Star power: Can nuclear fusion energy live up to high hopes?

Is nuclear fusion energy nearing its Kitty Hawk moment?

Fusion energy would produce electricity by fusing atoms — the process that powers the sun — rather than splitting atoms like today’s nuclear energy. After decades of slow, and some say quixotic, research, there were big announcements over the last year and renewed optimism about fusion’s viability.

In September 2020, a team affiliated with MIT and the start-up Commonwealth Fusion Systems published peer-reviewed papers describing a new reactor design. Unlike current fusion reactors in labs, Commonwealth’s model would be the first to reach the milestone of net energy gain — a reaction that produces more energy than is required to create it. The start-up aims to have a working prototype by 2025.

Commonwealth is just one of many projects racing to take fusion from the lab to the market.

With this momentum, in February 2021, the National Academies of Sciences, Engineering, and Medicine issued a consensus report endorsing a pilot program to create a reactor for commercialization.

By some estimates, humanity will need up to five times more energy by 2100 to support a population of 10+ billion and enable rising living standards for people in poverty. All that while addressing climate change and other environmental hazards — from ambient air pollution to indoor smog to nuclear waste and waste from renewables — that require transformations in energy systems.

Proponents say fusion can have a starring role in meeting these challenges. A fusion energy reactor would emit no CO2 to operate — only producing steam. It would create less radioactive waste than conventional nuclear and cannot have meltdowns. It could run on fuel derived from virtually limitless seawater while taking up relatively little land. Hence, many articles refer to fusion as the “Holy Grail” of energy — hard to get but marvelous once in hand.

How realistic is this fusion optimism?

Below I unpack the challenges and opportunities for fusion.

Overall, fusion is not quite a Holy Grail — there are still significant barriers to a practical reactor and more trade-offs than is sometimes reported. But there is progress in fusion research and a good chance we will see a net energy reaction by 2030. From there, pockets of commercial fusion can slowly emerge in some markets as the technology economizes.

By 2050, fusion could fill a niche in our energy portfolio, alongside other technologies, such as renewables and advanced nuclear fission, that move us toward an abundant and sustainable future.

Where is fusion at today?

There’s an old saying: Fusion is the future of energy… and it always will be.

Since early research with the Manhattan Project, there was hope fusion could become an energy source in 30–40 years — only for that timeframe to elapse and elapse again.

Producing fusion is relatively easy, but the brief reactions created in labs require immense energy input because artificial fusion is working uphill.

Scientists condense and heat plasma, usually of hydrogen isotopes. Plasma is a state of matter like gas, though hotter and with atoms stripped of their electrons so that nuclei move freely. But the nuclei are positively charged and, as with same-sided magnets, they repel each other due to electrostatic force.

For fusion to occur, the plasma must be extremely hot so that fast-moving nuclei can push through electrostatic resistance, allowing the strong nuclear force to pull the nuclei together. In the Sun, fusion is aided by the pressure of gravity. On Earth, scientists must heat plasma to over 100 million C° — six times hotter than the Sun.

When the plasma is hot enough, the nuclei fuse, and a new atom is created (hydrogen to helium). The difference in mass is released as energy.

Though new methods are emerging, scientists make fusion reactions in two ways.

The first is magnetic-confinement: plasma — which as an electrically-charged “gas” is responsive to magnetic fields — is heated, and squeezed by magnets inside doughnut-shaped containers known by the Russian acronym, tokamak.

A diagram of a tokamak
This is a tokamak in South Korea

The second type is inertial-confinement fusion: a pellet of fuel is compressed and heated using a laser.

Diagram of an inertial-confinement device

Though fusion researchers have yet to achieve net energy, there is progress. A key variable is the Triple Product, which measures the energy, density, and confinement time of fusion reactions. By this measure, fusion today is 1,000 times more efficient than 40 years ago. Researchers are now confident net energy is technically possible. They also hope to achieve “ignition” which is when a reaction produces enough energy to become self-sustaining without an outside heat supply.

The most prominent project aiming to achieve proof of concept is called ITER (International Thermonuclear Experimental Reactor), which is under construction in France — supported by a public sector consortium including the EU, US, India, Japan, and 35 other countries.

ITER in Saint-Paul-lès-Durance, France

ITER has been helpful for the conceptual development of fusion and nurturing a supply chain to build the most advanced reactor technology. However, ITER’s focus is basic research, not producing electicity. ITER has a conservative timeline, aiming to achieve net energy in the late-2030s. The program has plans for a subsequent reactor called DEMO that would produce electricity, but that project is decades away.

Given ITER’s long timeline, numerous fusion startups are working to create a reactor much faster — aiming for between 2025–2030 — by focusing directly on technology with the potential for commercialization.

Fusion bears

Some researchers are bearish about fusion.

Skeptical researchers agree net energy is possible, but argue fusion is still a long way from being a power source and has many downsides.

Physicist Daniel Jassby wrote in the Bulletin of the Atomic Scientists, “After having worked on nuclear fusion experiments for 25 years at the Princeton Plasma Physics Lab, I began to look at the fusion enterprise more dispassionately in my retirement. I concluded that a fusion reactor would be far from perfect, and in some ways close to the opposite.”

Here are some of the concerns raised by scholars about fusion’s viability.

Poor energy efficiency

Jassby notes fusion has a problem with parasitic power drain: “Fusion reactors… consume a good chunk of the very power that they produce… on a scale unknown to any other source of electrical power.” Fusion reactors require many energy-intense systems including liquid-helium refrigerators, vacuum pumps, and heating and ventilation equipment.

The most energy-efficient fusion experiment occurred in 1997 at the UK’s JET reactor. The widely reported figure is that the reaction required 24 MW to create and produced 16 MW — recouping 67% of the input.

That figure is accurate but not the whole story. The 24 MW input only accounts for the thermal energy used to heat the plasma, not the total energy required to operate the reactor. Overall, the reactor required of 700 MW (including electricity) to operate—thus recouping 2% of the input.

The issue of fusion “accounting” is important for what is meant by net energy, which measures the amount of heat that goes into creating a reaction relative to the heat produced. Measuring net energy that way makes sense in the context of a basic research experiment focused on the dynamics of plasma, but a practical reactor requires considering a reactor’s total energy economy.

ITER — aiming to exceed JET’s performance — usually reports that its goal is to produce 500 MW for 50 MW of input. But Dr. Thiery Pierre, a plasma physicist, cautions that ITER will also require at least 400 MW of electricity to produce the 500 MW of thermal energy.

With those figures, ITER would exceed net energy by a factor of 10 and then do slightly better than breakeven in total power. That would be an impressive accomplishment for plasma physics, but fusion projects aimed at commercialization will need to produce far more energy than ITER for economical electricity.

Tritium Troubles

The fuel for fusion is, in theory, almost limitless and can even be derived from seawater. However, the practical reality is quite different.

Fusion reactors run on the hydrogen isotopes deuterium and tritium. While fusion reactions are possible with deuterium alone, a mix of deuterium and tritium is 20 times more reactive.

Deuterium is easy to acquire, but tritium must be made artificially as a byproduct of nuclear fission.

Today tritium is one of the most expensive substances, costing $30,000 per gram, and has only been used in a few past fusion experiments.

Fusion projects aim to reproduce — or “breed” — their own tritium using the reactor itself. That would be done by lining the inside of the reaction chamber with a “blanket” of ceramics infused with lithium. The plasma reaction emits a stream of neutrons that bombard the lithium, and the reaction of these elements produces tritium.

But lithium blanket technology remains in the early stages of development and will need big improvements over past approaches. Prior techniques to recapture tritium still required an outside supply. At the Princeton Plasma Physics Lab, for instance, 10% of the tritium was lost to attrition.

To put that in perspective, the DEMO reactor (ITER’s planned successor) would require 300 grams of tritium per day. If 10% had to come from outside sources, the tritium cost would be $900,000 per day or $329 million per year. DEMO’s tritium tab would cost $1 billion in the first three years of operation — plus all other capital and operating costs.

As Dr. Werner Antweiler, an economist at the University of British Columbia puts it, “without effective tritium breeding, nuclear fusion has a tritium dilemma that will keep this power source expensive for a very long time.”

Degradation and radioactive waste

Fusion reactors have problems with long-term structural integrity and radioactive waste.

The plasma reaction itself damages the equipment by emitting a stream of neutrons that knock atoms in the container out of their lattice position causing the device to degrade over time. The reactor’s interior becomes highly radioactive and maintenance work can only be done with robots.

It’s true that fusion produces less radioactive material than nuclear fission. Fusion also cannot have meltdowns, since if there was a power failure the reaction would just stop.

But the degraded and irradiated equipment eventually needs to be handled and buried. Plus there is a history of radioactive tritium passing through certain solid materials and leaking into the environment. These problems require containment facilities and careful disposal practices.

Barriers to fusion in summary

Jassby cites numerous other problems with nuclear fusion. For instance, a fusion reactor would require a large, highly-trained staff to operate it. Fusion technology could also be used to make weapons-grade plutonium (though not easily) creating some proliferation risk.

No one drawback is disqualifying. Every power source has risks and trade-offs. But the problems with fusion will add to the cost and point to serious difficulties in scaling the technology.

How are researchers addressing challenges with fusion?

A practical fusion reactor would require exponential improvements over today’s lab reactors. Fortunately, there are important advances in the parts and processes that make fusion reactors.

These advances lowered the barriers to entry for fusion research, created new paths to net energy, and in time, may be effective enough to overcome the concerns raised by Jassby and others.

Better Magnates

Some companies are drawing on the well-established methods of magnetic-confinement fusion while using new technologies to make more efficient devices.

Commonwealth Fusion Systems is building a tokamak, called SPARC, that is conceptually similar to ITER but utilizes advances in materials science to create a strong magnetic field in a far smaller device.

SPARC visualization (Wikimedia Commons)

SPARC uses a form of high-temperature superconductor made from materials that were not available for past reactors (rare Earth barium copper oxide, or REBCO, to be precise). Using this new superconductor, SPARC can achieve a strong magnetic field in a device with a radius 70% smaller than ITER.

A smaller device reduces capital costs and could provide better energy efficiency. Advanced shielding is also being tested to protect the device from the plasma reaction’s neutron stream for improved durability.

Commonwealth aims to build SPARC and achieve net energy by 2025, producing between 50 and 140 MW — with 25 MW of thermal power needed for the reaction. However, it is not yet clear what the total required energy input will be.

Based on mostly favorable scientific reviews, SPARC has a good chance of at least hitting the net energy milestone. In the process, Commonwealth will gain experience to improve future devices, with plans to build a second device called ARC for commercial use by the early 2030s.

Better lasers

There is progress in inertial-confinement fusion, which uses lasers to ignite a fuel pellet.

Until 10 years ago, inertial-confinement relied on lasers that could only be used a few times a day — due to the need to cool the laser device.

But lasers have improved in recent years. One Czech Republic facility, which opened in 2018, can fire its laser 10 times per second rather than two times a day. Moreover, the energy efficiency of this method improved by a factor of 20 in the last ten years.

The advances could benefit companies like First Light Fusion and General Atomics that are exploring fusion reactors using inertial-confinement.

Advances in computation and simulations

Fusion research is benefiting from advances in computing and simulations. As physicist C. Wendell Horton of the University of Texas explains:

“We’re making calculations that were impossible just a few years ago and modeling data about plasma in three dimensions and in time… Now we can see what’s happening with much more nuance and detail than we would get with analytic theories and even the most advanced probes and diagnostic measurements. That’s giving us a more holistic picture of what’s needed to improve reactor design.”

Advances in computing are also enabling the Canadian company General Fusion to resurrect a once discredited approach. General Fusion’s design injects plasma fuel into molten lead and lithium. Pistons are then used to generate shock waves, compressing and igniting the mixture.

Pistons for the General Fusion device (Wikimedia Commons)

The U.S. Naval Research Laboratory experimented with that approach but gave up in the 1970s, unable to operate the pistons with the precise timing required. But General Fusion has developed algorithms and control systems to improve the performance of the pistons which may provide a path to net energy.

Are the advances enough?

Better magnates, lasers, and computers have accelerated fusion research and there are numerous other advances. With regards to tritium breeding, research estimates advanced reactors could generate 15% more tritium than required for the reaction. That surplus would eliminate the need for costly outside tritium replacement. The lithium blanket technology needed to achieve that surplus is still in development, but ITER, Commonwealth, and other programs are working to achieve proof of concept.

There are also experiments underway to replace tritium fuel with a hydrogen-boron mix that would be far cheaper to produce. Hydrogen-boron fuel faces a big barrier since its ignition temperature is 10 times higher than tritium fuel, but researchers believe there may be ways to reduce the ignition temperature.

Clearly, there are many what-ifs about fusion. But, given the progress so far, there is a good chance one project or another will achieve net energy by 2030. From there, the challenge will be improving and economizing the technology.

Could fusion be economical?

Venture capital firms have invested $2 billion in fusion startups in recent years, including investments from business leaders like Bill Gates and Jeff Bezos. The National Academies of Sciences, Engineering, and Medicine has also recommended public-private partnerships to accelerate the development of fusion energy.

For these investments to make sense, fusion must produce affordable and reliable electricity. Fusion must also have unique advantages relative to other energy sources.

While fusion is trying to break out of the lab, we are also seeing encouraging advances with other technologies.

Since 2010, the cost of solar power declined 82%. The cost of lithium-ion batteries fell by 87% in the last 10 years, which can provide better storage capacity to reduce some of the intermittency problems that have limited renewables. Nuclear fission companies like NuScale and TerraPower are working on small modular reactors that are far safer, smaller, and more efficient than plants like Byron. Carbon capture and sequestration (and better management of methane release) could help mitigate emissions from natural gas.

So what might fusion cost and how does it compare to other electricity sources?

Several studies assessed the cost of fusion electricity — estimating the Levelized Cost of Electricity, or LCOE, which is the cost to build and operate a plant divided by the energy produced over the projected lifetime of a reactor. Since fusion is still experimental, cost estimates are only rough approximations. But these estimates can provide some guidance on the costs of fusion electricity.

A 2018 study in the journal Energy examined a version of the DEMO reactor (the planned successor to ITER). The study estimated the price of fusion electricity would be 3.5 to 5 times higher than the market price of electricity from rival sources.

Other recent studies showed similar warning signs. Scientist and investment analyst Dr. Carly Anderson estimated that in 2040 fusion electricity would cost more than double the price of the next cheapest electricity sources: $0.11/kWh for fusion compared to $0.03–0.045/kWh for solar, wind, and natural gas.

However, fusion could still be viable, as further technological advances can cut the cost.

Fusion reactors should get cheaper as the industry matures. Energy notes that if fusion follows a path similar to other high-tech industrial projects, the 10th power plant would be up to 40% cheaper to build than the first.

Similarly, Anderson estimates that improvements in energy output and the longevity of reactors could cut the price of fusion electricity from $0.11/kWh to $0.052/kWh — much closer to prices for other sources.

What do these estimates mean for fusion’s role in the energy market?

Estimates point to a premium for fusion electricity, but with costs falling over time as the technology is commercialized. But even if fusion were more expensive in terms of the LCOE (which does not take into account all distributional costs), it may still be competitive in certain markets and use cases. A few advantages include:

  • Fusion would be useful in regions with reduced sunlight — like the far north or areas with frequent cloud cover — where solar is less efficient.
  • Fusion would be a power source with high energy density (the amount of energy in a given mass) — requiring relatively little land. In contrast, solar and wind have low energy density and require a lot of land. This could give fusion an advantage in countries with high population density. For instance, India is rapidly increasing its use of solar energy, which can and should scale further in the years ahead. But, given India’s large population and growing demand for electricity, overreliance on solar could eventually require an inordinate amount of land. Nuclear fusion — along with nuclear fission — could provide helpful complements.
  • Fusion would produce no carbon emissions from operations, though, like all energies, it would produce some emissions throughout its lifecycle, such as those to build the reactor. But fusion’s life cycle emissions — using conventional nuclear as a benchmark — should still be about 65 times lower than coal, 40 times lower than natural gas, and 4 times lower than solar. Fusion life cycle emissions would be roughly on par with wind.
  • Fusion has similar safety advantages with advanced nuclear fission since emerging small modular reactors are very safe. But while fusion does have its challenges with tritium, overreliance on fission could stretch the supply of uranium. As nuclear fission production increases in the years ahead (spurred in large part by increased nuclear use in China and Russia), the cost of uranium could triple due to demand pressure. Fusion energy could achieve similar goals as fission without adding to the pressure on the uranium supply.

In general, it will be important to have a diverse portfolio of energy technologies, which are more or less efficient depending on the region, use, and other local variables. Today’s investments in fusion are worthwhile — to give people another tool in the future to meet energy needs, under a wide range of local conditions.

Still, it will probably be at least 15 years until investors recoup any return. That is a long way through the “death valley of investment” — where projects perish because they cannot get to market soon enough.

Fortunately, mission-oriented VCs like Breakthrough Energy Ventures — supported by Bill Gates — are investing in fusion, taking a long-term approach to find solutions to climate change. The Department of Energy could also support fusion research, as it has done with advanced nuclear fission.

Fusion moonshots

The difficult, though plausible, path forward is that fusion can achieve net energy by 2030, launch commercial pilot programs by 2040, and, given its advantages, carve out a niche in certain energy markets by 2050.

Amid critical energy and environmental challenges, policymakers and investors must be prudent about what energy technologies to bet on. Fusion research will have to withstand ongoing scrutiny and may disappoint.

But for now, there is cause for optimism about fusion, which has the potential to play a meaningful role in creating an abundant and sustainable future.

Inertial-confinement device at the National Ignition Facility, Livermore, California

If you spot any technical mistakes, I am happy to correct them. Message me on LinkedIn.



Sharing research on technology, business, and public policy.

Love podcasts or audiobooks? Learn on the go with our new app.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store