Fusion taking a road less travelled
Vincent Chan
Editor-in-Chief, Journal of Fusion Energy
Most modern technology development follows a strategy that utilizes the principle of “economy of scale”. The concept is first demonstrated in a table-top device, followed by experiments of larger sizes leading to commercialization. Each step is carefully evaluated in order to move to the next step. This approach minimizes the startup costs and reduces the risks when more significant investment is required for the later stages. The strategy is highly attractive for private sector investment. Due to an important physical reason, magnetic fusion energy development cannot follow the “economy of scale” approach. Fusion reaction in the laboratory requires a temperature that is nearly ten times that of the sun! Even when the plasma, which is a mixture of deuterium and tritium fuel, is well-confined by the magnetic field, it requires a certain minimum size for the temperature at the center of the plasma to cool down sufficiently when in contact with a material wall. This defines the required size to construct a first prototype reactor. The cost of construction is roughly proportional to this size. Based on present technology, the cost of investment is too high for private sector initiatives. Furthermore, very few countries have the resources or the urgency to take this on although the potential of fusion as a clean source of energy is well recognized. For this reason, government investment in fusion energy had been heavily tilted towards scientific research for many decades, and significant progress has been made including reaching the temperature, density and confinement time required for fusion burn.
The turning point occurred at the turn of the century when seven international partners decided to join forces to construct the International Thermonuclear Experimental Reactor (ITER). Almost twenty years later, ITER is on the verge of operation [D. J. Campbell, J. Fusion Energ (2019), https://doi.org/10.1007/s10894-018-0187-9]. The success of ITER design and technology development has encouraged several countries to start planning for a test reactor, which is the step before the commercialization of fusion energy. They include the EU DEMO project [A. J. Donné, J. Fusion Energ (2019), https://doi.org/10.1007/s10894-019-00223-7], and the China Fusion Engineering Test Reactor [J. Li, J. Fusion Energ (2019), https://doi.org/10.1007/s10894-018-0165-2].
This road less travelled taken by fusion also has an attractive feature, namely once the first test reactor is demonstrated, innovations in science and technology can reduce the size of the subsequent reactors. One such technology is the development of high temperature superconducting magnets (HTS). According to Martin Greenwald, Deputy Director of MIT Plasma Science and Fusion Center, increasing the magnetic field from 6T to 9T, which is only accessible by HTS, can reduce the size of a test reactor by 50%, and with the same fusion gain. A recent announcement in MIT News [MIT News Office, Sep 8, 2021, https://news.mit.edu/2021/MIT-CFS-major-advance-toward-fusion-energy-0908] says that a full-scale magnet designed and built by Commonwealth Fusion Systems and MIT has demonstrated a record-breaking 20T, which bodes well for this approach. Another approach of operating at high β, defined as the ratio of plasma pressure to magnetic pressure, can also lead to a compact fusion reactor. These innovations have stimulated many private enterprises to enter into fusion development. They include Commonwealth Fusion Systems, Tokamak Energy, General Fusion, HyperJet Fusion and Lockheed Martin Skunk Works among others. We hope to hear a lot more from these companies in the coming years.
Another advantage with constructing a large size first prototype fusion reactor is that with the synergistic development of advanced diagnostics systems, a large scale database that truly reflects the complex interactions of an integrated power plant would be available. With the advent of big data, a new branch of fusion research utilizing machine learning (ML) to process research data has emerged [D. Humphreys, J. Fusion Energ (2020), https://doi.org/10.1007/s10894-020-00258-1]. ML will be first applied to ITER and then to the first test reactor. Beyond that, one could anticipate using ML to incorporate innovations to reactor design thereby speeding up the timescale and reducing the development cost to realize fusion energy.
Dr. Vincent Chan was the director of Theory and Computational Science in the Energy Group of General Atomics, USA until his retirement in 2014. Since 2014, he is an international visiting professor at the University of Science and Technology of China, where he serves as thesis advisor to doctoral students. His research interests include plasma heating and transport, and integrated modeling of burning plasma systems.
