I just posted this in comments over at NASA Watch, in response to a foolish comment there:
Space X boosters need to become reliable cargo transports before they can be trusted with manned space craft. And with only a 2 out of 5 success rate, they have a long way to go!
I guess this is the new mantra of the simple-minded commercial space bashers — the implication that Falcon 1 is only 40% reliable because of its “2 out of 5 success rate.”
Here is the real story. The vehicle had a flight test program as part of its development. The purpose of flight test programs is to wring out design issues, and it shouldn’t be surprising to have some failures in that process.
The first flight barely got off the pad. They figured out what was wrong, and did a second flight. That one had some slosh dynamics issues. They figured out how to fix that and had a third flight. It had a separation problem because they hadn’t accounted for the longer thrust tail-off of the upgraded engine. They fixed that with a minor software change, and the two flights since were perfect.
In other words, they had some teething problems, but have solved them, and now have a reasonably (certainly more than 90%, and probably high nineties) reliable vehicle.
Remember this as you see these morons continue to say “2 out of 5.”
[Afternoon update]
Clark Lindsey has further thoughts.
Rand,
It’s ok. Remember if the mantra is any more complicated than “Bigger is Better” you’re going to lose some people.
~Jon
People also have trouble accepting the fact the first two flights were test flights where reaching orbit was only a *secondary* objective on the predetermined objective list. They hence laughed at Elon when he labeled the flights as successful tests.
It was only after those two flights that Falcon 1 was declared operational and in that sense, flight 3 was the only true operational failure. People claiming the 2/5 record will just keep moving the goal posts, be it 2/6, 2/10 or whatever. I hope they don’t even get that chance with Falcon 9.
Well, flight 5 was the first successful commercial launch and even it didn’t go entirely flawlessly. But your bigger point is valid of course: the problems are getting smaller.
Regarding foolish comments: not all of the posters are foolish, some are simply disingenuous. Looking back on the past year I’ve come to the conclusion I’ve spent far too much time debating liars and hypocrites in the vain hope of convincing them. That makes me the fool, not them. I’m sure Jon will have a less confrontational and more generous point of view on that, but this is how I see it.
I’m not sure I would quibble with his numbers. When the average fail rate differs strongly from the marginal fail rate, then indeed one can argue the vehicle ought to demonstrate itself on more flights before being trusted with valuable cargo. This is why I don’t buy cars in the very first model year, nor the very first release of a new operating system. It’s reasonable.
Where you might take issue with him, however, is his disingenuous use of the passive voice in “they can be trusted.” Trusted by whom, eh? Isn’t that a very important question? There are certainly individuals who would be willing to ride Falcon 9 to orbit on its first flight, without a single test. For them, the risk is worthwhile. There are other individuals, of course, who would never ride a lit firecracker no matter how much you paid them. Most of us are in between.
But shouldn’t that be an individual decision? This fellow appears a disciple of the new collectivist insect mentality, where all of us (although he really probably means “a select superior subset of us, including myself as it turns out, who will represent all of us”) make decisions for each of us. You see this all over the social map, of course, from the foolish debate about how much “we” should spend on medical care — as if “we” have any damn business looking over someone’s shoulder as he writes a personal check to his doctor — or how much “we” should pay to save the environment, or, in this case, whether “we” should decide for some putative astronaut whether he is allowed to trust a particular rocket to boost his ass to orbit.
I’m sick of it. I’m sick of all this fucking “we” nonsense. There are a very few cases where indeed “we” have to make decisions, but this is America, and I thought we had agreed long ago that we would keep the we’ing to a minimum, that in general each “I” among us could decide stuff for himself. Long live individual liberty, and death to the Borg.
Let those who want to loft cargo or people decide each for themselves whether to trust a rocket or not. There is no need for a national consensus. And people who are not actually riding the thing themselves. or paying to put cargo on it, should STFU.
Well, flight 5 was the first successful commercial launch and even it didn’t go entirely flawlessly.
@Martin: Would you care to explain what was it that was off-nominal on flight 5?
The vibration issues that delayed the launch and didn’t leave any room for a secondary payload. It was not a major problem and it shows that the trend is ever smaller problems. It does mean they haven’t perfected everything yet, but then again it would not be fair to expect that of them.
Even seasoned space launch vehicle design teams make vehicles which have one failed flight at the beginning of the launch campaign (e.g. Ariane 5 V-89, Ariane 5 ECA V-157, Delta IV Heavy D310) before unanticipated issues are fixed. Atlas is one of the few vehicles escaping this state of things.
Falcon 1 has had many design changes. It can be considered that the launches were made by at least two different vehicles using Merlin 1A and 1C engines. The Merlin 1C engine version of Falcon 1 has had 2 successful flights out of 3. This is in line with figures from established teams.
Taking this into account the numbers for Falcon 1 reliability can easily be in the range Rand mentioned. SpaceX has however decided to abandon Falcon 1 for Falcon1e and I would not be surprised to see a launch failure in that first flight. Falcon 9 covers a lot of new ground for them (“new” second stage engine, lots of engines in the first stage) so it may be more bothersome. But once they work out the bugs the reliability will go up again.
Godzilla,
Only one failed flight? Let’s jump in the Wayback Machine and compare apples to apples. Let’s look at the beginning of the Atlas launcher and it’s success ratios;
Atlas A – 8 launches, 4 failures
Atlas B – 10 launches, 6 failures
Atlas C – 6 launches, 3 failures
Next let’s look at the beginning of the Delta (a.k.a Thor-Delta) launcher and it’s success ratio;
10 launches, 4 failures
For Falcon 1 to ONLY have a failure rate of 2 in 5 is PRETTY GOOD.
Falcon 1 has a failure rate of 3 in 5. And I don’t see where Rand’s certainty of reliability above 90% comes from.
Karl, if you didn’t read what I wrote, I don’t know how else to explain.
Not that I expressed a “certainty” of anything.
[Update a while later]
OK, I did express a certainty of 90+%. I stand by it, based on the last three flights.
Wow Rand, you didn’t even get a “GRATUITOUS INSULT” label from Keith for your “simple-minded commercial space bashers” comment…..
So, from a PR point of view, should Elon have spec’ed the mission to reenter before the first orbit on the first two launches? If he had, they would have been unequivocally test flights.
But then I guess people got excited about SS3 having an issue shooting touch and goes during early flight test.
Some people are just not reasonable, it seems.
I’d like to offer a distinction: the 2 out of 5 notwithstanding, Falcon’s short launch history shows SpaceX are on the right track. However, it is too early to tell what the reliability numbers will be. I don’t think it would be unreasonable for NASA to expect them to have done a bunch of absolutely flawless launches in a row before they start launching their people on it. And the first absolutely successful Falcon 1 has yet to fly, in fact it may never fly entirely flawlessly since it is about to be retired. Falcon 9 has never flown at all. All in all I’m optimistic, but there’s still a long way to go.
I don’t think it would be unreasonable for NASA to expect them to have done a bunch of absolutely flawless launches in a row before they start launching their people on it.
And yet that same NASA should have no problems putting humans on top of the Ares I after only 3 flights, one of which is a bottle rocket and not a real test? I fail to understand how some people seem to have double standards about this. It needs to be pointed out that no hardware planned to be flown on Ares I will have any flight history or heritage. It is a new vehicle and not any less so than Falcon 9.
If people are ever committed to a Falcon 9, it will be after at least half a dozen flights. The Dragon capsule itself would have nearly as many shakedown opportunities.
Ares I + Orion? Not so much.
And yet that same NASA should have no problems putting humans on top of the Ares I after only 3 flights, one of which is a bottle rocket and not a real test?
No, Ares I should be cancelled. 🙂 The same principle should apply to NASA if it were to build a new launcher, there shouldn’t be a double standard. Then again, NASA shouldn’t be in the launch business at all. It would be much better if it were structured the way NACA apparently was or the way I believe ESA is.
Perhaps Elon should have called his first three or four flights x-1 to x-4 then retroactively called his first success Falcon 1, flight 1. Then he could claim a 100% success rate.
Rand, I suppose I’m just being cranky. But I recall one of the rationalizations for Ares I was its supposed safety and that was based on assumptions like that the SRBs had a failure rate of much better than 1 in 3000. As far as I can tell, the historical record for SRBs has one failure in around 300 uses of the SRB. Prophesying high reliability rates based on little data looks like trouble to me. Hopefully you are correct though.
Prophesying high reliability rates based on little data looks like trouble to me. Hopefully you are correct though.
I think that my estimate is much more justifiable than 40% is. It’s possible that they were just lucky on the last two flights, but I doubt it.