28 thoughts on “Commercial Space In 2014”

  1. Is Falcon Heavy expected to -not- fly in 2014? Being left off of that overview…. Do we have any news?

    Because as much enthusiasm as I have for all the other newspace stuff, FH seems like a key element for so many useful paths. From immediate impromptu ‘depot’ through the dramatic changes to the list “Things -requiring- heavier lift.”

    2. The article focused on suborbital so maybe using the term commercial was a bit too broad. For Falcon Heavy I expect this to be the case,

      “”Rule number one of projects is, it will take longer and cost more than you planned…doubly so for advanced aerospace vehicle projects,” he said.”

  2. Al,
    I dunno, I’ve always felt that FH was far less important than F9R. I’d much rather live in a world with a fully-reusable F9R than one where SpaceX can launch really big payloads in one piece.

    ~Jon

    1. I think the goal is a fully reusable heavy, which would be the best of both worlds. In fact, a reusable heavy should be easier, at least for the outer cores, with their lower staging velocity.

      1. Even an affordable “heavy” lift would be good.. but government is a terrible customer to bank on. Both the Air Force (which buys the launches for the other agencies) and NASA are more likely to make you jump through hoops than pay you deposits. SpaceX was right to go after commercial markets first.. then COTS came along.

        1. “but government is a terrible customer to bank on.”

          Trillions in government contracts cant be wrong. The government makes a great customer if you have the skills and resources to deal with all the crap that comes with it.

          There is a key difference between SpaceX and ULA. ULA isn’t in the launch industry, they are in the government services industry. SpaceX is in the launch industry.

  3. I suspect Heavy falcon is related to Falcon-9 reusable first stage.
    So one would want to be able to recover the 27 rocket engines of Falcon heavy.

  4. The last I heard of the FH design was use of cross-fed stages based on F9 first stage. But what if SpaceX plans for FH are really different? Three clues I pay attention to regarding FH: Grasshopper flight test successes, SpaceX dropping out of Stratolaunch because of claimed production disruption, Raptor engine change to methane propellant and described high thrust.

    What if the FH uses standard production F9R lower stages instead of unique cross-fed stages ? Combined with a much larger Raptor engined upper-stage to make up the difference in FH performance for full recovery of all the lower stages?

    1. You lose a lot of performance if you don’t cross feed, and cross feeding makes it much easier to recover the outer cores, due to the lower staging velocity. Ability to cross feed isn’t that big a deal in terms of stage design.

  5. Did you notice Sir Richard Branson did not mention anything about commercial flights this year in his quotes on yesterday’s flights? All he did was state his prediction they would make it to space again, 10 years after the X-Prize…

  6. The focus on the, thus far, non-existent suborbital tourism business is indeed odd as the major upcoming story anent commercial space for 2014 is clearly SpaceX and its LEO/GEO orbital efforts.

    Depending upon which source one trusts best, SpaceX appears to have at least a dozen flight operations to orbit scheduled in addition to the two Dragonrider abort tests, which are also commercial, but not intended to reach orbit. Half or more of these missions are to LEO orbits, giving SpaceX the potential of seven or eight more opportunities in 2014 to recover the F9 v1.1 first stages employed. This includes all of the four scheduled 2014 Dragon cargo flights to ISS.

    SpaceX has already shown the ability to operate at close to the projected once-a-month ops pace from its Canaveral AFS pad. Some have called the 34-day pad turnaround between SES-8 and Thaicomm 6 launches a record, but it is not. Arianespace managed sub-30-day pad turnarounds at Kourou once each year from 2008 through 2012. One can fairly note, also, however, that Arianespace has never launched more than 7 Ariane 5’s in a single year. If SpaceX can put up even eight more Falcon 9’s this year, they’ll take the commercial annual launch record for a single booster family.

    I was a bit surprised to find that the Russians never launched more than 8 commercial Protons in a single year, though their annual launch record total for this booster was 14 in 2000. They have managed 10 – 12 total launches of this booster in each of the 5 years preceding 2013. Roughly half of these launches were commercial comsat shots for foreign customers.

    The Russians are less constrained than other commercial launch providers as they have more than one Proton launch pad in service. SpaceX also has two extant pads, but no more than two of their missions for 2014 seem to be scheduled out of Vandenberg AFB. So the Canaveral pad is still on most of SpaceX’s critical mission paths.

    As an indication of which way the commercial space winds now blow, the most significant event of 2014 may already have occurred when Sky Perfect JCSat signed with SpaceX to put up one of their GEO comsats in 2015 just four days after SpaceX’s Thaicomm 6 success. Sky Perfect JCSat has previously been a user of both the Russian ILS and Arianespace launch services. Their last lost payload was riding an ILS Proton. Their last three successful launches were on Ariane 5’s. Yet here they are knock-knock-knockin’ on Hawthorne’s door.

    SpaceX can clearly now land all the satellite launch business it can handle. The main question that I think the remainder of 2014 will see answered is, “Exactly how much business can you handle, there, Elon?” I have a feeling the answer is not going to make anyone in Russia or France very happy.

    The unhappiness is likely to spread to certain domestic precincts as well, given that the Thaicomm 6 mission will, in all likelihood, be certified as being the third consecutive success required by the DoD to qualify SpaceX as a bidder for USAF/DoD/NRO payload launches. Affirmative word on this will probably come down sometime in 2Q 2014.

    Now that SpaceX has significant non-NASA revenue coming in, it is likely to finish 2014 nicely in the black. In the event Mr. Musk should need more capital to accelerate any one of his many parallel efforts underway at SpaceX, he can easily borrow against his personal holdings in Tesla and Solar City, the value of which have roughly quadrupled his wealth in 2013. He’s not quite Jeff Bezos-rich yet, but he’s definitely getting there.

    SpaceX is well and truly out of the weeds. Now it’s a matter, mainly, of seeing how well they can service a very ambitious 2014 launch manifest and which of the incumbent launch services providers takes the biggest hit to its own future book of business.

    I follow the suborbital tourist companies too and wish both Virgin and XCOR all the best in 2014. I’d be delighted if both were in revenue service by year’s end. But that’s probably not the way to bet the rent. SpaceX is the commercial space story for 2014.

    1. That was a wonderful piece of writing, better than the linked article. 🙂

      You mention that barrier of turning over a pad in less than a month, which limits the flight rates. Several months ago I commented quite a bit on a system of launching rockets while laying on their side, since once you get a flight vehicle in the air the orientation is just a matter of doing what all flight vehicles do, which is change their orientation freely and frequently, while on the ground long short things are cheap to service and tall thin things are a nightmare, and at some point touch labor will dominate the pricing.

      To bump SpaceX out of the lead will require a cheaper engine design and a cheaper configuration, but my suggested cheaper configuration can’t work without a much cheaper engine design. If you have a much cheaper engine design in your head, along with a cheaper configuration, the market looks like a big swimming pool when you’re on top of the high dive, getting the shakes, walking back and forth and wondering if you should just climb back down the ladder.

      1. Thanks for the kind words, George. I doubt any of the rest of us who comment here will ever erase our “Wow! Neat comment!” balance of payments deficit with you, but it feels good to know I’ve put a little something on account.

        I do indeed recall the sideways launch discussion. Design/trade studies to assess its technical and cost merits relative to SpaceX’s fairly conventional Falcon approach lie well outside my competence even at the back-of-an-envelope level, but I’d love to see the idea pursued.

        My feeling, though, is that SpaceX’s horizontal final assembly on a transporter/erector/launcher (TEL) followed by a “stand-up” at the pad is plenty good enough and gives a ground infrastructure advantage of considerable magnitude to purely liquid-fueled architectures such as Falcon. The main reason is that empty liquid-fueled boosters simply don’t weight much compared to vehicles with fully or partially solid motor elements. This allows use of pretty much off-the-shelf hydraulic systems developed for heavy construction and/or mining equipment in the fabrication of the TEL.

        There are limits to everything, of course. Vertical assembly and crawler/transporters were probably the right call for Saturn V. Tilting up something large-diameter and 363 feet tall is way beyond tilting up a Falcon 9, even the new stretched v1.1. For something like SLS, with most of its empty weight in a pair of 5-segment SRB’s, the TEL/tilt-up approach is completely out of the question.

        This is one good reason, by the way, for supposing Ariancespace is going to find that its proposed all-solid-lower-stages notion for the Ariane 6 is not going to yield SpaceX-like economies in practice. Vertical assembly and transport will be even more mandatory for the notional Ariane 6 than for the extant Ariane 5. Arianespace may be able to repurpose the Ariane 5 vertical assembly infrastructure for Ariane 6, but some elements may need non-trivial beefing up too.

        In the context of reducing pad turnaround times, it’s also worth noting that TEL’s like SpaceX’s (and Orbital’s for Antares) are nimble and quick compared to the Shuttle/SLS creeping crawlers and their smaller counterpart in Kourou. I have a feeling that SpaceX can fairly easily figure out ways to take many additional days out of their recently set 34-day mark. Even with the limits of vertical assembly and a slow crawl to the pad, Arianespace’s best effort to-date was 25 days. I suspect SpaceX can handily beat this by building a second horizontal assembly building and another TEL at Canaveral so that two vehicles can be in different stages of prep at once.

        I concur that beating SpaceX’s cost structure would be best accomplished by use of a fundamentally cheaper engine design. With respect to the extant Merlin and in-the-works Raptor that would seem to mean doing without the expensive and temperamental turbo machinery. I have no idea what you’ve got in mind here, but if such a project was in my hands, I’d probably be talking to XCOR about scaled-up versions of their piston pumps. I’d probably build a combined power/pump cylinder block in which some of the cylinders constitute an automotive-type internal combustion drive engine running on the RP-1 or liquid methane fuel and vaporized LOX and the rest are used to move fuel and LOX from tanks to combustion chamber.

      2. Regarding a simpler pump: Take a small LH2/LOX thruster with an exhaust velocity of 4000 m/sec burning 1 kg/sec of propellants, and stick it in a valley next to a tall mountain with its exhaust aimed upwards. Then arrange a bunch of hoses around the engine’s nozzle and spray 39 kg/sec of propellant into the exhaust gases (which is just supersonic steam). Conservation of momentum dictates that the final velocity of the steam and propellant will equilibrate at 100 m/sec, and will travel upwards in an arc to land in a white oak barrel 1600 feet up on the side of the mountain. At the bottom of the bucket is a copper pipe which runs back down to the valley, and the flow out of the pipe is 40 kg/sec with 1600 feet of pressure head.

        That’s my thought experiment to show it’s theoretically possible to build a high volume, high-pressure pump with no moving parts. As an added bonus, if your propellant is cryogenic, the steam freezes into ice and gets injected into your combustion chamber along with the propellant, so you get a staged combustion cycle by default.

        Oddly, I went through a bunch of thermodynamic analysis of the idea, along with shrinking the pump into a small lightweight package, and was planning to test it with water sprayed into the exhaust of a small solid fuel hobby motor, and only then thought of the bucket on the mountain analogy to answer the engineers who kept saying “It’s impossible to build a pump with no moving parts.” ^_^

        But the solid motors have a very short run time and I didn’t think I could get reliable steady-state data with them, so I started doodling on a simple liquid fueled thruster I could build, and that led to the idea of making the stainless steel chamber liner into a conveyor belt so it only gets heated for tens of milliseconds before it goes back through a cooling bath, giving me a duty cycle. I wrote a bunch of transient heat transfer code to explore the idea, and frankly you could build an engine with a chamber liner made out of aluminum foil if you don’t push the chamber pressure up too high. If both concepts pan out, engines will get really, really cheap.

        Anyway, as an aside, by serendipity there have been a lot of studies dating back to the 60’s of spraying propellant into a supersonic gas stream (including droplet trajectory modeling) for scramjet projects.

        1. A computer “Heat Pump” is pumping the heat via the motion of the boiling and condensing liquid inside. No actual pump required – simple concrete example is just an evacuated and sealed cylinder with “the right amount” of water.

          Very easy to reimagine that same pump in formats that gets the fluid flowing inside actual tubing, Tesla’s one way valve with no moving parts also helpful to thinking about this.

          There’s also no mechanical pump in my coffee maker. And yet, it pumps.

        2. Yes, but bubble lift pumps can are generally only good for a foot or less of head pressure, and I need a couple thousand feet.

          Interestingly, the bubble lift pump shows up as one method to move water in dynamic aquaria with algal turf scrubbers. One of the issues they identified was that our conventional high-speed centrifugal aquarium pumps generate enormous shear forces. Algae can survive the trip through the pump, but diatoms and tiny little algae eaters can’t, making it impossible to establish a natural food chain unless using a completely different pumping method. At one point I was diddling with making a radial-engine piston pump out of toilet plungers, but I’ve also thought of using a large gear pump.

          To me, a turbopump is just an extremely roundabout way of accelerating a fluid to high velocities. You have a burner that runs off-mixture to keep the temperature down, which drives a set of turbines, which are connected to a shaft, which has to be sealed between the exhaust gases and the propellant, and the shaft turns an impeller which accelerates the fluid radially, and then the fluid dumps into a volute so it slows back down to turn the kinetic energy into pressure. We use it because Von Braun needed a pump for his rocket and walked down to the local fire chief and asked how fire truck pumps work, since they obviously produce high volume, high-pressure flows.

          My idea, ironically, is that if you need to pump high volume high pressure propellant for a rocket engine, just use a smaller rocket engine to blast the propellant to the required velocity for pressure recovery.

          Strangely enough, there’s a related application in engine testing. If you moved some distance down from the nozzle of a rocket engine under test and sprayed water into a tube, the water would accelerate to the velocity determined by the water/exhaust mass ratio and exhaust velocity. You can then convert that velocity into potential energy (such as shooting it upward, though that’s not what I would do). Once the water is under pressure it can drive a conventional large water turbine which are about 90% efficient at turning pressure head into mechanical rotation. It wouldn’t affect the engine test because it all occurs downstream of a supersonic flow.

          If your rocket engine is about 40% efficient at turning chemical energy into kinetic energy, and conservation of momentum turns that into potential energy as fluid pressure or height, then you can turn that into mechanical energy at about 36% efficiency, which is competitive with a lot of gas turbines and more efficient than most coal fired power plants. In a simple case you just blast water from the bottom of a dam back up into its reservoir. You sell the electricity on the grid and the fuel costs of a long series of engine tests is essentially paid for by the electricity you sell.

          My other engine test idea is that if you always test two engines at once, aimed in opposite directions, and link their controllers, they would push against each other instead of the giant steel monstrosity you have to build to withstand the thrust of an unopposed rocket engine.

          So your test stand construction costs plummet, and your fuel is largely paid for by selling electricity, so you can test the heck out of your engines while remaining financially solvent. I’d ask why the government never thought to do this, but the question answers itself.

          To me this is all kind of interesting. The big government approach is building a giant rocket re-using engines designed in the 60’s and 70’s, based on work done in the 40’s and 50’s, and it’s costing a fortune. There are entirely different ways to approach the same design problems that might be much simpler, much cheaper, and just as efficient, but they weren’t the ideas we started with and ran with in our rush for ballistic missiles and the space race.

          However, as a caveat, thinking through all the geometry of my rolling combustion chamber liner may have caused me some brain damage. It’s worse than Rubik’s cube crossed with Soduku. I’m reluctant to talk about it because if it works as well as I think it might, I really don’t want unstable rogue regimes knowing anything about it. That said, if you’re familiar with pumps or wood shop equipment, you probably can guess how I would turn kinetic energy into pressure without putting an oak barrel on a mountainside unless you’re standing naked in a cyclone or drowning in a volute.

    2. Good summary. We’ll see how well SpaceX handles their manifest this year. Their use of horizontal integration may be the driving factor on reducing pad turn-around times. For their latest launch, the rocket and payload were rolled out, erected and launched in less than 24 hours. It’s possible that as they gain experience and confidence with their rockets and crews, they’ll streamline their launch procedures, such as eliminating the static fire test. No other launch company does that before every launch, at least not to my knowledge.

      1. Thanks Larry. As noted in my reply to George, above, the easiest way for SpaceX to accelerate launch tempo on its single extant pad at Canaveral is to duplicate its horizontal assembly and prep facilities there. Once they have LC-39A retrofitted for Falcon ops also, then launch tempo can potentially be stepped up even further.

        One thing I would not expect from SpaceX is elimination of the pre-launch hot fire tests. The hot fires have been crucial in the detection of problems with several of the eight F9’s launched to-date – as they were intended to be. SpaceX wisely designed in easy recycleability to enable diagnostic hot fires for this very reason. Giving their rockets a thorough shaking before the real launch lets anything that might rattle loose and cause trouble to be found in advance and put right if need be. Sometimes it’s been rocket stuff not up to snuff, sometimes it’s been pad stuff. But the hot fires have been critical in avoiding any mission failures to-date. Ariane 5 may be the long-term reliability champ now, but its early history was a lot rockier than Falcon 9’s is turning out to be. I think the crucial difference is the hot fires.

  7. The recent rise of commercial spaceflight is a bit like the history of organic chemistry. The last few years have been a long and thorough debunking of the previously widely held theory of government vitalism as applied to spaceflight. As it turns out spaceflight is mostly just a matter of engineering, as we see more and more pictures of control rooms that are just banks of workstations manned by engineers that fact may start to sink in. Our future in space is no longer inexorably hitched to multi-billion dollar multi-decade programs, just like everything else it’s dependent on technological and industrial advancement from the private sector.

    1. I don’t think much of it, frankly. The anonymous Wikian who doubts SpaceX’s ability to undercut Russian and Chinese launch prices is, doubtless, sincere, but provides no objective basis for his skepticism.

      The best evidence that SpaceX’s posted prices are both real and much lower than those of all other competitors has been the growing tendency of senior management at ILS (Proton), Arianespace (Ariane 5) and ULA (Atlas V/Delta IV) to make public statements about initiatives to reduce costs over the past year. Such talk has spiked even higher since SpaceX’s recent trifecta of Falcon 9 v1.1 launches including back-to-back successful GTO missions.

      SpaceX has completely frustrated its competitors’ best efforts to sow fear, uncertainty and doubt (FUD) about its systems by stubbornly refusing to blow up any rockets. By a comparable point in its launch history, Arianespace had lost two Ariane 5 missions.

      SpaceX continues to book new business even as it swings into a much higher rate of flight ops to begin working down its already impressive backlog. The company is on track to launch more missions in 2014 than it has in the preceding four years.

      SpaceX has two extant operational pads – one on each ocean coast. It should shortly be concluding a lease deal for LC-39A at Cape Kennedy which will give them three, after a year or so of remodeling/refurbishment. They should also shortly be concluding arrangements to build a new launch facility – and possibly a new manufacturing facility – near Brownsville, TX. I suspect the new facility will have at least two pads. In 24 to 30 months, SpaceX will have at least five pads in service, some or all with duplexed assembly/prep buildings and TEL’s. They should be capable of launching 40 or more missions per year.

      1. Thanks!
        Can you comment on how easy it is for a space company to train new people to “do a launch”? I imagine there’s a lot involved in that (understatement). Say they had four launch pads tomorrow. How long would it take them to get them all working in parallel? Do they still manufacture the rockets in one place and transport them?

        1. Interesting questions.

          The training process for anything that is highly technical, detailed and intolerant of failure must always be protracted and difficult. Elon Musk came from the software engineering world, Thus he had, so to speak, a good basic “attitudinal platform” upon which to build managerial and training approaches at SpaceX.

          I speak here from no first-hand knowledge of SpaceX internal protocols, but let’s look at the question of how SpaceX might reasonably train up multiple cadres of launch control people; the ones who sit in front of the displays with earpieces and microphones on during countdowns. The first thing to note is that, up to now, anyway, being a launch control staffer at SpaceX has been a decidedly part-time job. The company has only conducted 13 launch attempts over the past eight years of its 12-year history. The SpaceX staffers who man the consoles in Hawthorne at launch time have other duties on their “day jobs”.

          My guess is that these folks are mostly engineers and developers of the subsystems they serve as controllers for during launches. Avionics hardware and software engineers, for instance, serve shifts at the guidance and navigation controller’s console, propulsion engineers monitor engines, etc. As SpaceX expands its launch facilities and accelerates its ops tempo, launch-specific jobs, like console controllers, will probably take more and more of the time of the people assigned, but I’m not sure it would ever make sense for them to be exclusively console jockeys. Remaining active in design and development activities would keep part-time launch controllers up to speed on the subsystems they’re responsible for on launch days. Controllers need such deep knowledge of their particular subsystems to do proper jobs at their consoles.

          The model just described for launch controller personnel development would seem to allow SpaceX to pretty quickly ramp up their controller cadre as needed to handle both the upcoming increase in ops tempo at the current two pads and the somewhat longer-term facilities expansion to four or five active pads.

          As for the logistics of rocket stages, all SpaceX manufacturing is still done in Hawthorne, CA. It is my understanding that all stages are trucked from Hawthorne to McGregor, TX and run for a full mission duration static test fire on a test stand before being repacked and shipped on to Canaveral or back to Vandenberg. This uncovers any manufacturing/assembly flaws or anything that rattled loose in shipping. After the entire vehicle is integrated and stood up on its pad, it is subject to a two- or three-second hot fire test to identify any issues that may have crept in on the first stage’s travels from McGregor to launch site.

          Building a second production facility at the prospective Brownsville spaceport would seem a good way to streamline SpaceX’s logistics as Brownsville-to-McGregor is a much shorter road route than Hawthorne-to-McGregor. Given that most of SpaceX’s pads will be two to three thousand miles from Hawthorne, I would expect any new manufacturing facility in Brownsville to fairly quickly be built out to exceed Hawthorne’s capacity and the latter may well be downsized or even eliminated at some point.

Comments are closed.