- Capacity factor measures how hard a power plant actually works. It’s the electricity a generator produces over a year, divided by the maximum it could have produced running at full power nonstop.
- The gaps are enormous. U.S. nuclear plants run at about 91% capacity factor; utility-scale solar manages about 24% — so a 100-MW nuclear plant delivers roughly 3.7× the annual energy of a 100-MW solar farm.
- Capacity factor is essential context for every gigawatt announcement. “We’re building 1 GW” means very different amounts of electricity depending on the technology.
The simplest definition in energy
Every power plant has a nameplate capacity — the maximum power it can produce, measured in megawatts (MW). But no plant runs at full power all the time. Turbines need maintenance, nuclear reactors refuel, the wind doesn’t always blow, and the sun sets every evening.
Capacity factor captures this in one number: the actual electricity a generator produced over a period, divided by the electricity it would have produced running at full nameplate power for every hour of that period. There are 8,760 hours in a year, so the formula is:
Capacity factor = actual annual generation ÷ (nameplate capacity × 8,760 hours)
A plant with a 100 percent capacity factor would run flat-out, 24/7, all year — which no real plant achieves. The U.S. Energy Information Administration (EIA) publishes capacity factors for the entire U.S. fleet in its Electric Power Monthly, and the numbers tell a story that nameplate capacity alone never could.
The numbers: solar, wind, coal, gas, and nuclear
Here are the annual capacity factors for U.S. utility-scale generators, from EIA’s Electric Power Monthly data (2025 figures are preliminary):
| Technology | 2024 (final) | 2025 (preliminary) |
|---|---|---|
| Nuclear | 90.8% | 91.0% |
| Natural gas | 60.5% (combined cycle) | 58.4% |
| Coal | 42.6% | 48.7% |
| Hydropower | 34.6% | 35.3% |
| Wind | 34.3% | 34.2% |
| Solar PV (utility-scale) | 23.2% | 24.4% |
Sources: EIA Electric Power Monthly, Table 6.07.A and Table 6.07.B. Small-scale (rooftop) solar systems run lower still, typically 10–25 percent.
Read the table as a measure of utilization, not quality. A nuclear plant near 91 percent isn’t “better” than a solar farm at 24 percent — it’s simply doing a different job. Nuclear reactors are designed to run continuously for 18 to 24 months between refueling outages. A solar panel, by contrast, produces nothing at night through no fault of its own.
What it means in practice: the 100-MW thought experiment
Take two hypothetical 100-MW plants using the 2025 U.S. averages:
- 100 MW of nuclear at 91%: 100 × 8,760 × 0.91 ≈ 797,000 MWh per year
- 100 MW of solar PV at 24.4%: 100 × 8,760 × 0.244 ≈ 214,000 MWh per year
The nuclear plant delivers about 3.7 times as much electricity from the same nameplate capacity. To match one 100-MW nuclear plant’s annual output with solar, you’d need roughly 370 MW of solar panels — before accounting for the fact that the solar output arrives only during daylight hours.
This is why capacity announcements need a capacity-factor translation. When a developer announces a 1-GW solar farm and a utility announces a 1-GW gas plant, they are announcing very different amounts of future electricity. Journalists, investors, and policymakers who compare nameplate gigawatts across technologies without adjusting for capacity factor are comparing incomparable things.
Why the numbers differ — and why they move with the seasons
Capacity factors differ for two broad reasons: physics and economics.
Physics governs the weather-dependent sources. Solar’s capacity factor is capped by the simple fact that panels generate only in daylight, and less in winter. EIA’s monthly data makes this vivid: U.S. utility-scale solar ran at just 16.4 percent in January 2025 and 32.4 percent in July 2025 — a single fleet, performing twice as well in summer as in winter. Wind is seasonal too: the U.S. wind fleet hit 44.2 percent in March 2025 but only 22.9 percent in September.
Economics governs the dispatchable sources. Coal and gas plants can run at high capacity factors, but they often don’t because it isn’t profitable — cheap gas, renewables, or weak demand can idle them. That’s why the coal fleet’s factor drifts in the 40s rather than the 90s. Nuclear sits near 91 percent because reactors are expensive to build but cheap to run, so operators keep them online as much as physically possible; the dips you see — like the fleet’s drop to about 80 percent in October 2024 — are refueling and maintenance outages.
What capacity factor doesn’t tell you
Capacity factor is about how much electricity a plant makes, not when it makes it — and in electricity markets, timing is money. A megawatt-hour delivered at 6 p.m. on a hot August evening, when air conditioners are straining the grid, is worth far more than the same megawatt-hour at 2 p.m. on a mild spring Sunday when solar is flooding the grid.
This is the metric’s main blind spot. Solar’s midday output can be so abundant that its market value collapses at the very hours its capacity factor is highest — the dynamic behind California’s “duck curve” and its overhaul of rooftop solar credits. Conversely, a gas peaker plant with a capacity factor under 10 percent can be extremely valuable precisely because it runs only during the scarcest, most expensive hours.
Energy planners have a separate concept, capacity credit, for how much a generator can be counted on during peak demand. But for a first-pass reality check on any energy claim, capacity factor remains the indispensable tool: it turns the abstract gigawatts of a press release into the concrete terawatt-hours that actually power homes.
