Originally published October 15, 2013. Updated September 28, 2026.
Capacity factor is how much of a power plant’s maximum possible output it delivers over a year, and in 2024 that ranged in the US from 90.8% for nuclear plants to 23.2% for utility-scale solar. Wind fell in between at 34.3%. The gap decides how many megawatts your utility has to build for the same electricity, since a 100-megawatt solar farm and a 100-megawatt reactor carry the same nameplate rating but don’t come close to producing the same amount of power over a year.
What capacity factor means
The Energy Information Administration defines capacity factor as the ratio of a plant’s actual output to its maximum possible output over the same period, expressed as a percentage.
A 100-megawatt plant running flat out for all 8,760 hours of a year would produce 876,000 megawatt-hours. If it produces 300,000 megawatt-hours instead, its capacity factor is about 34%. A 100-megawatt solar farm at the 2024 US average of 23.2% produces about 203,000 megawatt-hours, roughly two-thirds of that.
The same math applies to a home system. A 6-kilowatt rooftop array running at that 23.2% average produces around 12,200 kilowatt-hours a year. That’s more than the average US home’s 10,791 kilowatt-hours in 2022. That’s why installers size a system to your roof and your annual use instead of your single hottest afternoon in August.
Capacity factor is a different thing from efficiency, which describes how much of a fuel’s energy a plant turns into electricity. A wind turbine can be in perfect working order and still post a low capacity factor, because it can produce only about as much power as the wind happens to give it. A plant’s nameplate capacity is the ceiling on what it can produce at any moment, and capacity factor measures how much of that ceiling it fills across a whole year.
Capacity factor is also not the same as availability factor, which measures how much of the year a plant is physically capable of running, whether or not it does. A plant sitting offline for maintenance or repairs produces nothing during those hours, no matter how strong its output is the rest of the year.
US capacity factors by source
The figures below are the latest full-year US averages from EIA data. Most of them are for 2024, and the solar and hydro averages are for 2023.
| Source | Capacity factor |
|---|---|
| Nuclear | 90.8% |
| Geothermal | 64.6% |
| Natural gas (combined cycle) | 60.5% |
| Coal | 42.6% |
| Wind | 34.3% |
| Hydroelectric | 34.6% |
| Solar (utility-scale) | 23.2% |
Source: US Energy Information Administration, Electric Power Annual, Table 4.8.A and Table 4.8.B, 2024 data.
Nuclear plants lead this table by a wide margin, which is part of why they keep coming up in arguments about a low-carbon grid, alongside the questions people still ask about nuclear safety. When your utility retires a coal or gas plant and replaces it with wind or solar, this table is why the replacement often lists more megawatts of nameplate capacity than the plant it’s replacing, even though it burns no fuel.
Hydroelectric numbers move with the weather. EIA’s Electric Power Annual shows the US fleet moving between 34.6% and 36.3% from 2022 to 2024 depending on how wet each year was.
How wind got more productive
US wind turbines produce a lot more of their nameplate rating than they used to, mainly because the machines themselves got bigger. The average new US turbine’s tower rose 83% between 1998 and 2023, reaching about 103 meters tall, high enough to put its blades well above the choppy air near the ground, with blades sweeping a circle more than 133 meters across. That height is part of why the US wind fleet’s average capacity factor stood at 34.3% in 2024.
A wind-heavy grid feels that 34.3% figure directly. It takes about three watts of installed wind capacity, on average across the year, to reliably deliver one watt to your outlet, which is why the grid still needs gas plants or batteries standing by for calmer hours.
Location matters as much as equipment. British government statistics show UK onshore wind ran at a 25.7% load factor in 2024, while offshore ran at 38.7%. Per megawatt of rating, Britain’s offshore wind turbines produced about half again as much electricity as its onshore ones.
Country averages differ in the same way. Wilson’s analysis put German wind at about 18% and Danish wind at about 30%. The World Economic Forum drew on similar figures to show that 10 gigawatts of wind in Britain delivers about 2.6 gigawatts on average, against 1.8 gigawatts in Germany.
Solar in the US, the UK and Germany
Solar’s capacity factor depends mostly on how much sun a site gets. US utility-scale solar averaged 23.2% in 2024. The UK, much farther north, posted a solar load factor of 9.8% in 2024, well under half the US figure. If you live somewhere with that kind of weather, a home solar system needs more panels than one in Arizona to produce the same yearly total, which is part of why installers size Northern European systems larger relative to the roof.
Germany sits close to the UK. In an analysis first published on this site in October 2013, Robert Wilson put German solar below 10%, against about 20% in Arizona. Fraunhofer ISE reports German solar produced 87 terawatt-hours in 2025 from 116.8 gigawatts of installed panels, which computes to about 8.5% of full output across the year. That’s likely a slight undercount, since some of that capacity arrived partway through 2025, but German solar still runs at around a tenth of its rating.
An average like that hides big swings. Wilson pointed out that German solar output peaked at 22 gigawatts in May 2012, about seven times its average output of roughly 3.2 gigawatts. A grid has to absorb those peaks as well as cover the gaps.
Germany has built a large solar fleet anyway. It’s a good example of how a low capacity factor raises the amount you have to build without stopping a country from building it.
How much capacity it takes to replace a reactor
The same formula shows why capacity factor decides how much a grid has to build. A 1-gigawatt nuclear plant at 90.8% in 2024 produces about 7,954 gigawatt-hours a year. Getting that much energy from solar at 23.2% takes about 3.9 gigawatts of panels. Getting it from wind at 34.3% takes about 2.6 gigawatts of turbines.
So replacing a retiring reactor with solar means building roughly four times its rating in panels, and replacing it with wind means about two and a half times. If a reactor near your town retires, that’s the arithmetic your utility runs, and neither solar nor wind shows up exactly when demand does. Even then, matching the annual total still leaves the question of what supplies power at night and on calm days.
Storage is starting to close the gap wind and solar leave open
Wilson made this argument using 2009 US figures and the grids of Denmark and Germany, and the same tradeoff shapes your grid today if it leans on wind or solar. His point was that a grid built on low-capacity-factor sources needs far more installed capacity, plus somewhere to send the surplus when the wind and sun peak together. He estimated that past 70% wind, Denmark would export or store about 30% of its output, and that once solar supplies about 40% of Germany’s power, it would need to curtail or store roughly half that output.
Coal has already lived through a smaller version of that problem, for a different reason. US coal plants ran at 63.8% in 2009, versus 90.3% for nuclear that same year. Coal’s capacity factor has since fallen to 42.6%.
Wind and solar’s low capacity factor is the same kind of problem, and storage is the tool now closing it. US utility-scale battery capacity reached 43.6 gigawatts by the end of 2025, up from about 26 gigawatts a year earlier. Batteries give a solar farm somewhere to send its midday surplus instead of curtailing it, the same surplus problem Wilson described in Germany, and EIA expects developers to add about 26 more gigawatts in 2027 alone.
