Nuclear Fission: The Reactor, the Fuel Cycle, and the Numbers

The Chain Reaction

Uranium-235 is unusual: struck by a slow neutron, it becomes unstable and splits within a fraction of a picosecond. The fragments fly apart carrying most of the released energy as kinetic energy, which becomes heat as they slam into surrounding atoms. Two or three neutrons come with them, and if on average exactly one of those goes on to split another nucleus, the reaction sustains itself at constant power.

That word 'average' is the whole of reactor control. A fraction of the neutrons are not released instantly but seconds to minutes later, from the decay of certain fission fragments. Those delayed neutrons are a small minority, and they are the reason a reactor responds on a human timescale rather than instantaneously. Without them, controlling a chain reaction with mechanical rods would be impossible.

Natural uranium is only about 0.7 percent uranium-235; the rest is uranium-238, which does not fission with slow neutrons. Most reactors therefore enrich the fuel to 3 to 5 percent, and the same centrifuge technology taken much further produces weapons material - which is why enrichment is the most closely supervised step in the whole industry.

As in fusion, the energy comes from the binding-energy curve, just approached from the other side. The scale differs from chemical combustion by roughly a factor of a million, which is why a fuel assembly the size of a person runs for years.

From Heat to Electricity

What a reactor produces is heat, and everything downstream is conventional. Water carries the heat away, becomes steam, drives a turbine and turns a generator. The thermal efficiency of that cycle is around 33 percent in a typical pressurised water reactor, set by steam temperatures rather than by anything nuclear - the same energy conversion limits that govern a coal plant.

The dominant design worldwide is the pressurised water reactor, where water under roughly 150 atmospheres stays liquid at over 300 degrees and passes its heat to a separate steam circuit. Boiling water reactors let the coolant boil directly. Both use ordinary water as moderator and coolant, which gives them a useful safety property: if the water is lost, the chain reaction stops, because the neutrons are no longer slowed enough to sustain it.

Decay heat is what remains. Even after the chain reaction stops, the fission fragments continue decaying and release several percent of full power immediately after shutdown, falling away over days. Removing that heat without external power is the design problem behind most reactor safety engineering, and it is what failed at Fukushima in 2011.

Output, Waste and Land

A reactor's defining operational number is its capacity factor - the fraction of theoretical maximum output actually delivered over a year. Across the United States fleet this has exceeded 90 percent for two decades. For comparison, utility-scale solar typically lands between 25 and 35 percent and wind between 35 and 50 percent, not because the equipment is unreliable but because the resource is intermittent. Fission is one of the few sources whose output does not depend on the weather, a property it shares with geothermal and, in most configurations, hydropower.

The waste question is genuinely two questions. By volume, spent fuel is remarkably small: all the spent fuel ever produced by US commercial reactors would cover a single football field to a depth of about ten metres. By hazard duration, it is the opposite: some isotopes remain dangerous for tens of thousands of years, which is longer than any human institution has lasted.

Finland has moved furthest on the second question. The Onkalo repository at Olkiluoto places spent fuel in copper canisters 400 metres down in crystalline bedrock, and it is the first deep geological disposal facility anywhere to reach operational status. Most other countries still store spent fuel on site, which works but was never intended as the final answer.

Land use runs the other way from renewables. A gigawatt-scale nuclear station occupies a few square kilometres; the same average output from wind requires a far larger area, though the land between turbines remains farmable.

Cost, Construction and What SMRs Are For

Fission's economics are dominated by construction. Fuel is cheap and operations are cheap, but building a large reactor in the West has repeatedly taken longer and cost more than planned - Flamanville in France and Vogtle in the United States both ran years over schedule. The interest accumulating during a decade of construction can rival the concrete and steel.

This is not universal. South Korea and China have built reactors closer to schedule and budget, largely by building the same design repeatedly with a stable workforce. That observation points at the diagnosis: the problem is first-of-a-kind construction, not the technology.

Small modular reactors take that diagnosis seriously. By shrinking output to tens or a few hundred megawatts, the components become small enough to build in a factory and ship to site. Factories learn; construction sites forget. The bet is that dozens of identical units will be cheaper per megawatt than one large bespoke plant, despite losing the economies of scale that made reactors large in the first place.

Whether that bet pays is not yet known. Several designs have regulatory approval, a handful are under construction, and the first commercial fleets will be the test. The question is one of manufacturing and supply chains rather than reactor physics, which has been settled since the 1940s.