More than 40 years have passed since Freeman Dyson suggested that advanced technological civilizations are likely to dismantle planets in their solar systems to harvest all of the energy their stars wastefully radiate into space. Clearly this was an idea that was ahead of its time. Since that time, dozens of SETI searches have been conducted and almost all of them have focused their attention on stars which by definition cannot be the advanced civilizations that Dyson envisioned. I will review the data that created the confusion between Dyson spheres and Dyson shells. The sources that disprove Dyson spheres while still allowing Dyson shells will be discussed. The use of outmoded ideas that have biased the few searches for Dyson Shells that have occurred will be pointed out. An update of the concept of Dyson shells to include our current knowledge of biotechnology, nanotechnology and computer science will be explored. Finally, an approach to setting limits on the abundance of Dyson shells in our galaxy using existing optical astronomical data and future optical satellites will be proposed.
Keywords: Dyson spheres, Dyson shells, evolution, megascale engineering, optical SETI, nanotechnology, technological civilizations
To date searches for these so-called ``Dyson Spheres'' have been relatively limited. The most prominent being those conducted by Drs. Jugaku and Nishimura [3,4,5,6]. In part this may be due to the difficulties in distinguishing such advanced civilizations from naturally occuring dust enshrouded stars that also produce large amounts of infrared radiation. This was pointed out by Sagan & Walker [7] and reiterated by Harwit [8]. Unfortunately, many astronomers are unlikely to be aware of the distinguishing criteria pointed out by Kardashev in 1981 [9].
The purpose of this paper is to recap the history of the development of ``Dyson Shells'', point out where misconceptions may have arisen and correct these so that future astronomers may be more aware of the possibilities that exist.
From the above (calculations) it follows that the construction of a spherical shell around a star from the material present in its planetary system is an impossible task. What is possible, however is to have a large number of independent space structures in orbit around the star, but these would intercept only a relative fraction (1%) of the star's radiation. Consequently such stars, would display a normal spectrum with only a small excess in the infrared.[10]This analysis is based on a highly anthropocentric view, that Dyson Shells must consist of O'Neill style habitats [11], designed for liquid water based life forms. This perspective may have been suggested by Dyson himself, who in his original paper stated:
One should expect that, within a few thousand years of its entering the stage of industrial development, any intelligent species should be found occupying an artificial biosphere which completely surrounds its parent star.[2]Whether the statements of Dyson and Papagiannis are in conflict depends on ones interpretation of the term ``completely surrounds''. Dyson continues:
The most likely habitat for such beings would be a dark object, having a size comparable with the Earth's orbit, and a surface temperature of 200 deg. to 300 deg. K. Such a dark object would be radiating as copiously as the star which is hidden inside it, but the radiation would be in the far infrared, around 10 microns wavelength.[2]In Dyson's view the star is ``hidden'', while from the Papagiannis perspective it is not. A significant problem with both positions is they assume that the temperature of a ``biosphere'' must be that required for ``liquid water''. Yet, 11 years later, during the joint Soviet-American conference on CETI, the flaws in Dyson's perspective were pointed out by the Artificial Intelligence pioneer Marvin Minsky:
Since radiation at any temperature above 3°K is wasteful and a squandering of natural resources, the higher the civilization, the lower the infrared radiation. We should look for extended sources of 4°K radiation. There should be very few natural such sources.[12]Papagiannis performed his calculations and reached his conclusions 12 years after Minsky's comments were published. We may reasonably assume that Papagiannis must have read them but that he must have been enamored with the space colonization ideas favored by Gerard O'Neill [11]. Or he might have limited his vision to that suggested by Dyson's response to letters criticizing his idea [13]:
A solid shell or ring surrounding a star is mechanically impossible. The form of ``biosphere'' which I envisaged consists of a loose collection or swarm of objects traveling on independent orbits around the star. The size and shape of the individual objects would be chosen to suit the inhabitants. I did not indulge in speculations concerning the constructional details of the biosphere, since the expected emission of infrared radiation is independent of such details.1Dyson may be forgiven because in 1960 computers were a relatively new, and ideas such as solar power satellites and such fields as artificial intelligence, artificial life and genome engineering would have been difficult to predict. But Papagiannis, who had the benefit of the extensive CETI discussions in 1971[14], should have considered the possibility that a civilization might envelop its star in lightweight solar power collectors if only to power a long-distance CETI transmitter! The reasons for his emphasis on the burdensome mass requirements of primitive space habitats is unclear. He further failed to consider the development of robots[15] or ubiquitous computing [16] that would enable Dyson Shell architectures quite different from his vision.
Perhaps unknown to Papagiannis, was a detailed analysis of the characteristics of Dyson Shells done by Dr. K. G. Suffern of the Department of Applied Mathematics at the University of Sydney that was published in 1977.[17] While the general tone of the analysis is to draw conclusions similar to those of Papagiannis, Suffern concluded:
Of course this is not an objection in principle to optically thick Dyson spheres because in principle the radiation from any black body with a temperature T > 2.7°K is thermodynamically useful. It just points out a practical difficulty that thermodynamics imposes upon optically thick biospheres.Suffern's calculations showed that O'Neill style colonies would not effectively hide a star, but his perspective is somewhat more open minded in that he at least allows that there may be architectures that could hide the star.
It is safe to say that Dyson's use of the term ``dark object'' and the subsequent popularization of ``hollow balls'' by Science News Letters [13], rather than his intended ``shells'' [18] was the source of much confusion. These misunderstandings may have lead Davydov to argue that spheres were impossible because materials were not strong enough.[19] He considered orbiting shells, but thought they were unfeasible as well due to orbital spacing requirements. Later work by Pokrovskii [20,21,22] and Ulubekov [23] linked Dyson's ideas to Konstantin Tsiolkovsky's much earlier ``Ethereal Cities'' and suggested that shells were indeed feasible. They felt the outer layers could have temperatures ranging from 84-180°K. Because these ideas were discussed primarily in Russia, the knowledge was not widely circulated and recently published books by very serious engineers make such statements such as ``Dyson Spheres are impractical'' and ``the implausibility of the Dyson Sphere is so extreme that it has even been criticized by science fiction writers.'' [24].
Dyson Shells are implausible or impractical only if you envision them as true spheres or primitive human habitats. And some science fiction writers, by stretching their imagination to include engineering methods such as force transfer using high speed circulating fluid streams, have described plausible Dyson Spheres [25]!
The Papagiannis and Suffern calculations are based on the assumption that we are limited to the available ``construction material'' present in our solar system. Papagiannis uses a figure of 10 Earth masses, while Suffern uses a figure of 30 Earth masses from Bracewell [27]. The author estimates that the solar system contains minimum of 3-10 times the amount of construction material considered by Suffern and Papagiannis (100 Earth masses; 5.9×1026 kg). Why is this?
During the last 25 years, it has become clear that the capabilities of traditional construction materials (iron, aluminum, glass) is somewhat limited, and that structures will ultimately be constructed out of materials with more robust physical properties. These include diamond and sapphire as well as ceramics including carbides, oxides and nitrides. So substances such as methane, ammonia, carbon dioxide and carbon monoxide, which are abundant in the bodies with orbits beyond Saturn are more useful for construction purposes than was previously thought. Water ice is also present in significant quantities in these bodies and may be considered a structural material if it is kept significantly below its freezing point. In our solar system, the excess of hydrogen and oxygen over carbon (and other metals) dictates that the most efficient materials usage must use water in lieu of other materials as much as possible. In addition, the most abundant elements, hydrogen and helium, as highly pressurized gases or possibly in liquid or solid form, may be used as radiation shields, leaving other materials available for more important uses.
The classic O'Neill habitat contains 4% aluminum, 2% glass, 10% water, and 82% soil, rock and construction materials [11]. How can we substitute advanced materials into these habitats? The best mass utilization would result from constructing all structural elements out of carbon (as diamond or buckytubes), sapphire (AlO), silicon carbide (SiC) and hematite (FeO). Metals would only be used in situations requiring great ductility or electrical conductivity. Diamond has an elasticity 5× that of high carbon steel and 13× that of aluminum alloys. Replacing the aluminum and steel in the original designs with diamond and sapphire (which are also more abundant) reduces the structural material mass requirements by 10 times or more.
For example, advanced colonies could contain diamond walls of sufficient strength to hold an atmosphere, surrounded by silica aerogel insulation to keep heat in and cold out, surrounded by ice containing bubbles filled with high pressure hydrogen gas as a radiation and meteor shield, coated with ultrathin high efficiency solar cells that efficiently harvest incoming visible and infrared photons and covert them into electricity.
The traditional agriculture methods envisioned by O'Neill are not those that would be used by an advanced civilization. The soil requirements are presumably those required for growing food. Yet we know that traditional foods may be grown quite effectively hydroponically. That would eliminate the soil requirement completely and replace it with a water requirement. Water as previously noted is one of the most abundant materials in the solar system. If the population adapts to non-traditional foods, or food processing can manipulate basic ingredients into traditional forms, one may envision very thin orbiting pancakes (agrisats) with glass surfaces facing the sun and ceramic radiators facing away from the sun filled with water containing algae like organisms that are bioengineered to manufacture and store a completely balanced set of nutrients. These would be very thin structures, perhaps only a few cm thick. These agrisats could periodically dock with habitat colonies where the nutrients would be harvested and the agrisats refertilized with carbon dioxide, ammonia and trace elements.
Advanced societies would attempt to make the most efficient use of the available energy as possible. Initially, this could be based on multi-layer solar cells (30-40% efficiencies). Ultimately it may use solar concentrators with high temperature heat engines that radiate at the background radiation temperature. The temperature differential (2900°K - 3°K), potentially allows 99% efficiencies. Harvested power could be used to produce hydrogen ions (H) that are utilized by mitochondria- or bacteria-like bioreactors to synthesize ATP and subsequently carbohydrates and proteins. Ultimately the losses involved in these chemical processes might drive civilizations to engineer individuals to be able to utilize sunlight or electricity directly. There are only mental barriers that would prevent extraterrestrial bioengineers from producing chloroplast-like energy conversion units in their own skin. Humans unfortunately have too little skin area to power their 100W metabolisms. They would require the addition of wings with increased surface area, or lower operating temperatures to pursue this approach.
Thus if we divide the multi-purpose O'Neill style habitats into powersats and agrisats that harvest most of the available sunlight and habitats that house populations at New York or Tokyo density levels then the resources of a solar system may be utilized much more efficiently.
So, only primitive civilizations (with late 20th century technologies) would limit themselves to collecting less than 1% of the solar output of their star. Using advanced materials and technologies, advanced Type-II civilizations would harvest all of the power produced by their stars as envisioned by Dyson. As the civilization evolves, becoming increasingly efficient, they would radiate heat at ever lower temperatures, potentially approaching the background temperature as Minsky and Suffern have suggested.
Life supporting chemistries that do not require liquid water have been postulated [29,30], and NASA scientists are beginning to engineer artificial life [31]. This suggests that life forms existing outside the temperatures ranges envisioned by Dyson can be discovered or engineered. Ultimately advanced civilizations should apply molecular nanotechnology[32,33] which will allow life-forms based on machines operating over temperatures ranging from 3°K to 2900°K or higher.
The time spans for technological civilizations of intelligent beings to evolve from very primitive states (e.g. Homo erectus) to the limits imposed by classical laws of physics are very short ( years) compared with the galactic time scales ( years). Work by the author has shown, that a Type II civilization could rapidly form an highly integrated meta-mind with a computational capacity times that of a human brain.[34] This meta-mind would occupy the computers of star-enveloping solar power satellites that do effectively hide the star. Conclusions regarding the forms Dyson Shells will take, the failures of SETI efforts thus far[35], and the Fermi Paradox[36] are unlikely to be valid if one assumes an anthropocentric starting point as most individuals unconsciously do.
Discussions about advanced civilizations are likely to be most valid, only if one assumes capabilities and capacities at the limits of known (experimentally validated) laws of physics. Unfortunately, astronomers and SETI researchers continue to look for signs of life assuming human perspectives (i.e. visible stars) rather than perspectives dictated by the limits of the laws of physics. Only whole sky surveys, with long exposure times in the mid-far IR ranges are likely to detect low temperature Dyson Shells. Unfortunately, even for expensive telescopes, like SIRTF,[37] the detectors for these wavelengths, typically, Si:As, Ge:Ga and stressed Ge:Ga are still very primitive.2 It is open to debate whether very advanced telescopes, to be launched over the next decade would have the capabilities required to provide definitive evidence for or against these objects.
Dyson Spheres are either impossible or highly difficult to construct. Dyson Shells are feasible and may be constructed by civilizations only slightly more advanced than ours. This will cause a star to shift its radiation emission from the visible into the infrared and eventually ``disappear''. There do appear to be methods by which the natural and artificial sources of infrared radiation may be distinguished.
Advanced civilizations are unlikely to have the temperatures of ``biospheres''. Depending on the stage of their construction, and the materials at their disposal, they may be much hotter or much colder than liquid water. If we had the technology to conduct extensive surveys for brown dwarfs,5 we might find that many of them are being disassembled to supply fuel and construction materials for advanced meta-minds. The unresolved problems of the missing baryonic dark matter[39,40] and the gravitational microlensing observations6 hint at the possibility that such entities may exist.
Unexplained or incompletely studied astrophysical phenomena such as odd star populations of the galaxy NGC 5907[42] or the asymmetry of increases and decreases in the brightness of long-period variable stars[43] provide us with a number of locations that may be studied for signs of Dyson Shells. If we free ourselves from anthropocentric perspectives and combine the ideas of Dyson, Minsky and Suffern as well as the technological progress of recent decades, we can envision advanced civilizations at the limits of physical laws. Observations directed towards stars decreasing in visual magnitude or searching for stellar occultations by large cold dark objects, merit serious consideration as future strategies in optical SETI.