Hacker News new | past | comments | ask | show | jobs | submit login

What's also interesting about Falcon 9 is that it didn't need to figure out landing and re-use before it was performing commercial missions in an expendable fashion. Once the payloads had separated, SpaceX were free to experiment with the boosters; until they eventually figured out reliable landings, and could start re-launching used boosters.

I imagine a similar thing will happen with Starship: once it reaches orbit, I wouldn't be surprised if each subsequent mission was full of Starlink satellites (and perhaps other payloads, from brave customers!); even if it takes a long time to get reliable re-entry, re-fuelling, landing, etc.




Yes, OP is a particularly ludicrous comment in the context of this article where Starship is getting compared to SLS. SpaceX's long term, self-set ultimate economic goals will certainly take iteration and a lot of work. But that shouldn't disguise that "merely" going up at this point doesn't look that big a deal at all, which is itself a remarkable thing. Once they're getting to orbit it can start doing useful work and generating money while they work on reuse to make it cheap, making ongoing testing an iteration itself profitable. And they still have no competition in even medium-lift reusable rockets let alone heavy lift, which in turn means they have enormous economic leeway. They can afford to charge $/kg far far above their ultimate cost goal and beat the market anyway. As long as they can land Super Heavy, throwing away Starship every single time would still be market shaking for LEO.

Where they will need full reusability is to make in-orbit refueling work out economically long term, and they need that to use Starship in deep space or other high energy work. Though even there if we're comparing to SLS it still looks good because SLS is just that stupid bloated.


That assumes they will even find customers who want to launch that much stuff.

For a long time, the only fully loaded Starships will be carrying Starlink satellites, and maybe fuel for an orbital tank farm. The Starlink launches will be physically full, but not boosting even close to the advertised cargo weight capacity, because those Starlink sats don't pack in tightly enough. Eventually, people may start to think of things that can use up that much load capacity or volume (rarely both).

Suborbital freight will be a tricky business. It can take quite a lot more than 45 minutes to round up and pack enough freight to fill up a Starship, and to get it unpacked and distributed at the other end, so the 45 minutes aloft doesn't save as much time as you might expect. Same for passenger service: if it takes four hours to get everybody secured, and two hours to get everybody out to the curb after, you have to go pretty far to make it save enough time to be worth the extra cost.

So, in order to take advantage of the new capability, a lot of boring stuff has to be invented and fielded. Practical air freight needed, besides the aircraft, a custom shipping container design delivered to everybody with freight to ship, and cargo bay floors with rollers built in to get them on and off fast enough.


This is certainly an issue, launch capacity is likely to dwarf demand for a while, but on the other hand generic cubesat technology has been commoditised and easily available for a while. A basic kit costs well under $10k. This is partly why starlink has been able to ramp up so quickly, the tech in the satellites themselves doesn't have to be super ground-breaking. The constraint at the moment stopping every educational institution from many many high schools up having their own small sat, or even a fleet is launch costs. I can easily see Starship launching tens of thousands of cubesats a year for organisations all over the world within a few years.


Falcon 9 reuse means expendable Starships probably will never be a thing (IMHO). Why? Economics. The Falcon 9 can get ~22t to LEO. Starship is planned for ~100t. If the net cost of a Falcon 9 launch (factoring in reuse) is $20m then that puts an effective cap of $100m on the cost of a Starship launch.

You can still launch beyond that but it means you're doing so for non-economic reasons. This could be range (eg geosynchronous or Moon orbit instead of LEO) or larger payloads. Falcon Heavy launches seem to be pretty much exclusively military payloads to high orbits that the Falcon 9 just can't carry.

People forget that payloads are deeply coupled to the launch system. Starship will ultimately allow a much larger payload (eg think of JWST 2.0 but with way less complexity) but that will take years for the industry to adapt to, possibly a decade or more.

Starlink exists in part because the launch demand for getting payloads into space just doesn't exist otherwise. Cheaper launch costs will ultimately create more use cases (because they're now economic) but this will take time too.

Starships may be lost in attempting to land but I honestly believe that attempted landing and reuse will be there from day one.


Falcon 9 has a launch velocity limit imposed by the throwaway second stage. Starship can be more expensive and still make sense if you want to launch a lot of payloads quickly.

That said, SpaceX is targeting $10M for a Starship launch early on, with aspirations of bringing that down towards a single million. https://www.businessinsider.com/elon-musk-spacex-starship-ro...


>that puts an effective cap of $100m on the cost of a Starship launch. You can still launch beyond that but it means you're doing so for non-economic reasons.

That assumes you don't have any mission payloads larger than 22t (or 63t for heavy). This is more extreme for Mars transfer orbit 4 & 16t respectively. Starship is advertised as being capable of 100t to mars transfer.


> Starship is advertised as being capable of 100t to mars transfer.

Only with a yet undeveloped in-orbit refueling. Starship can get a lot of mass to LEO or is can get much less mass to a higher orbit or Mars transfer.

The only way it can get a lot of mass out of LEO is for a second Starship loaded with fuel to refill its tanks. This capability is likely still a ways off, SpaceX need to get Starship flying and landing reliably before they attempt a large scale in-orbit fuel transfer.


> Falcon 9 reuse means expendable Starships probably will never be a thing (IMHO)

Sure, customers won't be fronting the entire cost of a Starship. My point was that Starship can perform revenue-generating commercial work through much of its testing.

SpaceX will still be losing a fortune for each RUD; just a slightly smaller fortune than if they stuck with mass simulators until everything's perfected.


Are there any independent assessments that back up the F9 reuse costs? It's hard for me to parse what is Musk hype vs. what is reasonable and objective estimates.


"Cost" is a hard thing to determine of anything this complex. Obviously there's a development cost that has to be amortized across launches so how many do you expect? 10? 100? Famously for years the 747 was a huge profit center for Boeing because it had blown away any estimates of how many they'd sell.

But it's not just straight fixed vs variable costs. You need assembly lines that have a production capacity. How mjch capacity do you want? Do you want the ability to ramp up production if there's sufficient demand?

But other launch vehicles had relatively straightforward math. The vehicle costs $Xm to produce and launch so you charge $(x+y)m for that (where y is the amortized development cost and profit). Production capacity is still a factor (as a variable cost) of course. But a resuable vehicle just upends all that math. Part of the math is then how long does it take to recondition a vehicle for a future flight? How many times can it be reused? Does each reuse make the reconditioning cost more? A lot of these questions will only be answered with time and experience.

And obviously that's a key commercial secret for SpaceX. But they can pretty much charge what they want because there's literaly no competition at that price point. I imagine this is a big part of what is funding the (significant) R&D costs for Starship and Raptor.


I agree. Although the space industry does have various ways of measuring the cost uncertainty, including the cost covariance in a complex system. (A good book on this is Probability Methods for Cost Uncertainty Analysis)

>A lot of these questions will only be answered with time and experience.

I guess that's a bit of what I was poking at. In my mind, there's still a lot of uncertainty with the cost claims, but I see so many people talking about how it's going to bring launch costs down X% and they seem to say it with relative certainty.

I have faith that SpaceX will make great strides in this, but at the moment it seems to be a big unknown, at least in terms of magnitude. My biggest worry is that every misstep will layer additional quality oversight that will drive up cost. An example being the strut failure from a few years back that was related to a supplier material issue. As a remedy, SpaceX now does additional supplier quality control checks. I can imagine a few decades of similar issues being fixed may drive their processes to be more costly.


> Why? Economics. The Falcon 9 can get ~22t to LEO. Starship is planned for ~100t. If the net cost of a Falcon 9 launch (factoring in reuse) is $20m then that puts an effective cap of $100m on the cost of a Starship launch.

By this logic the cost of mail should be much lower given I can get a package 200x larger delivered to my house for 10x the price.


>think of JWST 2.0

An expendable starship with fairings can launch a (dimensionally) even larger telescope. Or a starlink like array of large telescope modules.


That was my point.

The JWST was so complex in part because the mirror was so large it had to have the complicated unfolding mechanism to even fit on the choosen vehicle. When Starship is in service, I imagine a replacement for JWST with the same mirror size oculd be developed for much cheaper because it wouldn't have to be so complex. Or it have an even larger mirror.

But JWST development went hand in hand with the capabilities of the launch vehicle for a decade or more. You can't just put that same payload on another rocket. It's part of the design process. It will take time for the capabilities of Starship to flow into the design process for even larger payloads that otherwise just aren't possible today.


Nah, they plan to utilize starship to the max with a similar folding mechanism as in JWST. Called LUVOIR-A. What I meant is that with this mirror folding scheme and a starship stage with fairings they could launch a telescope even larger than LUVOIR-A. Like an enormous orbital telescope.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: