Fusion is one of the most ambitious technical feats attempted in human history. Realisation of what is a complex engineering system will require the integration of knowledge and hands-on expertise from several different fields. Integrating this experience and understanding for the realisation of fusion energy is at the core of Kyoto Fusioneering’s (KF) mission, which is to accelerate the development of high performance, commercially viable reactor technologies that will be required for a fusion power plant, specifically those associated with heating and current drive systems, power generation, and the tritium fuel cycle, to support the rapid expansion of the budding fusion industry. We do this by leveraging the collective centuries of subject matter expertise of our staff, comprising global fusion leaders in both industry and academia, and through connections to global experts and institutions.

KF, which was established in 2019 as a spin-out from Kyoto University, is therefore different by virtue of its approach from the majority of other privately funded fusion start-ups that have emerged in recent years, in that the company is focused on the development of technologies contrasting with and supporting those developing novel reactor configurations most of whom are, predominantly, focused on generating a net power-producing and cost-effective fusion plasma, by one of a range of approaches to plasma confinement. Broadly, KF’s technologies are those that will, when twinned to those fusion machines, provide a solution for the fuel cycle (tritium breeding, isotope processing, and tritium management) and be used to produce usable energy from fusion (energy conversion, via the tritium breeding blanket). To our knowledge, KF is the only company to specifically focus on solving engineering challenges for fusion power plants.

KF has three arms to its business. The first of these is the in-house research and development, aimed at advancing new technologies into future products that will be required for the success of the fusion industry in the coming decades. This feeds into the second arm, which is focused on core products, such as gyrotrons, which are required for fusion experiments today, and which require technological progression for a commercial fusion industry. The third is consultancy services that allow our scientists, engineers and innovators to apply their expertise for the benefit of the wider fusion ecosystem of public and private organisations who, collectively, are tackling various challenges in realising fusion energy. Ultimately, our goal is to consolidate these activities into an overarching knowledge, understanding and experience of fusion power plant engineering in line with our company mission. This will allow KF to work with fusion developers to bridge the gap, to couple that power-producing fusion plasma with a useful and commercialisable form of energy. Whilst the company is revenue generating, it is also funded by venture capital and government grants.

The scope of this review paper is to detail the major challenges to the establishment and scale-up of a fusion industry from the perspective of Kyoto Fusioneering, as will be detailed in Sect. “Challenges to Fusion Technology”. Although many challenges discussed are industry-wide issues, a view of the industry has been presented in this paper that highlights issues specific to the niche in which KF sits. Accordingly, in Sect. “KF’s Technologies & Solutions”, KF’s proposed solutions to these technological challenges are detailed and discussed. Finally, Sect. “Commercialisation” provides an objective assessment of the changes in the fusion ecosystem, including the barriers and opportunities in the commercialisation of fusion energy, and the mostly complementary roles of the public and private sector. This is considered an important inclusion for this special issue, as it details KF’s perspective on the now rapidly evolving fusion ecosystem and, specifically, what KF’s position is within it. Section 5 closes by summarising and discussing some of the near- to mid-term goals for KF.

Challenges to Fusion Technology

As Sect. “Introduction” highlights, fusion is a technologically complex undertaking. KF contends that this complexity makes the supporting systems of plant components critical path technologies. Whilst a breakthrough in plasma science realising a high net energy gain (the ITER project plans Qfusion = 10 [1, 2]) will be a key milestone in fusion, since it would demonstrate the energy-generating potential of fusion, it would be only one contributing factor in the quest to realise fusion as a commercially viable energy source. The vast range of technology that would be required for a fusion power plant, from fuel extraction to energy conversion systems, are so fundamental that their development cannot be an afterthought. Most of these technologies are currently at a low Technology Readiness Level (TRL) and cannot be developed serially; they must be developed in parallel with the work on fusion reactors to ensure that they are available when they are required.

Producing Fuel: Breeding and Handling Tritium

Tritium is a vital fuel for nearly all fusion reactor designs currently touted for commercialisation in the near future. However, there are only small amounts of tritium commercially available, and there are very real risks to supply [3,4,5]. Pearson et al. find that the further the schedules for D-T (deuterium–tritium) fusion experiments are delayed, for example with ITER, the greater the uncertainty of supply. This will affect the whole industry not just large-scale reactors and demonstrates why tritium breeding is key for a viable reactor. It is not reasonable to expect that commercial fusion can rely on external sources of tritium and remain a proposition that can be rolled out in any significant capacity on a global scale. Tritium breeding, the generation of additional fuel from a lithium-based blanket around the reactor core, is therefore expected to be ubiquitous amongst D-T commercial fusion designs.

Whilst many commercial plants have considered the role that breeding—and more broadly blankets—will play in their reactor designs, the blankets themselves, as well as the support systems for tritium extraction and management are mostly early-stage technologies, not yet ready for a fully scaled fusion power plant. Without the rapid development of these technologies, there is a risk that there will not be sufficient tritium made available via breeding for a large-scale rollout of fusion.

Depending on the concept, and whether the plasma is low density or high density, and the corresponding achieved fusion performance, a fusion reactor may only achieve a low fuel burnup, in some cases around 1% or less, and will have a high flow of exhaust gases, which will contain plasma enhancement gases, helium ash, and unburnt fuel [6]. It is important that the exhausted tritium is recycled into the reactor chamber and, as with the bred tritium, extraction and management systems are required. For both sources of tritium, specific components such as tritium-compatible pumps will also be required. A schematic of the fuel cycle, showing both the inner and outer fuel cycle systems, and how these systems and technologies are connected, is provided in Fig. 1.

Fig. 1
figure 1

Fuel cycle schematic for a candidate fusion power plant, based on [5]

Tritium management has been mentioned twice in this section and is important from two perspectives; safety and optimisation of the Tritium Breeding Ratio (TBR) [7]. Tritium is a radioactive gas that presents a safety risk, not only directly to humans, but also in the damage it causes as it permeates into materials. Tritium releases and its permeation into materials also both count towards losses from the overall inventory in the system and will decrease the overall effective TBR.

Energy Conversion

In a world of increasingly competitive renewable energy [8], a potential nuclear fission renaissance [9], and increasing interest in enablers like interconnectors and energy storage technologies [10, 11], the electricity generation will be a challenging sector to break into. By 2040 or so, the transition to greener electricity sources may be largely complete [12]. Whilst electricity generation may present an attractive first market for fusion [13], it could provide valuable forms of energy—in some cases uniquely—for deep decarbonisation outside of electricity generation. It is KF’s goal to enable fusion to expand into these markets.

Consideration of the conversion of energy into a commercially useful form should be paramount for organisations involved in the development of fusion. Despite the advancement of renewables, it is reasonable to assume that fusion can have some role in contributing to the electricity generation mix. As shown in Fig. 2, for electricity generating power cycles, higher turbine inlet temperatures in excess of 650 °C allow the use of Brayton cycles operating at higher efficiencies. However, achieving very high temperatures, in the region of 850 °C or perhaps higher, would give fusion the flexibility to tap into broader markets, such as process heat and hydrogen production, whilst simultaneously enabling higher-efficiency electricity production. KF argues that all of these markets will be key to the realisation of fusion’s commercial and decarbonisation potential.

Fig. 2
figure 2

Approximate energy conversion efficiency and power cycle applicability at different turbine inlet temperatures with curves taken from [14]. The possible efficiencies of a fusion reactor with various blanket concepts the SCYLLA© blanket is (ITER TBM, Li, FLiBe, SCYLLA©) as well as fission reactor concepts (PWR, AGR, HTR) are indicated

Heat generated must be extracted from the reactor itself. Whilst historically, primary focus in fusion has been around the tokamak-type reactor, there are many configurations of plasma confinement machine, and designs therein, currently being explored. Each envisages a slightly different approach to heat extraction and tritium breeding. Accordingly, a raft of breeding blanket concepts have been suggested to date, ranging from blankets using solid breeder materials cooled by gases (typically helium), liquids (chiefly water), as well as some advanced blankets that use liquid metal or molten salt blankets as the main heat carrier, with no need for additional coolant, i.e. they are self-cooling. The number of candidates for breeding blankets adds another level of complexity when scaling an experimental plasma demonstrator to a commercial fusion power plant, and the design decisions made have implications for both the operational parameters of the reactor as well as the energy conversion systems. If adopting one type of blanket coolant, for instance, and later wanting to pivot to another, may necessitate a different suite of technologies altogether in order to function. Multiple options should be considered in parallel and integrated in the demonstration plant stage.


A common thread running through many of the challenges associated with fusion is the lack of suitable materials that are required to withstand the fusion nuclear environment, whilst being commercially attractive, i.e. low-cost, manufacturable, and have ample in-service lifetime and reliability. The materials immediately surrounding the fusion reactor suffer an incredibly harsh environment in many respects, particularly with regards to the high levels of neutron irradiation (high flux and high energy), high magnetic fields, high temperatures and the corrosion from interaction with coolants. Other challenges include those related to handling at end of life (recycle or disposal), risk of radioactive contamination, and the ability to meet the exacting standards typical of any highly regulated industry, as is expected for fusion akin to the nuclear or aviation industries, for example [15]. A significant number of technological challenges in developing fusion could be solved by using, or indeed by newly developing, appropriate materials. Many novel material options have been considered, including steels that have been adapted to minimise the number of radioactive isotopes produced by neutron irradiation, so-called reduced activation ferritic/martensitic (RAFM) steels. Oxide dispersion-strengthened variants of ferritic steels are also in development and would offer improved resistance to irradiation damage and a higher maximum operating temperature. Novel alloy systems such as vanadium alloys are also being considered as fusion structural materials.

RAFM steels such as EUROFER97 and F82H are generally considered the most technologically mature of the existing candidate fusion structural materials [16, 17]. However, the most well-characterised RAFM variants also suffer from a relatively limited and low operating temperature range of around 350–550 °C [18]. Oxide Dispersion-Strengthened (ODS) steel can both widen and raise the operating temperature window, which would result in a more efficient fusion reactor [19]. Furthermore, recent advances in fabrication and machining mean that ODS steel could be used in flat blanket first walls [20]. Vanadium alloys, considered a less mature technology [21], can also offer improved high temperature performance relative to RAFM steels [19] but also suffer from high levels of tritium retention and are susceptible to low-temperature irradiation embrittlement [22].

Another candidate is silicon carbide fibre-reinforced silicon carbide matrix composites, hereafter referred to as SiCf/SiC. SiCf/SiC offers the highest maximum operating temperature of the aforementioned materials, and may also survive for longer in the fusion environment than other candidate materials. Indeed, SiCf/SiC has long been seen as an advanced candidate material for fusion due to its excellent mechanical properties at high temperatures [23], resistance to corrosion [24], low electrical conductivity, and promising activation properties [25]. For this reason, the material is finding strong utility in the fission industry, too. For fusion, however, SiCf/SiC is less well-characterised than RAFM steel and requires significant development before it can be used on the scale that commercial fusion will require. Greater understanding of performance following irradiation, corrosion resistance and suitability for different purposes is required. Several irradiation studies of SiCf/SiC have already been conducted [26, 27] but data on performance in a representative fusion environment (high neutron flux at energies up to 14.1 MeV) are not available. There is known to be a considerable degradation of thermal properties of SiCf/SiC after irradiation of the order of 1 dpa [23, 28]. However, SiCf/SiC should not be considered as only one material, and with several variants in existence and new ones under development, including at KF (see next section), further investigation and validation of different grades of SiCf/SiC is necessary. In addition, the processing and disposal of irradiated material remains a concern and approaches to end-of-life decommissioning have not been considered in any great detail. Despite this, SiCf/SiC has attractive activation properties relative to other candidate materials [25].

It is important to highlight the advancement and uptake of SiCf/SiC in other industries over the past two decades; SiCf/SiC is now used commercially in aerospace and is being trialled in the fission industry. This industrialisation of SiCf/SiC and resulting developments can also be taken advantage of by the fusion industry [29], as there are parallels to be drawn, particularly as regards manufacturability. It is for these reasons that KF is pursuing development of SiCf/SiC as a breakthrough material for fusion (see Sect. 3.1) [30].

A broader issue for the fusion industry relates to material scarcity, and availability or maturity of supply [31]. This is particularly problematic considering the long timescale before which commercial rollout is likely, which is probably not until after 2040. At this point, it is likely that the current supply to demand ratios of many materials, especially those that are already scarce at present will significantly increase in price and become harder to acquire. The demand for many materials is already much higher than current supply and this will only be exacerbated as supply diminishes. The fusion industry will have to contend with bottlenecks in supply caused by geopolitical conflict in production regions where there is a concentrated supply and competition from other industries, in particular renewables, and, in this regard, fusion is not immune from becoming like any other industrialised technology. Indeed, KF sees such matters as a factor in considering materials such as SiCf/SiC which, alongside potentially beneficial properties like high temperature and radiation tolerance, is made up of abundant elements compared with, for example, RAFM steels, which rely on the addition of vanadium and tungsten; both of which are sourced predominantly from China, and for which the required volumes of the materials to support a fleet of fusion plants vastly outweighs current supply [30]. Indeed, where maybe SiCf/SiC is beneficial in this regard, there are always trade-offs, and the development of the material comes with its own set of challenges, particularly regarding manufacturability.

Heating and Current Drive

Gyrotrons and neutral beam-injectors are the two prevailing technologies for Heating and Current Drive (HCD) in modern magnetic confinement-type fusion reactors (they are not required for inertial methods) [32]. Gyrotrons have some advantages over neutral beam injection systems. Firstly, their microwaves can be transmitted to the plasma through a series of waveguides and mirrors. This allows them to be housed some distance from the reactor, even in an adjacent area. Neutral beams cannot be transmitted in this way and must thus reside in close proximity to the reactor, occupying valuable space nearby and requiring a substantial number of ports to inject power into the reactor. Neutral beam injectors must also share the reactor’s vacuum, whereas gyrotrons need not. Additionally, the maintenance burden for neutral beams is considerably greater than for gyrotrons [33]. These latter two points present serious reliability concerns for neutral beams.

Gyrotrons, however, are not without challenges. To date, they remain an experimental technology. Their relative nascence and the considerable power demands from the vacuum tubes themselves, and to a lesser degree their cooling and ancillary systems means that they represent by far the largest parasitic load on a fusion power plant. The requirement that the heat generated by a fusion reactor be converted into electricity at efficiencies well below 50%, and that it is this electric power output on which the parasitic load of the heat and current drive system is placed, means that even modest improvements on gyrotron efficiency can have a substantial positive effect on the net power output of a fusion plant [34].

Gyrotrons, despite their relatively higher TRL when compared to many other fusion-related technologies, are a technology that requires further development to achieve longer pulse lengths approaching continuous operation, as well as a scale-up in manufacturing and repeatability, before they can be considered a commercial product. To date, gyrotrons remain a largely bespoke technology and must be designed for their application. Their frequency, for example, must be tuned according to the strength of the magnetic field of the reactor on which they will be used. In addition, vital ancillary components such as coolant systems and matching optics units must be designed for repeatability and built at scale. At current, typically, these are sourced from different suppliers.

A final drawback is the limited production capacity for gyrotrons globally. Estimates vary, and whilst figures are not made publicly available by the organisations in question, production is of the order of ten to twenty gyrotrons per year. This is a severe supply bottleneck. Experimental reactors require only a few gyrotrons, but demonstration plant designs envisage requiring up to a few hundreds of megawatts of power injection, and commercial scale plants might require even more. Given typical gyrotron power ratings of around a single megawatt, this implies several hundred gyrotrons will be required per commercial plant. As a consequence, ostensibly, a failure to scale up gyrotron performance and also manufacturing represents a major barrier to the rollout of commercial fusion reactors.

KF’s Technologies & Solutions

The plant technologies and systems that will make up a commercial fusion reactor are shown in Fig. 3. Those the KF has or intends to build as core competencies are highlighted in orange. In order to keep this discussion concise, only a few flagship technologies under development by KF will be placed centre focus in this paper. These are the primary loop systems (with a focus on blankets), the heat injection system (with a specific focus on gyrotrons), the ancillary fuel cycle systems, and the materials research that will underpin these technologies.

Fig. 3
figure 3

Conceptual diagram showing most plant components of a fusion reactor, and the level of involvement in KF’s Research and Development (R&D) activity: limited (grey), partial (light orange), or almost complete (orange). It should be noted that this diagram is not meant to represent the radial build of an actual reactor; it shows key systems and approximate position in relation to the plasma

Primary Loop Cooling and Tritium Breeding Systems

As detailed previously, operating at the highest possible temperature (to maximise reactor efficiency) presents enormous challenges, not least to materials selection and power management and handling in the design of a breeding blanket. Myriad coolant systems and approaches to tritium breeding have been proposed. Self-cooled breeding blankets have gained increased interest in recent years as being commercially attractive, for several reasons including the fact they avoid the need for a pressurised secondary coolant such as helium or water, and they can achieve higher operating temperatures than can typically be reached [35]. This need for high temperature, high efficiency and low cost inspired the development of the SCYLLA© (Self-Cooled Yuryo Lithium Lead Advanced) blanket concept by KF [30]. Self-cooled lithium lead blankets hold many advantages over other designs, including those that were chosen for the ITER Test Blanket Module (TBM) programme, for example, which, due to their lower operating temperature of 300–500 °C, convert thermal heat to power with a lower efficiency [36,37,38]. However, no self-cooled blanket has been developed past the conceptual design phase (e.g. [35]). Due to the perceived challenges in developing high-temperature self-cooled blankets, particularly in relation to the materials challenges, other lower-temperature blanket configurations are now significantly further along in development by comparison (here again referring to those selected for the ITER TBM programme). Self-cooled blankets may prove to be much simpler in design; a factor which is likely to improve commercial viability of the product in comparison to some other designs such as the complex helium-cooled pebble bed concept, for example, which uses two types of ceramic breeder, two different streams of helium (one for cooling, one for tritium extraction) in a complex array.

As the name suggests, the SCYLLA© design uses a lithium lead eutectic simultaneously as a breeder and coolant (and also as a neutron multiplier, facilitated by the lead atoms). Whilst the tritium breeding potential of lithium lead is lower than for pure lithium, for example, there are considerable materials compatibility advantages. The use of lithium lead as a breeder medium reduces the corrosion rate of SiCf/SiC [39], reactivity with air, and tritium solubility compared to pure lithium [40], albeit that it has a slightly higher melting point. Lithium lead designs typically have a lower TBR than pure lithium designs, yet with the self-cooled design of SCYLLA©, the high breeder volume fraction can lead to a sufficient TBR, which can further be increased by tuning the level of lithium-6 enrichment. It may also be plausible for the Pb:Li ratio to be continually adjusted during operation to control the TBR value (with the trade-off being that the eutectic temperature will fluctuate). This is important, as being able to control the quantity of tritium produced in a fusion reactor is a key challenge from the perspective of tritium handling, management and potentially regulation. Using lithium lead means that tritium can be extracted without an additional medium such as helium which decreases design complexity; although it should be noted that helium will still be produced in the blanket and therefore will still require separation.

SiCf/SiC is a key structural material used in the design of SCYLLA©, and offers specific advantages over metallic structural materials, as detailed in the previous section. How materials behave in a magnetic environment and combating the possible effects has to be considered—especially for blanket materials. SiCf/SiC provides partial electrical insulation which therefore reduces the magnetohydrodynamic (MHD) effects on the LiPb when it is flowing through the magnetic field. This is beneficial as higher pumping power is needed if the MHD effects cause a pressure drop in the LiPb flow [41]. The true effects will be able to be tested once KF’s UNITY project (discussed later in Sect. 3.2) has been completed.

KF is developing two different grades of SiCf/SiC for use in the SCYLLA© blanket. Using different grades of SiCf/SiC means its properties can be optimised to the requirements, solving some material compatibility issues. The grade of SiCf/SiC for the heat exchanger must be highly hermetic and have good thermal conductivity, but as it will not see neutrons, radiation damage is not a concern. By contrast, SiCf/SiC as a blanket material has different requirements. SiCf/SiC is compatible with LiPb, which has already been tested for temperatures in the range of 300–900 °C [42, 43]. The SCYLLA© design allows a thin-walled structure, which results in low neutron absorption loss compared with steel, which would further likely need a protective coating. This improves the ratio of breeder to structural material. However, the lack of data about SiCf/SiC’s performance under fusion specific conditions, and current inability to manufacture it in a cost-effective manner adds to the risk and uncertainty around its use. In particular, component manufacturing based on prepregs (impregnated fibre sheets that have not been fully cured and remain malleable) as used in the SCYLLA© design is not a commercial-off-the-shelf product and therefore does not have a well-established supply chain as other materials such as SS316 steel.

In addition to the R&D work on the SCYLLA© blanket, KF is advancing the readiness of other lithium-based coolants. FLiBe is a low-TRL but promising coolant option with the potential for a self-cooled blanket design. FLiBe provides strong tritium breeding (due to the presence of beryllium as an effective neutron multiplier), has good thermal performance, lessened chemical reactivity when compared with lithium, supports higher temperature operation of up to 700 °C for better power conversion efficiency (limited by the available FLiBe-compatible materials at current), and is not expected to be impacted by MHD effects as in liquid metals. On the other hand, there are issues surrounding the availability, supply and cost of beryllium resource that is a key constituent in making FLiBe an effective breeder [30]. KF has recently commissioned a test loop operating with a 1 kg FLiBe inventory (about 0.5 L) at temperatures up to 650 °C. The loop is designed for experiments on coolant purification, corrosion under flow conditions and tritium extraction (although it is expected that deuterium will be used as a proxy). FLiBe-compatible materials that could be used outside of the bio shield such as Inconel as well as ceramics and alloys with lower nickel content will be studied for use in the high-neutron flux environment.

Lithium is also being actively studied as a coolant for a self-cooled liquid breeder blanket. Its advantages include the possibility to make a design with sufficient breeding (high TBR) without the need for 6Li enrichment and with the possibility for relatively high temperature operation, if a compatible material can be found. Several designs have been created for a lithium-based blanket, envisaging operations at various temperatures [44,45,]–[46]. Vanadium alloy (V-4Cr-4Ti) is often used as the structural material of choice for these designs; however, its manufacturability has limited its advancement as a fusion material to date. Key challenges are its affinity to elements such as C, N, O [47] resulting, inter alia, in oxidation during manufacturing and operation and creep-induced rupture at temperatures above 700 °C [48]. To take the application of vanadium alloys as a blanket structural material further, a large high-temperature lithium loop (on the scale of UNITY introduced in Sect. 3.2) with vanadium alloy components is necessary but is estimated to be very costly.

Instead KF is testing other options for use as a structural material. KF has designed and is currently constructing a Li loop. The total volume of Li in the pipes is 0.7 L with a drain tank of 7 L capacity, corresponding to a maximum Li inventory of 3.7 kg. The structural material for the loop is 9Cr-1Mo ferritic steel and nickel-free steel (SS430) to enable higher temperatures than previously operated loops in support of the International Fusion Materials Irradiation Facility [49] or experimental fast breeder reactors of up to 600 °C. The designed flow rate is 10 L/min. The loop contains several test cells for testing corrosion under flowing conditions. This enables testing of material compatibility and sealing technology. The system includes hot trap sections where hydrogen and nitrogen are trapped to control the impurity levels during operation. The design process of the loop was completed in March 2022 and construction will be completed by early 2023.

Plans for Plant Component Testing in UNITY

KF is developing a number of plant components that will be required for a commercial reactor. These are shown in Fig. 3, and can be broadly split into two categories: thermal cycle and fuel cycle technologies. Thermal cycle technologies comprise the blanket and energy conversion systems and are discussed in Sect. 3.1. The vast majority of fuel cycle technologies are at relatively low TRL. Some are actively under development in public programs, including pumping technology for upcoming experiments, whilst others, such as detritiation systems for fusion were designed decades ago. Despite this, in many cases, knowledge of such systems has been (or is actively being) partly lost due to retirement of the involved engineers, which is in many cases being exacerbated by the delays in the commencement of the construction of tritium components for the ITER tritium systems, as one of the only fusion projects that will use power plant-relevant quantities of tritium.

At KF, tritium fuel cycle components are being developed for integration into a world-first fusion plant technology testing facility, named UNITY, which stands for Unique Integrated Testing Facility [50]. A conceptual diagram of UNITY and the general layout is shown in Fig. 4. KF has started construction of the facility in Japan in order to demonstrate the closed fuel cycle and power generation from blanket heat. Several components for both the fuel cycle and for the power loop have been tested at small scale at Kyoto University and more recently at Kyoto Fusioneering. With UNITY, those components will work together at a scale relevant to a commercial fusion reactor.

Fig. 4
figure 4

Artist’s impression of the UNITY facility showing the plasma simulator (top left), the plasma heating (lower left), main loop (middle) including the blanket test section (top middle), several fuel cycle components, and the control centre (right) [50]

Although there are some expected experimental outputs from 2023, UNITY is scheduled for completion in 2025 and will provide a new avenue for the completion of R&D work on critical challenges regarding materials such as SiCf/SiC and the MHD effects of LiPb under magnetic fields and the effects this has on pumping power requirements. Additionally, a small-scale SCYLLA© blanket module will be tested including all features from tritium extraction to electricity generation. The key focus areas of UNITY will be blanket/divertor integrated testing, gyrotron testing, pump testing, high temperature heat extraction, some aspects of tritium fuel cycle demonstration, and energy conversion. This is shown in Fig. 5. UNITY may also serve as a useful facility for component testing for the wider fusion community.

Fig. 5
figure 5

Design of the UNITY for breeding blanket and primary loop, including electricity generation and tritium recovery [50]

UNITY will be the world's first total integrated testing facility for fusion power plant equipment that would meet some of the need for R&D on integrated power cycles beyond lab scale. In doing this, UNITY will seek to help the fusion industry at large understand the practical challenges associated with coupling the multiple highly complex systems that will comprise a fusion power plant. Additionally, it will significantly help tackle the aforementioned challenges such as lack of performance data on SiCf/SiC, including KF’s own in-development grades. It will also provide a way to look at a way to scale production of components made from this material in a cost-effective way. By inducing magnetic fields on the loop, the effects of MHD and how it impacts other requirements such as pumping power will be studied. The eventual aim of UNITY will be to demonstrate all Kyoto Fusioneering products for the fusion industry and allow continuous testing and improvement of these products. The data arising from testing will provide a more holistic understanding of the challenges faced by fusion power plant designers.

As discussed in Sect. 2, tritium extraction systems are vital for the viability of commercial fusion. The batch processing of exhaust gases designed for ITER; which contain unburnt fuel, helium ash, plasma enhancing gases, and impurities; is not sufficient for the continuous operation of a commercial fusion reactor. Timescales for extraction, separation, and injection of deuterium and tritium must be low, in the order of a few hours or minutes. The time scale has direct influence on the TBR [5] and tritium inventory [51] requirements of the reactor, as can be seen in Fig. 6.

Fig. 6
figure 6

Influence of DIR on tritium inventory (at a fusion power of 2 GW and a burn-up fraction of 1%). Data from [51]

One way to achieve low time scales is a process called Direct Internal Recycling (DIR), in which hydrogen isotopes are directly removed from the gas pumped from the exhaust via a membrane free from heavier elements. Using this method, separation into a deuterium-rich and tritium-rich stream can be achieved, the DT fuel mixed to a 50:50 ratio and injected back into the vacuum vessel without going to the isotope separation system. Using DIR, the inventory requirements can be reduced by about a factor of 3 and it has been estimated that a demonstration plant will maintain about the same tritium inventory as the ITER plant [51]. UNITY offers a test bed that will allow KF to develop such technologies to the point where they are ready for commercialisation.

Figure 7 shows some of the pump systems currently available and under development at Kyoto Fusioneering. The lithium diffusion pump can be used for primary pumping, i.e. evacuating the vacuum chamber. The pump uses lithium as the working fluid, instead of mercury as considered for the European demonstration plant designs. This is important because in 2022 an amendment of the Minamata convention was passed that prohibits the sale and manufacture of mercury-based vacuum pumps [52]. Conventional oil-based (organic) diffusion pumps are also unviable due to the fact that tritium will react with the hydrocarbon-based fluids. KF is developing a roughing pump (KFRP) that is reciprocating, oil-free, and tritium-compatible and which can be employed in the hydrogen gas gun for pellet injection and for rough pumping at the exhaust. Separately, KF is developing a proton conductor pump (PCP) which provides a large area permeable membrane with a conductive field to extract hydrogen isotopes and separate D and T via the isotope effect.

Fig. 7
figure 7

Pump systems available by KF and use cases in the fuel cycle of a fusion reactor

It should be noted that the UNITY facility will not have sufficient testing capability to bring all components to commercial-off-the-shelf products. Likely a bigger next-step facility, which may include integration with a plasma source, will be needed after the completion of the current planned programme. Firstly, the size of UNITY is still a fraction of the size required for a full commercial plant (area of around 400 m2 compared to 420,000 m2 for the ITER platform) and while some components may more or less simply need scaling for mass production, others cannot be scaled easily, e.g. when the efficiency depends on the number of stages as in columns used for chemical separation. Furthermore, there are limitations in simulating conditions of a real fusion reactor. For tokamaks for example, one area of concern is simulating the energy deposited on the blanket by the plasma. In a reactor, the plasma heats the walls of the reactor via surface heat flux but also deposits energy as volumetric heating, which is created by neutrons interacting with the breeder material. Effects and influence on operation due to the presence of tritium and other sources are studied separately, outside of UNITY. Specifically, KF is exploring ways to tackle tritium-related aspects and testing including via collaboration with institutions in the US, the UK, Canada, the European Union and Japan.


As mentioned in Sect. 2.4, the gyrotron has significant advantages over alternative HCD systems. As a result, many experimental fusion reactors, including ITER, the Japanese reactor, JT-60SA, and several private sector developers pursuing a range of magnetic confinement devices (not only tokamaks) may use gyrotrons as their main, or even only source of HCD [53]. KF intends to support gyrotron development by focusing on two areas: improving gyrotron efficiency and applicability and developing the gyrotron itself into an off-the-shelf commercial offering [54].

In order to improve gyrotron efficiency, KF is researching a path to reducing power consumption whilst maintaining power output of millimetre wave radiation from the gyrotron output window. Figure 8 shows a simplified power loss diagram for the HCD and energy conversion systems in a hypothetical tokamak-based fusion plant. Here, a power conversion efficiency of around 30% is assumed (although higher efficiencies may be achievable as shown in Fig. 2) for a 3 GWth fusion reactor with a Q of 20.

Fig. 8
figure 8

Power losses of a generic 3 GWth tokamak-based fusion plant using gyrotrons for plasma heating

An efficiency of 50% in the HCD system from the 300 MW consumed at supply to the power injected into the plasma is assumed, which is optimistic given the current state of the art, but not unattainable [55]. This takes into account losses in the power supply, transmission line and in the gyrotron itself. A further 300 MW is assumed to be consumed by the parasitic loads for the whole plant, including cooling requirements (including for the HCD system) and reactor containment. In this case, the energy conversion losses and the amount of power consumed by the HCD system, only 400 MWe is output by the 3000 MWth reactor. This illustrates that HCD and plasma maintenance efficiency improvement is paramount to improving the overall efficiency of a commercial fusion plant. It should be noted that the power values shown in the figure are highly speculative and this is intended to be a representation of the problem only.

In order to create a product that is applicable to different reactor designs, KF is working on alternative approaches, such as step-tuneable gyrotrons, whilst further increasing stable frequency output to 236 GHz and beyond, if necessary for advanced concepts.

At the same time, KF is working on turning the gyrotron into a commercial product. Some aspects of this are tied to R&D work, such as the development of the step-tuneable gyrotrons. Other aspects are related to consolidating the supply chain for gyrotrons and associated ancillary components, thereby creating a commercial, ready-for-market product that does not need to be custom-designed for each new application. This would improve repeatability, streamline the manufacturing process, and allow for the scale up of global production, reducing a potential supply bottleneck. At the same time, KF is actively exploring alternative commercialisation opportunities for gyrotron technology, such as microwave drilling, beamed energy propulsion, nuclear magnetic resonance, and space solar power generation, and the company is fostering an open innovation philosophy [56].


A Paradigm Shift in Fusion: the Birth of A new Look Ecosystem

In the past, fusion development programmes were mostly the remit of government laboratories. Such programmes tended to focus on the front-end of innovation, i.e. on fusion science and technology R&D [57]. Here, following a so-called linear innovation model—where broadly first the science must be understood before the technology can be developed, and only when the technology is understood can it be developed into a product for market—public fusion programmes have indeed progressed our understanding of the science and technology. However, a predication on such a linear model, government laboratories, supported by public funds, are not focused on developing a commercial product. Moreover, they necessarily cannot tolerate a high risk of failure in their exploits to develop a technology and, again, are principally seeking the advancement of scientific understanding. Somewhat ironically, this linear model can lead large technology projects—a specific example in fusion is the ITER project—to suffer delays and cost overruns, to run the risk of programmatic failure, even where the technology risk is low due to the intensive R&D that has been carried out in support. The ITER project, still many years from completion, has not yet proven that its fusion physics and technology work in an integrated manner, which has pushed back the full design and build of next step demonstration machines.

In recent years, the fusion sector has undergone a fundamental change, which is principally due to the emergence of a raft of privately backed fusion developers. These developers are pursuing fusion reactor concepts that mostly originated from university programmes or laboratory projects. In contrast to public programmes operating on a linear model, these developers are aiming to demonstrate their technology as a product that is suitable for market as quickly as possible; they want to accelerate commercialisation of fusion energy. With significant but limited funding—and a commitment to return investment to their shareholders—they must necessarily proceed with risk [57]. To fulfil this, they are embracing uncertainty, and accepting a higher level of risk, with the hope of progressing faster. Their approach evokes a statement made by the Founder of Intel, Gordon Moore, who said “it’s not science becomes technology becomes products. It’s technology that gets science to come along behind it” in reference to a model of innovation that is not linear, and focuses on engineering and products, not on science.

For private fusion companies, it is an acceptance of risk that is allowing them to proceed with development of ideas despite often having lower scientific maturity. For this approach to work, an agile philosophy is needed: a focus on getting to a minimum viable product as quickly as possible [57,58,]–[59]. This means not getting bogged down in seeking a full understanding of the science. It means no late changes, no cost overruns, and no (or limited) mission scope creep. Fusion companies operating on this model thrive on quick decision-making, a reduced number of internal processes and the layers of bureaucracy that are often necessary in larger organisations. Indeed, underpinning the approach in general is a different mindset whereby failure is considered to be essential for innovation.

Whilst privately funded fusion developers offer an exciting and different approach to innovation, they cannot exist in a vacuum. Government funded fusion programs tend to have stricter requirements whereby it is more difficult to explore high-risk, novel technology opportunities beyond fundamental research (where it may be possible to develop intellectual property). However, in such programs there is a world-class pedigree of knowledge and scientific capability, as well as a range of experimental facilities. Today, several public–private partnerships for accelerated fusion development have been commissioned with the goal to combine and coalesce the best of the fast-moving agile-natured private companies with the technological might of government laboratories and innovative universities. A new, collaborative environment is being forged. This coming together is key to developing fusion within a timeframe that will help the energy crisis. An awareness of the need for industrialisation has increased the sharing of knowledge and lessons learnt which in turn is giving rise to what will become the fusion industry ecosystem. In short, governments are pivoting their fusion strategies to allow for joint working between public and private sectors, and the private sector wants to get the best people and organisations working to support their ideas, so is seeking to leverage government backed leviathans such as national laboratories and specialist institutions such as universities.

All of this is important for KF’s mission for the following reasons. Firstly, KF is a company backed by private venture capital and thus adopts the same agile innovation approach as other fusion developers, only KF’s focus is on developing reactor system products, not plasma machines. Secondly, KF is a company trying to accelerate fusion energy for commercial and societal impact; which is ostensibly the same vision as all other fusion developers. Finally, KF is a company that sits squarely between most private sector developers focused on developing a power-producing fusion plasma machine and publicly funded institutions that are sitting on decades of R&D into fusion technology, which KF is now keen to leverage, spin-out, and develop into a product for market.

KF’s Strategy

As noted, the vision of KF is the same as other private sector fusion developers: to develop fusion as a practical energy source, on a timescale relevant to make an impact on current global problems. However, KF’s mission—and business model—is different. KF is focused on developing components related to a power plant, rather than producing power-producing fusion reactions, and thus KF’s challenges are not associated with achieving a net power-producing fusion reaction itself (plasma). It is therefore in KF’s interest to support other developers, public and private, in achieving their missions. Indeed, to fulfil KF’s mission, the company must also focus on building a successful and stable business that will be viable for the company to first establish, then grow, its role in the fusion ecosystem. It must have a suitable business model to achieve this.

An example of developing a capability (and technology) that allows KF as a company to establish its business model—and at the same time establish itself as a key player in the fusion industry—is with its gyrotrons. As detailed earlier, gyrotrons are well understood as a technology, but to turn the technology into a high-volume, commercial product means several challenges must be tackled. In short, as an R&D device, the way that gyrotrons are currently developed, including the cost of manufacturing and the performance they deliver, is sufficient. For large scale deployment of magnetic confinement fusion, however, perhaps hundreds of gyrotrons will be needed per reactor. Developing for manufacturing in large volumes, whilst still ensuring high precision to meet the specification for fusion applications, remains an open challenge and is something that KF must tackle.

As detailed in Sect. Introduction, KF has three main arms of its business, which are broadly: internal R&D, service provision, and consultancy. This split simultaneously uses the knowledge and experience of scientists, combined with the development of individual components, whilst interfacing directly with “the market”, i.e. fusion developers. This gives KF the ability to better focus its technology development decisions on the needs of the industry, and to make improvements that are commercially oriented. As a nascent industry, much of fusion development also remains in the remit of scientists and engineers who have been, until now, focused on R&D mostly for publicly funded fusion programmes like ITER. So in addition to working directly with customers to understand their requirements, forging collaborations with world-leading universities and laboratories to help transfer the technology for commercial application represents a key part of KF’s strategy to help industrialise fusion. In this way, KF fulfils a separate niche to other fusion developers, and is able to contribute and aid in solving the challenges encountered by the industry as it expands.

In the next decade or so the fusion industry will naturally progress and change. Whilst developing technology to realise fusion remains the priority—the company contains “fusioneering” in the name, after all—any high-tech company must also consider open innovation opportunities (see previous section). Companies working in advanced fusion R&D have some clear opportunities to expand their technology to be utilised in other industries or markets, such as the space sector or the broader advanced energy sector. KF is considering the use of gyrotrons for mining applications for deep geothermal energy. Another synergy is in high-temperature engineering, where advanced fusion technology R&D may open new avenues; operating at high-temperature will require novel materials, use of advanced unconventional fluids (liquid metals and salts) and necessitate new, advanced energy conversion technologies, all of which could be transformative for non-electric energy applications in general. KF’s blanket technologies may be relevant here. Such applications are highly important due to the need for deep decarbonisation, which is not simple to achieve through the greening of electrification alone. Specifically, a technology like KF’s SCYLLA© blanket may yield breakthroughs that can be later applied to a problem that has nothing to do with fusion, but may be equally societally or commercially useful.

In order for any commercial fusion organisation to grow and survive in what is a nascent industry, the performance of technologies – and products – must be continually improved as that industry progresses. For example, at KF, several innovative design changes for gyrotrons are already in progress, with key focus on attaining higher frequency, higher efficiency and lower cost. Additionally, R&D improvements are being made to allow testing of longer duration operation as this will be necessary in future commercial prototypes. In the tritium area, KF is one of the only companies in the world focused on realising a self-sufficient fuel cycle for tritium. It is, for example, developing advanced components for air detritiation systems and pumps, and will make improvements via testing in UNITY as well as on separate development programmes. For both these cases, however, the requirements of the technology must be considered to be driven by the needs of others in the fusion ecosystem; experimental programmes or technology demonstrators. Accordingly, the need for well-established component supply chains will become increasingly important. Here, KF aims to establish itself in this supply chain. Ostensibly the status quo in fusion is that whilst there are some existing, viable solutions to the problems ahead, many are at very low TRL. This means they are either not yet proven, not yet technologically advanced enough or not yet commercially viable. This is the status of the majority of technology in the fusion industry and KF considers tackling this challenge core to its business; one which, if it can be solved, will give the company a competitive advantage in what is a budding new sector.

Summary and Next Steps

As highlighted at the beginning of this paper, there are many challenges that the fusion industry has yet to tackle. These range from tritium breeding to structural material choices, both of which greatly affect the design of the whole fusion plant. Throughout the industry, solving these challenges is the key focus, with different companies trying to solve them in different ways. Whilst this paper does not cover every challenge that the industry is facing, it provides an overview of the challenges and opportunities ahead for the development of several major systems that are critical in realising a functioning, commercially attractive fusion power plant. These are highlighted as being, broadly: heating and current drive via gyrotrons, fusion-grade materials, tritium breeding, tritium handling, and energy conversion. Kyoto Fusioneering is tackling many of the issues by focusing on developing high performance reactor technologies linked to the fuel cycle and power generation, and is developing R&D capability to test them. To this end, KF considers that component integration is equally as important as individual component design. UNITY will soon provide useful data to support fusion development in this area. By focusing on plant technology as its central focus, Kyoto Fusioneering is able to develop components that will be key to fusion becoming commercially viable as well as providing testing data that can also be used to further progress the activities in development at other private fusion companies. As mentioned, this separates KF from other companies that are mostly focused on developing a plasma confinement machine, which itself comes with a range of challenges that require dedicated focus. This leaves a gap whereby few others are actively focused on bridging the gap between experimental success and a viable commercial plant, a gap KF intends to fill with the technologies it is developing.

The goal of this paper is to underline key challenges ahead on the path to developing fusion energy specifically from Kyoto Fusioneering’s perspective, as well as to demonstrate the fundamental ideas behind KF's technology. The paper links both of these topics and illustrates how the company sits within the fusion ecosystem and how it is helping to move fusion closer towards industrialisation. The next steps for the company will be to focus on continuing to collaborate with other fusion companies and strengthening relationships in order to further establish Kyoto Fusioneering’s place in the budding fusion ecosystem. Completing the build of UNITY and starting to test and produce data under specific fusion conditions will be vital and hence is a key focus within the business. KF will continue to publish data and more detailed information about its specific technologies. From an organisational point of view, in late 2022, KF expanded its business into the US in order to facilitate further technology and knowledge exchange globally, as well as continuing to grow and upskill its workforce. Eventually, the aim is to have functionally diversified R&D facilities at each base of the company. These facilities will benefit from having the knowledge of lessons learnt from the build of UNITY, and as such, these goals are further in the future of KF’s development.