EN

Nuclear Power Plant Output Calculator

The figure on a nuclear plant's nameplate — its rated electrical power — is never delivered continuously all year round: scheduled refuelling outages, decade maintenance overhauls and output adjustments mean actual production is worked out through a capacity factor, not simply "power × 8,760 hours".

Like any thermal plant, a nuclear plant converts only part of its thermal power (released by fission) into electricity: the rest is necessarily rejected as heat, a direct consequence of the second law of thermodynamics applied to the conversion cycle (the Rankine cycle).

This page calculates the output of an existing plant from its rated power — not the physics of fission (critical mass, reactivity, radioactivity), which belongs in the Physics category.

Rated power and operation

The plant

Commercial reactors worldwide range from about 900 MW for older units up to roughly 1,600 MW for the latest EPR design.

Share of the year at rated power. Typical nuclear fleets: around 75 to 90 % depending on the year (maintenance outages); some national fleets with heavier maintenance schedules run lower, around 65 to 80 %.

Fraction of thermal power converted into electricity. Pressurised water reactors (PWR): roughly 33 to 37 %.

Reactor2,727 MW thTurbine +generator900 MW elecElectrical grid1,827 MW thWaste heat (cooling)
Principle diagram (example values). The diagram redraws with the entered values after the calculation.

Reactor jargon with an asterisk is unpacked in the glossary at the end.

Nuclear capacity factors year to year

The capacity factor* is actual energy sent out over a period, divided by what the plant would have made if it had sat at full rated power the whole time. For a nuclear fleet that is not fighting a maintenance backlog, the long-run figure is typically around 75%, and a quiet year can push it toward 85%. It is not a score for how “good” the reactor is. It is a diary of how the year went.

Three things move it.

  • Outage scheduling. Every unit comes offline to refuel and to let inspectors into the core. Those outages are planned, but they do not all last the same number of days — a few weeks for a straightforward reload, several months when a decade overhaul is on the ticket.
  • Weather and river (or sea) temperature. A station that rejects heat to a river can be told to back off in a heatwave, to keep the discharge inside its thermal consent. Coastal plant is less exposed, not immune.
  • The grid. When wind and hydro are pouring in, the operator may turn the reactor down on purpose. That is load-following, and it knocks the annual factor even if the kit itself was available.

A 900 MW unit at 75% is about 5.9 TWh in a year. Same unit at 65% after a long outage: closer to 5.1 TWh. The nameplate never changed. The factor is typed; the year writes it.

Eyr = Pelec × 8,760 h × capacity factor

Refuelling outages and decade overhauls

A pressurised-water reactor* typically swaps out about a third of its fuel assemblies every 12 to 18 months. A standard refuelling outage lasts 30 to 40 days, during which output is zero. That already takes roughly 8% to 11% off a perfect year before anyone mentions extra work.

The real dent is the decade overhaul — in UK terms, the major statutory outage that sits alongside the periodic safety review. Every ten years or thereabouts the vessel internals are inspected, steam generators may be replaced, and hydraulic tests plus heat-up trials have to be signed off. Those campaigns can keep the unit offline for 3 to 6 months. Fleets stagger the big outages; the whole system then does not sag in the same calendar year.

  • Inspect the reactor vessel and internals for integrity.
  • Replace major components (steam generators are the classic item).
  • Run hydraulic tests and heat-up trials before the unit is allowed back on the bar.

Estimating next year’s TWh means asking whether this is a “normal reload” year or a “vessel open” year. Mixing the two averages is how people over-promise.

The Rankine cycle and the thermodynamic limit

A nuclear plant is, first, a thermal plant. Fission heats the primary coolant; that heat makes steam in a secondary loop; the steam drives a turbine coupled to a generator. The cycle is Rankine*. The second law puts a ceiling on it, set by the hot-source temperature (around 300 °C in a PWR) and the cold sink (ambient, around 15 °C). That ceiling is the same Carnot efficiency that bounds any heat engine.

ηmax = 1 − Tcold / Thot (temperatures in kelvin)

ηmax ≈ 1 − 288 / 573 ≈ 0.497, i.e. roughly 50% before irreversibilities

In practice, friction, leaks and pumping work pull real PWR efficiencies down to about 33% to 36%. The EPR, running a little hotter and at a slightly higher pressure, aims around 37%. That is why a 900 MW electrical unit is sitting on roughly 2,700 MW of heat: two-thirds of the fission energy never becomes a kilowatt-hour. It leaves as waste heat, which is the next section.

There is no tuning past that ceiling. A hotter steam cycle would help; the fuel and the pressure vessel set how hot the plant is allowed to go.

Open-loop cooling and cooling towers

The cold sink of the Rankine cycle is the condenser cooling system. Two families exist worldwide, and the UK uses both ideas — seawater once-through on the coast, towers inland where a river cannot take the heat.

  • Open-loop (once-through). Water from a river, lake or the sea is pumped through the condenser and returned 5 to 15 °C warmer. Better for plant efficiency; hungry for a withdrawal and thermal-discharge permit, especially in a dry summer.
  • Closed-loop with cooling towers. The water recirculates. Heat leaves as evaporation in the big hyperbolic towers (the ones that grow a vapour plume). Far less freshwater taken, a slightly warmer sink and a slightly lower efficiency, and a visible cloud on a cold morning.

Coastal UK stations — Sizewell, Hinkley, Hartlepool, Heysham, Torness — reject heat to the sea. Inland plant lives or dies by residual-flow and temperature rules. Neither arrangement changes the Rankine arithmetic. They only change how close the plant sits to the cold-sink temperature assumed.

Nuclear capacity factors versus wind and solar

Capacity factor is not a quality badge. It is a statement about the fuel. Nuclear stores its fuel on site and can run when asked. Wind and solar wait on the weather.

  • Nuclear: long-term average around 75% (up to 85% in a good year). Dispatchable, fuel on the pad.
  • Onshore wind: 24% to 28% in a temperate climate. Offshore does better, around 40% to 45%, because marine winds are stronger and less fidgety. See the wind turbine power calculator for the cube of wind speed that sits behind those figures.
  • Solar PV: 12% to 16% as an annual average in the same climate, with a summer midday peak and a winter slump.

Those gaps are why nuclear is treated as baseload and why wind and solar need storage or flexible plant beside them. A low nuclear factor in a given year is usually an outage story, not a “the wind didn’t blow” story.

Reactor size classes

Commercial light-water units come in a handful of familiar electrical ratings. The numbers below are the ones on nameplates and in this form.

  • 900 MWe class — the workhorse of older PWR fleets. Net efficiency around 33%.
  • 1,300 MWe class — a step up, efficiency around 34.5%.
  • 1,450 MWe class — fewer units, around 35%.
  • EPR (European Pressurised Reactor) — about 1,650 MW gross, target efficiency around 37%. Extra safety redundancy and a slightly hotter steam cycle.

Those are gross figures at the generator terminals. Net sent-out is a few percent lower once the pumps, HVAC and the rest of the house load are subtracted. Sizewell B, the UK’s only operating PWR today, sits near 1,200 MW. Hinkley Point C’s two EPRs are in the 1,600 MW class each. The remaining AGRs* are smaller, typically around 600 MW per unit — a different family, covered further down.

Load-following

PWRs can move. Not like a gas turbine, but they are not stuck at 100% either. Typical range is 20% to 100% of rated power, with ramps of about 1% to 5% per minute.

  • Boric acid in the primary circuit — slow, campaign-length reactivity, the “long handle”.
  • Control rods — the fast trim, in real time.

Load-following shaves the capacity factor because the turbine is no longer sitting at its best efficiency point all year. French PWRs do this routinely around the daytime peak and when wind surges. The GB fleet historically has not.

“Can” is not “will”. The fuel cycle, xenon after a drop, and the grid contract all have a vote. Modelling a UK station as if it load-followed like a French 1,300 MW unit will understate the TWh.

What fission physics adds to this output calculation

The civil engineering sets the envelope. The nucleus sets the energy per gram. This calculator does not replace a core follow; it converts a rated electrical power, a capacity factor and an efficiency into TWh and waste heat. The physics is why those three numbers are the ones that matter. Broader physics pages cover the underlying quantities; this page stays with output.

  • Fission yield. Each fission of a 235U nucleus releases about 200 MeV, i.e. 3.2×10−11 J. Burnup is quoted in MW·d/t (megawatt-days per tonne of uranium). A 900 MW reactor uses roughly 25 tonnes of enriched uranium a year for a thermal power of about 2,700 MWth.
  • Conversion to plutonium. Some neutrons are captured by 238U and make 239Pu, which itself fissions and can contribute 20% to 30% of the energy toward the end of the cycle. That stretches the fuel’s useful life.
  • Temperature feedback (void coefficient, Doppler broadening). Negative in a PWR: the reactor fights its own power rise. Core codes fold that in when they predict the power that can actually be asked for today.

Next Tuesday’s megawatts need the core model, the outage diary and the thermal-hydraulic limits. A year’s TWh from a nameplate is this page’s job.

Hinkley Point C and Sizewell C

Hinkley Point C, on the Somerset coast, is two EPR units — the same 1,600 MW class that can be typed into the form. Combined capacity is around 3.2 GW. Sizewell C in Suffolk is the proposed copy: two more EPRs, same thermal island, same seawater cooling idea. Both are meant to run for decades at a capacity factor well above the old AGR average.

Run this page on 1,630 MW, 90% and 37% and a sense of what one of those units is meant to send out in a settled year appears. The construction programme is not the calculation. The calculation is what the nameplate is for, once the unit is actually on the bar.

A first-of-a-kind outage programme will not look like year 15. Treat early-life capacity factors as a guess, not a promise. The waste-heat figure will be honest from day one: at 37% the plant rejects about 1.7 MW of heat for every megawatt sent to the grid.

AGR versus PWR: the UK's two reactor families

The UK did not build a PWR fleet. It built advanced gas-cooled reactors: graphite moderator, carbon-dioxide coolant, a different temperature and a different outage culture. The steam cycle remains Rankine. Sizewell B (1995) is the exception — a Westinghouse PWR, the one unit that looks like the 900–1,300 MW class above.

AGRs typically sit around 600 MW electrical. Their thermal efficiency is a little better than a PWR’s because the gas is hotter. Availability has been the sore point: graphite-brick inspections, boiler issues, and long statutory outages have pulled several units well below the 75% textbook factor.

When this calculator is used on an AGR, a PWR’s 85% should not be pasted in. Look at the last few years of sent-out TWh and back-calculate the factor. The new-build programme — Hinkley, Sizewell C, and whatever SMR design eventually gets a licence — is PWR or PWR-like. The AGR fleet is a closing chapter.

Rated electrical power is a licence figure, not a guaranteed TWh. An AGR held off for graphite work, or a PWR in a decade overhaul, produces nothing that month whatever is typed in the box.

The Office for Nuclear Regulation

Output is not only a physics number. In Britain the Office for Nuclear Regulation* licenses the site, the operating rules, and the outage. If ONR holds a unit off after a finding — graphite, a steam-pipe weld, a safety-case gap — the capacity factor for that year is whatever is left. No calculator can see that coming.

ONR does not set the megawatts. It sets whether the megawatts are allowed. Thermal-discharge consents sit with the Environment Agency in England (Natural Resources Wales, SEPA in Scotland); a hot river in July can cut output even when the nuclear safety case is clean.

A “rated 1,200 MW” plant is 1,200 MW only while both regulators are happy. Sizewell B’s nameplate is not a promise to National Grid. It is a ceiling.

Baseload on the GB grid

The electricity system operator has, for years, treated nuclear as baseload: run it, don’t fiddle with it. That is partly economics — high fixed cost, cheap fuel, every MWh wanted — and partly the AGR design, which is less keen on daily ramps than a French PWR. Sizewell B can load-follow. It is rarely asked to.

As the wind fleet grows, that bargain gets harder. A 3.2 GW Hinkley on a calm, sunny Sunday in 2032 may be asked to come off its plateau. Until then, modelling a UK station at a high capacity factor (80% plus, outages aside) is the honest default.

Load-following is a switch turned on when the contract and the safety case both say yes — not a number shaved off “because nuclear is inflexible”. It isn’t, always. Here, it has mostly chosen not to be. The 33% to 36% thermal efficiency does not care what the grid asked for: two-thirds of the heat goes to the sea or the towers.

Glossary

Capacity factor
Actual electrical energy sent out, divided by rated power × hours in the period. A nuclear fleet in good order sits around 75% over the long run; a quiet year can reach 85%. Outages, thermal consents and load-following all pull it down.
Rankine cycle
The steam cycle of a thermal power station: heat makes steam, steam drives a turbine, the condenser rejects the rest to a cold sink. A PWR’s hot source is about 300 °C, and Carnot already caps efficiency near 50%; real plant sits around 33% to 36%.
PWR
Pressurised-water reactor. Water in a primary loop, kept liquid under pressure, carries heat to a steam generator. Sizewell B is the UK’s operating PWR; Hinkley Point C and Sizewell C are EPRs, a PWR development.
AGR
Advanced gas-cooled reactor, the UK’s graphite-moderated, CO2-cooled design, typically around 600 MW electrical. Hotter steam than a PWR, but availability has been pulled down by graphite and boiler outages. The fleet is closing.
ONR
Office for Nuclear Regulation. It licenses the site and the operating rules. A unit held off after a finding produces nothing that month, whatever the nameplate says. Thermal-discharge consents sit with the environment regulators, not with ONR.

Related calculators