Elon Musk on Why His Rockets Are Faster, Cheaper and Lighter Than What You’ve Seen Before

On Wednesday, I was on a call with Elon Musk, CEO of both Tesla Motors and space exploration company SpaceX. The latter company was announcing what it’s hailing as the “largest single commercial launch deal ever” — a $492 million contract to carry into space the mobile telecommunications satellites of Iridium Communications, beginning in 2015.

Though SpaceX has launched just two three rockets successfully in its history — including sending one of its Falcon 9 rockets into a 155-mile orbit earlier this month and, last July. delivering to orbit a satellite for Malaysian satellite maker ATSB — it has over $2.5 billion in revenue under contract. In addition to its partnership with Iridium, the company has booked orders to carry to orbit a giant satellite for Loral Space & Communications and to conduct 12 “resupply” flights to the International Space Station on behalf of the National Aeronautics and Space Administration.

Whether it manages to successfully complete all its missions remains to be seen, but on the call, Musk spoke at length about why SpaceX is far more cost-efficient than heritage aerospace companies; I thought the detail he offered was pretty interesting. If you’re a space nut, you might too.

The answer to that is complex and I don’t like to give sound bites because it oversimplifies the thing and it sort of comes across as wrong to me, if I were to put myself in the shoes of someone hearing that bite. It’s sort of like when they say: why is Southwest Airlines so much more efficient, and you could say, well, it’s because they use 737s. Well, if it were that easy, then anyone could do that.

If you go to heart of it, SpaceX operates with really with a sort of Silicon Valley operating system and DNA as applied to problem of space transport. And in that sense, culturally, the way things operate are closer to an Intel or Google or Apple in the way that it functions.

We’re quite vertically integrated, which increases the number of problems you have to solve, but you also have fewer middlemen. I think the aerospace guys got into this craze of outsourcing everything, then even the suppliers would outsource, and eventually it got to where you had to go four companies down before you found someone cutting metal. And there’s an old maxim that if you want efficiency, you have to cut out the middleman, and so we’ve done that at SpaceX.

Also, to the degree that we inherited legacy components, we inherited the legacy cost structure, and that necessitated rebuilding almost the whole market in order to achieve significant breakthroughs. Because if you look at the cost of our market, if you break it down, it’s the cost of the engines, the structures, the electronics, the launch operation, and the overhead of the business, and in order to make a significant breakthrough, you have to really see improvements across the board, systemwide.

Like if somebody gave you free engines or free structure or free electronics, you’d still only make maybe a 20 percent improvement in the actual cost of launch; to have a 50 or 60 or 70 percent improvement, you have to see improvement in all of those areas.

So I think we’ve come up with significant technology innovations in all of those areas; we’ve come up with process innovations. We’re vertically integrated but there’s a strong focus internally on communications — everyone is in a cube, including me. We have the factory collocated with engineering, which helps tighten the loop between production and engineering, because however you design a vehicle, that sets the best possible cost number for that vehicle.

Some of what we’ve done is really just common sense — for example, using the same propellant in the upper and lower stages means that operationally, you only need to have one set of fuel tanks. If you can imagine a situation where you have a kerosene first stage, hydrogen upper stage, and solid rocket side boosters, you’ve just tripled your cost right there.

Also, the upper stage of a Falcon 9 is simply a short version of the first stage. That may seem pretty obvious, but nobody else does that. They tend to create upper stage in a totally different way than they create the first stage.

The Merlin engine — we used it on the upper stage of Falcon 9, on the main stage of Falcon 9 and on the first stage of Falcon 1. So we get economies of scale in use of the Merlin engine.

Our tanks are friction stir welded, [aluminum] skin and stringer designed as opposed to machined aluminum, [giving us] a 20 fold advantage in the cost of materials, and our stage ends up being lighter …because geometrically, we can have deeper stringers.

So, yeah, those are some examples.

Related Posts

23 Comments

  • I believe that should be “friction stir welded” a relatively novel technique the yields superior stength at lower cost.

  • [...] SpaceX Falcon 9 Relatively Cheap Compared To NASA’s New Pad June 19th, 2010 06:27 admin Leave a comment Go to comments Hello there! If you are new here, you might want to subscribe to the RSS feed for updates on this topic.Powered by WP Greet Box WordPress PluginAn anonymous reader writes with this excerpt from Motherboard.tv: “As debate over the future of spaceflight rages on — and as the axe all but falls on NASA’s mission back to the moon and beyond — the successful launch of SpaceX’s Falcon 9 two weeks ago proved at least one of the virtues of the private option: it’s a heckuva lot cheaper than government-funded rides to space. In fact, the whole system was built for less than the cost of the service tower that was to be used for NASA’s proposed future spaceflight vehicle (yup, the service tower is finished, but the rocket isn’t, and the whole program may well be canceled anyway).” CEO Elon Musk spoke recently about some of the ways SpaceX finds to cut costs in the construction of their rockets. [...]

  • hey this is solid gold, I am studying economics and this is great stuff to bring to classes.

    After reading this, ot is quite clear that NASA must reinvent themselves or die.

  • Friction stir welding is not exactly new – the solid rocket boosters on the space shuttles used the same method.

  • “If you go to heart of it, SPaceX operates with really with a sort of Silicon Valley operating system and DNA as applied to problem of space transport. And in that sense, culturally, the way things operate are closer to an Intel or Google or Apple in the way that it functions.”

    Geeze. Pure marketing babble. Or poetry, judging by the number of nonspecific references.

  • ah, thanks, phschmidt. fixed.

  • How much can it lift and into what orbit. How many successful launches ( 2 ?). Can it move a payload out of low earth orbit ? I believe the answer is no. Can the design scale up to heavy lift. I think they have another design they are working on with plans to test soon. What are its specs….

    The US space program is dead. All of our eggs now ride on private enterprise. My guess is that Nasa will find a way to so burden them with regulatory baggage and liability the industry will be still born.

  • NASA, and space programs in other nations, solved many problems that SpaceX can now engineer easily. They are going to low Earth orbit. Try landing humans on the moon and bringing them back.

    Re cost saving:

    A man-rated space vehicle (IE, deemed safe for humans to fly) is more expensive that one that will only lift machinery.

    NASA used a materials attribution process in Apollo and may do the same today. In Apollo, they could tell you which load of ore from which mineface in a mine, on which day, that the steel in a bolt came from. That made the bolt more expensive, but it enabled them to trace failures and fix problems if they happened. Each early launch was a big science experiment- they needed to have very strict traceabilty.

    SpaceX may not need to do this- maybe they can use materials such as sheet aluminum for the skin from a cheaper source than NASA would allow, at least for a man-rated vehicle. That certainly would reduce the cost 20X.

    One story in “Angle of Attack” by Mike Gray describes the failure of a titanium tank (on the service module(?) filled with nitrogen tetroxide. The tank worked properly at the manufacturer but blew up when tested at North American in Downey. They eventually worked out that the manufacturer of the Nitrogen Tetroxide was making it purer for North American “because it’s for Apollo”. The discovery: greater than 99% pure NT attacks titanium and caused the tank to fail. Materials science has come a long way because of the space program.

    SpaceX has and will solve real and large problems, but I suspect that they won’t have to go down as many blind avenues as the pioneers. I’m glad customers want to uses their product.

    NASA is sometimes tasked with going to the moon, or Mars, or maybe just LEO. It changes every 4 years. No wonder they waste billions.

    While it may seem that I support NASA 100%, I don’t. But you can’t really compare the two organizations. They have different missions.

  • Not aiming for human flight probably saves a lot. But then, that’s sort of the point – we don’t need humans in space. Want, yes, but it’s more of a fun bonus – com sats are more in the “hard to go without” category these days. NASA approached (and still approaches) it as basic science which is wonderful but very expensive. SpaceX approaches it as a business. The research angle of figuring out what genetic combination for an apple is the most appealing to humans is much different then the business of producing x barrels of apple for $y. Space tech is, in other words, maturing and sliding from high-tech throw the gold-plated kitchen sink at it and even that may not be enough to lower-tech “Git R Done”. Same as computers, nuclear power, cell phones, inkjet printers, etc, etc.

  • Very interesting indeed. It’s funny though – the one comment I expected to see — that SpaceX was saving big bucks by using off the shelf parts — was nowhere to be seen. Musk’s comments about vertical integration was also unexpected. Traditional wisdom, as he points out, is that outsourcing is the way to save money. If SpaceX does indeed get real US hardware into space cheaper than government transports, however they do it, more power to them.

  • Good article, Connie. Actually, SpaceX has launched 3 rockets total ie. 2 falcon 1`s and 1 falcon 9, this last one. Falcon 9 is man rated. It uses the same standards as Ares 1 and and any of Nasa`s man rated rockets. It will be will be safer than those others though as Elon plans to build its launch escape system into the side walls of the Dragon Space Ship which will give astronauts the ability to escape all the way to orbit thus they won`t need to discard to discard the launch escape system, another danger if for some reason you can`t discard it.

  • The Ares was/is basically a solid rocket booster from the shuttle program that has been “man-rated” once you fire it off you’re going, or you’re blowing up, there isn’t much in between although there has been work on a crew escape vehicle. When you consider how much energy there is in one of the shuttle SSRBs I’m not sure I’d want to be sitting on top of it, escape system or not. The Falcon 9 has fully controllable engines that can be throttled off and the vehicle not launched for up to 5 seconds after launch due to the clamps at the base of the rocket. In my mind this is a superior system.

    The fact is that NASA has done, and will continue to do great things. However if we ever want to truly have a space economy where your average person can be somehow involved we need people like Elon Musk and companies like SpaceX to bring costs down.

  • Correction, the Falcon 9 is NOT man rated yet. It was designed to be man rated, but has not completed all the required testing.

  • Falcon 9 was designed to launch SpaceX’s Dragon capsule, which itself is built to haul both cargo and people into LEO. This is a cost savings over even the Soyuz system, since Russia produces the Soyuz for carrying people and Progress for carrying cargo. Both have a lot of commonality but they are still unique vehicles. Dragon will be more like a transport plane in being able to be converted to crew or cargo in a short amount of time.

    RE: SpaceX launches: they actually have launched the Falcon 1 four times, the first two launches were failures for different reasons which were resolved. Due to the high commonality between Falcon 1 and 9, lessons learned on Falcon 1 more easily translate to Falcon 9 reliability.

    I am really looking forward to SpaceX lowering the cost of getting into orbit. This will have NASA a lot of money, money they can focus on actual space exploration beyond earth orbit, and developing heavy lift launch capabilities. Let private enterprise handle the stuff thats already proven and routine.

  • [...] mal in Russland und der Ukraine, heute aber nirgendwo mehr. (Space X Press Release 16.6.2010; peHUB 18., Space News, Space Today, 17., Spaceflight Now, Space.com, Nature Blog, Discovery, Spiegel 16., [...]

  • “NASA, and space programs in other nations, solved many problems that SpaceX can now engineer easily…. SpaceX has and will solve real and large problems, but I suspect that they won’t have to go down as many blind avenues as the pioneers.”

    This is a really important point. As much as NASA can benefit from some of the innovative/entrepreneurial thinking of SpaceX, the latter wouldn’t even be conceivable without the massive investment in institutional knowledge and expertise.

    The real trick for us (humans, that is) will be to figure out how to get these two kinds of entities to intentionally operate in a virtuous cycle. That means evolving past the kind of “Company vs Govt” competition, which is a false idea anyway.

  • Dan,

    You can go to SpaceX’s website and download both the Falcon and Falcon 9 guides – they have tons of information.

    Falcon 9 can put about 4500 kg into geosync orbit (well, geosync transfer orbit). Falcon 9 heavy – a variant with two extra booster stages on the side – would more than triple that, putting it in the same realm as the heaviest Atlas or Delta variants.

  • I hate to be the guy to point this out, but you have an error right in the second paragraph.

    SpaceX has launched 6 rockets. The first 5 were Falcon 1′s. Of those, the first three launches failed. The second two reached orbit, one which carried a dummy payload and the other which carried a paying customer’s satellite. Falcon 9′s first flight, as you know, successfully reached orbit.

  • thanks, roga. someone else also pointed out that i’d missed a successful flight. fixed.

  • Friction stir welding is not exactly new – the solid rocket boosters on the space shuttles used the same method.

    You sure about that? FSW is mainly used on aluminum. The shutle SRB casings are made of maraging steel, which has a much higher melting point.

    Perhaps you meant the shuttle’s external tank?

  • Yes, he probably meant the ET (the manufacture of which began to use FSW in 2001):

    http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20080009619_2008009118.pdf

    Interestingly, FSW looks attractive for welding in space (see link).

  • “How much can it lift and into what orbit.”

    “How many successful launches ( 2 ?).”

    One so far for Falcon 9.

    “Can it move a payload out of low earth orbit ? I believe the answer is no.”

    The answer is very much yes. F1 can do lunar and F9 is a far more capable machine. F9 has a published lift capability to Geo Transfer Orbit of 4680 kg. Falcon 1e (an updated F1 with some engine improvements), set to launch later this year is slated to put 480 kg into GTO.

    “Can the design scale up to heavy lift. I think they have another design they are working on with plans to test soon. What are its specs….”

    Yes, the F9 Heavy will have the highest lift capability of any launcher flying when it is built. It would rank fourth after Saturn V, Saturn INT-21, and Energia for highest lift to LEO with similar comparative capability to GEO and GTO. It would be able to lift 7% more to Geo Transfer Orbit than China’s proposed Long March V moon rocket. However, Ares V and likely whatever replaces it in the US HLV category will outpace it.

    I’ve only heard Musk reference a timeline once, which at that point was 2 years after the standard F9. That would be around 2012. They do have a lot of bookings as well as the Manned Dragon project, however, so other new dev programs may be slowed a bit.

    “The US space program is dead. All of our eggs now ride on private enterprise. My guess is that Nasa will find a way to so burden them with regulatory baggage and liability the industry will be still born.”

    Musk has made it plainly clear since SpaceX’s very founding that he plans to do manned space with or without NASA support or contracts. The only way they could be stopped is via bankruptcy (which given their current manifest is unlikely at least in the next 5 years) or via blanket regulations for flying any human into space (as opposed to restrictions on flying NASA astronauts only). I’m really doubting that will be the case either. I expect regulations that are strict and difficult to meet, but totally possible for any comapany serious about manned spaceflight. There is some liability in that all SpaceX launches are currently to be conducted from US Air Force owned facilities (Vandenburg, Kwajalein, and Kennedy) which means the USAF says whether or not the vehicle is safe to fly.

  • Today my eight year old daughter goes to science camp to launch rockets built during the week. I was so proud to show her videos of the successful SpaceX launch and tell her it was built by entrepreneurs and venture investors. How inspiring for kids to see what the world holds for those with entrepreneurial appetites.

Leave a Reply

PEHUB Community

Join the 12493 members of peHUB to make connections, share your opinion, and follow your favorite authors.

Join the Community

Psst! Got any hot tips?

  • This field is for validation purposes and should be left unchanged.

Look Who’s Tweeting

Reuters VC and PE feed

RSS Feed Widget

Groups