Development of steady-state fusion reactor by Helical Fusion

Helical Fusion Co., Ltd. in Japan was established at the end of 2021 aiming for the implementation of a commercial helical fusion reactor for base load electric power to society, and Helical Fusion USA, Inc. is a wholly owned subsidiary of Helical Fusion Co., Ltd. (jointly referred to as the HF group). The HF group is developing a commercial helical fusion reactor based on the design studies of the force free helical reactor (FFHR) series,¹ which have similar magnetic configurations to that of the large helical device (LHD),² incorporating the latest technologies of high-temperature superconducting (HTS) magnets and liquid blankets. Here, we choose the LHD-type heliotron configuration generated by two continuous helical coils because of the capability for steady-state long pulse operation, high density limit, and large apertures for maintenance ports between magnet coils. LHD is the only facility in the world to date that has achieved ion temperature higher than 10 keV,³ electron density higher than 1 × 10²¹ m⁻³,⁴ and plasma duration time longer than 3000 s,⁵ although these were achieved individually. These achievements are comparable to or better than those of large tokamaks. In particular, the operable density limit and plasma duration time are far superior to those of tokamaks with similar device size and magnetic field strength.⁶ The necessity of plasma current is the fundamental difference between tokamak and helical. It is not necessary to drive a plasma current inside the helical plasmas since the nested magnetic surfaces confining the high-temperature plasma are completely generated by external magnet coils. Helical plasma is quite stable and does not require sophisticated plasma position control as in tokamaks, even in plasma startup and/or shut-down phases. Helical plasma is free from harmful plasma current disruptions.

The first-of-a-kind (FOAK) fusion power plant envisioned by the HF group is a 100 MWe-class steady-state helical fusion reactor. The electricity output is relatively small compared with conventional fusion reactor designs. This is one of the distinguished points of the HF group's vision. The small-output fusion reactor can be commercialized not only as a small-scale distributed power source but also as a large-scale centralized power source by combining multiple reactors together. The small-output fusion reactor is applicable as, for example, a local power source for factories such as aluminum or titanium smelters, remote islands, or large ships. The construction cost of the FOAK power plant is estimated to be around USD 5 billion. After more than one year of continuous operation, maintenance is to be completed within 3 months to achieve a high availability of more than 80%.

Before proceeding to the construction of the FOAK power plant, a fusion pilot plant (FPP) is planned to be constructed and operated to demonstrate steady-state net-electricity generation. At present, two options named FFHR-HESTIA (or simply, HESTIA, HElical reactor with the surface flow of TIn alloy) and FFHR-SOARHER (or SOARHER, Superconducting Optimally ARranged HElical Reactor) are being considered as candidates for the FPP. HESTIA adopts the modular-type Liquid Metal (LM) blanket⁷ and flexible HTS WISE (Wound and Impregnated Stacked Elastic tapes) conductors for magnet coils.⁸ The first wall of the LM blanket including the divertor strike zone is covered and protected by LM free-surface flow, and therefore, individual divertor systems are not required in HESTIA. Electron cyclotron heating (ECH) is adopted for plasma heating. Neutral beam injection (NBI) and ion cyclotron heating (ICH) are not required. By skipping the development of individual divertor systems, NBI, and ICH, which are widely recognized as the most difficult issues in fusion reactor development, a fast-track plan is possible in the case of HESTIA. On the other hand, SOARHER explores a more conservative development strategy by employing either the molten-salt¹ or ceramic pebble blanket,⁹ the tungsten divertor,^10,11 and rigid HTS STARS (Stacked Tapes Assembled in Rigid Structure) conductors for magnet coils.^11,12 Both HESTIA and SOARHER have a device size twice that of LHD, a magnetic field strength of ∼3 times that of LHD, and a maximum electrical output of 50 MWe.

In this manuscript, we would focus on HESTIA. Section II provides an overview of HESTIA. Section III describes the development plan for HESTIA. Discussions on strategies to secure fusion fuels are given in Sec. IV. Finally, these are summarized in Sec. V. While the development plan discussed here provides a rough sketch of a commercial helical fusion reactor, they are only the envisioned goals that must ultimately be achieved to demonstrate the economic feasibility of fusion. Our current focus is on technical feasibility, i.e., showing that a fusion power plant can indeed produce net electricity.

In the case of the LHD, m = 10, $ℓ$ = 2, γ_c is variable in the range of 1.13–1.33 by changing the current ratio in the three helical coil layers, and α = 0.1. In HESTIA-Primary and HESTIA, m and $ℓ$ are the same as LHD, but γ_c = 1.20 and α = 0.0 are adopted as the first candidate of magnetic configuration, because of the larger blanket space and the better confinement property.^13,14

HESTIA is a deuterium–tritium (DT) fusion reactor, where tritium is self-produced using liquid metal (LM) blanket systems. The maximum net electrical output is 70 MWe, with the following core plasma parameters: central ion temperature, T_i0, of ∼8 keV, central ion density, n_i0, of ∼4 × 10²⁰ m⁻³, central plasma beta, β₀ (ratio of central plasma pressure to magnetic pressure), of ∼3%, and fusion power, P_fusion, of ∼260 MW. ECH is supposed to be applied with the heating power of P_ECH = 20 MW, and therefore Q ∼13. The HTS magnet coils require a current density, j_c, of ∼85 A/mm². The stored magnetic energy of the HTS magnet system, W_mag, is 66.2 GJ. These and other design parameters of HESTIA are summarized in Table I, together with the reference parameters obtained in LHD and those expected for a prototype device named HESTIA-Primary and a FOAK power plant after HESTIA, which will be discussed in Sec. III.

In Table I, R₀ and a₀ are the major radius and the minor radius of the magnetic axis in vacuum, respectively, B₀ is the magnetic field strength at R₀, $− ı$_2/3 is the rotational transform $− ı$ = ι/(2π) = 1/q (q is the safety factor in tokamaks) at the normalized minor radius ρ = r/a = 2/3, where a is the toroidally averaged plasma minor radius in the magnetic equilibrium determined by the plasma beta profile, I_c is the helical coil current, n_e0 is the electron density at the plasma center, T_e0 is the electron temperature at the plasma center, (P_dep/P_dep1)_avg is the line-averaged value of the volume-integrated heat deposition profile, P_dep(ρ), normalized by the total heating power, P_dep1 = P_dep(1), γ_DPE is the confinement improvement factor given by ((P_dep/P_dep1)_avg,reactor/(P_dep/P_dep1)_avg,exp)^0.6 (Ref. 15), H is the additional confinement improvement factor expected in the optimized magnetic field configuration, τ_E is the energy confinement time, τ_E^ISS95 and τ_E^ISS04v3 are the energy confinement times predicted by empirical scaling laws of ISS95¹⁶ and ISS04 with the renormalization factor f_ren = 0.93,¹⁷ respectively, n_edge is the edge electron density at ρ = 1, n_Sudo is the Sudo density limit,^18,19 Γ_FW,avg (MW/m²) is the averaged neutron heat flux on the blanket first wall, Γ_div,avg (MW/m²) is the averaged divertor heat flux, P_NBI is the NBI heating power used in the reference discharge, f_α is the contamination ratio of helium ash, η_α is the ratio of the confined alpha particle, η_EG is the conversion efficiency of electricity generation from the thermal energy absorbed by the blanket, η_ECH is the conversion efficiency from the wall-plug electricity to P_ECH, η_cryo is the conversion efficiency from the wall-plug electricity to the cryogenic power for HTS coils, P_net is the net electric generation, P_gross is the gross electric generation, Q_eng is the engineering fusion gain defined by the ratio of P_net to total electric power for reactor operation, C_direct is the direct cost for construction other than land, cite preparation, and licensing, and τ_life is the reactor life at full power operation. Note that the C_direct in Table I is estimated by our original cost evaluation model. This cost model is based as much as possible on the actual performance of LHD and ITER at the end of the 1990s, but it does not take into account inflation and other factors over the past 20 years. In fact, prices in Japan have hardly changed at all over the past 20 years. On the other hand, prices have more than doubled over the past 20 years in other countries. Therefore, if this inflation is taken into account, C_direct must be modified by a factor of 2 or more. The parameters of HESTIA-Primary, HESTIA, and the FOAK power plant in Table I are estimated using the systems code HELIOSCOPE²⁰ and the integrated transport analysis code TASK3D.²¹ 

To realize HESTIA and the FOAK power plant, however, we need to overcome the technology gaps that remained in the categories of (1) reactor core plasma design, including enhancements in energy confinement in heliotron-type stellarators, (2) HTS magnet, including more flexible HTS conductor suitable for construction of complex coil shapes, (3) LM blanket, including LM that can provide a single solution for both plasma exhaust and tritium breeding, non-ferromagnetic structural metals, and efficient LM circulation pumps, (4) ECH, including high-power, high-frequency gyrotrons, (5) fueling and evacuation, including high-frequency, high-speed pellet injectors, and direct recycling system, and (6) electricity and H₂ generation, needed for a novel startup strategy including self-generation of startup tritium using the electricity generated from green H₂.

Detailed discussions on reactor core plasma design, HTS magnet, LM blanket, ECH, fueling and evacuation, electricity generation, and H₂ generation are given in the Subsections II A–II F.

In stellarator/heliotron devices, including LHD and W7-X,²² the parameter dependence of energy confinement time is known to follow the predictions of a physical model called the gyro-Bohm model, which is also true for tokamaks.²³ In the gyro-Bohm model, the energy confinement time is proportional to the 0.6th power of the density, the 0.8th power of the magnetic field strength, and so on. Miyazawa et al. have developed the DPE (direct profile extrapolation) method, which extrapolates the experimentally observed plasma density and temperature profiles in the minor radius direction to the reactor condition using the gyro-Bohm property.^24,25 This enables highly accurate prediction of fusion power in fusion reactor core plasmas.^13,26

The Sudo density limit¹⁸ for the case of HESTIA is also shown in Fig. 2(a). Note that n_Sudo corresponds to the edge density limit.¹⁹ Although n_edge exceeds n_Sudo in HESTIA, as in Table I, the difference seems to be not critical in Fig. 2(a). If this difference between n_edge and n_Sudo appears to be critical, it will become necessary to add an off-axis ECH to increase n_Sudo without largely modifying the beta profile. It should be noted that the beta profile changes if the heating power is increased, and the magnetic axis moves outward due to the Shafranov shift. In the LHD-type heliotron configuration, the outward shift of the magnetic axis results in the increase in the neoclassical transport. Therefore, the off-axis ECH that has a weaker impact on the beta profile compared with the on-axis ECH will be favorable.

In designing the core plasma parameters of HESTIA shown in Table I, we assumed a modest confinement improvement factor of H = 1.3, together with the heating profile effect of ECH and the alpha heating profile localized at the plasma center. It has been observed in LHD that the energy confinement apparently improves in the cases with central heating compared with the cases with broad or hollow heating profiles. This heating profile effect has been already formulated and incorporated into the DPE method using (P_dep/P_dep1)_avg.¹⁵ As can be seen in Fig. 2(e), the heat deposition profile in HESTIA is extremely peaked at the plasma center. This results in a large (P_dep/P_dep1)_avg of 0.710, which is larger than that of the reference data of 0.536. The confinement improvement due to this center-peaked heating profile is taken into account using a factor of γ_DPE = (0.710/0.536)^0.6 ∼1.18. Although the extremely center-peaked heating profile has a possibility to trigger the formation of an internal transport barrier (ITB) that is not taken into account at this moment, because the physics scenario of ITB formation is still unknown.

Instead of assuming confinement improvement due to the ITB formation, we are now trying to find out the optimized magnetic configuration to achieve an additional confinement improvement of H = 1.3, other than γ_DPE, by slightly deforming the helical coils from those of LHD.²⁷ It is known that the inward-shifted magnetic configuration with R₀ ∼ 3.6 m in LHD conforms to a σ-optimized field,²⁸ where the magnetic field strength at helical ripple bottoms is similar along a field line.^29,30 However, there is still ample potential for optimizing the magnetic configuration of LHD further,¹⁴ because we have gained experimental and theoretical knowledge of plasma physics and have established numerical analysis tools during the 25 years of LHD operation. We are now searching for an optimized magnetic configuration that can achieve high MHD stability and reduced neoclassical transport at the same time. Although it is expected that the neoclassical heat flux of Q_tot^neoS can be compensated with P_tot as shown in Fig. 2(f), it is marginal at ρ ∼ 0.5. Therefore, it is favorable to reduce neoclassical transport further to secure a margin for turbulent transport. If the neoclassical transport cannot be reduced and the turbulent transport is not negligible at ρ ∼ 0.5, additional heating power should be added inside ρ < 0.5 to compensate for the turbulent heat flux. The off-axis ECH without largely modifying the beta profile will be favorable for this, as was explained already.

As for the high-energy particle confinement, we assume η_α = 85%, according to the simulation results in the former study,¹³ in which it is shown that the ratio of confined alpha particles in the case of the magnetic configuration similar to HESTIA is larger than 85% at ρ < 0.7, as long as alpha particle loss boundary is set on the blanket first wall, and the reentering effect is taken into account. Note that p_α at ρ > 0.7 is negligibly small as shown in Fig. 2(e). It is expected that the high-energy particle confinement property can be also improved in the optimized magnetic configuration with reduced neoclassical transport.

Our reactor core plasma design does not rely on the non-thermal ion population. Instead, we plan to use only electron heating by ECH and alpha heating. The so-called energy equipartition by Coulomb collision between electrons and ions will take place, and the ion temperature is expected to become similar to the electron temperature. HESTIA and the FOAK power plant will be operated at high density as shown in Table I, which promotes electron-ion collisions. The equipartition effect is calculated using a theoretical model and taken into account in our reactor core plasma design.³¹ 

Although LHD is the world's first large superconducting helical device and our helical reactor design largely relies on the knowledge accumulated in LHD, we need further R&D. Especially, there is a large difference that LTS (low-temperature superconducting) magnet coils cooled to 4 K by liquid helium that are adopted in LHD, while HTS magnet coils cooled to 20 K by helium gas that are adopted in our helical reactors. There are several reasons why HTS magnets must be used. First, the critical magnetic field is higher with HTS magnets than with LTS magnets. Second, the coil current density can be higher than that of LTS magnets. A high coil current density is essential to ensure a large blanket space. Third, the amount of helium required as a coolant can be significantly reduced compared to the LTS magnets. In recent years, the global supply of helium has been depleted. In our helical reactors, gas helium, of which the density at 20 K is 1/50 of liquid helium at 4 K, will be used to cool superconducting magnets to address this problem. The 20 K helium gas will be produced using a helium refrigerator, and the inlet temperature could be precisely kept at 20 K using the heat exchange with liquid hydrogen that is generated by solar power as will be discussed in Subsection II F. A cooling efficiency of 2% is assumed for the cryogenic system for HTS magnets in HESTIA. The cooling efficiency is supposed to be increased from 1.5% to 2% in the step-by-step development from the prototype to HESTIA.

The maximum magnetic field strength on the magnet coils is expected to increase from 6 T in LHD to ∼20 T in HESTIA. The high current density of 85 A/mm² is also required for the HTS conductor of HESTIA, which is higher than that of LHD of 34 A/mm² at the maximum. The HF group is developing a new HTS WISE conductor⁸ using stacked REBCO tapes. The WISE conductor is quite flexible to be wound to fabricate the HTS coils with complicated shapes, such as the helical coils. After the winding procedure is completed, the volume and the WISE conductor inside the coil body are impregnated by a low-melting point alloy to fix and reinforce the WISE conductor. It is supposed to fabricate the HTS helical coils for HESTIA without insulation. Such a no-insulation HTS coil is expected to have a high resilience to the quench accident, although the time constant of the coil increases.^8,32 If a normal transition takes place locally in the no-insulation HTS coil, the coil current flowing the normal part is automatically bypassed to the neighboring HTS tapes which are still superconducting, and further progress of normal transition can be avoided.

A schematic view of the LM blanket in HESTIA is shown in Fig. 3(a). The LM blanket is composed of nine main modules as shown in Fig. 3(b) and one test blanket module at each of ten similar sections.⁷ Our LM blanket is designed aiming at easy construction and maintenance. Although 90 blanket modules in total are used in HESTIA or the FOAK power plant, it is possible to replace them within 1 month, if three or four modules are replaced in one day. Because fusion reactors have high radiation levels immediately after the operation, access to the inside of the reactor is difficult. All LM blanket cartridges in HESTIA are designed to be maintained from the upper port using a three-point suspension crane. It should be noted that no pipe cutting or welding is necessary for our LM blanket.⁷ These eliminate the need for a robot to enter the reactor during maintenance. After 1 year or more of continuous operation, a maintenance period of about 3 months is supposed. Then, the availability is larger than 80%. The maintenance period consists of 1 month for decreasing the magnetic field and for cooling the radioactivity and tritium degassing of blanket materials, 1 month for inspection and replacement of the blanket as a necessity, and 1 month for increasing the magnetic field and for preheating of blanket modules for the next operation.

The blanket first wall is surely one of the key components in magnetic fusion reactors. It is reported from the long pulse discharge experiment in LHD that the particle recycling changes with a slow time constant of a few tens of minutes.³³ The slow variation of particle recycling can cause difficulty in density control in a fusion reactor. In HESTIA, this problem is solved by adopting the LM first wall, i.e., the first wall of our blanket is covered and protected by the LM free-surface flow. Since the first wall is always refreshed by flowing LM, the hydrogen recycling condition is expected to be kept stable. To implement the LM first wall, the plasma-facing sides of each blanket module are composed of porous material. The LM leaks out from the porous material and forms an LM free-surface flow on the first wall.

Since the free surface of the LM is exposed to the plasma, the vapor pressure of the LM must be low enough not to affect the plasma performance. Tin is one of the low melting point materials and has a very low vapor pressure.³⁴ Therefore, we plan to develop a tin–lead–lithium alloy LM breeder material for fusion reactor blankets, which is composed mainly of tin mixed with lead for neutron multiplication and lithium for tritium production. Here, lead can be substituted by other materials if only it results in a better tritium breeding ratio (TBR) without largely affecting the LM characteristics such as the melting point. Since the hydrogen retention in the pure tin is as low as ∼0.1 at.%,^35–39 hydrogen isotopes including the DT molecules hitting the LM and T atoms generated inside the LM will be immediately emitted to the vacuum vessel (VV) and exhausted by the cryopumps. Hydrogen retention in the tin-lead-lithium alloy, which may differ from pure tin, will be investigated in future studies. Tin is highly corrosive to steel and therefore special consideration for corrosion protection is required for structural materials and the porous material for the blanket of HESTIA. Research on corrosion protection is also considered in our research plan.

Contamination of the main plasma by heavy impurities such as tin and lead should be avoided. At this point, we expect the impurity shielding effect of the ergodic layer surrounding the last-closed-flux-surface (LCFS) in the LHD-type heliotron configuration. In LHD, it has been observed that the friction forces inside the ergodic layer that push the impurity ions in the direction from LCFS to divertor become dominant over the thermal forces that work vice versa.⁴⁰ It was also reported that this impurity shielding effect of the ergodic layer is enhanced at higher density and for heavier impurities.

On the other hand, we are also planning to develop new structural material. Although RAFM (reduced activation ferritic martensitic) steels are widely adopted in the fusion community as the structural material for fusion reactors, they are ferromagnetic. In magnetic confinement fusion devices, the use of ferromagnetic materials that produce error fields should be avoided. Therefore, we are going to develop a new non-magnetic material. One of the possible candidates is a non-magnetic reduced activation high-manganese steel. In the past, high-manganese steel was considered for use in fusion reactors, but the first choice was given to RAFM because the high-manganese steel has problems of large decay heat and swelling caused by neutron irradiation.⁴¹ These weaknesses of high-manganese steel, however, can be greatly mitigated in low-power fusion reactors with fusion power of several tens of MW, such as the HESTIA and the FOAK power plant we are envisioning.

Additive manufacturing or other 3D printing method using pure titanium, titanium alloy, or the new high-manganese steel will be adopted to fabricate the porous material equipped on the blanket's first wall. In the case of titanium porous material, its surface will be already covered by a strong titanium oxide layer that will be effective for corrosion protection. In the case of high-manganese porous material, it will be necessary to form the anti-corrosion oxide layer on its surface. In both cases, the first wall side of the porous material should have a high wettability to form the LM free-surface flow. Research on wettability is also planned in our research plan.

Independent divertor devices are not needed in HESTIA or the FOAK power plant of the HESTIA-type, because the divertor strike zones of the blanket first wall are already covered and protected by the LM surface flow. No special heat exchange equipment for the divertor is needed either since all plasma energy reaching the divertor is absorbed by the LM of the blanket. The averaged divertor heat load in HESTIA is expected to be ∼4 MW/m². This can be easily managed by the LM free-surface flow while keeping the metal vapor pressure low enough. These factors have contributed greatly to the simplification of the reactor system and the reduction of the development program, construction time, and total cost.

We also need a reliable LM circulation system. Because the mechanical pump with high-rotation speed has low reliability, and the electromagnetic pump has only poor driving efficiency, we will newly develop an original reciprocating pump driven by pressurized gas as shown in Fig. 3(c). This system is composed of multiple reservoir tanks containing LM. High-pressure gas is alternately introduced to each tank to push the LM out to HESTIA. The LM is introduced to blanket modules through the top of the VV. The LM is heated inside the blanket modules and exhausted to the lower pool. The high-temperature LM then flows into the heat exchanger and finally returns to the reservoir tanks again. It should be noted that the electric power needed for the new LM circulation system is quite unknown at this moment. In Table I, the electricity for LM circulation is estimated assuming mechanical pumps. We are expecting that our new LM circulation system can be operated with smaller electricity compared with mechanical or electromagnetic pumps. The new LM circulation system will be applied to the prototype and then to HESTIA.

ECH is adopted in HESTIA as the only auxiliary heating method to enable a year-long continuous operation without maintenance. As shown in Table I, P_ECH = 20 MW is required in HESTIA. The resonance frequency of ECH in HESTIA and the FOAK power plant is as high as 250 GHz, due to the high magnetic field of around 9 T at the plasma center. It becomes difficult to design a gyrotron as the oscillation frequency increases. In LHD, 154 GHz gyrotrons were developed and applied in the experiment. The 154 GHz gyrotron could put 1 MW of ECH power for only a short period of shorter than 2 s and the output power becomes limited to lower than 0.5 MW at continuous operation. No gyrotron working at 250 GHz with 1 MW of output power at CW (Continuous Wave) is available at this moment. Therefore, we will develop a 250 GHz-1MW-CW gyrotron by ourselves. Although it is designed to operate continuously at 1 MW, we will prepare 60 gyrotrons of 250 GHz-1MW-CW for HESTIA to ease the requirements for the gyrotron. These gyrotrons will be operated alternately to inject 20 MW of ECH power continuously. The 60 gyrotrons are evenly distributed at ten outer ports of VV, six at a port, as shown in Fig. 4. ECH is vertically injected from the outside of the horizontal cross section of the helical plasma.

The 250 GHz-1MW-CW gyrotrons we develop would have a high wall-plug efficiency of 50%, i.e., 40 MW of the wall-plug electricity is required to inject 20 MW of ECH power into HESTIA. Although the wall-plug efficiency of ECH in LHD is around 30%, it is possible to increase the efficiency to larger than 50% depending on the design of the gyrotron.

As for the fueling apparatus, we will develop 30-barrel pipe-gun DT-ice Pellet Injectors (PIs) together with a direct fuel gas recycling system.^42,43 It will be required in HESTIA to inject high-speed DT pellets at a few Hz of injection frequency. To obtain high-speed pellets, the pipe-gun pellet injector is the best candidate among the various pellet injection systems. On the other hand, the difficulty of pipe-gun pellet injectors is that it requires at least one minute to form a pellet in the pipe. This is the reason why we need to develop 30-barrel PIs. We will set ten PIs at all ten outer ports of HESTIA as shown in Fig. 5, i.e., 300 barrels in total. If a pellet is formed within 100 s, then we can inject pellets at an injection frequency higher than 3 Hz. In LHD, a 20-barrel PIs has been developed and used in the experiment. We are aiming to increase the barrel number from 20 to 30.

The PIs are to be coupled with the direct fuel gas recycling system, where the DT gas exhausted from the VV is directly supplied to the pellet injectors after being purified and pressurized in the DT gas conditioning system. By this, it becomes possible to suppress the tritium site inventory. Multiple cryopumps as shown in Figs. 1(b) and 5 will be mainly used for vacuum pumping in HESTIA. Cryopumps are accumulating vacuum pumps and must be periodically regenerated to expel the accumulated hydrogen isotopes. Steady-state operation is achieved by alternately repeating pumping and regeneration operations using multiple cryopumps.

In HESTIA, a supercritical CO₂ (SCO₂) gas turbine generator⁴⁴ will be adopted to achieve an electricity generation efficiency higher than 50%. The working temperature will be 800–1200 K. High energy conversion efficiency and relatively small size compared with ordinary steam turbines are the merits of SCO₂ gas turbines. We will develop an SCO₂ gas turbine power generation facility for HESTIA.

Large electricity is required at the startup phase of a fusion reactor, to cool the superconducting magnets and to preheat the blanket modules. We will use H₂ produced by solar power and an H₂ gas turbine generator to provide this startup power. We will also develop a solar-powered H₂ production facility and an H₂ gas turbine power generation facility. These systems can be also applied to HESTIA-Primary which is planned to operate without DT fusion reaction or power generation. The volume of H₂ can be reduced by ∼1/730 times by liquifying the H₂ gas. Furthermore, the cold heat of the 20 K liquid H₂ is available to stabilize the inlet temperature of the HTS magnets.

HESTIA is supposed to operate in a steady state for 1 year. As shown in Table I, the fusion output of HESTIA is 260 MW. The gross electric power generated from this fusion output will be ∼140 MW. Subtracting a circulating electric power required for reactor operation, the net electricity of 70 MW can be supplied to the power grid or electricity consumers. The majority of the circulating electric power is used in the cryogenic system for HTS magnet cooling and the ECH system for plasma heating.

As discussed in the Subsection II A, the HF group will adopt an optimized magnetic configuration. To verify the improved plasma performance, it is planned to construct and operate a prototype reactor named HESTIA-Primary. It is possible to examine if the energy confinement is improved or not, without the DT fusion reaction. Therefore, HESTIA-Primary will be operated in the non-nuclear condition, where only H₂ gas is used without D or T. If the parameters listed in Table I for HESTIA-Primary, which are estimated using the DPE method with the confinement improvement factor of 1.3, are achieved, that can be strong evidence showing that the improved energy confinement is achieved in the optimized magnetic configuration. Then, we will apply these results to HESTIA, and the parameters listed in Table I for HESTIA that satisfy the FPP requirements will surely be achieved with high prediction accuracy.

The HF group aims to demonstrate the functionality of the key elements, such as the HTS magnet and LM blanket, by the early 2020s, and to integrate them into HESTIA-Primary, which is expected to be constructed within roughly 3 years and to begin operation by the end of the 2020s. This is a fairly difficult schedule, but not necessarily an impossible one, given the budget and human resources. Of course, if the necessary budget cannot be secured, the schedule will be pushed back. HESTIA-Primary will be about 1/3 the size of HESTIA to minimize construction costs as low as USD 480 million. As was discussed in Sec. II, this cost evaluation is based on the prices at the end of the 1990s, and if inflation during these 20 years is taken into account, the construction cost must be modified by a factor of 2 or more. The size of HESTIA-Primary is large enough to simultaneously demonstrate the functionality of the key elements and the high plasma performance. The construction and operation of HESTIA-Primary as a prototype before HESTIA can reduce the development risk.

After the successful operation of HESTIA-Primary, HESTIA will be constructed. The construction period is expected to be 5–6 years. Therefore, in the fastest case, we can start the operation of HESTIA by the early 2030s. In HESTIA, a net electricity generation of 50 MWe at the maximum and a steady-state operation for more than 1 year will be demonstrated. Thereafter, maintenance will be completed within 3 months, and the next operation will be resumed to demonstrate the availability of more than 80%. Our FOAK fusion plant assumes net electrical output of 100 MWe in a device of the same size as HESTIA. The thermal/neutron load handling of the in-vessel components and the energy extraction from the reactor core can be realized by similar systems as HESTIA. On the other hand, the neutron flux to the HTS coils will increase. Therefore, it is necessary to improve the neutron shielding performance by, for example, expanding the blanket space by further optimization of the magnetic field configuration and/or developing efficient neutron shielding materials, to achieve a sufficient plant lifetime from the viewpoint of plant economy. As listed in Table I, the central beta of the plasma needs to be increased from 3.4% in HESTIA to 5.4% in the FOAK power plant to increase the fusion output from 260 to 340 MW and the net electricity from 70 to 103 MW. A high central beta larger than 5% must be demonstrated in HESTIA even for a short time of, for example, 3 h, which is short enough not to influence the HESTIA's life significantly.

Based on the knowledge obtained in HESTIA and further progresses in the magnetic configuration optimization and material studies, the FOAK power plants of 100 MWe net electricity will become available by the late 2030s to the early 2040s.

HESTIA is a DT fusion reactor and strategies to secure the initial tritium inventory at startup and enriched ⁶Li supplies for reactor operation are quite important. Our strategies for these issues include a so-called Deuterium–Deuterium (DD) startup operation and a precise 3D neutron transport simulation.

The initial tritium inventory is generated by the HESTIA itself by the DD startup operation, where the DD fusion reaction is used to produce fast neutrons to produce tritium in the blanket via a nuclear spallation reaction of ⁶Li atoms by 2.45 MeV DD neutron. The ratio of D at the very initial stage of the DD startup for the first operation of HESTIA is 100%. The ratio of T gradually increases with reactor operation. It is possible to shorten the period of DD startup operation by mixing the tritium atoms produced in the blanket into the fuel gas. After the DT ratio reaches 50:50, the excess T is separated, recovered, and stored for the second and subsequent operations of HESTIA or for the FOAK power plant. A cryogenic distillation system similar to that of ITER⁴⁵ will be adopted for tritium separation. Although detailed and accurate estimations of the DD startup scenario will be performed in the very early phase of our development plan, it is roughly estimated that a few months of DD operation is enough to produce the tritium inventory needed for a few hours of DT operation.

The use of isotope-enriched lithium is known to be effective to increase the TBR of the blanket in a fusion reactor. Also in our case, the isotope-enriched lithium that includes 80 at. % ⁶Li and 20 at. % ⁷Li is supposed. The accurate enrichment factor of ⁶Li from the natural isotope ratio of ∼8 at. % ⁶Li and ∼92 at. % ⁷Li will be determined at the initial stage of our project by the 3D neutron transport simulation using an accurate 3D-CAD model of HESTIA.

The FPP concept we would propose is a 50 MWe-class steady-state helical fusion reactor named HESTIA. The major radius of helical coils is ∼8 m, and the magnetic field strength at the plasma center is ∼9 T. HESTIA is a DT fusion reactor, where tritium is self-produced using LM blanket systems. Since plasma current drive power is unnecessary, HESTIA can be operated at a low fusion gain of ∼13, and steady-state operation is principally possible on the order of a year. The construction cost of HESTIA is estimated to be USD 5 billion. The availability will be larger than 80%.

HF group has started research and development on magnetic configuration optimization, HTS conductor, LM blanket equipped with the LM first wall, 250 GHz-1MW-CW gyrotrons, 30-burrel pipe-gun DT-ice pellet injector, SCO₂ gas turbine generator, solar-powered H₂ generation system, H₂ gas turbine generator, and HTS magnet cooling system using liquid H₂. These will be integrated into a prototype device named HESTIA-Primary. After the integrated technology demonstration by HESTIA-Primary, HESTIA will be constructed to demonstrate 70 MWe of net electricity generation by the fusion reaction. A long pulse operation for more than 1 year and a demonstration of high availability of > 80% will be also conducted in HESTIA. Finally, in the fastest case, the FOAK power plants of 100 MWe net electricity will become available by the late 2030s.

The parameters, technological issues, and schedules discussed here are our targets that must ultimately be achieved to demonstrate the economic feasibility of fusion. Our current focus is to demonstrate the technical feasibility that a fusion power plant can indeed produce net electricity.

The author is grateful to members of the HF group (T. Taguchi, N. Yanagi, T. Shimozuma, A. Nishimura, and A. Sagara) and collaborators in the United States (D. Andruczyk (Univeristy of Illinois), [A. Khodak and R. Maingi (PPPL), C. Kessel, S. Smolentsev, B. Pint, L. Baylor (ORNL), and R. Yamaguchi (PARC)] and in Japan (Y. Sugawara (Aoyama Gakuin University), M. Sakama (Tokushima University), R. Kasada (Tohoku University), Y. Narushima, Y. Hamaji, and G. Motojima (NIFS).

The authors have no conflicts to disclose.

Junichi Miyazawa: Conceptualization (lead); Data curation (equal); Formal analysis (equal); Funding acquisition (lead); Investigation (lead); Methodology (equal); Project administration (lead); Resources (equal); Software (supporting); Supervision (lead); Validation (equal); Visualization (equal); Writing – original draft (lead); Writing – review & editing (lead). Takuya Goto: Data curation (equal); Formal analysis (equal).

The data that support the findings of this study are available within the article.