Nuclear Reactor Engineering

Why reactor engineering matters

Nuclear fission supplies about 10% of the world's electricity and roughly a quarter of its low-carbon electricity, from a fuel so energy-dense that one uranium pellet the size of a fingertip — about 7 grams — releases the energy of nearly a tonne of coal. A single 1000 MWe unit runs at capacity factors above 90%, higher than any other generating source, and refuels only every 18 to 24 months. That density and reliability is why reactors anchor baseload grids, propel submarines and icebreakers, and are being scaled down into small modular reactors (SMRs) for distributed power.

• Baseload power. Steady, dispatchable output independent of weather, unlike wind or solar.
• Low carbon. Lifecycle emissions near 12 gCO₂/kWh — comparable to wind, far below gas.
• Naval propulsion. Compact, high-enrichment cores run a decade without refueling.
• Process heat. High-temperature designs can drive hydrogen production and desalination.
• Safety-critical design. Defense-in-depth, redundancy, and negative feedback are the engineering core.

How a reactor works, step by step

Every thermal reactor is built from four functional ingredients working together inside the core:

1. Fuel. Ceramic uranium dioxide (UO₂) pellets enriched to 3–5% U-235, stacked in zirconium-alloy cladding tubes bundled into fuel assemblies. When a thermal neutron is absorbed, the U-235 nucleus splits into two fission fragments, ~200 MeV of kinetic energy (mostly deposited locally as heat), and an average of 2.43 fast neutrons at ~2 MeV.
2. Moderator. Light water, heavy water (D₂O), or graphite slows the fast neutrons by elastic scattering down to thermal energy (~0.025 eV at room temperature). This matters because the U-235 fission cross section is ~585 barns for thermal neutrons but only ~1 barn for fast ones — slowing the neutron makes the next fission hundreds of times more likely.
3. Control rods. Rods of strong absorbers — boron-10, silver-indium-cadmium, or hafnium — slide into the core to soak up neutrons and lower reactivity. Withdrawing them raises k; inserting them lowers it; a scram drops them fully in 2–4 seconds to shut down.
4. Coolant. Water (or in advanced designs, liquid sodium, helium, or molten salt) flows past the fuel to remove heat and carry it to the power-conversion system.

The steady state is a bookkeeping balance on neutrons. Each fission makes ν neutrons, but some leak out of the core, some are absorbed without causing fission, and some are captured by control materials. If exactly one neutron per fission survives to cause the next fission, the population is constant and the reactor is critical. Operators nudge k up to bring the reactor to power, then trim it back to 1 for steady running, using control rods and — in a PWR — boric acid dissolved in the coolant as a slow, uniform absorber.

The heat then goes to a Rankine steam cycle. In a PWR, pressurized primary water (155 bar, ~325 °C, never boiling) passes through steam generators that boil a separate secondary loop; that steam spins the turbine. In a BWR, water boils directly in the core at ~75 bar and the steam goes straight to the turbine. Either way, condensed water returns and the cycle repeats.

Criticality and the six-factor formula

The behavior of the chain reaction is captured by the effective multiplication factor:

k_eff = (neutrons in generation n+1) / (neutrons in generation n)

and the closely related reactivity:

ρ = (k_eff − 1) / k_eff

where ρ (rho) is dimensionless, often quoted in pcm (per cent mille, 10⁻⁵) or in dollars ($1 = β, the delayed neutron fraction). At k_eff = 1, ρ = 0 and the reactor is critical (steady). k_eff > 1 is supercritical (power rising); k_eff < 1 is subcritical (power decaying).

For a large reactor k_eff is broken into the six-factor formula:

k_eff = η · f · p · ε · P_FNL · P_TNL

Symbol	Factor	Meaning
η (eta)	Reproduction factor	Fast neutrons produced per thermal neutron absorbed in fuel
f	Thermal utilization	Fraction of thermal neutrons absorbed in fuel vs. everywhere
p	Resonance escape probability	Fraction of neutrons that slow past U-238 capture resonances
ε (epsilon)	Fast fission factor	Boost from fast fissions in U-238 (~1.03)
PFNL	Fast non-leakage	Fraction of fast neutrons that stay in the core
PTNL	Thermal non-leakage	Fraction of thermal neutrons that stay in the core

Power kinetics follow the reactor period T, the e-folding time of the neutron population: P(t) = P₀ e^{t/T}. Near critical, with delayed neutrons dominating, T ≈ (β − ρ)/(λ ρ) for small positive reactivity, where β ≈ 0.0065 is the delayed fraction and λ ≈ 0.08 s⁻¹ is the effective delayed-precursor decay constant. Push ρ above β (past $1) and the reactor becomes prompt critical: the period collapses to the microsecond prompt-neutron lifetime and power runs away. Operators therefore stay well below prompt criticality at all times.

Delayed neutrons: the reason control is possible

If a reactor ran only on prompt neutrons, the mean generation time would be about 10⁻⁴ to 10⁻⁵ seconds. A tiny 0.1% reactivity insertion would then double power in milliseconds — impossible for any mechanical rod to catch. Fission of U-235 is saved by delayed neutrons: about 0.65% of neutrons come not from the fission event but from the beta decay of certain fission products (bromine-87, iodine-137 and others) seconds to minutes later. Averaging prompt and delayed populations lifts the effective generation time to roughly 0.1 second — slow enough that control rods, boron injection, and even human operators can respond. The engineering rule is absolute: keep ρ < β so the chain reaction depends on the sluggish delayed neutrons, never the prompt ones.

Thermal-hydraulics and decay heat

Producing neutrons is only half the job; getting the heat out safely is the other half. A 3000 MW-thermal core packs a volumetric power density around 100 MW/m³ in a PWR, and the fuel centerline can reach ~1400 °C while the cladding surface stays near 350 °C. The limiting phenomenon is departure from nucleate boiling (DNB): if the heat flux exceeds the critical heat flux (~1–2 MW/m² for a PWR channel), a vapor film blankets the cladding, the surface temperature spikes hundreds of degrees, and the fuel rod can fail. The margin to DNB — the DNBR (departure-from-nucleate-boiling ratio) — is a primary safety limit.

The most under-appreciated hazard is decay heat. When the reactor trips, fission stops within seconds, but the accumulated fission products keep decaying. Immediately after shutdown, decay heat is roughly 6.5% of full thermal power; a common engineering approximation (Way-Wigner) is:

P_decay/P₀ ≈ 0.0622 · (t^−0.2 − (t + t₀)^−0.2)

with t and t₀ in seconds after shutdown and after startup. That means a 3000 MW-thermal core still makes ~200 MW the instant it trips, ~30 MW an hour later, and ~15 MW a day later — enough to melt the fuel if cooling is lost. This is exactly what happened at Fukushima Daiichi in 2011: the reactors scrammed successfully, but the tsunami knocked out the decay-heat removal, and the residual heat melted three cores over the following days.

Common misconceptions and failure modes

• "A reactor can explode like a bomb." No — 3–5% enriched fuel cannot go prompt-supercritical fast enough for a nuclear yield; the real risks are steam/hydrogen explosions and core melt.
• "Shutting it down makes it safe instantly." Decay heat requires cooling for days after trip; loss of that cooling is the classic severe-accident path.
• "More reactivity is fine as long as rods can compensate." Exceeding prompt criticality (ρ > β) makes power uncontrollable in milliseconds — the Chernobyl RBMK reactor's positive void coefficient did exactly this.
• "The moderator and coolant are separate things." In light-water reactors the water is both — which is a safety feature: lose coolant and you lose moderation, so k drops (a negative void coefficient). The RBMK's graphite moderator plus water coolant gave a dangerous positive void coefficient instead.
• "Xenon is negligible." Xenon-135, a fission product with a ~2.6 million-barn absorption cross section, poisons the core after shutdown and can trap a reactor in an unrestartable "xenon dead time" for hours.

Worked example: reactor plant efficiency

Take a PWR that fissions enough U-235 to release 3000 MW of thermal power and delivers 1000 MW of electrical power to the grid. The overall plant thermal efficiency is:

η_th = P_electric / P_thermal = 1000 / 3000 ≈ 0.33 (33%)

This ~33% is limited by the Rankine cycle's steam temperature. The Carnot ceiling for hot-side 325 °C (598 K) and cold-side 30 °C condenser (303 K) is η_Carnot = 1 − 303/598 ≈ 0.49, so real plants capture about two-thirds of that ideal. Now estimate fuel burn: each fission yields ~200 MeV = 3.2×10⁻¹¹ J, so 3000 MW = 3×10⁹ J/s needs about 9.4×10¹⁹ fissions per second — roughly 3.2 kg of U-235 fissioned per day (about 3.7 kg consumed once neutron capture is included, matching the ~1 g-per-MWd rule of thumb). Comparing the two dominant designs:

Parameter	PWR	BWR
Coolant pressure	~155 bar	~75 bar
Core outlet temp	~325 °C (no core boiling)	~285 °C (boils in core)
Loops to turbine	2 (steam generators)	1 (direct cycle)
Turbine steam	Non-radioactive secondary	Slightly activated
Reactivity trim	Rods + soluble boron	Rods + recirculation flow
Thermal efficiency	~33%	~33%