Power Density of Energy Sources
A nuclear power plant produces roughly 300 W/m² of power averaged over its site. A solar farm produces 5–20 W/m². Wind produces 1–3 W/m². A hydroelectric reservoir produces 0.5–2 W/m². These differences in power density translate directly into the land footprint of an energy system — and bound how much renewable build-out is feasible in any given country.
Key insights
Renewables are diffuse by physics
Solar irradiance averages ~170 W/m² at Earth's surface across all latitudes and weather. Wind energy averages ~1 W/m² of swept area integrated over time. PV panels and turbines extract a fraction of that flux. There is no way around the physical upper bound; renewables systems have to use more land to deliver the same total energy as fossil or nuclear systems.
Land use is the constraint where land is scarce
Replacing the US's 4,000 TWh/year of electricity with solar alone would require ~80,000 km² of panels — about the area of South Carolina. Land in the US, Australia, China is abundant. Land in the Netherlands, Singapore, Japan, Korea is not. This is one reason offshore wind, rooftop PV, and nuclear retain strategic value in some geographies.
System integration is a separate problem
Power density tells you how much land. It does not tell you how much grid, storage, or backup is needed to integrate intermittent sources. A 100 W/m² solar farm averaged over a year is delivering 0 W at night, requiring transmission and storage to translate raw output into deliverable power. These are separate constraints; both bite.
Power density by energy source
Watts per square metre of site area (illustrative middle of ranges)
Key Finding: Fossil and nuclear sources are roughly two orders of magnitude denser than wind and solar.
Land use per TWh — selected sources
Square kilometres needed to produce 1 TWh of electricity per year
Key Finding: Solar and wind need 30–100× more land per TWh than nuclear or gas — though much can be co-located with other uses (rooftops, farmland).
Methodology & caveats
Defining 'power density'
Power density can be measured at the panel/turbine level, the site level (including roads, spacing, substations), or the system level (including transmission and storage). Headline numbers vary by an order of magnitude depending on the boundary. Smil's site-level figures are the most commonly cited and what these charts use.
Capacity factor enters here
A 100 MW nameplate solar plant at 20% capacity factor produces 20 MW averaged over the year. Power density figures are usually annual averages — nameplate density × capacity factor. Nuclear capacity factors of 85–92% are why nuclear's effective power density is so much higher than its nameplate density.
Co-location and dual use
Rooftop PV uses no incremental land. Agrivoltaics combine solar with crops. Offshore wind uses sea area but not arable land. The 'how much land' question always depends on how strictly land is defined and which uses can be combined. Pure greenfield deployments at the lowest densities are the worst case, not the baseline.