Fusion Reactors and Net Energy Gain: Revolution or Mirage?

Let’s talk about fusion reactors, some of the most exotic and poorly understood technologies that have been proposed for solving our biggest energy and ecological challenges.

The Basics of Fusion

At a very simple level, nuclear fusion is a process where lighter atomic nuclei combine to form heavier nuclei, releasing energy along the way (see Figure 1 below). Fusion is the fundamental basis for the energy we get from the Sun, which converts about 600 million tons of hydrogen into helium every single second. The high temperatures of the Sun are absolutely critical for fusion, because electromagnetic forces would otherwise prevent the nuclei from combining. High temperatures deliver enormous kinetic energy to these nuclei, and that kinetic energy is enough for them to overcome their mutual electrostatic repulsion and combine into a new atom. The resulting atoms from the fusion process weigh slightly less than the sum of the individual lighter atoms that combined in the first place. That mass difference is converted to energy, usually in the form of high-energy photons like gamma rays, which can then spend hundreds of thousands of years bouncing around inside the Sun, losing energy along the way before finally making it to the solar surface, where many of them escape as visible radiation and eventually come to greet planet Earth.

Figure 1: A basic visual explanation of how nuclear fusion works in the context of hydrogen isotopes fusing into helium, from the International Atomic Energy Agency.

In principle, fusion offers the prospect of generating vast amounts of energy, like it does in the Sun, so many nations and scientific groups have been interested in harnessing its power. In the past few decades, dozens of experimental fusion reactors have been built around the world. It’s important to emphasize the word experimental. There are no commercially viable fusion reactors anywhere. In other words, there’s no fusion reactor out there which is actually producing heat that’s driving a turbine for electricity generation. In a practical and commercial sense, fusion is very much an untested technology, and so much of what's described in this post is fancy theoretical guesswork about how fusion technology is supposed to work.

It’s generally estimated that we are at least another decade away from having even one commercially operational fusion facility, and even that timeline might be very generous. If fusion has a major impact at all on our energy mix, it likely won't be until the second half of the century, which could be too late to concretely address the major challenges we face. Many governments around the world have been funding fusion projects, most notably the International Thermonuclear Experimental Reactor (ITER), with the goal of generating a sustained fusion reaction and making other technological breakthroughs that could allow fusion to be commercialized. In this post, I will avoid delving into the subject of cold fusion, as there is no reliable evidence that it’s ever been done, and a high-profile flop in 1989 ended most of the interest in the field. To our best scientific understanding, generating fusion requires high-temperature conditions. At a practical level then, cold fusion is just pseudoscience.  

Fusion Reactors and Net Gain

Generating a sustained fusion reaction is a major technological challenge because we need to produce a highly controlled environment of high temperatures, high pressures, and intense radiation. It's not easy to find the right combination of materials that can survive such an unforgiving environment for a long period of time. Fusion reactors can come in many different types, shapes, and sizes, but they are generally intended to work as follows: we start off with a specialized vacuum chamber, turn on the confinement mechanism (like magnetic fields), inject the nuclear fuel into the chamber, generate extremely high temperatures and pressures to create an ionized plasma, trigger and sustain the fusion reactions of the plasma, and finally take the resulting heat and use it for something (the obvious commercial use case would be taking the heat and driving a steam turbine to produce electricity). Some of the most popular reactor designs are tokamaks, torus-shaped structures that basically looks like giant donuts, and stellerators, which are similar to tokamaks but confine the plasma in a slightly different way (see Figure 2 for a conceptual diagram of a tokamak). Fusion reactors can control and heat the plasma through a variety of methods, with two popular ones being the application of magnetic fields generated from electric currents (magnetic confinement) or simply using high-energy lasers (inertial confinement). But whatever the method of confinement, the holy grail in the fusion industry has long been to produce something known as “net energy gain,” so let’s explore what this is all about.  

Figure 2: Diagram of a typical tokamak reactor, showing the magnetic fields that confine the plasma (pink) in place, from the Department of Energy.

As a physicist, one of the things I find very frustrating are the many misguided claims that fusion reactors will produce more energy than we put into them. This is a common mistake that appears in numerous media articles and stories. If you remember your high school physics days, you learned about something called energy conservation. It’s one of the most important principles in science: energy can change and be converted to different forms, but the total amount of energy always remains constant in any given physical process. What goes in must be equal to what comes out. When people therefore make these kinds of reckless assertions about fusion technology, you could be forgiven for thinking that somebody has broken the laws of physics and stumbled on a magical device. Of course, that’s not really what’s happening at all.

When fusion supporters talk about net energy gain, they are measuring something very specific: whether the energy released from fusion reactions exceeds the direct energy used to start and sustain the reaction in the first place. In 2022, for example, the National Ignition Facility conducted a famous experiment that produced 3.15 megajoules of energy through fusion from only 2.05 megajoules of laser energy that entered as an input. Is this an earth-shattering result? Does it mean that we’ve somehow violated or overcome the laws of physics? Absolutely not, because there’s a critical factor these narratives ignore: the fuel source itself.

When you light a match and start a campfire, the energy of the fire is greater than the energy of the initial spark from the match, and you could therefore claim that a net energy gain has been achieved, just like fusion supporters do. But to really get the fire going and expanding, you also need a source of fuel, like wood. Only by repeatedly converting the chemical bond energy stored in the wood into the heat energy of the fire do we actually get a sustainable fire. And once the fire burns through all the wood, we all know what happens: no more fire. Fusion reactors have a fuel source too, and it’s typically things like deuterium and tritium, which are isotopes of hydrogen. These components themselves have lots of internal mass-energy, and once you include that energy with the energy of the magnetic fields or lasers that are driving the reaction, then energy conservation still holds in the end. So no, fusion does not violate the laws of physics and it doesn’t somehow produce more energy than what we put in, provided we’re clear on the critical point that the energy input includes the nuclear fuel itself, otherwise you can’t get any fusion to happen at all.

The reason why I wanted to heavily emphasize this part is because fusion is sometimes presented as a miracle technology that's somehow detached from all the pesky natural constraints to which other technologies are subject. The International Atomic Energy Agency even claims that fusion offers the potential for "limitless energy production," an absurd and pseudoscientific statement that makes one wonder who was in charge of writing this stuff. The reality is that establishing a large global fusion industry would require the extraction, transportation, refinement, industrial processing, and distribution of vast quantities of natural resources and raw materials, a massive undertaking that would drastically scale up energy use, resource consumption, and GHG emissions.

The Limits of Fusion

The nuclear fuels that are necessary to sustain fusion reactions don't just magically fall out of the sky; they have to be extracted, refined, and transported. All of that takes energy. For an imperfect though useful analogy, many people don't realize that a substantial portion of the uranium in a typical nuclear reactor has to be replaced about once a year or two. That's because as the fission process keeps going (nuclear reactors rely on the fission, not fusion, of isotopes like U-235), it leads to a reduction in the available stock of nuclear materials necessary for further fission. Even the Sun, as brilliant and radiant as it is today, will stop shining in the future because it's going to run out of fuel. Now, the Sun has enough nuclear fuel to keep shining for a few more billion years, but humanity doesn't have the same luxury. Deuterium is certainly plentiful in nature and that's great, but tritium is not and has to be produced from synthetic sources and processes. Many fusion reactors are being designed precisely with that goal in mind.

When deuterium and tritium fuse together, they produce a helium atom along with a high-energy neutron, and these free neutrons strike the inner shell of the containment chamber, usually called the blanket, at tremendous speeds. They strike with so much energy that the blanket becomes very hot. Water passes through the blanket and picks up this heat, and the idea is that we can use the steam from this hot water to drive a turbine and produce electricity. But blankets are meant to have a dual purpose: not just to heat up water but also to create more tritium itself, usually called breeding in the nuclear industry. That's why current plans call for blankets in fusion reactors to be made out of lithium and lithium compounds. When the high-energy neutrons strike these lithium atoms and compounds, one of the byproducts of those reactions is tritium, which can then be fed back into the initial fusion reactions as a fuel source. One can easily see where the problem is: to produce more tritium, we need to produce more lithium or some other elements and compounds, which is guaranteed to be a very energy-intensive process with a finite timespan. And the supply of tritium is not the only issue. Even though deuterium is easily found in nature, it still needs to be extracted and transported on a vast scale if the fusion industry ever achieves commercial success, a process which in and of itself would require vast amounts of energy and resources.

Supporters of nuclear energy tend to think that we can deal with scarcities either through substitution (let's just use thorium instead of uranium) or that we can simply shift extraction strategies (let's just get uranium from the oceans instead of mining it). These strategies can merely delay the inevitable, and not for long either. That's because the absolute natural availability of a resource on planet Earth is not the only factor that determines how much of it we can actually use. The extraction process itself is also a crucial factor, and it's simply becoming more difficult and more expensive to extract so many of our most important resources, from oil to uranium, regardless of their available quantities in the natural world. This problem is especially critical for uranium because we've already extracted most of the high-quality ores that are easy to mine, so what's left now are increasingly low-grade deposits that are getting harder to extract, contributing to a recent surge in uranium prices. And going to the oceans isn’t a magical solution either, because uranium ocean concentrations are very low and the challenges of large-scale extraction would be significant.

The same basic thing will eventually happen for whatever nuclear fuel we decide to use in fusion reactors. And this is without delving into a million other problems with relying on fission-based or fusion-based nuclear energy for our civilizational needs, a subject which I plan to cover in a future post. Although it's great that the latest fusion experiments have achieved net energy gain, defined in a very narrow and technical sense, this fact by itself is not a compelling reason to carry out a massive energy-intensive transition towards fusion, which would almost certainly come at the expense of pursuing other more effective strategies for dealing with our bionomic challenges.