[some0x80070005]

I can’t tell if this is satirical, but rockets have not proven their reliability much better than 94% for an unmanned rocket and 99% for manned. Airplanes have reliabilities on a per journey basis around 99.9999%. Even if SpaceX is full of 10x engineers who can produce rockets 10x safer than before, that gets them to 99.9% reliability.

The airline industry, for all of its faults, is multiple orders of magnitude safer than rockets. Perhaps rocket based vehicles are the future, but rocket failures must first become drastically more rare or at least less prone to everyone dying onboard in the case of a failure to be used in lieu of traditional airplanes.

Rockets - https://space.stackexchange.com/questions/8566/what-is-the-s...

2020 Airline safety - https://www.reuters.com/article/uk-airlines-safety/aviation-...

[btilly]

No, rockets have not proven their reliability. Reliability is something that gets developed through volume of operations combined with a safety and feedback procedure that slowly catches ever rarer problems. And then gets proven through an even larger volume of successes. Airplanes have gone through that learning curve, rockets have not.

There have been around 6000 orbital launches, in total. That's across all countries, all companies, and all time. This is simply insufficient volume to learn how to do it safely. When airplanes were at a similar level of maturity they crashed regularly for every reason from mechanical failure to the fact that people didn't know how to fly into a cloud and not fall out of the sky. (It turns out that in about 15 minutes the inner ear gets confused and you're almost certainly trapped in a spiral if you don't have the right instruments.)

Judging from their public comments, SpaceX is hoping to get 3 launches per day from each Starship, and is hoping to produce hundreds of starships per year. Which means that after a few years, they intend to get to more orbital launches per day than we have had orbital launches in all of human history.

If they do so, there is every reason to believe that they will learn how to make rockets safer. Not by a factor of 5 or 10 because they have good engineers. But to the level of, say, car travel. Because they will develop the volume of operations that will let them find the things that go wrong one flight out of 10,000.

(It will take rockets a long time to get to the volume of operations that enables them to match airplanes. But in the long run there is no reason that rockets can't become that good.)

[sonofhans]

> (It will take rockets a long time to get to the volume of operations that enables them to match airplanes. But in the long run there is no reason that rockets can't become that good.)

You know, I've read about easy rocket transportation in science fiction for decades. I'm reading _The Man in the High Castle_ to my son now, and it's an example of this.

But I honestly have trouble buying it, and I'd love if you could tell me why you believe it.

For one thing, it will take ages. I'm having trouble finding good numbers, but I think a reasonable SWAG is that there have been on the order of 100 million airplane flights in human history. Rockets have decades of catchup.

And rockets have an inherently worse failure mode. Many airplanes that fail during flight can still land. I don't think this is true of rockets. Failures are vastly more likely to be catastrophic. Is it possible to ameliorate this at all?

[madamelic]

We have had commercial flights for 80 years. We have had commercial rockets for 0 years.

The same arguments could've been made for train travel against planes at first. In my opinion, rockets are likely safer than planes in a non-fuel incident since they are built to have parachutes and other safety devices built in from the beginning beyond just gliding with the hopes rudders and flaps aren't affected.

Below a few thousand feet (if I recall), both rockets and planes have similar risks in my opinion. The most deadly planes crashes are in the first or last few minutes of the flight when they have the least altitude and speed to figure out a plan; similarly, if an incident occurs with a rocket in the first 90 seconds, there is a good chance of high fatalities.

Maybe I would argue that rockets _could_ be safer since they will spend less time in this dangerous altitude but I am neither an aerospace engineer nor a rocket scientist.

[btilly]

Fundamentally with a machine, if it goes right it works. It is just a question of making sure that it goes right often enough. Planes have the advantage that there are more ways for them to fix things that go wrong. Rockets have the advantage of being simpler so less can go wrong. If we had equal experience in both, it isn't obvious which would be safer.

And yes, it is possible to make rockets safer. See https://en.wikipedia.org/wiki/Launch_escape_system for one of the ways. But there are limits. If you're next to something that decides to explode, that explosion is a problem.

[Gene_Parmesan]

> Judging from their public comments, SpaceX is hoping to get 3 launches per day from each Starship, and is hoping to produce hundreds of starships per year. Which means that after a few years, they intend to get to more orbital launches per day than we have had orbital launches in all of human history.

This sort of talk has always seemed to be really infeasible to the point of being nonsensical. What could possibly drive the need for 6000 orbital launches per day? Where are the fuel resources going to come from?

This sort of pie-in-the-sky thinking I don't find motivating at all, I just find it to be so excessive that I have no choice but to immediately dismiss it.

The double-standard with eco-consciousness gets me too. Green cars, green home power, green everything... but 6000 orbital launches per day. What?

I'm not trying to be a SpaceX hater - although I admit to being very sick of the Musk fanboyism - I just genuinely don't understand where these claims are coming from.

[btilly]

SpaceX is operating on the theory that if you build it, they will come.

That said, rapid delivery of materials from one spot on Earth to another is inherently valuable. I'm sure that most of their flights will actually be suborbital instead. As in point to point travel between two places on Earth. The ability to do transcontinental trips in an average of 30 minutes is of interest to a lot of people, and doing so with physical packages is acceptable at higher risk than doing so with humans.

[m4rtink]

It will be definitely interesting too watch Starship development progress - even just because it's actually in a big part a lot simpler vehicle than an airliner.

You see, there are those very top notch and quite a bit complex raptor engines powering it, but modern efficient jet engines are also pretty complex. The rest is basically "just" two cans of propelant and some very basic cold gas thrusters and big but simple aerosurfaces.

Also while indeed the energies Starship needs to handle at any given time are much bigger, it needs to handle them for minutes at a time (launch, landing) versus an airliner fighting air resistance and 10 different types of weather during a 12 hour transcontinental flight. Not to mention passenger and crew amenities you dont need to lug with you with a ~30 minute hop time.

Rocket engines also have their own oxidizer on board so FOD ingestion or birdstrike is not an issue.

And for launch and landing you need an isolated but rather simple and small area compared to the many kilometers of runways, taxiways and other airport infrastructure.

Lastly they run on liquid oxygen & methane, which might be actually cheaper and easier to get than jet grade kerosene, at least over time.

So indeed, Starship definitely is rocket science but it also does not have many pain points of modern airliners.

[sudosysgen]

Airliners can glide. Rockets can't. That's your big difference right there.

Wings are also way simpler than rockets.

You also have way more time to try to fix or workaround issues.

Reliability wise, airplanes will always be much better than rockets, especially if you include a propulsive landing.

[m4rtink]

Sure, gliding helped to avoid many catastrophes (Gimli Glider, Hudson miracle, etc.) but the thing is - does it make sense for a rocketship ? On a point to point flight you are going to land (or impact) somewhere and you need a couple working engines to land safely, the same as you need working control systems to glide to an airport.

Still, Starship does not really need an airport for emergency landing - anything reasonably level will very quickly become a landing pad once the raptors get going for landing.

Also as for wings - those seem very complex to me on a modern airliner - flaps, spoilers, ailerons, wingless, integral fuel tanks, etc. In comparison to that starship is a lightweight water tower with a couple metal barn doors attached - much simpler structure with far less moving parts.

[sudosysgen]

Wings are certainly not complex. They are passive, and ailerions/flapse/spoilers are hydraulically operated.

[benhurmarcel]

There are requirements for commercial flights which would be extremely hard to achieve with a rocket architecture.

For instance, you cannot have a single failure (whatever its probability) that leads to a catastrophic condition. Meaning, for every component and at every moment, you have to assume the component fails and still have to ensure you don't kill your passengers.

[paulsutter]

SpaceX scheduled to make a new Starship every 2.5 days by 2025

You might want to look at the early days of airplane safety, not the 2020 numbers. Look up reliablity learning curve, safety comes from doing a huge volume of flights. All existing rockets have a small number of flights

EDIT: am a SpaceX shareholder and I’ve met with Elon more than once, so not sure what you mean by parasocial

[staticman2]

Whenever something related to Elon Musk comes up it becomes hard to sort out what is a reasonable claim versus an unreasonable one fueled by a parasocial relationship with Elon Musk.

[chrononaut]

As an aside, I do get amused when Person A makes an argument that may not have explicit references, Person B rebuttals with explicit references, and Person A responds by instructing Person B to go look up their own references for Person A's own argument.

Reminds me of the "All Birds are Cats" sketch by Clarke and Dave: https://www.youtube.com/watch?v=eIl1VuGTk3g (Relevant bit starts at 51s)