The pressure to fix SLS and get to a launch for Artemis 2 will build. If there is another rollback in April or new leak or valve issue in May or long repairs. The program will get closer to reaching political exhaustion. This pressure will increase if there are clean launches of SpaceX Starship in March and April and SpaceX reaches full recovery and likely reusability of the version 3 Super Heavy Booster and Upper stage Starship. Reviewing the Artemis 1 history and what is happening so far in Artemis 2 is a history of helium and hydrogen leaks. They were one after another for Artemis 1 and have been one after another so far for Artemis 2.
Current Situation (as of Feb 24, 2026).
Successful 2nd Wet Dress Rehearsal (Feb 19) that fixed the earlier hydrogen leaks.
There is a new issue (Feb 20–21). Interrupted helium flow to the ICPS upper stage. This repeats a similar Artemis 1 problem. Possible filter, valve, or umbilical quick-disconnect problems.
March 6–11 window is officially dead.
Rollback to Vehicle Assembly Building begins Feb 25 (weather permitting) for full access and repair.
Earliest realistic launch is April 1–6 or April 30 but only IF the fix is straightforward and verification testing goes smoothly.
There have been 8 total failed WDR or scrubbed launches. If there are one or two more or long repairs then this could finally lead to the canceling of SLS/Orion.
Both Artemis 1 and so far for Artemis 2 missions they repeatedly hit the same two culprits — LH₂ leaks at the ground-to-rocket tail service mast umbilical and helium issues in the ICPS upper stage. Continuing these problems can easily cause months of delays that can lead to the final cancellation of SLS and Orion.
8 total failed WDR or scrubbed launches.
6 failed WDR or scrubbed launches for Artemis 1 and 2 for Artemis 2 so far.
How hard was the previous Artemis 1 fix of a helium problem? It was moderately difficult. It was straightforward technically, but logistically painful.
The check valve was not accessible while the rocket is on the launch pad. The only way to reach and replace it is inside the Vehicle Assembly Building (VAB) with proper platforms and access.
They rolled the entire SLS/Orion stack back from Pad 39B to the VAB (rollback began April 26, 2022) and they are doing the same for the similar problem now.
In 2022, they removed the stuck valve, cleaned out all contamination, replaced the valve and any affected seals/O-rings, and re-verified the entire helium system.
Artemis 1 helium fix time required: ~5–6 weeks total impact on the schedule.
Rollback: late April 2022, repair and restacking inspection May, back the pad June 6, WDR, then launched in November.
Artemis 1 passed its final (fourth) Wet Dress Rehearsal on June 20, 2022, but it was not fully clean or ready for immediate launch. The test ended successfully at T-29 seconds after demonstrating most objectives, but NASA had to use workarounds for a persistent liquid hydrogen (LH₂) leak at the tail service mast umbilical quick-disconnect on the Mobile Launcher.
The entire stacked SLS/Orion vehicle was rolled back from Pad 39B to the Vehicle Assembly Building (VAB). They had to replace worn Teflon seals and bellows in the LH₂ quick-disconnect, verify the umbilical, fix a separate gaseous nitrogen purge issue, and complete other final launch preparations.
This alone pushed the earliest launch target from late July/early August to August 29, 2022
There were two scrubs on launch attempts. RS-25 Engine #3 would not chill down to the required temperature. Faulty sensor reading. Later confirmed not a real engine problem. There can and were launch scrubs after good wet dress rehearsals.
The hydrogen leak problem did not stay fixed.
5 week repair would mean end of March to get out for a WDR and then April 30 launch. April 1-6 would be too tight.
Another failed WDR or launch scrubs would push to May/June. Long repair of 8+ weeks would also push to May/June for launch.
Trump, NASA and Isaacman have made return to the Moon a signature goal to beat China. Missing the entire 2026 calendar year for Artemis 2 would be politically toxic and hand China a PR win.
Each extra month costs hundreds of millions in standing army + pad time. The FY2026+ budgets already direct funds toward commercial alternatives. Repeated failures give Isaacman and Congress all the cover needed to say we are accelerating the transition now.
Three+ major cryogenic/upper-stage issues in a row on the same vehicle would trigger a full stand-down and independent review board (Columbia-style). The heat-shield, separation bolts, and Boeing quality issues would all get re-litigated.
Artemis 2 would still fly (even if delayed to late 2026 or early 2027) because the hardware is built and the milestone is too valuable.
Artemis 3 (lunar landing) would be de-scoped or fully re-architected. Orion/SLS replaced by human-rated Starship (or Dragon + Starship stack) for lunar-orbit crew transport, with Starship HLS as the lander. Blue Origin Blue Moon as backup.
NASA has full confidence the Artemis mission is safe. Independent experts and some former NASA engineers feel that the Orion heat shield problems are still a risk. They have not fixed it but just reduced the time that the heat shield is under extreme temperatures by changing to steeper and faster angle for re-entry.
Critics like former NASA heat-shield experts Charlie Camarda, Ed Pope, and others call Artemis 1 a major failure that was downplayed.
The models still don’t fully predict behavior. Trajectory change is unproven on this exact hardware.
Some estimate catastrophic risk as high as 1-in-5 to 1-in-50 .
If the modified profile does not work, then the worst-case is catastrophic heat-shield failure during the final 5–10 minutes of re-entry — no abort option, no backup capsule, crew loss.
Less-bad problem is heavy char loss that again requires major post-flight investigation, but the crew survives.
IF there is loss of crew then again no question SLS and Orion get cancelled.
If there is heavy charring and loss of chunks of the heat shield combined with the delays in April or later then the Artemis 3 change is possible. SLS/Orion cancelling after Artemis 3 would be pretty guaranteed.
If there are leaks and problems during the mission, it will not matter after 2 hours. SLS will separate from Orion after 2 hours into Artemis 2. The heat shield is the main concern after 2 hours into Artemis 2.
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 don’t understand the need to send humans. We can get to the moon faster with robots. And soon Optimus Prime robots will be ready and we can accomplish everything needed without risking human life.
[ me honoring(/saluting?) NASA for their documents and ‘educational-for-citizens/people’ efforts (being material science, explained and published, science concepts, calculation methods, project descriptions and missions/explorations to the planets) from the 60’s ongoing, ’til these days(?)
If that’s competitive compared to private companies?, some of it seems bigger than that, for society? (thx) ]
During Trump’s record long (1:46m) SOTU speech last night, he never mentioned NASA or the space program. No wonder. Both are an embarrassment at this point. So is Starship, which has had two launches shy of failures equal to number of successes of the Apollo program, if you discount Apollo 1 (3 deaths) and 13 (explosion, then missed the Moon landing & had to turn back). And of course, Starship is no where near Apollo 8’s orbit of the Moon & return to Earth.
Apollo 7 was in October 11, 1968. Less than a year later, Apollo 11 landed two men on the Moon and returned them safely to Earth.
It can’t just be more money; by now Artemis and Starship combined must have cost what a comparable number of flights cost Apollo back in the late 1960s, adjusted for inflation. The most expensive rocket and the one closest to the Saturn V is the SLS, and it’s also the closest to success. But a fuel-leaking rocket? I wouldn’t fly on that. Would you?
In the 1960s we had no personal computers, no cellphones, no color flat screens. Electronics were heavy and less reliable. Materials science was relatively primitive. Yet, the right stuff men, and a handful of women behind the scenes, made everything work, even the ill-starred Apollo 13. WTH happened to America since then?
NASA is only paying $3.5 billion for Starship. SpaceX is paying the vast majority of developing Starship. SLS and Orion are over $100 billion and that does not include $10 billion or so for Constellation which was the first false start. SpaceX is doing iterative development of Starship and is on Version 3 of the rocket and version 3 of the Raptor engine in order to get to 200+ tons of reusable payload. SLS is using a lot of old hardware. The $100 million RS25 engine which was reusable for the Space Shuttle but is tossed for every SLS mission.
Saturn V development cost: Approximately $6.4 billion in then-year dollars (appropriations from 1964–1973 for R&D and the 13 flights), equivalent to ~$34.5 billion in 2024 dollars.
Each Apollo flight (Saturn V + spacecraft + mission ops) in today’s dollars:
The direct cost of a Saturn V Apollo mission launch in the 1969–1971 timeframe was ~$185 million nominal, or ~$995 million–$1 billion in 2024 dollars. Full Apollo program context: Total U.S. spending on Project Apollo (1960–1973) was $25.8 billion nominal ≈ ~$257 billion in 2020 dollars
Development cost: SpaceX has invested at least $5 billion privately in Starship/Super Heavy/Raptor systems and Starbase facilities through 2023 (with ~$2 billion spent in 2023 alone); estimates for total program R&D to reach reliable operational capability are $5–10 billion (some analysts put it at ~$9–10 billion including ongoing work). This is dramatically lower than Saturn V ($35–66B adjusted) or SLS (see below). Raptor engine development was a small fraction of this (early USAF contracts were only tens of millions; the bulk is iterative private testing/production scaling). NASA has added ~$4 billion in Human Landing System (HLS) contracts for Starship variants.
Per-launch cost: Current test-flight hardware/build cost is estimated at ~$90–100 million per full stack (heavily driven by the 33–39 Raptors at $0.5–1 million each in current production). Operational marginal cost (with full reuse and mass production) targets $2–10 million per launch
Saturn V’s payload ~140 tons (expendable, one-time use).
Constellation program (2005–2010, precursor to SLS/Orion) Spent ~$9–12 billion nominal before cancellation by Obama admin.
Adjusted: ~$15–20 billion today.
This funded early Ares I/V designs, Orion capsule work, and ground systems. Most hardware/concepts were directly repurposed into SLS (Block 1 core from Ares V elements, Orion, 5-segment SRBs from Shuttle heritage). The sunk Constellation costs are part of why SLS/Orion feel like “older gear” rebooted at huge expense.
SLS costs (up to 2026, including heritage/Constellation roots)
SLS is expendable, Shuttle-derived (RS-25 engines + 5-segment SRBs), with Orion capsule on top. Development + production through Artemis I (2022): ~$23.8 billion for SLS rocket itself (Planetary Society; NASA official lower ~$11.8B by excluding some formulation/pre-Artemis).
Orion capsule (2006–2022, includes Constellation spend): ~$20.4B.
Exploration Ground Systems: ~$5.7B.
Annual spending 2023–2025: $2.5–2.6B/year for SLS. By early/mid-2026 (Artemis II/III era): SLS cumulative ~$35–40B+ nominal; full vehicles (SLS + Orion + ground) **$70–80B**; broader Artemis program (vehicles + support) on pace for $100B+ (OIG/Payload Research estimates through FY26).
Per-launch cost: Recurring rocket ~$2B+ (OIG/Payload 2024).
Full early Artemis mission (incl. Orion, ground, overhead) $4.1B+ (NASA OIG; Trump 2026 budget docs called SLS “$4B per launch and 140% over budget”).
The Saturn V 140 ton to LEO number is apples and oranges because it has Clive’s the dry mass of the S-IVB stage. More like 110 tons actual.
It includes the dry mass. Autocorrupt.
This is also apples to oranges comparison:
“Operational marginal cost (with full reuse and mass production) targets $2–10 million per launch. (Starship)
Saturn V’s payload ~140 tons (expendable, one-time use).”
Just launching Starship won’t do any useful work. It has to have a destination, and to be fully realized it also has to carry a large payload. There’s no point in using it for the ISS; the Falcon series does a better, cheaper job of that. OK, so the PAY in “payload” pays for some of the cost as far as SpaceX is concerned, but it’s still a cost to NASA and the taxpayers. And the payload (100 ton? 50 ton? A banana? (so far, the only payload ever launched)) adds fuel depletion and cost. Already, it seems Starship in its current configuration can’t achieve LEO with the promised 100 ton payload. Musk keeps elongating the rocket, but it would be easier to land if he just widened it…but then Mechazilla would have to be reconfigured too, and the launching pad. More cost!
Starship has to be at least a Moon rocket to make any sense. I believe Musk has privately given up on it ever reaching Mars. The refueling-in-space challenges are so immense that it may not ever do either safely, let alone affordably. Just getting more fuel to LEO, let alone midpoint to Mars, is an enormous logistical challenge (it would be safer just to dock a fully fueled second and maybe unique third stage onto the remaining upper part of Starship. There’s a reason Saturn had three stages, not just two). But that may be also almost impossible in space, and to get it there is almost as hard as launching a crewed rocket.
All of this will add billions, maybe 10s of billions, and too much time.
Starship is far away from being practical for its intended function, let alone safe.
China will almost certainly put people on the Moon first, probably a colony too. By PPP, China is now richer than America, and is diligent, serious and committed to the project, you know, in the way the U.S. used to be. And Trump is outright anti-science.
P.S. could you make it so that when someone responds to a comment, the user is notified by email? I generally check back a day later like this, but a notification would be much better.
The money payloads are all Starlink. High speed internet, Direct to cellphone and AI.
One million per year per satellite and rising.
V3 STarship is at 100+ tons to LEO in reusable mode and as they make some improvements over 2026 it will get towards 150 tons to LEO in reusable mode.
V4 STarship in 2027 at 200+ tons to LEO.
27+ V3 dummy satellites already launched. Obviously, operational satellites can and will be launched.
ISS and other missions are 10% of SpaceX revenue and falling. By 2027-2028, the NASA stuff will be 1% of SpaceX revenue or less.
Starship will be huge success, there will be 40,000+ V3 communication satellites up by 2030. There will be over 1 gigawatt of SpaceX AI datacenter in space before 2030.
SpaceX will land thousands of tons of mission onto the moon by 2032. The scale will be over 100 times what China will be able to do.
By 2035, millions of satellites.
Starship booster has already been caught 3 times and relaunched twice. SpaceX knows how to land rockets. They have done 600+ times for falcon 9. You have a bizarre smarter than SpaceX disease when you do not even bother to keep up with what has happened and why. Do this other mission and optimize for a client or mission that will represent less than 1% of your future business.
I do feel for NASA, they have on their hands an obsolete, unreliable, under-powered, and dangerous craft that is expected to restart the manned space flight exploration mission begun over sixty years ago and then abandoned by politicians in the early 1970s. A pause is warranted, to allow the funding wasted on the SLS/Orion boondoggle to save many of the current NASA science projects going into the trash pile.
I am also heartened with the way Jared Isaacman is setting out to refashion NASA and bring back into house the overly outsourced engineering positions. Good engineering is not an accident, something that is very missing on this terribly haphazard project!