The Columbia Disaster

The latest issue of Space Safety Magazine is dedicated to it. I disagree with Andrea’s take here, though:

The focus of commercial space is very much on cost-cutting, while vague assurances are made about safer vehicles. Sometimes safety is even presented as a stubborn obstacle to industry development and progress [I plead guilty as charged - RS]. The commercial human spaceflight industry needs to remember that the primary goal of the Shuttle Program was cutting the cost of transportation to orbit by an order of magnitude, a goal at which it failed miserably. As with the supersonic Concorde, the Shuttle was doomed by being both expensive and unsafe. Being expensive made it in turn unaffordable to undertake any further development or safety modification. But even being expensive to operate did not stop either the Shuttle or Concorde from operating for about 30 years. What ultimately ended these programs was their inadequate safety.

Probably true for Concorde, but not for Shuttle. As I write in the book:

While many mistakenly thought that it had quadruple redundant computers because it was a manned vehicle, the presence of crew was irrelevant to the need for its reliability – it had to be reliable because they cost about two billion dollars each, and they only had four of them at most operational, sometimes only three, at any given time, with production lines for components shut down in the eighties. The fact that it wasn’t reliable to “n” nines (where “n” was at least three, or would fail less than one in a thousand times) meant that the program was in fact a failure. Ultimately, this was why it was canceled, not because it killed fourteen astronauts who are, in comparison to just four Shuttle orbiters, and at the ever-present risk of sounding cold hearted, if not a dime a dozen, certainly available in surplus numbers.

After Challenger, the crew didn’t need a fire-pole escape system, or ejectable crew capsule, or any of the other impractical changes that some wanted to make to improve the Shuttle’s safety, any more than airline passengers need parachutes – they needed a more reliable vehicle. And there was no way in the end to provide it with the Shuttle, because its reliability was too compromised by its intrinsic design resulting from its pinch-penny development. That’s why the program failed and was canceled, not because it was unsafe per se. We simply couldn’t afford to continue to lose orbiters.

The corollary to this is that any reusable vehicle that is sufficiently reliable to make economic sense is sufficiently reliable to carry valuable cargo, including humans, simply because the vehicle itself is so expensive, and there is nothing about space vehicles that changes this intrinsic economic argument. Note that this is almost completely independent of what we think a human life is worth (unless we believe that the value is infinite, which many behave as though they do). This is why “human rating” (the modern PC version of “man rating”) is an archaic relic of early and unreliable expendable vehicles, and is a phrase that should be banished from the policy and engineering lexicon.

If what they were doing in space had been important, a one-in-a-hundred chance of loss of crew might be just fine, particularly if we had a large enough fleet. As I note in the book, a four percent loss of aircrew per mission in WW II was considered acceptable by the Army Air Force. The real problem with government spaceflight is that it isn’t important. And for private and commercial spaceflight, both service providers and customer are going to have to make their own assessments of what is an acceptable risk. It’s none of the government’s damned business. I’d also note that commercial industry’s assurances of safety are no more “vague” than NASA’s have ever been.

By the way, Andrea is supposed to be reviewing my book. I’ll be interested to see what he says.

[Afternoon update]

One other important difference between Concorde and the Shuttle. One was an airliner, with a high expectation of passenger safety. The other…wasn’t. Or at least shouldn’t have had such an expectation, particularly after Challenger.

10 thoughts on “The Columbia Disaster

  1. George Turner

    Your post makes me wonder how much of the increase in reliability of fighter aircraft is driven by their extreme cost compared to earlier ones like P-40 Warhawks and F4F Wildcats.

    For a large agency like NASA you could easily caluclate the expected annual death rate by summing up the average rate of murder, auto accidents, bath tub accidents, heart attacks, and workplace oopsies and it would certainly be much higher than the number of astronauts they lose in space flight, though that’s also a reflection that they also put very few people into space.

    1. Casey

      Actually, it’s the other way around; they cost more because they’re more reliable.

      That is, a better-designed, better-built vehicle costs more than a cheaply-built vehicle. Shocker, I know. :)

  2. Rand Simberg Post author

    Your post makes me wonder how much of the increase in reliability of fighter aircraft is driven by their extreme cost…

    Most of it, I’d say. They’d be designed to the same reliability even if they were unpiloted.

  3. Leland

    As I note in the book, a four percent loss of aircrew per mission in WW II was considered acceptable by the Army Air Force.

    You could also note that in his book, Failure is not an Option, Gene Kranz stated they expected to lose at least one of the 7 Mercury astronauts.

  4. Trent Waddington

    The reason Shuttle was a failed program was because it was a government program. Period. It served no purpose, as is often the case with government programs, except those purposes that were made up after the money had already been spent.

  5. Edward M. Grant

    Concorde wasn’t grounded because it was unsafe, it was grounded because it was a thirty-year-old supersonic cruiser that was uneconomical to maintain given the passenger numbers. British Airways restarted commercial service after making safety improvements — just as any other airliner involved in a fatal crash — and only stopped flying them in the post-9/11 airline slump.

    Richard Branson offered to continue operating them as part of Virgin Atlantic, but I don’t know whether he was serious or just causing controversy as advertising for his airline. I suspect the latter.

    1. Leland

      I agree mostly, but there were some environmental factor as well. A few, but loud, members of the public complained of the dangers and its environmental/carbon foot print. At the time when climate changers were in full envangelism, the Concorde crash (Air France btw, not British Airways) came at a convenient time to squash that decadent aircraft. The uneconomics came mostly because the enviros made it so.

    2. Rand Simberg Post author

      I think that safety was a factor as well. Like Shuttle, it did have some inherent design flaws. Of course, the joke at the time was surprise that the French would have shut down such a useful system, after they’d finally figured out how to kill Germans.

  6. K Lundermann

    It certainly should have been the Shuttle’s cost that lead to its termination, but I wonder whether that really was the reason. Politicians and NASA fear astronaut deaths to an irrational degree. Do we know this fear wasn’t the dominant factor among the decision makers? After all, the Shuttle’s high costs were due to its supporting standing army, and what space-state politician doesn’t like standing armies?

Comments are closed.