You know, I can think of very few people in whose opinions I have less interest right now, with the possible exception of Mike Griffin, than Scott Horowitz. The sad thing is, he probably actually believes this:
To help address the safety and reliability issues, SAIC was commissioned to evaluate the potential hazards of the first stage solid and the overall reliability of the vehicle. The SAIC study (SAICNY05-04-1F) showed that a worst-case scenario of a catastrophic case burst of the first stage (extremely unlikely) would result in a maximum overpressure at the crew capsule of approzimately 1 psi. This overpressure is well within the design characteristics of the capsule. SAIC also conducted an initial Probabilistic Risk Assessment (PRA) evaluation of the reliability of the launch vehicle and estimated a launch vehicle failure rate (LOV) of 1 in 483 and a loss of crew rate (LOC) of 1 in 3,145 at the mean of the estimated uncertainty distribution. Recent NASA PRA estimates for the current configuration predict a LOC of approximately 1 in 2,500. This compares to the LOC for Shuttle of 1 in 88.
“1 in 3,145”?
Really? Not “1 in 3,144” or “1 in 3,146”? And did they include all of the TBD gimcrackery that was going to be incorporated to keep the vehicle from shaking itself and Orion apart?
And how many centuries did he expect this vehicle, that would cost billions of dollars each flight, to fly in order to determine whether the genius rocket scientists at SAIC got the numbers right?
It would have a little more credibility if he had at least said one in three thousand (though not a lot), and not made the elementary first-year physics student’s mistake of overprecision.
But you don’t get safety by doing PRAs. If we learned anything from the Shuttle, we sure as hell should have learned that. You get safety from flying. A lot.
But the other thing that is disturbing are the requirements:
Do not compromise crew safety for cost, performance, or schedule
This requirement, taken to its logical conclusion, would keep us on the planet forever. Don’t compromise it for cost? OK, then it becomes unaffordable because we can’t compromise “safety.” Don’t compromise for performance? OK, then the job doesn’t get done. Don’t compromise for schedule? So when do we fly?
This was a system of the astronauts, by the astronauts, for the astronauts. Which says to me, we need a different kind of astronaut, one who takes their job, and its purpose, seriously.
Life, and engineering are compromises. If safety becomes the ultimate value, then you might as well stay in bed (assuming that someone doesn’t decide to pump poison into your bedroom, or a meteroid doesn’t come crashing through the roof). And the irony, of course (again, as we should have learned from the Shuttle) is that you don’t get safety by spending billions in the attempt. The only way to get some level of safety is to do something a lot, make mistakes, and learn from them. In fact, if Shuttle were still a fully-operational program, with new parts being produced, and a reasonable flight rate, it would be safer now than at any time in its history, because we learned a lot from Challenger and Columbia. But it’s not, and we can’t afford to fly it enough to make it truly safe.
The lesson here is that if you want safety, don’t avoid reusables — reuse them, a lot. But the nonsensical lesson that many (including the CAIB) seemed to take away from Columbia, in addition to unrealistic safety requirements, is that we should return to throwing away the vehicle so that each flight is a first flight, and then put a heavy, expensive escape system on it when it fails anyway.
People with such an attitude are fundamentally unserious about space. The unwillingness to risk the lives of astronauts says that what we are trying to accomplish in space is unimportant. As long as that is the case, it will remain unaffordable, and we will accomplish little. Today’s announcement is, I hope a first step toward a more sane, and realistic approach to human spaceflight.