TCS Daily

Not Unsafe At Any Speed

By Rand Simberg - November 19, 2004 12:00 AM

Alexander Tabarrok's pessimistic and sobering assessment of the prospects for the space tourism industry is seemingly well argued and persuasive.

It is also quite wrong.

It rests on false premises (though understandable ones, since most of our culture has been inculcated with many mistaken myths about space and space technology over the past four-plus decades).

Professor Tabarrok assumes that there are characteristics of spaceflight that render it intrinsically unsafe. He also assumes that SpaceShipOne and its competitors are on a continuum with past efforts, and that their safety and cost can be extrapolated from those.

These are not unreasonable assumptions, given the history of rocketry that he ably analyses and presents. Nonetheless, they are mistaken. Spaceflight is no more intrinsically unsafe than it is intrinsically expensive (another myth), or at least it does not have to be as dangerous and costly as one might rationally conclude from the limited empirical data.

I say limited data because one of the reasons (in fact, the key reason) that rocketry is unsafe and expensive compared to other forms of transportation is because we have so little experience with it, compared to other forms of transportation. More people take off in a single flight of a Boeing 747 than have flown into space to date. Any major airport will have more takeoffs and landings in a single day than the number of non-munitions rockets launched throughout history. When it comes to spaceflight, we are still at the bottom of a steep learning curve. It thus shouldn't surprise us that we aren't very good at it yet, but that doesn't mean that we cannot become so, and much faster than would be indicated by Professor Tabarrok's mistaken extrapolations.

Specifically, there is a mistaken and quite misleading extrapolation when Tabarrok writes here:

"The space shuttle has a slightly better record of safety -- it was destroyed in two of 113 flights. There are lots of millionaires willing to spend one or two million dollars for a flight into space but how many will risk a two to five percent chance of death?"

It might seem reasonable to infer the safety of a suborbital passenger vehicle from the safety record of the Space Shuttle, but it is in fact absurd to do so.

The Space Shuttle is an orbital vehicle. The difference in velocity between its flights and those of a suborbital vehicle, such as SpaceShipOne is a factor of four or five. The difference in energy (and power) on a unit basis is that number squared. The much lower requirement for the suborbital vehicle allows much greater margins, in performance and structure, which leads to robust and safe designs, until we gradually understand how to take such vehicles higher and faster to orbit down the road.

Moreover, the Space Shuttle is in fact a quite unsafe design, incorporating hazardous expendable elements, and is not necessarily representative of a tested, truly reusable vehicle, orbital or otherwise. The fact that it's one of the few real-world benchmarks that we have doesn't make it a good one. In fact, the Russian Soyuz vehicle has a much better safety record, particularly in recent years, though it's not an exemplar either. But there's no reason to think that launch systems much safer than the Shuttle and Soyuz can't be built, particularly given the much greater knowledge base that we have now, relative to three decades ago when they were designed. Thus, it's not sensible to attempt to infer the risk of flying in a twenty-first-century space tourism vehicle from either of them.

Further, when he points out that

"In its first trip...a control system failed and in another flight SpaceShipOne rolled 29 times."

he fails to realize that he's actually making the case for the safety of SpaceShipOne. It suffered a control failure, and went into a series of rolls, and yet the pilot recovered from these failures in both cases, achieved the necessary altitude to win the prize, and returned to earth completely unharmed with vehicle intact. This is the mark of a robust, safe design with the ability to recover from a failure, not an unsafe vehicle, and it in fact has a perfect safety record.

Of course, the most misleading extrapolation is the one that predicts airline-like safety for rocketry over two hundred years from now. This assumes that the learning is continuous, and continuously slow. But in fact, SpaceShipOne and its brethren represents a complete discontinuity from the old ways of doing business. The rockets of yesteryear (including much of the Shuttle itself) are expendable, and each flight is a first flight (and a last), with associated infant mortality. The new rockets will be fully (or at least mostly) reusable, and cheap enough to operate that many test flights will be flown to fully understand their characteristics and performance with gradual envelope expansions (as was the case with SpaceShipOne).

It is an entirely new and different approach than the one of the traditional space industry, and while we can't say for certain how safe vehicles developed in this manner will be, we can confidently predict that their safety record will bear no discernable relationship to the safety record of converted, expendable munitions, and that extrapolations such as those presented by Professor Tabarrok are less than useless, because they present a false sense of the risks to potential investors.

Richard Branson is many things, but few would accuse him of being unwise with his money. He has been observing this industry for many years, and has only now stepped forward to invest in it. He is doing so because he understands that everything changed last month when a private astronaut affordably and safely flew above the atmosphere, in the very same vehicle that had performed that feat previously. It was the beginning of a new paradigm for vehicle development in which, rather than building something that goes to orbit and then figuring out how to make it cheap and safe, we'll be building affordable and reliable vehicles, and then gradually expanding their capabilities to get them to orbit. Further, this will be done with a number of test flights that will dwarf all of the numbers of launches to date, completely throwing Professor Tabarrok's logistics curve into a cocked hat.

This new industry will not be risk free -- no new industry, particularly a new transportation industry is, or can be. But the risks are understandable, and manageable, and the danger to the industry is not the safety of its vehicles per se, but zealous overregulation and underinvestment based on false fears. I remain hopeful, however, that it will overcome these hurdles as well, despite well-meaning columns like those of the good professor.

Rand Simberg is a consultant and entrepreneur in commercial space, space tourism, and internet security. He publishes a weblog, Transterrestrial Musings.


TCS Daily Archives