I doubt airlines are actually interested in committing tens of billions of capital, today, to get a dozen % improvement in fuel efficiency a decade from now.
And why would a private investor take such a huge risk without guarantees of making it back several fold?
So it doesn't happen.
Putting my cynical cap on, my hot take is this: Everybody now realizes that Emirates and Clark are exactly right. As was Airbus.[1] Demand will blow past supply (i.e. runways and gates). Ticket prices will sky rocket. But from the perspective of the plane and engine manufacturers this translates to, "if everybody sits on their hands and does nothing, we're going to enjoy a windfall of profits". Clearly, IMO, Boeing, all the engine manufacturers, and now (somewhat despite themselves) Airbus, are going to passively collude to bring about this windfall. All they need to do is sit tight and wait things out. The threat of competition isn't significant enough to break the collusion; there are too few players, the barriers to entry--regulation, technical and logistical expertise, etc--too steep, and like with housing little prospect that political reforms will make it easier to build enough new runways and airports to change the capacity trajectory.
Similar calculus applies to most of the established carriers. There's much more competition among them, but if the manufacturers are expected to continue to coast there's no angle to exploit in that regard, and so the carriers also get to look forward to windfall profits, notwithstanding that they'll be forking over much of it to the manufacturers. Clark just isn't cold-blooded enough to want to play that game; he's got too much of a fighting spirit. (Or perhaps more realistically it's also because if Emirates doesn't keep growing it won't be viable in this prospective future reality?)
[1] Boeing wasn't wrong, either, but point-to-point only bought a couple of decades of supply expansion.
[2] I think there's an economics term for this, but I don't remember it.
Watching the Comac C919 shitshow though (14 years and still not in revenue service, and this is for a 737/A320 clone), I'm not holding my breath. The twin-aisle C929 is still a decade away at best, and it'll take a good long time until any Western airline trusts it; even Chinese turboprops (Xian MA-60/600/700) have so far only managed to find buyers in Africa and the poorer bits of Asia.
I agree that the C919/929 pose no threat as is, but the Chinese are clearly playing the long game here and their massive domestic market coupled with government arm-twisting lets them scale: the C919 has already racked up over 1000 orders.
That said, watching the Comac C919 shitshow (14 years and still not in revenue service, and this is for a 737/A320 clone), I'm not holding my breath.
Me: What's the official economics term for passive collusion, when players in a market avoid competing with each other without any explicit agreement not to?
GPT-4: The official economics term for passive collusion, when players in a market avoid competing with each other without any explicit agreement not to, is "tacit collusion." Tacit collusion occurs when firms in an industry indirectly coordinate their actions, such as price setting or output levels, without explicit communication or agreements. This behavior can lead to higher prices and reduced competition, similar to the outcomes of explicit collusion or cartels, but it can be harder to detect and regulate.
There is a non-trivial chance that point to point suborbital flights totally disrupt the airline industry. Hong Kong to Los Angeles in 30 minutes for the price of a first class ticket is pretty hard to argue with.
Commercial airlines have to deal with going over countries that have air forces which can intercept them. Intercepting satellites is quite a bit harder. And attempting to gets everyone mad at you because of how much garbage it creates. In fact people are STILL unhappy with China because of the fallout from their attempt to TEST their capacity to do that in 2007!
First you need to include price of the floating megaports at each destination in the world. The cost of a single such human rated platform divided by a number of physical rockets using it per year will be many orders of magnitude higher than a price of airport divided by the number of planes serviced per yer.
Second you will need a lot of highly expensive maintenance crew members at each location (more expensive per person than airplane maintenance crews). Plus much much bigger staff at the dispatcher tower in each location, compared to a handful of aircraft dispatchers in an airport servicing hundreds of planes simultaneously.
Third you will need a private (and thus unprofitable) public transportation lines/ships to each of the rocket starts.
4th - you will need a lot of training facilities for the prospective passengers. And staff for them.
5th - you will need low orbit rated space suits in a variety of the sizes at each location. And maintenance and repair facilities for them. Dragon low orbit suit clock at about 0.5 mil dollars today. 4 suits and that's more that entire fuel cost for the launch :) (theoretical cost, with best outcome). Even is economies of scale will help to lower costs 2-3 times (doubtful) it still doesn't scale to a global E2E operations.
I'm probably missing some other significant costs but you can get the picture. It will never be a first class ticket price because it is impossible.
And that's not even counting other blockers for this idea.
First, you're doing a ton of double counting. The estimated $2 million/launch already includes necessary infrastructure, servicing, and so on.
Second you're assuming a ton of totally unnecessary gold plating. For example you talk about "floating megaports". In actuality they've been taking barges (available for about $1 million) and retrofitting them. Starship needs a bigger and stronger ship, with infrastructure built on it. Suppose that it is about $10 million. (Which is roughly the cost of a basic containership.) You could put out a couple of hundred of these for the average cost of a new airport. It would only make sense to go to a megaport if the cost per passenger were less than it is now.
Third, you're ignoring the possibilities of automation. For example your "army of dispatchers" is solving a problem which will certainly be solved by computers. In fact it has to be - because most of the problems will be encountered in the possibility of collisions in space. And we're having to track the entire constellation of everything up there, ranging from rockets to pieces of space junk.
And fourth, you're ignoring costs that already exist for airplanes. For example airplanes have expensive pilots while rockets don't.
First of all - nobody including Elon himself can estimate eventual Starship launch cost. Second - Elon have said that maybe it can come down to 2 mil and that most of that price would be fuel. So he was wildly optimistic as for production and maintenance of the actual hardware. In that interview iirc he was going from an assumption that Starship is already at the high end of reusability (multiple tens of flights or more) and flying regularly.
No one including himself published any total esimations for the total costs of the E2E system as whole and I'm pretty sure that 2 mil does not include any costs for platforms, transit to them, spacesuits, training facilities and dispatchers (new ones, outside of the existing in USA). Maybe he included start gantry, and some preflight checks. Maybe not even that.
Regarding platform cost, I think your estimate is way off. You are forgetting that unlike Elon's barges it need to be a) human rated, b) civilian human rated, c) internationally civilian human rated. It would need a lot of facilities for human on board, it will need some elaborate start rescue systems for 100 space suited passengers. Platform would need extensive protections from them the launch of this insanely powerful rocket (repeated and low on maintenance). Platform would need some kind of armored bunkers for the cryogenic fuels or an extra powered platform which can come for fueling and go away (x2 costs for two platforms). Some kind of evacuation facilities for hundreds of humans in case of a platform fire. A billion of things really. Each one platform will go to hundreds of millions, definitely not a mere 10 mil.
Fortunately I work in IT, so nope, any amount of automation added will simply require the same amount of even more experienced humans supervising them PLUS new people manufacturing, debugging and maintaining said automation.
Starship pilot would cost much more than an airline pilot. Maybe on par in the best case.
And I have to repeat myself - that's ONLY cost considerations. There a lot of other problems.
International permissions for airspace
Total time for the end to end trip (platforms will be VERY far from the cities, exponentially depending from the city size). As soon as that time exceeds 10-12 hours this whole idea is junk.
Total time to train civilian passenger. Total time to train old aged civilian passenger. Tha includes both for rocket, for escaping it, for using a spacesuit etc.
G forces. Why would older billionaires suffer that, potentially risking some aneurism or trombosis or worse?
100 passengers together. So you think 100 billionaires would voluntarily be packed liked sardines in the same confined space with one another? Lol, high society doesn't work like a lowly public bus to the factory.
100 billionaire passengers planning the same trip weeks/months in advance to the same city at the same time? Business does not work like that. They need to go anywhere at the moments notice, when they want, without waiting for a bus... erm.. Starship to fill up. And doing this not as a stunt once a year, but many times a day? Forget about it :) .
Things missing from the regular flight - comfortable private seat in the 1st or a jet (remember price would be so high that a even a big jet price would be cheaper, with all the seats together still cheaper). Comfortable talking and working conditions. Comfortable transit on land, not on a choppy sea. Comfortable suit instead of spacesuit. Gourmet food and alcohol. Chosen company. A lot of things.
First, the specs for Starship mean it can hold about $1 million in fuel in its tanks. So we can externally estimate fuel costs. His aspirational goal is $2 million for a launch. His stated initial target is $10 million. There is room in his aspirational target for lots of things other than fuel.
Second, you keep insisting on a ton of people doing unnecessary things. SpaceX already has a tracking system in place to keep track of everything in space and make sure that none of their about 4000 StarLink satellites hit anything. Adding a few hundred Starship rockets to this system does not require an army of dispatchers. You're now talking about rocket pilots despite the fact that essentially nobody bothers with rocket pilots, and they can't be used in SpaceX flights because human reaction time is too slow to do anything useful.
Third, you've talked over and over again about packing billionaires in like sardines. But the target passenger is NOT billionaires. It is executives in companies who currently fly first class, but would rather avoid a 20 hour flight from Los Angeles to Dubai. Every day, literally tens of thousands of people fly first class over an ocean. The Concorde demonstrated that a sufficient volume of them are willing to pay more for a shorter ride in a less comfortable plane to enable regular commercial flights.
Fourth, your set of requirements for civilian rating should be seen as negotiable. Elon's position is that Starship IS a rescue system. He has a point. For example external analysis has found that if the Space Shuttle had simply put the shuttle on top of the rocket, nobody would have been hurt in either accident. If he provides a compelling service, but only offers it where local authorities allow, authorities are likely to figure it out. Just look back on the history of how the Concorde was allowed to land in the USA despite our laws against sonic booms.
And fifth, I DO NOT AGREE that I should accept, on your authority as an IT person, that automation cannot reduce the cost of people for a given task. History is full of counterexamples that represent far better evidence than your claim to authority. Granted, Jevon's paradox is common. We reduce the cost, which increases the demand, which leads to more overall usage. But on the per unit economics, the cost really does go down. Were it otherwise, nobody would be particularly interested in automation.
The problem they have is that there is a large fixed cost to flying the plane from A to B, and a small marginal cost of adding you and your luggage to the plane. All airline tickets cover that marginal cost, but they vary in how much of the fixed cost they cover. The minority of expensive first class tickets carry a very large fraction of the fixed cost. Effectively they subsidize the rest of us.
If those people go elsewhere, then you aren't covering the fixed cost of the flight. And now everyone else has to pay a lot more. Which makes the pool of people less.
If Boeing begins the design of a new plane now, they would be planning on finding customers in 10 years who expect to use the planes for the following 20.
But in 5 years we should have a good sense of whether Starship is viable. If it is, then 10 years from now Boeing won't have a market.
Even if you give Elon only a modest chance of success, it makes sense to stretch out the sales cycle for existing planes and delay starting new development for 5 years.
Also it will not be 30 minutes from actual city to actual city and it will not be a first class price ticket (10-20k) for that, but actually more like x10 times higher. And that IF Elon's fantasy Starship price of 2mil/launch happens. If actual launch price will be 20 mil, then this theoretical E2E ticket will be x100 times more expensive than fist class ticket.
And that not even mentioning several dozen independently blocker issues to this zany idea. How come people still repeat it on the tech site?
Then Elon made an insane sounding suicide slam burn. And he made that reuse routine.
Now to your claim. Elon's "fantasy" is that a rocket launch and turnaround can be made on relatively similar economics to airplane launches. If so, the cost of launch is $1 million, not $2 million. And so charging $2 million is adding headroom for things like accidents, depreciation and profit margins. It is ambitious to be sure. But based on the materials cost, this doesn't look like a fantasy.
not that far from the theoretical thermodynamic efficiency limit
No. Even if an engine got to the Carnot efficiency, you can still improve it. Massively. You can push a certain quantity of air with a velocity v or 4 times that quantity with half the velocity. The second option produces twice the thrust, for the same fuel consumption. That's the reason engines are getting larger and larger, and going for higher and higher bypass ratios.
Separately, the maximum theoretical (Carnot) efficiency depends on how hot the engine runs. GE has introduced turbines made of silicon carbide, which can withstand absurdly high temperatures. Silicon carbide melts at more than 2800 Celsius, but they don't push it to such limits. But they push it to more than 1300 Celsius [1]. Military jet engines run much hotter though. Exact numbers are classified, but according to [2] the temperature inside F35's engine might reach 2000 Celsius (the actual number is obviously classified). Of course, military jet engines are many times more expensive than civilian ones, with higher maintenance costs, and lifespan about 2 orders of magnitude shorter.
Still the fact that they can run this hot means improvements for the civilian jet engines are possible. And they will happen.
[1] https://www.ornl.gov/news/ceramic-matrix-composites-take-fli... [2] https://www.researchgate.net/publication/285727346_Fahrenhei...
These planes might not make much sense for many routes, but for trans-Pacific routes and other very long-haul routes, a huge plane like this is great.
This is why a 4-engined plane like the A380 isn't necessarily less economical for having 4 engines instead of 2. Airbus and Emirates have always argued, and arguably Emirates has demonstrated, that all things considered the A380's 4 engine configuration is still cost competitive. However, these factors cannot overcome the aged engine technology. When the A380 debuted, its engines were already half a generation behind the latest, greatest wide body engines, and now they're even further behind. Moreover, IIRC, even older 777 models have received engines upgrades so now the A380's engines are technologically behind aircraft that preceded it.
Airbus and Emirates weren't asking Rolls Royce to deliver anything fancy; just a replacement engine sporting updated blade designs, improved bypass, and other ubiquitous advancements since the mid 1990s; but nothing bleeding edge, like new gearing technology.
EDIT (wiki links)
Isn't SpaceX going to use Starship for international travel. That might motivate these engine makers.