Chapter 26 — Nuclear Energy: Reactors, Fuel Cycles, and the Future of Fission

"The energy produced by the breaking down of the atom is a very poor kind of thing. Anyone who expects a source of power from the transformation of these atoms is talking moonshine." — Ernest Rutherford (1933)

"The Italian navigator has just landed in the new world." — Arthur Compton to James Conant, confirming the first self-sustaining nuclear chain reaction, Chicago Pile-1, December 2, 1942

Chapter Overview

In Chapter 20, we studied fission as a nuclear reaction: the liquid-drop barrier, the mass distribution of fission products, the energy release of ~200 MeV per event, and the concept of a chain reaction sustained by the 2–3 neutrons released per fission. We derived the multiplication factor $k$ and saw that $k = 1$ defines the critical condition.

This chapter takes that physics into the engineered world. We will see how the abstract condition $k = 1$ is maintained — minute by minute, month by month — in a working nuclear reactor; how the nuclear fuel cycle transforms uranium ore into reactor fuel and manages the radioactive waste that results; how three major accidents exposed the consequences of getting the physics wrong; and how a new generation of reactor designs aims to make nuclear energy safer, more flexible, and more economically competitive.

Throughout, we maintain the perspective of this book: the physics is primary. Policy debates about nuclear energy are important, but they can only be evaluated honestly if you understand the underlying nuclear physics. By the end of this chapter, you will.

Fast Track: If you are familiar with basic reactor types, skim Sections 26.1–26.2 and begin at Section 26.3 (fuel cycle) or Section 26.5 (accidents). The essential new physics is in Sections 26.2.3 (xenon poisoning) and 26.4 (delayed neutrons and control).

Deep Dive: Section 26.7 (the nuclear energy debate) synthesizes physics, economics, and policy data. It is written to be read alongside the case studies, which provide extended treatments of the three accidents and of SMR economics.

26.1 Reactor Physics: From Chain Reaction to Controlled Power

26.1.1 The Multiplication Factor Revisited

Recall from Chapter 20 that the effective multiplication factor $k_{\text{eff}}$ is the ratio of the number of neutrons in one generation to the number in the preceding generation:

$$k_{\text{eff}} = \frac{\text{neutrons in generation } (n+1)}{\text{neutrons in generation } n}$$

Three regimes define the reactor's state:

A power reactor operates at $k_{\text{eff}} = 1$ during steady-state operation. To increase power, operators briefly make the reactor slightly supercritical ($k_{\text{eff}} \gtrsim 1$); to decrease power, slightly subcritical ($k_{\text{eff}} \lesssim 1$). The art of reactor control is maintaining $k_{\text{eff}}$ at exactly 1 — or deviating from 1 by precisely controlled amounts.

The reactivity $\rho$ is defined as:

$$\rho = \frac{k_{\text{eff}} - 1}{k_{\text{eff}}} = 1 - \frac{1}{k_{\text{eff}}}$$

At criticality, $\rho = 0$. Positive reactivity means the reactor is supercritical; negative reactivity means subcritical. Reactivity is often expressed in units of pcm (per cent mille, $10^{-5}$) or in dollars ($\MATH2k_\infty = \eta \, f \, p \, \varepsilonMATH3\eta = \nu \cdot \frac{\sigma_f^{\text{fuel}}}{\sigma_a^{\text{fuel}}}MATH4\eta_{\text{nat}} = \frac{2.43 \times 0.0072 \times 584}{0.0072 \times (584 + 99) + 0.9928 \times 2.68} \approx 1.34MATH5f = \frac{\Sigma_a^{\text{fuel}}}{\Sigma_a^{\text{fuel}} + \Sigma_a^{\text{mod}} + \Sigma_a^{\text{other}}}MATH6p = \exp\left(-\frac{N_{238} I_{\text{res}}}{\xi \Sigma_s}\right)MATH7\varepsilon \approx 1.02\text{–}1.08MATH8k_\infty = \eta \cdot f \cdot p \cdot \varepsilon \approx 2.02 \times 0.92 \times 0.87 \times 1.03 \approx 1.67MATH9k_{\text{eff}} = k_\infty \cdot P_{\text{FNL}} \cdot P_{\text{TNL}} = \eta \, f \, p \, \varepsilon \, P_{\text{FNL}} \, P_{\text{TNL}}MATH10P_{\text{FNL}} \approx \frac{1}{1 + B_g^2 \tau} \qquad P_{\text{TNL}} \approx \frac{1}{1 + B_g^2 L^2}MATH11k_{\text{eff}} = 1 \implies B_g^2 = \frac{k_\infty - 1}{k_\infty(\tau + L^2)} = \frac{0.67}{1.67 \times 47.3} \approx 0.00848\,\text{cm}^{-2}MATH12R_{\text{crit}} = \frac{\pi}{\sqrt{B_g^2}} = \frac{\pi}{0.0921} \approx 34\,\text{cm}MATH13n = \frac{1}{\xi} \ln\frac{E_0}{E}MATH14n = \frac{1}{0.920} \ln\frac{2 \times 10^6}{0.025} \approx 20 \text{ collisions}MATH15\alpha_v = \frac{\partial \rho}{\partial \alpha_{\text{void}}}MATH16\text{SWU} = P \cdot V(x_P) + W \cdot V(x_W) - F \cdot V(x_F)MATH17B = \frac{\text{thermal energy produced}}{\text{initial heavy metal mass}}MATH18\beta = 0.0065 \quad (\text{i.e., } 0.65\%)MATH19\bar{\ell}_d = \sum_i \frac{\beta_i}{\beta} \frac{1}{\lambda_i} \approx 12.7 \text{ s}MATH20T_{\text{prompt}} = \frac{\ell_p}{\rho} \approx \frac{10^{-4}}{0.001} = 0.1 \text{ s}MATH21\ell_{\text{eff}} = (1 - \beta)\ell_p + \sum_i \beta_i / \lambda_i \approx \ell_p + \bar{\ell}_d \cdot \beta / \beta \approx 0.1 \text{ s}MATH22T \approx \frac{\bar{\ell}_d}{\rho / \beta - 1 + \bar{\ell}_d \cdot \sum_i \lambda_i \beta_i / \beta}MATH23T \approx \frac{\bar{\ell}_d \cdot \beta}{\rho} \approx \frac{12.7 \times 0.0065}{\rho} = \frac{0.083}{\rho} \text{ s}MATH24\rho [\$] = \frac{\rho}{\beta}$$ At $\rho = 1\$$ (i.e., $\rho = \beta$), the reactor is prompt critical: the chain reaction is self-sustaining on prompt neutrons alone, and the delayed neutrons become irrelevant. The reactor period drops to milliseconds, and control is lost. Every reactor protection system is designed to ensure that $\rho$ never reaches $1\$$ under any credible scenario. > **The central insight.** Delayed neutrons are a 0.65% minority. But they slow the reactor time constant from ~0.1 s (uncontrollable) to ~10–100 s (easily controllable). Nuclear reactor control is, quite literally, built on this 0.65%. ### 26.4.3 Control Mechanisms **Control rods** contain materials with large thermal neutron absorption cross sections: | Material | Absorbing isotope | $\sigma_a$ (barns) at thermal | |----------|-------------------|-------------------------------| | Boron (${}^{10}\text{B}$) | ${}^{10}\text{B}(n,\alpha){}^{7}\text{Li}$ | 3,837 | | Cadmium | ${}^{113}\text{Cd}(n,\gamma){}^{114}\text{Cd}$ | 20,600 | | Hafnium | ${}^{177}\text{Hf}(n,\gamma){}^{178}\text{Hf}$ | 373 | | Silver-Indium-Cadmium | Alloy (80%Ag-15%In-5%Cd) | ~2,500 (effective) | Inserting a control rod increases $\Sigma_a^{\text{other}}$ in the thermal utilization factor $f$, reducing $k_{\text{eff}}$. In a PWR, control rod assemblies (containing ~20 individual rods) are held above the core by electromagnets. On a **SCRAM** (emergency shutdown), the magnets de-energize and the rods fall by gravity into the core — a fail-safe design requiring no power or operator action. **Soluble boron (chemical shim).** In PWRs, boric acid dissolved in the coolant provides slow, uniform reactivity adjustment. Boron concentration is adjusted over the fuel cycle: high ($\sim$1,200 ppm) at the beginning of cycle (fresh fuel, excess reactivity) and reduced toward zero at end of cycle. This allows all control rods to remain nearly fully withdrawn during normal operation, producing a more uniform neutron flux. ### 26.4.4 Temperature Coefficients: Inherent Safety from Physics A **negative temperature coefficient of reactivity** means that an increase in temperature causes a decrease in $k_{\text{eff}}$. This provides inherent, passive, physics-based negative feedback that stabilizes the reactor without any operator or control system intervention. The temperature coefficient has several components: 1. **Fuel temperature (Doppler) coefficient $\alpha_T^{\text{fuel}}$:** As fuel temperature rises, thermal motion broadens the ${}^{238}\text{U}$ resonances (Doppler broadening), increasing the effective resonance integral and decreasing the resonance escape probability $p$. This is always negative and is the fastest-acting feedback mechanism (responds in microseconds, as fast as the fuel heats). $$\alpha_T^{\text{fuel}} = \frac{\partial \rho}{\partial T_{\text{fuel}}} \approx -2 \text{ to } -4 \text{ pcm/°C}$$ 2. **Moderator temperature coefficient $\alpha_T^{\text{mod}}$:** As water temperature increases, water density decreases, reducing moderation. In a PWR, this is negative (less moderation $\rightarrow$ more leakage, lower $f$). Typical values: $-10$ to $-50$ pcm/$°$C. 3. **Void coefficient:** The limit of the moderator temperature coefficient — complete loss of liquid water. As discussed in Section 26.2.4, this is negative for PWR/BWR and dangerously positive for the RBMK at low power. The net effect: if a PWR's power increases for any reason, the fuel and moderator temperatures rise, $k_{\text{eff}}$ drops, and the power returns toward the setpoint. The reactor is a self-regulating system. This is not a design choice that could be omitted — **it is a regulatory requirement** in every Western nuclear licensing regime. ### 26.4.5 Xenon Poisoning: The Invisible Hand on the Throttle ${}^{135}\text{Xe}$ has the largest thermal neutron absorption cross section of any known nuclide: $$\sigma_a({}^{135}\text{Xe}) = 2.65 \times 10^6 \text{ barns}$$ This is roughly 4,500 times larger than the fission cross section of ${}^{235}\text{U}$. Even at concentrations of parts per billion, ${}^{135}\text{Xe}$ absorbs enough neutrons to significantly affect $k_{\text{eff}}$. **Production and removal of ${}^{135}\text{Xe}$:** ${}^{135}\text{Xe}$ is produced by two routes: 1. **Direct fission yield:** $\gamma_{\text{Xe}} \approx 0.003$ (only 0.3% of fissions produce ${}^{135}\text{Xe}$ directly) 2. **Decay of ${}^{135}\text{I}$:** $\gamma_{\text{I}} \approx 0.061$ (6.1% of fissions produce ${}^{135}\text{I}$, which beta-decays with $t_{1/2} = 6.57$ hours to ${}^{135}\text{Xe}$) ${}^{135}\text{Xe}$ is removed by: 1. **Radioactive decay:** ${}^{135}\text{Xe} \to {}^{135}\text{Cs}$, $t_{1/2} = 9.17$ h ($\lambda_{\text{Xe}} = 2.10 \times 10^{-5}$ s$^{-1}$) 2. **Neutron absorption (burnup):** ${}^{135}\text{Xe}(n,\gamma){}^{136}\text{Xe}$, rate $= \sigma_a \phi X_{\text{Xe}}$ The rate equations are: $$\frac{dI}{dt} = \gamma_I \Sigma_f \phi - \lambda_I I$$ $$\frac{dX}{dt} = \gamma_{\text{Xe}} \Sigma_f \phi + \lambda_I I - \lambda_{\text{Xe}} X - \sigma_a^{\text{Xe}} \phi X$$ At equilibrium ($dI/dt = dX/dt = 0$): $$X_{\text{eq}} = \frac{(\gamma_I + \gamma_{\text{Xe}}) \Sigma_f \phi}{\lambda_{\text{Xe}} + \sigma_a^{\text{Xe}} \phi}$$ At high flux ($\sigma_a^{\text{Xe}} \phi \gg \lambda_{\text{Xe}}$), the equilibrium xenon concentration saturates: $$X_{\text{eq}} \to \frac{(\gamma_I + \gamma_{\text{Xe}}) \Sigma_f}{\sigma_a^{\text{Xe}}} \approx \text{constant}$$ and the equilibrium xenon reactivity worth is typically $-2,500$ to $-3,000$ pcm ($-25$ to $-30\%$ of $k_\infty$) in a typical PWR. **The xenon transient after shutdown.** After a reactor shuts down (or reduces power), the neutron flux drops but ${}^{135}\text{I}$ continues to decay into ${}^{135}\text{Xe}$. Since the ${}^{135}\text{Xe}$ is no longer being burned by neutrons ($\phi \to 0$), the xenon concentration *increases*, peaking roughly 11 hours after shutdown at a value that can be 2–3 times the operating equilibrium. This **xenon peak** can make the reactor impossible to restart — the negative reactivity from xenon exceeds the available control rod worth. The reactor is said to be in a **xenon dead time**, which can last 24–48 hours until the xenon decays away. This phenomenon has major operational consequences: a reactor that trips unexpectedly may not be restartable for a day or more. > **The Chernobyl connection.** The operators of Chernobyl Unit 4 on April 25–26, 1986, attempted to raise power after an extended period at low power, during which xenon had built up to very high levels. Their attempt to override the xenon poisoning by withdrawing nearly all control rods set the stage for the disaster (Section 26.5.2). --- ## 26.5 Three Accidents: The Physics of What Went Wrong The history of nuclear energy includes three major accidents at power reactors. Each is a case study in how nuclear physics, engineering design, and human factors interact. Understanding the physics of these events is essential — not as an argument for or against nuclear energy, but because the lessons shaped the safety regime under which all modern reactors operate. ### 26.5.1 Three Mile Island (1979): A Partial Meltdown with No Health Consequences **The reactor:** TMI-2, a Babcock & Wilcox PWR (880 MWe), near Harrisburg, Pennsylvania. **What happened:** 1. A feedwater pump failure caused the steam generators to stop removing heat from the primary loop. 2. Pressure in the primary loop rose, opening a pilot-operated relief valve (PORV) on the pressurizer. 3. The PORV stuck open after pressure dropped, creating a small-break loss-of-coolant accident (LOCA). But the control panel indicator showed only that the *signal to close* the PORV had been sent — not that the valve was actually closed. 4. Operators, believing the system was at high pressure, reduced emergency cooling water injection to prevent (they thought) the pressurizer from going "solid" (completely filled with water, which can damage piping). 5. With inadequate cooling, the core partially uncovered. Zirconium cladding reacted with steam at high temperature ($\text{Zr} + 2\text{H}_2\text{O} \to \text{ZrO}_2 + 2\text{H}_2$, exothermic above ~1,200$°$C), generating hydrogen and causing partial fuel melting. About 45% of the core was damaged. **The physics lesson:** The accident was not caused by an exotic failure mode. It was caused by a stuck valve, a misleading indicator, and operators who overrode the automatic safety systems based on an incorrect mental model of the plant state. The containment building performed its function — despite a hydrogen burn inside the containment, no significant radioactivity escaped. **Health consequences:** The maximum dose to any member of the public was estimated at $< 1$ mSv — less than the annual natural background radiation dose. Epidemiological studies have found no statistically significant increase in cancer rates in the surrounding population. **Regulatory consequences:** The TMI accident led to the creation of the Institute of Nuclear Power Operations (INPO), mandatory control room improvements, enhanced operator training, improved emergency procedures, and the requirement for post-accident monitoring instrumentation. ### 26.5.2 Chernobyl (1986): Positive Void Coefficient Meets Operator Error **The reactor:** Chernobyl Unit 4, an RBMK-1000 (1,000 MWe), near Pripyat, Ukraine (then USSR). **What happened:** 1. On April 25, 1986, operators began a test to determine whether the momentum of the turbine generator could provide emergency electrical power during a station blackout. The test required reducing power to 700 MW(th). 2. Due to an unplanned power reduction to ~30 MW(th) — far below the intended test power — xenon accumulated to extreme levels, poisoning the reactor. Operators withdrew nearly all 211 control rods to compensate, violating operating procedures that required a minimum of 30 rods inserted. 3. With the reactor at very low power (~200 MW(th)) and almost no control rods inserted, the operators began the test by closing the turbine steam valves, which reduced coolant flow. 4. Reduced flow caused increased voiding (boiling) in the coolant channels. Due to the RBMK's **positive void coefficient**, the reactivity increased. 5. The power surged. The operators pressed the emergency SCRAM button (AZ-5), but the RBMK control rods had graphite tips (graphite displacers at the bottom of each rod). When the rods began to insert from their fully withdrawn positions, the graphite tips initially *displaced water* in the lower core, briefly *increasing* moderation and reactivity — exactly the opposite of what was needed. 6. Power surged to an estimated 100 times nominal within seconds. The fuel disintegrated, steam pressure ruptured the pressure tubes, and the 1,000-tonne reactor lid was blown off by a steam explosion. The graphite moderator caught fire (1,700 tonnes of graphite at ~1,200$°$C in air), burning for 10 days and lofting radioactive fission products into the atmosphere. **The physics in summary:** $$\underbrace{\text{Positive void coefficient}}{\text{RBMK design flaw}} + \underbrace{\text{Excessive rod withdrawal}}}} + \underbrace{\text{Graphite-tipped rods}{\text{Design flaw}} \to \underbrace{\text{Prompt supercriticality}}$$ The reactor reached prompt criticality ($\rho > \beta = 1\$$). At that point, the delayed neutrons were irrelevant, and the power excursion occurred on the prompt neutron timescale (~milliseconds).}

Consequences: 31 plant workers and first responders died of acute radiation syndrome within three months. The long-term death toll is debated: the WHO estimates ~4,000 additional cancer deaths in the most contaminated populations; the UNSCEAR 2008 report found statistically significant increases only in thyroid cancer among children exposed to ${}^{131}\text{I}$, with ~5,000 cases (mostly treatable). A 30-km exclusion zone remains in place. The total release was estimated at 5.2 EBq ($5.2 \times 10^{18}$ Bq), representing ~3.5% of the total core inventory.

26.5.3 Fukushima Daiichi (2011): When the Design Basis Is Exceeded

The reactors: Units 1–3, GE BWR Mark I design (460–784 MWe each), Fukushima Prefecture, Japan.

What happened:

1. On March 11, 2011, the Tohoku earthquake (magnitude 9.0) struck off the coast. The reactors automatically shut down (SCRAMmed) — the control rods inserted correctly, and the fission chain reaction stopped within seconds.
2. The earthquake destroyed off-site power connections. Emergency diesel generators started and provided cooling power.
3. Approximately 50 minutes later, a 14-meter tsunami overwhelmed the 5.7-meter seawall, flooding the diesel generator building and electrical switchgear in the basement. All AC power was lost: station blackout (SBO).
4. Without power, the reactor cooling systems (which require pumps) could not remove the decay heat (~6% of operating power at shutdown, declining over hours). Even though the chain reaction had stopped, radioactive decay of fission products continued to generate ~20 MW(th) per reactor.
5. Over the next three days, the water level in the reactor pressure vessels dropped, fuel was uncovered, zirconium-water reactions generated hydrogen, and all three reactor cores partially or fully melted.
6. Hydrogen accumulated in the reactor buildings and detonated, destroying the buildings above the containment (but not the primary containment vessels) of Units 1, 3, and 4 (Unit 4 received hydrogen via a shared vent).
7. Venting of containment and hydrogen explosions released radioactive material, primarily ${}^{131}\text{I}$, ${}^{134}\text{Cs}$, and ${}^{137}\text{Cs}$.

The physics lesson: The reactors shut down correctly. The chain reaction stopped. But decay heat — an immutable consequence of fission product radioactivity — continued for weeks. You cannot "turn off" decay heat any more than you can turn off radioactive decay. The failure was not nuclear but electrical: the loss of all power to drive cooling pumps.

Health consequences: No deaths from radiation exposure. The UNSCEAR 2021 report concluded that no discernible health effects are expected among the general public or the majority of workers. Roughly 154,000 people were evacuated; the social, economic, and psychological impacts of evacuation were severe and arguably caused more harm than the radiation itself (an estimated 2,300 evacuation-related deaths from stress, disruption of medical care, and suicide).

Design consequences: Post-Fukushima safety improvements include: passive cooling systems that require no electrical power (gravity-driven, natural-circulation), extended battery life, hardened venting systems, additional portable emergency equipment ("FLEX" strategy in the U.S.), and raised flood barriers. The accident accelerated the development of Gen III+ reactor designs with passive safety features (AP1000, EPR).

26.6 Advanced Reactors: Generation IV and Small Modular Reactors

26.6.1 Generation IV Concepts

The Generation IV International Forum (GIF) identified six advanced reactor concepts for development. Four are the most actively pursued:

Sodium-Cooled Fast Reactor (SFR)

• Coolant: Liquid sodium (melting point 98$°$C, boiling point 883$°$C at atmospheric pressure).
• Key physics: Fast neutron spectrum (no moderator). Can "breed" new fissile material (${}^{238}\text{U} \to {}^{239}\text{Pu}$) at a rate exceeding fuel consumption — a breeder reactor with a breeding ratio > 1. Can also burn long-lived actinide waste (transmutation).
• Advantages: Efficient use of uranium (potentially 60-100x more energy from the same ore), actinide waste reduction, high outlet temperature (550$°$C, better thermal efficiency).
• Challenges: Sodium reacts violently with water and air, requiring careful engineering. Historical cost overruns. The French Super-Phenix and Japanese Monju prototypes both experienced sodium leaks.
• Status: Russia's BN-600 (operating since 1980) and BN-800 (2016) are the only operating commercial-scale fast reactors. China's CFR-600 is under construction. India's PFBR is being commissioned.

Molten Salt Reactor (MSR)

• Fuel/coolant: Nuclear fuel dissolved directly in a molten fluoride or chloride salt mixture (e.g., LiF-BeF$_2$-UF$_4$ or NaCl-UCl$_3$). The fuel is the coolant — no fuel rods, no cladding.
• Key physics: Can operate with thermal or fast spectrum. Can be designed for thorium fuel cycle (${}^{232}\text{Th} + n \to {}^{233}\text{Th} \to {}^{233}\text{Pa} \to {}^{233}\text{U}$, another fissile fuel). Online fuel processing: fission products can be continuously removed, and fresh fuel added, without shutting down.
• Advantages: Operates at atmospheric pressure (no pressure vessel needed — the salt's boiling point exceeds 1,400$°$C), inherent safety (if the salt overheats, a freeze plug melts and the fuel drains by gravity into a subcritical dump tank), reduced waste.
• Challenges: Corrosion of structural materials by molten fluoride salts at high temperature. Tritium management (${}^{6}\text{Li}(n,\alpha){}^{3}\text{H}$). No mature fuel processing technology at industrial scale.
• Status: Oak Ridge National Laboratory operated the Molten Salt Reactor Experiment (MSRE) from 1965–1969 successfully. Multiple startups (Kairos Power, Terrestrial Energy, Moltex, ThorCon) and national programs (China's TMSR project) are developing MSR designs. Kairos Power began construction of its Hermes test reactor in Tennessee in 2024.

High-Temperature Gas-Cooled Reactor (HTGR)

• Coolant: Helium gas.
• Fuel: TRISO (TRi-structural ISOtropic) particles — tiny (~1 mm) spheres with a UO$_2$ or UCO kernel coated in layers of pyrolytic carbon and silicon carbide, a robust micro-containment. TRISO fuel can withstand temperatures exceeding 1,600$°$C without releasing fission products.
• Key physics: Very high outlet temperature (700–950$°$C), enabling industrial process heat applications, hydrogen production, and high thermal efficiency (>45%).
• Advantages: Meltdown-proof by design — the TRISO fuel retains fission products at temperatures far above any conceivable accident scenario. The reactor can lose all coolant and the core temperature stabilizes below fuel damage thresholds through radiative and conductive heat loss alone.
• Status: China's HTR-PM (two pebble-bed modules, 210 MWe total) began commercial operation in December 2023 — the world's first commercial HTGR.

Lead-Cooled Fast Reactor (LFR)

• Coolant: Lead or lead-bismuth eutectic (LBE). Lead has a boiling point of 1,749$°$C (enormous margin to boiling), is chemically inert with air and water, and provides natural shielding.
• Key physics: Fast spectrum, potential for breeding, high outlet temperature.
• Challenges: Lead is corrosive to steel at high temperatures, heavy (requiring robust structural support), and has a high melting point (327$°$C for lead, requiring heating systems to prevent solidification).
• Status: Russia has decades of experience with lead-bismuth reactors in submarine propulsion. The Westinghouse LFR concept and Newcleo's design are in development.

26.6.2 Small Modular Reactors (SMRs)

SMRs are reactors with electrical output below 300 MWe, designed for factory fabrication and modular construction. The key SMR concepts as of 2025:

NuScale VOYGR (PWR-based)

• 77 MWe per module (previously 50 MWe, uprated in 2020); up to 12 modules per plant.
• Integral design: the steam generator is inside the reactor pressure vessel (no external piping loops).
• Passive safety: Natural circulation cooling — no reactor coolant pumps needed. In an emergency, the reactor vessel is submerged in a large pool of water that removes decay heat by natural convection and boiling for at least 72 hours with no power and no operator action.
• Received NRC design certification in January 2023 — the first SMR to achieve this milestone in the U.S.
• The Carbon Free Power Project (CFPP) at Idaho National Laboratory, the lead customer, was canceled in November 2023 due to cost escalation (from $5.3 billion to $9.3 billion for 462 MWe). NuScale continues to pursue other customers internationally.

GE-Hitachi BWRX-300 (BWR-based)

• 300 MWe, based on the proven ESBWR design but simplified.
• Natural circulation cooling, no recirculation pumps.
• 60% reduction in building volume compared to conventional BWR.
• Under construction: Ontario Power Generation's Darlington site (Canada), with planned operation by late 2029. Selected for projects in Poland, the Czech Republic, and the UK.

Other notable designs:

• X-energy Xe-100: Pebble-bed HTGR, 80 MWe, TRISO fuel. Selected by Dow Chemical for industrial process heat.
• TerraPower Natrium: Sodium-cooled fast reactor with molten salt energy storage, 345 MWe. Under construction in Kemmerer, Wyoming (DOE co-funded), planned operation ~2030.

The SMR value proposition:

• Lower capital cost per unit (though not necessarily per MWe): factory fabrication should reduce construction risk and schedule.
• Scalability: Add modules as demand grows.
• Siting flexibility: Smaller footprint, potentially suitable for remote communities, industrial sites, or replacement of retiring coal plants (many coal plant sites have existing transmission infrastructure and cooling water).
• Passive safety: Simplified designs with fewer systems that can fail.

The SMR challenge: No SMR has yet demonstrated economically competitive electricity production. The fundamental economic challenge is that nuclear power plants have historically shown positive economies of scale — larger plants produce cheaper electricity per MWh. Whether factory learning curves and simplified construction can overcome this remains unproven. The cancellation of the NuScale CFPP project underscored this uncertainty.

26.7 The Nuclear Energy Debate: Physics and Data

Nuclear energy is one of the most polarizing topics in energy policy. As physicists, our contribution to the debate should be data and analysis, not ideology. Here are the key dimensions:

26.7.1 Climate

Nuclear power produces essentially zero direct greenhouse gas emissions during operation. Life-cycle emissions (including mining, construction, enrichment, and decommissioning) are estimated at 5–12 g CO$_2$/kWh — comparable to wind (~7–15 g CO$_2$/kWh) and far below natural gas (~400–500 g CO$_2$/kWh) or coal (~800–1,000 g CO$_2$/kWh).

As of 2024, nuclear energy provides approximately 10% of global electricity (~2,700 TWh/year) and avoids roughly 1.5 Gt CO$_2$/year compared to a counterfactual gas-fired replacement. The International Energy Agency's net-zero scenarios generally require a doubling of nuclear capacity by 2050.

26.7.2 Safety: Deaths per TWh

Despite three major accidents, the safety record of nuclear power, measured in deaths per unit of energy produced, is the best of any major energy source:

Source: Markandya & Wilkinson (2007), Lancet; updated by Our World in Data (2023) using Sovacool et al. and WHO data.

The data are striking: coal kills roughly 800 times more people per TWh than nuclear. The fear of nuclear energy is real and understandable (radiation is invisible and the word "nuclear" carries profound historical associations), but it is not supported by the mortality data.

26.7.3 Waste

The volume of nuclear waste is small compared to waste from fossil fuels (coal ash, CO$_2$). But the radioactivity is intense and long-lived:

• Fission products (${}^{137}\text{Cs}$, ${}^{90}\text{Sr}$): Dangerous for ~300 years (10 half-lives)
• Long-lived actinides (${}^{239}\text{Pu}$, ${}^{237}\text{Np}$): Dangerous for ~100,000–1,000,000 years

A 1 GWe PWR operating for one year produces approximately 20 tonnes of spent fuel, containing:

• ~950 kg of fission products
• ~250 kg of plutonium (all isotopes)
• ~18.7 tonnes of uranium (mostly ${}^{238}\text{U}$, with ~0.8% ${}^{235}\text{U}$)
• ~25 kg of minor actinides (Np, Am, Cm)

The total high-level waste from 70 years of global nuclear power (if not reprocessed) is roughly 400,000 tonnes — a quantity that sounds large but is physically small (nuclear fuel is extremely dense). The waste problem is not primarily technical — the Finnish and Swedish repository programs demonstrate that deep geological disposal is feasible — but political and institutional.

26.7.4 Cost

Nuclear power has a well-documented history of cost overruns and schedule delays in Western countries:

The contrast between Western projects and the UAE, South Korean, and Chinese construction programs is striking. South Korea's APR-1400 (the design used at Barakah) was built on time and on budget, suggesting that the cost problem is not inherent to the technology but related to regulatory frameworks, industrial capacity, supply chain continuity, and institutional learning. Countries that maintain continuous construction programs build cheaper reactors.

Levelized cost of electricity (LCOE) for new nuclear in Western countries: ~$90–160/MWh, compared to ~$25–50/MWh for onshore wind and utility-scale solar (before integration costs). However, comparing nuclear to intermittent renewables on LCOE alone is misleading: nuclear provides firm, dispatchable, 24/7 baseload power. The system cost of high-renewable grids (including storage, backup generation, and grid reinforcement) narrows the gap substantially.

26.7.5 Proliferation

Every nuclear reactor produces plutonium. The question is whether the plutonium is accessible for weapons use:

• Spent fuel from a PWR/BWR contains reactor-grade plutonium (~60% ${}^{239}\text{Pu}$, with substantial ${}^{240}\text{Pu}$, ${}^{241}\text{Pu}$, ${}^{242}\text{Pu}$). A nuclear explosive device using reactor-grade plutonium is theoretically possible but technically much more difficult, less reliable, and produces a lower yield than one using weapons-grade plutonium (~93% ${}^{239}\text{Pu}$).
• The enrichment route is arguably a greater proliferation concern than reactors themselves: a centrifuge cascade that can produce LEU (3–5%) can, with reconfiguration, produce HEU (>90%).
• The IAEA safeguards system monitors nuclear materials through accounting, containment, surveillance, and inspection. The system is effective but imperfect — it can detect diversion but cannot physically prevent it.

The proliferation risk of nuclear power is real but manageable, and it is worth noting that the five nuclear-weapon states acquired their weapons through dedicated military programs, not through civil power reactors.

26.8 Chapter Summary

Nuclear reactors are engineered systems built on the physics of the fission chain reaction, neutron moderation, and radioactive decay. The key physics concepts are:

1. Criticality is maintained by balancing neutron production (fission) against neutron loss (absorption + leakage), described by the four-factor and six-factor formulas.

2. Delayed neutrons (0.65% of total for ${}^{235}\text{U}$) slow the reactor time constant from milliseconds to seconds, making mechanical control possible. Prompt criticality ($\rho \geq \beta$) means loss of control.

3. Negative temperature coefficients — particularly the Doppler broadening of ${}^{238}\text{U}$ resonances — provide inherent, physics-based, self-regulating negative feedback. The RBMK's positive void coefficient was a design flaw that had catastrophic consequences.

4. Xenon-135 poisoning ($\sigma_a = 2.65 \times 10^6$ barns) affects reactor operation, restart capability, and contributed to the Chernobyl accident sequence.

5. The nuclear fuel cycle — from uranium mining through enrichment, irradiation, and waste management — determines the economics, environmental impact, and proliferation risk of nuclear power.

6. The three major accidents each teach a different lesson: TMI (operator error + misleading instrumentation), Chernobyl (fundamentally unsafe design + operator violations + positive void coefficient), Fukushima (beyond-design-basis external event + station blackout + decay heat).

7. Advanced reactors (Gen IV, SMRs) aim to address safety, waste, cost, and flexibility through passive safety features, alternative coolants, and modular construction — but remain largely unproven at commercial scale.

Nuclear energy is the densest, most carbon-free, and statistically safest source of baseload electricity available today. Whether it will play an expanded role in the world's energy future depends not on the physics — which is well understood — but on economics, public acceptance, regulatory efficiency, and political will.

In Chapter 27, we turn to nuclear medicine — where the same radioactive decay physics that makes nuclear waste a challenge becomes a life-saving tool for diagnosis and therapy.