← All ebooks
Fusion Reactors : Star Power
Biology-Humans

Fusion Reactors : Star Power

For informational and educational purposes only. Not professional advice. Users 21+ only.

Every second, the Sun fuses hydrogen into helium and releases the energy that makes life on Earth possible. Fusion is the most proven power source in the universe, and humanity has been trying to copy it for more than seventy years. This book asks why it is so hard, how close it is, and what a working reactor would actually look like.

Why stars can do it and Earth cannot. Fusion needs atomic nuclei to get close enough to overcome their mutual electric repulsion. A star solves this with gravity: its enormous mass squeezes the core and holds it together. The Sun's core sits at roughly 10 to 15 million degrees Celsius [1]. Earth offers no such gravity, so a reactor has to make up for it with heat, and the target is above 100 million degrees, more than six times hotter than the Sun's core [1]. No solid material can touch a plasma that hot, so it has to be confined by magnetic fields, by the inertia of an implosion, or by other tricks [1][2]. Getting more energy out than goes in depends on temperature, density and confinement time together, and every design is a different way of balancing those three.

Where things stand. The physics has already been shown to work once, in one kind of machine. In December 2022 the National Ignition Facility in California fired 2.05 megajoules of laser energy at a tiny fuel target and got 3.15 megajoules of fusion energy back. By April 2025 it had reached 8.6 megajoules from 2.08, a target gain of about 4.1 [3]. That counts the energy that reached the target, not the far larger amount of electricity the lasers consumed from the grid, so it is a scientific milestone and not a power plant. On the magnetic side, the Wendelstein 7-X stellarator in Germany set a record in 2025 for the key "triple product" over a 43-second plasma, and tokamaks in France and China have held plasmas for more than a thousand seconds [4][5].

The race has changed. For decades fusion was run by governments. ITER, the giant international tokamak under construction in France, has been delayed repeatedly. Its latest schedule starts research operations in 2034 and deuterium-tritium operation in 2039, after a cost increase of about €5 billion [6]. Meanwhile private companies raised a record $4.48 billion in the twelve months to July 2026, bringing the total since 2021 to $14.24 billion across 56 companies [7]. Commonwealth Fusion Systems is building its SPARC tokamak and aims to demonstrate net energy gain, with first plasma now expected in 2027 after earlier targets slipped [8]. Helion has a contract to supply Microsoft with electricity from its Orion plant, but it has softened its public language about the 2028 date, and an executive said full power would not come before 2029 or 2030 [9]. These are company targets, not results, and this book treats them that way. As of the sources used here, no private machine has demonstrated net energy from fusion [10].

What makes it hard. Fusion is not a single problem but a stack of them. A plasma of more than 100 million degrees has to be held stable, and instabilities can end a pulse in an instant. The fusion reaction used by most designs releases a torrent of high-energy neutrons that damage the reactor wall. And the fuel, tritium, is so rare in nature that a plant has to breed its own from lithium, which means a breeding ratio above one in a machine that does not yet exist [11]. Add cost, regulation and the need to run reliably for decades, and the question changes from "can fusion work?" to "can it work cheaply, safely and continuously?"

What nuclear weapons research tells us. Fusion was first achieved at scale in weapons, long before anyone could make it useful. That history is worth telling, at the level of public record, because it shows what the physics can do and what it cannot. A weapon releases its energy in a single, uncontrolled burst. A power plant must produce steady, controlled power, pulse after pulse, for years. The two uses of the same physics have been linked for decades: the National Ignition Facility serves energy research and also supports the U.S. stockpile mission, and its director's 2026 briefing describes using ignition work to close stockpile capability gaps [3][12]. This book covers that history and policy, including arms-control and proliferation questions. It does not cover weapons design, and nothing in it should be read as a guide to building anything.

What changes if the reactor is somewhere else? Stars do fusion in space, so it is natural to ask whether a reactor would work differently outside Earth: floating on its own in space, attached to a space station, or sitting on the Moon or Mars. The short answer is that the core physics stays the same and the engineering changes, mostly for the worse.

What does not change. A star holds its hot gas together with its own gravity. A reactor is far too small for that, so it needs magnetic fields or lasers, and those forces are electromagnetic, not gravitational [1][2]. They work the same anywhere. Vacuum gives nothing extra, because a tokamak is already a vacuum vessel that the ground builds for itself. Weightlessness does not change how a 100-million-degree plasma behaves inside a magnetic field, and the Moon's gravity (about one-sixth of Earth's) and Mars's (about 40 percent) are too weak to matter for confinement. (These points are the author's reasoning from the physics, not claims from a single source.)

What does change. In a vacuum there is no air or water to carry heat away, so waste heat can leave only by radiation, which means large radiators [13]. For nuclear-electric spacecraft the radiator alone can account for 40 percent or more of the power system's mass [14]. Weightlessness removes the natural circulation of hot and cold fluid, so coolant has to be pumped. It also makes heavy structures lighter to hold up, which is one real advantage. There is no air to absorb radiation either, so shielding has to be designed in. Repairs are remote and slow. Everything has to be launched, and ground machines like ITER are the size of a building [15]. Assembling a large reactor from many deliveries removes the limit of fitting on one rocket, but every tonne still has to be lifted out of Earth's gravity. (This is an engineering judgment, not a sourced finding.)

Place by place.

• Free-floating in space: nothing to build on and nothing to cool it except radiators. It also needs a way to use or send out its power, and repairs are very hard.

• Attached to a space station: it shares the station's structure, so size, vibration and crew safety become limits. Heat still has to be radiated away.

• On the Moon: it has ground to build on, some gravity, and rock that could be piled over the reactor as shielding, though that idea is unproven. There is no air, and the lunar night lasts about two weeks.

• On Mars: it has ground and some gravity too. The air is so thin, under 1 percent of Earth's pressure, that it helps cooling very little. Dust is a problem, and the trip from Earth takes months.

In every case, the places change the engineering, not the physics, and none of them removes the hard part. The best case for a reactor in these places is powering things that are already there, such as a spacecraft or a Moon base, and not producing electricity for Earth. Fusion hardware that is small, light, cooled by radiation and reliable without maintenance is a harder target than a ground plant, not an easier one.

Where space does make sense. The strongest case for fusion in space is propulsion and onboard power, not electricity for Earth. A fusion engine could in principle give much higher exhaust speeds than chemical rockets. Companies including Pulsar Fusion and Helicity Space are working on concepts, and Pulsar has reported a first plasma in its Sunbird exhaust test system, using krypton gas and not fusion fuel, with a demonstration mission hoped for in 2027 [15][16]. Those are early-stage company claims, and no fusion engine has yet produced net energy anywhere. The famous space fuel, helium-3 from the Moon, adds its own doubts. Reacting deuterium with helium-3 needs temperatures roughly ten times higher than deuterium-tritium fusion [17], critics argue it is not truly neutron-free because side reactions between the deuterium nuclei produce neutrons [18], and lunar mining at scale is unproven [17].

The star already in space. There is one fusion reactor in space that works, and it is the Sun. Orbital solar farms that beam energy to Earth skip the problem of building a reactor and use a proven one, although critics call the economics very difficult and the engineering untested at scale [19]. It is a different technology from fusion, but any honest comparison of "star power from space" has to include it.

Where will it happen? The most likely outcome, in the author's reading of the evidence, is that the first commercial fusion electricity will come from Earth-based grid plants. That is where the money and the announced siting agreements are: CFS plans its ARC plant in Virginia in the early 2030s, and Helion's Orion is being built in Washington State [8][9]. Fusion in space is a later, narrower question, mostly about propulsion and power for spacecraft, and it depends on the same breakthroughs on the ground first. This book weighs the options, including standalone reactors, large grid plants, fusion in orbit and space propulsion, and says plainly where the evidence ends and speculation begins.

How to read this book. The goal is a fun deep dive, not a sales pitch and not a prediction. Where a claim comes from a company, it is labeled as a company claim. Where something is a record, a delay or an unproven target, the text says which. Fusion has a long history of being "thirty years away," and a good book on it has to hold both ideas at once: the progress is real, and so are the obstacles.

References

1. Stanford Understand Energy, Fusion (Sun's core 10 to 15 million °C; Earth requires more than 100 million °C; no gravitational confinement). https://understandenergy-prod.stanford.edu/news/understand-fusion

2. Stanford Understand Energy, Nuclear Fusion (magnetic and inertial confinement). https://understand-energy.stanford.edu/energy-resources/nuclear-energy/fusion

3. Lawrence Livermore National Laboratory, 2025 annual report, National Ignition Facility (8.6 MJ from 2.08 MJ, target gain 4.13); R&D World on the December 2022 result (3.15 MJ from 2.05 MJ). https://annual.llnl.gov/sites/annual/files/2026-02/llnlannual25-3-national-ignition-facility.pdf and https://www.rdworldonline.com/?p=78774

4. EUROfusion and Max Planck Institute for Plasma Physics, Wendelstein 7-X triple-product record (May 22, 2025, 43 seconds). https://euro-fusion.org/eurofusion-news/wendelstein-7-x-sets-world-record-for-long-plasma-triple-product/

5. IAEA Fusion Energy Conference 2025 proceedings (EAST up to 1066 s; WEST up to 1337 s). https://conferences.iaea.org/event/392/papers/36141/files/13971-Litaudon_Proceeding_FEC%202025-vf.pdf

6. Science|Business, ITER confirms more delays and €5B cost overrun (2024); Physics World. https://sciencebusiness.net/news/iter-fusion-project-confirms-more-delays-and-eu5b-cost-overrun and https://physicsworld.com/?p=115383

7. Energy Live News, fusion sector secures record $4.48bn (Fusion Industry Association 2026 report); World Nuclear News. https://www.energylivenews.com/2026/07/22/fusion-sector-secures-record-4-48bn-in-funding/ and https://world-nuclear-news.org/articles/fusion-industry-raised-usd45-billion-over-past-year-report-says

8. CFS and trade coverage on SPARC (first plasma 2027, ARC in Virginia in the early 2030s); schedule slippage noted by an independent summary. https://www.aip.org/scilights/sparc-positioned-to-demonstrate-fusion-power-breakeven and https://www.inkl.com/news/fusion-power-nearly-ready-for-prime-time-as-commonwealth-builds-first-pilot-for-limitless-clean-energy-with-ai-help-from-siemens-nvidia and https://mentat-s3r.pages.dev/finds/cfs-sparc-first-plasma-2026

9. Axios Pro, Helion and the 2028 date (September 2026); TechCrunch and NEI Magazine on Polaris. https://www.axios.com/pro/climate-deals/2026/09/08/helion-energy-fusion-net-electricity-date-year and https://techcrunch.com/?p=3092675 and https://www.neimagazine.com/news/helion-advances-fusion-energy-program/

10. The Next Web, 17 fusion startups have raised over $100M each (no one has hit breakeven). https://thenextweb.com/news/fusion-startups-13-billion-private-funding-race-commercial-energy

11. Commercial Fusion Power Faces 3 More Epic Tech Hurdles (tritium breeding ratio above one; neutron damage). https://gs.asiams.net/article_details/commercial-fusion-power-faces-3-more-decades-of-hurdles

12. LLNL, NIF Update, February 2026 (director's briefing on stockpile capability gaps); okdiario on NIF's defense role. https://lasers.llnl.gov/sites/lasers/files/2026-03/Gordon-Brunton-NUG-2026-FINAL.pdf and https://okdiario.com/techy/en/?p=4605

13. NSS, Thermal Management in Space (only radiation rejects heat in a vacuum). https://nss.org/settlement/nasa/spaceresvol2/thermalmanagement.html

14. NASA MSFC, Lightweight, High-Temperature Radiator for Space Propulsion (radiator 40% or more of power-system mass for nuclear-electric spacecraft). https://ntrs.nasa.gov/api/citations/20130001608/downloads/20130001608.pdf

15. Live Science, fusion propulsion (Pulsar Fusion, Princeton and Helicity Space; reactor must fit on a rocket; ITER is the size of houses). https://www.livescience.com/space/space-exploration/everything-we-know-about-space-travel-is-going-to-change-within-a-decade-the-fusion-breakthrough-that-could-unlock-a-path-to-the-stars

16. Tech Briefs and Pulse on Pulsar Fusion's Sunbird first plasma with krypton (company claims). https://www.techbriefs.com/component/content/article/55030-advancing-nuclear-fusion-space-propulsion and https://www.pulse.bot/spacetech/news/pulsar-fusion-ignites-first-fusion-rocket-plasma-paving-way-for-faster-mars-trips-b3565b60-81c2-44af-a460-723cf48ec93c

17. Space Odyssey Hub, helium-3 and Interlune (D-He3 needs roughly ten times hotter plasma; mining unproven). https://spaceodysseyhub.com/articles/helium-3-moon-mining-interlune-2026

18. MIT Technology Review, Mining the Moon (critic on helium-3 fusion feasibility and deuterium side reactions). https://www.technologyreview.com/2007/08/23/223985/mining-the-moon/

19. Metavert, Nuclear Fusion vs Orbital Solar Farms; Live Science on space-based solar power and its critics. https://metavert.io/nuclear-fusion-vs-orbital-solar-farms and https://www.livescience.com/space/were-not-a-competitor-to-fusion-we-are-fusion-inside-the-race-to-harness-unlimited-solar-power-in-space

A Note on Sources

This book leans on primary and institutional sources where it can: Lawrence Livermore National Laboratory's annual report and briefing, EUROfusion and the Max Planck Institute, the IAEA's 2025 conference proceedings, NASA technical reports and Stanford's Understand Energy pages. Industry and policy developments come from trade and news coverage such as Science|Business, Physics World, Axios, TechCrunch, Live Science and MIT Technology Review. Claims from companies are labeled as company claims, and the book treats announced dates as targets, not results.

Some sources are weaker, and they are flagged where used. Source 8 includes a schedule summary hosted on a pages.dev site and a syndicated copy, source 11 is an article on a little-known site, source 12 includes a news-site summary of NIF's defense role, and source 19 includes a commentary site. Source 9 includes Axios Pro, a subscription service, and source 18 is from 2007. The reasoning about reactors in space, marked in the text, is the author's engineering judgment and not a sourced finding. Primary announcements and filings from the fusion companies, the IAEA and ITER would be better where they exist. As of these sources, no private machine has shown net energy from fusion. This is the best record that could be assembled so far, and you are encouraged to check any claim against the source named beside it.