This post is a brief interlude from our normal discussions on virtue, ethics, and community. For a moment, let’s talk about rockets. They’re really big, and really cool, and they’re about to change human history forever.

Check out this Vine:

What you’re looking at is the first stage of a Falcon 9 rocket failing to land perfectly on an autonomous spaceport drone-barge in the Atlantic Ocean. Let me break that down for you and give you a bit of background just in case you don’t follow developments in rocketry.

Who is SpaceX, and a bit of history

SpaceX is an American aerospace company based out of California, founded in 2002 by Elon Musk (previously a co-founder of PayPal) with the stated goal of lowering the cost of access to space, and in the long term, colonizing Mars. They have several future rockets in development, but their current work-horse rocket is the Falcon 9. It’s the first and largest of the Falcon 9’s two stages you see crashing in the above Vine. Since the Falcon 9 was introduced to the market in 2010 it has filled its launch manifest with contracts from both NASA and the private sector, business that came at the expense of existing rocket companies like Boeing and Arianespace, by offering access to Low Earth Orbit (LEO) and Geosynchronous Earth Orbit (GEO) for substantially lower prices than the competition.

The Falcon 9 however  is an “expendable launch vehicle”, or ELV, which means that each rocket launch is a one-use event. The rocket is not recoverable, and a new one is built each time. Needless to say, this is a very expensive way to do business. Imagine what airline tickets would cost if they threw away the plane after each flight.

Of course the problem of expendable launch vehicles being wasteful and expensive has been known for a long time. This was obvious during even the dawn of the rocket age immediately after World War II. There was some work done on reusability back then, but those projects were quickly scrapped and the funding was moved to expendable design. The reasons for this were two-fold: (1) the military needed rockets to deliver nuclear payloads, and those sorts of rockets are a one-way trip. Reusability is a wasted feature on an ICBM. And (2), President Kennedy wanted to reach the Moon on an incredibly ambitious time schedule in order to beat the USSR at something. Expendable rockets are simpler to build and fly, and the technology easier to develop, so they could meet the politically imposed schedule that reusable rockets could not. To the extent the Apollo program intimidated the Russians and helped prevent the Cold War from escalating into a hot one, it was probably worth it, but rocket reusability was lost for a generation.

The next attempt to build a reusable system was the Space Shuttle, and it was a partial success, but mostly a failure. The Space Shuttle Orbiter (the white quasi-plane looking part) was partially reusable, but not rapidly so. It required significant refurbishment between each mission (especially on the heat-absorbing tiles), which was very expensive. Further the solid rocket boosters on each side basically had to be rebuilt each time they were used (a bit like rebuilding the engine of you car between each trip to the grocery store), and the big orange external fuel tank was lost entirely. The end result is that the Space Shuttle flew infrequently and cost between $1 and $2 billion per launch, depending on how you chose to do the accounting. Not very frequent, and not cheap. (By comparison, Falcon 9 may have a smaller capacity than the Shuttle, but a new one can fly every month and for only $65 million per launch).

NASA fiddled around a bit more with reusability in the 90’s by experimenting with the DC-X and VentureStar concepts, but both of these projects were eventually abandoned by NASA despite their technical merits and NASA has not done anything significant in the RLV space since. Ever since that time NASA has (in my opinion) been fatally captured by the special interests in Congress more interested in funneling government money into local-district jobs than reaching space, and the Constellation-come-SLS programs have been a sinkhole of money and engineering talent from which no useful rocket will ever fly.

SpaceX picks up where NASA left off

Now it’s precisely into this morass of despair that mankind will never develop cheap access to space that Elon Musk ventured when founding SpaceX. We know from interviewing engineers who started the company with him that Elon demanded that reusability be built into the Falcon’s Merlin engines from the very beginning, even if the rockets themselves weren’t reusable yet. Elon had learned the lesson of the Space Shuttle – it wasn’t enough for a rocket system to be “sort of reusable”. If access to space was going to be cheap, reusability had to be rapidly reusable, or RRLV. Basically – fly, land, refill the tank, and fly again within a day or so. No six-month refurbishment cycle, no strap-on boosters that need replacing.

SpaceX is advancing quickly towards that goal, and has already made more progress than NASA ever did. As I mentioned earlier, the Falcon 9 only flew for the first time in 2010. The “Grasshopper”, a modified Falcon 9 used in Texas to test reusability, was built only a year later in 2011 and flown between 2012 and 2014. Here’s some videos of those test flights. In 2014, following an ordinary rocket launch (if any of them are ordinary yet), SpaceX caused the Falcon 9’s (since upgraded to v. 1.1) first stage to decelerate and briefly hover over the Atlantic Ocean before landing in the water. The stage wasn’t recovered but it proved deceleration to a controlled landing was possible. And so it was that last November SpaceX revealed its autonomous spaceport drone-ship, a football field-sized drone-ship that a rocket could conceivably land on.

(As a side note, the ship itself is really cool. You think remote-controlled drones that Amazon sells are cool? Or that Google’s self-driving cars are the bees knees? Well this is a drone ship the size of an oil rig that can hold position in stormy seas and a rocket can land on. What a great synthesis of recent innovations in offshore platforms, sensors, and drone technology.)

Now as you saw in the above Vine, the landing was not a success. That’s shame, but I bet the folks at SpaceX are really happy with it anyway. For one thing, this wasn’t an expensive test-flight rocket built just to land on the barge. This was a fully functional (and paid for) Falcon 9 that had already delivered a Dragon cargo spacecraft into its docking orbit with the International Space Station. This flight had paying customers satisfied with the performance; SpaceX is basically having NASA pay for its test flights.

More importantly though, this was only the first attempt to land on an free-floating platform at sea that’s ever been attempted by a vertically-landing rocket (or any kind of rocket, that I know of). It’s amazing to me that the rocket even hit the barge at all. For a rocket coming in from 80 km up and traveling at Mach 10, finding a 100′ by 300′ target in the middle of the Atlantic Ocean and landing on it even as gently as was seen is incredible.

And of course this isn’t where the story of SpaceX’s reusability program ends. The reason this rocket crashed is already known (it ran 10% short of the necessary working fluid in the hydraulic system controlling its wings). A bit more hydraulic fluid will be added to the next attempt (in just three weeks!), and maybe this time the Falcon 9 will stick the landing. Or, as Elon admitted on Twitter today, maybe the rocket will blow up for another reason entirely. But the point is they keep learning and trying. Eventually I have faith they’re going to get this, and the Falcon 9 will become the first rapidly reusable space transport system the world has ever seen – and probably some time this year, as they have twelve flights on their manifest to keep trying with.

But why should anyone besides a rocket nerd care? 

Okay, so possibly some time this year SpaceX may land a rocket, fill the tanks back up, and fly again. What’s this mean in practice to the average Joe?

This is where economics comes in. Currently, a Falcon 9 costs $65 million / flight and puts 28,000 lbs of cargo into Low Earth Orbit; that’s about $2,100 per pound. The Falcon Heavy (the next generation SpaceX rocket to fly for the first time this year, but using mostly proven Falcon 9 technology) should bring the price per pound down to around $1,000. Not bad at all, compared to contemporary and historical examples. But that’s based on the ELV design where the rocket is thrown away. What happens when we reuse the rockets? Well, prices come down – a lot.

Rocket fuel is pretty cheap. According to Elon, the fuel bill is less than 2% of the cost of a launch, or about $20/lb for the Falcon Heavy. If the cost of the rocket can be amortized over 20 flights or more, you’re looking at costs to reach orbit down around $100/lb or less. Assuming a human and all his luggage weighs 500 lbs, that’s a mere $50,000 to reach orbit. Compare that to the $70 million per person that NASA is currently paying Russia to reach the ISS. Quite a savings.

Further, consider that SpaceX is currently building (and throwing away) a new Falcon 9 every month. Imagine if instead of throwing them away, they flew them again the next month (or week (or day)). Within a year SpaceX could have a fleet of rockets making daily trips to orbit. Currently there isn’t even demand for that level of space access, but that’s because the current market is built around paying Boeing $50,000 per lb instead of paying SpaceX $500 per lb or less. When price falls, demand rises.

Once space access is cheap, what happens? Here’s some ideas-

  • You think the Hubble telescope is cool? With cheap space access, private Universities or smaller national programs could put multiple Hubble telescopes in orbit each. And bigger, too.
  • Private robotics teams could afford to send rovers to the Moon, Mars, Venus, or, who knows- Ganymede.
  • CubeSats, already cheap, could become college-project cheap.
  • Bigelow Aerospace could put inflatable space stations in orbit larger than the ISS, for much less money, and lease the space to national space programs and corporations.
  • A private company could send men back to the Moon.
  • Communication and weather satellites could be larger, and more numerous, improving life on Earth in numerous ways.
  • We could mine the asteroids for precious metals such as gold and platinum. These metals aren’t just for jewelry, but have useful industrial purposes too (like being used in fuel cells). One large asteroid can have more platinum-group metals than have been mined in Earth’s entire history. We could see within 10-15 years these metals be demoted from precious to bulk commodity, just as aluminum was a little over a century ago. The consequences of such a price change is hard to predict, but you know engineers and entrepreneurs will come up with something.
  • We could go to Mars.

And those are just the near-term ideas that people are already working on. Ultimately what’s worth noting is that Earth is only a small part of the solar system. There’s more of everything in space. Most of the Sun’s solar energy does not hit Earth. There’s hundreds of times more physical resources available in the asteroid belt than we have ever used on Earth. There’s 200 million cubic km of water just on Ceres, and then there’s the rest of the asteroid belt and all the comets. Habitable real estate is not so common, but you can make it if necessary.

Previous technologies I would compare this to would be the advances in navigation and ship design that propelled European explorers, settlers, and colonizers to every corner of the globe in the 1500-1700s, the invention of the rail roads connecting the American continent and the Russian frontiers, and the advent of container shipping which really globalized the world economy. Each of these inventions made what had been previously unreachable, available and cheap. The European settlement of the Americas essentially quintupled the physical resources available to Western civilization, a fact that we are still in the process of developing. The containerization of trade (along with a steadily improving regulatory climate in East Asia) made Japan, Taiwan, South Korea and China into economic dynamos. And so giving humanity access to 1000x more energy and physical resources that it previously had will have … unknown effects. But you can bet they’ll be big. Get ready. We’re about to see an expansion of human economic activity that only comes along every few centuries. It’s going to be awesome.

10 thoughts on “Rockets!

  1. Pingback: Rockets | Transterrestrial Musings

  2. Kelly Starks

    Ok, a few factual errors. Yes the shuttles were labor hogs, but their cost per ton to orbit for NASA IS LOWER THEN ON THE SPACEX LAUNCHES. Both the total cost per, and the margin cost per. Yes the total cost per launch of the SpaceX craft is about 1/3rd that of the shuttle ($440M vr $1.2b [not $2b]), but the shuttles lift 4 times as much tonage, and a crew of 7.

    Second. SpaceX flight operate under SAR contracting rules, rather the the FAR rules the shuttles are operated under which quadruple costs ( largely due to buracratic expenses). The company (Boeing L/M joint operation) that operated the shuttle fleet repeatedly offered operate the shuttles for 1/4th the cost if allowed to operate under same SAR. Or about $300m a flight per shuttle, vrs $440m per Falcon/Dragon.

    Third it’s historically generally cheaper to build a reusable system then a expendable.

    Dirty little secret of space flight, is that it’s not the physics or cost of flying it, that make the flight costs. It’s all the overhead and upfront costs over the nearly nil flight rate. All the cost of developing and owning aircraft, with about a millionth the fleet flight rate. I.e. No economies of scale.

      1. Kelly Starks

        CBO report. See the CRS budget is $1.6, but that excludes the NASA funding to develop and support the craft to be flown in CRS. SpaceX supporters like to ignore those upfront and overhead costs. Of course such costs dominate all space flight costs, so it gives a very squeezed vision.

      2. While were discussing this

        NASA 2012 report ”“; report only included $3.5
        billion in launch fees to Orbital and SpaceX in the COTS/CRS program, not the rest of the COTS/CRS costs for
        the flights including the R&D and overhead. The full COTS/CRS program costs which as Page 1 of
        shows was projected at $6.4B through 2016. Your $3.5B in fixed-price contracts for flights through 2016 might
        not all be paid for by 2016 (or be from a different account) but should fit in that $6.4B. SpaceX got more money
        for development and other upfront costs from the $5.4 as I outlined in the posting I mentioned that you didn’t
        want to look up. (See below)
        In May 2011 Congress issued a report listing $850 million (more then the total dev cost of both Falcons and
        Dragon, according to SpaceX)
        congressional document
        “Commercial Cargo Will Cost More Than Shuttle-Delivered Cargo Says Congressional Document”
        SpaceX got
        $278 million for three demonstration flights of the Falcon 9 launch vehicle and Dragon capsule,
        $258 million in milestone payments for completing 18 of 22 COTS milestones. Please see Appendix 1 for
        SpaceX’s schedule milestone chart.
        The COTS CRS refs
        1 of 3 7/23/14, 8:44 PM
        $185.6 million for milestones tied to four CRS missions
        $128 million toward additional risk reductionǁ‖ milestones
        Or SpaceX had receaved $850 for COTS as of May 11 (NASA only spent $1.25B in total COTS!!)
        CCDev 2 awards
        $75 million as part of to develop a revolutionary launch escape system and SuperDraco engine ( ;
        Now SpaceX’s press release 5-14-12 ( on
        page 3 states “..To date, SpaceX has received $381 million for completing 37 out of 40 milestones..” Which is
        another $132 million since the above congressional report said them were paid $258m for 18 of 20 COTS
        milesstones. I.E. about $132m for the Dragon to ISS test flights.
        said “paid SpaceX $381 million in an agreement to help pay for the design, development, and testing of the
        Falcon 9 rocket and Dragon spacecraft.. SpaceX has spent $1.2 billion to date, including public and
        private capital.”
        They seem to be just refering to the COTS milestone payments, not all the other fees / Awards paid to spaceX
        the development adn testing of the Falcon’s & Dragon’s, but whats interesting is the $1.2B number.
        Totaling up the above, SpaceX has received $1.057B total from NASA as of May 2012
        Musk said total dev costs for Falcon and dragon were $800 M and the SpaceFlight now article says “SpaceX
        spent $1.2 billion to date”, and Musks quoted as having invested $100m of his money in SpaceX. That implies
        hes only gotten about a $100M in investor money?
        They might actually have gotten a lot less then that given they must have gotten some money in advance fees
        from the launch contracts they got (Bellow) which I can’t sort out.
        So is this all SpaceX has gotten from the feds? Itcertainly doesn’t sound commercial!
        Am I missing something?

    1. Mike E

      Ok, so shockingly, if you amortize Shuttle development costs over 100 flights and Falcon over 12 Shuttle is (barely) cheaper. Another dozen flights and Falcon is already clearly cheaper – looking through those docs, the marginal costs per flight on Shuttle at peak flight rate was about $1B.

      1. Kelly Starks

        Not quite.
        Granted amortorizing the development over a tenth as many flights gives the shuttle a edge. But those development shuttle costs only come to about $300m per flight at this point (SpaceXs comes to about a third that much currently – again a little higher per ton).

        the shuttles margin cost per flight is only about $50m per flight – not a $1B.

        Your ignoring the elephant in the room, that shuttle is forced to operate under FAR contract rules, rather then the SAR rules SpaceX is given. Operating shuttle under SAR rules, the companies could operate the the fleet for lower absolute cost per flight then SpaceX can fly Falcon/Dragon’s, while carrying 4 times as much cargo, and more then a Dragon 2’s flight of crew. (The maned Dragon obviously requiring a while new development program. SpaceX is already 6 months behind in their engineering of those craft.)

  3. Mike, Kelly likes to inflate launch costs for Falcon 9, by including everything that is not launch, that NASA also buys. In cargo missions to ISS, NASA insists on a new Dragon every time, instead of reusing them. He includes all development of Dragon that was done under Space Act Agreement work, no matter that those were paying for faster development than SpaceX and others would have done on their own.

    His statement that there were offers to run shuttle outside FAR Cost+ contracts may even be true. He ignores in this that NASA’s overlords in the Appropriations Committee view space launch as a way to get money spent in Florida, Alabama, Utah and Texas. They want *more* money spent on each program they have already paid for, politically, with favors to other members of Congress. That’s why the SLS/Orion coalition came into being. It’s why SpaceX has been attacked by that coalition at every opportunity. There was never a hope that those pols would let shuttle pay less.

    Then, he also speaks as if there is no price/demand curve for spaceflight as well. He assumes that *only*and*always* the present customers will be the future customers, whenever talking about any non-government space vehicle. In short, his experience inside spaceflight over the last 40 years are the future as well. Of course, if Musk agreed with that, he never would have started SpaceX as a way to settle Mars, and the rest of the inner Solar System.

    1. > …. NASA’s overlords in the Appropriations Committee view
      > space launch as a way to get money spent in Florida, Alabama,
      > Utah and Texas. They want *more* money spent on each
      > program they have already paid for, politically, with favors to
      > other members of Congress. ….

      Partly true, but more self serving given you blame congress and a couple states. The program was designed to fund something in as many districts as possible, given its VOTERS who must profit, given thats where 90% of voter support comes from.

      >..Then, he also speaks as if there is no price/demand curve
      > for spaceflight as well….

      Not at all, I’ve frequently stated the opposite, but in this case (and others when there wee small temporary price drops,) it failed to generate any market growth. Market analysis suggests You really would need a ten fold increase to see much growth. Since we’re in the most extreme part demand of the demand curve.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.