Electricity from Space: The 1970s DOE/NASA Solar Power Satellite Studies

Image 1 (see callout in text). Image credit: NASA.
Of all the many spaceflight concepts NASA has studied with any degree of seriousness, probably the most enormous was the Solar Power Satellite (SPS) fleet. Czech-born physicist/engineer Peter Glaser outlined the concept in a brief article in the esteemed journal Science in November 1968, and was awarded a patent for his invention on Christmas Day 1973.

Glaser had noticed that a satellite in geosynchronous Earth orbit (GEO), 35,786 kilometers above the equator, would pass through Earth's shadow for only a few minutes each year. It was well known that a satellite in equatorial GEO moves at the same speed the Earth rotates at the equator (1609 kilometers per hour). This means that, for people on Earth's surface, the satellite appears to hang motionless over one spot on the equator. Glaser also understood that electricity did not have to travel through wires; it could be beamed from a transmitter to a receiver.

Glaser mixed these three ingredients and came up with a satellite in equatorial GEO that would use solar cells to convert sunlight into electricity, convert the electricity into microwaves, and beam the microwaves at a receiving antenna (rectenna) on Earth. The rectenna would turn the microwaves back into electricity, then wires would carry it to the electric utility grid.

The great advantage an SPS enjoyed over a solar array on Earth's surface was, as mentioned, that it would spend almost no time in Earth's shadow. Earth's rotation meant that an Earth-surface solar array could make electricity at most about half the time. The rest of the time it would sit dormant under the night sky.

NASA and its contractors displayed low-level interest in the SPS concept as early as 1972. Early work took place at the Jet Propulsion Laboratory and NASA Lewis Research Center (now NASA Glenn), as well as at Arthur D. Little, a Cambridge, Massachusetts-based engineering firm of which Glaser was a Vice President. The level of effort increased in 1973, after the Organization of Petroleum Exporting Countries imposed an oil embargo to punish the U.S. and other industrialized nations for their support of Israel in the 1973 Yom Kippur War. By 1976, NASA Johnson Space Center in Houston, Texas, and NASA Marshall Space Flight Center in Huntsville, Alabama, led SPS studies within the space agency.

In June 1975, NASA and the Energy Research and Development Administration (ERDA) signed a Memorandum of Understanding calling for joint SPS research. ERDA began to plan an SPS study with NASA at the beginning of Federal Fiscal Year 1977 (October 1976), in the waning days of Gerald Ford's caretaker Presidency. The three-phase study began in July 1977. Total cost of the joint SPS studies, which were meant to last for three years, was $15.6 million, of which the DOE paid 60% of the total.

Energy shortages coupled with the Three-Mile Island nuclear accident (March 1979), made the mid-to-late 1970s a fertile environment for alternative energy research. A month after the ERDA/NASA studies began, President James Carter made ERDA a part of the new Department of Energy (DOE). Creation of the DOE was part of a policy package aimed at U.S. energy independence and "clean energy."

After Apollo, NASA had, despite its best efforts, found itself without a clearly defined mission for its piloted program other than development of the Space Shuttle. SPS supporters in the aerospace community saw in the concept an irresistible opportunity for NASA to contribute to the solution of a pressing national problem.

Development, deployment, and operation of SPSs would confront NASA with engineering problems far beyond any it had tackled before. If an SPS was to contribute a meaningful amount of electricity to the interlinked U.S. utility grids — and, by DOE's reckoning, "meaningful" meant gigawatts — then it would have to be colossal by normal aerospace engineering standards. The SPS silhouetted against the Sun in the NASA artwork at the top of this post (Image 1) is typical: it would have measured 10.5 kilometers long by 5.2 kilometers wide and had a mass of 50,000 tons.

Paired with a rectenna a couple of kilometers across, such an SPS would contribute five gigawatts to the U.S. electricity supply. DOE estimated that 60 such satellites with a total generating capacity of 300 gigawatts could contribute meaningfully to satisfying projected U.S. electricity demand in the 2000-2030 period.

Image 2. Image credit: Boeing/NASA.
There was, of course, no way that NASA could launch such huge satellites intact, or even in a few modular parts. It would need to construct the SPS fleet in space, most likely in GEO, from many parts. This called for an armada of highly capable space transport vehicles and an army of astronauts and automated assembly machines.

The red, white, and blue "Space Freighter" pictured in the Boeing painting above (Image 2) was, as its name implies, meant to serve as the main cargo launcher for SPS construction. Fully reusable to cut costs, it would have comprised at launch an automated, delta-winged Booster with a piloted, delta-winged Orbiter on its nose. After separating from the Orbiter, the Booster would have either landed downrange (if it were launched from a site in California, Arizona, New Mexico, or western Texas) or would have deployed turbofan engines and flown back to its launch site.

Image 3. Image credit: NASA.
Had it been built, the Space Freighter would have utterly outclassed all other launchers. Its Orbiter would have delivered up to 420 metric tons of cargo to a staging base in low-Earth orbit (LEO). For comparison, the largest single-launch U.S. Earth-orbital payload, the Skylab Orbital Workshop, weighed 77 metric tons. Skylab was launched on a two-stage Saturn V rocket on 14 May 1973.

Engineers speak of "gross liftoff weight" (GLOW) when they describe large launchers. The Space Shuttle had a GLOW of about 2040 metric tons and the three-stage Apollo Saturn V, about 3000 metric tons. Estimated GLOW for the Space Freighter was a whopping 11,000 metric tons.

Alert readers will notice discrepancies in the paintings that illustrate this post. These occur because the images are based on design concepts developed by different engineers in different phases of the multi-year SPS study. The delta-winged Boeing Space Freighter design, for example, is different from the NASA Space Freighter design depicted in the illustration above (Image 3).

The NASA Space Freighter has a Booster with some resemblance to a Saturn V S-IC stage; both the Booster and the Orbiter have skinny main wings and forward canard fins. The Orbiter payload bay is located near its front; not, as in the Boeing design, at mid-fuselage. Despite these differences, the NASA Space Freighter would have had the same capabilities as the Boeing Space Freighter.

Image 4. Image credit: NASA.
The NASA painting above (Image 4) depicts a hexagonal LEO staging base with a central "control tower." Access tubes link the control tower to docking modules at the hexagon's six vertices. Between the access tubes are color-coded triangular "marshaling yards" with socket-like bays for storing standardized NASA Space Freighter cargo containers.

The staging base control tower has mounted on its roof a "space crane" descended from the much smaller Space Shuttle Remote Manipulator System, which was under development at the time DOE and NASA conducted their joint SPS study. The control tower space crane is positioning a cargo container so that an automated chemical-propulsion Orbital Transfer Vehicle (OTV) can dock with it. After docking and space crane release, the OTV would automatically transport the container to a construction base in GEO.

Another, smaller space crane rides a track around the edge of the hexagon. It is shown unloading a cargo container from the newly docked Space Freighter Orbiter.

The painting includes many other details. It shows, for example, what appears to be a conventional Space Shuttle Orbiter approaching the staging base in the background. Rockwell, prime contractor for the Space Shuttle, proposed that second-generation Space Shuttle Orbiters serve as dedicated crew transports for the SPS program. The company envisioned that replacing the Orbiter payload bay with a pressurized crew module would enable it to transport up to 75 astronauts at a time.

Next to the crew transport is a cluster of cylindrical modules for housing the staging base crew and astronauts in transit between Earth and GEO. A piloted OTV for transporting astronauts to and from the GEO SPS work-site — identical to the automated OTV, except for the presence of a pressurized crew module — is shown docked with the LEO staging base at lower right.

Image 5. Image credit: NASA.
Image 6. Image credit: NASA.
In the SPS study, NASA sought to balance automation and astronauts. Automation was, its engineers noted, good for repetitive actions such as fabricating the tens of kilometers of trusses needed to support SPS solar cell blankets.

The basic "beambuilder" depicted in the upper image above (Image 5) would turn tight rolls of thin aluminum sheeting into sturdy single trusses. The more complex multiple beambuilder system in the lower image (Image 6) would combine and link together single trusses to make the major structural members of the satellite.

Astronauts would supervise and maintain the beambuilder robots and join together the trusses they fabricated. Automated OTVs would deliver thousands of aluminum rolls to the GEO work-site, which the astronauts would then load into the beambuilders.

DOE and NASA expected to added two SPSs to the "fleet" in GEO each year starting in 2000. Each SPS would need about 200 Space Freighter launches and hundreds of OTV transfers between the LEO staging base and GEO. Propellants for the OTVs, as well as 50 metric tons of orbit trim propellants for each SPS per year, would demand even more Space Freighter launches.

Image 7. Image credit: NASA.
Despite extensive reliance on automation, the 30-year SPS project would require the presence of nearly 1000 astronauts in space at all times. Most would be based in GEO (Image 7).

In addition to construction workers, personnel needed in space would include physicians, administrators, OTV pilots, life support engineers, general maintenance workers ("janitors"), cooks, space suit tailors, and computer technicians. Personnel needed on the ground — at the launch/landing site, at the rectennas, and at widely scattered factories for manufacturing SPS parts, OTVs, spares, foodstuffs, and propellants — would outnumber astronauts by at least 10 to 1, NASA and DOE estimated. Building and operating the SPSs could become a major new U.S. industry.

Image 8. Image credit: NASA.
As beambuilders and astronauts completed trusswork sections, automated OTVs would begin to deliver rolls of solar cell "blankets" to the SPS work-site. The NASA painting above (Image 8) shows in the background an automated OTV laden with bluish rolls of solar cell blankets (upper right).

Meanwhile, an automated system feeds blanket sections to a piloted "cherry picker" at the end of a small space crane. The cherry picker's "pilot" — who wears only shirt-sleeves in his pressurized cab — uses manipulator arms to link one end of a solar cell blanket to a truss.

More than 50 square kilometers of solar cell blankets would be spread over the trusswork of each SPS in this way. The end result of this intensive human and machine labor is depicted in idealized form immediately below (Image 9).

Image 9. Image credit: NASA.
Image 10. Image credit: NASA.
The lower painting above (Image 10) shows Glaser's invention at work. The intense sunlight of space strikes solar cells, which are hidden from view (the image does, however, provide a good look at the backside of a completed SPS). Millions of silicon or gallium arsenide cells efficiently convert the sunlight into electricity.

The kilometer-wide steerable microwave transmission antenna at the lower end of the SPS converts the electricity into microwaves and focuses the microwave beam on a rectenna on Earth, nearly 36,000 kilometers away. The beam appears in the illustration as a ghostly cone; in reality, the microwaves would be invisible.

DOE and NASA envisioned building the 60 rectennas (Image 11) required for the SPS system from coast to coast along the 35° latitude line. Cities on or near that line include Bakersfield, California; Flagstaff, Arizona; Albuquerque, New Mexico; Amarillo, Texas; Oklahoma City, Oklahoma; Little Rock, Arkansas; Memphis and Chattanooga, Tennessee; and Charlotte, North Carolina. If one flew between these cities, one would overfly rectennas on the ground in different settings — forest, farm fields, mountains, swamp, desert — every 50 kilometers or so.

The 1970s saw growing awareness of environmental problems and the dangers of terrorism. DOE and NASA took pains to seek public input so that they could attempt to calm public fears. Most people polled worried about the microwave beams linking the SPSs with their rectennas on Earth. Some expressed concern about the environmental impact of the beams, while others feared that terrorists might seize control of an SPS and turn its beam on a city.

Image 11. Image credit: NASA.
NASA pointed out that the beam would be de-focused to reduce risk to the Earth's upper atmosphere, aircraft, and people working at the rectennas. As depicted in the painting above, limited agriculture could take place under the rectennas, directly in the path of the microwave beams. In addition, the microwave transmitter on the SPS could be designed to shut off automatically if its beam drifted. DOE and NASA expected that each rectenna would have around it a "buffer" zone of uninhabited land so that if the beam drifted a small distance before it turned off automatically, only the ring-shaped buffer would be affected.

In this final image of this post (Image 12), we see the SPS fleet near the end of 2015; that is, halfway through the 30-year construction program, when 30 satellites would form a bright line across the southern night sky as viewed from the contiguous United States. A DOE document explained that each satellite would shine a little brighter than Venus. The satellites would appear about as far apart as the three stars making up Orion's belt. Widely available 7 x 50 binoculars would reveal each satellite's rectangular shape to Earth-bound observers.

Image 12. Image credit: NASA.
The string of satellites would remain still against a background of moving stars and planets. In reality, of course, the stars and planets would remain still relative to the rotating Earth and the SPSs would keep up with Earth's rotation.

Every six months, at the time of the spring and autumn equinoxes, each SPS would pass through Earth's shadow near midnight for several days in succession. During its brief shadow passage, a satellite would not produce electricity. One by one, starting with the eastern satellites, the SPSs would redden and grow dark. After about 10 minutes in eclipse, each would return to its full brightness.

The DOE/NASA SPS studies continued into the Administration of President Ronald Reagan, who took office in January 1981. In August 1981, the Congressional Office of Technology Assessment (OTA) published a review of SPS work performed since 1976. The OTA's assessment of the viability of the concept was generally favorable. The Reagan Administration was, however, not enthusiastic about electricity from space or, indeed, from any but conventional sources.

The DOE/NASA SPS studies constituted only a tiny, low-priority portion of the space agency's total activities. The first Space Shuttle test flight in April 1981, the first American piloted space mission since July 1975, was, of course, of far greater consequence.

With the first Shuttle flight under its belt, NASA redoubled its efforts to build support for a Shuttle-launched Earth-orbital space station. The agency sought to portray the Space Station as a space shipyard, a marshalling yard for space tugs and payloads, and a laboratory for exploitation of the unique qualities of space.

Ultimately, only the laboratory function would gain support, in large part because it would be less costly than the shipyard and marshalling functions. Even that support was grudging; though Reagan approved the Station in January 1984, it would undergo a series of redesigns and would not be completed for more than 20 years.


"Power from the Sun: Its Future," Peter Glaser, Science, Vol. 162, 22 November 1968.

Feasibility Study of a Satellite Solar Power Station, NASA Contractor Report 2357, P. Glaser, O. Maynard, J. Mackovciak, and E. Ralph, February 1974.

Memorandum of Understanding Between the Energy Research and Development Administration and the National Aeronautics and Space Administration, 23 June 1975.

The Solar Power Satellite Concept: The Past Decade and the Next Decade, JSC-14898, July 1979.

Some Questions and Answers About the Satellite Power System (SPS), DOE/ER-0049/1, U.S. Department of Energy, Office of Energy Research, Satellite Power System Project Office, January 1980.

Satellite Power System Concept Development and Evaluation Program, Volume I: Technical Assessment Summary Report, NASA Technical Memorandum 58232, NASA Lyndon B. Johnson Space Center, November 1980.

Solar Power Satellites, Office of Technology Assessment, U.S. Congress, August 1981.

More Information

Think Big: A 1970 Flight Schedule for NASA 1969 Integrated Program Plan

Evolution vs. Revolution: The 1970s Battle for NASA's Future

NASA Johnson's Plan to PEP up Shuttle/Spacelab (1981)


  1. Hi David.
    In the UK we've recently seen the cost to construct & run a nuclear powerplant over it's projected 60 year lifetime reach an estimated £37 billion. It's output is expected to be 3200Mwe. I wonder how this compares with the projected figures for the SPS scheme?

    1. I have a document on SPS cost estimates, but I haven't looked at it yet. My plan is to get this onto this blog (a different version appeared on my WIRED blog Beyond Apollo), then poke into the details. I'm not sure that a meaningful cost estimate could emerge from the 1970s studies. So many technical challenges existed. Had studies continued at the same level into the 1980s, then I suspect that we would have seen accelerated advances in low-cost Earth-based solar power that would have rendered SPSs undesirable. I might find out something different as my research continues, however. :-)


    2. Thanks David, I look forward to that. I should perhaps add that the £37 billion doesn't include the cost of decommissioning the plant & disposing of the radioactive end of life material.

    3. I got a PDF about this
      Estimation for entire program with build Prototype are between $39 to $52 Billion in 1976
      thats $168 to $224 billion in 2016 or around 137~183 billion British pound...

      Source (NTRS)
      Initial Technical, Environmental and Economic Evaluation of Space Solar Power Concepts
      Volume II - Detailed Report
      NASA, JSC, august 31, 1976

    4. Kerrin:

      Thanks for clarifying that. I'm bowled over by the sums of money we're talking about here. I wonder how much decommissioning the plant and disposing (more like storing) of the waste would add to that cost?


    5. Michel:

      Thanks for finding that. The costing document I have is from 1980. It would be interesting to compare the 1976 and 1980 documents.


  2. were studies made what to do with manhattan size satellite, once it run toward end of it's life time ?
    mean like Solar cells are so radiation damage that nearly produce any power to transmit.
    what to do with 50,000 metic tons of scrap metals in GEO ?

    1. Such studies were indeed conducted. I haven't delved into them yet, however. As I mentioned in my response to Kerrin, this post is meant as a general intro, then I'll dive into the details. In addition to disposal, lifetime cost, and offshore rectennas, I plan to write about mining the moon to build the satellites and the connection with space colonies and space disposal of nuclear waste. Should be fun. :-)


    2. Thanks for info

      yes, the combination of SPS and O'Niel space colony with Moon as material source,
      is most notorious...

    3. I found the SPS disposal study from 1980
      they estimate it would take 2 years, to disassembly a SPS.


    4. Michel:

      Again, the document I have is from 1980, but it's different. Again, thanks for looking this up. I'll have a look.


    5. We have an individual who pushed for SPSS on thespacereview--but Dwayne Day naysays them. Just some demonstrators would be nice--they could always be converted to solar electric tugs later.

    6. I often find myself in agreement with Dwayne Day. And I should probably say that I don't endorse everything I write about on this blog. That being said, I find the SPS concepts to be very interesting from the standpoint of characterizing spaceflight in the late 1970s.

      There's a tendency for more practical things to "spin off" these big concepts. Thinking about SPS contributed to concern about orbital debris, for example. And SPS construction is an early example of human/robot partnership in space.

      It's not a bad thing, in my view, to dream big. All kinds of space proposals are inspiring, though they might never be built. Space is very big, and, assuming we get our act together and really get out into it, will eventually need some very big engineering.


  3. Fascinating read. Thanks a lot for writing it up - I'm sure there's more to come!
    I can't but help thinking about the economics though. You'd be spending billions upon billions to create an orbital infrastructure that... makes electricity, one of the cheapest commodities?

  4. How does the dramatically lower cost of heavy-lift launch from SpaceX (and soon, others) and the lower cost and higher efficiency of solar panels affect the feasibility equation?

    1. I don't think it would make much difference. SpaceX hasn't demonstrated low-cost heavy lift. Assuming that they can, the payloads they talk about are small for a project of the scale of SPS. That means lots of launches, so the cost would still be enormous. I know we've been conditioned to think of SpaceX and commercial space in general as the solution to all problems, but that's one of the things commercial entities do - they pitch their products. Which is OK, but we have to be informed consumers.

    2. The funny thing is, a Small variant (12.5m diameter) Series Burn VTHL Heavy-Lift Launch Vehicle from the SPS study has similar 120t to Starship, yet only 4000t mass when fully fueled and ready - which is less than simply the fuel mass of Starship! Apparently, aluminium construction/winged flyback/hydrolox upper stage is a pretty efficient design!


I like hearing from my readers. No rules except the obvious ones - please keep it civil and on topic.

Advertiser comments have led me to enable comment moderation.