The terrain around Barstow, California is a sun-blasted desert. It is a country of low hills, where scrubby brush manages somehow to stay alive, along v nth an occasional Joshua tree. It is monotonous country, hot, worthless for farming; among its most important uses is to provide much of the 300 miles that separate Los Angeles from Las Vegas. It is also the location of the Goldstone tracking station of the Jet Propulsion Laboratory.

In a recent series of tests there, one of Goldstone’s tracking antennas was aimed not at a spacecraft millions of miles away but at a low hill a mile away. On that hill was a 100-foot tower with an array of receiver panels 25 feet high. Below the tower was a bank of seventeen lights, each about the size of an auto headlight.

Red lights flashed in the area of the antenna, 85 feet in diameter, as it swung upward upon its pedestal until it was pointing directly at the tower. Loudspeakers warned that a power beam was about to be turned on. Richard M. Dickinson, electrical engineer and project manager for the experiment, switched on the power. With the antenna zeroed in on the receiver panel a mile away, Dickinson ordered the power brought up slowly in increments of 25 kilowatts. There was no sound and at first, no glow from the lights.

Then at 50 kilowatts, the lights showed dimly under the glare of the midday desert sun. As the power was brought up to the peak of 400 kilowatts the lights brightened to full intensity, flicked on and off as the antenna was shifted slightly from left to right, or up and down.

This experiment represented the first high-power field trials of the use of a microwave beam to transmit electric power, a system which in years to come may provide a major new energy source—power from space. Such a system will provide the means by which electricity, generated in space from the intense sunlight available there, can be efficiently sent down to Earth.

This concept, power from space, represents one of the most important and newest ideas to appear in recent years. Arthur C. Clarke once noted that there are two types of inventions: There are achievements such as the airplane and the moon rocket, which had been anticipated centuries in advance. There are other inventions, such as radar and the laser, which had a considerable aura of surprise about them and which were predicted (if at all) only slightly in advance of their realization. Many of the latter class of inventions have grown out of progress in physics or electronics, and the transmission of power by microwave is definitely one of them. Indeed, Clarke, in his 1968 prediction of the future out there, The Promise of Space, completely failed to note the possibility of power sent down.

Microwave transmission of power rests upon two recent advances in electronics: a means for efficiently generating large quantities of microwave energy, and a means of directly converting this energy back to direct-current electricity.

The microwave generator is a somewhat remote descendant of the tubes first used in World War II to generate radar beams. These tubes were wasteful of power, short-lived, and prone to heat up when high power levels were demanded. But the basic principle of operation has not changed over the years. Electrons are made to spiral around a magnet, to produce the microwave energy.

Over the years, these tubes were improved. They are now mass-produced in the hundreds of thousands per year for use in microwave ovens. A major advance came with the introduction of a new material for the magnets, samarium-cobalt. This not only permitted the magnets to be reduced in size by 90 percent; it also meant great improvements in efficiency, since the electrons would spiral in farther and give up more energy before striking the magnet. Another advance came through the use of platinum to supply electrons through the process called secondary emission. In this process, a small amount of electricity is made to stimulate a flow of electrons from the thin layer of platinum.

The result of these advances is a device known as the Amplitron. A single Amplitron, only a few inches in diameter, can produce up to 5000 watts of microwave power at efficiencies approaching 90 percent. If a given amount of energy is fed into a bank of Amplitrons (as direct-current electricity) only 10 percent of that energy will be rejected as waste heat. The rest will be converted into microwaves.

Microwaves are easy to form into a beam and can travel long distances with very little absorption in the atmosphere. They readily penetrate even the thickest clouds and rain to arrive at the receiving antenna, or rectenna. The heart of the rectenna is a system of small dipole antennas, similar to the rabbit ears of a TV set. Each is connected to a device called a Schottky-barrier diode. Microwaves, collected by the dipoles, are converted to direct-current electricity within the Schottky diodes with an efficiency of over 80 percent.

In the Goldstone experiments, the seventeen receiver panels on the tower mounted 4590 of these dipoles. Each one was T-shaped, and about four inches long, with the T-arms held vertically to discourage birds from roosting. The receiver panels were not designed to intercept the entire beam, but rather covered only about 10 percent of the beam area. But they demonstrated a maximum efficiency of 82.5 percent in collecting their portion of the microwave beam and converting it to electricity. Other experiments showed overall efficiencies for the total system of 54 percent, so that of the power initially fed into the Amplitrons, 54 percent was subsequently recovered from the rectenna. This is not the limit, however. Engineers such as Richard Dickinson talk confidently of achieving an overall efficiency of 60 or even 70 percent.

In an actual application, the rectenna would be several miles in diameter. The small antenna elements and Schottky diodes would be needed in the millions. They would be mounted on panels which would be set upright at a convenient angle. There would be no need for precision adjustment of these panels or of close tolerances and high accuracy in their assembly. The rectenna could be placed wherever land is cheap: in desert country, in rocky or hilly terrain, or even in the ocean. A floating rectenna out at sea, its panels bobbing up and down, would be a distinct possibility. The electricity garnered from the large number of individual panels would be fed through solid-state devices known as inverters to convert it from direct current to alternating current at the usual 60-cycle frequency. It would then be fed directly into the nation’s power grid, being transmitted up to hundreds of miles by standard overhead power lines or by an undersea cable.

The earliest application of this technology could come about in as little as ten years. This would involve the power relay satellite, proposed by Krafft Ehricke of Rockwell International, as a means of transmitting large blocks of electric power across continents.

Ehricke’s idea is based upon his observation that there is no practical way to transport electricity across oceans. Nevertheless, there are good reasons to do so. It would be possible to put nuclear power complexes in Greenland or to build nuclear power complexes on a remote island. The island of New Guinea has swift-flowing rivers which could develop enough hydroelectric power to run much of Australia—if the power could be transported across the Arafura Sea. Or the intense solar energy which falls wastefully upon tropical deserts—the Sahara, the Kalahari in Southwest Africa, the Atacama in Peru and northern Chile—could be used as a source of electricity which otherwise would find only the most limited local market. It could even become practical to generate electricity in tropic oceans, using the large difference in temperature between the sun-heated surface waters and the cold waters of the depths.

Even without these new means of generating energy, a method of transporting electricity globally would have important benefits. The most efficient way to generate electricity is to run the generating plant at full load, continuously. But electricity is not used at a constant rate. There is a peak in usage in the afternoon and early evening, and the rate of use drops off to a very low level in the early hours of the morning. So power companies must build the relatively expensive “peaking” plants which are shut on or off to cope with the changing loads. If excess power could be sent to Europe or Japan, or their nighttime power sent here, then most plants could be built as the less expensive “base-load plants,” which’ do indeed run continuously.

The electricity produced at the earth’s surface, by whatever method, then would be fed into the transmitter array, six miles square. This array would be subdivided into 1,000,000 transmitter modules each 30 feet square. Within the overall array, which would be nearly as large as Manhattan, the generated power (as much as 10 million kilowatts, which is enough to light Manhattan) would be subdivided by a network of substations and transmission lines to feed about ten kilowatts into each module. This power would feed an Amplitron, and the resulting microwave energy would be channeled into arrays of small antennas spaced every foot or so.

Each Amplitron would have attached to it a phaseshifter. The phaseshifters would control the microwave energy from the modules to ensure that the energy would be transmitted in the form of coherent waves. Without the phaseshifters, the energy would be incoherent: each module would radiate independently, and the energy would spread out uselessly. The modules would behave like a crowd of people milling around. The phaseshifters would keep the microwaves from each module properly adjusted, so that the modules would be like an army marching in step. This would produce a powerful well-focused beam, tighter than the beam of any searchlight. This beam would rise into space and reflect off the power relay satellite, which would return it to the rectenna on the ground.

The power relay satellite would be in orbit, 22,323 miles up. There it would take exactly twenty-four hours for each revolution, thus remaining always above one spot on the earth’s surface, in the well-known synchronous orbit used by communications satellites. It would be a mile in diameter with a mass of 500 tons. It would serve simply as a reflector. The best reflector would not be a mirror but a mesh of thin wire, similar to ordinary window screening. This reflector would have to be maintained almost perfectly flat. It could deviate from its calculated curve by no more than a twentieth of an inch across the entire mile of its diameter. This could be done by building the reflector out of some one thousand subreflectors, each accurately flat, and each capable of being shifted in position so as to give the overall reflector its proper shape.

The satellite would be a gossamer, incredibly fragile structure, a spider web in space. It could be assembled only in orbit 200 or 300 miles up. It would then be maneuvered to its higher orbit using ion thrusters. It would need weeks or months for that trip, accelerating at very slow rates to avoid damaging the structure. Once in orbit the satellite would be a rather simple system needing to be visited only rarely, if ever.

The total system of transmitter array, satellite, and rectenna would cost some $15 billion. During a thirty-year period, the usual planning time of power companies, it would transfer so much electricity that the cost to the users could be as low as seven-tenths of a cent per kilowatt-hour. It would then be no more costly than a system of overhead transmission lines a few thousand miles long—lines that could never be built across the ocean.

And it could be built within the next decade or so. The satellite would be quite large in comparison to what has been orbited, and the need for assembly in space would certainly be a new thing. But it would be only four times the weight of the Apollo lunar systems carried to orbit on the way to the moon.

The power relay satellite does not offer the promise of opening up fundamentally new major energy resources. It may permit more effective use of our existing power plants or permit the use of energy resources which until now have not been regarded as worth developing. But it would still be basically limited by the constraints which restrict our energy future. These limitations can be overcome by generating the electricity to be transmitted not on the surface of the earth, but in space.

It is nearly a century now since the first practical solar-driven engine was demonstrated at the Paris Exposition. Since then, and particularly in recent years, there has been no shortage of inventors to focus sunlight upon blackened pipes, boiling water to run a turbine or other engine. But the field of solar energy still awaits its Edison. No one yet has demonstrated the combination of low cost and high performance which would make solar power a common thing. This is particularly true for solar generation of electricity. While sunlight is free, the machinery needed to use it is not, and the equipment must necessarily spread over a large area. No solar electricity plant will ever be built as a small, compact structure with its attendant efficiencies in operation and maintenance.

What is worse, the sun does not always shine when it is wanted. In the sunniest parts of Arizona, the sun goes down long before the peak hours of electricity use are over. In the winter months not only are there fewer hours of sunlight, but the sun stays low in the sky.

In space all these problems vanish. There is at least six times as much insolation (solar influx per square foot, per year) as in the sunniest desert. There are no day-night cycles in space and no low angle of the sun in the sky. A solar collector can always be aimed directly at the sun. There are no clouds or haze. In space solar energy is available at its most powerful: constant, unfiltered, and at full strength. The problem is to use it.

The first serious proposal for a solar power satellite was made by Peter Glaser, a vice-president at Arthur D. Little, Inc., a research firm in Cambridge, Massachusetts. The original description was in an article (“Power From the Sun: Its Future”) in Science late in 1968. With support from NASA, he teamed up with William C. Brown of the Raytheon Company, inventor of the Amplitron, and with a group of engineers from Grumman Aerospace Corporation. The group of specialists proceeded to study the power-satellite concept in considerable detail and came up with their conclusion: it can be built, but not soon.

Glaser’s powersat design is like many other satellites which have been launched because it uses solar cells to generate electricity—and there the resemblance ends. It consists of two large panels, each over three miles on a side, with the entire powersat seven miles long and weighing 25 million pounds. This is twenty-five times as much as the power relay satellite. The panels are not completely covered with solar cells. Instead, mirrors are used to concentrate the sunlight so that only half the areas of the panels must be covered.

The panels always face the sun. Mounted between them, free to turn to always face the earth, is the transmitting antenna, 3000 feet in diameter. Within this antenna are the Amplitrons. In the vacuum of space they need no glass tubes to enclose them. The microwave energy produced by the Amplitrons passes down hollow aluminum tubes or waveguides, escaping through slots in the direction of the earth. Phaseshifters serve to produce a tight well-focused beam.

An important result of the work of Glaser and his associates was the devising of a way to prevent the power beam from wandering off the rectenna. Their solution calls for a small portion of the microwave energy to be reflected back up to the satellite as a pilot beam, thus providing a reference signal for controlling the phaseshifters. If the power beam were to wander off, this pilot signal would be lost and the phaseshifters would fail to keep the beam properly focused. The beam would spread out, dissipating its energy harmlessly over the entire earth and surrounding space. The spread-out beam would then be no more harmful than the signal from a radio station and would be as weak.

The power beam would be deliberately designed to spread out slightly on its way to the ground, to meet environmental restrictions on allowable power per square foot. Radiation leaking from a microwave oven is pulsed and can be dangerous, but power-beam radiation would be continuous, which is safer. Animals wandering into the beam, or birds flying through, would find their bodies warming up slightly. They would be in no danger of being cooked. Even in the heart of the beam, the microwaves would have less intensity than ordinary sunlight. Airplanes would find that the microwaves would bounce off their aluminum skins.

The U.S. standard for exposure to microwaves is ten watts per square foot. In the center of the rectenna there would be ten times this limit, but at the outer edges only one-tenth the limit. The main safety feature would be a chain-link fence to prevent people from wandering into the rectenna area. Microwaves are not a penetrating, ionizing form of radiation, as are X rays or the radiation from radioactive substances. They do not cause cancer or genetic mishaps, but merely warm the body.

Inevitably, some microwave energy will leak from the transmitting antenna. The leakage will be at levels far below even the strictest medical limits but will be quite sufficient to cause radio interference. It will be necessary to assign certain radio frequencies for the purpose of power transmission only. The ten-centimeter wavelength is particularly desirable for this, since it is associated with particularly high efficiencies of the Amplitron and the Schottky diode and penetrates the atmosphere well. Of course, any time there is a new allocation of radio frequencies, some people are upset. Radio astronomers will be particularly unhappy. But a space program which can build powersats can also build very large new radio telescopes deep in space and shield them from radio interference.

Unfortunately, we are no more ready to build Glaser’s power satellite today than we would have been ready to build a Boeing 747 in 1935. One of the major problems lies in the solar cells. Many spacecraft, such as Skylab, have displayed impressive arrays of solar cells. But all these arrays have been assembled by hand from small individual cells, each laboriously cut and polished. If a powersat were built using such cells, its electricity would cost over a hundred times as much as its customers would be willing to pay.

The search for low-cost solar cells is under way, but it’s a frustrating business. The best material is pure silicon—easily produced—but it must be available in the form of single crystals. At present, such crystals are grown slowly from molten silicon, then cut and polished by hand. A method is needed for growing the crystals rapidly in a thin ribbon or strip. Tyco Laboratories, a small firm outside Boston, has demonstrated just such a method in which ribbons of silicon are shaped with the aid of a die. Unfortunately, the silicon is so reactive that it dissolves the die. Then the ribbons which are grown are not particularly useful for power generation, because they are contaminated by material from the die.

It is quite possible that there will be breakthroughs leading to cheap solar cells, hundreds of times less costly than at present. The breakthrough may come about quite suddenly, transforming the outlook for solar cells overnight. But for now, among solar-cell scientists, there is the frustration which P. A. Berman of Caltech has expressed: “It seems almost in- conceivable that such a simple thing as a solar array upon which cells are mounted in some simple, economical fashion, having no moving parts and using no exotic materials, cannot be made for a few dollars a square meter, rather than the thousands of dollars per square meter experienced in the space program.”

Power from space need not rely upon such breakthroughs. It is quite possible to generate electricity by the old Paris Exposition method of focusing sunlight onto a boiler to run a turbine. Gordon R. Woodcock of the Boeing Company has pursued this approach. His powersat design uses arrays of thin plastic film, coated with aluminum, to reflect sunlight onto a hot spot which is heated to several thousand degrees. Helium gas flows through the hot spot and drives turbines. Each turbine is hooked to a generator and the electricity produced is transmitted the same way as in Glaser’s arrangement.

Woodcock’s approach has the advantage of relying on the well-developed technologies of turbines, generators, and hot gases. In fact, a very similar arrangement (though without the solar mirrors) has been installed in a power plant in Oberhausen, West Germany. When such powersats are built they will be easily seen, especially on a dark night. Near midnight they will gleam more brightly than any star, as sunlight momentarily reflects from individual panels of plastic film.

It is likely to be quite a while before such powersats are in orbit, to ornament the night sky. If we wanted to build them, we would need to press our technology to the limit. There would be the classic conflict of requirements: to keep costs to the lowest level yet to use the most advanced designs. We would need structures far lighter than any yet built for use in space. The turbines would have to be guaranteed for thirty years of operation in space, yet run at temperatures hotter than any yet used for producing electricity. The nation would need to build new rockets to ferry people and equipment at one-tenth the cost of the most advanced rockets now being built. And with all these required advances, all the risky new designs, the electricity from these powersats would barely manage to be cheaper than the most expensive electricity available today. With so many new inventions needed, it is almost certain that some things would go wrong; there would be cost overruns, and the program could not pay its way.

The required launch vehicles would tax the rocket-builder’s art to the limit. From the outside, they would look like enormous Mercury capsules, 167 feet in diameter and 192 feet tall. At liftoff, a launch vehicle would weigh 12,000 tons and would rise upon 20 engines, representing the thrust of 4 Saturn V moon rockets. The engines would gulp fuel (kerosene and liquid oxygen) at the rate of sixty tons per second, lifting a vehicle the weight of a naval cruiser. The power of ten such cruisers would be needed, to run the pumps of such a rocket.

To keep costs to the lowest level, the rocket would be fully reusable. All propellant tanks would be enclosed within the framework of the huge structure. The broad flat surface at the base of the rocket would give protection while re-entering the atmosphere. The rocket engines set into this base would need special protection. This would be provided by carrying water in the base. During re-entry the water would boil into steam which would flow to the outside of the base and shield the engines. There would also be a number of smaller rocket engines to fire just before touchdown and provide a safe landing.

Such a launch vehicle would resemble a ship in more than its size and weight. It would take off and land on water—on an artificial lagoon at Cape Canaveral three miles across. A creature of water, air, and space, it would never touch the land. After each flight it would be towed to a berth or drydock to be made ready for the next flight. It would carry 500 tons of cargo to orbit. Yet it would be necessary to launch such a rocket nearly every day to build one powersat a year.

One of the earliest projects would be to build in orbit, 250 miles up, an assembly facility for powersats. It would have its own electric power supply, living quarters for 100 people, communications, and other major functions. Its framework would support a system of tunnels, propellant lines, docking ports for spacecraft, and an assembly area. There would be a propellant depot and room for the loads of equipment as they arrive.

The assembly facility would serve primarily to build the thermal engine systems for the powersats. The first powersat parts to arrive would be the panels for the hot spots, or thermal absorbers. For reasons of safety there would be a manned tug to meet each load at some distance, with a pilot to guide the load to a docking port at the assembly facility. Other tugs would disconnect panels from these loads and transfer them to the appropriate locations in the assembly frame, clamping them in place. The panels would also be attached to each other. Additional modules would be attached, with elements of the turbines, generators, and heat exchangers. Collapsible frames would be mounted and extended to form a structure for the radiators. When the radiator panels were attached, the resulting relatively compact systems within the assembly frames would represent over half the total weight. Like the hull of a clipper ship, this system would be ready for its superstructure. The solar reflectors, resembling the masts and sails of a clipper, would be built up. Workers in specialized pods would attach collapsed support arms, extending them to form the structure supporting the reflectors. The reflector panels would be added last of all.

The first power produced would serve to ferry the powersat to geosynchronous orbit. Propellant tanks would be attached as well as electric-propulsion engines. In these engines the electricity would operate a number of carbon arcs to provide high temperatures. Hydrogen gas would flow through the arcs and be heated, providing several thousand pounds of thrust. The powersat would slowly accelerate, reaching its final orbit several weeks later. A few of the workers would stay aboard to make sure that it was functioning normally. They would then return to the assembly facility in a small space tug carried along.

There is no doubt that in time these things will all be possible. It will even be economically feasible to do them. Right now, nothing of the sort is at hand. At the current rate of progress in space technology, they may become feasible sometime in the next century. They will represent a twenty-first-century solution to a twentieth-century problem: the need for new energy sources.

Fabricating powersats on the ground, then sending them up the earth’s 4000-mile-deep
gravity hole to space, one by one, is a needlessly expensive way of solving the energy problem because the cost of overcoming gravity is like a tariff which nature levies upon space flight.

The powersat components could be built in a space colony rather than on Earth. The colony would require a single large effort in order to be built, using resources obtained—but only in part—from Earth. Using resources from the moon, the colony can turn out powersats in space—as many as would be required. Instead of paying the “gravity tax” every time a new powersat is built, this “tax” need be paid only once—at the outset of the colonization program.

The colonization effort would begin by building in Earth orbit a “construction shack” resembling the orbital assembly facility. In addition to possessing complete assembly facilities, the construction shack would also have ore-processing facilities to extract metals from lunar rock. It would also have a large power plant delivering energy for ore-processing.

While the construction shack was being built to hold 2000 workers, a smaller crew of 100 or 200 would establish a lunar mining base. The major facilities would be a nuclear power plant, to allow operations through the lunar night, and a mass-driver. The mass-driver would accelerate payloads of lunar rock and soil, launching them to space, where they would be collected by a mass-catcher.

The construction shack would be moved to the site of the colony and the lunar work crew would begin sending up a stream of material. This material would be transported to the construction shack to be processed into aluminum, magnesium, titanium, iron, glass, and oxygen—all of which are abundant in lunar rocks. The metals in turn would serve to build the colony and to build powersats as well. A few powersat components, such as turbine blades and Amplitrons, might still be brought from Earth. But over 90 percent of the powersats’ mass might be furnished by the metals obtained at the colony. Moreover, it would be easy to propel the powersats from the colony site into geosynchronous orbit. This maneuver would call for a velocity change of less than half that needed to transport a powersat from low Earth orbit to geosynchronous orbit.

There would be other benefits as well. Building powersats at a space colony means that it would not be necessary to strain the limits of technology. Instead of requiring the lightest possible structures and the hottest turbine operating temperatures, it would be perfectly feasible to use the types of designs which are available at present. When Gerry O’Neill was discussing this point with Gordon Woodcock, he asked what kind of power plant weights Woodcock was looking for. Woodcock replied, “Five kilograms per kilowatt. We can reach that by 1990.” O’Neill replied, “With my system, all we need is ten kilograms per kilowatt. When could we have that?” Woodcock answered, “Why, we could have that today!”