Japanese to pursue space-based solar power plant

The folks at Space Canada must be loving this. Just six days before the start of its three-day conference in Toronto, where scientists and engineers will get together to discuss space-based solar power, the Japanese government has disclosed that it’s prepared to spend the equivalent of $21 billion (U.S.) to build a 1,000 megawatt solar PV plant in space that would orbit the earth and beam back power 24-hours a day. Mitsubishi Electric and Mitsubishi Heavy Industries have both joined a research group that’s taking on the ambitious project. The group estimates that four square kilometres of solar panels would be needed to do the job. Planned operation: Sometime before 2040.

It might seem like a lot of money, but consider that it would amount to about $70,000 for each home supplied with 3 kilowatts — that’s about double what you could get with a rooftop solar system. But because you’d get the power 24-7, theoretically, it would actually work out to be cheaper on a kilowatt-hour basis. Now, this conveniently ignores the fact that there will be huge energy losses that come with beaming solar energy back to earth and then transmitting it through a grid over presumably long distances. There’s also the fact that, as with most megaprojects, you can bet that $21 billion is lowballing the true cost. Plus, I still can’t get around the cost and fossil fuels that will be required to get these solar panels into space. Wouldn’t it just defeat the purpose of building this clean-energy plant? The Japanese government admits that for this monster project to take off it will have to figure out a way to dramatically reduce the cost of getting panels into space, but that still doesn’t address the CO2 that would be emitted getting them there.

But hey, this is the land that gave us Godzilla. Ye must have faith. One thing for sure: this will be the buzz in Toronto next week. See my earlier Clean Break column on the upcoming conference.

15 thoughts on “Japanese to pursue space-based solar power plant”

  1. Tyler;

    The SPACE Canada crew were indeed very excited to see the Mitsubishi announcement! There will be several people attending our Symposium next week from USEF and JAXA, the Japanese government agencies involved in their space based solar power efforts, so we’ll be able to hear the latest from them.

    The points you raise in your article are all good ones. You’ll be happy to see them being addressed next week. Here are quick responses to a couple of your points:

    You wrote, “this conveniently ignores the fact that there will be huge energy losses that come with beaming solar energy back to earth and then transmitting it through a grid over presumably long distances.” One of the most startling facts about microwave power beaming is that it is actually *highly* efficient; as long as the antennas are sized appropriately for the link distance (and you get all the technical details right), the end-to-end efficiency can be rather better than 50%…regardless of the distance over which the power is beamed! So, most of the electrical power generated on-board the solar power satellite gets delivered as usable electricity on the ground. Also, one of the virtues of this approach is that you can locate the receiving antennas close to the cities that will be consuming the power, actually *shortening* the distance through the grid that the power has to travel (and reducing the amount of real estate that’s needed for new electrical power distribution lines).

    Certainly there’s an energy cost in putting the necessary equipment into space, in the form of rocket fuel, and the energy cost of manufacturing the rockets. However, that’s a one-time cost, after which a solar power satellite will deliver clean power for decades. An accounting is needed to figure out what the net benefit is. One of the main messages from next week’s Symposium will be that this potentially large-scale power supply option deserves serious study and R&D work; that net-energy-cost accounting is one of the things that would come out of such studies.

    – Kieran A. Carroll, Ph.D.
    VP Technology
    SPACE Canada

  2. The Shuttle’s main engines use Liquid Hydrogen and Liquid Oxygen. The exhaust is water vapour. The next generation of NASA vehicles are going to use the same launch technology as the current shuttle program.

    There is also this exciting research (http://www.popsci.com/military-aviation-amp-space/article/2009-08/fly-me-moon-water-ice) that is testing aluminum powder and water ice as rocket fuel. Far more environmentally friendly then kerosene based rocket fuels. So by the time the Japanese are ready to launch, the only CO2 could be from making the panels and the rockets.

  3. Since part of the sunlight is reflected into space when it hit the atmosphere, taking that energy out there to “force it” directly to earth, will it increase the amount of energy received on earth ? Hence increasing the warming ?

    I know that this is probably much better than CO2 emissions, but lets consider this before feeding the entire planet with this kind of energy source….

  4. I just don’t see this technology happening with all the other avenues for renewable energy to pursue that is, ah, closer to home. But, LOL, the story was worth it just for the picture;-)

  5. OK, I just did the math on your other point. Assuming a 1 GW solar power satellite has a total mass of 15,000 tonnes (1,000 to 2,000 tonnes for the solar arrays, the rest of the mass for supporting structure, mirrors, and power beaming equipment), it would take something like 1000 launches of a SpaceX Falcon 9 Heavy (http://en.wikipedia.org/wiki/Falcon_9) to place it into its operating orbit (in pieces, obviously, that would need assembly in orbit). A Falcon 9 Heavy weighs about 885 tonnes at launch; I’m going to assume that roughly 400 tonnes of that is fuel (essentially kerosene). So, 1000 launches would burn 400,000 tonnes of kerosene. The energy value of that kerosene is about 4,500 kW-hr/tonne, so 400,000 tonnes of kerosene have an energy value of about 1.8 billion kW-hrs—that is, if that kerosene was burned in an oil-fired power plant, it would produce about 1.8 billion kW-hrs of electricity.

    How long will it take that 1 GW solar power satellite to “pay back” that energy investment? About 1,800 hours (1.8 billion kW-hrs * 1,000 W/kW / 1 billion W), or 75 days. So, after 2-1/2 months of operation a solar power satellite would have paid back the CO2 production investment needed to launch it; for the next 20-40 years of operation, it would be offsetting about 2 million tonnes (about 12 million barrels of oil) worth of CO2 production per year from carbon-burning electrical power plants.

    (Caveat—I don’t entirely trust any math that I do after 11:00 PM…if anyone reading this finds an arithmetic error in the above, please let me know… 🙂

    E. Martin raised the question of whether importing power from solar power satellites would add heat to the Earth, making global warming worse. The answer is “no,” because it will be displacing other, less-efficient forms of power generation. a 1 GW thermal power plant (be it oil-, gas-, coal- or uranium-fired) generally produces something like 1 to 2 GW of waste heat, in addition to the 1 GW of electricity that it generates (which also ends up as waste heat, after being used by consumers), for a net heat contribution to the environment of 2-3 GW (in addition to the global warming effect caused by any CO2 released). A solar power satellite, on the other hand, delivers the 1 GW to the Earth, but is converted with very high efficiency to electricity (less than 20% waste heat), for a net heat delivered to the environment of around 1.2 GW, less than 1/2 of the amount added to the environment from the other sources (and also contributing no CO2).

    Paul C. made the point that other forms of renewable power are “closer to home”, presumably easier, cheaper, etc. True…unfortunately, they don’t scale up usefully. Wind power doesn’t work when the wind doesn’t blow. Terrestrial solar power doesn’t work at night. Installations of this type need to be super-sized (made something like 10x larger) to make up for the times when they have no generating capacity, and need to have large battery banks added to store power for when they aren’t generating. Batteries are *expensive*! So expensive that it’s cheaper to put the darned solar arrays in orbit, where the Sun always shines, to avoid the need for batteries.

    If you come to the Symposium next week, you’ll see these sorts of issues discussed in depth, and get to meet the people who’ve been studying this approach for many years, and have invented many of the elements of this solution.

    – Kieran A. Carroll, Ph.D.
    VP Technology
    SPACE Canada

  6. Kieren, I’m sold — I do hope to attend though can’t attend all days. Which day would you suggest is most important — Day 1? Also will you be around to chat?


  7. Tyler;

    The Symposium is really 2 conferences in one.

    Day 1 is an overview of high-level issues related to space based solar power. A main objective of Day 1 is to be a “summit meeting” between leaders of the global climate change community, and of the space based solar power technical community. It will begin with a talk on the global warming issue (by Prof. Richard Peltier of the University of Toronto, a principal author of recent Intergovernmental Panel on Climate Change reports), to remind us of the problem to be solved. Bryan Erb and John Mankins, both of them ex-NASA solar power satellite engineering experts (Bryan’s experience includes being one of the top-level Canadian designers and engineering managers for the Apollo lunar spacecraft in the 1960s), will then provide an overview of what space based solar power system designs look like and how they would work, and critically analyze the pros and cons of this technology versus other large-scale global power supply options. Following this will be a live demo of a microwave power beaming system operating a rover vehicle, then sessions exploring space policy, media, and university/industry collaboration angles.

    Days 2 and 3 are a technical conference, where most of the world’s leading space based solar power researchers will gather to present papers on their recent research. (For example, I’ll be giving a paper on the possibility of using microwave power transmission to convey large-scale power from Labrador to Newfoundland.)

    From the perspective of issues of interest to the general public, Day 1 is the best day of the Symposium to attend. (Not to detract from the importance of the technical conference! However, the topics discussed there will be more specialized.) Also, I believe that most people attending Days 2 and 3 will also be there on Day 1, whereas some of the Day 1 attendees will likely not stay for the technical conference, so there will be more people to meet and talk to on Day 1.

    I’m looking forward to seeing you there! I’ll be there all 3 days, and will be very happy to chat with you.

    – Kieran

  8. I also caculated the energy cost elsewhere and found it to be quite fast a payback!

    The first problem is raw dollar cost. 21 billon per GWe, that’s 21 dollars per Watt. Lowest cost land based solar farms (big thin film free field powerplants) are around 3 dollars a Watt last I checked and it is declining fairly rapidly. Space solar won’t be 24/7 power, ie no 100% capacity factor, due to the shadowing effect of the earth problem. 50% would be the theoretical maximum above the equator. The system would have to be situated above the poles to get 24/7 power to a fixed location on the earth, which makes transmission to non-polar countries rather difficult. So basically a big HVDC grid has to be built as well. I think the economics of high resource solar with HVDC to low solar resource areas is actually a lot better and less risky in terms of cost.

    There is also the inherent problem of weapons implications, since the microwave intensity beamed towards earth is big enough to fry your brain in mere minutes. This risk is not to be underestimated or waved at. Frankly, space solar scares the shit out of me for this exact reason.

    Bottom line, IMHO this is 21 billion better spent elsewhere (land based renewables, energy efficiency, and electric transport, to name just a few).

  9. Also, “installations like that” – solar in high resource areas – don’t need a 10x buildout, only a 3x buildout compared to peak capacity. This is due to demand not being constant. 3x peak is enough for over 90% of US grid plus electric vehicle transition. See ausra study:


    Japan doesn’t have such a good solar resource, though it isn’t nearly bad enough to justifiy the abominable economics of space solar. Don’t give me the “it will get cheaper in time” argument, that applies to land based solar and storage as well. Even at today’s CSP, PV and storage cost of thermal storage and underground pumped hydro land based solar is way cheaper than 21 dollars per Watt.

    The weapons implications are even more severe than the economics, and is reason enough to abandon the idea of space solar power altogether. I’m not usually immediately opposed to new concepts but space based solar is just not a good idea.

  10. Cyril;

    Thanks for the ausra study link; I’ll give it a look. Personally, I’m primarily a spacecraft engineer, and while I know probably more than most people about how to design solar power satellites, I’m very much still in learning mode when it comes to the details of how the electrical power utility business works. That being said, others who’ve been in this field for many years longer than me have done numerous comparisons of the economics of space-based solar power against those of pretty much every other available large-scale power source, and have certainly taken peak versus baseload power into account; their analyses produce pretty attractive-looking cost estimates (and many of these are *very* impressively smart and accomplished people!).

    Too bad you weren’t at the symposium this week. You have a couple of misconceptions about space-based solar power (ones which are are quite common, we’ve found), and you would have found numerous experts there who would have been able to address them. I’ll have ago here:

    You wrote, “Space solar won’t be 24/7 power, ie no 100% capacity factor, due to the shadowing effect of the earth problem. 50% would be the theoretical maximum above the equator.” If the Earth’s equatorial plane were in the same plane in which the Earth orbits around the Sun (the ecliptic plane), then what you wrote would be true. However, the Earth’s equator is inclined at 23.5 degrees to the ecliptic plane, and so a satellite in an equatorial orbit stays out of Earth’s shadow for most of the year. It’s true that an equatorial-orbit solar power satellite would (as do all geostationary communications satellites) experience eclipses twice per year, for a few weeks around the spring and fall equinoxes. These eclipses are fairly brief—at worst 72 minutes long, once per day during those periods. Fortunately these are at local midnight, when power demand is near its lowest point of the day.

    While it’s a bit hard to visualize at first, geostationary solar power satellites actually *can* provide power 24/7, for most of the year. This greatly offsets the cost differential between space-based solar power and terrestrial solar power, since the latter has such poor availability—about 10% of the availability of power in space.

    This also means that space-based solar power does *not* require a long-distance grid to deliver power to its customers; indeed, one of its big advantages is that the receiving antennas can can be built very close to the customers, and power beamed from space almost to their doorstep (i.e., within a few 10’s of kms).

    You wrote, “There is also the inherent problem of weapons implications, since the microwave intensity beamed towards earth is big enough to fry your brain in mere minutes. This risk is not to be underestimated or waved at. Frankly, space solar scares the shit out of me for this exact reason.” This one is even harder to develop an understanding of, without first studying electromagnetic theory a bit, so I’m not surprised at how common this misconception is. Fortunately, a microwave power transmission system of the type we’re discussing is *physically incapable* of being made into the sort of “death ray” that you’re scared of.

    The physics of the situation is that over long distances (such as the distance from geostationary orbit to the Earth’s surface), the width of a radio beam (or a light beam) expands as it travels from its transmitter. This means that by the time a radio beam reaches the Earth’s surface, it is much less intense than it was up at the transmitter antenna. By “less intense,” I mean that the amount of power flowing through each square meter of cross section of the beam is smaller down at ground level, than it is high up in space. For a given size of transmitter antenna up in space, it is physically impossible to focus a radio beam on the ground to smaller than a certain size. For a typical solar power satellite design, the transmitter antenna would be 1 km ni diameter, and at an altitude of 36,000 km the spot size on the ground would be about 10 km in diameter. For a given amount of power being pumped into the beam up at the satellite, then, you can figure out the absolute maximum power density that the beam will have at ground level. If the satellite collects (say) 1 GW of electrical power, and sends all of that out the transmit antenna as RF power, then the average power density am when it reaches the ground will be 1 GW divided by the beam’s area of pi*10*10/4 = 78.5 square km, which is 12 Watts per square meter—about 2% of the intensity of sunlight on a bright day in the desert, and about equal to Health Canada’s (pretty conservative) limit for microwave exposure from cell-phones and the like. (The intensity will be a *bit* higher than average at the center of the beam, still a perfectly safe level, but to meet Health Canada regulations that’d be fenced off to keep the general public away from that area.) So, unless you’re one of those people who thinks that their cell-phone is emitting death-rays (in which case I can’t help you :-), you’ll see that the microwave beam from a solar power satellite is at a very safe intensity level.

    You could take my word for it when I say that it’s not possible to focus the beam more than that, for a given transmitter size and distance. If you’d like to check up on that, I’d be happy to point you towards relevant electrical engineering textbooks and/or technical papers (which is where this sort of specialized knowledge is to be found).

    Hope this helps…

    – Kieran A. Carroll, Ph.D.
    VP Technology
    SPACE Canada

  11. @ 12 Watts per square meter, the energy density would be inferiour to land based solar.

    You can tell me all about smart people, but 21 dollars per Watt is not competitive even @ 100% capacity factor.

    The problem with solar is cost. Approaches that increase the cost and risk are not helping.

    I just wish all the smart people would get together on solving energy and climateproblems in my lifetime.

  12. Kieran,

    I am an electrical engineer and work in microwave communications. One thing I can not understand is how the microwave beam would get through the atmosphere? (what if it is cloudy or there is a lightning storm). I have had huge attenuation on microwave beams that are only 10 to 20 km apart because of weather. I under stand that you would be in a higher freq range to transport power but that would just lead more attenuation and the signal would get so distorted that it would not be useable.
    I don’t want to sound argumentative but when ever I hear about space based solar I always think of it as junk science due to the problem of getting through the atmosphere, what am I missing?

Comments are closed.