≡ Menu

Project Icarus: Finding the Fuel

Project Icarus, introduced to the IAA at last year’s Aosta conference, made quite a splash yesterday at the International Astronautical Congress in Prague, with four presentations by Icarus team members and related work on the FOCAL mission by Claudio Maccone. Icarus is the attempt to re-examine the Project Daedalus starship study of the 1970s in light of technological developments in the intervening years. A joint project between the British Interplanetary Society and the Tau Zero Foundation, it’s now in fully operational mode.

Fueling Up a Starship

There is much in these papers worth comment, but today I’ll home in on the issue of helium-3 and where to find it, presented yesterday and drawing on the work of Andreas Hein, Andreas Tziolas and Adam Crowl. Daedalus was envisioned as a fusion mission using deuterium and helium-3 as fuel, a reaction that has advantages over deuterium/tritium but one that has yet to be demonstrated in a working reactor. Assuming we do figure out how to light this reaction, we still have problems in that helium-3 has a very low abundance on Earth, which is why Project Daedalus’ planners came up with the idea of entering Jupiter’s atmosphere to collect it.

The idea is audacious, involving huge factories moving through Jupiter’s clouds and orbital ferries that would carry the precious cargo to Callisto, outside the worst of the planet’s radiation belts. Moreover, 128 such factories deployed over 20 years were needed to mine the needed 30,000 tons of helium-3. The Icarus designers note both the scale and the attendant problems of the idea, although uses of helium-3 fusion in Earth-based power generation and planetary defense could come into play to offer additional incentives to acquire helium-3 resources.

Image: The Daedalus starship at second stage firing. Credit: Adrian Mann.

But what about Jupiter itself? The strong radiation belts here rule it out for the Icarus team, which also eliminated Saturn because of its own radiation problems and high escape velocity. Uranus emerges as the planet of choice, with returning the processed helium-3 to orbit presenting perhaps the biggest technical hurdle. Double-walled hot-air balloons are chosen to keep the processing unit stable in the atmosphere, with a single-stage nuclear-thermal rocket emerging as the best solution for return to orbit. But the Icarus study is young, and the team is also considering tether concepts for atmospheric mining that do not involve braking and descent into the atmosphere at all.

The Outgoing Nuclear Option

Safety and political concerns would make a nuclear-thermal rocket unlikely for the outbound leg from Earth to Uranus, which the team thinks better handled in any case by nuclear electric propulsion, with aerodynamic braking or aerocapture used for orbital injection upon arrival. And we can’t rule out interesting solar sail possibilities. However, NEP remains the most attractive option to date. The paper on which Hein’s presentation was based lays out the benefits analysis and notes the high cost of gathering helium-3. What to do? Let me quote from the paper:

…we have to consider the consequences of the results for space exploration and an interstellar probe using D-He3 propulsion. For the annual rate of 1500 tonnes of He3 required for the Daedalus probe, it is reasonable to assume a civilization which has a much higher need for He3 for energy generation or interplanetary transportation. Taking current energy growth estimates, this will take many centuries. A possible solution to this problem would be a drastic reduction of the payload mass of the probe. A reduction by two orders of magnitudes would require an annual mining rate of 15 tonnes. Nevertheless, the course of the Icarus study will show, whether this kind of reduction of the probe size is feasible.

Thus we run into energy growth estimates again (see the discussion of Marc Millis’ paper on same in relation to the first launch of an interstellar probe). Reducing the payload mass helps, but we’re also looking to a civilization that has established a clear need for helium-3. The Daedalus design, of course, ran into constraints like this, but it falls to Icarus to take them to their logical conclusion, even as we note again that deuterium/helium-3 fusion has yet to be demonstrated. Must we wait for other fusion concepts and, possibly, a Uranus space elevator to supply the technologies that would fuel Icarus, pushing back its launch by centuries?

First Estimates and Their Uses

A search of the literature shows that deuterium/helium-3 fusion has been analyzed in terms of inertial confinement, electrostatic confinement and magnetic confinement, but the higher energy and temperature requirements are obstacles that seem to favor deuterium/tritium fusion in the short run. Hein examined whether deuterium/helium-3 could be competitive with the more researched deuterium/tritium alternative, coming to this conclusion:

A rough estimate shows that the maximum order of magnitude cost of Helium 3 must be between 1 – 10 Billion$ per ton, in order to be competitive. This is a very important result, because it will limit the allowed cost per ton of He3 obtained in our mining operation, if fusion reactor economics is the dominant factor.

The technical and commercial feasibility of deuterium/helium-3 remains in question, because Hein’s numbers exceed the maximum cost. But if a workable fusion propulsion technology using helium-3 becomes available, then its uses within the Solar System would be clear, especially in terms of dealing with dangerous objects like Earth-crossing asteroids. A planetary defense grid could well take advantage of this kind of fusion, allowing the delivery of fast payloads to these objects in response to perceived threats. This coupled with possible Earth-based reactors could impel society in the helium-3 direction and would represent the best scenario for lowering costs.

Image: Andreas Hein giving his presentation at IAC 2010. Credit: Pat Galea.

If that scenario plays out, helium-3 mining could become a long-term strategic investment in technology and security. A system delivering helium-3 to the L2 Earth-Sun Lagrange point could thus generate political and financial support. All of this could reduce the cost of helium-3 to within acceptable limits, making deep space missions like Icarus more likely. How likely is all this to occur? Reports like these deliver first estimates and are not meant to be definitive, but they do sketch out the extraordinary challenges implicit in the idea of putting together a near-term interstellar mission.


Comments on this entry are closed.

  • MattC September 28, 2010, 12:44

    If large fusion reactors were made practical could not h-3 be manufactured easier then it could be mined?

  • Tulse September 28, 2010, 12:46

    Is D-He3 that much better than D-T? Or, if one wants aneutronic fusion, p-B11? Surely the problems with these alternatives are easier to solve than setting up industrial gas mining operations at Uranus.

  • John Hunt September 28, 2010, 14:52

    128 factories by Uranus?

    Sorry guys, but are you really sure that beamed propulsion isn’t a better idea than fusion? And if beamed propulsion is more economic then why is so much human resources being put into a reworking Daedalus?

  • JoeP September 28, 2010, 15:14

    I’m skeptical that mining on a scale like that is ever feasible. Sure, maybe in a millennium or two, assuming the solar system is ever fully inhabited by space-faring, interplanetary, and very industrious humans.

    Better yet — isn’t there plenty of helium-3 locked up in lunar soil? Much easier to process on the moon with automated robots, and its so close — and the gravity well is nothing difficult to deal with. I’d say that is in the realm of current capabilities (if not cost).

    Even better still is to get a little more funding and focus on the Polywell concept. Ten million bucks is probably all they need to determine if that is the way to go. Cheap.

  • Pat Galea September 28, 2010, 16:48

    Guys, remember that we are still in the trade study phase. We are looking at lots of options. No decisions have been taken yet on what systems will actually be used in the Icarus design. (Other than that it will be mainly fusion based propulsion.)

    So during this phase you will hear a lot from us that may not necessarily sound immediately optimal. If we later decide to keep any of these concepts in the design, then we will have to justify them. But at this stage we are just exploring the options.

  • Robotbeat September 28, 2010, 20:45

    I think you’re getting ahead of yourself by deciding on fusion…. well, at least for a mission to Alpha Centauri.

    Mining from gas giants isn’t feasible. Mining on Earth is.

    Interstellar missions will not happen for centuries if they require:
    1) giant lasers with ENORMOUS structures for focusing light (can’t use sparse array, remember!)–or accelerators (assumes solar system-spanning infrastructure, thus solar system colonization)
    2) antimatter (too inefficient to produce, too volatile!)
    3) mining hundreds of thousands of tons from gas giants (assumes infrastructure of colonization of the whole solar system)

    A fission fragment rocket is completely feasible for interstellar propulsion if combined with an interstellar brake and if operating at .1-.15c or under and if targeting something like Alpha Centauri. The exhaust velocity for a fission fragment rocket is still within an order of magnitude of the acceleration-phase delta-v, so you aren’t getting crushed by the exponential rocket equation, just gently slapped. Fusion rockets may improve the exhaust velocity slightly, but now you have to mine the gas giants!

    We can mine plenty (thousands of tons) of fissionable fuel like thorium or uranium well within current budgets. It may be expensive to make it fissionable by feeding it into breeder reactors, but not impossible. If you can design your rocket to breed the fuel in situ, even better.

    If a manned mission lasting 40 years can be fit into a 100 ton spacecraft, then even a manned mission is feasible without assuming the colonization of practically the whole solar system first.

    Just my two cents. Sorry about my fission fragment rocket hobby horse.

    Even more important is that interstellar braking (perhaps combined with stellar wind braking as you approach the destination) is too useful to ignore, even if Zubrin’s approach is wrong.

  • Michael September 29, 2010, 0:47

    Could you not mine Jupiter of He-3 at the poles then super heat it until it is a plasma and send it via its magentic field to be collected in the radiation belts, no need for transporters just a large scoop to capture it from the radiation belts

  • JohnHunt September 29, 2010, 1:03

    > giant lasers with ENORMOUS structures for focusing light (can’t use sparse array, remember!)–or accelerators (assumes solar system-spanning infrastructure, thus solar system colonization)

    You’re probably thinking of something like the old StarWisp concept. But Jordan Kare’s SailBeam would accelerate miniature sails to 0.1c in 3.5 seconds. This would only require about 1/4 the distance to the Moon.

  • ljk September 29, 2010, 8:08

    Unless some technological miracle or cosmological find happens in the next two decades, our best bet to reach the stars in just a few human lifetimes while we still exist is by using Orion, which was proven in the 1960s but then abandonded like so many other promising space projects (such as NERVA).

    Orion doesn’t need an elaborate fueling station at Jupiter or Uranus, which will be nearly as arduous an undertaking as interstellar travel in its own right. And as Carl Sagan once said in Cosmos about Orion, what a great nonlethal use for numerous nuclear bombs.

    I get the feeling that thirty years from now there will be yet another paper study of the successor to Daedalus and Icarus. Have to think of a new Greek mythological name too.

  • Ronald September 29, 2010, 9:35

    People mention the moon as an alternative source of He-3. However, it seems that the density of He-3 in the moon regolith is quite low even in the best places, maybe even making it hardly or not feasible from the point of view of EROEI (energy return on energy investment), after all you have to heat the stuff to quite high temps, almost baking it.
    I read somewhere that even natural gas on earth contains an interesting amount of He-3, but I have no idea whether that would ever be a feasible concentration. Anybody else?
    John Hunt mentions beamed propulsion as *an alternative*, however, this would also require enormous amounts of energy. Surely solar energy would do, but the required collection area would be enormous, so maybe fusion would still be more feasible even in that scenario.
    Honestly and with all due respect, a Daedalus/Orion sized approach seems almost oldfashioned in this age of miniaturization, at least for an early stage of interstellar.
    I would rather expect something like a giant laser installation (‘leave the fuel and engine behind’), either on the moon or free floating in space, spitting out large numbers of very small interstellar probes (payload in the order of 10 kg). Braking would be virtually impossible but the large number would make this redundant and the probes cheap and disposable. Not all your eggs in one basket, once you have the (very expensive) installation, the variable cost per probe is almost negligible in comparison. Of course the decisive factor in this would be the required sail area.

  • Eniac September 29, 2010, 15:36

    To avoid both a reactor technology we don’t even know is feasible and mining operations at Uranus, the most attractive option for fusion appears to be tritium bred from lithium. Yes, there will be neutrons, but they will be captured and recycled for the breeding (which will generate some He3). The lithium blanket does not really contribute to the dry weight of the ship, because it is fuel. In the end, if all of the energy is made to emerge at the tail end, no matter what form, the efficiency cannot be all that much worse than with He3.

    Sails won’t work, I think, because they would inevitably be eroded by the relativistic impact of the ISM. Miniaturization, not, either, for the same reason.

  • Ronald September 29, 2010, 18:36

    @Eniac: ‘Sails won’t work, I think, because they would inevitably be eroded by the relativistic impact of the ISM.’
    Well, this problem may be reduced by accellerating the ship fast enough by means of laser, that is during the first fraction of the journey, after which the sail is redundant.

  • ljk September 29, 2010, 21:09

    Ronald said:

    “Honestly and with all due respect, a Daedalus/Orion sized approach seems almost old fashioned in this age of miniaturization, at least for an early stage of interstellar.”

    So something old equals no good, eh? :^) Well, Orion has a few things over these new-fangled fancy-dancy concepts everyone keeps going on about, like, oh, having a propulsion method that actually exists and we have and can build more of. And has been tested (the test object hangs in the Smithsonian Air and Space Museum).

    Plus Orion does not require the complex space infrastructure that most of these other concepts require. We have not sent a single human past Earth orbit since 1972 and the only space station at present has a crew of six which spends most of their time repairing it. NASA has abandoned its manned lunar colony plans and no one else is even close to replacing Constellation. And the promise of putting humans on Mars has been pushed all the way out to the 2040s at the earliest.

    And you expect there to be a space infrastructure in our Sol system with lots of working gigawatt laser beams just floating about (think having a nuclear bomb in space is touchy?) and factories floating in the atmosphere of Jupiter and Uranus pumping out Helium 3 for a power source that hasn’t even been developed on Earth, let alone in a space vessel that is supposed to travel all the way to Alpha Centauri.

    And didn’t we already discuss whether a bunch of tiny star probes could survive a relatively fast trip between suns what with the radiation and debris impacts impeding them – unless they were housed in a large well-shielded craft. Hmmm.

    And as for warp drives and wormholes – where’s Scotty the Miracle Worker when we need him?

    I know I sound terribly pessimistic but what I am trying to be is pragmatic. And it frustrates the heck out of me that while some focus on concepts that won’t happen for many decades or even centuries and require technologies and infrastructures that also won’t be ready for ages, we have an interstellar vessel idea that could be built NOW (or four decades ago) and get us to the nearest stars in just a few human lifetimes.

    Yeah, Orion is Old School. But it is real and it works and we could have it if we wanted it. In our lifetimes.

  • kzb September 30, 2010, 7:32

    Robotbeat, thanks for the mention of fission fragment rockets -there is actually a Wikipedia page for those interested. I had thought that this was going to be a non-starter (except of course for the rather primitive Orion concept), because of the requirement for a critical mass.

    However I now see there are credible advanced design concepts for this, which could result in exhaust velocities up to about 5% c.

    I think the main barrier to development would be political -testing the prototypes on Earth would result in horribly contaminated test chambers and the dangers of fission product release to the environment. Also, I wonder at what distance from Earth it would be deemed acceptable to turn on such a motor?

  • ljk October 1, 2010, 0:16

    China’s western deserts would be a great place to test Orion as hardly anybody lives there. Plus it would not surprise me a bit if they were the ones to take up the challenge while everyone else continued to debate and hem and haw.

    They already have one of the key players in a major part of any interstellar probe, the AI. They’ve given this guy millions to conduct his research while he was essentially ignored in just about every other place on Earth:


  • Tom D October 1, 2010, 18:38


    The old Orion concept is cool, but it’s unlikely to be efficient enough for “practical” interstellar propulsion. It would be very interesting for use in our solar system, but it just doesn’t give you as much propulsive energy per kilogram as fusion or antimatter. It’s just too heavy to get to Alpha Centauri any time soon. Of course a practical fusion rocket doesn’t exist yet, but that shouldn’t stop us from speculating :-).

    Atomic Rocket is a fun site to check out for the details of many types of propulsion:

  • Kilroy October 2, 2010, 22:16

    Guys/Gals… shouldn’t we master our neighborhood first (Moon, Mars, etc…) before jumping out to the deep water? I like space exploration but until we can control our own little corner of the solar system I dare say we should stay here!

  • kelvin October 3, 2010, 19:04

    Interesting comments guys. I echo the point made by Pat Galea that the Project Icarus Study Group is just looking into options for He3 and we would be remiss if we didn’t at least examine the gas giant mining options. All fuels and all sources of fuel are an open option currently. If we didn’t look at the gas giants to see what the numbers say, then we would be critisized for that. So we have to do our homework so we can at least defend any design decisions we do make later on which won’t be until 2011 at the earliest.
    Regards fusion propulsion, this decision was really made for three reasons (1) linearage and evolution with Daedalus (2) it is a good propulsion candidate for interstellar performance, although one of several (3) the propulsion element is a big issue and we decided to constrain the design team so that we could progress this particular option. It is not neccessarily the view of the Study Group that this is the best option, although it happens to be my personal view that it is right up their, along with other schemes like Project Orion.
    It is quite likely that in another 30 years time there will be a successor to Project Icarus and this is the philosophy behind the Tau Zero Foundation, “ad astra incrementis”. Long term thinking is what is required to progress interstellar research, especially in an unfunded climate where we are all doing this in our own personal time, self-funded, and sacrificing time with our families. Dedication to the cause of a positive future for humans expanding into space is what is required if we are to be successful.

  • ljk October 4, 2010, 5:15

    Kilroy said on October 2, 2010 at 22:16 :

    “Guys/Gals… shouldn’t we master our neighborhood first (Moon, Mars, etc…) before jumping out to the deep water? I like space exploration but until we can control our own little corner of the solar system I dare say we should stay here!”

    Somehow this reminds me of the comments made by the people who think we need to solve all human problems on Earth first before we can do anything else.

    Might as well have never left the caves and trees with that kind of thinking.

  • ljk October 5, 2010, 22:57


    Noteworthy new reports from the Congressional Research Service include the following (all pdf).

    “The Helium-3 Shortage: Supply, Demand, and Options for Congress,” September 21, 2010.


  • ljk October 6, 2010, 1:40

    Next Big Future looks at the latest on various proposed interstellar spacecraft propulsion systems here:


  • Paul D. October 13, 2010, 17:04

    If there is an undiscovered Earth-sized, or even Mars-sized, planet in or beyond the Edgeworth-Kuiper belt, it could be cool enough to retain helium in its atmosphere. Such a body might be the best place in the solar system to mine large amounts of 3He, assuming we had the fusion rockets needed to reach that far out in reasonable time.