☢️ Reactor Physics: Criticality, Control, and Safety
A chain reaction is a knife edge between two exponentials, in a machine nobody can see inside — it should be far beyond human control. It isn't, because about 0.65% of the neutrons show up late. Build
What you’ll learn
- Two Hundred Million Volts, and Two and a Half NeutronsEstablish what a single fission event produces, and frame the whole course around the number 2.5.A thermal fission of U-235 releases about 202.79 MeV — some 85% of it as the kinetic energy of two recoiling fragments — and, crucially, about 2.5 free neutrons. Since it takes only one neutron to cause a fission, a self-sustaining chain reaction requires that exactly one of those 2.5 survives to cause the next one. Everything in reactor physics is bookkeeping on that sentence.
- The Barn: A Cross-Section Is a ProbabilityCorrect the central misconception about cross-sections and introduce the barn as the currency of neutron bookkeeping.A neutron cross-section has units of area but is not a target size — it is a probability of interaction, expressed as the area a nucleus would need if it behaved like a solid disc. Measured in barns (10⁻²⁴ cm²), cross-sections vary wildly with neutron energy and isotope: U-235's thermal fission cross-section is 583 barns against just 1 barn for fast neutrons.
- Moderation: Slowing a Neutron Without Eating ItDerive why light nuclei make good moderators and explain the trade between slowing-down power and absorption.Slowing a neutron means having it collide with something. Billiard-ball mechanics dictates that energy transfer is greatest when the masses match — so hydrogen, whose nucleus is a single proton, is the most efficient moderator per collision. But hydrogen also absorbs neutrons, which is why heavy water and graphite exist: they moderate less efficiently but eat far fewer neutrons, and that trade determines what fuel a reactor can run on.
- The Ledger: k-effective and the Six FactorsAssemble the neutron economy into k = η·f·p·ε·P_FNL·P_TNL and show that criticality is an accounting identity.The six-factor formula decomposes k into the sequential fractions that survive each stage of a neutron's life: fast fission gain, leakage twice, resonance escape during slowing-down, competition for the thermal neutron, and neutrons produced per absorption in fuel. Typical thermal reactor values multiply out to almost exactly 1 — which is not a coincidence but a design requirement.
- Criticality, PreciselyFix the vocabulary: critical means steady, not dangerous — and supercritical is a routine operating state, not an explosion.Critical means k = 1 and the neutron population is constant: it is the normal, desired, boring state of every operating reactor. Supercritical means k > 1 and the population is growing — which is simply what a reactor does whenever it increases power, several times a week. The words describe rates of change, not hazards, and the popular usage inverts them.
- The Rounding Error That Makes Control PossibleDeliver the course's central idea: delayed neutrons lengthen the effective generation time by orders of magnitude, and that is why reactors are operable.About 0.65% of fission neutrons are not emitted promptly but appear seconds to a minute later, from the decay of a few unstable fission fragments. Because the chain must pass through those slow links, the reactor's effective generation time rises from ~10⁻⁴ seconds to ~0.1 seconds — turning an ungovernable machine into one a human can operate. Prompt criticality, where k reaches 1 on prompt neutrons alone, is the line where that gift is withdrawn.
- The Three HandlesExplain control rods, soluble boron and burnable poisons as three deliberate answers to three different timescales.Operators hold k at 1 with three tools that differ mainly in speed. Control rods are fast and local — seconds, and the only thing that can scram. Soluble boron is slow and uniform — hours, used to follow the fuel's burnup over months. Burnable poisons are not controlled at all: they are absorbers designed to be consumed at the same rate the fuel is, cancelling out fresh fuel's excess reactivity automatically.
- Feedback: The Reactor That Argues BackExplain reactivity coefficients — Doppler, moderator temperature, void — as the reactor's built-in self-control, and why their sign is the whole safety case.Reactors are not held at criticality by operators alone; they are held there by their own physics. The Doppler effect broadens U-238's absorption resonances as fuel heats, cutting reactivity within milliseconds — faster than any control system. Combined with negative moderator temperature and void coefficients, a Western reactor pushes back against every power excursion. Chernobyl's reactor had a positive void coefficient, and that sign is the difference.
- Xenon: The Poison That WaitsExplain xenon-135 poisoning, the iodine pit, and why it is an operational trap rather than a mere curiosity.Xenon-135 has a thermal absorption cross-section of about 2.6 million barns — the most powerful known neutron absorber — and it is produced largely by the decay of iodine-135, which has a 6.57-hour half-life. So xenon keeps being made after a power reduction while nothing is left to burn it out, peaking about 11.1 hours later. The result is a reactor that cannot restart when the operator wants it to, and a temptation that has proved lethal.
- Why a Power Reactor Is Not a BombEstablish rigorously that a commercial reactor cannot produce a nuclear explosion, and be equally precise about what it can do.A nuclear explosion requires a fast chain reaction in highly enriched material assembled faster than it can blow itself apart — tens of generations in under a microsecond. A power reactor is thermal, enriched to a few percent, moderated by a material that leaves if things get hot, and has a prompt negative feedback that acts in milliseconds. The failure modes are real and serious; a nuclear detonation is not among them.
- Three Mile Island: The Machine Told the Truth BadlyAnalyse TMI as an information failure, not a physics failure, and show what it changed.At Three Mile Island the physics did everything right: the reactor scrammed correctly and the accident was pure decay heat management. What failed was information — a relief valve stuck open while its indicator light reported only that a solenoid was powered, and a pressurizer level reading that told operators the system was full while it was emptying. It took about 80 minutes to diagnose; roughly half the core melted; containment held.
- Chernobyl: Every Coefficient Pointing the Wrong WayExplain Chernobyl as the physics of this course running in a machine built with the opposite signs, compounded by organisational failure.The RBMK had a positive void coefficient because graphite moderated and water mainly absorbed, so boiling raised power. Its control rods had graphite tips that briefly *added* reactivity on insertion — a positive scram effect known since 1983 and not disclosed to operators. A reactor deep in a xenon pit, at low power, with most rods withdrawn, was then scrammed: the design's own shutdown system triggered the excursion.
- What the Physics Buys, and What It CostsClose the through-line and hand the reader the ability to reason about nuclear power without handing them a conclusion.Reactor physics is a chain of hostile facts arranged into a controllable machine: 2.5 neutrons reduced to exactly one, a 580-fold cross-section penalty answered by moderation, and a knife-edge made operable by 0.65% of neutrons arriving late. The safety case rests on the signs of coefficients rather than on operators. The costs — waste, capital, and a fat-tailed accident risk with a genuinely contested death toll — are real, and the point is to be able to follow the argument, not to be handed its conclusion.
Questions this course answers
About 85% of a fission's energy appears as:
Roughly 169 MeV of the 202.79 MeV is the two fragments recoiling from each other's electrostatic repulsion once the nucleus stretches far enough that the short-ranged strong force can no longer hold the lobes together. They travel a few micrometres, slam into the ceramic lattice and stop. A power reactor is literally a machine for stopping fast debris in a solid and collecting the friction. Neutrinos take about 3% and leave the planet — energy you can never have.
In reactor physics, the number 2.5 matters more than 200 MeV because:
The 200 MeV tells you the reactor will be hot; it says nothing about whether it will run. The 2.5 is the whole design problem: each fission hands you 2.5 chances to cause the next one, exactly one of which must land. Every material, dimension and rod in the machine is an answer to what becomes of the surplus — fission, absorption, or leakage, and there is no fourth option.
A reactor core has a minimum size because:
Leakage is one of the three fates of a neutron, and it is a surface effect while fission is a volume effect. Shrink the core and the surface-to-volume ratio rises until neutrons reach the edge and depart before they hit anything. Below the critical size no arrangement of that fuel will sustain a chain reaction — a fact about shape and dimensions rather than about the uranium, which is why reflectors that bounce neutrons back reduce the critical size.
The stable operating state of a chain reaction (k = 1) is unnerving because:
There is no basin of attraction: k<1 dies exponentially, k>1 grows exponentially, and only the exact value 1.000 runs steadily. A quantity balanced on a point, pushed around by a hundred physical effects, inside a machine nobody can see into, should not be controllable by a human with an analogue meter. That it is comes down to a rounding error — about 0.65% of the neutrons arrive late — which is the fact this whole course is built toward.
Xenon-135's absorption cross-section is 2.6 million barns — vastly larger than any nucleus. This proves that a cross-section:
If cross-sections were sizes, a Xe-135 nucleus would be about a millimetre across. The real definition is operational: it is the disc area a nucleus would need if it caught every neutron that struck it, in order to produce the interaction rate you actually measure. Slow neutrons behave as waves that need only overlap the nucleus, and when their energy matches a resonance the probability spikes by orders of magnitude. Think radio tuning, not archery.
U-235's fission cross-section is 583 barns thermal versus about 1 barn fast. The consequence for reactor design is:
Fission neutrons are born fast, where they are nearly useless at 1 barn, and must reach thermal energies where they are 583 barns of use. Slowing them by a factor of about ten thousand without eating them is the moderation problem, and it dictates the moderator, the lattice geometry and much else. Fast reactors are not impossible — they exist — but they must compensate for that 580-fold penalty with much higher enrichment and a very different neutron economy.
Grounded in trusted sources
- Wikipedia — Nuclear fission (202.79 MeV per U-235 thermal fission; ~169 MeV as fragment kinetic energy; ~4.8 MeV in neutrons averaging ~2 MeV each; ~7 MeV prompt gammas; ~3% in neutrinos; ~2.5 neutrons per thermal fission): https://en.wikipedia.org/wiki/Nuclear_fission
- Wikipedia — Neutron cross section (1 barn = 10⁻²⁴ cm²; U-235 thermal fission 583 barns versus 1 barn fast; U-238 absorption 2 barns; hydrogen scattering 20 barns; the 1/v law): https://en.wikipedia.org/wiki/Neutron_cross_section
- Wikipedia — Six factor formula (k = η·f·p·ε·P_FNL·P_TNL; typical thermal reactor values η = 1.65, f = 0.71, p = 0.87, ε = 1.02, P_FNL = 0.97, P_TNL = 0.99; criticality conditions): https://en.wikipedia.org/wiki/Six_factor_formula
- Wikipedia — Delayed neutron (β = 0.0064 for U-235; six precursor groups totalling ~0.00657, with half-lives from 55.72 s down to 0.230 s; delayed neutrons are essential to inherent reactor safety): https://en.wikipedia.org/wiki/Delayed_neutron
- Wikipedia — Prompt neutron (prompt neutrons emitted within 10⁻¹³–10⁻¹⁴ seconds; β = 0.0064 for U-235): https://en.wikipedia.org/wiki/Prompt_neutron
- Wikipedia — Prompt criticality (a critical assembly is prompt-critical if it is critical without any contribution from delayed neutrons; SL-1's delayed neutron fraction of 0.70% and its 4-millisecond period): https://en.wikipedia.org/wiki/Prompt_criticality
- Wikipedia — Iodine pit (Xe-135 thermal absorption cross section 2.6×10⁶ barns; I-135 half-life 6.57 h; Xe-135 half-life 9.2 h; peak Xe-135 about 11.1 hours after a power decrease; the RBMK's size making a non-uniformly poisoned core hard to control at low power): https://en.wikipedia.org/wiki/Iodine_pit
- Wikipedia — RBMK (positive void coefficient because light water is an absorber while graphite is the moderator; steam is about 1/1350 as dense as liquid water; graphite-tipped control rods with a 4.5 m displacer and a 1.25 m water column; the positive scram effect discovered in 1983 at Ignalina; post-accident modifications): https://en.wikipedia.org/wiki/RBMK
Every Wunder lesson is built from real, reputable sources — never invented.
Related Science courses
Wunder is a personalized learn-anything platform — tell it any topic and it builds a beautiful, fact-checked course in minutes, with narration, a knowledge check, and a college-style University track.
Browse more Science courses · All topics · Home
© 2026 Wunder Learning LLC · Terms & Privacy