[New] Space Launch System / Orion Discussion Thread

[tater]

 

 

So with no slip at all December, with any slip at all, 2021.

[jadebenn]

Halfway there:

 

[tater]

 

[jadebenn]

14 minutes ago, tater said:

 

There is absolutely no way that's accurate. At that price the EUS would cost a hair more than an entire SLS Block 1; There would literally be no reason to ever use it.

[tater]

47 minutes ago, jadebenn said:

There is absolutely no way that's accurate. At that price the EUS would cost a hair more than an entire SLS Block 1; There would literally be no reason to ever use it.

An entire block 1 costs more than that. The core alone is almost a billion. We know this for a fact. The SRBs and ICPS are a few hundred million more (each). The marginal cost of block 1 is way over a billion (not counting program costs or dev at all). The RS-25s for each core cost 500 M$ alone (not counting the amount we paid for them in the first place (another 200M)).

EUS is 4 RL-10, so that's pushing 100 million right there. An 8m tank? Likely not that different than the core tank, it's labor, not materials that drives core cost. Then it has RCS, avionics, etc.

Edited November 5, 2019 by tater

[tater]

BTW, I understand that SLS has been under development, and dev costs are not production, but shouldn't we then see the contract values for SLS core plunging moving forward? They are making a second core, right, that one should be pennies on the dollar, right? Is it? LOL.

This whole debacle is why separating dev costs from marginal costs on these programs is nonsense. Add total cost, divide by launches. That's the actual cost. If this was an aircraft or a ship the marginal cost would matter, since they'd dev the vehicle, then the per unit cost would be lower, and the per flight cost would become negligible (because aircraft are not thrown away).

Pro-SLS people, here's a thread idea. Make a new thread where the content is strictly for actual missions where SLS is the optimal launch vehicle you can imagine. Optimal in all ways, C3, price, etc. Make sure price is total expense divided by launches, and of course we'd compare total expense for other vehicles as well when comparing to be fair (so Atlas V dev costs footed by taxpayers, divided by launches gets added in, as does the gov money inputs for F9 dev, etc). That will be a sad thread, containing zero missions.

https://oig.nasa.gov/docs/IG-19-001.pdf

It would be really nice to see what we are actually paying---exactly---for each SLS vehicle.

Edited November 5, 2019 by tater

[sevenperforce]

2 minutes ago, tater said:

BTW, I understand that SLS has been under development, and dev costs are not production, but shouldn't we then see the contract values for SLS core plunging moving forward? They are making a second core, right, that one should be pennies on the dollar, right? Is it? LOL.

This whole debacle is why separating dev costs from marginal costs on these programs is nonsense. Add total cost, divide by launches. That's the actual cost. If this was an aircraft or a ship the marginal cost would matter, since they'd dev the vehicle, then the per unit cost would be lower, and the per flight cost would become negligible (because aircraft are not thrown away).

Pro-SLS people, here's a thread idea. Make a new thread where the content is strictly for actual missions where SLS is the optimal launch vehicle you can imagine. Optimal in all ways, C3, price, etc. Make sure price is total expense divided by launches, and of course we'd compare total expense for other vehicles as well when comparing to be fair (so Atlas V dev costs footed by taxpayers, divided by launches gets added in, as does the gov money inputs for F9 dev, etc). That will be a sad thread, containing zero missions.

Well, not zero missions. It would have one mission: launch maximal money on escape velocity into minimal pockets.

Playing around with possible mission configurations for Artemis has me pretty well convinced that distributed launch is hugely advantageous for almost anything BLEO. Going beyond LEO is significantly different than getting to LEO. Use an LEO-optimized LV like Falcon Heavy or Vulcan to get to LEO, and use a high-energy upper stage to go beyond. 

[tater]

I'm waiting for "Europa Clipper" to appear, but the only benefit of SLS's C3 is a faster transit (direct trajectory instead of a slingshot). Robot will be bored on a longer trip, so additional billions are really important?

[jadebenn]

1 hour ago, tater said:

An entire block 1 costs more than that. The core alone is almost a billion. We know this for a fact. The SRBs and ICPS are a few hundred million more (each). The marginal cost of block 1 is way over a billion (not counting program costs or dev at all). The RS-25s for each core cost 500 M$ alone (not counting the amount we paid for them in the first place (another 200M)).

EUS is 4 RL-10, so that's pushing 100 million right there. An 8m tank? Likely not that different than the core tank, it's labor, not materials that drives core cost. Then it has RCS, avionics, etc.

I really don't know what you're going on about here. All I know is that I will guarantee you the cost of EUS will not be 880 million dollars. It won't even be close.

[Bill Phil]

The worst part about this program is the waste of talent. There are so many skilled engineers at MSFC but they’re stuck developing SLS.

[tater]

I'm waiting for a demonstration that an entire SLS costs less than

3 hours ago, jadebenn said:

I really don't know what you're going on about here. All I know is that I will guarantee you the cost of EUS will not be 880 million dollars. It won't even be close.

If you include dev, it will substantially exceed that. Development cost is in the billions---for a tank, with extant engines. It's fundamentally nothing more than a short SLS core tank with different engines, you're right, it should cost nothing remotely close to 880 M$. I'd go so far as to say if it costs more than 1/10 of that, it would be criminal. Would you agree that a stage 2 costing more than 88 million would be criminally ripping off the taxpayer? (which will be hard with 4 engines that each cost 10s of millions. For reasons.)

It's OK for Orion to cost over a billion each, though? (we know the price of each Orion is nearly a billion not counting the Service Module).

BTW, how will we know what the cost of EUS is? Seems like we should just have that figure available, just like the wiki pagefor SLS should list the core stage cost to the penny---since we (taxpayers) should know this.

Edited November 5, 2019 by tater

[tater]

OK, what I could find on DCSS (and hence ICPS) was that the initial contract was for 1 ICPS at 175M$. Later add-ons closer to 150 M$, maybe. Assuming that it didn't involve any cost overruns (what % overrun has been typical for Boeing with SLS core, do we need to multiply by the same factor?)

ICPS costs (assuming on budget) $50,000 a dry kg.

Orion is ~86k$/kg (just capsule).

SLS core? No idea, but the engines alone are 508 million, so let's call it 600 M$? Fair? That's only $7,000 a dry kg.

EUS of course should cost much more like ICPS. if it costs the same as ICPS per unit dry mass... ~680 M$. Wow. If it's cheap like later ICPS might be, then only 550 M$ for EUS (a bargain, you're right, no where near 880M$). If the part that is the same mass as ICPS costs the same 175M$, and the remaining dry mass is only 7k$/kg, then that adds ~70M$, and the added engines add about the same. So 175M$+140M$= $315,000,000 as a lower bound on EUS cost (IMHO, YMMV).

If it costs less per dry mass, then they should never refly ICPS, and immediately make EUS, cause we're paying a lot for ICPS (because it's cheaper. For reasons.).

There's a simple way to settle this. Tell us what all the parts of SLS are actually going to cost. Someone must know, since someone is writing checks.

Edited November 5, 2019 by tater

[tater]

In 2014 NASA was saying that the marginal cost of an SLS launch would be 500-700 million. That's on top of the ~2.5 B$ annual program cost (which somehow spends all 2.5 billion in a way utterly disconnected to the cost of any vehicle launch? Great accountant, maybe he could do my taxes...).

Interesting number, since the engines alone on the core cost more than the lower limit. Core engines are 508 M$ (this number is known precisely, 127M$/engine), ICPS is maybe 175 M$. If the side boosters are shuttle era cost, that's what, 50 M$ or so? So minus the tank, or any labor at all, or any share of dev costs, or any pro rata share of GSE costs, we're at 733 M$. So the minimum possible cost in the awesome world where the bulk of the actual cost to make the product somehow are "free" (the 2.5B$ "program" costs) is somewhere north of 800 million for an SLS launch (no payload, just the stack with ICPS). With EUS the minimum possible marginal vehicle cost is about a billion.

[mikegarrison]

47 minutes ago, tater said:

Tell us what all the parts of SLS are actually going to cost. Someone must know, since someone is writing checks.

To be honest, I would actually be surprised if anyone did know.

It's very, very difficult to actually know what an airplane costs to build and deliver to a customer. Yes, the total amount of money being spent is known, but how much of it should be assigned to what activities? It's not actually as simple as it seems like it should be, and that's not because of any budget shenanigans. It's just that large, complex processes are not actually easy to cost out.

[tater]

6 minutes ago, mikegarrison said:

To be honest, I would actually be surprised if anyone did know.

It's very, very difficult to actually know what an airplane costs to build and deliver to a customer. Yes, the total amount of money being spent is known, but how much of it should be assigned to what activities? It's not actually as simple as it seems like it should be, and that's not because of any budget shenanigans. It's just that large, complex processes are not actually easy to cost out.

Yeah, I know this, but for aircraft at least there is a contract for X units. Each might be somewhat different (different fittings/versions of 787 for a large order, say), but at the end you'd have an airline buying 35 787s, at however much the price was, and we'd have an average price, right?

I can't imagine United saying, "2 billion, 4 billion, whatever you charge is fine, we don't need the details."

 

Edited November 5, 2019 by tater

[mikegarrison]

29 minutes ago, tater said:

I can't imagine United saying, "2 billion, 4 billion, whatever you charge is fine, we don't need the details."

United doesn't need any details on how much the airplane costs to make. All they care about is their payment.

That's what I'm saying: overall cost at the highest level, yeah, this is known. Detailed cost is not. Airbus or Boeing does not know exactly how much it cost to build that particular A320 or 737 that they just delivered to United. They know what United paid for it, but they don't know exactly how much profit or loss they earned from that sale. They do know overall how much profit or loss the whole company is making, however.

Edited November 5, 2019 by mikegarrison

[tater]

7 minutes ago, mikegarrison said:

United doesn't need any details on how much the airplane costs to make. All they care about is their payment.

That's what I'm saying: overall cost at the highest level, yeah, this is known. Detailed cost is not. Airbus or Boeing does not know exactly how much it cost to build that particular A320 or 737 that they just delivered to United. They know what United paid for it, but they don't know exactly how much profit or loss they earned from that sale. They do know overall how much profit or loss the whole company is making, however.

I agree, completely. In this case, however, NASA is United. The customer. All that matters is what is being paid. For no other reason than planning and budget requests NASA needs to know that an SLS B1 vehicle costs X, and a B1b LV costs 1.2X, etc.

The actual launch cost is not the vehicle cost, though, it's the stacked SLS cost, plus the "value added" of the people paid by NASA that do all that 2.5 B$ worth or "program" work. How much is planning, and how much is actual launch/flight related work that should be costed into comparisons with other LVs... that's more tricky (but it matters in cases like Boeing trying to compete using SLS with commercial launches for a lander).

Edited November 5, 2019 by tater

[mikegarrison]

1 hour ago, tater said:

I agree, completely. In this case, however, NASA is United. The customer. All that matters is what is being paid. For no other reason than planning and budget requests NASA needs to know that an SLS B1 vehicle costs X, and a B1b LV costs 1.2X, etc.

The actual launch cost is not the vehicle cost, though, it's the stacked SLS cost, plus the "value added" of the people paid by NASA that do all that 2.5 B$ worth or "program" work. How much is planning, and how much is actual launch/flight related work that should be costed into comparisons with other LVs... that's more tricky (but it matters in cases like Boeing trying to compete using SLS with commercial launches for a lander).

Hmm. I'm not terribly familiar with the SLS business model, actually, but I believe some of the hardware (engines) and some of the facilities and probably some of the people are actually government property and employees, right? And some are contractors? I suppose a lot of accounting issues could arise in figuring out what is the cost to the government of using a government-owned engine, for instance.

I agree that it is easier to deal with if just buying off-the-shelf rather than paying for a development and integration project.

[tater]

It ends up leaving the only way to assess cost being to add all the money ever spent on it, and divide by launches, else buy fixed price. This will make every SLS launch multiple billions (the realistic price to consider when comparing to anything else that could be bought fixed price for whatever that fixed price happens to be).

Edited November 5, 2019 by tater

[tater]

^^^this was for the Block 2 advanced side booster BTW, not the core stage.

Edited November 6, 2019 by tater

[tater]

A quote from the White House Budget Office:

Quote

The Administration would also like to take this opportunity to share its views regarding language provisions in the bill including:

NASA Europa Mission. The bill requires that NASA use the Space Launch System (SLS) rocket to launch the Europa Clipper mission. The Administration is deeply concerned that this mandate would slow the lunar exploration program, which requires every SLS rocket available. Unlike the human exploration program, which requires use of the SLS, the Europa mission could be launched by a commercial rocket. At an estimated cost of over $2 billion per launch for the SLS once development is complete, the use of a commercial launch vehicle would provide over $1.5 billion in cost savings. The Administration urges the Congress to provide NASA the flexibility called for by the NASA Inspector General and consistent with the FY 2020 Budget request.

That's for Block 1, BTW.

The actual letter for people inclined to disbelieve something merely because Berger reported it:

https://cdn.arstechnica.net/wp-content/uploads/2019/11/shelby-mega-approps-10-23-19.pdf

Pretty sure the guy commenting is not reading internet forums guestimating costs for NASA rockets, he was likely given the number by, you know, people at NASA (because where else would he get it).

[sevenperforce]

11 hours ago, tater said:

"Unlike the human exploration program, which requires use of the SLS"

They really mean "which we pretend requires the use of the SLS" but otherwise....

Distributed launch for Europa Clipper would be so fffffing easy. I haven't done the math in a while but IIRC you could toss it into LEO on a reusable Falcon 9 and send it direct to Jupiter, faster than SLS could, with a DIVH or Atlas V companion launch. Under $200M in launch costs.

[IncongruousGoat]

32 minutes ago, sevenperforce said:

Distributed launch for Europa Clipper would be so fffffing easy. I haven't done the math in a while but IIRC you could toss it into LEO on a reusable Falcon 9 and send it direct to Jupiter, faster than SLS could, with a DIVH or Atlas V companion launch. Under $200M in launch costs.

You don't even need distributed launch. Falcon Heavy + Star 48 solid kick stage gets you there with only a single gravity assist, and with much less chance of distributed launch-related failure. Docking in real life is nowhere near as easy as docking in KSP, and distributed launch is complicated from an engineering perspective.

[sevenperforce]

7 minutes ago, IncongruousGoat said:

You don't even need distributed launch. Falcon Heavy + Star 48 solid kick stage gets you there with only a single gravity assist, and with much less chance of distributed launch-related failure. Docking in real life is nowhere near as easy as docking in KSP, and distributed launch is complicated from an engineering perspective.

Docking in LEO is not harder than docking in cislunar space. What is the difference between staging distributed launch in LEO vs staging distributed launch at LOP-G? We need to develop it regardless.

Europa Clipper is going to be spending years in space anyway. What's the harm in throwing it to LEO a few months early?

[tater]

1 hour ago, sevenperforce said:

Europa Clipper is going to be spending years in space anyway. What's the harm in throwing it to LEO a few months early?

Or sending it to Jupiter via a slightly longer trajectory for that matter. They push for a direct flight simply because it's something SLS can do, not because it is required.