NASA will pay $146M for each SLS rocket engine | Hacker News Reader

NASA will pay $146M for each SLS rocket engine

arstechnica.com

103 pointselsewhen59 comments

Tim Dodd just released an excellent (and long) video comparing the costs of SLS with Falcon Heavy and (eventually, maybe) Starship. https://www.youtube.com/watch?v=KA69Oh3_obY&feature=youtu.be

SLS will cost upwards of a billion per launch. Compare that to $90 million for a reusable Falcon Heavy launch or ~$150 million for a disposable one.

Kind of off-topic, but trying to watch that video brought this to mind: https://news.ycombinator.com/item?id=22792243

Tim can repeat himself often, but he's not bad at what he does. His content is admittedly designed for children.

That’s a bit harsh, at the beginning of his videos he does a kind of outline of the content. But the content is there ! Kind of understand why you say that when, at one point, he mentions a future video.

And SLS Orion Launch:

4x RS-25: 520M

2x SRB: 200M

ICPS: 150M (EUS will be more)

Core Stage: (not sure) 30M+ given how long they work on it

Orion Capsule: 900M

That is all without ground infrastructure or people working at NASA. Given that it will probably not launch more then once a year. 100s of M extra.

In comparison, a SpaceX Raptor internal cost is 1-2M, Merlin less then 1M. The sale price of the BO BE-4 is 8M.

The purpose of the space program isn't to explore space, it's to consume federal funds, accruing influence and kickback contributions to the politicians who arrange it.

It's important you understand this primary purpose if you want to understand what happens at NASA.

If you are a space program supporter who is okay with the situation as long as NASA is getting money, then you are the reason they can get away with this. Thanks a lot.

I worked as a contractor on a NASA project for a couple of years. Your observation that the purpose of NASA is to consume funds, accrue influence, and generally perpetuate itself matches with many of my experiences of those years. This is what comes to my mind:

I saw sounding rocket missions with overall costs of several million dollars completely lost, in terms of science data, because the telemetry link was lost. Well, we were launching recoverable vehicles and overall data recording needs were just a few Gigabytes over a 15 minute mission, so why not just record that data on-board in case the telemetry link to the ground failed? Oh, I was told, doing that is far too expensive. Well, the cost of a recorder is not even close to the cost of launching a whole research rocket with payload, so the whole thing just struck me as a tremendous waste.

My immediate previous job was as an engineer for the US Navy. I was well familiar with the concept of "colors of money", so I get that sometimes you may be flush with cash for some things (staff, facilities) but devoid of cash for things like equipment purchases. But even in my Navy years I saw obvious stupidities like this eliminated with some financial management. The contractor and NASA management didn't care. They were fine with losing years of time and millions of dollars for want of a $10,000 crash survivable solid state recorder.

The contractor has since merged with a larger competitor, then merged again. That same entity and largely the same people remain in place, managing (and in my opinion, mis-managing) this program. The only people who really seemed to "lose" when mission data was lost, as far as I could tell, were the graduate students working for our Principal Investigators who lost the data they were depending on for their PhD theses...

it's also a high quality job creation program that employs 17K at NASA and who knows how many at contractors

it's also a military capability

it's also saved who knows how many lives with weather satellites and ambulance GPS and satphone rescues

it's also a meteor warning program

it's also a global warming monitoring program

it also begets companies like spacex and capabilities like GPS for all and inspires people to pursue science

sure there's fat, but we're employing highly skilled domestic tech and science people with that fat, would it be better if we left all those tax dollars to be spent on iphones and Starbucks? or the prison industrial complex? NASA and NOAA and the NIH and any pure science agency is super defensible

We can spend all the taxpayer money on creating 17k highly skilled jobs that actually do something.

NASA does some useful stuff, but we could be doing much more useful stuff with the same money, while still having 17k+ jobs and good military technology and weather satellites etc.

Can we though? Or have we demonstrated over time that any suitably large program sustained over a suitably long time will innevitably need to account for the enormous incentive there is to exploit that funding. What do you propoes to do? Ask the defence companies if they would stop exploiting the government contracts that only exist because of their lobbying in the first place?

It's so tiring that people are unwilling or unable to handle nuance and complexity. An organization can both be bloated and inefficient as well as bold and daring and on the cutting edge of science.

I'll take a bloated NASA that sends probes to Pluto and funds new exciting programs such as Commercial Crew over no space program any day.

If you'd like to provide a specific plan for effective reform, I'd be interested in that as well.

It's not that it's bloated and inefficient, it's that the basic mechanisms driving its behavior are not aligned with its ostensible stated purpose.

A politician doesn't usually care about science. He cares about votes and influence. So when he gives money to NASA he is doing so to produce those, not to produce science. If science gets done it's almost accidental.

The reason politicians get away with this is that the public lets them. They accept that what NASA is doing is ok, even when it's not. Apologists like yourself are ultimately to blame.

The way I understand it is that during the space shuttle years, companies made ties and lots of money off of federal contracts. The whole SLS program has since been doninated by those companies using political pressure to get those contracts back on new missions.

The SLS design was required by congress to maximize use of Shuttle hardware and contractors.

> The core stage is structurally similar to the Space Shuttle external tank,[24][25] and initial flights will use modified RS-25D engines left over from the Space Shuttle program.[26] Later flights will switch to a cheaper version of the engine not intended for reuse.[27]

https://en.wikipedia.org/wiki/Space_Launch_System

So they reused as much of the space shuttle as possible, down to using left over engines. Developing a new engine from scratch using a new fuel (methane) like SpaceX did would have been a very risky proposition. Like NASA, Blue Origin and ULA both did much more conventional things and have launch costs multiple times those of SpaceX. The relative cost of NASA's more conservative design decisions is more a testament to Space X's remarkable success with the Raptor engine than NASA being somehow incompetent.

Each Merlin Engine costs $1M.

A Raptor is expected to cost less than $1M each.

Blue Origins BE-4 is priced at $8M each.

So how did conservative NASA get cheaper RS-25 engines for only $147M each?

Those numbers are not measuring the same thing. Those are the marginal cost per engine. What did it cost Space X to develop the design, and what did it cost them to build the production facilities? The price NASA is paying is for delivery, not the cost to produces one more after billions have been sunk. Also, how many do they plan to produce? NASA probably only needs a few dozen RS-25s for the life of SLS, so probably not worth the billions Space X spent developing a new engine.

The Merlin was developed for the Falcon 1, at a total cost of $90M (including the rocket itself; and Kestrel engine for 2nd stage).

It was enhanced for Falcon 9, which has a total development cost of $300M for the first versions, including rocket, tanks, second stage, Merlin Vacuum.

The RS-25 was developed in the 70s for the Shuttle, and had a quoted cost of $47M at end of Shuttle life. This is literally the cost to make a few more after all design and testing was done. NASA has spent over $15B developing the SLS, why are they paying three times more for these engines?

Because they shutdown the production lines capable of making them for $47 million over a decade ago, and it costs a lot of money to recreate them and get the tooling and personnel needed. It's also the single most expensive rocket ever. It was designed to be the best, economics were ignored because they had the budget to do everything optimally. In terms of basic specs, RS-25 is comparable to Raptor with superior specific impulse, and is a highly proven 45 year old design. Reliability is really the most important stat, given the cost of the payload and other parts of the rocket. Obviously Space X spent a lot of time thinking about economics, with great results, but going for the expensive but sure bet option made a ton of sense for SLS.

Merlin is much simpler and smaller and not really comparable to the other two.

The RS-25 got new life with SLS, I would expect future NASA craft to use Space X engines.

It is difficult to assess risk without an indepth technical knowledge of the engine. Particularly when the engine could have a profound influence on how the rest of the rocket is designed.

I feel like we'd all be better off if the federal government just cut an equivalent no-strings-attached check to Alabama in perpetuity, cancelled the SLS, and freed up all the rocket scientists to build something useful.

They could probably even do it in Alabama! SpaceX could open a local branch and mop up the talent. The current "waste money on the SLS for political reasons" just feels like the worst of all possible worlds.

I feel we are better closing down instagram/<useless app>, and using freed up computer scientists to build something usefull.

I don't think you understand the difference between public and private money

I don't think you understand that entire industry sectors would not exist without Nasa. There would be no GPS, no Google-maps, and you wouldn't know when a hurricane is coming.

Richard Feynman was pulled in to investigate the space shuttle challenger disaster. He refused to sign the report unless his findings were included. They were eventually included as an appendix, which is well worth reading.

There is a section on the main engines (you can search down to the SSME section)

https://science.ksc.nasa.gov/shuttle/missions/51-l/docs/roge...

> Thus, it is not unreasonable to guess there may be at least one surprise in the next 250,000 seconds, a probability of 1/500 per engine per mission.

It's interesting to read this in the context of the STS-93 incident [1] where a broken SSME injector came close to causing a mission abort. Was this Feynman's "one more surprise"? He was writing this report 70 missions before it happened in 1999.

[1] https://en.wikipedia.org/wiki/STS-93

Thanks for sharing, this is a great read.

The bit about top-down vs. bottom-up design (Liquid Fuel Engine (SSME) section) is surprising to me. It appears for the Shuttle program, all the years of knowledge on how to successfully complete engineering projects with large CAPEX + large risk was just ignored.

Of course, the answer to why it is done, is the money flowing into many states involved in the production. Jobs are something every politician likes to create. And it is part of an economy, that the state spends some amount of the tax dollars to support a wide range and healthy economy. But for that, the jobs need to do something useful. Just overpaying for rocket engine is the quite opposite. The same amount of money could be spent to create jobs which have a longtime value and as a consequence create more jobs. Even if it was just used to accellerate the whole project by not wasting effort on something which can be bought at a fraction. Like in having the funds to make a real moon station, instead of creating a rocket, which goes to the moon once or twice and then gets cancelled because at those costs, you cannot afford a moon station.

I don't understand - what is the motivation here? The article is a little short, and it gives good insight into just how expensive this is (compared to other similarly classed potential space purchases). However, what I am failing to grasp is... why? Why spend so much when these other options are available? Surely not ignorance. Is it malice? corruption? Someone's ego?

This is a jobs program which happens to produce rockets. The rockets are not the key objective, the jobs are.

Playing devil's advocate here: wouldn't "preserve institutional knowledge about building rockets" also be it's goal?

> preserve institutional knowledge about building rockets

No it would be "preserve institutional knowledge about building OUTDATED rockets"

If you are a tech company are you going to run your own outdated servers at 10x the cost because you would like to preserve institutional knowledge about running outdated hardware ?

Your argument would be correct if the cost difference was not that huge, the fact that it is so huge indicated that the tech they are using is very very outdated. No point in preserving anything about it.

The cloud hardware metaphor is a good one.

Preserving knowledge is easy and cheap.

Preserving capability (that is to say, people and companies with knowledge and ability to use it) is not.

To use the metaphor, a huge part of the military procurement budget is dedicated to continue building, deploying, and operating on-prem servers.

Not because it's more efficient, but because at an international level, having a cloud provider (or in this case Russia or China) tell you "No" when you ask to renew or expand, is unacceptable.

Unlike tech, machine shops filled with trained people can't be provisioned with a click...

"Outdated" also means well-researched, well-tested, with very little unknown unknowns. I do not mean "safer", necessarily - but the data about how safe they are is much more reliable.

> If you are a tech company are you going to run your own outdated servers at 10x the cost because you would like to preserve institutional knowledge about running outdated hardware ?

As a startup CTO, I will think twice about using the freshest framework or programming language, which is sexy and cool - just because I haven't seen enough people use it long enough yet. I will build a hobby project with it and love it, but or a small product that is allowed to fail, but I will not start an important new product with something like that.

But do we really need those C*O jobs and increased shareholder value? Probably keeping many of rocket science jobs could have been cheaper for taxpayers. But no one cares about that today, our children will be paying our debt, not us.

It's not C-level jobs, it's thousands of normal blue collar and white collar aerospace jobs across multiple states for many years.

Those jobs exist whether we build the utter waste of the SLS or not, those people will just be actually building something useful.

I'd say there's two main justifications:

1. Lower Risk - based on 40+ year old proven rockets. NASA can't have failures if it wants to get future funding.

2. Jobs, Jobs, Jobs! - spreading out jobs means politicians are more likely to approve funding

40 year old rockets proven to kill 17 people and miss every cost and performance metric when they were operated.

Were any of those deaths due to a failing engine though? That's the only part they're really reusing afaik

They are reusing very similar SRBs, which killed one crew and complicates SLS abort modes since SRBs cant be turned off and Orions parachutes will melt if they come into contact with SRB exhaust.

The RS-25 also requires Hydrogen fuel, which in addition to being a terrible first stage fuel because of its low density and heavy tankage (which is why SRBs are required) also has been a causative factor in most of the SLS delays to date.

The RS-25 is viewed as the Ferrari of engines because of its complexity and high ISP (which is only because it uses H2 as fuel). The main thing it has proven us that it’s far too expensive for any useful application. When used on the Shuttle each RS-25 had to be entirely torn down and rebuilt between flights, a process that took months and great expense.

So the SLS takes these enormously expensive engines and burns them up after each use. The only reason RS-25s aren’t long ago put into junkyards is congress trying to favor existing space contractors.

A keralox or metholox first stage would not need super expensive SRBs, would be far easier and cheaper to build. Raptors cost $2M each in prototype form and are expected to cost under $1M in production. Each raptor also provides more thrust than an RS-25, using a far denser and easier to handle fuel.

As far as I can tell, this is the space industry's "Wile E. Coyote" moment[1]. A long one at that.

This kind of largely unquestioned cost structure ("cost plus" contracts[2][3]) has been going on for such a long time now that nobody in that part of the industry has really accepted that the party is over.

And of course industry players don't actually have to accept that the party is over as long as government pays those old-style prices, which have been outrageous for some time and have only gone up. (They seem to have gone up even more egregiously recently, which might be an indication that industry is actually very well aware that the game is up and is trying to rake in as much as possible while they still can).

And yes, there are justifications, some of which may also be reasons (jobs, know-how, whatever ...), but I am sure that inertia is a huge part of this. If this were being decided from scratch, I doubt the outcome would be the same.

And yes, the current industry is almost certainly doomed. Yes, Falcon and Falcon Heavy are already dramatically cheaper. But Starship, if it works, is something else entirely. Game over.

I did a back-of-the-envelope calculation that showed a fleet of 10 Starships, if reused as planned, having such an increase of capacity over what we have now that the current yearly launch capacity of the entire world becomes a footnote, a few percent of what those Starships will be able to haul into orbit by themselves.

And they're planning to mass-produce these.

Game over.

[1] https://en.wikipedia.org/wiki/Wile_E._Coyote_and_the_Road_Ru...

[2] https://en.wikipedia.org/wiki/Cost-plus_contract

[3] https://arstechnica.com/science/2017/07/elon-musk-knows-what...

Not that I like the answer, but the reason is assurance: it was not certain these alternatives would exist. And since NASA is beholden to politics with shorter cycles than rocket development cycles, it's also jobs program.

Last sentence: Politics.

> Men and women in all 50 states are hard at work..

> Why spend so much when these other options are available? Surely not ignorance. Is it malice? corruption? Someone's ego?

I went into my disappointment with Bezo's choice to focus on Amazon rather than take the lead at Blue Origin while Boeing is on its knees in both their commercial airliner and Aerospace program a week ago knowing something like his would occur. It just raised $25 Billion in a Bond sale [1] and now we're back to this stagnant entity clogging up the heavy lift system and we revert back to the same nepotism that has plagued aerospace for decades.

> SLS will cost upwards of a billion per launch. Compare that to $90 million for a reusable Falcon Heavy launch or ~$150 million for a disposable one.

But this helps clarify things, too.

1: https://finance.yahoo.com/news/boeing-raises-25-billion-bond...

I guess this is more target at people who want to keep up with NASA. NASA has to build SLS, the company knows that they are the only ones, so they can charge whatever they like.

Thankfully NASA is slowly moving away from insane stuff like this, but SLS/Orion are still done this way.

You just assume that there are other options that meet their requirements, which might not be the case

The requirements were made so that the SLS would be the only thing that would be rated to launch it

Is there another moon-capable rocket i dibt know about? And please dont say starship - its highly experimental, has never flows, and spacex has never flown people.

The motivation as proposed was to reuse already existing and certified STS (Space Shuttle) components in the next generation rockets, to reduce costs. The Falcon rocket family did not exist at the time that the SLS was proposed and green-lit.

Falcon Heavy didn't exist, but the Falcon family certainly did (and had flown).

Compared to $1m for a reusable Raptor that has more power at sea level.

There really should be painful monetary and career consequences for the people who makes these insane decisions. The reality is it doesn't matter a bit and they'll retire with a wall full of accolades.

Space isn’t easy. Going to the moon is significantly harder than LEO. I don’t think one can claim functional and performance equivalence between SLS to FH.

https://youtu.be/Y81vx__JngY

Now, if the argument is: Why go to the moon? Well, that’s a different conversation.

Jim Bridenstine sounds like a competent guy, but the fact is that NASA as an organization is no longer competent. They can't build rockets anymore. SLS has been a complete failure. The program was announced in 2011. The plan was to make an expendable vehicle based on existing RS-25 engines, existing SRBs, and the existing Space Shuttle external tank. This idea of reusing all of these systems was based on the previous Constellation project. Over the past decade NASA has spent $15 billion on SLS and they have yet to launch. Falcon Heavy went from a concept in 2011 to launching Elon Musk's car into interplanetary space in 2018. The development cost for Falcon Heavy was around $500 million. That is 30x cheaper and almost twice as fast.

Orion has also been in development for over a decade and has cost $16 billion. Current estimates put the total development cost at $23 billion.

If you add up the development costs of Constellation, SLS, and Orion, I am sure it would be close to $50 billion. That's for an expendable launch system that costs $1 billion per launch. For the price of a single SLS launch, you could develop Falcon Heavy and launch 5 of them. Obviously Falcon Heavy doesn't have all the capabilities of SLS, but SLS is vaporware. And by the time SLS is working it will have to compete with Starship. It seems totally insane to continue this program considering that in the best case, we will have a launch system that is 10x more expensive than its competitor.

I am not going to dispute the idea that NASA, just like any government enterprise, isn’t as efficient as private enterprise. Just look at what happened with CDC COVID-19 tests as a data point from another domain.

That said, having experience working both with SpaceX and NASA, I can tell you that companies like SpaceX would not exist without NASA’s research and experience. The simplest example of this that comes to mind, again, having worked with both, is NASA’s decades-long work in materials science for space. The cost of acquisition for such knowledge is absolutely incalculable —some of it measured in lives.

The real cost and development timeline of a FH would be dramatically higher and longer without NASA.

A fair question, then, might be: Why can’t NASA use this knowledge to develop products as efficiently as SpaceX then?

In order to answer this question you would have to have access to documentation spanning at least from MCR to CDR in order to understand what we are comparing against. A rocket isn’t just a tube full of fuel and LOX.

I am not defending anyone, just highlighting context I am privy to within the limits of what I can disclose. I can say we are working on a small portion of the Artemis Mission.

Firstly, this money is a pittance compared to the costs of development of F-22, which produced a habdfull of planes and thats it.

Second, the decision to use existing shuttle conoonebts was made basically by congress/ lobbyists. Congress gets insane amount of control over nasa, and can fund or de- fund individual projects. As a result, nasa has to tailor them to something congress will like and approve. This miopic managenent probably results in what weve got

A) Whataboutism B) The program cost for the F-22 was 67B. It produced more than "a habdfull of planes", it produced 187. The reason it ended up costing so much (per aircraft) is the development cost had to be spread across fewer airframes than planned. Truncating the purchase to 187 is perhaps one of the greatest military acquisition mistakes in my lifetime.

This should be the top comment, and that video interview from NASA should be mandatory to watch before commenting in this thread.

I had little hope for Bridenstine when he got the position, but I must say that I've been impressed time and again. That video was absolutely terrific.

Reply on news.ycombinator.com