The energy question.
A guide to the technologies that will run the next century — solar, fusion, storage, fission, hydrogen, geothermal — and the messy real-world transitions in between.
Where the energy comes from
The story is solar
Solar PV cost has fallen ~90% since 2010. Module prices crossed $0.10/W in 2024. Levelized cost of utility solar is now below $30/MWh in best sites — cheaper than the operating cost of paid-off coal. Kommersant, IEA, Bloomberg NEF, RethinkX all project the trend continues.
Why it ate everything else
- Wright's Law learning rate: ~28% per doubling
- Modular: deployable from rooftop to gigawatt
- No moving parts; falling balance-of-system costs
- Capacity factor now 18–28% with trackers
forecast Solar passes coal globally as the largest power source by 2027–28; passes everything else by 2035–40.
The 30-years-away technology, perhaps not
Fusion has been "30 years away" since 1955. The 2020s have changed the picture in two ways: NIF achieved net energy gain (Q>1) on shot N221204 in December 2022; and a wave of well-capitalized private fusion companies (CFS, TAE, Helion, Tokamak Energy, ZAP) target pilot plants in the 2030s.
| Player | Concept | Pilot target | Notes |
|---|---|---|---|
| ITER | Tokamak (D-T, superconducting) | First plasma 2034 (delayed) | ~$25B+ international consortium |
| Commonwealth Fusion (CFS) | Tokamak (HTS magnets) | SPARC ~2027 · ARC ~2032 | MIT spinout; HTS magnet bet |
| Helion | Field-reversed config (D-He3) | Polaris 2024–25 | Direct electrical conversion |
| TAE | FRC, p-B11 | Late 2030s | Aneutronic ambition |
| Tokamak Energy | Spherical tokamak | Demo 2030s | UK · STEP collaboration |
| ZAP Energy | Z-pinch (sheared flow) | Pilot 2030s | Compact, cheap if it works |
scenario First commercial-scale fusion electron added to a grid: optimistically 2032, more plausibly 2035–40. Fusion does not solve the 2030 climate problem.
How a tokamak confines plasma
Batteries, and beyond batteries
Fission's second chance
Public opinion on nuclear has reversed in many countries since 2020 — pre-pandemic skepticism gave way to climate-driven pragmatism. France ordered six new EPR2 reactors in 2022; Sweden reversed a phase-out; Japan restarted idled units; the US passed the IRA's nuclear PTC.
Three nuclear bets
- Conventional Gen-III+ — AP1000, EPR, APR1400. Proven, expensive, slow to build in West (Vogtle 3/4 12yr +$30B over budget).
- Small Modular Reactors (SMRs) — NuScale (cancelled 2023 first project), Holtec, X-energy, Rolls-Royce. Factory-built, target $100/MWh. First operating units 2028–32. scenario
- Gen-IV — Natrium (TerraPower, Wyoming), Aurora (Oklo), molten-salt (Kairos, Terrestrial Energy). 2030s timelines.
Fuel: HALEU bottleneck (high-assay low-enriched uranium). Russia is the dominant supplier; Western capacity ramping but tight through 2030.
The molecule with a marketing problem
Most hydrogen today (~95%) is "gray" — made from natural gas, emits CO₂. The clean alternative, "green" hydrogen from electrolysis, currently costs $4–7/kg vs $1–2 for gray. Bloomberg, IEA, Liebreich agree: hydrogen will play a real role in steel, ammonia, shipping, aviation — and will be a bad fit for cars and home heating.
Hydrogen Ladder · Liebreich
- Tier A: fertilizer ammonia, refining, methanol, hydrogenation
- Tier B: steel (DRI), high-grade industrial heat
- Tier C: long-haul shipping (NH3 carrier)
- Tier D: aviation (synfuel feedstock)
- Tier G: home heating, light vehicles, peaking power — bad fit
forecast Green H2 reaches $2/kg in best sites by 2032; mass deployment in tier A/B follows.
The dark horse
Enhanced geothermal systems (EGS) take the well-drilling expertise that the shale-oil revolution built and turn it on hot dry rock. Fervo Energy's Cape Station (Utah, 400 MW under construction 2024–28), Eavor's closed-loop concept, and the DOE's Frontier Observatory have moved EGS from concept to early commercial.
The promise: 24/7 dispatchable clean baseload, geographically widespread, drilling depths of 4–7 km. The hard part: cost-down to $50/MWh-equivalent, and inducing seismicity safely.
scenario 50–100 GW of EGS by 2035 in optimistic deployment scenarios. Could be the second large dispatchable clean source after nuclear.
Who to read
| Vaclav Smil | Energy scaling realism; How the World Really Works |
| Saul Griffith | Rewiring America; electrify everything |
| Hannah Ritchie | Our World in Data; Not the End of the World |
| Jesse Jenkins | Princeton ZERO Lab; transition modeling |
| Michael Liebreich | Hydrogen ladder; energy realism |
| Bret Kugelmass | Last Energy; nuclear advocacy |
| Daniel Yergin | Energy historian; The Quest, The New Map |
| Ramez Naam | Solar S-curves; technology forecasting |
"The energy transition is happening. It is also happening too slowly." — a paraphrase of every honest analyst.
Recommended source
Real Engineering · "The Truth About Solar" / "Why ITER Will Take Forever"
Brian McManus's deep-dive engineering channel. Pair with Just Have a Think (UK) and Practical Engineering for grid topics.
youtube.com/@RealEngineering →Kurzgesagt · "The Pursuit of Wonder · Fusion"
10-minute primer on why fusion is hard, where the milestones are, and what NIF actually accomplished.
youtube.com/@kurzgesagt →Three energy futures
Indicators · 2026–2030
- Solar deployment rate · 600 GW/yr → 1 TW/yr inflection
- Battery prices crossing $50/kWh
- SPARC first plasma · 2026–27
- First operating SMR (US, UK, Canada) · 2028–32
- Permitting reform in US, EU — grid interconnection queues
- Critical-mineral supply chains (Li, Ni, Co, rare earths) — recycling & sodium swap