Space-Based Solar Power: How It Works and What It Would Cost

A solar panel in orbit outperforms its twin on the ground four to five times over: full-spectrum sunlight 99% of the time in the right orbit, no clouds, no atmosphere, no night. Space-based solar power (SBSP) has been "twenty years away" since the 1970s — but reusable launch has finally moved the decimal point. Model it yourself with the free Space Solar Power & Data Station designer.

The three subsystems

An SBSP station is a collector (kilometres of thin-film array — at 0.5 kg/m², a gigawatt of collection masses a few thousand tonnes), a converter, and a link to the customer. The link is the hard part. Microwave beaming at 2.45 or 5.8 GHz penetrates weather and reaches ~50–60% end-to-end efficiency, but physics dictates kilometre-scale transmitters and rectennas. Lasers shrink the aperture but drop efficiency and stop at clouds.

The launch-cost pivot

At Shuttle-era prices ($20,000+/kg), a gigawatt station's launch bill alone exceeded $60B — dead on arrival. At $100–200/kg, plausible with mature reusable heavy lift, the same launch bill is $0.5–1B, and suddenly the array hardware, not the ride, dominates. This is the single variable that revived the field; the UK, ESA, Japan, China and the US military have all restarted programs since 2020.

Beam it or use it up there?

Here's the twist the calculators make vivid: beaming loses ~40–50% of the energy, but computing doesn't need to come home — only photons do. Allocate the station's power to on-board GPUs and downlink answers instead of watts, and the beaming penalty vanishes. The designer has a compute-share slider for exactly this trade, the Orbital Data Center Planner prices the datacenter half, and the Constellation Energy-Grid model scales it to fleets. Ground comparison: the solar + battery sizing math.

FAQ

Is the microwave beam dangerous?

Design intensities at the rectenna edge are comparable to sunlight; birds flying through the beam center experience gentle warming. It cannot be "focused into a weapon" — the aperture physics that makes it safe is fixed by the hardware.

Why not just build more panels on Earth?

For most demand, you should — ground solar is cheaper today. SBSP competes for baseload-critical, land-poor, or high-latitude demand, and for powering space industry itself.

When is it realistic?

Demonstrator beaming has already flown (Caltech, 2023). Gigawatt commercial stations need $100/kg launch at scale — credibly 2030s.