Beaming 24/7 solar from space

Start with what the thing is for. A solar array in a high enough orbit sees the sun continuously and at full intensity, unfiltered by atmosphere, so it produces around the clock without storage and without weather dependence. That puts it in the small category of clean firm resources alongside nuclear, deep geothermal and hydro. It also has a property none of those have: because the power arrives as a steerable beam, it is not tied to geography. Any country that can put collectors up and a receiver down has the same access to the resource, and power can be moved from one point on earth to another with no losses in between. Vijendran makes that the explicit comparison to fusion, which he says could never do the geographic part.

The intuitive objection is that something in that chain must be unproven physics. Vijendran’s answer is that none of it is. A space solar station is essentially a very large telecommunications satellite, and telecom satellites already convert sunlight to electricity, convert that to radio frequency, push it through an antenna and beam it to the ground. What is new is only scale, and the scale is the problem. Sunlight is capped at about 1.4 kilowatts per square meter even in space, so a plant large enough to replace a coal or nuclear station needs on the order of 10 square kilometers of collection area or more. Keeping the ground receiver down to a few kilometers across requires a transmitting antenna one to two kilometers wide, and at best hundreds of meters. That is thousands of tons of hardware, assembled robotically, held and pointed as a single structure. The largest object humanity has put in space, the International Space Station, is 100 meters across.

Which is why the question was always cost, and why the answer kept coming back the same. The 1970s joint studies by the US Department of Energy and NASA, and roughly every decade of reviews since, concluded that it was technically doable, useful if done, and too expensive. Two line items drove that: expendable launch and hand-built, low-production-rate space hardware. Both have changed. Reusable launch has already cut costs by almost two orders of magnitude from the shuttle era, with another order of magnitude plausible from heavy-lift vehicles in development, and mass production for low-earth-orbit constellations has brought hardware from hundreds of thousands of dollars per kilogram to roughly a thousand. Vijendran’s threshold is below $1,000 per kilogram to low earth orbit for the economics to be in the ballpark, and he says we are not far off. The same system design that would have cost around $100 billion on 30-year-old assumptions now prices at $1 billion to $10 billion depending on scale. He attaches his own caveat: robotic assembly, operations and maintenance are the pieces nobody has demonstrated, so those costs are the least well understood.

That sets up the actual turn in his argument, which is about attention rather than technology. Power beaming has been demonstrated only in the kilowatt range over kilometer distances, and he insists this is not because anyone tried the larger version and failed. Nobody has had a reason to try, because the full-scale application never justified the R&D and the terrestrial market for beaming power has only just appeared. So the pacing item is money, and his comparison is pointed: fusion, which government roadmaps do not expect before the second half of the century, receives billions per year, while space-based solar receives almost no appreciable investment. With a moonshot-scale push he puts first systems in the mid-to-late 2030s and meaningful scale in the late 2030s into the 2040s. Without it, he says, it could be a very long time before anything happens at all.