IEET > GlobalDemocracySecurity > Vision > Directors > George Dvorsky > HealthLongevity > Futurism > Innovation > Biosecurity
How to build a Dyson sphere in five (relatively) easy steps
George Dvorsky   Mar 29, 2012   Sentient Developments  

Let’s build a Dyson sphere! By enveloping the sun with a massive array of solar panels, humanity would graduate to a Type 2 Kardashev civilization capable of utilizing nearly 100% of the sun’s energy output.

A Dyson sphere would provide us with more energy than we would ever know what to do with while dramatically increasing our living space. Given that our resources here on Earth are starting to dwindle, and combined with the problem of increasing demand for more energy and living space, this would seem to a good long-term plan for our species.

Implausible you say? Something for our distant descendants to consider?

Think again: We are closer to being able to build a Dyson Sphere than we think. In fact, we could conceivably get going on the project in about 25 to 50 years, with completion of the first phase requiring only a few decades. Yes, really.

Now, before I tell you how we could do such a thing, it’s worth doing a quick review of what is meant by a “Dyson sphere”.

Dyson Spheres, Swarms, and Bubbles

The Dyson sphere, also referred to as a Dyson shell, is the brainchild of the physicist and astronomer Freeman Dyson. In 1959 he put out a two page paper titled, “Search for Artificial Stellar Sources of Infrared Radiation” in which he described a way for an advanced civilization to utilize all of the energy radiated by their sun. This hypothetical megastructure, as envisaged by Dyson, would be the size of a planetary orbit and consist of a shell of solar collectors (or habitats) around the star. With this model, all (or at least a significant amount) of the energy would hit a receiving surface where it can be used. He speculated that such structures would be the logical consequence of the long-term survival and escalating energy needs of a technological civilization.

Needless to say, the amount of energy that could be extracted in this way is mind-boggling. According to Anders Sandberg, an expert on exploratory engineering, a Dyson sphere in our solar system with a radius of one AU would have a surface area of at least 2.72x1017 km2, which is around 600 million times the surface area of the Earth. The sun has an energy output of around 4x1026 W, of which most would be available to do useful work.

I should note at this point that a Dyson sphere may not be what you think it is. Science fiction often portrays it as a solid shell enclosing the sun, usually with an inhabitable surface on the inside. Such a structure would be a physical impossibility as the tensile strength would be far too immense and it would be susceptible to severe drift.

Dyson’s original proposal simply assumed there would be enough solar collectors around the sun to absorb the starlight, not that they would form a continuous shell. Rather, the shell would consist of independently orbiting structures, around a million kilometres thick and containing more than 1x105 objects. Consequently, a “Dyson sphere” could consist of solar captors in any number of possible configurations. In a Dyson swarm model, there would be a myriad of solar panels situated in various orbits. It’s generally agreed that this would be the best approach. Another plausible idea is that of the Dyson bubble in which solar sails, as well as solar panels, would be put into place and balanced by gravity and the solar wind pushing against it.

For the purposes of this discussion, I’m going to propose that we build a Dyson swarm (sometimes referred to as a type I Dyson sphere), which will consist of a large number of independent constructs orbiting in a dense formation around the sun. The advantage of this approach is that such a structure could be built incrementally. Moreover, various forms of wireless energy transfer could be used to transmit energy between its components and the Earth.

Megascale construction

So, how would we go about the largest construction project ever undertaken by humanity?

As noted, a Dyson swarm can be built gradually. And in fact, this is the approach we should take. The primary challenges of this approach, however, is that we will need advanced materials (which we still do not possess, but will likely develop in the coming decades thanks to nanotechnology), and autonomous robots to mine for materials and build the panels in space.

Now, assuming that we will be able to overcome these challenges in the next half-decade or so—which is not too implausible— how could we start the construction of a Dyson sphere?

Oxford University physicist Stuart Armstrong has devised a rather ingenious and startling simple plan for doing so—one which he claims is almost within humanity’s collective skill-set. Armstrong’s plan sees five primary stages of construction, which when used in a cyclical manner, would result in increasingly efficient, and even exponentially growing, construction rates such that the entire project could be completed within a few decades.

Broken down into five basic steps, the construction cycle looks like this:

1. Get energy
2. Mine Mercury
3. Get materials into orbit
4. Make solar collectors
5. Extract energy

The idea is to build the entire swarm in iterative steps and not all at once. We would only need to build a small section of the Dyson sphere to provide the energy requirements for the rest of the project. Thus, construction efficiency will increase over time as the project progresses. “We could do it now,” says Armstrong. It’s just a question of materials and automation.

And yes, you read that right: we’re going to have to mine materials from Mercury. Actually, we’ll likely have to take the whole planet apart. The Dyson sphere will require a horrendous amount of material—so much so, in fact, that, should we want to completely envelope the sun, we are going to have to disassemble not just Mercury, but Venus, some of the outer planets, and any nearby asteroids as well.

Why Mercury first? According to Armstrong, we need a convenient source of material close to the sun. Moreover, it has a good base of elements for our needs. Mercury has a mass of 3.3x1023 kg. Slightly more than half of its mass is usable, namely iron and oxygen, which can be used as a reasonable construction material (i.e. hematite). So, the useful mass of Mercury is 1.7x1023 kg, which, once mined, transported into space, and converted into solar captors, would create a total surface area of 245g/m2. This Phase 1 swarm would be placed in orbit around Mercury and would provide a reasonable amount of reflective surface area for energy extraction.

There are five fundamental, but fairly conservative, assumptions that Armstrong relies upon for this plan. First, he assumes it will take ten years to process and position the extracted material. Second, that 51.9% of Mercury’s mass is in fact usable. Third, that there will be 1/10 efficiency for moving material off planet (with the remainder going into breaking chemical bonds and mining). Fourth, that we’ll get about 1/3 efficiency out of the solar panels. And lastly, that the first section of the Dyson sphere will consist of a modest 1 km2 surface area.

And here’s where it gets interesting: Construction efficiency will at this point start to improve at an exponential rate.

Consequently, Armstrong suggests that we break the project down into what he calls “ten year surges.” Basically, we should take the first ten years to build the first array, and then, using the energy from that initial swarm, fuel the rest of the project. Using such a schema, Mercury could be completely dismantled in about four ten-year cycles. In other words, we could create a Dyson swarm that consists of more than half of the mass of Mercury in forty years! And should we wish to continue, if would only take about a year to disassemble Venus.

And assuming we go all the way and envelope the entire sun, we would eventually have access to 3.8x1026 Watts of energy.

Dysonian existence

Once Phase 1 construction is complete (i.e. the Mercury phase), we could use this energy for other purposes, like megascale supercomputing, building mass drivers for interstellar exploration, or for continuing to build and maintain the Dyson sphere.

Interestingly, Armstrong would seem to suggest that this might be enough energy to serve us. But other thinkers, like Sandberg, suggest that we should keep going. But in order for us to do so we would have to deconstruct more planets. Sandberg contends that both the inner and outer solar system contains enough usable material for various forms of Dyson spheres with a complete 1 AU radius (which would be around 42 kg/m2 of the sphere). Clearly, should we wish to truly attain Kardashev II status, this would be the way to go.

And why go all the way? Well, it’s very possible that our appetite for computational power will become quite insatiable. It’s hard to predict what a post-Singularity or post-biological civilization would do with so much computation power. Some ideas include ancestor simulations, or even creating virtual universes within universes. In addition, an advanced civilization may simply want to create as many positive individual experiences as possible (a kind of utilitarian imperative). Regardless, digital existence appears to be in our future, so computation will eventually become our most valuable and sought after resource.

That said, whether we build a small array or one that envelopes the entire sun, it seems clear that the idea of constructing a Dyson sphere should no longer be relegated to science fiction or our dreams of the deep future. Like other speculative projects, like the space elevator or terraforming Mars, we should seriously consider putting this alongside our other near-term plans for space exploration and work.

And given the progressively worsening condition of Earth and our ever-growing demand for living space and resources, we may have no other choice.

George P. Dvorsky serves as Chair of the IEET Board of Directors and also heads our Rights of Non-Human Persons program. He is a Canadian futurist, science writer, and bioethicist. He is a contributing editor at io9 — where he writes about science, culture, and futurism — and producer of the Sentient Developments blog and podcast. He served for two terms at Humanity+ (formerly the World Transhumanist Association). George produces Sentient Developments blog and podcast.


Humanity may someday need that kind of energy, however it first must overcome some of the serious engeneering problems. The biggest problem is energy distribution; how do you get that much energy to Earth. Microwave transmission is possible, but from such great distances the receiving rectena would have to be the size of a contenent. Then you have the problem the Earth rotates, so you will need a rectena that circles the whole globe. Another problem is that consuming energy creates heat. How would we diapate that much heat. I think a Solar Power Satalite in GEO orbit is much more doable and just a few of them at 100 square kilometers each could supply most of the Earths power for many, many years giving us time to solve the bigger problems.

At GEO orbital distance the rectena only needs to be about 5 to 10 kilometers wide. The theoretical transmission efficiecy would be about 90%, and operate 24 hours a day.

A more likely sceario for a dyson sphere will be when we reach a point that we are building habitats in space, they may be placed in a dyson sphere orbit with their solal panels arranged as you suggest. The power would then be consumed locally and radiated back into space.

Nice idea, except for the “turn the solar system into computronium” mantra. Really? All the heavy elements we could ever need could be scooped out of the Sun. If we’re building a Star-Dam, then we could industrialise the Sun too. The ~2% metals in the Sun is 20 Jupiter masses of the heavy elements sorely lacking in the gas giants.

I think when they speak of turning all the matter in the solar system into computronium, to include our suns mass does not make sense.

The energy produced by the sun is very efficient; it would not be economical to harvest a burning sun for elements, when there are so many dead sun bits floating around. Every asteroid and planet was once part of a stars core that exploded as a supernova. There is an immense supply of asteroid and planet mass right here in the solar system, and then hundreds of trillions more in the Kuiper Belt and Oort cloud, and likely throughout interstellar space. If we started populating asteroids with bases or just expanding as computronium, hopping from one to another, we would eventually reach the next star, and then someday fill the galaxy.

I think a better question to ask is, why would such a super intelligence desire to saturate the entire cosmos? Eternal expansion is a biological desire; who can say our computronium cousins will share that desire. We will have an even weaker understanding of what drives their desire than an amoeba can fathom what drives human desire.

Hi Kelly
The concept of star-mining has been around for almost three decades. My point was that there were other options than chewing up the natural planets.

I agree with you about super-intelligence and the presumed expansion/super-sizing imperative. Perhaps such an entity/entities would rather span the Galaxy lightly via a vast network of observation drones. Running one’s clock speed ever faster seems so biological…

Here’s a comment that I posted in a later thread about NASA approaching George about his Dyson sphere idea.  Some stuff you guys may have already mentioned.

Just reading some of the comments from the links posted, some of the criticism and suggestions were that it would probably better to utilize materials from the asteroid belt which doesn’t have too much gravity or exposure to solar radiation (as well as more usable materials, may even lower the risk of asteroid impacts on earth) and that strip mining mercury and the sphere itself could disrupt the gravitational influences on the sun and the earth (or something like that).  It seems you can get a lot of good info and input from regular commenters.

Hi Christian
A smaller lighter segment of a Dyson sphere can propel thousands of starships per year, so why disassemble planets? That’s my main issue with George’s plan. Otherwise I think it’s a perfectly fine idea. Long before we break up Mercury, the whole Galaxy is within reach. Better to touch a trillion planets lightly than smash up the ones nearby.

“strip mining mercury and the sphere itself could disrupt the gravitational influences on the sun and the earth”

Aren’t the masses involved too large to be concerned about this?

@ Intomorrow

Just repeating what other commenters said.

It certainly sounds like a good idea. However, I feel that there are some remaining things we would need to clarify:

1. Let’s assume that conventional oil (not fracking or tarsands) is going to become uneconomic to extract i.e. it takes more energy to extract than it provides around mid-century (though Shell put it at 2026 the last time I looked). How do we then find an equivalently relatively low cost alternative to this commodity? I have not yet seen anything with the same MJ/kg or MJ/m^3. (It goes without saying that fracked oil and tarsand won’t substitute for conventional oil whatever the snake oil salesmen say.)

2. The above problem thus limits the possibility of extracting other forms of energy cheaply (gas or coal).

3. The above problem point 1 also limits the possibility of extracting materials, manufacturing and building/deploying other forms of energy.

4. I don’t see algae or E-Coli working since both need concentrated CO2 to work and they’d use up vast amounts of space.

To build a thorium reactor one needs to be able to cheaply extract the thorium. The same goes for any other form of nuclear fuel.

But reactors don’t make energy that can help to extract resources from a mine or transport it cheaply. Electrical motors don’t have that kind of power curve.

To build Dyson Spheres or thorium reactors then WE MUST START NOW! Waiting until our oil is no longer economic to extract (despite there being lots of it left in the ground) is not going to work.

So may I ask has anyone costed this up? And would thin film solar work? And what about beaming it back to the moon and then the earth? We need some costings.

YOUR COMMENT Login or Register to post a comment.

Next entry: James Cameron Breaks Solo Dive Record

Previous entry: Record Battery Energy Density in Context [Updated]