From Heat to Grid: Anatomy of a Nuclear Plant

Strip away the containment dome and the security fence, and a nuclear plant is a thermal power station: heat makes steam, steam spins a turbine, the turbine spins a generator, and the generator pushes current onto the grid. Everything distinctive about nuclear is packed into the first step.

The heat source is the reactor core: tens of thousands of fuel rods, each a zirconium-alloy tube about the height of a two-story house and the width of a finger, packed with ceramic pellets of uranium dioxide. Rods are bundled into fuel assemblies — a large pressurized-water reactor core holds roughly 150–200 of them. Water flows up through the assemblies, carrying away the heat of fission. The same water usually does double duty as the moderator, slowing neutrons so the chain reaction can sustain itself — an elegant safety coupling, because losing the coolant also means losing the moderator, which kills the chain reaction (though not, remember, the decay heat).

From the core outward, the path to the grid runs through a steam supply system (the next page covers the two American variants — PWRs make steam in a separate loop, BWRs boil right in the core), then the turbine: typically one high-pressure stage and several low-pressure stages on a single shaft, spinning at 1,800 rpm in most US plants. The shaft turns the generator, producing electricity at around 20–25 kilovolts, which the plant's main transformer steps up to transmission voltage (345–765 kV) at the switchyard. From there it's the same grid physics as any other power station.

Two numbers describe a plant's size, and the difference between them matters. Thermal power (MWt) is the heat the core produces; electrical power (MWe) is what the generator sends out. The ratio — thermal efficiency — is about one-third for light-water reactors, set by steam temperature and the laws of thermodynamics, not by any nuclear peculiarity. A "1,000 MWe" reactor is really a ~3,000 MWt heat engine rejecting ~2,000 MW of low-grade heat to the environment. That waste heat is why every thermal plant needs a heat sink: a river, a lake, the ocean, or the iconic natural-draft cooling tower. The towers emit only water vapor — the white plume is cloud, not smoke, and not anything radioactive. (Palo Verde, in the Arizona desert, famously solves the heat-sink problem with treated municipal wastewater — the only large nuclear plant in the world cooled that way.)

Around this steam engine, a nuclear plant wraps layers that fossil plants don't have. The core lives inside a steel reactor pressure vessel with walls around 20 cm thick. The vessel and the primary loop live inside the containment — a steel-lined, heavily reinforced concrete structure designed to hold the pressure and radioactivity of a worst-case pipe rupture. Multiple independent emergency core cooling systems stand ready to flood the core if normal cooling fails, backed by redundant emergency diesel generators because the cooling pumps need power even when the grid is gone. The safety page explains the philosophy — defense in depth — that organizes all of this.

Operationally, the striking thing about a nuclear plant is its steadiness. Fuel cost is a small fraction of its operating cost, and the machine is happiest at constant temperature, so US plants run flat-out around the clock and shut down only on an 18–24 month rhythm to swap fuel. On the map's history charts you can see exactly this signature: long plateaus at 100%, punctuated by rectangular notches of refueling. The capacity-factor and outage pages put numbers on it.