why we won’t build a dyson sphere soon

April 5, 2012

See, this is what I get for having my head buried in code for days on end; interesting stuff just slips by without notice even when it’s begging for a post. Though in this case, circling back to the story now will actually add a fun twist. Basically, my former Skeptically Speaking sparring partner, George Dvorsky, wrote about making an honest to goodness Dyson sphere out of Mercury and attracted the attention of PopSci and a blogger at Forbes who dialed up the Bad Astronomer himself, Phil Plait, to do the math on whether this was or wasn’t a viable idea. Not surprisingly, their answer was no, but the odd thing is that the reasons why Dvorsky proposed an unworkable plan weren’t included in their calculations. Instead, they worked out that it would take so much energy to disassemble Mercury, that a 100% efficient Dyson shell of satellites would take us some 174 years to balance the energy budget. Now this would’ve been fine if we were talking about warp drives and negative energy/mass constructs, but we’re not, and even after having it pointed out that Dvorsky was proposing a very energy amortized bootstrapping scenario, Alex Knapp was still sticking to his energy balancing guns. So if we don’t need to balance energy the same way we need to balance a checkbook, what’s really wrong here?

First and foremost, the problem is that assumed 100% efficiency. As pointed out in a previous post in which I crunched the numbers on Dyson spheres, the amount of collected energy lost in transmission would be so high that the whole project may not even be worth it. When you do the math on beaming solar energy from the comfort of Low Earth Orbit, you find that you’d need a receiver stretching from the Earth to Saturn to get real returns on your investment due to the limitations on how powerful your beam can be and the diffusion of the photons in your microwave laser. If you really want to amp up the power, you’ll create a death ray and the swift release of energy will take down the grid for which you’re aiming, overwhelming your ability to deal with excess electricity now surging through the wires. Collecting energy isn’t like drilling for oil. You can’t put a lot of energy in a barrel and store it for future use for years on end in capacitor banks. This is why trying to work with efforts required to build a Dyson sphere and treating Joules like we’d treat capital investment and measure the rates of return in terawatt hours doesn’t accomplish much more than throwing big numbers around. The big point is that we’ll lose the vast majority of the energy being beamed back and without another incredibly precise mega engineering project, we can’t even beam that much power back without essentially zapping ourselves with an immensely powerful ray which will be illegal under the same treaty discussed in this morning’s post.

The second big problem? Cost. When we discussed putting up a sunshade to curtail global warming with some real numbers, we found that the optimistic cost for blocking out 1% of the Sun’s energy with a reflective solar sail comes to some $800 trillion. The global GPP hovers around $65 trillion, so in other words, blocking out or absorbing 1% of solar energy in space will cost an order of magnitude more than literally all the money in the world. Now, we can’t just multiply a bootstrapped project by 100 and declare that has what we could call an astronomical price tag of $8 quadrillion and call it a day. Certainly if we start small and take the recursively building nature of this endeavor, we’d be looking at budgets… in the tens of trillions? How? Why? Because all those initial robots and mining colonies won’t build themselves for free and will need human assistance for a whole host of issues since in space, a lot of things can go really, really wrong. This is why we lose missions; they break far away from humans who can fix them. Would we really want to start the absolute biggest project in all of human history by at least six or seven orders of magnitude to then be willing to lose it all when things go wrong and we can’t come and fix them? But it will all be worth it if we amortize this over many years and all the energy we’ll get will more than make up for our economic sacrifices, right? If you said yes, allow me to go back to the paragraph above and reiterate that we will not be beaming all that much power back.

Who would want to commit to a project that would last many decades, require tens of trillions of dollars to just get off the ground, and then be unable to beam back the monstrous energy it would be designed to collect for global consumption? And this isn’t even going into the politics of building this monstrosity and who would pull for what percentage of the energy generated and agree to find what part of all this. If we want access to clean, plentiful, cheap energy, we need to heavily invest in very efficient solar panels here on Earth and keep working away on fusion and its exotic power-generating applications. That’s what will grant us more energy than we’d even need, and those are projects with extremely high payoffs and to which we’re even closer than the robotic miners and energy collectors Dvorsky and his sources say we’re on the verge of building. Dyson spheres are great in science fiction but in practice, they would fail to do anything but barely sustain a species with no other seemingly viable options to survive and the money we’d have to invest in building one would be far, far better spent on at least a few thousand different projects that would do far more to secure our energy needs for the foreseeable future and beyond, while also making us a highly durable, long-living, space faring species, and it’s those projects in which we should be putting our money, time, and no pun intended, our energy.

[ illustration by Khang Le ]

Share
  • Bruce Coulson

    Is a Dyson Sphere (leaving aside engineering, scientific, and political concern in building one) even necessary? (For purposes of counter-acting global warming.)

    “On December 20, 2009, Myhrvold appeared on CNN’s Fareed Zakaria GPS and discussed his patented idea to eliminate global warming/climate change using geoengineering. It involves using hoses suspended from helium balloons 25 kilometers (16 mi) above the Earth. The hoses would be placed near the North Pole and the South Pole and emit sulfur dioxide, which is known to scatter light. Myhrvold estimated that such a configuration could “easily dim the sun by one percent, and even do it in a way that wouldn’t be visible.”[24]”

  • HikerTom

    “….emit sulfur dioxide…”? I believe that is an component of acid rain. What would Mr. Mynrvoid have us do, trade one problem for another? Like the Dyson sphere idea in the post, it seems we should try to minimize the harm we cause in correcting other problems.

  • http://lukeparrish.rationalsites.com Luke Parrish

    Knapp was sticking to his guns on a major miscalculation, and one of the commenters called him on it. 100% of the sun’s energy is 4E26 Joules per second. You need 2E30 to overcome Mercury’s gravitational binding energy. Even dividing by three you can still do it in a matter of hours.

    Dvorsky’s plan involves ten years of exponentially reproducing solar collectors occupying the space near Mercury. If you grant the premise of self-replicating robots, there is really nothing fundamentally wrong with the scenario.

    As for it being anything like what is necessary to fix the earth’s energy shortage, well of course it is way overkill. And to get the desired amount of energy to the earth (which would be an insignificant percentage of even just the the Mercury Swarm phase) you’d probably have to do something like set up a huge mass driver that catches high speed particles sent from the collectors and generates current by slowing them down.

  • http://lukeparrish.rationalsites.com Luke Parrish

    After first demolishing the strawman of spending quadrillions directly on solar collectors (um, how?), you then go on to address the actual premise which is indeed bootstrapping:

    “Certainly if we start small and take the recursively building nature of this endeavor, we’d be looking at budgets… in the tens of trillions? How? Why? Because all those initial robots and mining colonies won’t build themselves for free and will need human assistance for a whole host of issues since in space, a lot of things can go really, really wrong. This is why we lose missions; they break far away from humans who can fix them.”

    There are two premises Dvorsky relies on: Sufficiently good autonomous robots and advanced materials tech. He is optimistic that these will become a reality in 20 years or less. If you dispute these premises, fine — just say so. Now, if you have autonomous bots that are capable of self replicating (whether individually or as a colony), you can afford to lose a few if need be, as long as they aren’t the minimum starter colony. You would just need to be sure they replicate rapidly enough to make up for the loss.

    So no, we don’t have to fund a human colony on Mercury to make this happen. Given the premise of individual self replication (even if this is inefficient compared to the replication of large scale colonies of specialized machines), a single “generalist” rover landed on Mercury in a metal-rich area with a small area of solar paneling and some basic welding and mining equipment could do it. NASA is apparently interested in doing something similar on the moon (which in turn gives us both experience and, potentially, a massive energy budget to invest in a similar project for Mercury).

  • Greg Fish

    you’d probably have to do something like set up a huge mass driver that catches high speed particles sent from the collectors and generates current by slowing them down.

    Wow… That’s an idea beyond bold. I assume a future society with no concept of safety would be willing to try it, but certainly not one in any foreseeable future…

    “So no, we don’t have to fund a human colony on Mercury to make this happen.”

    I didn’t say anything about a human colony on Mercury, just that humans would need to be able to to get to the planet if we hit a major snag. Robots are great but I assure you, generalist bots are still very much the stuff of science fiction and they can both break and make mistakes.

  • http://lukeparrish.rationalsites.com Luke Parrish

    Wow… That’s an idea beyond bold. I assume a future society with no concept of safety would be willing to try it, but certainly not one in any foreseeable future…

    Particle streams may not be the best choice, but something along the lines of kinetic energy applied to mass. 0.1kg slugs of iron would be small enough for safety. The moment a meteor that size hits the atmosphere, it compresses the air in front of it which generates enough heat to vaporize it. (Larger meteors greater than 2-3cm across are dangerous because their surface area to volume ratio is lower, preventing the heat from conducting to the center fast enough.)

    I didn’t say anything about a human colony on Mercury, just that humans would need to be able to to get to the planet if we hit a major snag. Robots are great but I assure you, generalist bots are still very much the stuff of science fiction and they can both break and make mistakes.

    Assuming you are correct that sufficiently good robots won’t exist in 20 years to do the job without human aid, that is a good argument to start with the moon. We can probably start on that more quickly anyway. And once the moon has large numbers of solar arrays and factories on it, sending missions (whether manned or not) to Mercury from the moon should be much less expensive.