Fusion's physics curve is bending upward and capital is following, but tritium scarcity, ITER's nine-year slip, and a 150x wall-plug gap show why the 2030-2035 pilot-plant cluster is a bet, not a schedule.
Key findings
Industry timelines cluster on the early 2030s: 35 of 45 surveyed fusion companies target a commercial-grade pilot plant between 2030 and 2035, with only 5 aiming before 2030. Peak Nano (citing FIA 2025 Global Fusion Industry Report)
MIT/CFS demonstrated a 20-tesla HTS magnet in September 2021 — the world record for a large-scale magnet — enabling SPARC-class tokamaks at a fraction of ITER's volume. MIT News
ITER's 2024 rebaseline pushes first operation from 2025 to 2034 and full deuterium-tritium burn from 2035 to 2039 — a nine-year slip plus a ~€5B cost increase on a project conceived in 1985. Physics World
JET's final 2023 D-T campaign produced 69 MJ over ~5 s from 0.2 mg of fuel, tripling its 1997 record and outpacing NIF's per-shot energy by 20× before retiring. EUROfusion
Since first crossing ignition in Dec 2022, NIF has hit ignition ten times, climbing from 3.15 MJ to a record 8.6 MJ yield in April 2025 (target gain 4.13). Lawrence Livermore National Laboratory
NIF's best 8.6 MJ shot used 2.08 MJ on target, but the laser itself draws ~300 MJ from the wall — so ignition is a physics result, not yet a power source. Nature
Cumulative private fusion investment reached $9.77B by July 2025, up roughly five-fold since 2021, with the latest 12 months alone adding $2.64B — the second-best year on record. Fusion Industry Association
Plasma confinement endurance is improving fast: France's WEST sustained plasma 1,337 s in Feb 2025, beating China's EAST 1,066 s set weeks earlier and WEST's own 824 s from Dec 2024. DIFFER
Tritium is the binding fuel constraint: civilian global stock is ~28 kg, a single first-of-a-kind reactor may need ~10 kg to start, and a 1 GW plant burns ~55 kg/yr against CANDU output of ~2 kg/yr. Federation of American Scientists
U.S. fusion now sits under a lighter regulatory path: the NRC voted in April 2023 to use Part 30 byproduct-material rules, Congress codified it in the 2024 ADVANCE Act, and the proposed rule was issued Feb 2026. U.S. Federal Register / Nuclear Regulatory Commission
The argument
Private fusion capital reached $9.77B by July 2025, roughly five times the 2021 base, with $2.64B added in the last twelve months alone.
JET's farewell shot delivered 69.26 MJ in a single 5-second pulse, triple its 1997 record and 20x NIF's per-shot energy.
NIF has crossed ignition ten times since December 2022, peaking at 8.6 MJ yield in April 2025, target gain 4.13.
MIT/CFS hit a 20-tesla HTS field in 2021, unlocking SPARC-class tokamaks at a fraction of ITER's volume.
WEST sustained plasma for 1,337 seconds in February 2025, beating EAST's 1,066 s record set weeks earlier.
Yet NIF's record shot needed ~300 MJ from the wall to deliver 2.08 MJ on target: a 150x gap from physics to power source.
ITER's 2024 rebaseline slips first operation to 2034 and full D-T burn to 2039, with ~€5B in fresh cost.
A 1 GW fusion plant would burn ~55 kg of tritium per year against ~28 kg of global civilian stock and ~2 kg/yr CANDU output.
35 of 45 surveyed developers still target a commercial pilot plant in 2030-2035, a timeline the fuel and grid economics have not yet earned.
This research is published in English and Spanish. Ver en español
La investigación detrás de esta presentación
La curva de la física de fusión se inclina al alza y el capital la sigue, pero la escasez de tritio, los nueve años de retraso de ITER y una brecha 150x en red eléctrica muestran por qué el grupo de plantas piloto 2030-2035 es una apuesta, no un calendario.
MIT/CFS demostró en sep. 2021 un imán HTS de 20 teslas — récord mundial para un imán a gran escala — habilitando tokamaks tipo SPARC con una fracción del volumen de ITER. MIT News
El reajuste de 2024 de ITER mueve la primera operación de 2025 a 2034 y la quema D-T plena de 2035 a 2039: nueve años de retraso y ~5 mil M€ más en un proyecto concebido en 1985. Physics World
La campaña D-T final de JET en 2023 produjo 69 MJ en ~5 s con 0,2 mg de combustible, triplicando su récord de 1997 y superando 20× la energía por disparo del NIF. EUROfusion
Desde la ignición de dic. 2022, NIF ha logrado ignición diez veces, subiendo de 3,15 MJ hasta un récord de 8,6 MJ en abril de 2025 (ganancia de blanco 4,13). Lawrence Livermore National Laboratory
El mejor disparo de 8,6 MJ del NIF usó 2,08 MJ sobre el blanco, pero el láser consume ~300 MJ de la red: la ignición es física, no aún una fuente de potencia. Nature
La inversión privada acumulada en fusión llegó a 9,77 mil M$ en jul. 2025, cinco veces más que en 2021, con 2,64 mil M$ añadidos en los últimos 12 meses: el segundo mejor año registrado. Fusion Industry Association
La duración del plasma mejora rápido: WEST mantuvo 1.337 s en feb. 2025, superando los 1.066 s de EAST y los 824 s de WEST de dic. 2024. DIFFER
El tritio es el cuello de botella: el stock civil global es ~28 kg, un reactor pionero puede necesitar ~10 kg para arrancar y una planta de 1 GW quema ~55 kg/año frente a ~2 kg/año de CANDU. Federation of American Scientists
La fusión en EE. UU. opera bajo un marco regulatorio más ligero: la NRC votó en abr. 2023 usar la Parte 30 de materiales subproducto, el Congreso lo codificó en la Ley ADVANCE de 2024 y la regla propuesta salió en feb. 2026. U.S. Federal Register / Nuclear Regulatory Commission
El argumento
Private fusion capital reached $9.77B by July 2025, roughly five times the 2021 base, with $2.64B added in the last twelve months alone.
JET's farewell shot delivered 69.26 MJ in a single 5-second pulse, triple its 1997 record and 20x NIF's per-shot energy.
NIF has crossed ignition ten times since December 2022, peaking at 8.6 MJ yield in April 2025, target gain 4.13.
MIT/CFS hit a 20-tesla HTS field in 2021, unlocking SPARC-class tokamaks at a fraction of ITER's volume.
WEST sustained plasma for 1,337 seconds in February 2025, beating EAST's 1,066 s record set weeks earlier.
Yet NIF's record shot needed ~300 MJ from the wall to deliver 2.08 MJ on target: a 150x gap from physics to power source.
ITER's 2024 rebaseline slips first operation to 2034 and full D-T burn to 2039, with ~€5B in fresh cost.
A 1 GW fusion plant would burn ~55 kg of tritium per year against ~28 kg of global civilian stock and ~2 kg/yr CANDU output.
35 of 45 surveyed developers still target a commercial pilot plant in 2030-2035, a timeline the fuel and grid economics have not yet earned.
Esta investigación se publica en inglés y español. Read in English
