When I'm watching SpaceX launches, my biggest concern is to avoid a big boom. Specifically, an explosion that would threaten the astronauts who will eventually ride the Falcon 9.

So around stage separation, I often find myself wondering how much I can relax now that the first stage is done?

It's easy to find examples of first stage failures. Most of the stories people tell result from first stage issues. But it's hard to find examples of second stage failures, especially filtering for ones that could harm a passenger, not just "failure to achieve specified orbit".

I'm looking for some kind of numbers or sense of the failure rate of first stages compared to later stages. At least some historical examples? Is my perception correct that the first stage is the most dangerous?

  • 3
    $\begingroup$ There is also the case of failures that originate in the second stage, but happen before stage seperation, or even before launch. The falcon 9 has had two such failures (amos-6 and crs-7) $\endgroup$ – Peter Green Aug 7 at 0:11
  • $\begingroup$ @PeterGreen Yeah, I guess that's a good point. Not sure what category to put that in. I guess what I care about most is failures that happen before the first staging event vs. after, no matter where the cause originated. I guess what I'm looking for is the point in time I can say "okay, well the worst is over!" $\endgroup$ – Nick S Aug 7 at 3:09

Many entries in this 1957-2015 Catalogue of Launch Vehicle Failures mention stage numbers. By my tally, the blame for launch failure goes to the first stage 42 times; the second, 60; the third, 36.

If you restrict the reasons for failure to those that might apply to Falcon 9 (one entry is a Falcon 9, btw), or if you discriminate between explosions and less dramatic events, your tally might differ.

  • 1
    $\begingroup$ Hard data! Awesome, +1 $\endgroup$ – Organic Marble Aug 16 at 16:43

The first stage filters out a lot of the mistakes.

  • Pretty much the entire stack is tested to some extent during the first stage. This includes the hardware being subject to vibration and force as well as software for guidance etc.

  • They are also subject to the most environmental factors, which are harder to predict and test than isolated hardware.

  • They are also bigger, with more or larger engines and rockets. Hence even if they are just measured engine for engine, a F9 with 9 S1 engines and 1 S2 engine will 9 times out of ten fail on the first stage.

All told it's not reasonable to expect there would be a similar number of failures across the stages. So 'yes', your intuition would seem reasonable.

  • 1
    $\begingroup$ Do you have any historical data on relative numbers of 1st / 2nd stage explosions? That is what the question is asking for. $\endgroup$ – Organic Marble Aug 16 at 13:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.