGC1,⁵⁰ which predict devices such as ITER to have an increased heat flux width relative to scalings.⁵¹ 

In summary, SPARC will be able to access regimes of operation that are largely unique compared to existing facilities. It will have a significant set of diagnostics and actuators to study these regimes and in doing so has the potential to provide invaluable data to empirical databases and for the validation of computational models. Section IV of this article describes some of the modes in which CFS may want to work with the broader scientific community and provide data for these pursposes.

Sections II and III of this article have described the capabilities of the SPARC facility, the parameter regimes to which it will have access, and the potential for SPARC to contribute to the publicly funded tokamak physics program.

While CFS is a private company and must maintain control over some proprietary information, there is significant motivation to facilitate the mutually beneficial public–private relationships that were discussed in the introduction of this article. This section describes the ways in which CFS has collaborated and will continue to collaborate with academic institutions, as well as what access to experimental SPARC data might look like. For the purpose of this article, “experimental SPARC data” generally refers to physics data that is gathered during a plasma pulse and is the data that would be shared with collaborators. This is in contrast to detailed engineering designs or manufacturing techniques used to build SPARC, which will largely be proprietary. More information can be found in an accompanying article in this collection.⁵² 

This section describes a range of scientific collaboration options, ranging from becoming deeply embedded team members to performing analysis on publicly available information.

In areas where the work of a research group is highly aligned with the interests of CFS and the SPARC project, it is possible that a long term collaboration or funded project makes sense, in which case the group would be embedded into the project. It is anticipated that this will only be a small fraction of the full engagement with the community, however, as it requires significant alignment of incentives and deep engagement.

In addition, there have already been several instances of publicly funded groups contributing to SPARC physics analysis via dedicated grants. CFS has been awarded several INFUSE grants from the US DOE, which have funded physics work at Princeton Plasma Physics Laboratory (PPPL), Lawrence Livermore National Laboratory (LLNL), Oak Ridge National Laboratory (ORNL), Columbia University, the University of California—San Diego (UCSD), and elsewhere.⁵³ In each of these instances, collaborators were funded by the US DOE to perform work that was related to SPARC or ARC physics and several resulted in publication. Groups at PPPL, ORNL, and General Atomics (GA) were also recently funded through standard DOE grants to work on topics related to SPARC physics,⁵⁴ and collaborative work is proceeding in the same fashion as with INFUSE grants. A similar model could very easily be extended to SPARC operations, though hardware contributions would most likely be more complex than analysis contributions. The agreement between the US DOE and Tokamak Energy is an example of how this might work.⁵⁵ 

It is worth noting that the government funding described in the preceding paragraphs is distinct from direct government grants or reimbursements to private entities, such as ARPA-E grants or the more recent DOE Milestone-Based Fusion Development Program.⁵⁶ While CFS has received a small quantity of direct government grants (such as through ARPA-E), the large majority of CFS funding comes from private investors. In contrast, all of the funding described in the preceding paragraphs would fund a government or academic researcher to do work that is related to SPARC or ARC.

In the longer term, it would also be possible that a larger institutional or government agreement could be made to set the operational priorities for SPARC later in the machine's lifetime. While early SPARC operation will be highly targeted toward specific goals and moving to ARC construction, later operation will likely be more exploratory. Similar to how the MST1 program in Europe funded several European tokamaks in exchange for the ability to determine the experimental program for some portion of the machine's runtime,⁵⁷ it is possible that an institutional agreement could be made with CFS for SPARC runtime, including access to all data from that operation. This type of agreement could be with a government funding agency (e.g., DOE or EUROfusion) or an individual university or research institute. The details of such an agreement would have to be worked out in the future.

Finally, it is likely that some information about SPARC will be made fully public, such as what has already been put on GitHub, and so researchers could perform their own analysis on that data (or other published information) without a direct connection to the SPARC project. This would be the loosest form of collaboration with SPARC.

First, and very central to the CFS attitude toward open science, the SPARC project plans to publish research work in peer-reviewed academic journals and to make these articles open access so that there is no paywall to read them. This topic is covered in detail in another article in this collection.⁵² 

A key part of collaboration with external scientific partners will be access to some experimental SPARC data, as access to actual data allows significantly more detailed study and integration with other datasets than interacting only with published papers. While SPARC will not start producing scientific data for a few years, the team recognizes that policies for access to these data are important to how the SPARC team will interact with the broader scientific community and so will attempt to lay out a framework here.

First, a few notes on the advantages and disadvantages of more open data access are presented. More open data access is advantageous in that it allows more equal access to information for the community and allows a larger set of scientists to use the data to advance the physics understanding of tokamaks. It also allows more complete combination of data into sets that can benefit those involved. More open data access also increases transparency and facilitates more complete scrutiny of results (though this can also be achieved through publication).

On the other hand, providing highly open data access requires considerable resources to maintain the hardware and software to facilitate this access. There is also considerable overhead in providing context to data such that those unfamiliar with its collection can analyze it productively. Fully open data access can harm the scientists that have put effort into planning for and collecting the data.⁵⁸ Finally, CFS is a private company and must balance the benefits of open data sharing with the benefits of proprietary information to the company.

It is worth noting that present policies regarding data access in the worldwide fusion community vary considerably from institute to institute. On the most open end, some facilities require only a loose collaboration agreement before allowing full access to all experimental data.⁵⁹ More recently, some publicly funded organizations have begun requiring non-government funded researchers to sign legal agreements assigning intellectual property rights to get access to experimental data, while publicly funded researchers are allowed to sign looser collaboration agreements.

The SPARC team anticipates that there will be several modes of sharing SPARC data that vary in their level of engagement. It is also anticipated that access will evolve over time as SPARC operation and the public funding environment evolves.

First, there will be a significant number of non-CFS researchers that will have full access to SPARC experimental data as members of the SPARC scientific team. These team members will have CFS computer accounts and direct access to experimental data from SPARC. Accounts and access, both physical and cyber-, will be managed through a “CFS Affiliate” program.

To date, access to experimental data from SPARC has only been assured for the founding partners of the project, CFS and MIT. Both of these teams have been involved in the SPARC project from the beginning and are deeply committed to its success. CFS sponsors basic and applied research at MIT that helps to advance both the SPARC and ARC projects. Access to SPARC experimental data is necessary for MIT to advance the scientific understanding of tokamak plasmas.

Prior to initial plasma operations, CFS expects to broaden commitments of SPARC data access to additional established contributors. For now, this likely includes only those funded directly by CFS, but if public funding changes in the future, the SPARC team may expand as well. It is possible that CFS will issue open calls to fund work at other institutions in the future, selecting the teams best suited to making SPARC successful. The SPARC team is also considering the possibility of third party validation of key results where there is significant value (such as Q > 1), though this has yet to be decided.

In addition to the set of deeply embedded, CFS-funded team members, it is anticipated that access to SPARC data will be provided to teams that make significant in-kind contributions. This may include construction, installation, and operation of diagnostics or actuators on SPARC, as well as significant in-kind contribution of analysis or computational work for SPARC or ARC. Publication preference for scientific results will be based on the type of work completed and its impact on the project.

While CFS is more than happy to work with in-kind contributors to the SPARC project and provide SPARC data access in exchange, the merit of any potential partnerships will have to be evaluated based on their contribution to the SPARC and ARC missions. In addition, the relative benefit of the work must be weighed against the cost of integrating hardware or facilitating analysis. To prioritize the mission of clean energy given finite resources and time, CFS will likely have to limit collaboration with in-kind contributors to only the highest priority work.

This mission focus will almost certainly leave significant areas of tokamak and plasma physics unexplored. For example, the initial set of SPARC diagnostics discussed in Sec. II is targeted toward the CFS-prioritized SPARC missions, but clearly leaves opportunity for additions to explore other areas of physics. If there was more significant public engagement with, and funding of, the SPARC program, at the government or institutional level, it may be possible to use SPARC for much broader scientific exploration and fully utilize its unique capabilities to push the frontier of physics understanding. If individual researchers are interested in pursuing exploratory physics work on SPARC, CFS encourages them to work with their government or institution to set up a larger agreement that could leverage the full benefits of a public–private partnership, including sharing SPARC data with researchers from publicly funded institutions. To fully exploit the capabilities of SPARC, it is likely that funding on the order of that provided to existing tokamak facilities would be necessary.

Finally, the SPARC team plans to release some select SPARC datasets to the public for any that wish to analyze it. This will likely be done for key SPARC results in the interest of transparency after researchers that collected the data have had the first opportunity to publish. These data would likely be hosted online via GitHub (similar to what has already been done for SPARC simulation data at https://github.com/cfs-energy/SPARCPublic) or in a similar fashion in another location online. In addition, the SPARC team plans to contribute data to multi-machine databases used to derive empirical relationships and validate simulations, such as those managed by the International Tokamak Physics Activity (ITPA). Over time, it is likely that more and more SPARC data will be made fully public.

The SPARC team believes that this set of data access policies balances the advantages of having a broader scientific team contribute to SPARC physics with the overhead that comes with increasingly open physics data sharing and the proprietary nature of the device.

To summarize, the SPARC tokamak is now under construction and when complete will be able to access plasma parameters that are unique compared to existing tokamaks. The facility will be equipped with a reasonable set of diagnostics and actuators, but has the capability for significant upgrades, even after DT operation. Early operation will focus on demonstrating Q > 1 and informing ARC design, with a particular focus on heat exhaust management, disruptions, and burning plasma physics. This overlaps with previously identified public science priorities.^24,25

SPARC data will be fully accessible to a team of long-term contributors at CFS, MIT, and other key partners. A larger set of collaborators will have access to SPARC data in exchange for in-kind contributions to the project, though CFS will be selective in choosing contributions that align with the SPARC and ARC missions. A larger agreement with an institution or government would likely unlock the full scientific potential of SPARC. Some specific sets of data may be made fully public for broader analysis and transparency. This data access program is very likely to evolve over time, both due to changes in SPARC priorities over time and in response to any changes in the publicly funded program. If researchers are interested in becoming more involved in SPARC research, they are encouraged to contact their funding agencies to discuss the possibility of larger institutional agreements, as well as to respond to possible future calls for CFS-funded work.

The SPARC team hopes to work collaboratively with the worldwide fusion community to advance toward a fusion power plant and toward a better understanding of the physical universe. Scientific results will be published to contribute to the broader physics understanding of tokamaks, and CFS will work with academic partners to find areas where the missions of the different organizations overlap. The team strongly believes that this type of collaborative work in public–private partnerships is the most effective way to simultaneously advance toward clean fusion energy and a better understanding of the universe.

SPARC is interested in and open to new forms of collaboration, including deeper public–private partnerships with the potential of more fully using SPARC as a part of the publicly funded program. This will be especially true after completing the initial SPARC mission of informing ARC design, though early operation also has the potential to contribute significantly toward the broad understanding of tokamak physics.

CFS is excited to engage in a new era of fusion research targeted toward near term power generation, while at the same time maintaining the highly collaborative environment that has prevailed in the community since some of the earliest successes in the field.

This work was funded by Commonwealth Fusion Systems.

The authors have no conflicts to disclose.

Alexander James Creely: Conceptualization (equal); Writing – original draft (equal). Dan Brunner: Conceptualization (supporting). Robert T. Mumgaard: Conceptualization (equal); Supervision (equal); Writing – review & editing (equal). Matthew Reinke: Conceptualization (supporting); Writing – review & editing (supporting). Michael Segal: Conceptualization (supporting). Brandon N. Sorbom: Conceptualization (supporting). Martin Greenwald: Conceptualization (equal); Writing – review & editing (equal).

Data sharing is not applicable to this article as no new data were created or analyzed in this study.