Giz Explains: What It Will Take To Farm Sunlight From Space

What It Will Take to Farm Sunlight from Space

Whether they're producing voltage directly from solar rays or focusing them to melt salt like Ivanpah, even Earth's biggest and baddest solar power plants are hamstrung by all this damnable atmosphere getting in the way. But a new kind of off-world solar energy plant could soon provide the whole planet with plenty of power -- we just have to finish figuring out how to build and operate it.

Energy Production in Spaaaaaaaace!

With the advent of silicon-based photovoltaic solar panels -- the kind that directly convert solar energy to electrical current -- some 60 years ago, researchers immediately looked to the skies as the ideal place to collect solar energy. Up there, you don't have miles and miles of atmosphere and clouds absorbing, scattering or blocking out the sun's incoming rays. That means photovoltaic panels should, conceivably, be able to operate at (or very near) their theoretical efficiency limits. Plus, if you position a solar power satellite (SPS) properly over the equator, it will only reside in the Earth's shadow for a few hours every year and thereby provide nearly non-stop energy.

The idea of space-based solar power (SBSP) was formalised in the seminal 1968 report, Power from the Sun: Its Future, by American aerospace engineer Peter Glaser. The paper set forth a conceptual system for collecting unhampered solar energy from massive extra-atmospheric arrays of photovoltaic cells set in geosynchronous orbit above the equator, and transmitting it wirelessly back to Earth where it would be used by terrestrial power grids. In theory, with enough orbiting "solar farms," the energy needs of not just the U.S. but the entire world could be met.

In his paper, Glaser argued that while building, launching and operating such a power plant was currently beyond the reach of scientific knowledge at the time, those technological advances would be within our grasp in the coming years and decades. So, are we any closer to freeing the entire world from its energy woes with orbiting solar farms than we were at the start of the Space Age? Sure, but we still have some work to do before that actually happens. Specifically, there are a number aspects that we need to iron out before something like this actually comes to fruition.

Launch It

What It Will Take to Farm Sunlight from Space

The first issue is the fact that a commercial-grade SPS would be simply gargantuan. In order to produce a GW of power, you'd need a massive collection area 0.5km long by 5.2km wide and weighing tens of thousands of tons. No matter how tightly you fold it up, there's simply no way to get a fully formed SPS from the surface of the Earth into orbit given our current launch capabilities that wouldn't be cost-prohibitive.

So, for example, let's assume that a standard solar panel weighs about 20 kg per kw. Not including the necessary support and transmission components, a 4GW capacity would weigh a whopping 80,000 tons. It would require nearly 9000 Atlas V rockets (each with a max lifting ability of 8900kg to GEO) to free that structure from Earth's gravitational grip, or at least 9000 trips to geosynchronous orbit and back, and cost somewhere in the neighbourhood of $US320 billion. That's just to get the solar panels into position, not to assemble them or operate them -- just to get them up there. Nor is that accounting for the environmental impact of all those rocket launches.

However, while reusable space launch systems like Space X's Dragon Capsule can only lift a fraction of what an Atlas V can, their low-cost nature could provide significant cost savings and drastically shortened turnaround times should the project be attempted today. Similarly, since we're not rushing to beat another nation to the punch (something of this scale would demand the financial and technological assistance of every nation on Earth), slower but more cost efficient delivery methods like ion propulsion could also be deployed to shuttle materials from Low Earth Orbit up to Geosynchronous Earth Orbit. Essentially, LEO would become a staging area where materials would be tugged up to GEO by a fleet of as-of-yet-uninvented space transport vehicles.

A secondary option proposed by American physicist Gerard O'Neill in the 1970s would have avoided the high cost of launching materials from Earth by instead constructing the SPS from materials mined on the Moon. This would have offered significant launch cost savings given the Moon's far lower gravity, but would have required NASA to invent and deploy mass drivers (electromagnetic rail guns designed to throw packages into space) on the Moon's surface. Though this seems like it would cost a hell of a lot more than just using rockets, a 1979 report by General Dynamics' Convair Division estimated that using lunar resources would be cost effective should we build out 30 or so 10GW SPS's -- for a total capacity of 300GW, or enough to satisfy projected US electricity demand in the 2000-2030 period.

Build It

What It Will Take to Farm Sunlight from Space

Concept image of a solar collector via

So even if we manage to get these tens of thousands of tons of stuff into orbit, the next issue would be putting it all together. This of course comes with its own set of challenges. The structure, for example, wouldn't need to support itself against gravity or the elements as terrestrial-based power plants do, but would have to defend against micrometeors and solar flares.

There's also the matter of who would build it. When NASA took a look at the issue in the late 1970s, it estimated necessary construction time at around 30 years. Three decades of build time. We can barely keep highly-trained astronauts out there on the ISS for more than a year, and a project like this would require either a veritable army of orbital workers (we're talking a New Deal-scale workforce) continually shuttled back and forth to the surface, or we'd need an army of robots to do the same.

NASA's 1970s solution was to use a fleet of "beam builder" robots to roll and assemble sheets of aluminium into trusses tens of kilometres long. This method would reduce the necessary workforce of humans to a supervisory skeleton crew, which in turn would minimise training, operating, and liability costs. However, even with generous estimates of mechanisation capabilities at the time, NASA estimated it would need at least 1000 full-time astronauts on hand at any given moment -- again, that's just counting astronauts, not the additional doctors, cooks, cleaners and other service workers they'd require to live in orbit, or the massive amount of resources (air, water, and food) that they'd consume.

NASA estimated that the number of support workers would outnumber the builders by a factor of 10 to 1. And though this would be a massively expensive undertaking, it would also open up a huge new industry for anyone brave enough to work and live 35,000km up.

Maintain It

Not to put the cart before the horse, but assuming we do somehow manage to construct an SPS, keeping it from falling out of the sky could be tricky. The ISS for example, the largest orbiting man-made satellite in existence, uses regularly refilled gas propulsion to keep its orbit from fatally degrading. But given the monstrous size of these power plants, we'd have to devise a new, more efficient means of keeping them aloft.

Solar light sails have been suggested as one solution, propped up either by the suns rays or by ground-based laser and radio energy. This energy would essentially counteract the planet's gravitational pull and push the SPS just hard enough to keep it from falling back to Earth. But we're still years away from such technology being readily available.

Another solution, which is a bit closer to reality, is to convert solar-generated, DC power into microwaves and beam that energy up to the satellite to provide operational power. Researchers have been playing with this technology since the 1980s, and JAXA (Japan's space agency) recently announced that a proposed small-scale SPS might use this method when it comes online in 2040.

Get It Back

By far, the biggest stumbling block for SPS technology involves getting it from space to your wall socket -- it's not like we can just run a huge extension cord up there. Instead, we'll have to rely on a neophyte power transmission technology known as "wireless power transmission" (WPT). WPT converts DC current to microwave frequency and shoots it to a distant receiver where it is converted back to electricity and added to the power grid -- essentially the reverse of what we'd use to keep the SPS aloft, as described above.

This technology is far closer to science than fiction. It was first demonstrated in 1964 when American electrical engineer William C Brown demonstrated a microwave beam-powered helicopter for Walter Cronkite on the CBS Evening News. Subsequent developments by Raytheon in the 1970s saw microwaves transmit 30kW of energy over the course of 1.6km with 84 per cent efficiency. And while a 5GW beam would require massive arrays of receiver dishes spread over large uninhabited areas of the planet, the UN's non-profit SunSat Energy Council has stated that this type of beam would be of such low density that it wouldn't be capable of harming plant and animal life. You wouldn't get a kitten in a microwave effect if you walked through this beam -- in fact, it would reportedly warm your skin less than the Sun's natural rays would.

While WPT technology is certainly possible, there are a number of necessary factors to make it plausible. Factors like how you would generate the microwave signal in the first place. In the 1970s, when NASA first looked at the issue, the state of the art still used vacuum tubes. Today, semiconductor amplifiers offer superior efficiencies at a lower price point, but at the 1GW scale, an SPS would need somewhere around the order of 100 million such devices to create a powerful-enough signal.

There's also the matter of the specific frequency the beam will take, lest it interfere with existing technologies. Somewhere in the 1-10GHz range (around 5.8GHz) is most likely, given the need to balance between antenna size and atmospheric penetration capabilities as well as accounting for existing band usage.

Then there's the issue of aiming the damn thing to hit a receiver dish 36,000 km away. You wouldn't be able to do it with a single antenna. An SPS would require a massive number of smaller coordinated and synchronised antennae (up to a billion per satellite by some estimates) each precisely aimed at a 3km wide rectifying antenna on the ground, and aimed with an accuracy of just 10 µrad (and an efficiency of about 85 per cent). That's an unprecedented level of accuracy -- not even the beam line tolerances at CERN are that tight. For all intents and purposes, it's beyond our capabilities at this time.

Where to Go from Here

While this may seem like just as much of a Herculean task as it was in the 1970s, SBSP could well become a viable energy source within our lifetimes. Japan has already announced plans to build its own SPS within the next 25 years. Given both the rapid development of renewable energy over the past decade and the shift from public to private spaceflight -- not to mention the growing need for more and cleaner power -- the stars could soon align in favour of this ambitious project. [IEEE - Wiki 1, 2 - Navy Research Lab - NSS - NRL - Wired 1, 2 - NOVA]