A System Perspective on Fusion Power for Electricity Generation

Fusion as a power source

The nuclear fusion of hydrogen is the process that creates the Sun’s energy.  Extreme pressure and temperature causes hydrogen atom nuclei to merge together into helium atom nuclei, accompanied by the release of copious amounts of energy.  Human creation of nuclear fusion was first achieved in so-called hydrogen bombs in the early 1950s.  Since that time, many people have promoted using fusion as a long-term solution for generating electrical power, while simultaneously getting away from using fossil fuels.  The latter (coal, oil, and natural gas) are nonrenewable and have serious detrimental effects on the environment.  Fusion is seen as having an inexhaustible source of fuel and would avoid many negative aspects of generating power by nuclear fission.  (In fission, heavy radioactive atoms such as uranium are split into two lighter atoms with the consequent release of energy).  The promise of fusion as a power source has spurred the investment of many billions of dollars in science and technology efforts around the world over the last six plus decades.

Recent advances

Recently an advance in hydrogen fusion has gotten widespread attention.  The National Ignition Facility (NIF), located at Lawrence Livermore National Laboratory in California, announced that a test on December 5, 2022 achieved controlled hydrogen fusion for the first time where more energy came out than went in.  This was described as a milestone in fusion power research; all prior efforts have required more input energy than was obtained as output.

Inertial containment fusion

However, this accomplishment needs to be put in perspective.  The NIF’s approach to fusion is based on inertial containment of the material to be fused.  In the NIF, a few nanosecond-long burst of light from an array of 192 extremely powerful lasers is directed onto a target about the size of the eraser on a Number 2 pencil. The target encloses a small quantity of deuterium and tritium isotopes of hydrogen.  The intense light pulse raises the pressure and temperature of the material in the target to a level sufficient to cause the fuel to fuse, in a chain reaction process termed ignition.  The process is discontinuous; each event requires providing a new target and charging the array of lasers, and substantial time elapses between events.

Inertial confinement fusion (ICF) techniques have been studied since 1957 and early on were regarded as a promising approach to harnessing fusion.  In addition to the NIF, a number of other ICF research facilities have been built around the world, but none are as large and powerful as the Livermore facility.  However, rather than being oriented towards researching the generation of power, the focus of ICF systems research has switched.  Now they are used for supporting nuclear weapons design and maintenance through studying the behavior of matter under the conditions found within nuclear explosions.

Magnetic containment fusion

The major alternative to ICF is magnetic confinement fusion (MCF).  In this approach, a plasma made up of deuterium and tritium hydrogen isotopes to be fused is confined and heated by intense magnetic fields produced by superconducting magnetic coils.  The coils are typically arranged around a very large doughnut-shaped reaction chamber.  Many facilities around the world are working on MCF approaches.  The most ambitious one is the International Thermonuclear Experimental Reactor (ITER), being built in France by a consortium that includes China, the European Union, the United Kingdom, India, Japan, South Korea, Russia, and the United States, with several additional partners.  It is the largest of more than 100 MCF experimental facilities existing or now being built around the world.  ITER has been described as the most expensive science experiment of all time (estimated costs at completion are between $45 billion and $65 billion) and the most complicated engineering project in human history.  It is not expected to be operational for a number of more years.

ITER is not intended to be a prototype of a power-producing system.  Its stated purpose is scientific research and technological demonstration of a large fusion reactor, without electricity generation.  It is planned that facilities based on a successor design, known as DEMOnstration Power Plants, will be constructed by several of the ITER consortium members to experiment with technology for actual grid power production.  DEMO units will need to be somewhat larger than ITER in order to produce net fusion power.

Concerns about the practicality of using fusion power for the electrical grid

Considering the results obtained to date, I have a number of concerns about fusion reactors to produce power for the electrical grid:

• At the most basic level, fusion reactors are extremely complex heat sources for steam turbines to drive the generators that actually produce the electrical power output. The basic technology of turbine-generator systems is well over a century old.
• Fusion reactors are likely to have an extremely high capital cost per kilowatt hour of energy generated (for example, in comparison with the cost of a large wind turbine installation or a large solar array).
• Fusion reactor designs do not appear to scale well to create larger and smaller systems.
• Fusion reactors are expected to be very finicky, where all the parameters have to be just right for the system to achieve net power production at all, let alone run at optimum performance.
• Fusion power systems are likely to require a very highly skilled technical workforce. All the reactor subsystems have extremely high technology elements to be managed.
• The fraction of time that a fusion reactor will be online and operating is hard to predict, but there are indications it might be relatively low. This will make it harder to use a fusion reactor as a primary source for power generation, as it may be offline a substantial fraction of the time.
• It is important to consider fusion power systems from the standpoint of net energy production over their entire life, including energy used in their construction, fuel creation, operation, dismantling, and materials recycling. Also, what non-renewable resources will be consumed in building and operating them?

Additional considerations

Following are some additional considerations to take into account when considering the prospect of large-scale fusion power systems:

• MCF reactors are likely to require very large amounts of scarce and expensive nonrenewable materials (e.g., for the wiring in the superconducting magnet coils).
• MCF reactors are described as being inherently safe against accidents like the meltdown of a fission reactor. In particular, the amount of material in the plasma is quite small.  However, consider the possibility of their massive superconducting coils reverting to normal electrical resistance in the event of a cooling system failure—that’s a huge amount of energy to dump instantaneously.
• Deuterium is plentiful and relatively easy to obtain from seawater, but tritium is extremely hard to come by and has a short half-life due to radioactive decay. Today tritium is mostly produced in specialized fission reactors.  If the isotope is to be produced in the fusion reactor itself, the reactor will need to be shut down periodically to harvest the tritium from the walls of the containment vessel.
• Magnetic confinement of the plasma in a MCF reactor only confines particles that are electrically charged. The neutrons produced by the fusion process cannot be confined and will impinge on the containment vessel walls and interior systems, resulting in radiation damage.  The life of interior components of a fusion reactor may be relatively low due to this effect.  While the damaged materials are going to be relatively low-level radioactive waste, it will be hard to recycle them for any other use.
• Finally, MCF reactors are not free of nuclear proliferation concerns (they could be used to create plutonium and tritium for nuclear weapons).

Some conclusions

My view is that celebrations for the dawn of an age of fusion-produced electrical power are premature, at the very least.  I am skeptical of the appropriateness of investing vast sums in additional R&D on fusion power at this time.  Other innovative means for generating, storing, and distributing electrical power appear to be a better investment choice for our society, as they will result in usable systems much sooner.  Fusion research should be continued, but perhaps at a much more modest level than at present.