SLS (Space Launch System) uses a 50 year old rocket engine design that now costs $145 million each (not including labor and testing) which is over 300 times more expensive than the more powerful SpaceX Raptor 3 engine. It is a flying museum piece that is using difficult liquid hydrogen which was the reason for the leak and delay on the wet dress rehearsal preparation for the second flight in 20 years for the post-Shuttle SLS-Constellation programs. $100+ billion for SLS, Orion and Constellation to get one flight so far. The thinking after the Space Shuttle had 135 flights costing over $200 billion was lets keep using the RS-25 engine because it works and we can go back to an Apollo style rocket configuration. This will be tested, reliable and save some money. Less technical risk?
The last Space Shuttle lost was Columbia during STS-107 on February 1, 2003. The orbiter disintegrated during reentry over Texas due to damage from foam debris striking the left wing during launch, killing all seven crew members. This was the second (and final) major loss in the program. The first loss was Challenger on January 28, 1986. The last Space Shuttle flight was 2011.
The 50 year old RS-25 engine design costs $145 million. Modern engines like SpaceX Raptor (less than $1M) or Blue Origin BE-4 ($8M) are cheaper due to high-rate manufacturing. They are also not flying museum pieces.
SpaceX Raptor 3.X engines have serial production is ramping up. Over 75 Raptor 3 engines have been made by early 2026, and 3-4 engines tested per day at peak. 71 RS-25 engines have been made over 50 years. Each sea-level Raptor 3 will produce ~280 tf, equivalent to approximately 2.75 million newtons (MN) of thrust. Vacuum variants are targeted at ~306 tf or ~3.00 MN. Mid-2026, Raptor 3.x variants are expected to push thrust to 300 tf (2.94 MN) with minor iterative changes, such as refined chamber designs or propellant ratios. This could debut in Flights 13+ if testing succeeds post-Flight 12. In 2027, the Raptor 4 is could reach 330+ tf or ~3.24 MN) and have further weight reductions and cost drops to less than $250,000 per engine. This would be about 600 times lower cost than RS-25. Raptor 3.X can be over 50% more powerful than RS-25 and Raptor 4 could be 60% more powerful.
One RS-25 engine costs more than the entire 45 engine SpaceX super heavy starship. The RS-25 per unit engine does not include the extra testing and other costs associated with hardware that has flown once in 20 years and has had a handful of wet dress tests and 1 flight.
The Tail Service Mast Umbilical (TSMU) design for the Space Launch System (SLS) draws significant heritage from the Space Shuttle program’s Tail Service Masts (TSMs), which date back to the 1970s–1980s Shuttle era (with conceptual roots even earlier in Apollo/Saturn V designs from the 1960s).
The recurring LH2 leaks at the TSMU interface (seen in Artemis I WDRs 2022 and Artemis II 2026) stem from cryogenic challenges (super-cold LH2 contraction/expansion on seals/quick-disconnects). These were known in Shuttle ops (minor incidents like STS-112 close calls) but have been more problematic for SLS due to higher flow rates, different interfaces, or ground support tweaks.
LH2 must be stored at extremely low temperatures (~-253°C / -423°F), making it highly cryogenic.
Modern designs prioritize methane because it offers a better balance of performance, cost, and operability—especially for reusable vehicles. Higher boiling point (~-162°C / -260°F) → much easier to store and handle cryogenically. Often only passive insulation suffices, with far less boil-off than LH2.
Methane has larger molecules so it is easier to seal.
The RS-25 engine (originally the Space Shuttle Main Engine or SSME) was developed by Rockwell International (later acquired by Boeing, then Pratt & Whitney Rocketdyne, and finally Aerojet Rocketdyne in 2013, which was acquired by L3Harris Technologies in 2023). Aerojet Rocketdyne/L3Harris is the current prime contractor, but earlier entities handled Shuttle-era work.
The engine’s involvement across programs is as follows:
Space Shuttle (development started 1972, operational 1981–2011): Primary use and reusable engines.
Constellation (2005–2010): Planned for Ares rockets. Limited R&D on RS-25 derivatives (RS-25D/E), but program canceled before production. No engines built or flown. Mostly component tests and studies.
SLS/Artemis (development started 2011, ongoing). Uses 16 refurbished Shuttle engines for first four flights (Artemis I–IV). New production for Artemis V onward.
Total 71 RS-25 Engines Produced
Shuttle: 47 built (46 flown across 135 missions. 1 unused spare).
Constellation: 0 full engines (some prototypes/components for testing).
SLS: 16 refurbished from Shuttle stock used/allocated so far. New production: 24 under contract (6 from 2015 restart + 18 from 2020 extension). First new engine (E20001) completed in 2024; assembly ongoing for Artemis V engines as of early 2026. None flown yet.
Grand Total: 47 historical + 24 new = 71 (with more possible if SLS continues).
Estimated Total RS-25 Engine Revenue Across Programs — Without Inflation Adjustments
~$10–15 billion (Shuttle ~$6–9 billion. Constellation ~$0.2–0.3 billion. SLS ~$4–5.5 billion to date). This is conservative. Actuals could be higher with inflation adjustments and unpublicized mods.
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
“I think the secret of manufacturing of F-1 engine has been lost.”
The F-1 rocket engine needs a closer inspection for future development. There are plenty of Original F-1 engines still around to start off research with.
Modern materials, fuels and modern manufacturing processes should be brought to bear on this fantastic, old, F-1 engine design. I am not proud to say that NASA’s trip to the moon started with using these NAZI monsters’ technology, but that is how NASA got there, and NASA is still using that German technology today. China is stealing NAZI initiated (and now primitive) Russian technology as it is easier to understand and implement. A shortsighted and limiting approach, America can do much better!
You can track the development of all modern jet airliners from the original Junkers design?……
https://www.zona-militar.com/2021/09/15/junkers-b-47
The F-1 was a dead end.
SLS is worth the money in terms of keeping hydrogen infrastructure alive for NTRs.
Compared to Apollo it is a bargain. Both Elon and Bezos have far more money available to them…and Marshall is still going to beat them, and China, and Russia to the Moon for a second time…first with Apollo, now this.
Let the kids have LEO
An adult BEO program looks like this:
https://www.researchgate.net/publication/399883556_Six_Future_Deep_Space_Exploration_Missions_Enabled_by_The_NASA_Space_Launch_System
New Glenn uses plant of hydrogen.
Plenty
Another telescope? Seriously?
Neither spaceship is close to ideal. But the SLS has at least made a remote controlled loop around the Moon. Starship can’t even complete an orbit without something landing in the ocean that was supposed to be reusable, or exploding, after 13 attempts. It is not only not lunar flight incapable, it’s far from human flight safe, and who knows how long it will be until it is. That’s especially true once you consider the never-before-attempted space-refueling(s) necessary to get to the Moon. And then of course, there’s the issue of a lunar lander; there isn’t any. It’s hard to track all the competitor-partner options, but none of them will be ready this decade, or without serious money thrown at the problem, which is the last thing the Trump administration will do. Musk is maybe months, or even weeks away from a Kessler catastrophe which might make space unusuable for ANYONE for years. SpaceX just swallowed money-losing Xai (which swallowed money-losing X a while back).
Meanwhile, China makes gradual but consistent progress to landing a crew on the Moon by 2030.
America has lost a lot more than just the ability to return to the Moon.
Yes! Thank you!
Claiming that Starship is cheaper when it actually achieved NOTHING of the planned objectives is a completely dishonest characterization of the issue.
One thing that is ALWAYS missing in these analyses is that SPACE MISSIONS COST A LOT BECAUSE EVERYTHING NEEDS TO BE EXTREMELY RELIABLE.
SpaceX does not have a culture of reliability.
“Move fast and let things blow up” is exactly the opposite.
And you end up with glaring failures like cargo bay doors that do not open… The mission fails, but hey, the engines were cheaper… (and spoiler alert, in the long run all the failures will make the program more expensive)
Tell the users of Falcon that SpaceX doesn’t have a reputation for reliability. It will come as a surprise to them.
You’re confusing a finished product with a product under development. “Push it until it breaks, fix the bit that broke, repeat” is a standard approach to engineering in many fields. Then when it stops breaking you use it under less stressful circumstances, and have a reliable product.
Demanding that a test article being pushed to the breaking point so as to find where it breaks be reliable is a category error. Reliability is for product, not test articles.
The problem with programs like the SLS is that there are so few launches that EVERY launch is a test article.
It is quite reasonable even at this point to be confident that Starship will be cost effective, because it is already reliable if treated as a disposable launch vehicle, and would be cheaper than Falcon used in that way. The current development is focused on reentry so it can be reused, not on getting into orbit in the first place.
in 2021 SpaceX got 2.9B USD and claimed they could deliver the Starship Human Landing System by october 2024. They literally blew up all the money and they have nothing to show.
The Falcon program is for low-earth-orbit delivery.
Falcon 9 reliability is high 99.3% success, but please note that soyouz which is literally soviet era hardware has 97-98% success rate and the best in terms of reliability is still the United Launch Alliance Atlas V with a 100% success rate.
So, again, Falcon 9 is not outstanding, And if what you are launching is worth billions you might not care about saving millions, but utilizing a less reliable vector
Now compare that to around 24 billion cost of SLS and it only has been flown once. And this is exactly what Brett says: every SLS spacecraft is a test article.
Indeed, they were going to fly people on the SECOND SLS launch two days ago, when they had to cancel because it was leaking fuel too fast. It will never fly enough times in total to justify the man rating they’ve already gifted it, and they were going to fly people on it, and return using a known defective heat shield.
And that is what SpaceX’s competition looks like.
Musk isn’t perfect, he does have a tendency to over-promise. But the idea that he’s running a horrific, ill managed space company that’s just accidentally lofting about 97% of all payloads to space is a fantasy.
I expect that Starship will carry its first real payload to space some time late this year, and be in regular use before NASA is far enough along with it’s Moon mission plans to actually NEED SpaceX’s contribution.
scott baker, I have frequented this website for over a decade and this is the single dumbest comment I have ever read on it. Including that old troll Luca as well. It’s like a series of bad takes melded with some one off their meds. Kessler syndrome is weeks away? Really? Even if they purposely tried it the altitudes degrade in months. The richest man in the world with Starlink making Billions a year is bankrupt? You are not ok sir.
“…SLS has at least made a remote controlled loop around the Moon. Starship can’t even complete an orbit without something landing in the ocean that was supposed to be reusable, or exploding, after 13 attempts…”
They were given $40 billion or better and decades to do that one little stunt and none of it is reusable. When Spacex has $40 billion and decades invested “then” it will be a equal comparison.
Why would you use the old Raptor specs there? R3 has a much higher thrust and T/W ratio.
I would have still gone with further developing the F-1. It has the growth potential that the others have no hope of ever matching. The big disaster with the Saturn-5 first stage is it can not be reused.
I think the secret of manufacturing of F-1 engine has been lost.