116 thoughts on “Space Solar Power”

  1. Using a 2003 JAXA reference model with a 1-gigawatt station weighing 10,000 tons, Sasaki says power would cost a prohibitive $1.12/kwh at a launch cost to low Earth orbit (LEO) of $10,000 per kilogram. That is in the ballpark of what space launch costs today. Cut that to $1,000 a kilogram—in the ballpark for a reusable launch vehicle (RLV)—and electricity from space drops to 18 cents/kwh.

    The list price for a Falcon Heavy launch is reportedly $110 million for 110,000 pounds to LEO, so $1000 a pound or $2200 a kilogram. If SpaceX succeeds with reusability, the price could potentially be well under $1000 per KG. Fly a large load into LEO, deploy some solar arrays and use electric propulsion to fly the unit to GEO. Once a big unit gets to GEO, either rendezvous with earlier units or fly them in formation allowing for better redundancy. Put two large assemblies in GEO separated by at least 18 degrees and you’ll never have both of them in solar eclipse at the same time.

    1. 10,000 tons at ~50 tons per flight is only 200 flights! Plus, just know the fuel to get it from LEO to its operational orbit.. but let’s just ignore all that for now. At the current launch rate of Falcon Heavy, how long will it take to do that many launches? Oh yeah, it’s not flying at all yet.. umm.. okay, what’s a feasible flight rate? 20 flights per year? That’s a pretty optimistic number, let’s go with that, shall we? How long to get this thing built? 10 years at a billion dollars per year? Yeah, that’s.. umm.. how were we making profit again here?

      Maybe you were thinking a higher flight rate? Like, 20 years from now? If so, why are we extrapolating from the Falcon Heavy?

      1. If they get reusability, I expect flight rate to go up and cost to go down, and fairly quickly. And even if it was a billion per year, that’s only ten billion. That’s the cost of a nuclear plant or three.

        1. The point was that you can’t take ten years to build it.. and 100 Falcon Heavy flights per year is at least 20 years away – more likely it’s a completely different kind of vehicle away.

          a 10,000 ton facility in LEO (if launched from Earth) is not right around the corner.. a 10,000 ton facility in GEO (or whatever your operational orbit is for SBSP) is even further away.

          1. 100 Falcon Heavy flights per year is at least 20 years away

            That is not at all obvious to me. With reusable cores they aren’t constrained by vehicle production rate. If they can turn them around quickly,they could launch a minimum of a couple times per week per launch site, and Gwynne says they plan several launch sites.

          2. I don’t they’re going to be able to turn around their current design as fast as that. There’s still a learning curve that has to be walked. It won’t happen overnight.

            Even they could, I don’t think the launch sites can support that. So that’s another learning curve.

            Finally, there’s the chicken-or-egg problem of getting enough payloads to pay for all this learning. It’s not going to be cheap enough to do it on their own dime, but they can certainly learn a lot with their program-formerly-known-as-Grasshopper.

          4. There’s not much point worrying about reusability unless you plan to fly a lot. I think the Saturn V studies showed there was no benefit to recovering and reusing the S-1c stage unless you were going to fly at least a hundred or so times.

            I’d say it really all comes down to how fast they can refurbish the stages and how low they can drive the cost. At $100 a pound, that space hotel starts to become affordable to the kind of people who currently take vacations to Antarctica, for example.

        2. Do you believe competition would be behind by 20 years? SpaceX is basically showing everyone at least one way, to achieve reusablity. I do not believe Russia and China will sit idlely by as the lose their entire launch market. Also, you see in a lot of transportation costs that get averaged out. 8 falcon heavies, 4 Falcon 9’s, 2 Atlas V, 2 Delta IV, 2 Proton et cetera. and just average the cost out for the various modules. Total launch capacity could do this alot faster, especially as prices start falling as the competitive pressure increases.

          2. I would kind of have to disagree. There is a current global launch market of about 70plus launches per year. SpaceX has a 5 billion dollar backlog, that appears to be an “actual market”. You also have commercial crew approaching and commercial stations and the cargo runs to supply both the commercial and ISS stations.

            The market is already here, it is just a question how fast it expands.

          3. Now now, counting government launches is unrealistic. There’s no way the Russian government is going to junk their launch vehicles and hire SpaceX to launch their payloads and the same goes for the US government. In terms of a launch market, there’s a lot less than is currently needed for a single high flight rate reusable launch vehicle. The launch market has to fundamentally change before reusable launch vehicles make sense economically, which is why the only groups working on reusable launch vehicles are doing it for non-economic reasons.

          5. There has to be a market for there to be a response.. I’ve yet to actually see any evidence that one is going to materialize. I think that, fundamentally, spaceflight has to become something completely different to what it is today before a “low cost spaceflight” market materializes, and it’s very far from given that this is going to occur.

            In short: wait and see.

          6. Sure, the Russian, Chinese and EU governments will still use their national launchers to put up official payloads. But SpaceX is going to take all of ILS’s commercial business, all of Arianespace’s commercial business and all of ULA’s U.S. gov’t. business in three to five years.

            As for the size of the launch market going forward:

            (1) There will be numerically more actual launches of comsats owing to the transition from chemical to electric propulsion for GTO-to-GEO maneuvering and stationkeeping. This makes the birds lighter and smaller. Right now, the bigger, heavier birds have to ride up two at a time on Ariane 5 to make it economically reasonable. The coordination problems of double payloads have been considerable; triple payloads would be worse. Successors to the birds that rode up in tandem on an Ariane 5 will, in future, ride up alone, each on its own Falcon 9. Arianespace currently has more of the GEO comsat launch market than anyone else – nearly all tandem missions. Take whatever their current launch rate is, double it and put it in SpaceX’s column.

            (2) Orbcomm and Iridium are building new LEO comsat constellations. SpaceX already has contracts with both. That CASSIOPE bird SpaceX launched last September may also have completely commercial siblings in future. People who order smallsats by the dozen of the gross will be going to SpaceX to get them into orbit.

            (3) Large LEO constellations aren’t only for comsats. There are several recently formed Earth imaging companies with sizeable constellations of smallsats/cubesats to deploy. SpaceX has some of this business and will likely get most or all of the rest.

            (4) As additional space business cases now close owing to SpaceX’s already felt effect on launch costs – and still more close in the next few years based on their reusability initiatives – SpaceX will be in a position to take virtually all of the increase in demand attributable to price elasticity.

            For a guy who’s supposed to be all about markets, you seem oddly disinclined to believe in their price elasticity for some reason.

          7. What I “believe in” is evidence, and there is none for a competitive launch market.

            Your belief in markets seems to deny the abundant evidence that the majority of current payloads – especially government payloads – don’t care about price. Oh, they make noises like they care about price, I’ll give you that, but they still pay more than they need to and they’ll keep doing that whether SpaceX starts flying a magical unicorn or not.

            As for the rest of your analysis, could ya get any more flimsy evidence to base it on? At least wait until there’s a single profitable LEO constellation (there hasn’t even been one for comms) before declaring victory there, eh? The only other “business cases” that SpaceX has helped close so far is government funded ride-alongs on Dragon. That’s your evidence for a pent-up demand?

            At least try to sell me the regular-launch, empty-or-full changing the nature of payload processing.. it at least has the virtue of analogy to other successful transport markets.. but the usual follow up to that is to suggest a government subsidy to kickstart the industry. Heck, you could suggest the President needs to make space solar power a national objective and fund the whole thing at Apollo levels.. get it over with.

          8. What I “believe in” is evidence, and there is none for a competitive launch market.

            Your belief in markets seems to deny the abundant evidence that the majority of current payloads – especially government payloads – don’t care about price.

            Might I venture to suggest that your thoroughly goofy thesis is pretty much demolished by merely citing the existence of Sea Launch and ILS in the commercial – if I may make so bold – launch market? Both Sea Launch and ILS have had what may be charitably described as checkered records of success. Yet they continue to attract a fair amount of business. Why? It’s obviously not their superlative reliability. If all the comsat operators were interested in was dull reliability at whatever the asking price, Arianespace would long since have driven Sea Launch and ILS into the ground. And yet, there they are. The obvious answer – to everyone except you, apparently – is that price does matter to launch services customers and it matters enough that they’re willing to run significant risks of individual mission loss in order to pursue it. Now SpaceX comes along with still lower rpices and better reliability. I don’t have to be a genius to see which way things are going to go.

            As to government payloads, price insensitivity is hardly uniform even there. Russia and China will cetainly remain autarkic in launch capability as long as they can for their own military and civil payloads, but I see no one claiming otherwise. Europe, too, will try to stick with their self-developed Arianespace cartel as long as they can, but, denuded of commercial business, it’s problematical how long that can last with the Eurozone creaking under unsustainable entitlements of every other sort in addition to a subsidized space industry. In the U.S., deliberate profligacy is hardly unknown, but it tends not to last once cheaper alternatives arise. ULA has had a nice, fat litle monopoly, but they’re about to flensed and rendered. Hard times are coming to Huntsville and Decatur, AL.

      2. Since the system is modular, you launch a 50 ton piece, fly it to GEO and start operations. When the next piece is ready, you do the same thing. You don’t have to wait until all the pieces are there to begin operations. Lather-rinse-repeat until finished. Starting early operations allows you to test your design, find any problems, implement fixes and generate revenue early. Why would anyone want to wait until the entire unit is ready to take it to GEO? Like you said, that would take years.

  2. It also depends on how long you can amortize the cost and maintenance costs. If you can build something that doesn’t require maintenance you might be about to amortize it to almost nothing.

    1. More so, if you can use high profit margin services to amortize the installation costs then everything else is gravy, and you don’t necessarily need to make the same $/kwh everywhere. For example, if you can sell power to the military, remote mining operations, and what-have-you then you can make a huge profit doing so fairly easily. And once that profit has paid off your capital costs then you will be in a position to sell power based on your ongoing costs for the remaining lifetime of the satellite (which could easily be many decades), enabling you to compete in lower margin markets, such as municipal power in the developed world.

  3. I keep coming back to using something like the ‘Solar Power Towers’ in space the second you move from ‘convenient power with price no object’ to ‘industrial scale electrical production.’

    1) Both a photoelectric approach and a heat cycle approach require significant area. The sheer area required translates to the part that statistically is going to take attrition. I’d rather the major attrition be in the (relatively) cheap mirror section than in the standard (not-so-cheap) photoelectrics. (The main turbine and resevoir area can have protection – also known as insulation, which is desirable anyway.)

    2) The waste heat dumping comparison. The mirror-collectors shouldn’t have significant issues – both because they don’t care as much about what temperature they are, and because their back sides should be sufficient. (And if it isn’t, design from different materials.) The ‘power tower’, has a -hot- side, but the core pieces are fundamentally designed for operating at the highest possible temperatures – which will be well above the temperature any photocell ever thought of surviving. But the waste heat has to be carried away radiatively. But even if you assume a backdrop of 100K, you can dump a hell of a lot of heat if you’re willing to allow your emitter to be reasonably hot. 200C (or 300C or wherever.) means the T^4 is in your favor. Yes, I’m aware of how difficult radiative cooling is normally. This isn’t on earth, and you’re already committed to having plumbing for hot liquids – it’s a matter of how much radiator do you need to pump it through. It won’t be even vaguely the size of the collectors.

    The part that’s a problem is getting it anywhere useful.

    1. I have long thought that a heat engine makes a lot more sense than photovoltaic panels when building at the scale of a solar power satellite. If the cold side is 3 Kelvin, the hot side only needs to be 300 Kelvin to bring the maximum conversion efficiency above 99% (compared to low teens for PV panels). The bulk of the mass of the satellite is then nothing more complex than a sheet of aluminum foil rather than much more expensive PV. Launch your heat engine and transmitter, then attach aluminum foil in a spiral around that central heat engine. The SPS could start beaming power after only a handful of launches.

      I can see SPS being used to power cities, but for me a more interesting application is beaming power to spaceships. For a terrestrial power application the antenna and rectenna must be large, but for beaming power point to point in space the microwaves can be a MASER. Virtually anything can be used as a propellant if heated to a plasma; the ISP is just ridiculous compared to chemical burning.

      1. It’s my understanding that although ‘deep space’ counts as 3K, the radiative background near earth is treated as more like 40K. And keeping the ‘cold side’ at even that temperature is still difficult. (We’re aiming for high energy flux, not just ‘how cold would a shadowed object get’) But you can balance that by not being quite so worried about the -high- end temperature (on earth, you’d be quite focused on how much heat was being lost by convection and conduction away from your ‘boiler’ because you wouldn’t have quite such an open supply.) And keeping the ‘cold side’ to a ‘mere’ 200K or 300K shouldn’t be too bad. But even 500K ‘cold side’ should allow a reasonable efficiency that completely crushes PV when you get into the mass and attrition analyses.

        1. Cassegain reflector with a hyperbolic primary.

          By using a hyperbolic primary, you can (mostly) use the light-pressure itself to hold the darn thing open. Think long-ribbon “Solar sail”.

          One Falcon Heavy = 369,000 square meters of “mirror” in LEO, not a good final destination but… that much light pressure means (to me) figure out how to fly the darn thing. A rolled heavy foil ribbon (where maximum width is limited by the fairing), heavily perforated so pebble-impactors just rip a 10cmx10cm square (or hex, whatever) completely out of the pattern. And even that is probably seriously over designed. You aren’t unrolling it ‘in vaccum’ (the removal of the coiling happens ‘in the box’), and it isn’t at super-brittle temperatures coming straight off the roll. Guywires for the edges, unless a slight thickening of the ribbon works. Now if you “just” hold the two ends of the ribbon into the solar wind, the even pressure across the length should cause a natural hyperbola along the length. -And- a natural hyperbola across the width.

          That’s 500MW of insolation from a pretty heavy foil. Or a -lot- more if you can get anywhere near the thinner foils. Make the primary cheap.

          No, you won’t get the 99.6+% excellent reflective properties of polished glass-backed Al mirors from Spectra Physics, but you do get a whole lot of raw collection for dirt cheap.

          Yes, the secondary and the turbine/resevoir are more complex than PV. But a significant chunk of the complex bits can be shipped “finished”. A Falcon Heavy-sized turbine should be more than enough.

          1. Literally dirt cheap. There’s lots of Aluminum on the surface of the moon. Titanium too.

      2. IANAE, but the idea of a simple heat engine has a “perpetual motion machine” flavor to me. But if it works, awesome.

        I’m concerned that trying to punch this amount of power through the atmosphere would cause problems, problems that wouldn’t exist with your idea of powering ships and stations. 🙂

        2. It is far from a perpetual motion machine. The opposite, in fact. It’s a heat engine. It needs an energy input (sunlight) and a radiative output. It borrows energy from the sunlight at the expense of increasing entropy at the radiator. It’s the exact same process that drives industrial electricity generation in coal or gas or nuclear power plants.

          1. Insolation is something like 125 million gigawatts (1000 W/m^2 times pi times the radius of the Earth squared.

      1. Well there is about 250 billion a year going into terrestrial renewables, Space Solar, well,
        perhaps a few million.

          1. Capital investment has declined. Markedly in solar.
            However Capacity keeps
            rising because the cells keep getting cheaper. It’s the horrible treadmill Memory
            companies were on in the early 80s. The more they made, the cheaper it got,
            but, they were growing capacity and market while revenues were stagnating.

            http://i0.wp.com/cleantechnica.com/files/2013/05/price-of-solar-power-drop-graph.jpg

            when you are skiing down this wave, investment drops dramatically even as the market grows.

        3. Funneling taxpayer money to political cronies and endless subsidizies hardly counts as an investment. It certainly isn’t a rational or sustainable one.

          2. Utility companies are terrified of being forced to incorporate unreliable and inconsistent generation capabilities into their grids at high prices. Wind output varies considerably and solar output can drop sharply when a cloud’s shadow passes over the arrays. This means the utility companies have to keep gas turbines online to level out the electricity to keep the grid stable. This makes for higher operating costs and lower reliability for their customers. Sure, the utility companies are terrified – by idiots pushing the physics of wishful thinking into an unforgiving environment.

          3. “Utility companies are terrified of being forced to incorporate unreliable and inconsistent generation capabilities into their grids at high prices.”

            Spoken like a man who knows nothing of digital control

          4. And you speak like a man who knows nothing at all except how to be a parasite and political whore.

      2. If there’s no night in space, why is space so dark at night? Why is space always so dark?

        There. I just saved you the anticipation of waiting for his reply. 😀

        Going by the numbers here, a good solar site like Albuquerque averages 5.7 kWhr per square meter per day of incident solar radiation. The same collector in space would be exposed to 32.9 kWhr per square meter per day, which is a 5.77 times greater. The difference is like having solar cells that are 30 percent efficient versus cells that are only 5.2 percent efficient. More importantly, the output of the space array is constant, becoming part of the base load instead of being intermittent power that requires a separate back-up unit.

        1. The Solar intensity is great above the atmosphere, and the lack of clouds is good.
          However beam transmission losses are a significant problem and the power needs to
          move on terrestrial grid.

          Space Solar is the product of the 1950’s and is trapped in the 50s

          1. Unless someone invents a wiz bang wireless energy transmission method, with little losses and small infrastructure investment, it is going to stay that way.

            If you lose a lot of energy on transmission plus you need to pay for a terrestrial receiving station which costs alone as much as the damned terrestrial solar panels would cost in the first place why bother putting them in orbit?

            I think space based solar power only makes sense for space applications. Someone should consider splitting power from having it on every satellite to a separate grid.

          2. Someone already invented that. It’s a small dipole antenna (two bent pieces of wire) and a diode, the same as an old AM crystal radio set. It’s called a rectenna and it’s over 80% efficient.

          3. “It’s called a rectenna and it’s over 80% efficient.”

            decent number but, you lose converting Electricity to micro-wave, then you
            lose some on the down link.

            You only lose 40% of terrestrial solar to clouds.

          4. You lose 100% of terrestrial solar to night. Unless you have very expensive tracking arrays, your terrestrial installations will also only capture a fraction of the amount of solar energy that falls on them as well due to incident losses. For other an point installations, transmission losses apply to large solar arrays as well.

    2. With space based solar power you get 24/7 energy output at full power (efficiency * insolation). With ground based power you get a tiny fraction of that, interrupted by night and reduced by clouds. With SPS the capital equipment easily lasts for decades with zero maintenance, with ground based solar power regular maintenance is required. With SPS the ground segment can be extremely inexpensive and can be used to receive power 24/7.

      1. “With SPS the capital equipment easily lasts for decades with zero maintenance, ”

        spoken by a man who knows very little about spacecraft.

        Lifespan of a GEO bird is 15 years.

        1. Hubble has been around for more than that (yeah I know it is in LEO) but it had people doing servicing. AFAIK one reason they de-orbit the satellites is because they run out of fuel to keep them on station.

          2. I have no idea where you got that number. LEO orbits are not 80% dark. That little matter aside, though, nobody is suggesting putting SPS in LEO anyway.

          3. LEO orbits are not 80% dark, but they are 50% dark.
            Worse they are over water or empty terrain some 80% of the time.

            Iridium discovered there was very little revenue in a leo constellation.
            This is the same problem only worse. Lots of reboost, rough ionizing environment
            and little revenue.

          6. I never “howl,” let alone “prefer” it, you moron.

            Are you so far out on the spectrum that you are utterly incapable of discerning the emotions of others?

            The service revenue to governments made up 23% of Iridium’s revenues in 2012

            Gee, that means that 77% of them came from elsewhere. But basic arithmetic, like everything else, is not your strong suit.

        2. Design life is usually 15 years but the actual operational lifespans for satellites often exceeds the design life by many years. Older comsats used silicon solar arrays that degraded about 3% per year which also limited the satellite’s life. Newer generation solar cells are not only more efficient, they degrade far less than silicon. Battery technology has also increased far beyond the old nicads with their memory issues. Propellant is still a limiting factor but the switch to electric propulsion has made that much less of an issue. One of the biggest factors today is that the technology will become obsolete or obsolscent in 15 years and they’ll want to replace the satellite with something better to make more profitable use of limited GEO slots.

        3. Lifespan of a GEO bird is 15 years.

          I didn’t realize all GEO birds were the same. Do they die of old age at 15 years? How is that DSP has had GEO birds that have outlived 15 years (are they just lying)?

          Is satellite life influenced by particular failure modes rather than an arbitrary hard deadline? Are the GEO satellites you describe as having a 15yr lifespan in any significant way different than satellites that do little more than collect and send energy to Earth?

          spoken by a man who knows very little about spacecraft.
          I think the pot is over here mouthin’ off.

        4. That’s because it takes about that long to use up all the stationkeeping fuel a comsat gets launched with. During said 15 years, though, no maintenance is necessary – or possible. Terrestrial solar is far more maintenance-intensive, mainly because of weather. No amount of maintenance can make terrestrial solar capable of becoming a baseload technology, however. Not so, SPS. SPS with useful lifespans of decades are straightforwardly possible from an engineering standpoint, it’s just that current launch costs render them uneconomic. As noted, that is in the process of changing. When launch costs fall to a point where the SPS business case closes for users willing to pay the still fairly high initial costs – U.S. and allied militaries plus remote resource extraction facilities – then you’ll see commercial SPS proposals floated and funded.

          1. “When launch costs fall to a point where the SPS business case closes ”

            Perhaps when Malia Obama is sworn in for her second term as President

          2. I think it’ll close for certain users (U.S. and allied military, remote resource extraction operations) within 10 years. How old is Malia? I understood one has to be 35 to be elected to even a first term as President.

        5. You’re confusing lifetime with operational service. Satellites are replaced frequently by more advanced versions because there are limited orbital slots. This is especially true over the last 2 decades due to the huge advancements in satellite technology that have happened. As a case in point, consider Marisat F2, which lasted 32 years despite a glitch that dumped half its propellant near the start of its life. It was only retired because it was running low on fuel. Modern satellites have even greater potential longevity due to the extensive use of electric propulsion. In almost every case satellites are retired due to fuel shortages not equipment failures.

          Because power transfer isn’t as succeptible to technological obsolescence, it is much more worthwhile to design a solar power satellite for a service life of many decades, something which is very possible with current technology.

    3. Also, you don’t have the filtering effects of the atmosphere on a space solar power station as you would on an earth-based one.

    1. I don’t think that makes as much sense as using the light directly.

      You’ll be directing light toward a solar farm where the cells are about 30 percent efficient, and where they probably only have 80% coverage (frames, maintenance paths, etc) and you’ll be illuminating them at a slight angle, probably 30 degrees. Multiply all that together and you’d have about 20 percent efficiency. Then, in part, you’d use a lot of that electricity for lighting at 10 to 30 percent efficiency, so you end up providing 2 to 6 percent as much light as you beamed down. if instead you just used the beamed light as illumination, replacing things like outdoor lighting (stadiums, streetlights, etc), you’d perhaps do far better.

      1. if instead you just used the beamed light as illumination, replacing things like outdoor lighting (stadiums, streetlights, etc), you’d perhaps do far better

        Some buildings do this by transmitting natural light via fiber optics connected to a collector on the roof. But the thing is most electricity consumed is not spent on lighting applications.

        Not to mention buildings already have windows and skylights, etc.

      2. Take a look at pages 66 to 68 of this US lighting market characterization PDF.

        Outdoor lighting consumes 118 terawatt hours of electricity per year in the US, which at 12 cents per kWh comes to $14 billion dollars a year in electricity costs, plus the purchase, installation, and maintenance of 178 million outdoor lights.

        Obviously direct lighting from space wouldn’t work on cloudy nights, but for some areas and applications it might be ideal, and the satellite just has to be a big reflector to be useful.

  6. It would seem to me that if one of these expensive power-sats broke down you couldn’t just junk it You’d have to go out there and fix it, either robotically or with people.

    That requires more of a spacefaring capability than we have right now (not technical capability – infrastructure). That could be a good thing.

    1. By the time launch is cheap enough for SPS to be viable there will be so much more stuff put in orbit in general that the genesis of on orbit servicing infrastructure will be pretty much a given. Right now it seems so impossibly difficult, but with launch at 1/10th the cost and the consequent increase in population of LEO the problem gets vastly easier. Imagine if you could run a company that had operational costs of $100 million/year and was capable of keeping half a dozen crew in orbit continuously, that sort of thing is going to be possible within the next 20 years or so.

      1. “By the time launch is cheap enough for SPS to be viable there will be so much more stuff put in orbit in general that the genesis of on orbit servicing infrastructure will be pretty much a given.”

        Maybe. But low cost alone does not guarantee infrastructure. You have to have a market..a reason, for people to spend money to build infrastructure….i.e. profit.

        I can see cheap launches resulting in at least two modes:

        1) cost of launch to LEO is so cheap you don’t bother with repair infrastructure – you just put another gizmo up.

        2) Or,as you posit, repair infrastructure is created. Whether that’s manned or not is another question.

        1. You can’t just put another gizmo up or the clutter becomes a problem. You need to do something with the old broken gizmo.

      2. “By the time launch is cheap enough for SPS to be viable”

        that’s the problem.

        Terrestrial stuff is cheap and scales. Space Solar gets about 6 times as much sun
        and 2 times as much capacity but, you will lose about 30% of the power to transmission
        losses and conversion and it’s going to cost about 100X as much for the hardware.

        I’m looking at a 5 KW array for my house. It’s going to cost about $20K w/out tax incentives.
        Thats $4/watt and it’s probably going to weigh 1,000 lbs.( I haven’t weighed it) or $20/lb.
        Even if space based systems are 10 times lighter. it’s still looking to cost $1,000/lb for transport.

        Unless you can get GEO Launch costs down to $100/lb it’s hardly even worth a discussion.
        Is that possible? in some scenarios, but, in the meantime, terrestrial solar will keep growing and grab the market.

        1. in some scenarios, but, in the meantime, terrestrial solar will keep growing and grab the market.

          Ignoring everything else, this is a fundamental error.

          The energy sector is a sector with no upper bounds. If you doubled how much electricity was available and have the market price slip lower, it would be used. Perhaps not immediately, but it inevitably opens new markets.

          Ignore brute-force manufacturing of useful fuels (octane) and home consumption. Mining has a long, long list of ores that aren’t “economically viable” … because the energy requirements to convert them to something more useful than the rock they started as is prohibitive.

          And then think desalination: the actual limiting reagent to population being potable water. More people leads to … more urea. No problem, we’ll fix it with … more power.

          The field of separation processes also has a ‘as near to infinite as plausible’ list of things to separate using … energy.

          We don’t know what markets exactly would open and be successful with more plentiful energy. But never doubt that there’s an essentially endless stream of things to spend the energy on.

          1. sure, there is a potentially infinite demand.

            but cheap renewables will keep scaling up.

            There won’t be a window for space solar.

          2. But that’s the problem, Dude, renewables don’t scale very well. The output of a terrestrial PV solar (or thermal solar for that matter) installation is mainly a matter of how much area it covers. Double the capacity? Then double the land area. Wind power is fairly analogous. There’s a limit to how close together you can put big wind turbines and doubling output requires putting up twice as many. Fossil and nuclear power plants also scale in roughly this way, but their power concentrations are much higher so a lot less actual land is used. Renewable energy is diffuse and needs a lot of dedicated land on which to host the collection mechanisms. Even ignoring the very real problem of inconstancy, the sheer acreage of land needed to replace even a significant fraction of current baseload generating capacity with so-called renewables is what renders it effectively an impossible project. Those are the real “Limits to Growth.”

            SPS, on the other hand, is intrinsically suitable for baseload requirements because it is constant and its land requirements (rectenna farms) are reasonable. With foreseeable reductions in launch costs, small SPS projects aimed at users who now have to truck or fly in fossil fuels to run portable generator plants in God-forsaken places should be market-fundable in a decade or so. After that, a combination of extraterrestrial resources coming on-line and additonal reductions in launch costs should keep the ball rolling until the economics are eventually competitive with completely terrestrial baseload energy.

        2. There isn’t a single market to grab. Terrestrial solar, for all the reasons previously noted here by others, is intrinsically unsuitable as a baseload power generation technology. In that market, it’s a non-starter. As small-installation adjunct power for single-family homes, apartment or office complexes, terrestrial solar has niche markets it can fit into. In the U.S. Southwest, air conditioning is a major electrical load. Terrestrial solar is, at least if the cost of equipment comes down a bit more, a reasonable way to meet peak air conditioning loads as it produces maximum output when there is maximum air conditioning demand. This use, while not a replacement for baseload generating capacity, can help whatever baseload capacity does exist work with less variability in demand and, thus, more efficiency overall. Even now, however, the economic case for such terrestrial solar deployments seems to be unmakeable absent government financial incentives/subsidies. I do not, personally, think this will always be true, but for now it is.

          Terrestrial solar at least has a potential niche in which it is neither an economic drag on or distortion to the larger economy, though it is some way from achieving this status at present. There is, regretably, no comparable natural niche I can see for terrestrial wind power. The idea that “renewables” are about to take over baseload generation is simply a goofy fantasy of the innumerate Left (but I repeat myself).

          2. If you mean acknowledging that terrestrial PV solar is not a complete waste of space, then you misunderstand either Rand or me – probably both. It isn’t subservience to “dogma” that causes both Rand and me to conclude that terrestrial PV solar is unsuitable as a replacement for fossil-fueled baseload generating technologies, it’s physics. I know it’s an article of faith in modern progressive circles that everything – everything – is subject to the dictates of politics, but King Canute couldn’t really command the waves and you and your fellow fantasists can’t make a baseload silk purse out of a terrestrial PV solar sow’s ear. Sometimes reality really does have a conservative bias.

        3. Unfortunately, “terrestrial stuff” – by which I assue you mean solar – is neither cheap enough yet, nor particularly scalable as your personal experience would seem to underline. If it was, it would already be in widespread use for chopping down the demand peaks of air conditioning load in the U.S. Southwest (and Deep South and Southeast for that matter). Absent government subsidies, it still isn’t economical even for this purpose, but probably will be within a decade. At that point, it will have a solid, but not world-changing niche as a form of peak-load generation. It’s never going to be a viable baseload technology.

          2. If you’re going to post a link, use one that goes to a whole document, not one isolated chart. Are you claiming that this chart shows the impact of terrestrial PV solar installations in California displacing baseload generated electricity during the hottest hours of the day? If so, it doesn’t show it happening, it shows it projected to maybe happen in a number of future years. I have no idea what significance to assign this as I have no idea who put these projections together or what their methodology was. Given the tendency of the environmentalist Left to pull numbers out of the ‘ole anal archive, I don’t assign this chart any greater credibility than I’d accord any of those spiffy “climate models” that purport to show us all frying in our own grease by Memorial Day.

        4. “It’s going to cost about $20K w/out tax incentives.”

          How many years for you to break even?

          1. Without incentives about 12 years. With Incentives, about 7.

            Obviously it depends, upon details of weather, usage.

        5. You’re not accounting everything correctly. To build a baseload power plant using terrestrial solar you need two components: the PV installation and the storage installation. That’s the only way to provide 24/7 power availability, and the storage requirement more or less doubles the cost of the installation, and reduces the efficiency as well. Only during daylight hours do you get to take advantage of the raw efficiency of the PV arrays (minus cloud coverage and sun angle losses, of course), half of the time you have to contend with the PV->battery->power cycle, as well as the capital expenditure and maintenance costs.

          Meanwhile, with SPS you can build a dirt cheap rectenna on potentially multi-use land and get power 24/7 with minimal end-to-end losses. And then in 20 years when your terrestrial PV array and storage system needs heavy maintenance your neighbor with their rectenna grid will just be humming along with no worries.

          1. “you need two components: the PV installation and the storage installation”

            Battery is getting cheap fast.

          2. Battery is getting cheap fast.

            No it isn’t, as a modicum of checking easily demonstrates. The best present-day battery technology in large-scale production is lithium-ion. Automotive battery packs are the biggest “chunks” this technology is manufactured in. Electric car battery packs seem to range from 25 to 85 kWh capacity. Right now, such things cost about $600/kWh. Optimists believe this will fall to $200/kWh by 2020. Pessimists think it will still be $400/kWh by then. Beyond 2020, projected drops in cost appear to slow down.

            For purposes of discussing Li-ion batteries as a storage mechanism to render terrestrial PV solar installations round-the-clock baseload capable, even the rosiest predictions of future cost drops are pretty much irrelevant. Even the best of the predicted numbers is three orders of magnitude (that’s 1,000 times) larger than the installed capital cost of a terrestrial PV solar plant. Most forms of electrical generation have installed capital costs in the range of $75 to $250 per megawatt-hour of capacity. There may be some storage mechanism out there that can at least make a run at transforming terrestrial PV solar into something approximating a real baseload power source, but it ain’t batteries. Not at $200,000/MWh of capacity in capital costs. I’ll just ignore the fact that current and projected production capacity is – and will continue to be – barely adequate to cover the needs of the electric car manufacturers who’ve already got dibs in.

            Once again, you simply don’t seem to understand the difference between a technology suitable for baseload generation and one that is not. Nor do you appear to understand the damned scale of the U.S. – never mind worldwide – baseload power grid. You remind me of one of those primitive tribes that used to turn up occasionally in some remote jungle with words for only four numbers: one, two, three and many. Sorry, not all numbers greater than three are the same.

    2. A modular system would have massive redundancy. If you lose a single module, you only lose a small percentage of the system’s output.

  7. i think something we’re overlooking is that such a system (if it uses conventional solar cells) will have a finite lifespan, and when it’s done what do you do with it? Do you just shift it out a little bit and let the parts slowly clutter up incredibly useful orbits like GEO, or do you expend as much fuel as it took to get it from LEO to GEO to de-orbit it? Given it’s deployed size, could you even safely get it past LEO for re-entry without a collision with a satellite or piece of space junk?

    2. The old silicon solar cells degraded about 3% per year out at GEO. Newer cells are not only much more efficient, they don’t degrade nearly so fast. Since a solar power satellite is too massive to be launched in one piece, it’ll have to be modular. As pieces degrade and fail, they can either be left in place while new pieces make up for the lost capacity or they can be removed and supersynched like they already do for GEO birds at the end of their useful lives. They’ve been supersynching GEO satellites for a long, long time. No one has ever considered deorbiting one because, as you point out, it takes far too much energy. They use the last of the propellant to push it out from the GEO belt and turn it off so it won’t interfere with anyone.

    3. Good reason not to build PV SPS. Big dumb reflectors can be deployed a lot lighter and cheaper. Biggest ones would have power turbines at their foci. At least one smaller reflector should have a resource recovery/recycling plant at its focus fed by a steady stream of electric propulsion cleaner-bots that rendevous, grapple and drag away all the big and medium pieces of junk on orbit. Orbital ISRU as it were.

        1. That depends on your temperature regime. If you’re running really hot, molten salt. If you’re running cool, Helium. Or Ammonia. Or a 50/50 mix of ethylene glycol and water.

          Until someone puts a test up in orbit, the design space is kind of wide open.

        2. Not sure. Water has a lot of heat capacity, but it’s also famously a solvent for just about everything, including plumbing. No noble gases; too leaky. Maybe ammonia? Maybe even some kind of high molecular weight fluorocarbon or fluorosilicon oil that can stand being vaporized at high temp without breaking down and that would recondense at a temperature appreciably above the liquifaction points of the “usual suspect” small molecules? Seems like that might make the radiator problem easier. George? What say you, man? Help me out here.

          1. It would be nice to come up with something that we don’t have to launch from earth. Ammonia could be manufactured from captured carbonaceous asteroids.

          2. I expect initial, relatively small-scale SPS will be launch-from-Earth propositions. At some point, when more cis-lunar infrastructure has developed, it makes sense to expect a transition to ISRU sources for more and more SPS mass, including working fluid. For structure, I’m guessing it’ll be a contest between asteroidal steel and lunar aluminum or titanium.

          3. You might want to stick with a Brayton cycle gas turbine (or an Ericsson cycle, which is a Brayton with heat exchangers providing isothermal compression and expansion, while external heat input and rejection occur at constant pressure) instead of a Rankin cycle, so that your working fluid never condenses. Having to deal with a mixed state flow in zero G is a complexity that you might want to avoid in an early attempt at a working engine.

            A gas-turbine Ericsson cycle (he also invented the ironclad Monitor) presents many advantages over both a conventional or recuperated Brayton cycle.

            Here’s an old Rand report.pdf arguing that we should really switch all our power plants to the Ericsson cycle. In their conclusion they liken it to the way aviation failed to adopt turbofans for a long, long time after their advantages had been pointed out.

            It’s fascinating reading, if you like reading about engine cycles, which fortunately I do. 🙂

  8. That solar power station concept art seem pretty dated. Pre ISS AFAIK. There are more recent studies and designs than that like the sun tower. Then there was that far out proposal from David Criswell to cover the Moon in solar panels manufactured in situ.

    There are also several wacky proposals for space solar power by Japanese conglomerates like Mitsubishi:
    http://www.mitsubishielectric.com/bu/space/rd/solarbird/index.html

    Or Shimizu Corporation’s concept which looks a lot like Criswells:
    http://www.space.com/23810-moon-luna-belt-solar-power-idea.html

  9. beaming power to spaceships

    Ding. ding. ding. We have a winner.

    This makes sense long before we can scale up to anything we would even notice in earthly energy markets.

  10. It’s disappointing that the article seemed to stop around 1974. It doesn’t mention Gerard K. O’Neill, who used SPS as a justification for bootstrapping space colonization, and whose work led to the creation of the L5 Society and the Space Studies Institute. Nor does it mention Colonel M.V. “Coyote” Smith. I posted more comments here.

    1. O’Neill’s application of Space Solar Power to the problem of “What does a civilization is space have to trade with Earth?” is about the only time that it makes sense. The cost of development is decoupled from the cost of launch and there’s no need to make a business case, any more than there’s need for the Amish to make a business case to sell their excess production.

      1. It’s entirely possible that you’re the first person to ever bring up the Amish in a discussion of future commercial space policy. Bask in the glory of it.

  11. Slightly off topic but no one has said anything about the facilities needed to maintain and refurbish Falcon 9Rs. Say they hit their peak production at 40 cores a year, then what? Where and how do they store all those cores in a cost effective manner? Is it as trivial as building giant warehouses?

    I am curious how this will play out operationally. It still may be cheaper than expendables but there has to be some diminishing returns with the increase of infrastructure.

    1. Ideally, you’d meet each returning booster stage with a strongback, load up, refold and latch the legs, return to a horizontal processing facility with maybe a pass through something like a giant car wash on the way in, mate with either new or reused second stage, add payload, deliver to pad, gas and go. The horizontal processing facilities are pretty much the same kind of pre-fab steel buildings used for warehousing anyway, so duplicating or otherwise building them out to any necessary degree wouldn’t appear to present any major difficulties, financial or otherwise.

  12. There are at least three more problems, not often discussed, with using Li-ion batteries as storage for wind and solar power generation. The first is that the life of such batteries is not infinite; they need renewing at fairly short intervals. The second is that large amounts of energy stored in the form of highly reactive metal is rather dangerous, as evidenced by the fairly frequent stories about laptop battery fires. The third is the hardest to get around, however. Where is all the lithium going to come from? Lithium is not all that common, and most of the deposits happen to be in places that are outright hostile to Western interests or politically unstable, or both.

    1. Quite so, sir. The projected life of Li-ion car battery packs is in the 5 – 7 year range. Seems likely to have an unfortunate effect on resale values. Not good for power generation infrastructure either. This is a business that typically plans in terms of decades for the lifespans of large capital-cost items.

        1. For baseload power, Li-ion or most other conventional batteries are inadequate. Molten salt batteries show promise – the hotter they get, the better they are at storing power, and they scale up nicely. Flywheels work, too.

      2. Actually a lot of suggestions have been that used car Li-Ion batteries will make up a large part
        of terrestrial small scale solar systems. While a battery may be inadequate for automotive use, it can be more then adequate for a small terrestrial solar installation.

        I personally think that won’t be the approach, terrestrial will aim for nickel-iron, cheaper materials
        and better chemistry, but, i think the market will figure this out.

        In general, battery is getting cheaper and better.

Comments are closed.