Dangerous business

SpaceX blew up on the launch pad this morning. There were no casualties, at least, and we do get a spectacular explosion video out of it, but this is an unfortunate setback.

Human beings sometimes sit on top of those kinds of infernal devices? I don’t think I could do that.


  1. Elladan says

    The actual human-rated version has escape rockets and plausibly could have saved the astronauts. Maybe. (As far as I know, all crewed rockets have had escape systems, except for the Shuttle)

    The closest example would be a Soyuz mission where the rocket exploded on the pad as their Soyuz capsule blasted them to safety. From Wikipedia:

    The two crew members were badly bruised after the high acceleration, but were otherwise in good health and did not require any medical attention. Upon being greeted by recovery crews [….] the cosmonauts were then given shots of vodka to help them relax.

    Ah Russia…

  2. Nerd of Redhead, Dances OM Trolls says

    This is really a rehearsal for the ground and launch crews. The second stage LOX tank was filling prior to the test firing of the booster engines. It appears to me the failure started at the top of the second stage. Any reductant with a large concentration of oxygen has two out of three sides of the fire triangle present. Once started, aluminum will burn readily, and it is used for most of the rocket.

  3. brett says

    Whenever a major launch failure like this happens, it reminds me of astronaut Don Pettit’s essay on how difficult launches are:

    Rockets at 85% propellant and 15% structure and payload are on the extreme edge of our engineering ability to even fabricate (and to pay for!). They require constant engineering to keep flying. The seemingly smallest modifications require monumental analysis and testing of prototypes in vacuum chambers, shaker tables, and sometimes test launches in desert regions. Typical margins in structural design are 40%. Often, testing and analysis are only taken to 10% above the designed limit. For a Space Shuttle launch, 3 g’s are the designed limit of acceleration. The stack has been certified (meaning tested to the point that we know it will keep working) to 3.3 g’s. This operation has a 10% envelope for error. Imagine driving your car at 60 mph and then drifting to 66 mph, only to have your car self-destruct. This is life riding rockets, compliments of the rocket equation.

  4. Matrim says

    Human beings sometimes sit on top of those kinds of infernal devices? I don’t think I could do that.

    Whatever, dude. If it were in any way financially possible for me I would be on a rocket so fast I might achieve escape velocity before launch.

  5. zetopan says

    For anyone interested, the delay from the visual start of the explosion to the resulting sound indicates that the camera was located 2 miles from the launch pad.

  6. says

    See… I’ve already decided a long time ago that I don’t want to live on this planet anymore, so… if I could be sure of it’s relative safety, I would definitely be on a one-way trip to Mars if I could…

  7. says

    I wonder how the death ratio compares to cars.

    It’s a lot higher. Your chances of dying in any one space launch are around 1 in 50. The average driving distance of any one car journey is 5.95 miles (from the NHTSA) and the death rate is currently 1.08 fatalities every one million miles traveled. That comes to one death around every 155,000 car journeys.

    So yeah, over 3,000 times safer to go by car, on average (though I guess if you calculate the risks using miles traveled instead, the numbers are a little closer…

  8. zetopan says


    Yes, I was somewhat incorrect since I didn’t actually use a timer but rather estimated the time interval at 10 seconds. Thank you for pointing out my error. However, like all science, getting more accurate and detailed results over time is important. Hence, you have forced me to point out that 13 seconds (which I agree with, since I used a clock this time) is closer to 2.8 miles than 2.6 miles, since the speed of sound at sea level and 68F is about 1,126 ft/s. Since Cape Canaveral is located at about 10 feet above sea level, using the sea level speed of sound seems reasonable.
    Correcting for temperature at that time and place (80F vs 68F) results in about 1138.5 ft/s. There is also an even finer adjustment to account for the humidity since the standard speed of 1126 ft/s is for dry air. However I don’t know the humidity at that time and place so I cannot correct for that variable, but I can point out that greater humidity results in an even higher speed of sound than dry air when the temperature is above about 44F. So taking into account what is known from the launch site it would appear that 13.0 seconds would amount to 2.803 miles, assuming that there was essentially a zero wind speed.

  9. zetopan says

    Sorry, I just realized that I failed to put the trailing “3” (in 2.803) in parenthesis to show that it is not a significant digit while being the result of the calculation.

  10. Moggie says

    I’m a little concerned that Kerbal Space Program has taught me to think of such explosions as awesome and funny, rather than a failure and possibly a tragedy.


    Well, that does it… I’m not going into space until they have the space elevator.

    I’m not going in the space elevator until they settle on the elevator music. Also, imagine how much of a terrorism magnet a space elevator would be.

  11. applehead says

    RAH RAH NASA is an ossified, useless Big Gubmint relic and only Visionary(TM) Innovator(C) Job Creator(R) Elon “Real-life Iron Man For Realsies” Musk can save space travel!!!1ONE

  12. wzrd1 says

    Rocket, a bomb designed to slowly explode in one direction.
    The tricky part, keeping that explosion in only one direction.

    An example of how tricky it is can be well displayed in that Apollo 13 still had pogo oscillation problems in the SII stage, which nearly damaged the spacecraft before the engines were shut down.
    Even today, that can become a problem.
    Then, there’s slosh dynamics, which even today, can become a major problem.

    The list goes on and on, with many things that can result in the loss of a spacecraft. All, requiring intensive modeling and as was outlined above, intensive testing and even test launches.

    Designing such devices, well, it’s as hard as rocket science.

    Wow! I heard that groan all the way over here! ;)

  13. birgerjohansson says

    The insanely small margin for payload means ultra-light designs are used, there is not the usual margin of safety in the engineering – I have read of the launch vehicles as being constantly perched “on the edge of disaster”.
    — — — — — — — — —

    This is a loong stretch, but what the hell….

    “How to talk to a woman wearing an astronaut’s helmet” http://www.thedailymash.co.uk/features/how-to-talk-to-a-woman-wearing-an-astronauts-helmet-20160901113135

  14. birgerjohansson says

    brett, sorry, I didn’t read your post before posting, you are making the same point..

  15. rietpluim says

    @tacitus #10 – If you calculate casualties per travel distance instead of journeys, how different would that be?

  16. anchor says

    @#10 tacitus: As of now, with a total of 546 people who have reached orbit (some flying more than once) having put in a total of 133 man-years on the clock and a total of at least 33 billion man-km on the odometer, that comes to an average of about 60 million km per flight with an average duration of almost 90 days. A total of 18 known fatalities have occurred during missions (all happening during launch or reentry/landing), with a fatality rate of 1 in 1.8 billion km, or once every 7.4 man-years.

    Comparing this with the traffic fatality rate for the US, which is listed as 1 per 100 million vehicle-kilometers and 14.5 per 100,000 (population assumed to be car users).

    In terms of distance traveled, this suggests space travel is about 18 times safer than traveling by car.

    In terms of time its difficult to estimate how much time Americans spend driving their cars around, but if anyone can estimate that they can compare it to the 133 man-years so far spent in space travel, then see how the fatality rate per unit time compares.

    In terms of the population: the fatality rate is 1 in 30 astronauts versus 1 in 6900 by automobile drivers or passengers.

    So, the scariness of a transportation mode depends on what parameter (time, distance, population) one is paying attention to. I’d be too nervous to enjoy anything.

  17. zetopan says


    “You win..”

    Not really, and that was not my intent in any case. I was merely showing that Asimov’s short essay on “The Relativity of Wrong” plays a part here. My 2 mile estimate had a greater error than your 2.6 mile estimate, and similarly a 2.8 mile estimate is closer yet to the actual distance. We were together converging on a yet more accurate estimate of the actual distance.

  18. Crimson Clupeidae says

    It happens, no one was hurt, and while it is a big expense, they can learn a lot from failures. Much like Real (TM) science, failure is part of the process.

    This shouldn’t affect the Osiris Rex mission that’s due to launch next thursday. I’m excited about that mission (return a sample from an asteroid). The mission lead is a prof from my alma mater (and bonus: he made a cool board game based on his experience called Xtronaut).

  19. aziraphale says

    brett @5, that 10% margin of error is not as frightening as it may appear. These things are controlled by computers, which do not let their speed drift up in the way a human driver might do. You can see this from the way they land the first stage on a tiny barge, coming in very fast and flipping out the landing legs at the last minute. A human pilot couldn’t do that.

  20. Nerd of Redhead, Dances OM Trolls says

    Brett #5, the Falcon 9’s kerolox engines are higher thrust than older designs, and SpaceX chills the RP1 and LOX to increase their density, allowing for a higher mass loading in the same tank volume. This Video (alert, autoplay) at 6:45 says that for a LEO mission, the propellants were 90% of the mass. SpaceX loads 1.1 million pounds of RP1 and LOX in about 30 minutes.

    The fact finding has started.

    SpaceX is scouring computer and video data for clues to the devastating launch pad explosion that destroyed a rocket and satellite.
    Thursday’s accident occurred during a prelaunch test, eight minutes before the engines on SpaceX’s Falcon rocket were supposed to briefly fire. The rocket was being fueled when a huge fireball erupted.
    On Friday, SpaceX said it has begun reviewing 3,000 channels of computer and video data, covering a time period of just 35 to 55 milliseconds. The trouble appears to have originated somewhere near the liquid oxygen tank in the upper stage.

    Elon Musk has said initially it was a fast fire, not an explosion.

  21. Crudely Wrott, lurching towards recrudescence says

    There is not much to learn as of five o’clock this morning but I do want to point out something. A close examination of the only available video as of earlier today suggests that the initial ignition point was outside of the Falcon rocket. Watch here:


    At 1min 10sec the fire starts. Watch carefully, slow the playback down to x .25. It seems that the flash point is between the Falcon and the strongback tower that holds the feed lines to the rocket. That is, the explosion may not have originated in the rocket but rather in the interface between the Falcon and the launch pad support mechanisms.

    This is not official but a close examination of the video is highly suggestive that the SpaceX rocket was not at fault for this damned flustercluck.

    The rocket can be easily replaced; the satellite, not so easily though insurance may, I say may, cover the loss. My worry concerns damage to the launch pad.

    Should investigation show that the root cause was not onboard the Falcon there is still the issue of launch pad damage. Perfectly functional Falcons cannot launch from a broken launch pad. That is the factor that will have the most limiting influence on future SpaceX missions. There are, or were, at least six planned launches scheduled for this year including the initial demonstration of the Falcon Heavy and the first use of an “already been flown” booster to put a commercial satellite in orbit.

    It will take some days or maybe some weeks to ascertain the cause of the mishap and to determine the state of the launch pad. Do stay tuned to sites such as SpaceFlightNow, SpaceFlightInsider and SpaceFlight101 if you are interested.

  22. wzrd1 says

    @Crimson Clupeidae #24, wow! I missed that news. Going to orbit is complex enough, to the moon, orders of magnitude, as the complexities increase, to an asteroid to collect and return a specimen, the complexities are an order of magnitude higher at the least.
    Much greater temperature extremes, no magnetic shielding from earth, approach, interaction and sample retrieval with a significant time lag, while any processing error causing a disaster and more, to name a few.
    Best of luck to that team!

  23. Nerd of Redhead, Dances OM Trolls says

    The Osiris launch appears unaffected by the SpaceX anomaly.

    Mike Curie, a NASA spokesman at the Kennedy Space Center, said a few hours after the SpaceX incident that he was not aware of any effects the explosion had on OSIRIS-REx. The agency confirmed that assessment later in the day.

    “Initial assessments indicate the United Launch Alliance Atlas V rocket and OSIRIS-REx spacecraft are healthy and secure in the Vertical Integration Facility at Space Launch Complex-41,” NASA said in a statement distributed through social media.
    – See more at: http://spacenews.com/launch-of-nasa-asteroid-mission-unaffected-by-spacex-failure/#sthash.RuPd0OC0.dpuf

    SpaceX will shift Cape Canaveral launches to SLC 39A, where has been reconstructing the Apollo/Shuttle site for their manned launches and Falcon Heavy.

    With its launch pad likely facing major repairs, SpaceX said it would use a second Florida site, called 39A, which is located a few miles north at NASA’s Kennedy Space Center and was used for space shuttle missions.

    The pad is on schedule to be operational in November, SpaceX said. The company had planned to use the pad for the first time later this year for a test flight of its new Falcon Heavy rocket.

    NASA spokesman Michael Curie said in an email that the site could be used for commercial and government flights, and SpaceX President Gwynne Shotwell in a May conference said one customer, SES SA of Luxembourg, had expressed interest in flying from the historic launch pad.

    I looked at the film frame by frame, and I concur with Crudely Wrott, the fire appears to be centered on an umbilical dangling between the strongback and rocket. The umbilical appeared to be venting, and I don’t know if it normally would be. Over three frames at 30 frames per inch, it went from no fire, to a large vertical flame (estimate 20-30 meters high, and 5-10 meters wide) to a fireball engulfing the top of the second stage.