What are the odds of a safe rocket launch?

What are the odds of a safe rocket launch?

In Rockets, Videos by Paul ShillitoLeave a Comment


After watching the SpaceX Starship explode, the space industry could be seen as high risk but which  are you more likely to be killed in, a car journey or a rocket going into orbit?

Well, if you are in the UK which is ranked 7th in the world for the safest roads, your chances are 1 in 200 or 0.5% lifetime risk.

The all-time failure rate for launches into orbit over the last 75 years is 1 in 50 or 2%, and that falls to 1 in 100 or 1% for total launch failures before reaching orbit.

That’s also assuming that there were no safety systems to launch the crew cabin away from an explosion on or near the launch pad or a parachute system if the incident occurred at a high altitude, which all human-rated vehicles had apart from the space shuttle.

Rockets rarely fail gracefully and when they do it usually happens very rapidly.

If the engine in a car stops you can just pull over, in many aircraft if the engine fails you can glide but if a rocket fails then it won’t be going anywhere fast except towards the ground and usually in few thousand pieces.

So having SpaceX Starship explode mins into its first launch is not as uncommon as you would think  for a new rocket system but quite a rare occurrence on an established one.

So let’s see just how safe rockets are and why things sometimes go wrong.

The use of rockets to get objects and eventually people into orbit officially started 4th Oct 1957 with Sputnik though they had been trying since the 1940s. This early era was when rockets were as likely to blow up the launch tower as get past it.

Not a lot is known about the Soviet failures because they rarely acknowledged them but on the U.S. side, everything was filmed.

The first attempt by the U.S. to launch a satellite was on the 6th December 1957 and was also broadcast on live TV. This was the Vanguard TV3 which was overseen by the U.S Navy. This managed to travel just over a meter before it stalled and fell back with the resultant explosion severely damaging the pad. The exact cause was never determined but the fuel system was suspected, and subsequent engines of the same rocket model were modified without further issue.

The lack of an exact cause would often be the case in the early days. There was much less in the way of monitoring equipment and sensors because many hadn’t yet been invented so it was as much down to the filming of the launch to give the engineers clues as to what happened.

It would take the arrival of reliable data acquisition and solid-state sensors before their widespread could be used and a much finer level of detail could be achieved to know what failed and when.

Determining the reliability of rockets, especially ones carrying expensive payloads became the job of small number of specialist actuaries to determine the risks involved to provide information for insurance companies covering the launches.

The Intelsat-1 of 1965 was launched into space by the Hughes Aircraft Company for COMSAT. It was the world’s first satellite to be covered by an insurance policy, underwritten by Lloyd’s of London to cover physical damages on pre-launch but not for the launch itself.

An actuary is someone that analyses the financial costs of risk and uncertainty, these are the people that will be able to tell you what the risk going into space will be.

According to the stats, on the first or second launch of a new rocket you would expect around 30% to fail but things start to get better afterward and by the time you’re up to the 10th flight that risk would be around a 5% failure rate or less.

On the first few launches, there will be a lot of effort put in to make sure everything is right but with new systems integrating engines, fuel and control you just won’t know how well these play together until they have several launches under their belt.

But there is a problem here. Spacecraft and rockets are built in very low volume and are often built to do one specific task, as such most are ones offs and they don’t benefit from the mass production that say cars do which makes them not only very expensive but more prone to failure without extensive testing.

If we look at cars, a particular model of car will have a certain set of parameters that will determine how it might act in similar situations, this is how we can derive a safety rating like Euro NCAP for each model.

The problem with the earlier space vehicles was that they were hand made and no two were identical even if they were meant to be. Modern robotized assembly can now make maybe not  identical but quite close copies

The Saturn 5 rockets that took men to the moon looked the same on the outside but over the period of the Apollo missions they were constantly being changed here and tweaked there as issues arose.

Because these were so specialist, there was little or no practical way to construct many of the parts other than my hand.

Repeatable computer-controlled manufacturing was still in its very early days and very limited. Items like the heat exchanger on the nozzles of the F1 engines were a huge plumbing system that was brazed together by hand.

The welds that fixed all the parts of the engine together under incredible stresses were done by hand, no robotic welders back then.  The fuel injection plates had over 2800 holes that had to be precisely drilled. If you made a mistake 2750 holes into the job then the whole thing might need to be scrapped.

The shape and pattern of the baffles fixed the plate and helped stopped damaging oscillations from building up were arrived by trial and error as time and launches went by because there just wasn’t the advanced computer modeling available.

With these variabilities, every fight was a test flight to a degree but the errors were narrowed down and the overall risk was reduced too.

But It wasn’t always like that. Before the Apollo 1 fire which killed the crew during a ground test there was a more cavalier attitude to the safety, design and building of the not only the Saturn rockets, command module but the to the crew’s safety although this was never admitted.

Up until that point they had gotten away with pushing things to the limit, they were up against a deadline of getting a man on the moon by the end of the decade and there was no guarantee that they would do it. Much of the engineering hadn’t been done before so the real test would be the launch itself.

For example, North American Aviation the contractor for the Command module said that NASA was wrong to use a 100% oxygen atmosphere for the Apollo 1 capsule.

Constant spec changes to the command module meant that it was now too heavy. To NASA, running a low-pressure 5 psi pure oxygen environment would be simpler, lighter and put less stress on the structure, saying that with a trained crew the risks were manageable.

But what they didn’t contend with was the Teflon coating used for the electrical wires which could be easily damaged and allow a short circuit to occur. Combine that with the pure oxygen environment plus many other combustible materials and a door that only opened inwards and in hindsight it was an accident waiting to happen.

NASA flight director Gene Kranz took personal responsibility for the accident and he, NASA and the American government made immediate changes to safety measures and security protocol after the incident, something that stayed with the program up until Apollo’s end in 1975. No other crew or vehicles were lost, even Apollo 13 made it safely back to Earth.

However, the next generation of NASA managers working on the Space Shuttle become came too focused on getting results much like the early Apollo ones and again that came back to bite them big time with the Challenger disaster.

A late running schedule was the catalyst for a series of decisions to overrule engineers’ safety concerns and go ahead with a launch when they knew there was a possibility of failure of the O ring seals on the solid rocket boosters on the freezing morning of Jan 28th 1986.

Of the 135 shuttle missions 2 were lost leading to a 1 in 67 failure rate which is not bad in the grand scheme of things, but the lack of an escape system meant that when things did go wrong the crew of seven paid for it with their lives, twice. Vehicles can be replaced but people can’t.

The risk of things going wrong with the very complex business of building and launching rockets increases dramatically when companies or organizations are under pressure to meet time and or budget targets.

This is something that invariably happens when new launch vehicles are introduced because they have investors or governments looking over the shoulders all the way but once the major bugs are ironed out it can become a much more reliable system.

The destruction of the Starship and the pad could ultimately put down to rushing the job out and yet the same company SpaceX now has the most reliable launch vehicle in space history with the Falcon 9 family which has launched 239 times with one partial failure and one total loss a failure rate of 1 in 120.

Soyuz, the Soviet and Russian rocket family has more than 1,900 launches across about a dozen variants of the booster dating back to 1957, but it’s had more than 100 failures which is about a 1 in 20 failure rate.

NASA and the SLS on the other hand have been very conservative, and as a consequence, although it was late, it was in one piece and successfully completed the first test mission.

The James Webb Telescope is another where the risks could be seen as very high. A one-off $10B project, 10 years behind schedule and packed with very delicate one-off parts and systems that could not be fully tested here on Earth.

It was guesstimated to be a 50-50 risk that it would make successful deployment, not because of the Ariane launch platform which is one of the most reliable in the world, but that everything would have to work perfectly once it was unpacked 1.6 million km from earth.

The JWST was tested like no other spacecraft yet launched and the results proved that you can make things reliable if there is enough time and resources put into it.

Space will always be a risky business to be in but if Spacex can make the Starship as reliable as the Falcon 9, then maybe going into space will become no more dangerous than jumping into the car and nipping down the shops.

Thank you for watching and I hope you enjoyed it, if you did please thumbs up, subscribe and share and I’ll see you in the next video.

Paul Shillito
Creator and presenter of Curious Droid Youtube channel and website www.curious-droid.com.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.