The Nuclear Fuel Cycle

Part I of The Nuclear Fuel Cycle: A Complete Primer

Chapters 1–3 lay the ground that the other twenty-four chapters stand on. They establish what nuclear energy is physically, where its fuel comes from, and why the industry runs on the fuel it does. The three chapters move from the nucleus to the mine to the road not taken. Each one sets up a fact that the rest of the book keeps returning to: fission’s energy density, uranium’s 0.7% U-235 content, and an industry built on infrastructure inherited from weapons programs.

Chapter 1 — Nuclear Energy

The safety question, answered first. The chapter starts with the question most people ask first and answers it with the life-cycle epidemiology (Markandya and Wilkinson in The Lancet; Sovacool and colleagues). Measured in deaths per terawatt-hour (TWh), coal comes to about 24.6, natural gas about 2.8, wind about 0.04 and nuclear about 0.03. The chapter then separates the three accidents that shape public perception. Chernobyl (1986) was a flawed design run in violation of basic procedures, with 31 direct deaths and a thyroid-cancer toll that is still being counted. Three Mile Island (1979) damaged the core but killed no one. At Fukushima (2011) no member of the public died from radiation, one worker’s later lung cancer was confirmed as occupational, and roughly 2,200 deaths came from the evacuation itself.

The denominator. A safety statistic means something only if the base it rests on is large. Today roughly 440 reactors operate in 31 countries and produce about 2,600 TWh a year, around 10% of world electricity. Cumulative output is about 85,000 TWh, over more than 18,000 reactor-years. The United States runs 93 reactors in 28 states, and the average plant has operated for more than 40 years.

Fission and energy density. In fission, the strong nuclear force stores energy in the nucleus. When a heavy nucleus splits, the fragments sit in a lower-energy configuration. The difference in binding energy appears through E = mc². Fully fissioned, a kilogram of uranium matches about 3,000 tons of coal. The book is explicit that this is a theoretical ceiling. The annual fuel for a 1,000 MW plant fits on one truck, while a comparable coal plant takes about 100 train carloads a day. Every US spent fuel assembly ever discharged would cover a football field to a depth of less than ten meters.

U-235 and the 0.7%. Natural uranium is about 99.3% U-238 (uranium-238) and 0.7% U-235. The two differ by three neutrons. U-235 is fissile: when a slow neutron strikes it, it splits and releases two or three more neutrons. U-238 mostly absorbs neutrons without splitting, but it is fertile. Light-water reactors need 3–5% U-235, and some advanced and research designs need up to 20%. That gap is why enrichment, conversion and much of fuel fabrication exist.

Chain reaction: reactor versus weapon. The chapter defines the multiplication factor k. At k = 1 the reactor is at criticality, the operating state of every power reactor. Below 1 the reaction is subcritical and above 1 it is supercritical. A weapon needs k far above 1, sustained, with material above 90% U-235 or plutonium above 93% Pu-239, compressed by explosive lenses within microseconds. At 3–5% enrichment, an excursion can melt fuel but cannot produce a nuclear explosion. The chapter notes the open technical debate over whether a prompt criticality excursion started the Chernobyl sequence. Either way, nuclear detonation physics was never involved.

Heat to electricity. The plant boils water. In a light-water reactor (LWR), ordinary water serves as both coolant and moderator. A pressurized water reactor (PWR) transfers heat to a secondary loop. A boiling water reactor makes steam in the vessel. LWR thermal efficiency is about 33–35%. Higher-temperature gas-cooled and molten salt reactor (MSR) designs reach 40–50%.

Fission or fusion. Fusion physics is understood, but the engineering is not solved. In December 2022 the National Ignition Facility reached ignition: about 3.15 MJ out from 2.05 MJ of laser energy delivered to the target. The lasers, however, drew roughly 300 MJ from the grid. tritium breeding, materials that can survive the neutron flux, and continuous operation all remain open. The chapter judges fusion unlikely to contribute meaningfully before 2050.

Radiation. The chapter covers the four types: alpha particles stopped by paper or skin, beta particles stopped by millimeters of aluminum, gamma radiation stopped by lead or concrete, and penetrating neutron radiation. Background dose averages 2–3 mSv a year, and some regions receive 10–50 mSv with no detectable effect. An abdominal CT scan gives about 10 mSv. Evidence of increased cancer risk begins above about 100 mSv. The linear no-threshold model (LNT) extrapolates risk below that level, where the evidence is absent or lost in statistical noise. acute radiation syndrome begins above about 1,000 mSv received quickly, and the untreated 50% lethal dose is about 4,000–5,000 mSv. Regulatory limits are 50 mSv a year for workers under 10 CFR (Title 10 of the Code of Federal Regulations) 20.1201, against a typical actual dose of 1–2 mSv, and 1 mSv a year for the public.

Baseload. The grid must match supply and demand at every moment. In 2024 the UK’s battery fleet (about 4.5 GWh against roughly 1,000 GWh of daily demand) covered 6–7 minutes. Germany’s (about 12 GWh against about 1,500 GWh) covered under 12 minutes. Covering one German winter week would take about 10,000 GWh, roughly 833 times the installed capacity. Round-trip efficiency (85–92%) is not the limit. Material quantity is. The book’s conclusion is that a reliable grid requires dispatchable generation, and that the non-fossil dispatchable options at scale are hydro, geothermal and nuclear. Of these, only nuclear is not tied to geography.

Chapter 2 — Uranium

Origin and abundance. Uranium formed in supernovae and neutron-star collisions, came here with the solar system about 4.6 billion years ago, and concentrated in Earth’s crust. It averages about 2.8 parts per million (ppm), roughly 35 times silver’s 0.08 ppm. The chapter’s point: “Uranium is not a rare element. It is a dispersed one.”

How deposits form. Economic ore grades start around 500 ppm (0.05%) for conventional mines. High-grade Athabasca deposits run above 1% and reach 15–20% in places. The governing mechanism is the redox boundary: uranium dissolves under oxidizing conditions as the uranyl ion and precipitates where oxygen is depleted. A roll-front deposit forms in porous aquifers, and that same geology makes in-situ recovery (ISR) possible. The chapter traces Kazakhstan’s roughly 40% share of world output to this endowment.

The isotope split, developed. The chapter introduces the breeding chain: U-238 captures a neutron and becomes U-239, which decays to neptunium-239 and then to plutonium-239 (Pu-239). Every uranium reactor produces plutonium this way. About 700,000 metric tons of depleted uranium (DU) sit at US enrichment sites. In a closed fuel cycle with fast breeder reactors, that material becomes fuel. The chapter states that its energy content would exceed all known US fossil reserves combined.

Resources and geography. Recoverable resources would supply the current fleet for well over a century on a once-through fuel cycle. Breeding extends that by a factor of sixty or more. The producers covered are Kazakhstan, Canada (McArthur River averages about 15%, and Cigar Lake is mined by jet boring), Namibia, Australia, Uzbekistan and Russia, along with Niger’s disruption after its 2023 coup. US production covers less than 1% of US requirements, which the book calls a policy and economic artifact, not a geological one. Shirley Basin returned to operation in April 2026.

Commodity and strategic material. Spot prices sat below $20/lb for most of 2014–2017 and reached $80–90/lb by 2025–2026. A plant without fuel shuts down rather than throttling back. secondary supply includes Megatons to Megawatts (500 metric tons of Russian weapons HEU converted to reactor fuel, 1993–2013), mixed oxide fuel (MOX) and re-enrichment of tails. That overhang is now largely gone. The Prohibiting Russian Uranium Imports Act (May 2024) phases restrictions in through 2028, alongside the ADVANCE Act (July 2024).

China. The US uses about 40–50 million lb of U₃O₈ a year across 93 reactors. China targets 200 GW of nuclear capacity by 2035. The WNA’s 2025 report estimated end-2024 inventories at 42,000 tU in the US, 40,000 tU in the EU and 65,000 tU in East Asia. Chinese state equity stakes in mines take supply off the open market.

Unconventional resources. Phosphate rock contains 50–200 ppm uranium, and Morocco holds about 70% of phosphate reserves. Seawater holds about 4.5 billion tonnes as the uranyl tricarbonate ion, replenished at about 30,000 t a year. Japanese adsorbent extraction works, but at $200–300/lb. The chapter calls it “a very large, very slow river.” The constraint is economic and infrastructural, not geological.

Supply story. Mine output does not equal supply. Inventories, government stocks, re-enrichment and reprocessed uranium give the system inertia in both directions. That inertia explains why the price response, when it finally came, was sharp.

Chapter 3 — Beyond Uranium

The road not taken. The uranium-centric cycle came out of 1950s–60s decisions, not physics alone. A sidebar defines terms the rest of the book depends on. Fast neutrons are born at MeV energies, moving at roughly 5% of light speed. Thermal neutrons are slowed to about 0.025 eV, about forty million times less energy. A slow neutron is far more likely to cause fission in U-235, which is the reason for a moderator.

thorium. Thorium is element 90 at about 9.6 ppm, roughly 3.4 times as abundant as uranium and more widely distributed. Th-232 is fertile, not fissile, so it needs a fissile driver. Its high-level waste returns to background in roughly 300–500 years, against tens of thousands of years for transuranic-dominated waste. The book calls this a structural difference, not a marginal one.

The transmutation pathway. Th-232 becomes Th-233, which beta-decays to protactinium-233 (half-life about 27 days), which decays to uranium-233 (U-233). Pa-233 absorbs neutrons parasitically, a problem molten salt designs can manage by removing it from the core. U-233’s η (neutrons produced per neutron absorbed) is higher in a thermal spectrum than that of U-235 or Pu-239. That allows breeding in a thermal reactor, while the U–Pu cycle needs a fast spectrum to exceed a breeding ratio of 1.0.

U-233’s companion. uranium-232 (U-232) cannot be eliminated. Its decay chain yields thallium-208 with a 2.6 MeV gamma, which forces heavily shielded remote handling, “a different industrial category.” The chapter names this as the main reason no commercial U-233 cycle exists. It also covers U-233’s weapons-usability and the counter-argument that U-232 makes weaponization difficult.

Plutonium and MOX. LWR spent nuclear fuel is about 1% plutonium, of which half to two-thirds is Pu-239. US spent fuel holds roughly 950 metric tons of plutonium. France reprocesses with PUREX (Plutonium Uranium Reduction Extraction) at La Hague and fabricates MOX at MELOX for about 24 reactors, roughly two-fifths of its fleet. That cycle is only partially closed: minor actinides go to waste, and the Astrid fast reactor meant to transmute them was cancelled in 2019.

The conversion fork. The front end is built around UF₆ (uranium hexafluoride). thorium tetrafluoride is a solid and has no place in that chain, so a thorium cycle needs a new industrial chemistry, not a retrofit.

India, China, Norway. India holds about 846,000 t of thorium and has limited uranium. After its 1974 test it was cut off from international nuclear trade. Its three-stage program runs as follows: pressurized heavy water reactor (PHWR)s on natural uranium, then fast breeders with thorium blankets (the PFBR (Prototype Fast Breeder Reactor) reached first criticality on April 6, 2026), then thorium–U-233 Advanced Heavy Water Reactors. China’s molten salt program builds on Oak Ridge’s MSRE (1965–1969), announced a Gobi prototype in 2021 and targets 2030 for a larger unit. Norway tested thorium-plutonium fuel at Halden.

Why uranium won. The weapons program came first. Oak Ridge enrichment, the Hanford plutonium reactors and Rickover’s naval PWR and BWR designs became the commercial base. Thorium does not readily yield weapons material, and in that era this counted against it. Alvin Weinberg, the main backer of the MSR, was removed in 1973. The book is careful to say the historical record supports several contributing factors. Its conclusion: uranium is not inferior, but “institutions are harder to change than physics.”

What the alternatives change. The US deferred reprocessing in 1977 and has since accumulated 95,000 metric tons of spent fuel. MOX changes the near term and thorium the long run. In the closed cycle the alternatives are components, not competitors.

Supporting appendices

• Appendix XI — Atomic Structure. Covers nuclear structure, isotopes and stability, and the strong force. It includes the binding-energy curve and the iron peak, which explain why both fission and fusion release energy. A section titled “Why U-235 and U-238 Behave Differently” explains the fission barrier: capture produces U-236, which splits, or U-239, which decays through Np-239 (about 2.4 days) to Pu-239. It also sets out the fissile/fertile distinction, which the book calls one of its organizing concepts.
• Appendix XII — Physics and Mathematics. Includes units (the eV, atomic mass unit, barn and sievert) and E = mc² with 1 u = 931.494 MeV/c². A worked example shows that one U-235 fission releases about 200 MeV, about 0.09% of the nucleus’s mass, roughly 49 million times the energy of burning one carbon atom. A comparison note reconciles the “3,000 tons of coal” ceiling with the once-through practical yield of about 20–22 tons of coal per kilogram, since only about 0.6% of uranium atoms actually fission. The appendix also covers fission reactions and Q-values, the U–Pu and Th–U breeding chains, the decay law (t½ = ln2/λ), the four-factor formula k∞ = η·ε·p·f (worked LWR example k∞ ≈ 1.39, brought to k_eff ≈ 1.00 by leakage), thermal efficiency and Carnot limits.
• Appendix XIII — Optional Problem Set. Layer 0 (0-1 The Unit Bridge, 0-2 The Mass Defect, 0-3 Energy Density — The Coal Comparison) and Layer 1 (1-1 Balance the Reaction, 1-2 The Moderator Choice, 1-S The Plutonium in Our Cask) check Part I’s claims from first principles. Problem 3-1, The Efficiency Ceiling, returns to Chapter 1’s efficiency figures.
• Appendix VIII — Glossary of Terms gives working definitions, with the chapter of first use, for every term marked here. Appendix IX — Regulatory and Policy Timeline places the 2024 legislation and the 1977 reprocessing deferral in sequence.

---

Longer technical notes and Top Stories tagged to Part I will appear on this page as they are published.