The Fusion Energy Landscape
An overview of humanity's Promethean quest to harness the power of stars
The Fusion Energy Landscape
Fusion energy holds the potential to create nearly unlimited grid electricity, cheaply, without emissions, in any geography or climate. The fuel for fusion is hydrogen, the most abundant element in the universe, and the specific isotope of hydrogen isn’t mined from deep in the Earth’s crust like uranium, but rather extracted from regular seawater. In a single gallon of seawater there is enough deuterium (a hydrogen atom with an extra neutron) that, if burned in a fusion reaction, releases the same energy as 700 gallons of refined gasoline. Unlike gasoline the emissions from burning hydrogen aren’t greenhouse gasses, but helium - the stuff of kids balloons.
The promise for humanity of the emergent fusion energy industry is therefore threefold:
Creates a clean, pollution-free energy source that is dispatchable - can be built anywhere, turned on and off as needed unlike solar, wind, and geothermal
Readily accessible reserves of fuel - in the ocean - that would last humanity millions of years without invoking the geopolitical disputes of oil and natural gas reserves
Harnesses the energy of nuclear power, millions of times more energy dense than hydrocarbons, without creating a risk of nuclear plant meltdowns or weapons proliferation
Fusion energy might hold the promise to save our planet, and indeed civilization, from ourselves - yet the newly emerged private fusion industry can seem inscrutable to those outside the specialized realm of plasma physics. This article explores the “energy landscape” of the emerging fusion industry - a description of how different companies are approaching the same problem from different directions, elaborating the relative strengths and weakness of each general approach from a physics-first perspective that is digestible by the non-physicist, with consideration given to enabling tailwinds that answer the most pressing question of a technology long-promised to be 20 years away - “Why now?”
The following sections make use of data collected from public information online on 32 different private fusion energy companies to delineate trends in technological approach, fuel type used, capital raised, total employment, megawattage of proposed reactor, and expected year of positive net energy production from a fusion reaction - the holy grail of “Q>1” that would usher in a new Fusion Age.
The Four Flavors of Fusion: Magnetic, Inertial, Magneto-inertial, and Electrostatic
The flavors of fusion technology are named for how they accomplish the fundamental task of a fusion reactor - to take two smaller atoms, and bring them to such a high temperature and density, and for enough time, that they overcome their mutual electrical repulsion and fuse into a single heavier atom, releasing enormous energy in the process. This combination of density, temperature, and time are known as the ‘triple product’ and succinctly separate both types of fusion reactors, and the types of fuel used. On one end, magnetic confinement uses lower density fuels but traps them for a longer time, like Type One Energy’s Stellarator, while on the other end of the spectrum inertial confinement has higher density and shorter times, like First Light Fusions inertial laser confinement. Additionally, like comparing diesel versus gasoline, different fusion fuels require different temperatures to burn, and release their energy in different ways - fusion fuels that burn at low temperature are more difficult to capture energy from, like in Commonwealth Fusion’s ARC Tokamak, while fuels that burn at high temperatures release energy in a form that can be captured directly, like Helion’s Trenta reactor.
The fuel for fusion isn’t a liquid, gas, or solid - its plasma, but not the kind in your blood - the kind that stars are made of. Plasma is like a gas, but it is so hot, the electrons circling atoms have been stripped away by collisions with neighboring atoms; separating the positively charged atomic nuclei from the negatively charged electrons such that each kind of charge flows as electric current. That plasma burns at millions of degrees makes fusion a difficult task, but that plasma consists of electric currents, and responds to electric and magnetic fields gives commercial entities a finger hold in controlling and confining this fundamental fourth state of matter to make fusion a reality.
Here is a brief overview of the primary approaches to fusion, with a more comprehensive discussion after introducing the types of fusion fuel.
Magnetic Confinement: By far the most common approach to fusion and the most supported by large-scale government funded national labs, and also the most common among private fusion companies. This approach uses superconducting magnets to trap plasma at lower densities but for longer confinement times in large-volume vacuum chambers like the three-story Tokamak at ITER in Grennoble, France. Magnetic confinement is typified by general designs like the Tokamak, Stellarator, and Spheromak, enabled by high-temperature superconducting magnets which can produce magnetic fields on the order of 5 - 10 Tesla - 200,000 times stronger than the Earth’s field and 10x stronger than an MRI.
Inertial Confinement: Using the ‘inertia’ or momentum of lasers, or more rarely acoustic waves, to generate plasmas at extremely high densities but for extremely short periods of time, most recently popularized by the result at Lawrence Livermore Lab’s National Ignition Facility, which demonstrating for the first time a net-positive energy release from a laser-inertial confinement fusion experiment. High intensity lasers spherically compress a frozen pellet of deuterium and tritium to densities, pressures, and temperatures comparable to the inside of a star for millionths of a second, igniting it to fusion. The majority of inertial confinement fusion research has thus far been focused on weapons research, though First Light Fusion is breaking new ground with a focus on commercial grid-power production.
Magneto-Inertial Confinement: The second-most popular approach in private industry, but an approach that has fallen out of favor in public research since the 1970s and 1980s compared to magnetic confinement. This approach blends both magnetic and inertial confinement concepts, where plasma is formed into an initially low density ‘bubble’, then compressed into high densities either by laser pressure, magnetic fields like at Sandia Laboratories Z-Machine, or by liquid metal like General Fusions Magnetized Target Fusion reactor. A key advantage of this approach is the ability to burn hotter temperature fuels, meaning energy capture might be easier than the challenge faced by larger, more steady state devices like Tokamaks.
While there are other approaches to fusion, such as electro-static confinement, beam-driven fusion, and other hybrid approaches, the total funding and employment of these alternative concepts amounts to less than $10m USD and 30 employees across 3 companies worldwide, and they suffer the additional drawback of not capitalizing upon large-scale, multi-decade, multi-billion dollar research efforts undertaken by publicly funded national labs. Notably, Avalanche Energy has raised $45 million to commercialize a compact electro-static confinement generator known as the Orbitron.
Types of Fusion Fuel
Most approaches to fusion burn deuterium and hydrogen, the so-called “DT reaction,” which releases energy as a high-velocity neutron which must be captured in thick radiation blankets and extracted by a heat pump. However there are other types of fuel worth considering - for example, proton-Boron-11 (pB11), which burns at temperatures 10x that of DT reactions but releases energy as fast-moving ionized particles. Despite its popularity, companies pursuing DT-fuelled fusion face the challenge of designing complicated radiation shielding and heat exchanger systems, which involve pumping highly reactive liquid metal like lithium through meter-thick metal walls bombarded saturated with radioactive isotopes. In contrast, fuels like pB11 or Helium-3-Deuterium demand higher temperatures but release energy in a form that can be directly converted to electricity, and don’t have the requirement for meter-thick metal radiation shields.
The following table outlines the most popular fusion fuel sources and some key information: the temperature required in millions of degrees celsius (MC), the energy released per-reaction in mega-electron volts (MeV), the fuel’s origin, and the primary pros and cons motivating its choice as a fusion fuel, as well as examples of companies using that fuel.
Approaches to Fusion
Magnetic Confinement Fusion (MCF)
Total Private Sector Funding: $3.76 bn
Number of companies: 15
Private Sector Employees: 1035
Average proposed reactor output: 240 MW
Average expected year of achieving net energy production: 2031
Exemplars: CommonWealth Fusion ($2.1bn); Tokamak Energy ($150mm); Type One Energy ($100mm).
Notable public-sector projects: DIII-D (Tokamak in San Diego), ENN (Tokamak in China), ITER (Tokamak in Europe)
Magnetic confinement is far and above the most heavily researched, and well-funded area of the fusion energy landscape. With 15 companies and $3.7bn dollars committed to date in private venture capital, as an approach to fusion it also benefits from the longest legacy of public-sponsored national lab projects, the foremost of which is the ITER project in France, a $35 billion dollar project involving tens of thousands of scientists, engineers, and private contractors developing a fusion reactor, with individual magnet coils standing several stories tall, intending to produce over 500 MW of power.
Magnetic confinement fusion schemes work by using magnetic fields to interact with the moving ions and electrons in a plasma - the plasma current - while also pouring energy into the plasma with radio waves and beams of high-energy particles. The stronger the magnetic fields, the more slowly the energy poured into the plasma leaks out. If the energy builds up to a sufficient level the ions in the plasma will fuse together, releasing enough energy to make the reaction self-sustaining - like a glow plug warming up the cylinders in a diesel engine.
Advances in plasma physics simulations, the development of high-temperature superconducting magnets that can generate record-breaking magnetic fields, and the advent of micro-second control systems that can respond in real-time to plasma instabilities, have carried state-of-the-art technology in magnetic confinement far past the early projects of US-Soviet collaboration in the 1970s and make this by a large margin the most promising, well-developed and mature set of fusion technologies in the market.
However, serious engineering challenges remain to make magnetic confinement a reality - the difficulty in trying to bottle a gas that’s millions of degrees hot with magnetic fields means most approaches use DT-reactions as fuel, which create considerable engineering challenges for shielding sensitive components from radiation, maintaining and accessing reactor components that are highly radioactive, and pumping hot liquid metal out of the reactor and through heat exchangers.. This bombardment of neutrons, or neutron flux, can create 5-10 megawatts of heating per square meter at the reactor walls and necessitates extensive metal shielding, or enough energy to power 3000 homes - directly in the vicinity of magnets which must be cryogenically cooled to 20 to 50 degrees above absolute zero.
Notably not all magnetic confinement approaches face the challenge of capturing high-speed neutrons from DT fusion - companies like TAE use a Field-Reversed Configuration or FRC, with the intention of reaching much higher magnetic fields over a much smaller volume, enabling them to confine plasma hot enough to fuse fuels like proton Boron-11 (pB11). Since pB11 is an ‘aneutronic’ fuel type, it releases its energy as ionized atoms which can deposit their energy directly into a magnetic field to be captured as electrical energy without any heat exchange process. TAE made history just last month by demonstrating pB11 fusion for the first time in a magnetically confined reactor design.
Inertial Confinement Fusion (ICF)
Total Private Sector Funding: $280 mm
Number of companies: 8
Private sector employees: 261
Average proposed reactor output: 540 MW
Average expected year of achieving net energy production: 2032
Exemplars: First-Light Fusion ($107mm) , Marvel Fusion ($65mm), Focused energy ($15mm)
Notable public-sector projects: National Ignition Facility (LLNL labs, Q>1 in 2022); Laser MegaJoule (France); Shenguang-III (China)
Inertial confinement as an approach to fusion is conceptually quite simple - to smoothly and symmetrically compress a small fuel pellet to a tiny fraction of its initial size using some kind of shockwave or pressure, most commonly laser power. As it compresses, the density and temperature increase drastically, igniting the pellet to fusion conditions. On the other hand, the reality of compressing a peppercorn-sized fuel pellet of frozen fuel in a smooth, controlled, efficient process using 100’s of powerful lasers pushes at both the limits of our engineering capabilities and ability to accurately model and simulate physics at the small scale.
In stark contrast to magnetic confinement, inertial confinement research programs funded by taxpayers dollars have always been directed towards weapons research - the National Ignition Facility was by no coincidence founded in 1997, one year after the United States became signatory to the Comprehensive Test Ban Treaty. Also in contrast to magnetic confinement, the physics describing inertial confinement fusion is still a hotbed of debate and activity, in trying to combine laser physics, plasma physics, with the physics of acoustic material compression on extremely short timescales.
Another issue is longevity of the reactor components. The interior of tokamaks like those developed by Commonwealth Fusion and at ITER are typically lined with heavy tungsten and lead shielding which protect equipment from radiation while also capturing useful heat energy, allowing external magnetic fields pass through these walls unimpeded. In contrast, laser inertial confinement requires direct line-of-sight from focusing optics to the fusion target itself, with replaceable shielding that is not cheap. Longevity of these optical elements is far below what’s required for an economically feasible fusion reactor - at NIF scientists recently celebrated being able to increase the lifespan of reaction-facing optical elements from six shots before replacement, to 20 shots before replacement, though considering the power output of a few megajoules per shot, NIF would need to operate at 16 shots per second to achieve even the smallest power output of proposed magnetic confinement designs.
One tailwind for the inertial confinement approach to fusion is the massive economies of scale steadily driving down the cost of laser systems around the world. High temperature superconducting magnets are a bespoke and rare technology; lasers are ubiquitous and constantly improving. Indeed, inertial confinement fusion is the only approach to fusion energy that both capitalizes on large public investment into basic feasibility research and massive private investments into a key enabling technology, lasers, that has broader consumer and industrial applications. Indeed laser-inertial confinement is the only approach with such tailwinds that could be miniaturized into small-scale power production - with an optimized reactor design, small fusion power plants could burn aneutronic fuel directly captured as electrical power, avoiding the complicated cryogenics required by HTS magnets and heavy shielding of neutron bombardment.
Magneto-Inertial Confinement (MCI)
Total Funding: $913 mm
Number of companies: 7
Private sector employees: 321
Average proposed reactor output: 100 MW
Average expected year of achieving net energy production: 2029
Exemplars: Helion ($522 mm); General Fusion ($325 mm);
Notable public-sector projects: Z-machine (Sandia Labs, New Mexico);
Magneto-inertial confinement attempts to combine the best of both worlds - the relatively well-understood plasma physics of magnetic confinement, along with the higher densities and temperatures of inertial confinement, to work with fusion fuels that are ideal for directly capturing energy. As a hybrid approach there are many different sub-genres in magneto-inertial confinement, but they all follow the same general concept: develop a relatively low-density cloud of plasma, and then compress it by several orders of magnitude. Commercial approaches generally differ in how they achieve this compression: TAE, Helion and General Fusion all accelerate two similar clouds of plasma towards each other to initiate the compression, but TAE and Helion compress using secondary magnetic fields while General Fusion uses a circulating liquid metal ‘wall’ which is smoothly collapsed around the plasma, compressing it to fusion.
Compared to magnetic confinement, magneto-inertial fusion is attractive because the pulsed-nature of operation gives less time for plasma instabilities to grow and disrupt the machine's operation. For example, in Tokamak at ITER, the amount of raw energy circulating as plasma current is 350MJ, which when disrupted produces an explosion equivalent to 200 lbs of TNT - much of ITER’s development effort has been in how to armor the interior of the reactor and stop plasma instabilities before they detonate the device. Compared to inertial confinement, magneto-inertial is also attractive because it doesn’t require very high implosion velocities and so gives rise to fewer poorly-understood physical phenomena - the task of iterating on a design via simulations and experiments is relatively simpler, though still a complex engineering feat. .
While magneto-inertial confinement benefits from advances driving forward magnetic confinement like high-temperature superconducting magnets, inertial confinement physics originating in weapons research programs is considered highly classified information and is not easily translatable into commercial avenues of research. Rather than continuing a long and cumulative line of public funded research into basic tokamak design and engineering, like Commonwealth Fusion, magneto-inertial companies are often forced to break new territory and innovate novel approaches. A large variety of approaches to generate the initial plasma - Z-pinch, reverse-field configurations (RFC), spheromaks, etc - coupled with a diversity of compression methods - magnetic fields, sound waves, laser beams, liquid metal - make magneto-inertial the most diverse approach to fusion, but reduce the limit of commercial entities to stand on the shoulders of publicly funded giants like ITER for magnetic confinement and NIF for inertial confinement.
Takeaways and Future Outlook
Despite the sudden rise to prominence in the modern news cycle, the commercial quest to develop fusion energy is a unique area of venture capital investments - private companies are almost always partnered with a government-funded national lab, like those found in the Department of Energy, and frequently hire physicists and engineers who had previously spent decades in public sector research, a pattern more similar to biotech and pharma than clean-tech energy. Yet the requirements for a fusion company to reach its goals dwarf those of biotech by several orders of magnitude, which itself is considerably more capital intensive than software: many fusion companies targeting grid-scale power estimate it will take upwards of $1 billion in funding to build their proposed reactor, with some designs consuming hundreds of millions of dollars in high-temperature superconductor magnets alone.
Fusion is also an area uniquely un-accelerated by the advent of machine learning and artificial intelligence, to paraphrase Sam Altman, CEO of OpenAI, where AI tools often produce far worse outcomes than human physicists. Rather the challenges of fusion are classical engineering problems, involving steel girders, megawatt power amplifiers, cryogenic cooling systems, meter-thick radiation shields and thousands of hours of building warehouse-sized devices to harness the most basic forces of physics - the classic stuff of heavy industry. Although, much like the advent of cheap intelligence, fusion energy would be a massive disinflationary force for good in the world by making almost-free the most basic factor of production for our modern material world, energy, and radically changes the cost-benefit analysis of almost every economic activity; for example we might irrigate the deserts with electrolyzed seawater, a concept currently prohibitively expensive because of the cost of grid power.
Regardless of the design ultimately successful or the company which first brings it to market, the quest for fusion is mankind's greatest promethean ambition yet - to harness nuclear fire from the stars, and use it to remake our home here on Earth; into one that is clean, abundant, and prosperous for generations to come.