The Case Against SpaceX

As an advocate of commercial spaceflight I can understand why many find it difficult to write objectively about SpaceX - which is arguably the poster child of this nascent industry. As a result, the majority of negative commentary about our darling comes from a horrible "journalist" like Andy Pasztor at The Wall Street Journal or a traditional aerospace mouth-piece like Loren Thompson at Forbes. An occasional coherent comment on a blog or space forum may be accepted by the space community as containing a nugget of truth, but these are easily filed in the don't-think-too-hard-about basket and forgotten. I've taken to thinking about the criticism I have heard, and after some long and rather arduous discussion with these critics I've processed it through what I hope are reasonable and constructive filtering. Here's the finished package.

The Circuitous Route To Reuse

Since SpaceX first announced the Falcon 9 they have claimed it is designed to be reusable, but they've yet to demonstrate how. For a number of years the answer has, apparently, been parachutes. Both the successful flights of the Falcon 9 have carried them and for a while we were told they had been deployed. Gwynne Shotwell, speaking at the Space Access conference this year was quoted "We have recovered pieces of the first stages." They were breaking up during re-entry, not giving the parachutes time to deploy.

Most recently, SpaceX has announced with fanfare the new overall approach with pretty graphics and a funky soundtrack. Clearly, they are still a long way away from a working vehicle. I asked Gary Hudson on The Space Show to provide us an educated guess at how fast the Falcon 9 may be going at first stage separation, and at what altitude - the kind of trivial information required to even visualize how such a vehicle could function. He declined.

While it is certainly true that SpaceX's engineers have a lot more information and no doubt have some idea how it is supposed to work, I find it more than a little disconcerting that arguably one of the most seasoned RLV veterans around today isn't able to speculate. At the same time, a new test program dubbed Grasshopper was announced to test vertical takeoff, vertical landing (VTVL) which is a critical part of the new non-parachute approach to reuse.

If that sounds familiar, it should. A subscale VTVL demonstrator has been considered the starting point for this kind of RLV ever since the DC-X program back in the early 1990s. More recently, Armadillo Aerospace and Masten Space Systems have been following this path, with considerably less money to play with. It's 2011 and it seems like SpaceX is starting all over again with reuse.

The Funding Crunch

There's another pathway to reuse: put wings on it. So far, we haven't seen any indication that SpaceX is pursuing that route but, then again, we saw no indication they were pursing VTVL a year ago either. A regularly advocated way to maintain revenue while pursing this route is to woo suborbital markets such as scientific research and tourism. This approach is best exemplified by XCOR and, to a lesser extent Virgin Galactic (as they still seem to have no orbital aspirations). So far, there is no indication that SpaceX is doing that either, but who knows what the future might bring. If the current suborbital providers are successful it may boost investor confidence so much that SpaceX begins to take an interest.

Instead, SpaceX intends to fund their RLV development by selling launch services on the expendable configuration of the Falcon 9. This is good in a number of ways, most notably that it gets into the orbital launch business early, establishing a record of success (hopefully), and has given SpaceX the rocket engine and other components necessary to even start thinking about making a reusable vehicle.

The alternative to both of these paths is to simply have enough up-front funding to buy rocket engines and components from existing providers. For example, the RL-10 from Pratt & Whitney Rocketdyne is considered one of the most reliable rocket engine families available with variants that have been tested rigorously for reuse. Many single-stage to orbit (SSTO) designs of the 1990s just assumed this engine, and the cost, most likely because it was used in the DC-X. SpaceX didn't have this option because their funding was meager by aerospace standards.

The Mars Dream

Elon Musk's plans to send humans to Mars are simply not realistic. Or, at least, that's what I'd say if I had any idea of the details. From all the times I've heard the dream I've managed to garner that basically he's adherent of Bob Zubrin's vision of men braving the perils of space to explore the red planet, with families of immigrants following close behind. This is complete with the heavy lift fetish.

Despite decades of examples that heavy lift can never be cheap, SpaceX has redefined the idea by claiming their upcoming Falcon Heavy launch vehicle will break the $1000/lb barrier and usher in a new age of cheap access to space. The aspirations for even bigger launch vehicles (presumably with even cheaper prices per lb) run deep.

Advocates of staging propellant in orbit, assembling and refueling deep space exploration vehicles which are launched on more modest sized rockets should not be surprised if they find heavy lift advocates counting SpaceX in their camp - but they often are. This defiles the traditional battle lines, with RLV advocates more commonly coming down on the side of propellant depot advocates, if not simply because one of the best uses for an RLV is filling propellant depots with propellants.

As such, it seems that the dream of Mars at SpaceX is essentially Mars Direct with a single heavy lift launch vehicle throwing a Dragon-sized capsule, with stir-crazy explorers, directly to Mars escape velocity. Zubrin has written of such a plan, claiming a Mars landing by 2016 is possible using the Falcon Heavy. It looks good on a cocktail napkin but the same old hand waving is required to shoo away the issues with those pesky human factors like radiation protection and artificial gravity generation.

NASA Assimilation

Practical and profitable space activities are much more effective for exciting public support than dreams of Mars exploration, but it is clear NASA is not going to industrialize space - it threatens the status quo - and today NASA remains SpaceX's greatest customer.

The goal of SpaceX is human spaceflight, and the greatest repository of knowledge about human spaceflight is NASA. As such, it would appear obvious that getting NASA to help you to fly humans safely is a good idea. The way to do that is with Space Act Agreements. This is what SpaceX did under the COTS program, and later under the CCDev program.. and they got paid for the privilege. As a result, the Dragon spacecraft will soon be fully qualified as safe for human habitation on orbit as it will be berthed to the ISS and have astronauts inside it.

The problem is that NASA is a precocious customer. They know what they want, they think they know even better than you do how to make it, and they feel no guilt about changing their mind halfway through the project. As such, Space Act Agreements just totally grind NASA's gears. They don't have enough control.

NASA money is like heroin.. once they start taking it, most people find it very hard to stop. There's a dependence that has grown between NASA and SpaceX, and although it is obviously a love-hate relationship, it's going to be very hard for SpaceX to let go.. but, inevitably, they must. The current needs of NASA are very different to the long term goals of SpaceX.

Promises, Promises, and Delays

SpaceX promises a lot more than they deliver. Over time those promises have changed, with the old promises being forgotten, and new promises being made with more showmanship. Failure is to be expected, with plans changing in response to the lessons learned, but doing so requires clear acknowledgement that there was a failure.

In September this year it was revealed that the second flight of the Falcon 9 had experienced an engine anomaly. While it later became apparent that the issue was minor and not unexpected, the immediate response by the space media was to pounce on what could be a hot story. Quite a number of people I talk to have expressed dismay at the way SpaceX handled the situation, including the lawsuit against Joseph Fragola earlier in the year. While I certainly don't subscribe to the view that SpaceX should be anywhere near as open as NASA with their proprietary information, I do agree that it is indicative of a deeper problem with their engineering culture.

Oh, and we're still waiting for a Falcon 9 flight in 2011.. seems it isn't going to happen.

The Business Case

Now I'd like to talk about the elephant in the room. Fundamentally, SpaceX has a shoddy business case which is best described as a house of cards.. that they're still trying to play poker with.. and there's dogs at the table, and they're smoking cigars! Yeah, metaphor.

The launch business is about volume. If you can get your launch rate up then you can charge less for each launch because the fixed costs will be spread over more launches. SpaceX hasn't done that yet, but they're already charging less than anyone else in the business. This is a common criticism of SpaceX, which most of us in the advocate community love to retort by saying something like: Elon says SpaceX has been profitable every year since 2007!

Okay, that's great. How? There's really only two possible answers: NASA's money, or booking fees. If it is just the former then SpaceX is destined to become just another NASA lackey. So we prefer to think it is the latter - but that means they're living on their seed corn. Eventually they're going to have to actually fly these payloads or give back the deposits. So the acid test will come when SpaceX is called upon to launch and turn a profit in the same year. At that time we will discover if SpaceX is getting the launch rate they require to amortize the fixed costs such that their revenues exceed their expenses. Only then will we know if their prices were realistic.

Suppose they're not. What options does SpaceX have then? Obviously, they can't rise their prices much - that will put them in the same market as the existing providers which have a much better track record (and much better ties to the biggest customers in the government). SpaceX is competing on price, so they will have no choice but to reduce their expenses or increase their flight rate. Reuse is their strategy for doing both of these approaches simultaneously.

There's only one problem: the inherent assumption that there is a market for cheap lift, and that this market can come online fast enough to provide the demand to both amortize their fixed costs and fund their reuse development. In the space community we have a name for this kind of faith: if you build it they will come [no really, watch it!].

If you ask Elon Musk why he is building something which is totally illogical, he will give you the story about the Mars dream. While I don't fully subscribe to the space-based solar power vision, preferring the human tended maintenance of geostationary orbital satellites variety of industrializing space, at least they have some practical idea of what might be economically valuable activities to do in space.

SpaceX is, unquestionably, a bold faced shot in the dark. It is going all-in on the boat hoping that the river will give you a straight flush. It's ballsy madness, and that's why we love them.


  1. Reuse: SpaceX is "starting all over again" (with "considerably more money"). So what? Neither Masten, or Armadillo are targeting orbit in the short term and it will cost "considerably more money" if they actually ... do. In fact, both Masten and Armadillo have been on record saying orbital is a long ways off and not in their current design plans (so in essence they'll be effectively starting over, even if their VTVL code is adaptable). As far as DC-X is concerned SpaceX will be further along as soon as they do a successful series of VTVL's that don't ... fail. As far as I'm concerned SpaceX is entering territory that hasn't been accessed before and they're putting a big stake in its success.

    Mars: "Heavy lift can never be cheap." Only if you dismiss reuse. You yourself pointed out to me that Falcon 9H could be great if you reused the first stages. How about a Mars mission on Falcon 9H with three reused first stages that have gone through something like 10 cycles each? They'll have been paid for several times over. A Mars mission, then, would be extremely affordable. Without reuse I of course can't disagree with you, but it seems as if you're dismissing a fundamental aspect of SpaceX's further space vision.

    As far as SpaceX is concerned such a mission would be Mars Semi-Direct, as it would still take several launches. But hell, if you're reusing everything the costs become ridiculously affordable.

    NASA Assimilation: SpaceX has not hid the fact that NASA's funding has sped up the development of SpaceX, however, it would've happened one way or another, taken decades, potentially, but it would've happened. Any business person knows that you start small and grow, which is why SpaceX had a strong limit on its number of employees for quite awhile there. Unless SpaceX is caught up in NASA-style cost-plus contracts that eat at them like a cancer, SpaceX should survive quite well. Half of their manifest, they like to tout, is already commercial space, and not NASA. Now, that is somewhat misleading because NASA is paying far more than commercial space is paying. But we can appreciate their "dependence" on NASA as merely a bureaucratic "speeding up" of their business, as opposed to simply looking at their NASA agreements as a relationship and dependency that is "hard to let go." Once R&D is finished, they will have no reliance on NASA in any way whatsoever. As it stands now NASA's control over SpaceX is minimal at best, which is why we have these debates about CCDev and how it acts as a virus.

    Delays: For a space startup they're irrelevant, so far they're about a year behind by my reckoning. If they achieve launch this Dec. and since they've compiled the two COTS Demos (COTS B/C) into 1, they will in fact be completely within the contract requirements (COTS-A-C require the contract to be fulfilled by the end of this year). Sure, if they weren't able to compile B/C one could make the argument that they were unable to meet their deadlines, but as far as I'm concerned they've been on target, which is remarkable given the difficult of achieving orbit.

  2. Business: SpaceX again has noted that NASA's money has helped them move forward as fast as they have, but Elon Musk has been on record saying that if NASA disappeared tomorrow they'd be OK. I believe him. Without reuse the business case falls apart. But by the same token, it falls apart for every *orbital* company that exists. It's asinine to just throw away million dollar machines every flight. Even if you could automate the whole process you're looking at millions of dollars of raw materials. Unless you're going into the Ultra-Heavy category like Sea Dragon it just doesn't make sense to throw away all those materials if you can afford not to do so.

    As far as a market for the kind of cheap lift that they expect; Elon Musk comes from the internet dot-boom age, and he survived. Arguably, imho, because of the kind of analysis necessary to know where the markets are heading. He halves, quarters, the cost of launch, then I think there's a market there. Particularly if he can get a hold of those Army launches, which is partially where he's obviously going with it. Then, again, I suspect the argument wouldn't be "NASA's paying them, the USAF is paying them!" but I can't really argue against that.

    From a business standpoint it shouldn't matter who the hell is writing the check so long as the checks are being written, and I don't think the SpaceX will have any issues with that in the future. The question really will be whether or not SpaceX allows itself to cede to NASA's bullcrap cost-plus pork-based jobs programs that would ultimately kill SpaceX in the end. If SpaceX is forced to triple its workforce in order to apease NASA and if they don't see that it's a disaster, they do have a risk of becoming a niche. But, there's no evidence whatsoever that SpaceX is going to allow itself to fall into that trap. If they succeed that reuse then they will obliterate all the other space launch companies, and the USAF and NASA will be potentially giving SpaceX free reusable first stages to play with, reducing costs dramatically, beyond anyones dreams.

  3. Anonymous2:57 AM

    Good summary of all things SpaceX.

    I bet if you ran a Monte Carlo simulation of their business plan you would find a significant percentage chance that they would fail. I won't guess what that percentage is, but I can imagine the business conditions that would lead to it.

    Of course I could also run a Monte Carlo simulation for any NASA HSF program and see a significant percentage too, although the main difference is that the government can provide additional money to recover from significant failures easier than companies can.

    So really this boils down to acceptable risk, and acceptable rewards, and this is the type of thing that Elon Musk knows about managing. In the startup world you don't get investors to invest unless you can convince them that there is a HUGE profit potential, and the only way you get that is by coming up with a new paradigm for making money. Kind of the Go Big Or Go Home thing.

    So it is with SpaceX, which is going big with a set of technologies and techniques that run counter to the existing industries (i.e. disruptive technology). And really the bottom line in these types of situations is how they manage risk, which usually boils down to management decisions about everything from design choices to hiring choices.

    No doubt finishing COTS and starting CRS deliveries will be key to their short-term profitability, so next year will be key. I'll be watching with fingers crossed like everyone else.

  4. A good article, and necessary, I think. The reason why the other side has gotten away with claims of "amateurism" and "hobby firm" is perhaps because SpaceX has skimped at the answers because the claims were so obviously wrong. So I think SpaceX need to be asked some good, constructive questions and by answering these questions they will make their case obvious for all. I mean, noone calls Boeing or ULA for hobby firms.

    So go for it, Trent.

    One thing I would like to know was their reaction to the suddenly disappearing market for Falcon 1e. At one time they were planning for a lot of orders using this rocket, but that market suddenly disappeared. Why?
    Another one is their launch manifest. A lot of "cheap" Iridium's in there. Isn't it so that when a new rocket launches expensive satellites, the price of the launch itself doesn't really matter? And especially so when factoring in insurance costs? Isn't a cheap rocket dependant on the need for cheap satellites?
    And I would also like to hear what they are planning to do after COTS is completed. What are the long- term thoughts of the market they will cater for? NASA or Satellites?

  5. Great replies, thanks everyone.

  6. Good article Trent. One serious point I think you should have considered was their failure to deliver cost reductions on COTS. Yes NASA gave them bureaucracy light (for gov contracts), which should have lowered their costs, but by the time GAO added up the money - SpaceX and orbital cost them more per pound to ISS then shuttle or Progress. Their failure to meet dates, their obvious failure for their first RLV concept to work out (whatever it was), and a failure to deliver cost reductions on the extremely high profile COTS program (which they hoped to use as a stepping stone to crew carry contracts to the ISS, all have to make commercial customers nervous. Its certainly not wining friends in congress, and NASA has used it to prove FAR programs must be used in the future.

  7. GAO never did any such analysis.. Congressional staffers did, but they're all idiots who can't even add. Happy to be proven wrong with a link to a GAO report.

    Everything else you mentioned is already addressed in the article.

    BTW, thanks again for our discussions.

  8. True that the report I was talking about was congressional, not GAO.

    But given the GAO's Christina Chaplain (Director, Acquisition and Sourcing Management) it seems a fine point. And its not hard math. [Did you have some dispute with the numbers?]

    Also its interesting to see how Orbital and SpaceX reacted Differently. As I remember, SpaceX screamed about how unfair it was etc, but agreed the numbers were correct - but given the situation....

    Any time with the discussions - hope it helped.

  9. The complaints in the document about NASA being premature in making a procurement for commercial resupply services mirror my own complaints about commercial crew services. They certainly did jump the gun.

    As for the numbers, I assume you're talking about page 5.. which has the caveat:

    Space Shuttle, calculated assuming four missions per year with a capability to deliver 16 metric tons (35,264 pounds) to the space station at a total annual program cost of $3.0 Billion. $3,000,000,000 ÷ (4 flights × 35,264 pounds/flight) = $21,268 per pound.

    Even if you're willing to accept that the aging shuttle fleet could have been kept flying four missions per year, they have never launched more than 28,104 pounds to the ISS. And don't ignore how horrible a round three billion is as an estimate for the cost, or the fact that it means doing nothing else but resupplying the ISS - with human crews at risk no less. So yes, it's just stupid.

    But if you care to reword your criticism as saying that NASA spent more on COTS/CRS than they needed to, I'll happily agree with you. It's not like they care about getting the best price anyway.. the best way to do that is to just pay Russia. When I'm being generous I suggest that NASA is paying SpaceX/Orbital more than the Russians to deliver less supplies because it means they can stimulate the local industry.. but the truth is that they just don't have any motive to lower costs and never have.

  10. There would be NO problem keeping the shuttles flying. And there is no other option for flying crews on the horizon as safe as the shuttles.

    Shuttle never lifted more then 28K per flight? The ratings I saw were 6 ton max. to ISS.

    So your math and their's is $21K per pound to ISS as the report listed.

    As for NASA having to pay more on COTS, I'ld say no. (I mean they are NASA, they spend 4 times as much on shuttle then they need to as well, but COTS got a lighter load, so they got a unfair advantage if anything. That's what it took to get SpaceX and Orbital up to the point that they had ships to support COTS with. I'E. a development program which was not what COTS was supposed to be. Which didn't win friends in Congress. Its been suggested NASA played COTS to disprove commercial launch services the way they played X-33 to discredit SSTO and RLV's.

  11. Oh, speak of the devil

  12. You write well and put a lot of thought into it.

    Let's start with reusability, since it is the key to everything. The video from SpaceX represents the 100% optimum solution. They may not get there. It wouldn't surprise me if second stage recovery is just too difficult for the foreseeable future. Bit the second stage only has one engine, compared with nine of the first stage. So what if they can't reach 100% reusability. If they can recover the first stage then they've got 90% of their engines back.

    First stage recoverability is very achievable. Ok, so once again the video shows the optimum solution that's maybe a bit of a stretch. But add a drogue chute (to save on fuel for all but the final descent) and land on a barge in the ocean rather than execute a return to launch site, and it's not a problem to get the first stage back.

    Assuming that SpaceX can recover the first stage of the Falcon 9, then that changes everything. For one, it means that there is an insignificant cost difference between flying a Falcon 9 vs a Falcon 9 Heavy (with two Falcon 9 cores serving as boosters). SpaceX has already reduced US launch costs by a factor of 10. With just first-stage reusability they will reduce it by another factor of 10.

    A two order of magnitude reduction in launch also drops the price of the cargo. If launch costs are $100/kg, one no longer needs to spend $100,000 for titanium parts that shave a kilogram off the payload. No longer does one need triple redundancy, since the mission doesn't cost a billion dollars any more. There's a chain reaction of effects that drops the mission price much more than just the rocket's cost drop.

    TL;DR: Don't get hung up about the difficulty of 100% reusability. 90% will still change the world.

    1. This comment has been removed by the author.

    2. Hi Kyle, thanks for your comment.

      One can criticize without "hating" or being "jealous". Perhaps you should go listen to Elon Musk's interview with Kevin Rose where he expresses the importance of seeking out and listening to negative feedback. He goes on to say that most people avoid it because it can be painful to hear, but ultimately it's the only thing that will make your product better.

      Good luck on the work, we all wish you the best.

    3. This comment has been removed by the author.

  13. Great analysis. Thanks.

  14. > 90% will still change the world.
    And 50% will ensure that it has enough business income to provide all of the money it needs to pursue greater reusability. So I think that SpaceX has already reached the tipping point. It's very competitive costs means that it is lining up future customers more quickly than others. Signed future contracts gives investors & loaners (not just seed corn) considerable confidence and so I would bet that they are home free and so are on a sustainable path even without more NASA investment.

    1. According to SpaceX's manifest, they are mostly gov launches. According to the customers - they are not really cheaper then the established providers - so new commercial clients list them as one of the providers they booked a reservation on. I.E. SpaceX is a back up provider -- or a wedge to get the main provider to give a lower charge.

      With Congress losing patience with SpaceX (SpaceX is the poster child to congress of why all consideration of Commercial crew should be dropped) and their inability to meet deadlines and cost projections; Complaining that tax money should not be used as venture capital grants, invested in pet competitors in a market, SpaceX could lose over half their investor and client base at a shot. Such a drop would likely scare off the bulk of their commercial clients as well..


Post a Comment

Popular posts from this blog

Disabling OS-X Device Removal Warnings In Yosemite

A 5mb binary blob in the kernel?