Compact spherical tokamak

Towards a compact spherical tokamak fusion pilot plant

A. E. Costley

The question of size of a tokamak fusion reactor is central to current fusion research especially with the large device, ITER, under construction and even larger DEMO reactors under initial engineering design. In this paper, the question of size is addressed initially from a physics perspective. It is shown that in addition to size, field and plasma shape are important too, and shape can be a significant factor. For a spherical tokamak (ST), the elongated shape leads to significant reductions in major radius and/or field for comparable fusion performance. Further, it is shown that when the density limit is taken into account, the relationship between fusion power and fusion gain is almost independent of size, implying that relatively small, high performance reactors should be possible. In order to realize a small, high performance fusion module based on the ST, feasible solutions to several key technical challenges must be developed. These are identified and possible design solutions outlined. The results of the physics, technical and engineering studies are integrated using the Tokamak Energy system code, and the results of a scoping study are reviewed. The results indicate that a relatively small ST using high temperature superconductor magnets should be feasible and may provide an alternative, possibly faster, ‘small modular’ route to fusion power.

This article is part of a discussion meeting issue ‘Fusion energy using tokamaks: can development be accelerated?’.


Research with tokamaks has been ongoing for more than 50 years and for most of the time it has generally been considered that in order to generate net fusion power tokamak fusion reaktors will have to be large and powerful; major radius of ≥6 m, plasma volume ≥ 1000 m³, and operation with fusion power ≥1 GW, typically being considered necessary. The large scale ITER device currently under construction in France is the latest device in this line of approach, and designs of even larger and more powerful demonstration (DEMO) reactors are underway.

Recent work, however, has shown that an approach based on much smaller and lower power devices may be possible. The approach is based on a re-evaluation of the empirical scaling of energy confinement time with machine parameters such as size and field, and the adoption of a relatively new technology, high temperature superconductors (HTSs) for magnets. The shape of the plasma is important too. The work indicates that much smaller devices based on the spherical tokamak (ST) configuration, perhaps with a major radius of 1.5-2.0 m, volume of 50-100 m, volume of 50-100 m³ and operating at relatively low power levels, 100-200 MW, may be feasible. Smaller devices would open the possibility of a modular approach to fusion power; that is one where single or multiple relatively small, low power devices would be used together to achieve the required power. Smaller and less expensive fusion modules would enable faster development cycles and thereby speed up the realization of fusion power.

Fig1: Schematic of conventional and spherical tokamaks. The aspect ratio A=R0/a and the elongation κ=b/a.

Spherical tokamaks have a much smaller ratio of plasma major radius (R0) to plasma minor radius (a) than conventional tokamaks such as JET and ITER; they resemble the shape of a cored apple rather than the more conventional tokamak shape of a doughnut (figure 1). Research has shown that STs have beneficial properties from a reactor standpoint such as operation at high plasma pressure relative to the pressure of the confining magnetic field, and the generation of higher levels of self-driven current within the plasma. This aspect is especially important. Auxiliary current drive systems are inefficient and thus can lead to substantial amounts of re-circulating power, and that power could be a major drain on the potential economics of a fusion reactor. There are also indications that STs have higher levels of energy confinement relative to conventional shaped tokamaks. STs share many of the challenges experienced in the development of the larger devices, for example the handling of the plasma exhaust in the divertor region where the power loads will be at the limit of available materials, and the installation of shielding on the inboard side necessary to protect the central column from the intense neutron and gamma radiation. The technical solutions being developed for the larger devices can be adapted and used on STs. The positive performance characteristics combined with potential solutions to the technical problems make STs particularly attractive for the compact approach.

In this paper, the work that is ongoing to realize this alternative approach to fusion power is reviewed. First, the question of size is addressed in general terms from a physics perspective. It is shown that it is not just size that is important; magnetic field and shape are important too, and the interplay between these parameters is developed. The question of fusion power is also important since that determines loads on the internal tokamak components and will limit the minimum possible device size. As shown in previous papers, the two key reactor performance parameters, i.e. the fusion gain, which is the power produced divided by the input power, and the fusion power are found to be directly linked. Using the latest empirical scalings for the energy confinement time, it is shown that the power needed for a useful fusion gain is three to four times lower that previously thought necessary. Taken together these findings indicate that smaller fusion devices based on the spherical tokamak should be feasible.

The realization of a relatively small, low power fusion module will depend on satisfactory solutions being developed to several key technical challenges such as the superconducting magnets that provide the plasma confining magnetic field, the inner shield that protects the central column from neutron and gamma radiation that potentially could cause material damage, and the divertor that handles the plasma exhaust.

The fusion triple product

The most important figure of merit of a fusion plasma is the product of the density (n), temperature (T) and energy confinement time (τE), nTτE. This is known as the fusion triple product and is derived from the work of John Lawson in 1957. For net fusion power, nTτE must be greater than 1×1021 m-3keVs. The progress towards fusion can be measured with nTτE.

Fig2: Improvement of the fusion tripel product (relative units) with time.

Figure 2 shows how the triple product has increased with time as larger tokamaks operating at higher magnetic field and higher plasma current were brought into operation. As can be seen, the rate of progress was very rapid from the late 1960s through to about 2000 but has slowed since, partly because of delay with ITER. Insight into key aspects of achieving net fusion power with tokamaks can be gained by looking closer at the fusion triple product.

The density and temperature are straightforward parameters but the energy confinement time is complicated. The energy confinement time characterizes the rate at which heat is transported from the hot central core of the plasma to the relatively cold surrounding material surfaces.  Within a tokamak plasma there are multiple, interacting phenomena occurring simultaneously on a wide range of temporal and spatial scales. These interactions lead to the transport of heat through processes that are essentially turbulent. While graet progress has been made in understanding these processes it is not yet possible to determine the transport of heat through the plasma by a first principles approach. This is not an unfamiliar situation. In many areas of physics and engineering situations are too complex for a ‘first principle’ approach. In such situations, it is common to perform experiments on devices or structures of different scale and to determine how the parameter  of interest scales with device parameters.