Jonas Lauren Norr

Building Power Stations in Orbit

Jonas Lauren Norr

By Jonas Lauren Norr

A power station built in orbit could collect sunlight while the city beneath it sleeps, then send that energy to a receiving field connected to the grid. I’m interested in what that would change: electricity with less dependence on local weather, and a new way to connect the resources of space with life on Earth. The compelling possibility is a vast, lightweight machine assembled where it will operate, with robots adding and replacing its parts.

From Sunlight to Electricity

The proposed energy chain is straightforward. Photovoltaic cells turn sunlight into electricity. Electronics convert that electricity into microwaves, and a coordinated transmitting array directs the beam toward Earth. A receiving antenna field, often called a rectenna, converts the incoming energy back into electrical power. Every conversion loses energy, so the performance of the whole chain matters more than any impressive component.

Geostationary concepts seek long periods of sunlight and a consistent relationship with a ground receiver. They still need to account for eclipses and interruptions. The scale is substantial: ESA describes concepts with both satellites and receiving fields extending across kilometres. Beam control, land requirements and compatibility with aircraft, wildlife and other systems belong inside the engineering problem. [1]

Two Pieces Are Becoming Real

Caltech’s SSPD-1 mission tested several ingredients in orbit in 2023. Its MAPLE experiment transferred power wirelessly to onboard receivers and directed a beam toward Earth. Other experiments tested solar cells and an ultralight deployable structure. These were component demonstrations, not a power station supplying a grid. The mission also exposed practical weaknesses, including deployment jams and degradation in transmitting elements. Those results make the next design more informed. [2]

A separate advance makes the construction question more interesting. On November 5, 2026, Caltech reported that its ASTRA robotic demonstration had assembled a truss section in space. Two sections had launched already built; the robot constructed a third. The planned reflector remained incomplete, and the team described problems that required troubleshooting from Earth. This was a small assembly experiment for a reflector, not a solar power installation. [3]

I see a useful connection between these efforts. Large collectors and transmitting surfaces need structures that can grow beyond the package that fits inside a rocket. Robotic assembly offers a plausible route from launching a finished machine to launching parts and building the machine in orbit. Extending that capability to electrical connections, inspection and replacement would be a further breakthrough. It has not been established by assembling one structural section.

What Would Make the Power Valuable

The economic question is whether the entire system can deliver electricity worth more than its lifetime cost. NASA’s 2024 assessment examined two conceptual systems beginning operation in 2050 and found space solar power more expensive than terrestrial sustainable alternatives under its assumptions. It identified gaps in assembly, maintenance, autonomy and efficient power transmission, alongside launch and manufacturing costs. That is a demanding starting point for development. [4]

I would focus on useful electricity delivered over a system’s life per kilogram launched. A lighter collector helps, but only if it survives. Better conversion efficiency could reduce the collecting area and supporting mass needed for a given output. Replaceable modules could extend productive life, provided replacement flights and repair machinery do not consume the savings. These improvements interact; cheap launch alone does not establish a viable power station.

Consider an illustrative target, not a forecast: a system delivering one gigawatt at its ground connection for 90 percent of a year would supply about 7.9 terawatt-hours. At an assumed average realised price of $100 per megawatt-hour, gross annual electricity revenue would be roughly $790 million. That is before financing, operations, replacements, ground infrastructure and all other costs. The example shows why both utilisation and the cost of capital matter so much for a machine that must earn back a large initial investment.

Its strongest market might be electricity delivered during hours when other low-carbon supply is scarce. That is a hypothesis to test against alternatives offering comparable reliability, including terrestrial generation, storage and transmission. A receiver still needs land and a grid connection. The value of predictable delivery must be measured at that connection, rather than inferred from how much sunlight reaches the spacecraft.

The Next Useful Demonstration

I would like to see the construction and energy experiments meet: a modest orbital array assembled from modules, with measured electrical output at a ground receiver and a demonstrated ability to isolate a failed module. Publish the mass, conversion losses, assembly time and degradation over repeated operating cycles. Those measurements would let engineers and investors test the same proposition. The ambition is to make a power station that can grow in space and earn its place in the energy system on Earth.

Sources

[1] ESA, Wireless Power From Space. November 9, 2022. Accessed November 16, 2026.

[2] Caltech, Space Solar Power Project Ends First In-Space Mission With Successes and Lessons. January 16, 2024. Accessed November 16, 2026.

[3] Caltech, Caltech Robotic Space Demo Successfully Builds in Space. November 5, 2026. Accessed November 16, 2026.

[4] NASA, New Study Updates NASA on Space-Based Solar Power. January 11, 2024; page updated June 22, 2026. Accessed November 16, 2026.

Comments