Worldbuilding Space Megastructures: Beyond Dyson Spheres

Space stations, spaceships, and starships are fundamental to space opera and many other subgenres of science fiction; vehicles and vehicular communities traveling through space feature heavily in these worlds and stories set in them, for the simple reason that any travel between points in outer space necessarily involves vehicles, unless you’re using wormholes or some other sort of portal. Indeed, Earth itself can be thought of as a self-sustaining more-or-less self-contained vehicle weighing in at one Earth mass; “Spaceship Earth” is a term in our lexicon for good reason.

In more advanced or speculative settings in science fiction, ships gradually become larger and larger, far beyond the sizes of the largest spaceships ever actually launched from Earth. When going beyond vehicles the size of the Apollo capsule or the International Space Station, the first change is that centrifugal force to simulate gravity starts to be viable. This requires a centrifuge to do so, so spaceships, space stations, and space habitats past this range will be more spherical, cylindrical, or disk-shaped, much like the famous space habitats proposed by Gerard O’Neill and many others over the years. An exploration of these space habitats would be fascinating enough, but the question we will explore in this post is how large could these structures and even larger structures be made? What is the ultimate limit?

Everything’s bigger with Carbon

The O’Neill cylinders as originally proposed were to be built out of steel, but carbon nanotubes, if mass-produced, would enable much larger structures. How large? Tom McKendree of NASA proposed in 2000 that an O’Neill cylinder built out of carbon nanotubes could easily be 580 miles in diameter and 2900 miles long. That’s big. For perspective, the surface area on the inside of such a cylinder would be 5 million square miles, almost as much area as all of Russia put together. The theoretical limit of carbon nanotube construction is around 1200 miles in diameter and 6000 miles in length, yielding a surface area of 24 million square miles, greater than Eurasia’s area. It wouldn’t take very many of these things to equal all of Earth’s surface area, both land and ocean: 196 million square miles.

The sheer size of these things is illustrated by the fact that the diameter and length, and thus volume, of such a large structure is actually bigger than many of the smaller planets. The mass, however, would be smaller as their interiors would be filled with air as opposed to rock and ice. One interesting feature is that the deep interiors of these cylinders, near the spin axis, would be depleted of air and be a near vacuum due to the centrifugal force pulling the air down toward the surface.

Along similar lines as the McKendree cylinder, Forrest Bishop proposed in 1997 that a Stanford torus, a rotating wheel-shaped installation big enough to be a space habitat, could be scaled up using carbon nanotubes to 1200 miles in diameter and 300 miles in width. The surface area thus produced would be 1.2 million square miles, comparable to Argentina or India. Perhaps the most striking feature of this habitat, named the Bishop ring, is that it is so large that unlike the Stanford torus the roof wouldn’t even need to be enclosed. Atmospheric retention walls 120 miles in height would together with the centrifugal force be enough to hold in all of the air. Open-air living in space might sound crazy, but it is technically achievable (albeit not affordable) even using current technology, let alone what will become available in the future.

“That’s no Planet! It’s a Space Habitat.”

Forrest Bishop himself may have been inspired by a similar concept in Ian M. Banks’s Culture novels, which was even larger than the Bishop ring. These structures were called “Orbitals”, designed to rotate once every 24 hours and provide a centrifugal force of 1g on the surface, requiring them to be an incredible 1.9 million miles in diameter, sporting over a thousand times as much area as the Earth. Like the Bishop ring, the Orbitals were also open-air space habitats. According to the truly excellent Orion’s Arm worldbuilding project, “Banks orbitals”, as they are known in that setting, require exotic matter in their construction; no normal material can stand up to the stress imposed by such a large rotating structure. Obviously we won’t be constructing such a habitat any time soon. Nevertheless such exotic matter may one day be available in real life, and would be trivial to introduce even into a hard science fiction world.

These megastructures are around the size of a planet; something around the size of the Bishop ring, containing volume comparable to the smallest planets, is about the biggest structure that would be technically feasible for us to build now or in the near future. However, looser definitions of “structure” enable us to go far beyond even this range; swarms of smaller structures may even be more efficient than having one or a few big ones, after all, and the surface-area-to-volume ratio on these rings and cylinders is so efficient that an inner solar system could easily become a thicket of these habitats.

Dyson Spheres: the original Megastructure

The most famous “structure” of this sort is undoubtedly the Dyson sphere. Although popular culture often supposes a solid sphere around a star to capture all of its energy, this would actually be a very inefficient way to go about harvesting all of a star’s energy; a slicker method would be to amass a swarm of solar power satellites gradually until the vast majority of the star’s output is converted. This swarm concept was actually Freeman Dyson’s original proposal in his 1960 paper that made the idea famous, though it was first described by Olaf Stapledon in his 1937 novel Star Maker.

One of the primary reasons Freeman Dyson floated the idea is that such a sphere, or more accurately swarm, would convert a star’s normal energy into waste heat, thus leading to a heavily infrared output that would in principle be an easily detectable sign of advanced alien civilizations. To date no such signature has been definitively detected, though there are tantalizing hints here and there among the stars. In recent years whole swaths of the sky have been scanned with the aim of detecting galaxies that may be covered in Dyson spheres. Galaxies were chosen because we can see whole galaxies out to a very long distance, far further than individual stars can be detected, thus the results would be more representative of the universe than a search limited to nearby regions of our own galaxy. The searches have virtually ruled out the possibility that there is a single Dyson-swarm-covered galaxy remotely near us.

This has been taken by some as ruling out the possibility of any nearby Type III civilizations on the Kardashev scale, but all it really proves is that there are no galaxies near us that consist predominately of stars covered by Dyson swarms. A civilization advanced enough to control an entire galaxy would almost certainly have scientific knowledge vastly superior to our own, and it would be foolish to assume there is no power source out there better than Dyson spheres. After all, one of the most promising power sources we know of, from powering starships to sustaining a civilization in the distant future of the universe, is Hawking radiation from black holes, a mechanism that as recently as the 1960s was unknown to our science. Analogous discoveries could easily have been made by a Type III civilization when they became more advanced than us, perhaps obviating Dyson swarms as a cost-effective power source.

Nicoll-Dyson Beams: turning a Star into a Laser

Nevertheless, the concept remains interesting, even if observation seems to point toward the conclusion that it is a technological dead end, assuming one accepts the premise that advanced alien civilizations are out there. Much more recently it has been pointed out that a Dyson swarm can in principle direct the energy collected from the star outwards toward a target; in this way an entire star’s output can be turned into a laser beam. This could be used for spacecraft propulsion, where truly titanic energies could be utilized for beam-powered propulsion, or for destructive purposes. This is called a “Nicoll-Dyson beam”, after James Nicoll and of course Freeman Dyson. Such a powerful laser could easily cause planetary-scale destruction at interstellar range, and is one candidate for the ultimate directed-energy weapon.

Nicoll-Dyson beams could also be used, as in the Orion’s Arm setting, to send laser messages out to great distances unreachable by conventional means. One use I haven’t heard about is the amplification of sunlight to distant targets, a similar use as the weapon but less intense; after all, the same sort of laser that can fry a target out to ten light-years would merely make a bright light at, say, fifty light-years. There may be uses for such intense illumination in the interstellar void. Outer parts of solar systems may also be heated with less-concentrated versions of the same concept.

The ultimate cooling Laser: a real Freeze Ray?

Most exotically of all, though, would be the possibility of using a Nicoll-Dyson beam for laser cooling. Laser cooling works by making a atoms absorb and re-emit a photon, thus changing its momentum; for a group of atoms the more homogeneous the velocity the lower the thermodynamic temperature. Applying a laser to compress the velocity distribution of a group of particles thus lowers their temperature; in the laboratory this is done on small groups of atoms to cool them down to near absolute zero. Laser cooling is usually associated with atomic-scale manipulation, but in principle it can work on any number of atoms. In 2007 at MIT and 2010 at CalTech objects a micrometer across (vastly larger than a few atoms) were successfully laser-cooled down to below 1 Kelvin.

If objects a micrometer across can be cooled with laboratory lasers, there is no particular reason to believe that much larger objects, such as a man or a tree, couldn’t be cooled down to cryogenic temperatures using much larger lasers. Indeed, this would bring to life the science-fiction concept of a “freeze ray” that is so often derided as being unrealistic. In principle an entire planet could be laser-cooled; this could be very effective wherever cooling is needed, such as certain inner-solar-system environments that may be out there. Terraforming of Venus is one application where such a laser would prove very useful.

Of course, the weaponization potential of such a laser cooling system is quite obvious, and may prove far more effective than an ordinary heating laser; after all, it’s usually much easier to put heat into something than to take it out, and an Earth-like planet frozen by a laser would have a much easier time recovering to a condition suitable for the conquerors’ use than an Earth-like planet turned into molten rock. That’s something I hardly ever see in science fiction or space opera, especially the harder versions. It might make a hook for a very interesting story.

Dyson swarms around a single star aren’t the only megastructure that could be engineered; theoretically Dyson swarms could encompass all the stars in one galaxy or even an arbitrarily high number of galaxies, gathering energies dozens of orders of magnitude greater than modern human civilization. Nicoll-Dyson beams as well can be compounded with billions of stars and even galaxies. Obviously we do not live in such a universe (or such a neighborhood of our universe anyway) because we don’t see dark infrared-dominated galaxies sending out galactic-scale laser beams hither and yon, but it is perfectly feasible according to the laws of physics and would make for a fascinating setting for a space opera story.

Dyson swarms scaled up aren’t the only sort of “structure” that could be built; assuming sufficiently strong materials could be engineered or discovered, space habitats need not be limited to the size of a small planet or even a Culture-style Orbital. Robert Freitas in his work “Xenology” outlines much larger structures that would take multiple solar systems’ worth of material to build.

Truly interstellar-scale Space Habitats

“Megaring” is a Banks Orbital or Bishop ring scaled up to twenty light-years in diameter, a truly interstellar-scale structure. The habitable surface is a million kilometers wide, with atmospheric retaining walls a titanic 60,000 kilometers in height. Interestingly the structure is so massive that it has 0.001g worth of gravity (henceforth we’ll call it a milligee) just from its own mass. The ring would rotate at a peppy 10% of light speed; as rings grow in size they need to rotate faster at the edge in order to generate artificial gravity. The rotation speed of the Banks Orbital, for example, is around 134 kilometers per second. At a scale of twenty light-years, even this near-relativistic acceleration only generates 0.03 milligees of gravity; the primary function of the gravity at that point would be to hold in the atmosphere rather than provide Earth-like gravity.

The structure’s atmosphere, because of the low gravity, would only fade out to 0.5 atmospheres worth of pressure at an altitude of 6000 kilometers. A man that could jump a meter on Earth could jump a kilometer on a Megaring, and it would take him seven minutes to come down to the surface; the gravity would be low enough that humans could fly thousands of feet up into the sky unaided.

330 solar masses would be required for the frame, and 3130 solar masses worth of material would be required for the atmosphere. Stars could be put in orbits around the ring set up so as to provide all parts of it Earth-like illumination and climate; this would take 126,000 solar masses. Megaring would be a titanic project, but for a galactic or multi-galactic civilization the cost may be worth it to build a structure with as much living space as a trillion Earths.

An even bigger Megaring would be possible. Freitas’s example of a Big Megaring is a thousand light-years in radius, spinning at 10% of light speed to generate a minuscule gravity of a few tens of microgees, leading to the atmosphere not dropping off to 0.5 atmospheres of pressure for tens of millions of kilometers, comparable to the distance between Earth and Mars. To imagine that much of an area being breathable and flyable is mind-boggling; realistically such a habitat would be more of an air-world dominated by atmospheric living rather than anything on the surface, perhaps resembling the environment in Larry Niven’s Integral Trees novel, the Smoke Ring. In particular the near-weightlessness would be a common point between the two settings.

The mass needed to build this structure is 34 million solar masses for the structure and a truly monstrous 3 billion solar masses for the atmosphere. This is starting to be a non-negligible percentage of a Milky-Way-sized galaxy’s entire mass! 130 quadrillion Earths of living space in terms of surface area would be provided for in such a structure.

The ultimate Space Megastructures

Freitas’s last and largest concept is the Megasphere, a sphere 200 light-years in diameter, sporting a surface area of more than 130,000 square light-years, an incredible twenty-two orders of magnitude more surface area than the Earth. Surface gravity would once again be in the milligee range, and 6000 kilometers is where the atmosphere fades out to 0.5 atmospheres of pressure. 1300 galactic masses are required to build the structure in total. The energy required to maintain such a structure would seriously tax the abilities of even a galactic civilization; a supercluster-scale civilization or ideally one even larger would probably be required to sustainably build such structures.

This Megasphere is perhaps the ultimate megastructure, since building larger structures introduces some side effects that are rather interesting. A Megasphere that is 2000 light-years in diameter could be constructed with 4000 galactic masses of material. However, if the surface atmospheric pressure is greater than 0.01 atmospheres the structure is smaller than its own Schwarzschild radius and becomes a black hole!

Because the structure is extremely large the gravitational stresses involved wouldn’t be particularly punishing; it would be an example of a “gentle” supermassive black hole. Nevertheless, an event horizon will form around the structure and any trips inside will be one-way, as the escape velocity is greater than the speed of light. If one is willing to accept this, there may be no limit to the size of space habitats that could be constructed.

Indeed, if there were a halo of similarly dense objects around galaxies and further out into the void it would look rather similar to the mysterious “dark matter” that surrounds us! Black holes have been suggested as dark matter candidates, and for all we know they may be artificial, hosting who-knows-how-many aliens shielded behind impenetrable event horizons. This doesn’t seem very likely, since one would assume at least some of these advanced civilizations would not have completed the process of sequestering themselves in black holes yet and would be more easily visible, but it is an idea I hardly ever see mentioned that is intriguing as far as science-fiction worldbuilding is concerned.

It also raises the question of why an advanced civilization would want to sequester themselves inside a black hole in the first place, but there may be “basement universe”-style possibilities, or even access into actual basement universes, inside black holes. Many have pointed out that we know nothing about what is actually behind the event horizons of black holes, though there are of course many educated guesses. The one-way nature of travel into the black hole would seem to be a potentially fatal handicap, but for all we know it may be possible to use wormholes to link the interior of a black hole with the wider universe or the interior of other black holes. If that is the case, or if while worldbuilding one assumes it to be the case, then black holes might be great places to live and the ultimate option, short of wormholes, for traveling in style throughout the universe.


We have seen in this brief exploration of megastructures in space that there are many possibilities that go well beyond the tried-and-true O’Neill cylinder and Dyson sphere range, and that are much more interesting to boot. Some of these possibilities can even easily be turned into weapons or artificial illumination for various purposes, and others are truly exotic, such as the habitats behind event horizons.

The Megaspheres and especially Megarings are very interesting because it is readily imaginable that some civilization somewhere, perhaps in another galaxy, more comparable to present-day humanity may run into one of these structures that is in ruins or otherwise abandoned by its creators and make use of it themselves, perhaps without even understanding how it works or having any knowledge of its origin or purpose. That would make an interesting premise for a science fiction story, and would provide almost limitless possibilities for worldbuilding. Sometimes this is done with Dyson spheres, but the Megaring makes a Dyson sphere look like a pipsqueak by comparison.

Many of these structures could be similarly used by science fiction worldbuilders, writers, and artists as elements in settings, perhaps the missing link to tell a compelling story, or perhaps as an inspiration to create new stories in some of the most imaginative and also most difficult and seldom-used genres in all of fiction, space opera and the ultra-speculative parts of science fiction.

Leave a Reply

Your email address will not be published. Required fields are marked *