Will SS/SH be unreasonably cheap?

[Lar]

A recent thread started with the premise "What if Starship is unreasonably cheap" ... but people forgot that when posting in "what if" threads, one has to assume the premise... some folks started arguing against the case

See https://forum.nasaspaceflight.com/index.php?topic=49071

This thread has been carved out to allow that debate to continue without taking the "what if" thread off focus.

NOTE: the header post used to be something else, don't worry about the date being earlier than the thread these posts came from

NOTE2: If you see some I missed, or see some here I shouldn't have moved, PM me

[M.E.T.]

Surely SH/SS should at least cost more or less what a F9 costs to build. After all, it uses more material and a lot more engines.

The paradigm shift of SS has always relied on its full reusability. Not on its per unit construction cost.

I reckon ballpark $50-$100m per SH/SS stack remains the realistic target. But at 150 tons to LEO the mid range cost estimate of $75m still means a very cheap $500/kg to orbit. For a FULLY EXPENDABLE launch.

This is compared to what, $10,000/kg for SLS? Or is that $20,000/kg? I can’t keep track anymore.

So that’s a factor of at least 20 times cheaper. Without any reuse.

Once it is reused 10 times, even with moderate refurbishment after each launch, the cost drops to maybe $100/kg. That is just insane. So a paradigm shift anyway you look at it.

But costs of $7m per SS are taking things too far, and appears unrealistic. And we don’t even need such fantastical numbers. $100/kg to LEO beats a space elevator for cost, and opens up the solar system to mankind.

[Neopork]

To me, costs could decrease by using roll steel and having only one vertical weld per ring.  Less welding, less welders.  Eventually having a large building like the VAB with all building inside to minimize weather problems such as rain and wind, steady building could improve output and maybe save time also.  Having pre-cut nose pieces would minimize re-dos and make for quicker welding also.  Many things could improve speed of building with less labor.  However, the sunk costs of a huge multi-million dollar building would increase $/kg.  Horizontal construction in a horizontal building might be even cheaper.

Good thoughts. For something as large and rare as SS/SH I am not convinced that having a manufacturing facility with automation and robotic welding, etc would actually be that cost effective - simply because it is hard to imagine them reaching the production volume necessary to really justify the infrastructure cost and additional tooling required. I do think it would make sense to automate portions of the production, even without a new manufacturing facility.
For example:
A machine that can bend the single-sheet ring and complete the weld rapidly that could pump out rings very quickly (They might already have this - it is unclear how automated the new single-weld ring process is in Cocoa).
A robot that could to the circumference welds between the rings in a stack rapidly and perhaps validate/QC the welds in real time.
Presses to cut/bend the various pieces of a bulkhead and/or nosecone.
General incorporation of lean manufacturing principles that are being undoubtedly developed as these two builds progress. Getting components or modules finished and QC'd right before they need to be combined with another module or component so there is minimal waiting time (e.g. can't complete Boca Chica stack until the nosecone is repaired). Staging materials to minimize travel or work site congestion.

I am really interested to see what the "final" SS/SH manufacturing process looks like. I think it will blow minds how quickly one of these can be put together once all the kinks have been worked out and processes validated.

[docmordrid]

Elon Musk ✓ @elonmusk

This will sound implausible, but I think there’s a path to build Starship / Super Heavy for less than Falcon 9


https://twitter.com/elonmusk/status/1094793664809689089

[wannamoonbase]

But costs of $7m per SS are taking things too far, and appears unrealistic. And we don’t even need such fantastical numbers. $100/kg to LEO beats a space elevator for cost, and opens up the solar system to mankind.

Agreed, $7M is too low.  At least at this point in development.  The unseen engineering and business costs are likely more than the materials and labor at this point.

I wouldn't even guess what the cost to develop and install a heatshield the size of SS is going to cost.

They need a better facility, ability to work indoors and out of the elements.

There are easier ways to build up and weld the components.  This is fun to watch out in the open and I suspect its being done for cost and so that the world can see.  (Free publicity)

One thing worth considering though, is that if they can build a SS for some lower number of millions, that it maybe worth building an expendable version with a reduced number of Raptors and over all mass. 

The next couple of years are going to be very interesting.

[M.E.T.]

Elon Musk ✓ @elonmusk

This will sound implausible, but I think there’s a path to build Starship / Super Heavy for less than Falcon 9


https://twitter.com/elonmusk/status/1094793664809689089

I am very aware of that quote, and in fact had it in mind when formulating my post above. To me that puts the optimistic end state manufacturing cost target of SS/SH at F9’s equivalent cost.

That is the benchmark he is chasing. Not F1’s cost. Beating F9 would be the massive end state stretch goal being pursued. And that would be amazing, considering it has 6 times the payload capacity and is a fully reusable vehicle.

[RedLineTrain]

To me that puts the optimistic end state manufacturing cost target of SS/SH at F9’s equivalent cost.

That is the benchmark he is chasing.

Don't think he is chasing any benchmark of the sort.  The cost to manufacture this rocket will be as close to zero as possible as soon as possible.

Musk has had an epiphany regarding manufacturing over the last year-and-a-half.  It started with tents.  Now it often doesn't even include tents.

This process was hard-won.  It nearly broke him. 

[pochimax]

It is impossible for starship to be so cheap. Impossible to be cheaper than Falcon9.
Are you joking here?

[AC in NC]

It is impossible for starship to be so cheap. Impossible to be cheaper than Falcon9.
Are you joking here?

The consensus is that it is very FAR from Impossible.  We're all a long way down the path of discussions that have raised that possibility.  You might want to review if not already familiar.

The differential manufacturing processes and materials make it eminently possible.  Probable?  Who knows?


edit for spelling

[RotoSequence]

It is impossible for starship to be so cheap. Impossible to be cheaper than Falcon9.
Are you joking here?

The consensus is that it is very FAR from Impossible.  We're all a long way down the path of discussions that have raised that possibilities.  You might want to review if not already familiar.

The differential manufacturing processes and materials make it eminently possible.  Probable?  Who knows?

It's not a consensus. There is the assertion by Elon Musk that they have a way to do it, but it's not safe to say that a majority of people believe that there is a path to it, let alone the idea that they've already done so.

[pochimax]

There is no way a raptor engine could be cheaper than a Merlin. If you can do Raptor very cheap the same can be done for making a Merlin.

[spacenut]

Several reasons it may be cheaper:
1. Raptor was designed to be reused many times.
2. Methane is a clean burning fuel that doesn't coke like kerosene can.
3. They are using off the shelf stainless steel.  Far less expensive than composite or aluminum.
4. Stainless steel can take a lot more heat on re-entry than composite or aluminum thus safer and less heat shielding required. 
5. Many of the parts for Raptor are 3D printed parts. 

There may be many more. 

The $/kg delivered to orbit is what is lower in cost or as low as F9.  Yes it is a larger rocket and may cost more than a F9, but 100-150 tons to orbit reusable is a lot more than 10-15 tons to orbit reusable.  EM said that F9 would have to be re-conditioned after 10 flights.  Starship is supposed to be 100 flights or more.  Lower long run costs. 

The Merlin costs about $1 million each.  Raptor is to cost about $2 million each at almost 2-1/2 times the thrust.  Even at 2 times the thrust, the cost is the same for $/kg delivered to orbit.   

[AC in NC]

The consensus is that it is very FAR from Impossible.  We're all a long way down the path of discussions that have raised that possibilities.  You might want to review if not already familiar.

The differential manufacturing processes and materials make it eminently possible.  Probable?  Who knows?

It's not a consensus. There is the assertion by Elon Musk that they have a way to do it, but it's not safe to say that a majority of people believe that there is a path to it, let alone the idea that they've already done so.

I stand by what I said (depending upon how you define "very FAR" w.r.t. Impossible.  Since I'm depending on my defintion, I'm on firm ground).

Elon didn't say that "have a way".  No one said anything about an "idea they've already done so" (not even sure what that means).

Also, I have the likely imprecision of pochimax's declaration (SS) vs. Elon's implausible sounding conjecture (SS/SH).

What's not safe to say is that there is a consensus it's impossible, or effectively impossible.

Here are examples of the point in graph form:

Graph #1:  Consensus Impossible or Effectively Impossible (including close to)
Graph #2:  Far from Impossible

[pochimax]

If SS/SH is so incredible cheap to build it will be because it will be incredible easy to made. In that case it will be very easy to copy, everywhere in the world.

Any concept which involves 7000 people working on it or extraordinary knowledge of doing it, implies much more costs and that it will be not-so-extraordinary cheap to made.

You can' t have both things.    SS can' t be both unreasonable cheap and exclusive.

In case you think about reducing costs because of lots of reusing, you have to explain what are this merits and again apply the same reasoning. If its design is so good it lets a very cheap operation why not everybody out there will not copy the same design and the way to operate it.

Cheap = easy.

[Cheapchips]

If SS/SH is so incredible cheap to build it will be because it will be incredible easy to made. In that case it will be very easy to copy, everywhere in the world.

Cheap = easy.

You keep missing that it's 'easy' for SpaceX but it's not for anyone following.  Many companies could replicate the cheap and mature part of building the steel structures. What they can't duplicate easily or cheaply is the years invested in vertical landing, engine development and operating reusable boosters.

Blue's the exception, but development for them is all about New Glenn for the next couple of years. We don't know if their next vehicle will use lessons learned from Starship.

[DistantTemple]

If SS/SH is so incredible cheap to build it will be because it will be incredible easy to made. In that case it will be very easy to copy, everywhere in the world.

Cheap = easy.

You keep missing that it's 'easy' for SpaceX but it's not for anyone following.  Many companies could replicate the cheap and mature part of building the steel structures. What they can't duplicate easily or cheaply is the years invested in vertical landing, engine development and operating reusable boosters.

Blue's the exception, but development for them is all about New Glenn for the next couple of years. We don't know if their next vehicle will use lessons learned from Starship.

Several organizations have demonstrated small scale take off and landing using engine vectoring. OK these are far from orbital, but TVC of the engine is not a big problem .... maybe the hyper-sonic retro-propulsion is.

However the knowledge that it works, and of the visible technologies is a big leg up! 

[Cheapchips]

If SS/SH is so incredible cheap to build it will be because it will be incredible easy to made. In that case it will be very easy to copy, everywhere in the world.

Cheap = easy.

You keep missing that it's 'easy' for SpaceX but it's not for anyone following.  Many companies could replicate the cheap and mature part of building the steel structures. What they can't duplicate easily or cheaply is the years invested in vertical landing, engine development and operating reusable boosters.

Blue's the exception, but development for them is all about New Glenn for the next couple of years. We don't know if their next vehicle will use lessons learned from Starship.

Several organizations have demonstrated small scale take off and landing using engine vectoring. OK these are far from orbital, but TVC of the engine is not a big problem .... maybe the hyper-sonic retro-propulsion is.

However the knowledge that it works, and of the visible technologies is a big leg up!

I ignored them for the sake of simplicity.  Those organisations demonstrating landing are very far from having orbital engines, let alone orbital vehicles.  They'd need an Uncle Bezos cash injection to transition. It would still take them years to be flying before they could start cheap production.

[pochimax]

No. Easy = cheap. Complex or difficult = expensive.

If only your organization can do it, it won' t be cheap to do it, because you need an organization to achieve the goal. It means hugh costs.

The main reason not everybody is copying vertical landing is for the complex and costly global architecture of implementing a rocket. If the rocket and engines are now so cheap, everybody will attempt to do vertical landing experiments.

[pochimax]

You can't have a cheap and difficult to made engine.
If the engine is $200k per unit that means it will be an incredible easy-to-made rocket engine.

Again, it can't be both cheap and rocket-science.

[speedevil]

It may be "cheap" and "easy" to copy the rocket body. (Lets ignore the GNC and computing.)

But despite the designs on NSF showing schematics of Raptor, re-producing such engines without engineering details etc etc etc would be difficult, time consuming and expensive. So the engines are a significant hurdle. Clearly other engines could be used, and if the organization was not focused on Mars, then Methane might not be their choice. However now the rocket is no longer a copy of Starship ....

I would argue that reproducing raptor is hard.

If you are thinking solely of cost of launch vehicle, and have seen a nice shiny stainless steel vehicle get to orbit with a reusable booster stage, you may reasonably consider what happens if you go even simpler.

As a silly example, consider Black Arrow - H2O2/Kerosene from the 70s - Gamma 8 - ISP 260, 20 tons, 70:1 thrust ratio.

If you replace raptors with Gamma 8s in SS, you get something that can barely make orbit - where in the same condition SS (reusable) can launch 150 tons. (yes, you need 20-50 of them).

Now consider three stage, ~8000 tons liftoff mass, and you can pretty much do 150 tons payload, even with 1970s peroxide engines.

Peroxide is an extreme case, as it alone costs $10M for the amount needed for a launch, but it is sort-of-plausible to imagine a very low tech approach like this hitting - if not SS costs - F9 current prices - if the right lessons are learned - for a 150 ton to orbit class launcher.

Picking the right lessons from SS is key.
'It must be methalox' is not the right lesson.