On Starship Prices

The world has finally seen a launch system that can find out what the price-demand elasticity curve looks like. Elon wants to maximize flight rate and revenue, because he wants to drive costs down to make Mars more affordable.

63 thoughts on “On Starship Prices”

    1. He has always said other people need to step up and that he will just solve a few of the problems.

      More stuff like this on the horizon? https://spacenews.com/vast-space-intro/

      “Vast’s innovations will serve the role of a research platform, which is what the ISS did historically,” McCaleb said. “But we also want to be a machine shop where national and private sector astronauts can iterate and prototype things in orbit. Ultimately, our contributions will enable something akin to a way station for human habitation that orbits the moon – maybe even Mars.”

        1. Probably not. Lathes tend to produce a lot of shavings. That wouldn’t be cool on 0G, especially if you’re turning metal. It’s unlikely they’d send a wood lathe into space. Metal shavings would cause all sorts of problems unless you can create a foolproof containment system. Besides, additive manufacturing would be much more efficient for space applications.

          1. Don’t make NASA’s mistake and confuse efficiency with efficacy. Modern CNC already uses an enclosed cabinet to keep fluid and chips inside, and there could be a simple way to adapt that to zero-g. If Starship does its job, then hauling bulk stock to orbit may become affordable, and additive manufacturing still has quite a way to go to create really strong materials.

          2. For many things, objects built in space don’t need to be as strong as the same object built on earth then launched into space. The objects built in space won’t have to survive the same acceleration and vibration loads as terrestrially built objects. Reducing mass in structure pays huge dividends in reducing propellant mass needed for interplanetary missions.

            Lathes and CNC machines employ reductive manufacturing. You put in a large mass of material and remove that which isn’t needed for the final design, similar to how a sculptor starts with a large stone and removes everything that doesn’t look like the final sculpture. The problem with reductive manufacturing in space is that you have to launch more mass than necessary. Even with Starship potentially radically reducing the cost per kg to orbit, there’s no need to be wasteful. Additive manufacturing is far less wasteful, and it allows for greater flexibility in design than can be produced by reductive manufacturing.

          3. I know people in the additive manufacturing business. It is perfectly possible to make 3-D printed metal parts that are stronger than comparable parts machined from billet stock.

            Then there’s the matter of NASA having recently developed a metal-ceramic composite technology that allows manufacture of parts for rough service in very high temperature regimes – say, rocket or aircraft engines – that have eye-wateringly better creep resistance at very high operating temperatures than anything now in service.

            Said technology pretty much requires 3-D metal printing as the fabrication technology, though there seems to be some possibility of using it for hot isostatic pressing (HIP) applications as well. HIP is not exactly an additive manufacturing technology, but it is a near-net-shape technology so, in terms of materials usage, it has more in common with additive manufacturing than with conventional stock removal techniques.

            Additive manufacturing is certainly not a universally applicable technology terrestrially. But it will have far wider cost-effective application in space because of its relative minginess anent material mass and its very low wastage of same.

  1. This does assume that it is going to hit its’ price point, flight rate, and predicted cargo capacity. And in a timely manner. IMO, going for the whole size, stainless, methane, and full reusability in one shot carries a hubris that can get damned expensive.

    I believe a smaller precursor vehicle would have been a good development insurance policy as well as a medium lift replacement for Falcon9 as the technology matures. Finding all the gotchas in the largest LV in history might get expensive in time, money, and capability.

    I think it’s going to work, but take longer and cost more than most believe.

    1. I keep wondering whether a “chopped” (shorter) version of Starship as the Lunar Starship in Artemis might be a reasonable design, given the vast difference in cargo capacity between SpaceX’s current design and the not-yet-and-maybe-never-selected competing designs for an Artemis lander.

          1. That may happen at some point but I don’t thing anyone else currently in the business thinks as big as Musk.

        1. Say it costs 5 times more than SpaceX second stage and you make 10 of them- so it’s only 5 times the cost.
          Whether you do it, will depend on the Starship test flights. And you going use it for LEO. Or doesn’t have re-entry from the Moon or Mars.
          Though once those profiles are tested- maybe it works for them.
          Or Musk has been changing his design, and probably will continue to change it, as things move forward. That aspect of it, favors using stainless steel.
          Or instead Carbon fiber, how about using Titanium?

          1. Titanium is not as insanely more expensive than stainless steel as is carbon fiber, but it’s still quite a bit more expensive, especially recently. It has a lot of fabrication issues not present with stainless. It is not as high-temperature resistant as stainless steel.

            Finally, the most prolific producer of titanium is Russia. So not selecting titanium as Starship’s basic material has allowed SpaceX to avoid availability and price issues that both would have developed at a very inconvenient point. Even without a crystal ball anent the Russo-Ukraine War, however, Elon is always keenly aware of sourcing issues for materials he uses a lot of, going all the way back to the mines. Elon was never going to entrust any part of his enterprises’ essential supply chains to Russians after his original experience with them that led to the formation of SpaceX in the first place.

        1. NASA way of making rockets has many problems.
          It builds the Space Shuttle with general idea that
          it will be privatized- and it never happens.
          And if you are private, one doing this all the time- or there is no moments of transitions, there are just different business deals.

    2. In some sense the current vehicle is the “smaller precursor”, in that the original design was 12 meter (downsized to 9 meter). Still quite the leap from the 3.7 meter Falcon 9.

      If Starship/Superheavy works as planned, perhaps they will take another look at larger vehicles, 12 meter or 18 meter. Sea-platform launch would be a virtual necessity for the latter I guess.

    3. There’s a difference between hubris and ambition. I have noticed that many people associated with OldSpace seem to have no grasp of that distinction while also embodying considerable hubris themselves in terms of a seeming belief in the impossibility of doing much better than they have done over the last five or six decades.

      SpaceX has a particular highest-priority application in mind for its next-generation spacecraft. As such, it is not unreasonable to go as directly as reasonably possible for that goal.

      In any case, it’s not as though SpaceX has entirely eschewed smaller-scale experiments before getting down to full-scale brass tacks. The ancestors of the current Raptor engines were initially fabricated and tested at sub-scale until enough confidence was obtained to go to full-scale designs. And the engines were, by far, the riskiest part of the project and where the most advance over prior art was being sought.

  2. Well, any answers to the question posed in the link? Are any science missions prepared to take advantage of Starship and Super Heavy? The scale of what can be done leaves a lot of room for creativity but even just doing rideshares opens up a lot of possibilities.

    Every college should be launching something and why not high schools too?

    1. My guess is there may be some early discussions about it but no one is ready to devote serious effort and money towards designing payloads until Starship is proven successful.

  3. At $10/kg, serious space colonization becomes a realistic option. Rotating habitats included. Is anyone doing serious design work on updated O’Neill Cylinders, or Bishop Rings?

  4. If you ever saw the Stephen King miniseries Rose Red, you might remember the character Emory, who, when he saw ghosts, closed his eyes and repeated “Not there, not there, not there.” And that’s exactly what just about everyone is doing with Starship. Still planning around scarce, expensive EELV-class launchers. It will be at least 2030 before we see science missions that take advantage of Starship’s capabilities. Even if there are scientists and engineers out there who are interested, the suits are telling them to get real and get back to work and shave another gram off that instrument.

    1. I’ve seen mentions (possibly here) that Musk and astronomers such as David Rubin have been kicking around the idea of building a space telescope into the structure of a modified Starship upper stage.

        1. Rand, if you are replying to me (impossible to tell with this system), I certainly hope you are right. Forty years of SSDD makes me pessimistic.

  5. $10 might be a reasonable capital cost fraction, but if even a tiny fraction of people are willing to pay one million to spend a week in orbit that puts the vale delivered per flight up in the hundreds of millions. So to get to the low price point one would have to ramp demand. But there are likely a major multiple of people who would be willing to pay $100,000. If you can pack 853 like an airbus, that’s still generating $85 million per flight. And one can see how lower prices still keep the value in the $40+ or $20+ million range. Unless there’s some kind of miraculous ramp up like 50,000 us planes per year from the US in WW2, it may take years and entry by several daily manufacturers to get to $10/lb much less $10/kg. If the price gets down to $12,000 a seat they might attract suborbital traffic to replace fast military deliveries and long haul aircraft. NASA might buy 100 2nd stages a year to send on one way trips if the capital costs drop enough. So maybe 20-30 years to $10/kg retail.

    A nice problem to have vs. not enough demand at the cost point.

    1. That chart claims the Saturn V launch cost (per unit of mass) to be only twice the Falcon 9, and certainly 1/10 that of the SLS/Shuttle.

      Was the Saturn V really that inexpensive, or is this the unadjusted-for-inflation price/cost in 1969?

      1. Those numbers look correct – in general, if you are throwing things away you want the largest possible, lowest tech vehicle. Saturn V was a good match.

        1. 1. Saturn V with its hydrogen-fueled upper stages was not that low tech.

          2. If those prices are for-real in inflation-adjusted terms, developing the Shuttle had a simply enormous opportunity cost in terms of what a production line cranking out Saturn V’s could have done over the same historical period? True, it is twice the cost of Falcon, but Falcon wasn’t a thing until very recently.

    1. How tall? Huge. Compare: he volume of the LH2 tank in the Shuttle’st External Tank was about 2.7 times the volume of the LOX tank; the volume of the LCH4 tank in Starship is only about three quarters of the volume of the LOX tank. True, H2/O2 has a higher specific impulse (about 30 percent, IIRC), but I suspect that engines for the colder hydrogen would be more complicated and more massive, and the additional thermal insulation would be a problem, too. Besides, LH2 is a bitch to use. The folks at SpaceX made the right choice.

      1. My impression is that the upper stages of Saturn were H2 because of the greater exhaust velocity. Greatly simplified, a rocket becomes less efficient when its speed is larger than the exhaust.
        OTOH, it’s possible that SpaceX looked at all the problems associated with H2 and making it reusable and found it was cheaper overall just to make it bigger.

        1. There are many advantages of using Methane- it’s the cheapest rocket fuel available on Earth. And when you plan to launch cheap enough that the cost of rocket is significant factor in launch costs.
          Plus it has it higher temperature of boiling- which is significant issue if want rocket fuel when get to Mars.
          In terms of Moon, the moon could have enough frozen CO2, and that could make Methane a cheaper rocket fuel than LH2. But if LH2 is more expensive on the Moon, it could better to pay more and use LH2.
          The Moon doesn’t have Earth’s problem with LH2.
          On earth and particular for the first stage, you need a lot thrust. On the Moon, you have a lot less gravity loss as compared to Earth, and less than Mars. And lower gravity, also mean need less thrust.

  6. Currently Starship is heavy on engine and plumbing and light on everything else. Just like the original Stephenson Rocket locomotive. As it should be. But I wonder how hard it would be to adapt Starship to hold essentially what amounts to the interior of the Dragon capsule into the design? Seems to me mix and match would be the quickest way to get from here to there.

  7. What about an expendable i.e. non-returning third stage? It could be a simple as a cylinder totally encased in Starship acting as its fairing and then ejected out the front via the nose. Your orbital workshop and who knows what else? I wonder if lacking a nose cone is really a problem for the belly flop maneuver if the empty pipe is vented? Calling Kelly Johnson….

        1. That’s the most popular conception, and that makes sense if your final destination is back on Earth. But to me it seems a waste to put all that plumbing and engine tech permanently out of commission on a one-way trip for a permanently orbiting laboratory. Basically you are putting yourself back on the expendable curve by doing so. The one big difference is that this rocket is being mass produced, as such once unit costs dip below a certain point maybe it IS desirable to just bite the bullet and park the extraneous hardware. You could certainly save something by dispensing with the TPS. But somehow I don’t think the economics work. A precedent was WWII. At the end of the war we had lots of surplus transport aircraft, etc, but I don’t recall people generally sending transports out on one-way missions. I’m sure there were a few, but not many, mostly to be used as museum pieces I suspect. Starship is reusable for a reason. But if the cargo need not return, nor return intact, then you implement the Pez dispenser or equivalent third-stage technology.

  8. Starship, but… Aldrin Cyclers!
    Starship, but… nuclear rockets!
    Starship, but… too big!
    Starship, but… different!

    It does seem like we’ve found a new use for the phrase, bitter clingers!

    And it’s funny that only the Chinese are reacting directly to the impending reality of Starship. Maybe the Russians would too, if circumstances were different. The rest? Europe responds by subsidising the national buggywhip industry. US OldSpace by saying, “Nonsense! Cars require imaginary physics. But a buggywhip now, that’ll get the job done!” Maybe India? Funny how Starship is beinging accepted by three of the letters in BRICS. And Brazil has an equatorial launch site. That’s four.

    1. Starship, but… Aldrin Cyclers!
      No, Starship, and Aldrin Cyclers — additional life support, living volume, and radiation storm cellar that accompanies a “Mars Colonial Transport” Starship but doesn’t have to land and doesn’t require fuel (except for occasional orbit corrections) once established. It can even be a Starship, or several Starships. Think of it as an ocean liner that never docks, with passengers that board and disembark via yachts at ports of call — a much more comfortable transatlantic voyage than you’d get if you had only a yacht.

  9. $10/kg vs $58,000/kg. Quite a difference.

    But say that neither camp can deliver. That the true cost rises by …500%. Which is how NASA seems to work these days….

    Why, that would be $50/kg, vs $290,000/kg.

    Why are we wasting air on SLS?

    1. The worst case, expendable SuperHeavy/Starkicker, would be around 3% the cost of SLS on a per-launch basis (no amortization of developments costs, just write them off). Best case is so bizarre most people are convinced it can’t happen. And yet.

    1. Most people would have a hard time figuring out how to drive an original Model-T. First off, you have to figure out what the middle pedal of the three does and how far to depress it. Also, the right most pedal doesn’t work the way you think it does. Try the nearest “turn signal” instead. I figure it’s probably millennial theft-proof. Especially since you have to crank it, but don’t forget to switch on the magneto first.

  10. Walt Disney had his head frozen, so some say. James Doohan really did have his creamted ashes scattered from orbit, to late burn (futher?) up during re-entry. What would Starship allow?

    Imagine at death the body contributes a DNA sample — not much more than the mosquito in amber from the Jurassic Park movie. Preserve it, shield it from radiation, make a capsule, wrap the capsule with micro-engraved text of the cell sample’s donor’s biography — life, accomplishments, contributions to humanity. So, capsule with DNA sample weighs maybe an ounce?

    Build a “Voyager” style rocket powered probe intended to loop around and around from Earth orbit to Jupiter to parts unknown. Fully fueled, about 800 Kg. Give half total over to the DNA and history payload. So 400 Kg or about 14,000 1 ounce memorial capsules that can reach orbit. Probe and all orbits for $8000 in chump change. But you charge the hopeful romantic grieving heirs $1000 apiece to put Grandpa Walt’s amber-DNA on the way to the stars. You have $14 million to work with, can you build and control a Voyager-like probe for maybe $10 million and put $4 million profit in the bank?

  11. I worked for Bellcore in the ’80s. We were about to roll out ISBN digital phone access, and we struggled to understand what kind of a market there would be for it. Would enough companies see the value to them? Then the internet came along.
    I guess we’ll see.

  12. As Elon never falls for the sunk cost fallacy, I’d say what he’s doing is close to what he currently thinks is optimum. Size of the vehicle doesn’t matter. Complexity does and a smaller vehicle is going to be just as complex (and likely have a MUCH lower payload). Besides there are scale advantages to making rockets bigger.
    Making Starship out of carbon probably doesn’t save any mass. What you care about is mass of heatshield + structure. Carbon will need a heavier heatshield and be much harder to fabricate.

    1. Carbon fiber is as heat resistant as steel, about as strong as steel and about a quarter of the mass of steel.
      It’s harder to fabricate.
      And can’t recycle it.

      Second stage Starship is has dry mass of 100 tons,
      so one could probably reduce it by about 50 ton.
      Or increase payload by about 50 tons.

      1. Most carbon composites have polymer matrixes and are nowhere near as heat-resistant as steel. Carbon-carbon fiber composities have superior heat resistance, but would be insanely expensive to fabricate at the scale required for Starship.

  13. “Carbon fiber is as heat resistant as steel”
    The carbon might be but the resin isn’t. Add in the mass of the resin, then the thicker heat shield tiles to keep the carbon fiber – resin cool enough. Elon talked about this.
    Carbon also has a repair/damage detection problem, particularly on Mars.
    Elon also doesn’t care about efficiency and elegance except when it comes to minimizing cost per ton to orbit. Choosing stainless steel was an inspired decision. I’d love to know how that came about in detail.

Leave a Reply

Your email address will not be published. Required fields are marked *