11

Air travel is the safest method of transportation, with 0.07 deaths every billion passenger-miles. While space travel certainly has a higher fatality rate per-space-traveller (~3%)1, 2, where does it rank when we look at passenger-miles, especially given the vast distances travelled (people on the ISS for long periods of time)?

The statistics for other methods of transportation appear to only consider fatalities, so that is good enough for me – although I'd be interested in how many non-fatal injuries there actually are in space travel!

Nathan Tuggy
  • 4,566
  • 5
  • 34
  • 44
w8ite
  • 119
  • 3
  • You may take a look at comparision like this one – Manu H Apr 05 '18 at 05:48
  • 4
    Deaths per passenger-miles skews the result in favor of space travel because speeds are so high. Deaths per flight hour may give a better comparison. – Hobbes Apr 05 '18 at 08:22
  • 3
    Generically, I think the per-passenger-mile comparison is only useful for destination-to-destination travel (i.e. what's the safest way to get from A to B). Space travel doesn't really fit because its B destinations are generally unique (unless you want to say e.g. flying the Space Shuttle is a shockingly unsafe way to get from LC-39 to the CCAFS runway). – Erin Anne Apr 06 '18 at 19:47
  • 3
    @ErinAnne It's not even a fast way to make that trip! ;) – Russell Borogove Apr 06 '18 at 20:14
  • @Hobbes For a comparison, maybe, but the comparison is apple and potatoes anyway. The absolute best metric for space travel safety is probably death per flight. – Antzi Apr 15 '22 at 05:12

3 Answers3

10

According to WP:

The number of deaths per passenger-mile on commercial airlines in the United States between 2000 and 2010 was about 0.2 deaths per 10 billion passenger-miles

The ISS is by far the largest contributor to crewed spaceflight time in the 21st century, being continuously occupied by 3-6 people for over 17 years since November of 2000, during which time it's achieved somewhere between 8 and 16 billion passenger miles†. Shuttle flights from 2000-2011 make up something like 10% as much as that; Soyuz flights are accounted for by my ISS crew estimate, and other crewed spaceflight missions are negligible.

Over the same time frame, 7 astronauts have died (the crew of Columbia on STS-107).

That makes space travel about 20-40 times more dangerous than airline travel by passenger-mile, but it's worth noting that all the astronaut fatalities in the 21st century occurred in that single incident, so we're clearly dealing with statistical outliers here; the "real" hazard rate could be much higher or much lower. In the decade 1991-2000, there were no astronaut deaths, making space flight infinitely safer than air travel; in the decade 1981-1990, there were 7 deaths in the single Challenger incident with far fewer space-crew miles flown to amortize over.

As Hobbes notes, it might be more appropriate to consider the risk by passenger-hour rather than by passenger-mile. In this case, assuming an average of 500 mph for airlines, we get ~0.01 deaths per million passenger-hours. There have been somewhere between 500,000 and 1 million passenger-hours spent on ISS, and again 7 deaths. This makes space travel 700-1400 times as deadly per passenger hour.

† 26500 miles per orbit * 1440 minutes per day * 6360 days / 92.65 minutes per orbit = 2.6 billion miles per crew member.

Russell Borogove
  • 168,364
  • 13
  • 593
  • 699
  • Are you using the ISS as your only reference point because it's too difficult to determine how many miles manned rockets have traveled or because you expect the contribution from manned rockets to be too low to be significant? If the former, you should make that an explicit assumption which would result in deaths per mile being overestimated. If the later, you should make an attempt to provide some evidence. As it is, all I see is ISS miles traveled divided by deaths in rockets, which doesn't seem to make sense since those are two different vehicles. – Ellesedil Apr 06 '18 at 16:07
  • The contributions from other crewed flights are small in comparison with ISS. All the shuttle flights from 2000 on totaled a little more than 1 year in space with 6-7 crew each, working out to something like 10% of the ISS numbers (and much of that already accounted for by ISS crew time); all the Soyuz flights were taking crew to/from ISS, so are accounted for by those numbers; China's crewed program amounts to noise in comparison. As I'm Fermi-estimating over a 2:1 range anyway, none of those matter. – Russell Borogove Apr 06 '18 at 16:23
  • 1
    Made that assumption more explicit. – Russell Borogove Apr 06 '18 at 16:29
  • 1
    Thanks. I'd argue that a 10% difference in miles travelled is significant, especially if you assume the ISS has allowed for 16 billion passenger miles travelled, but it doesn't change the conclusion much since the difference is already so vast. – Ellesedil Apr 06 '18 at 16:36
  • 1
    Because other components in my estimate vary over a 2:1 range, 10% isn't significant. https://en.wikipedia.org/wiki/Fermi_problem – Russell Borogove Apr 06 '18 at 16:38
  • 1
    Nice answer, my intuition was wrong on this one. – Organic Marble Apr 06 '18 at 19:36
  • 2
    @OrganicMarble We just fly a ridiculous number of airplanes -- 90,000 flights a day full of people compared to ~5000 orbital launches ever and only a few hundred crewed. – Russell Borogove Apr 08 '18 at 02:30
  • Unpowered flight (the ISS) is vastly different from powered flight. Comparing the two is frankly risable. – Pieter Geerkens Aug 18 '21 at 23:17
  • 2
    @PieterGeerkens Another answer, based on powered-flight time, would be welcome, I'm sure. Remember to omit STS-107, Soyuz 1, and Soyuz 11 from the fatalities. – Russell Borogove Aug 18 '21 at 23:36
  • this is stupid... ISS inhabitants are not traveling anywhere, their passenger-miles is zero. it might be reasonable to calculate passenger-miles for spacecraft trips to and from ISS, presumably with horrible results (compared to airliners) – szulat Aug 20 '21 at 21:22
  • 1
    @szulat Feel free to write an answer based on other criteria. – Russell Borogove Aug 20 '21 at 21:27
  • "0.01 deaths per million passenger-hours" I take this with a grain of salt. Deaths due to catastrophic accidents, maybe. But for this statistic to be true, the normal mortality rate indicates your specimens have an average lifespan exceeding 11400 YEARS, and only died of natural causes, no accidents. Or does boarding a plane infer complete immunity to death from strokes, heartattacks, emoblisms, aneurisms? (all of which should be higher than normal, in a reduced air pressure, shaking, noisy can with limited personal movement space!!) – CuteKItty_pleaseStopBArking Aug 22 '21 at 09:47
  • @CuteKItty_pleaseStopBArking None of that seems relevant to the comparison of air travel and space travel, but feel free to write an answer accounting for natural deaths. – Russell Borogove Apr 13 '22 at 20:06
3

This is an interesting topic but fraught with well-known difficulties. It is as well to look at the aviation example in more detail before plunging into trying to fit space travel into the same analysis.

The use of the "per passenger mile" is a notoriously poor comparison for a topic that is surprisingly hard to find a single fair comparison, but nevertheless its used in airline marketing because they think people will just accept what they say.

Most air accidents occur on taking off or landing and insurers act accordingly in their rating as this New Scientist article 2014 mentions.

One might argue that if the whole point of aviation is to travel distance then there is no alternative to looking at the distance traveled. However there are usually several ways of achieving an acceptable outcome for whatever motivated the travel in the first place and so the fundamental reasons are always a balance of the greater cost and greater inconvenience. Holidays to sunny destinations are a good example, where the objective is just "somewhere sunny".

At the practical level of the calculation for short haul air travel one typically has to journey for longer than the flight itself just to get to and from the airport, but the motor accidents that occur around busy airports aren't counted against air travel.

Finally on to space travel. The Wiki article posted by @Manu H provides this table wchich it cites as for the UK 1990-2000 (where Space Shuttle obviously isn't from the UK!): https://en.wikipedia.org/wiki/Aviation_safety#Transport_comparisons

The use of the Space Shuttle in this table is obviously going to skew the results one way or another, just consider the following points

  • It was withdrawn by its own sponsor, NASA, for being unsafe
  • There are no other systems like it, now or in history. i.e. it is not representative of spaceflight as a whole
  • Just on calculations, should one only count the astronaut time/km spent in the shuttle rather than the time/distance travelled by those astronauts who stayed on the ISS and came back on a later flight? No one would count an accident on the beach as part of the air travel to get there, so there would have to be a metric that discounted those ISS hours/km from the Shuttle's assessment. The same goes for Soyuz/Mir, Soyuz ISS, Dragon/ISS etc.
Puffin
  • 9,514
  • 33
  • 55
0

The study was made, and published, by NASA at the insistence of Feynman after the Challenger disaster in 1986 (my emphasis).

Conclusions

If a reasonable launch schedule is to be maintained, engineering often cannot be done fast enough to keep up with the expectations of originally conservative certification criteria designed to guarantee a very safe vehicle. In these situations, subtly, and often with apparently logical arguments, the criteria are altered so that flights may still be certified in time. They therefore fly in a relatively unsafe condition, with a chance of failure of the order of a percent (it is difficult to be more accurate).

Official management, on the other hand, claims to believe the probability of failure is a thousand times less. One reason for this may be an attempt to assure the government of NASA perfection and success in order to ensure the supply of funds. The other may be that they sincerely believed it to be true, demonstrating an almost incredible lack of communication between themselves and their working engineers. ....

Challenger was shuttle flight number 51 of 133 total. Feynman predicted a failure rate of about 1%; and of the 82 subsequent flights 1 failed. The distinction between flights before and after Jan. 1, 2000, is a red herring: as no meaningful changes in regards safety were ever made to the program's safety between its start and conclusion.

Pieter Geerkens
  • 117
  • 4