I wonder why nobody ever proposed a space launch system like Starship. Methane as a fuel and all the theory behind are well known, so why?

Is it solely because of IT progress that made returning the stages possible?

  • 1
    $\begingroup$ Just to clarify, you are specifically asking about Methane as a fuel source? Because if not, there are similar proposals--for example the Sea Dragon--which never got anywhere for a variety of reasons $\endgroup$ – Dragongeek Mar 23 at 19:10
  • $\begingroup$ @Dragongeek no, about the whole system, which many consider "revolutionary" and "impossible". $\endgroup$ – Anixx Mar 23 at 19:11
  • 3
    $\begingroup$ @Anixx Has anyone besides industry competitors claimed the system is "impossible"? $\endgroup$ – Russell Borogove Mar 23 at 19:30
  • $\begingroup$ en.wikipedia.org/wiki/McDonnell_Douglas_DC-X $\endgroup$ – Polygnome Mar 24 at 7:57
  • $\begingroup$ OP WHAT? Nothing about starship is new in the slightest. Everything implemented in it has been proposed before $\endgroup$ – Hobbamok Mar 24 at 9:32

I wonder why nobody ever proposed a space launch system like Starship.

What exactly do you mean by "like Starship"? Systems like Starship have been proposed before, although differing in the details, going back to Von Braun's mid-1950s Mars expedition concepts. Starship is ambitious in several ways, but it's more evolutionary than revolutionary.

Starship is fundamentally a large, two-stage-to-orbit launcher with full reusability of both stages.

It's about 60-70% larger by liftoff mass than the largest successfully flown orbital launcher (the Saturn V), but much, much larger rockets have been seriously proposed (e.g. Sea Dragon).

Two-stage-to-orbit with full reusability of both stages was seriously considered for the US space shuttle program (as described in The Space Shuttle Decision), but the budget didn't allow for development of a fully reusable booster. The fully reusable shuttle concepts relied on winged, horizontal landing rather than vertical propulsive landing, but the pros and cons of the two approaches are endlessly debatable. Propulsive vertical landing has been demonstrated and proven at smaller scales (Surveyor, Apollo LM, several Mars landers, DC-X, Falcon 9).

Large methane rocket engines and full-flow staged-combustion engines haven't been flown before, but both concepts have had development work done previously.

  • 2
    $\begingroup$ A lot of projects of the past indeed proposed full reusability via airstrip landing or parachutes. It seems though the vertical landing is the only option, compatible with travel to other planets. $\endgroup$ – Anixx Mar 23 at 19:58
  • 25
    $\begingroup$ As I see it, the most "revolutionary" aspects of Starship are 1: they're not relying on some super-advanced technology that will cost billions to develop, and 2: they're actually executing on it, in a field where the established players were comfortable with the status quo. Much the same goes for another business of Elon's: Tesla isn't taking advantage of some revolutionary advance that has just enabled electric cars, it's just executing on something the established car makers didn't care to do and in fact were actively resisting. $\endgroup$ – Christopher James Huff Mar 23 at 21:17
  • 10
    $\begingroup$ @RussellBorogove I would say the dramatic success of SpaceX is that they didn't try to go high tech. $\endgroup$ – Loren Pechtel Mar 24 at 1:45
  • 8
    $\begingroup$ @LorenPechtel or did so in a very constrained way. The Raptor is pushing the boundaries of the state of the art, and they're using some high-end robotic welding machines. However, those machines are welding stainless steel with pretty bog-standard techniques, and they didn't even start on that until they were sure that Raptor would do the job. $\endgroup$ – Christopher James Huff Mar 24 at 2:10
  • 3
    $\begingroup$ Their engine is hi-tech, their material are hi-tech, their code is hi-tech (they use C++ and Javascript!), their sensors (MEMS) are high tech and did not exist like 20 years ago. But nothing is completely innovative, automated landing? done by CCCP in a time where computer where big as rooms. VTVL (vertical take off and landing)? explored and demonstrated in the 1960 (apollo landing!) and multiple other rocket later. Even blue origin had Goddard before spacex's grasshopper. What spacex did was to manage to concrete those idea in a product in a economical way, an incredible feat by itself $\endgroup$ – Lesto Mar 24 at 14:44

Fundamentally, it's because of economics. There simply wasn't any demand for a large rocket between today and the space race.

Let's analyze what (I think) makes the Starship concept special:

  • Size: Starship is one of the biggest (if not the biggest) rockets ever constructed.
  • Reusability: One of Starship's core design goals is to be reusable comparable to airplanes
  • Budget: Being cheap and simple is at the core of Starship's philosophy ("the best part is no part")
  • Mars compatibility: Starship is intended to be refuelable via ISRU on Mars

Now, let's go over why nobody's seriously suggested something like Starship before (although there are exceptions, which I'll get to later).

Size: A bigger rocket is more efficient, primarily due to the square-cube law reducing non-propellant mass in the rocket the bigger it is and certain elements, such as avionics, which do not scale significantly in mass when the rocket gets larger. Large rockets have been proposed and built before (Saturn V, N1, SLS) however the rockets have always been built to suit demand. The Saturn V and the N1 were built for the space race and the SLS is being built because NASA wants to send heavier payloads to the Moon. In the past, large rocket proposals have mostly faltered against the question "But why do we need a big rocket when a smaller one will do just fine?" Only recently, with burgeoning interest in space exploration and Elon Musk's vision, has manned, beyond-orbit, spaceflight become attractive again to a wide audience, which enables the building of bigger rockets.

Reusability: Arguably, Starship is not new in this regard because it's aiming to achieved the dream that the Shuttle Program was designed for but never achieved: The spacecraft lands, the astronauts get out and high-five the next crew, which boards, and then they take off again after quickly refueling. Originally, the Space Shuttle was supposed to be just that, a shuttle but for reasons that are out of the scope of this answer, it never happened. Starship is attempting something similar, with Elon's goal of multiple reflights in a day being very ambitious. Again though, incorporating reusability in a rocket system really only makes sense when there's enough demand that the financial investment it requires makes sense compared to disposable rockets.

Budget: One of the iconic hallmarks of the Starship program is that it's being essentially welded together in a field while most other spacecraft manufacturers work in clean rooms while wearing in bunny suits. This is, in my opinion, an advantage that commercial enterprise has over government development: since SpaceX doesn't need to justify themselves to politicians and ultimately taxpayers, they can get away with a lot more. If a NASA rocket explodes, people will complain about their tax dollars being "wasted" but if SpaceX's prototype blows up... who cares? SpaceX only needs to appease their limited and hand-picked investors, which is much easier than justifying tax dollars. This institutional pressure to get it right on the first try leads to a cautious approach so that if something goes wrong, it's nobody's fault. Why haven't we seen this before? It's because successful commercial launch companies (that aren't defense contractors) are very new.

Mars compatibility: This is again, in my opinion, due to the traditionally conservative nature of space exploration. Typically, things are done in steps. For example, first, we send an orbiter, then we send a lander, and finally a rover to make sure everything will work. Building Starship with ISRU in mind is jumping over several steps and, again, not something that a government-funded program would be able to justify to a taxpayer easily.

Now, to the exceptions and what's been proposed/done before:

  • The original goals of the Space Shuttle program are remarkably similar to the goals and approach of the Starship program
  • The Sea Dragon proposal from the 1960's exemplifies the "Budget" and "Size" traits of Starship
  • Plenty of other proposed very large space vehicles like Project Orion
  • 4
    $\begingroup$ So nothing like it has been proposed, except for all the things like it that have been proposed? $\endgroup$ – Russell Borogove Mar 23 at 19:43
  • 8
    $\begingroup$ "There simply wasn't any demand for a large rocket between today and the space race" – Just compare the launch cadence of Falcon 9 and Falcon Heavy to get a feel for the difference between the market for heavy lift and superheavy lift vehicles. $\endgroup$ – Jörg W Mittag Mar 23 at 20:10
  • 14
    $\begingroup$ There still is no demand. At least not more demand than there was for the iPhone before it came out. Both products are pushed into existence by entrepreneurs driven by a vision. $\endgroup$ – Peter - Reinstate Monica Mar 24 at 9:00
  • 10
    $\begingroup$ @Peter-ReinstateMonica the iphone seems an odd example; there were a bunch of big touchscreen phones out there running crufy old windows operating systems, and people bought and used them and would have liked something better. But perhaps you were drawing a parallel between two charismatic leaders with reality distortion fields and an uncritical cult following ;-) $\endgroup$ – Starfish Prime Mar 24 at 11:02
  • 2
    $\begingroup$ @user253751 sure! the original iphone was a bit peculiar, but it highlighted how poor the user experience was on all other smartphones of the same era and how much better it could be done. Cue iphone 2 and android, and everything else bit the dust. Anyway, I could talk about techbro cultists and paradigm changes all day, but not on someone else's post which isn't particular related to either subject. $\endgroup$ – Starfish Prime Mar 24 at 11:25

As other answers imply, the reason is likely not technical but economical. May be this question could attract some interesting answers at this sister site. I think it is a good example of an entrepreneur venturing into a field that has been always controlled by governments and exercising opportunities to "fail fast" and the availability of capital, both unencumbered by politics. In a sense, this is agile development at a cosmic scale.


Building a space ship is rarely anything other than a capital joint venture, meaning it demands approval and agreement of many people from various different disciplines, having a wide range of authorities, responsibilities and things at stake. Starship is an exception in that most executive power lays with one single person. It is that key component, that makes things that should for obvious reasons be possible, but never got of the ground, actually happen.

As intelligent as some people may be, mankind as a whole is brutally stupid. The sad truth about EVs for example, is that they could have been introduced 50 years ago, if it wasn't for the entire species allowing one single financial interest (not a person or group of persons mind you, no no, just money !!) to obstruct progress for half a century.

It becomes even more clear when considering things that could or should obviously exist, but for some reason haven't been developed yet. For instance, the classic understanding of aerodynamics demands fixed wing airplanes to have a rudder. By looking on any airport, we see that every certified airplane does. This in spite of the fact that no flying animal has one, proving it is not strictly necessary. It is not so strange to see that any revolutionary progress in aviation, has to come from a source other than aviation itself, since aviation's own development is utterly stuck.

The beauty of quad-copter drones is that they fly with more agility than any other man made object, using nothing other than four different ways of pushing air down. They do not have any control surfaces. No aircraft factory would have ever been able to produce anything remotely close to that. Even in the development of electric aircraft, they are still stuck on the same old principals, resulting in ridiculously inadequate, but nevertheless certifiable contraptions.

So why did it take so long? Some rocket science is just too much rocket science for rocket science i guess. It took someone who is willing and able to send a roadster into orbit. Apparently, obviously and evidently, that is what it took.

  • 1
    $\begingroup$ This is just plain wrong. EVs WERE introduced more than a century ago (e.g. en.wikipedia.org/wiki/Baker_Motor_Vehicle ), but the battery tech didn't allow sufficient range. It wasn't until the demand from portable electronics made lithium batteries available that EVs became semi-practical. Likewise drones &c: they just don't scale up. It's that unfortunate square-cube law in action. $\endgroup$ – jamesqf Mar 26 at 16:37
  • $\begingroup$ If what happened a century ago is something you want to describe as 'introduction' of EVs, then yes, you're right. My idea is more like i can go to the next shop and buy one, which even now is only beginning to become fact. Lithium batteries became available for portable electronics due to limitations in the patent. It is this limitation that allowed for the construction of cars that are powered by what basically is a whole lot of batteries from portable electronics. $\endgroup$ – Berend Mar 27 at 18:15
  • $\begingroup$ Concerning drones, the laws that supposedly rule aerodynamics are embarrassingly incomplete. This massively constrains the extend to which they can work in anybodies favor. Same argument, different case. Drones don't just scale up, but they do scale up. $\endgroup$ – Berend Mar 27 at 18:16
  • $\begingroup$ Again, just wrong. If lithium batteries had been available a century ago, the patent would have long since expired. WRT drones (those small, multi-rotor electrically-powered flying things, which is really a misuse of the term) it is mostly a matter of scaling - you can likewise have small, electrically-powered model aircraft. It's the same reason why larger birds don't fly like hummingbirds, the largest flying birds, like eagles & condors, mostly soar, and the largest don't fly at all. $\endgroup$ – jamesqf Mar 28 at 0:29
  • $\begingroup$ Of course the real drawback of those quadcopter designs, even if they did scale, is what happens when one of the motors quits. With a conventional plane, you can glide for quite some time, which gives you time to pick out a place to land with minimal damage. See e.g. en.wikipedia.org/wiki/Air_Transat_Flight_236 en.wikipedia.org/wiki/Gimli_Glider and many others. $\endgroup$ – jamesqf Mar 28 at 0:39

The most important aspect of Starship is that it is (supposed to be) reusable. For most of rocket history it was almost impossible*.


Because of lack of good enough computers. You would need to put a pilot in it and the first few landings would end in RUD. The Space Shuttle got around this by landing as a plane.

Only after someone demonstrated it's possible to land an unmanned rocket (some time before SpaceX) did the idea of the reusable rocket move from a dream to something that can be actually done.

* There are other ways to achieve at least partial reusability that could be done sooner (parachute for engine and aerial capture), but there was no market for it (if the government is paying for a few rockets a year regardless of price, you as the supplier don't want cheaper rockets).

  • 3
    $\begingroup$ Buran was able to fly and land autonomously over 30 years ago. So I disagree with your assessment that computers were not good enough for most of space history. $\endgroup$ – DrSheldon Mar 25 at 14:19
  • $\begingroup$ It could not land engines first. It landed as aircraft which was understood. $\endgroup$ – Alpedar Mar 25 at 14:22
  • 3
    $\begingroup$ The DC-X landed engines first autonomously over 20 years ago. $\endgroup$ – 16807 Mar 25 at 16:42
  • 3
    $\begingroup$ The Surveyor spacecraft landed engines first (on the moon) in the 1960s. What you miss here is that on Earth "engines first" only works for 1st stages, that don't need to survive the heat of reentry. $\endgroup$ – jamesqf Mar 25 at 17:02
  • 1
    $\begingroup$ @Alpedar I'd love to give you a nicer welcome, but this answer earns a downvote instead. Your claim that vertical landing is harder to automate than horizontal is completely unsupported and fairly dubious. In fact, landing vertically is conceptually easier because touchdown occurs at essentially zero velocity. All the automation needs is "kill rates using thrusters, vector thrust to kill horizontal velocity, throttle to achieve zero vertical velocity at touchdown". Apollo-era computers can easily do that. The hard part is making engines throttle deep enough to be able to execute this. $\endgroup$ – TooTea Mar 26 at 9:49

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.