Gerard K. O’Neill, President
Space Studies Institute
The rationale of the plan which follows rests on three premises. They are based on realities. Those realities are: A plurality of nations now possesses or soon will acquire overall technical capability, including, but not limited to, the space arena, which equals or surpasses that of the U.S. Therefore, no U.S. governmental space plan can be written or judged in isolation from those of other nations and non-governmental entities.
The economic deterioration of the Soviet bloc over the past 10-15 years, and the consequent rapid political changes occurring in the Soviet bloc, do not ensure that there will not be a major shooting war, but do make the conventional Cold War a thing of the past. Therefore, no space effort based on a posturing either to show technological force (Apollo) or to thaw ideological barriers (Apollo-Soyuz or a hypothetical U.S.-U.S.S.R. manned trip to Mars) makes any sense in the world of 1990 and beyond.
The major new threat is the highly aggressive and successful penetration by the nations of the Pacific Rim in recent years into U.S. and global markets. The 1992 economic unification of Europe will increase market pressures on the United States.
These realities combine to suggest strongly that over the next decades the United States will have neither sufficient capital resources nor a military reason to subsidize a governmental space program that is not market driven. Space is a high technology arena, and the U.S. governmental space program must be reoriented radically to address the primary problem of U.S. industrial competitiveness in high technology.
The consequent first premise is that a U.S. national space program must be highly cost effective, must open up substantial export markets, and must be planned in conjunction with a cooperative global program which both draws from and benefits many nations. There is no place in such a program for expensive “flags and footprints” spectaculars, though such events were justified in the very different world of the 1960’s. Nor is there any place for expensive programs, alleged to be economically justified, but which are, in fact, transparently spurious. Zero-gravity manufacturing is an obvious example.
A corollary to the first premise is that in the highly competitive decades to come, the U.S. cannot afford expensive mistakes which waste time and allow competitors to gain commanding positions. The scrapping of the very successful 1970 U.S. fleet of expendable launch vehicles in favor of the Space Shuttle is now generally recognized, even in the pages of so friendly a journal as the AIAA’s Aerospace America, as a disastrous mistake which cost the U.S. at least fifteen years. We cannot afford further such mistakes.
The second premise is that the time scale for substantial accomplishment (measured as substantial economic or scientific return) must be shortened from the leisurely pace which characterized an isolated, protected program funded by a government dominant in the world to the rapid pace necessary to outrun today’s fierce global competition. That means five years maximum, rather than 20 to 40 years, for significant return on investment.
A corollary to the second premise is that a large fraction, though by no means all, of the personnel now in NASA are not well-suited by talent and inclination to a program capable of meeting U.S. space needs in the real world of the next decades. Fortunately, an exception is that NASA has an excellent Administrator.
The positive side of the corollary is that a time-urgent, goal-oriented program will attract to NASA vigorous, talented, aggressive young people who will be the 1990’s counterparts of the 1960’s tiger team. NASA has great need of those people.
The third premise, which enables large-scale results to be obtained, even within the constraints of the first two premises, is that a great deal of cost can be saved, and the time scales for all space activities drastically shortened, by making the maximum use of resources which are already located at the top of Earth’s gravity well. Those resources are primarily the intense solar energy which is available full-time everywhere in space except in the shadows of planets, and the abundant resources of oxygen, silicon and metals available on the surface of the Moon. We need to pay the price of launching out of the Earth’s gravity well only those items which can leverage themselves to a high payback: information, intelligence (computerized or in the human brain, depending on the mission) and sophisticated tools – NOT materials (such as oxygen) or heavy structures. That is the way we opened the New World of the Americas, and that is the only practical way to open today’s New World of space.
It is not generally recognized, except by rocket specialists, how very large is the beneficial effect of using materials available at the top of Earth’s gravity well. Up to now, all materials used in space have had to be lifted from the Earth’s surface, which is at the bottom of that well. The ratio of a rocket’s liftoff weight (made up mainly of propellants) to final weight (made up of the remaining upper-stage rocket vehicle structure plus payload) is typically about sixty to one, for weights lifted to geostationary orbit or an escape trajectory. Deducting the vehicle structure, the useful payload is typically only about 1/100th of the liftoff weight.
By contrast, in the case of a rocket lifting a payload to escape from the Moon, the payload plus remaining vehicle structure can be as much as 60% of the lift-off weight, instead of less than two percent as in lift from the Earth. The overall advantage for launch from the Moon is 35-fold, an enormous factor. In addition, the fact that the Moon is in vacuum allows the use of efficient launching machines located on the Moon, operating on solar electric power and not requiring rocket fuels, to launch materials into space.
Programs to reduce launch costs from the Earth have been tried and have failed and are now being tried again. If a new program of that kind is successful, it will reduce somewhat, though not to an important extent, the cost of placing machinery on the Moon to bring materials out in quantity and at very low cost. But for any space program it would be a mistake to concentrate solely on the issue of launch costs from the Earth. Waiting for reduced launch costs would delay us still longer, and would be all too likely to reduce the pressures on us to think intelligently and design for high leverage from the payloads we do lift. Recent history suggests that our global competitors will not be so shortsighted.
We will be well served if we design the U.S. national space program consistently with the three premises just listed. We will be served even better if we build general capabilities rather than insisting upon a plan which is to remain rigid over decades. We are no longer alone in the world, and the ability to respond quickly and massively to changing market opportunities and market threats will be our greatest asset. As we learned to our cost over the past fifteen years, the world will not wait while we scrap general capability in favor of a narrow, parochial program based on a false premise.
The U.S. national space program, meshing with a cooperative global program, should include three components. They are resource prospecting, manufacturing of economically productive structures in space, and space science.
For all three components, cost effectiveness dictates that we develop modular systems which can be tried, proven and successively improved, and which can then be built in substantial production runs. That is the way to achieve low costs, high reliability and rapid response. The issue of whether a module – for example, a rocket vehicle – is fully reusable or not is a red herring, not a key to success. An inexpensive module which can be mass-produced and thrown away may be more cost effective than a fully reusable module which is fragile and pushes the limits of technology in order to achieve equal performance.
Surveying the inner solar system for its material resources, for use in refueling spacecraft and in building structures to be used in space, is an activity wholly absent from today’s space program. It is also, unfortunately, virtually absent from NASA’s most recent internally generated proposal as to its own future. Most space scientists would see it only as an activity subtracting in a zero-sum manner from their own pet projects. Yet it is urgently needed by our nation to build the knowledge base that will enable us to develop a cost effective space program.
Institutionally, our government should establish a continuing advisory group made up of scientists and engineers who have an established reputation in the field of space resources. To give muscle to that group’s recommendations, the government needs to direct that some fixed small percentage of NASA’s total budget be devoted to resource prospecting missions. The necessary missions are modest in number, size, technology and cost, and all of them can be unmanned. But they need to be flown, and without an institutional change none of them will be. For comparison, even the most demanding of those missions is technologically easy compared to the 1970’s Viking landing missions to Mars, or the forthcoming Galileo mission to Jupiter. Some of these missions may carry scientific experiments as well, but decisions as to whether they are flown must not be put in the hands of the space science community.
The list of resource prospecting missions, in each case best carried out as a series of flights by relatively simple spacecraft, includes:
Lunar polar orbiters, using remote sensors to search for water-ice in the permanently shadowed frozen craters near the lunar poles. If ice is found, the Moon can become a refueling base not only for oxygen, which constitutes 86% by weight of rocket propellant and is already known to be the most abundant element on the lunar surface, but also for hydrogen, which is the remaining 14% by weight of rocket propellant. The first such mission, if started now, could be designed, built, flown and return information within five years, around 1995.
Lunar landers to probe the surface at locations identified by the orbiter missions. These landers can be far simpler than spacecraft which must travel beyond the Moon; the reason is that they can be teleoperated by commands from the Earth. That method is practical at lunar distance, where the round trip signal time lag is only 2.7 seconds. it is impractical at any of the planets, which are about 1,000 times farther away. The time scale for lunar lander missions should be around 1995 to 1997.
Spacecraft equipped with small telescopes capable of searching, in the infrared or optical wavelengths, for asteroids which approach or cross the orbit of the Earth. Already, several dozen asteroids with such orbits have been found by scientists using telescopes at Mt. Palomar. Even very modest additional funding for that work could increase greatly the yield of newly discovered Earth-crossing asteroids. It might result in the discovery of an asteroid from which materials could be returned to high orbit above the Earth at a particularly low cost in propellant.
But any location on the Earth’s surface is less than optimum for searches of that kind. Telescopes in orbit over the Earth, observing at night, and in orbit around the Moon, in periods when the dark side of the Moon is not exposed to reflected Earth-light, can observe with especially high sensitivity.
A particularly interesting trajectory for asteroid search missions is a circular orbit slightly closer to the Sun than is the Earth. Theoretical work indicates that there may be material trapped stably in the Earth’s own orbit around the Sun. If such material were found, it could be returned to a high orbit over the Earth at almost no cost in propellant. For scale, an asteroidal fragment of diameter only 40 meters would contain about 100,000 tons of material. To lift that much from the Earth, even if launch costs were reduced 20-fold from those of today, would cost 90 billion dollars.
To search for such material, a telescope in circular orbit with the sun at its back, looking outward toward Earth orbit, would be ideal. The time scale for asteroidal search missions of these kinds should be about seven years. That is, if a decision were made now to undertake them, their first flights and first information return could occur by around 1997. Each such mission, including development and first flight, would be in the 20 to 70 million dollar range, a tiny fraction of the cost of any major space science mission such as Viking or Galileo.
It is appropriate to discuss space science at this point, because like resource prospecting it is observational. While space science is not fundamental in the sense of addressing basic questions of the structure of matter or of life – as are, for example, molecular biology, particle physics or astrophysics – it is a legitimate and worthwhile field of study, comparable philosophically to the descriptive voyages of exploration which ended around the year 1900. It is also of general public interest, and is certainly a proper subject for any nation’s space program.
The main changes which should be made in the U.S. governmental space science program are two: first, institutional and policy barriers should be swept away so that foreign scientific sensor packages can fly on “U.S.” missions, and so that U.S. scientists will be entirely free to place their sensor packages on spacecraft of other nations, in cooperative programs.
Second, there should be many more scientific space missions, flown much more frequently, and each mission should not be overloaded, as are those of the present day, with far too many experiments. By a series of well-intentioned errors, we have arrived at a regime in which a scientist may have to wait ten to fifteen years before his experiment is flown. When it is flown, it is one of ten or twenty experiments being flown on the same spacecraft. By then the entire ensemble costs a billion dollars, and because of budgetary constraints there is often no backup. With only one shot at the mark, the scientist also has no chance to send a second experiment whose design is improved to zero in on an unexpected first result.
The sensible approach would be to offer many more flights on spacecraft which are turned out on assembly lines. There should be missions launched to each planet of interest every one or two years, depending on launch window opportunities. Information returned from the first spacecraft of a series should arrive in time to affect the design of experiments flown later. Launch costs are not, incidentally, an issue in this strategy. Even present-day launch costs are no more than a small fraction of the total cost of one of today’s multi-experiment missions.
The earliest acquisition of space resources should be the retention and maintenance in Earth orbit of discarded Space Shuttle external tanks. Currently, the tanks are brought to 99% of orbital energy, then are dumped back into the atmosphere, wasted. Each tank weighs considerably more than a maximum Shuttle payload, and its materials, mainly aluminum, are well known. The tanks are usable with almost no alteration as pressure shells for habitable working space (as was Skylab, which was built from an Apollo upper stage). The addition to an external tank of lightweight insulation is an undemanding job and would convert the tank into a long-term storage medium for propellants. Most Shuttle flights are volume-limited rather than weight limited. Therefore, each Shuttle flight could carry to such an orbital tank farm, at little cost, propellants not needed for that flight itself, which could be transferred to the long-term storage tank.
The uses of such stored propellants include maintaining or raising the orbits of space stations and tank farms, fueling scientific spacecraft destined for voyages to other planets, and, especially, fueling shuttle craft operating between relatively low-Earth orbit and orbit about the Moon.
Uses of external tank materials for construction include aluminum fabrication into structural beams, counterweights, gantries, storage warehouses, solar cell array backings, electrical conductors, and specialized pressure shells for space stations. In a cost effective space program, space stations should be fabricable at relatively short notice and at low cost from the ready reserve of well-understood materials available in orbit from stockpiled external tanks. The time scale for initial developmental trials in space of these techniques should be the early 1990s, for example, 1993.
The material needed earliest from the lunar surface is oxygen. It can be used to resupply propellant to the Earth-Moon shuttle craft referred to earlier, and to lunar landers which can dock in orbit with those shuttle craft, as did the descent and ascent modules of the Apollo spacecraft more than thirty years earlier.
For the past ten years there have been experimental studies, partly under NASA sponsorship and partly under private sponsorship, on optimum methods for separating oxygen from the lunar surface soils. There are a number of such techniques at the laboratory proof of concept stage. Funding so far has been tiny, but as billions of dollars of potential savings are at stake, this activity should be strengthened greatly. There is no need to wait for lunar prospecting missions, as oxygen is found universally on the lunar surface. The time scale for setting up a small pilot plant on the Moon, running on solar energy and operated only during the lunar day, should be the late 1990’s, for example, 1997. A plant of that kind would weigh at most a few tons and should be capable of generating many times its own weight in liquid oxygen each year. That plant, and duplicates of it landed later, should be teleoperated from the Earth. The plant could use Shuttle external tanks, insulated both by multi-layer foil insulation and by coverage with lunar soil, for long-term storage of liquid oxygen.
Iron is relatively abundant on the Moon, and sintered iron technology is well understood already for the fabrication of high-precision, strong components primarily for machinery. Iron can be separated to a high level of concentration by the use of magnetic fields, because much of the iron on the Moon is in the form of small particles which have arrived over eons of bombardment by meteors.
Other lunar materials usable early on for construction include the natural lunar glasses, from which strong composites can be made. Also in the lunar soil and necessary for fabricating solar cells is silicon, the second most abundant material on the Moon after oxygen.
Experimental studies under private sponsorship have shown during the past ten years that the lunar surface materials can be separated into their component elements, including metals such as aluminum, titanium and magnesium, by chemical techniques similar but not identical to processes that are used on the Earth. Those studies were carried out by a large aerospace firm. They were based on experimental measurements rather than on paper calculations and indicate that a plant operating on the Moon or in space could process about 100 times its mass each year.
The time scale for the more complex processing methods, such as the full separation of lunar soils into component elements, should be the late 1990’s to the early 2000’s, for example, 1998-2005.
There are a number of locations on the Moon favorable for scientific or industrial activity. Those of special interest include the entire near side, especially easy for communications with and teleoperation from Earth; the lunar farside, for its radio quiet; and the polar regions if they are proven to contain water ice and other frozen volatiles – the elements necessary to sustain life without resupply from the Earth.
In a cost effective program several of these locations may be suitable for the construction of bases equipped at least with life-support habitats, but it may not be cost effective to maintain people in all of those habitats all of the time. For energy-intensive activities, such as the operation of chemical processing plants or of electromagnetic launching machines (mass-drivers) to launch lunar materials to precise points in space, it may be most economical to build two or three identical facilities, spaced 180 or 120 degrees in lunar longitude apart, and operate each one only when it is in full lunar daylight, receiving intense solar power. Maintenance personnel should probably be based at a single location and be equipped to make service calls when needed, as the shutdown of one operating facility and the startup of another follows the slow movement of the lunar sunrise in its 28-day cycle.
Lunar bases can best be constructed of glass-glass composites made out of locally available materials or from Shuttle external tanks from the stockpile in orbit. It does not make sense to fabricate away from the Earth complex lightweight products such as computers, radios, sensors, space suits, vehicles or life-critical precision components such as docking ports. It does make sense to use local materials to build pressure shells for enclosing large volumes, to use lunar soil for cosmic ray and thermal shielding, and to use lunar silicon for large solar cell arrays. Mass-driver machines, referred to earlier, consist mainly of aluminum ring-shaped conductors in a repetitive structure. Because their components are simple and repetitive, mass-drivers are good candidates in the long run for construction on the Moon from locally derived materials.
For the construction of large structures to be used in space (such as steerable radio telescope arrays, large ships of exploration, space workers’ shielded habitats, and power satellites, which will be described in the next section) cost effectiveness follows the same logic as just described. Heavy, simple, repetitive components should be built from materials available from the Moon, while lightweight, complex components should be built on and lifted from the Earth. The choices for location for construction are two: the lunar surface or free space itself. Cost and size should dictate the choice for each particular product. Studies over the past 20 years indicate that for products to be used in space, space itself is the best location for construction. The reasons are that it is cheaper to deliver construction machinery from the Earth to free space than to the Moon; that it should be very cheap to deliver raw materials from the Moon to free space by mass-driver; that processing plants and fabrication machinery in free space can run on full-time solar power rather than being cut off by the lunar night; and that structures which would be much too large, heavy and fragile to lift off the Moon can be assembled easily in the zero gravity of space.
The current level of understanding of mass-driver machines for launching materials from the Moon is indicated by the fact that a mass-driver of the same diameter (coil size) as required for a high-capacity lunar launcher was designed using a computer-aided design program in 1981-82. It was built and tested in 1983 and obeyed its design program within one percent. Its acceleration was 1,300 gravities (g’s).
The concept of self-replicating machines was addressed particularly by John von Neumann. It ties in closely with leading-edge thinking in the industrial field, and is therefore one of many points where developments carried out for the alternative U.S. space program could benefit U.S. industry in a real and clearcut way. Currently, the practical application of self-replicating machinery has been exploited to its most advanced state in Japan.
In its early application in space, self-replication is likely to be employed for three general purpose facilities: mass-drivers, processing plants to generate industrial feed-stock from lunar materials, and general purpose, teleoperated “Job shops” capable of building more mass-drivers, processing plants and job shops. As noted earlier, it is not cost effective to carry self-replication to the 100% level. Many of the components of all three types of facilities are complex but light in weight and therefore inexpensive to lift from the Earth. Self-replication should be confined to the heavy, repetitive components of the production facilities.
The logic of self-replication for production facilities in space is that a “seed” facility consisting of a mass-driver, processing plant and job shop on the Moon and a processing plant and job shop in space could replicate itself in the sequence 1, 2, 4, 8, 16… A series of seven doublings, starting with an initial set of facilities built on the Earth and weighing less than fifty tons, would lead to the capability of processing about 100,000 tons of lunar material per year into completed structures in space. Because of its urgency in terms of potential payback and its derivation from already existing scientific work and profitable industrial practice, a realistic time scale for the earliest pilot plants which are partially self-replicating industrial systems on the Moon and in space is ten years – the year 2000. As doubling times would be as short as two months, the years 2002-2005 could yield a full-scale system processing 100,000 tons per year or more.
Large scientific telescopic arrays, radio and optical, large ships of exploration, and safe, comfortable habitats for construction workers and their families are among the products that make sense for a construction facility in space based on lunar materials.
Missions more distant in the future than programs which return direct economic benefit to the Earth are those of manned space science. The ultimate space science missions should be manned and multinational, and should be flown by large ships built in space and fueled by propellants from the Moon. The missions should be multi-year, each to orbit around a particular planet. Because of their duration, these missions should provide to their crews normal Earth gravity by rotation, and should be shielded to a level comparable to that of the Earth’s surface. The ships should be large enough to accommodate scientists, crew members and their families. Total personnel should be sufficiently numerous to include many scientific specialties and medical doctors working with fully adequate facilities.
From these ships probes can descend into the atmospheres of planets such as Venus and Jupiter. Unmanned and manned landers can descend for exploratory visits to the surfaces of Mars and the Jovian moons. The ships themselves can dock at the Martian moons, Phobos and Deimos. Refueling can take place from the moons both of Mars and of Jupiter.
These voyages will be closely analogous to that of Darwin’s “Beagle”. The size and technology of the ships will require that they follow logically and chronologically the prior development of space manufacturing from lunar resources, and the development of orbiting laser arrays which can beam intense concentrated power to the ships, derived from the permanent solar energy of free space near the Earth-Moon system. Their earliest realization could be in the decade 2010 to 2020.
There are potentially much larger products which can generate large revenues and which address the two greatest issues which transcend politics and challenge the entire world at this time. Those issues are how to obtain the energy necessary to power the world economy of the next decades and how to avoid further buildup of the greenhouse effect caused by the burning of fossil fuels.
Among many solutions which have been studied, only one, solar power satellites, has the ability to generate all the power needed, does not employ radioactivity in any form, and does not inject pollutants into the Earth’s atmosphere. The basic idea, invented by Dr. Peter Glaser in the United States in the late 1960’s, is to collect solar energy where it is available nearly full-time, in geostationary orbit. There it is converted to electricity by conventional solar-cell arrays, and the electricity is converted to low-density radio waves which are directed to a collector array located in a fenced-off area on the Earth. Using present technology, the collector converts the radio waves to ordinary electricity at an efficiency of more than 90%.
Studies carried out by NASA, the DOE and private foundations show that each power satellite would have a mass of about 100,000 tons, and could supply to the electrical grid on Earth an amount of power equal to the continuous output of ten nuclear power plants. Those studies also indicate that at least 99% of the mass of a solar power satellite, and probably a still higher percentage, could be obtained from lunar surface materials. That is crucial, because those studies also indicate that solar power satellites can undersell coal or nuclear power plants only if the satellites are built from lunar materials.
Much work needs to be done before solar power satellites are a working reality, but that work is straightforward engineering, requiring no scientific breakthroughs and no materials or components that do not already exist.
The potential market, if solar power satellites are carried through engineering development and are built in quantity, is about $250 billion per year simply to satisfy present needs for new and replacement central station generators. Full success, from a global environmental viewpoint, would consist in a shift over the next several decades from fossil fuels to fuels synthesized to store cheap electric power in their molecular bonds. Those fuels are already well understood and are compounds such as propane. With full success in the world of 2040, half a century away, the market for solar power satellites would be six trillion dollars per year.
Japan and nations in Europe are already pursuing research into solar power satellites. With so large a potential market, we in the United States must move aggressively to regain a leading role in its development. Above all, we who live in the biosphere cannot afford to let it die.