Random Thought: Should Safety be the Top Priority?

[Note: I just wanted to share a quick semi-baked opinion, and it was long enough that if I broke it up into a series of tweets, Ben Brockert and Will Pomerantz would probably remind me of this blog thing I supposedly run...It's probably not that new, profound, or even correct, but as I said, think of this as a blog equivalent of a series of tweets...]

One of the things that really strikes you about all the conversations between NASA and Congress about NASA’s attempt to help you know, follow its charter and “seek and encourage, to the maximum extent possible, the fullest commercial use of space” by funding commercial development of crew transport vehicles is the emphasis on safety. Shuttle ended up killing two crews out of 135 flights, which is actually about what you’d expect to get from flying crews on EELV-class vehicles without a launch escape system of any sort, yet in almost every Congressional hearing, you hear a ton of hand-wringing about whether these vehicles will be safe enough for NASA’s astronauts. And you can tell that NASA has taken these inputs very seriously, with all the requirements (and referenced requirements, and requirements referenced in referenced requirements, and requirements referenced in requirements referenced in referenced requirements), paperwork, overhead, and with their attempt to force things into a FAR-based mold closer to how NASA does major programs. It’s pretty clear that NASA and Congress both see safety as the top priority for commercial crew. I know this may be heretical, but I’m wondering if this is a misplaced priority.

Maybe I’m wrong, but here’s my concern:

  1. NASA really wants at least two independent, self-sustaining, affordable ways of getting people to and from the ISS. Having this capability means that if anything happens to one system, you don’t get the standdowns like what you had with the Shuttle program.
  2. Having at least two affordable and healthy competitors also means more price competition, and more incentive to innovation.
  3. There’s no chance that Orion on SLS is going to be anything within spitting distance of “affordable” for routine crew rotations.
  4. As NASA has been openly admitting for almost as long as this blog has been around, they know that they can’t afford to go beyond LEO if they can’t offload all of the ISS crew and cargo needs to commercial providers using firm, fixed-price contracts.
  5. But NASA only wants to buy about 8 seats per year (two rotations of four crew each) from commercial providers, in order to meet their ISS obligations.
  6. You’re only likely to get two affordable and healthy commercial crew providers if they have enough demand to spread their fixed costs out over (and if they can keep those fixed-costs within reason).
  7. I can only see a few ways of doing that (though there may be others):
    1. Have the commercial crew vehicles be affordable enough that they can enable significant non-NASA crew, cargo, and recoverable freeflyer (like DragonLab) services.
    2. Having the commercial crew vehicle be similar enough to a commercial cargo vehicle that each provider can actually get a decent number of flights per year out of a mix of crew and cargo.
  8. Only the first of those two options avoids the challenge of a NASA/commercial crew monopsony scenario, where the ISS is the only thing keeping the commercial crew providers afloat.
  9. While there is a small, but non-zero, chance that you could get sufficient demand from what Bigelow calls “sovereign clients” to get non-NASA crew/cargo demand even at the old $20M/seat Soyuz price, the best analysis I have seen with the existing data (pgs 43-53 of this presentation) suggests that the price point commercial crew needs to get in order to reach a tipping point is $5M/seat max, and possibly as low as $1-2.5M/seat.
  10. While it may be barely possible for NASA to eke out a minor victory by getting two independent¬† and semi-healthy commercial ISS crew providers who also do ISS cargo deliveries on unmanned versions of their rockets/delivery vehicles, even this minor victory is only possible if the fixed cost of the crew capability isn’t too excessive.
  11. With only two flights per year worth of crew demand, there might not even be enough demand for one commercial provider unless they can find synergies with ISS cargo deliveries, or more preferably non-NASA customers.

I guess my big concern is that it doesn’t appear as though NASA or Congress are being realistic about how to properly prioritize safety. Ultimately you can always spend extra money on safety (one more test, one more certification, one more sign-off, one more review, etc)–the only way to have 0% chance of losing a crew on an ISS mission is to not do the mission. If you are actually going to fly, there’s a point where you have to accept some risk, and you have to say at some point that you’re only willing to spend a certain amount of money to potentially buy down tiny fractions of a decimal point safety-wise. If you have to make that decision anyway, then it makes sense to do it in the framework of the big picture of the mission risks and overarching goals.

This is something for instance that the Constellation program utterly failed to do–the core justification for Ares-I was that it’s launch ascent safety was supposedly going to be so darned good (a 1 in 2106.4823910293 chance of losing a crew on ascent, at a 50% confidence interval…), but in the light of a program that expected a 2% or greater chance of losing a crew on a given lunar mission, it’s pretty clear that spending money to go from a 1 in 1000 probability on existing LVs versus spending a decade and $10-20B on a new launcher to buy that risk down a bit was money very foolishly spent. The problem is I worry we’re going down the same path with commercial crew.

While I don’t personally have any really sage advice on how best to ensure safe operations while still keeping the overhead low enough to keep commercial crew provider costs low enough to give a realistic shot at enabling a new market to emerge, I am worried that the current balance is a well-intentioned disaster waiting to happen (see also Wayne Hale’s previous warning on this topic).¬† If NASA and Congress continue down the path they’re going with safety, there’s a very real chance that they’re going to make commercial crew commercially unviable. And that would be the ultimate Pyrrhic Victory–having one or two “commercial” crew providers that in the end that are flying, but are so expensive that only NASA can afford them.

This entry was posted in Commercial Space, COTS, NASA, Space Policy. Bookmark the permalink.

10 Responses to Random Thought: Should Safety be the Top Priority?

  1. Blackjax says:

    I really think basing your market estimates on that old T-Space report is dubious. First, it does not do any projections in the $6m-$20m range and lets face it, that will be the market that gets tapped in the first year or two *before* providers start heading towards tapping the $5m-$1.5m pool. Prices will fall as the market at a given pricepoint is tapped out. Sure, at the higher prices the pool is much smaller, but then you don’t need hundreds, you just need 24-36 people for the first year or two to subsidize the startup of operations. Also, bear in mind that while it speculates on the amount of net worth people would be willing to spend, it is only an estimate, and there will be outliers. It also does not account for repeat business, and as Simonyi showed, that may happen.

    Second, it only addresses the tourist market because it only looks at those who would be paying entirely from their own resources. As Richard Garriot showed, this does not necessarily have to be the case, there is the possibility of hybrid funding. Further, a Bigelow station would need Bigelow employees, which would constitute a predictable baseline demand. In addition, researchers with human tended experiments would need tenders (Astronauts4Hire staff for example). Beyond that, flights to a Bigelow station might in some cases substitute pressurized cargo riding along with people (effectively paid carryon luggage) in place of some seats. There may also be cases where peoples trips are paid entirely by the government, academic institution, or a company; contest winners and researchers who must be onsite rather than using automation or another human as a proxy for example. There is also the chance of companies doing some sort of non-research thing there, a publicity stunt, a movie, or some other endeavor that they would foot the bill for. I don’t know if any of the above could be relied upon to provide a healthy market alone, but in aggregate and combined with NASA demand and a few sovereign clients, I think it’d close the business case. In any event, I think the market model needs to be far broader and more nuanced than what is captured in that report you cite.

  2. Blackjax,
    Oh, if you read through their methodology for those numbers it gives you a way to update them based on more recent net worth numbers, which may be able to give you more granularity in the $6-20M range. I still think the old t/Space reanalysis of the Futron study has some utility, even though it is now almost a decade old.

    A couple of quick points:
    1- While there may be several good reasons for it, the Futron passenger estimates have been significantly higher than the historical record so far.
    2- Prices can only come down if costs are low enough–and my concern is that with the currently planned NASA safety overhead, it’s going to be really hard to actually get down to the $5-10M/seat price at the passenger rates you’re likely to get at those prices.
    3- More importantly, the market elasticity around $20M/seat is such that the drop to $10M/seat may result in an increase in passengers low enough that you don’t actually make more revenue at that point (ie 1/2 the revenue per seat, but maybe only a 1.8-1.9x increase in passengers). The reason I was interested in the $1-5M range is that’s where you get into the virtuous cycle where a reduction in seat price leads to a sufficiently higher increase in passenger sales that your total revenue increases.
    4- While it may appear that the most optimistic cases show that commercial crew could make it even in spite of the distortions caused by NASA’s currently planned safety processes, that’s still in the most optimistic cases. Entrepreneurs and investors need to be concerned not just with the best-case scenarios, but with the most likely ones. And the most likely ones point to a pretty weak level of commercial demand at the price points the providers can likely deliver if put through the safety certification ringer that NASA and Congress want them to go through.

    ~Jon

  3. Blackjax says:

    Wouldn’t the NASA rigor only apply to the flights that in some way service NASA itself? If there really is a commercial market out there that is sustainable, then what would stop the commercial providers, once they exit CCiCap with a flight ready system, from siloing operations and charging NASA a higher rate based on their higher requirements, and operating the purely commercial side with more sensible requirements and charging a lower rate for it. Alternately, if NASA was just too onerous, they could walk away from them altogether (if there really is a commercial market out there that is sustainable).

    Personally I have always viewed NASAs role as primarily that of providing the seed funding for the systems development and to a lesser extent as priming the pump with some modest demand in the very early years, not as ‘the market’ or even a major component of it. This is especially true given the budget situation we are likely to see get much worse as time goes on.

    If this industry is going to happen, it is going to have to stand on its own two feet outside of NASA business fairly quickly after the capabilities come online.

    Personally, I think people overlook the market which might come from people like Garriot (and Olsen?) who are wealthy and would love to fly multiple times, but who are not quite wealthy enough to make it happen. I see them signing on with outfits like Astronauts4Hire and agreeing to share the costs of their flights with projects who need their services. The projects win because they get a partially subsidized astronaut, the wealthy astronaut wins because they get to satisfy their passion without footing a bill they could not otherwise financially justify. There is also the possibility that some types of research subsidy (particularly the human factors stuff where the passengers body is effectively the experiment) where most passengers could see their costs mitigated to some extent. Just as NASA is piggybacking some data collection purchase on GLXP competitors, I could easily see many commercial passengers being paid for bio data collection on their bodies reactions to zero-g. It wouldn’t even close to pay for the trip, but it might take a little of the edge off the financial hit. I could also see numerous other ways for the ‘nearly wealthy enough’ to find modest ways to mitigate the costs to the extent that they could do it. Flying then auctioning stuff they bring with them, shooting a small film/documentary while onboard (as apparently Garriot tried to do), speakers fees either while on orbit or at events afterwards, playing with/testing their own HSF related invention ideas that could later be productized and sold back to the commercial HSF market, etc. There are a lot of things that don’t have a strong case as a standalone funding strategy, but which might serve as a cost mitigator to someone who was primarily motivated by a passion for space, but who only had some of the money necessary to do it.

    Your point #1 is probably true, but you don’t need a huge number, you just need enough volume to keep the providers operating and driving their costs down.

    Point #3 is also quite valid, but again, again you don’t need a huge number, you only need enough. Businessmen like Elon and Bigelow are smart enough to project out and look for sweet spots in the price/volume equation. Elon might be willing to take a chance on a lower (or even negative) revenue/profit portion of the curve if he sees that beyond it is a volume level that enables him profitably achieving his dream of opening up space. He and the other wealthy company founders involved at this stage are philanthrocapitalists not really pure capitalists, which means the way they evaluate risk/reward has some significant differences from what you’d see from a pure capitalist.

  4. Rand Simberg says:

    Thanks for this, Jon.
    I’m working on an Issue Analysis for CEI on this topic, and will probably steal liberally from it.

  5. Chris (Robotbeat) says:

    There’s an even bigger problem…

    With Ares I, it would’ve never flown often enough to have reached anywhere near its theoretical safety. Spending too much money on safety can be counter-productive to the actual eventual per-flight safety since you are using a unique system whose only purpose is to launch NASA astronauts at great expense and rarely. Any system which has other uses (i.e. cargo or science missions or non-NASA missions) with a far more modest investment in safety can exceed the safety of one which spends so much money on safety that it never flies more than a dozen times. Even if you don’t buy the idea of engine-out, etc, SpaceX’s crewed Dragon’s first manned flight will probably have be much safer than Ares I’s first flight, since it will have had at least a dozen or so unmanned flights beforehand. And because it’s used in unmanned launches too, if a quality control problem does turn up, most likely it will occur first in unmanned launches, as the recent Proton failure showed. This applies to EELVs as well, but not as much to the non-Dragon spacecraft since they won’t be used for lots of cargo launches before launching manned.

  6. Adam G. says:

    I agree with Blackjax that if you really can reach the tipping point price, then you tell NASA to bugger off.

    The issue, however, is that trying to reach that tipping point is a risk and its a risk NASA should be partly assuming instead of being at best irrelevant with overly-safe, overly-expensive commercial crew.

    IMO, we shouldn’t really blame NASA, though. This is mostly Congress and a lack of executive leadership.

    Finally, when arguing this point, be sure to remind NASA and everyone else that high flight rates are a much better way of identifying and reducing risks than more studies and more features.

  7. Andy Clark says:

    Nice piece Jon. However, no one wants to seriously address the issue of risk and what level of risk can be considered acceptable. In the early days of the Shuttle program a loss rate of 1 in 25 missions was being talked about at Michoud. Unfortunately Challenger proved that to be a really accurate prediction. So, what happened: the hardware system was made safer, the management system was not changed significantly but now has become extremely averse to risk of any sort. This has become endemic in the agency and the political class who do not understand the issues but are always on the lookout for a way to get publicity. The Columbia accident only added to this sense of risk aversion. The Shuttle was an experimental craft and flight test involves risk. We attempt to minimize the risk in any venture but it can never be reduced to zero. At some point the law of diminishing returns sets in and costs, mass and added complexity begin to take their own toll of the reliability numbers.

  8. Peterh says:

    To amplify the point Chris (Robotbeat) made, flights are needed not just to catch bugs in the hardware, but to also build proficiency in the crews building and operating the systems. So excessive pressure on “safety” can paradoxically make the vehicle less safe by reducing flight experience.

  9. JohnHunt says:

    NASA is going to insist on a high level of safety for any craft visiting the ISS. But for cis-lunar operations with lunar landings, could NASA just pay for cargo deliveries and leave manned use of the OTV/lunar landers to the commercial firms by having the workforce there be private? NASA could pay a set amount for delivery of lunar water to L1 and LEO cracking fuel depots. NASA could hold off on applying safety requirements until they were ready to send their own crew in their own vehicle to an asteroid, a martian moon, then to Mars. But operations delivering lunar-water derived fuel for orbital servicing would be a strictly commercial affair. As for private lunar astronauts, their commercial value would be to fix equipment and exploit the regolith to construct simple but bulky telerobotic parts. Occasional lunar landings would provide the high-tech, difficult to produce but small parts for an ever-expanding telerobotic workforce. Circum-lunar and lunar surface tourism could also come in due season.

  10. Pingback: Selenian Boondocks » Blog Archive » Random Thought: Is Commercial Crew Having to be “More Catholic Than the Pope”?

Leave a Reply

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

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>