NASA's mighty SLS megarocket for Artemis moonshots 'unaffordable' for sustained exploration, audit finds

the silhouette of a sea bird perched on a wooden outcrop from the water, watches with its beak agape as the flames from a massive rocket launching in the distance reflect on the surface of the water.
NASA's Space Launch System (SLS) and Orion spacecraft launch from Kennedy Space Center on Nov. 16, 2022, kicking off the Artemis 1 moon mission. (Image credit: Space.com/ Josh Dinner)

NASA needs to be more transparent about the costs and schedules associated with its Space Launch System (SLS) moon rocket, a newly released audit has found. 

SLS is NASA's rocket of choice for its Artemis program, which aims to establish a permanent, sustainable presence on and around the moon by the end of the 2020s. But that ambitious goal may be tough to achieve given how expensive SLS is, according to the audit, which was conducted by the U.S. Government Accountability Office (GAO) and released on Thursday (Sept. 7).

"Senior NASA officials told GAO that at current cost levels, the SLS program is unaffordable," according to the GAO report, which was delivered to the congressional committees that have jurisdiction over NASA's budget. The SLS program's cost "exceeds what NASA officials believe will be available for its Artemis missions," the report's authors added. 

Related: Is SLS worth the cost? NASA's new megarocket comes with a mega price tag

NASA has spent $11.8 billion since it began developing SLS in 2011. After many delays, the huge rocket debuted on Nov. 16, 2022, when it launched NASA's successful uncrewed Artemis 1 mission to the moon. 

An additional $11.2 billion was allocated in the White House's 2024 federal budget request for future work on SLS from 2024 through 2028. NASA plans to use these funds to develop core stages, rocket engines and other components for SLS, ultimately to increase the vehicle's efficiency as well as the amount of cargo that can be delivered to the moon for Artemis.

However, the baseline costs and schedules for this future work have not been established despite GAO's nearly decade-long concerns, according to the report, which was put together after interviewing NASA officials and reviewing the agency's current activities, SLS documentation and future plans. 

"NASA does not plan to measure production costs to monitor the affordability of the SLS program," the report states. 

Artemis 1 blasts off November 16, 2022.  (Image credit: Space.com/Josh Dinner)

A 2021 revision in NASA's policy had allowed the agency to stick to five-year production and operation cost estimates after the SLS first demonstrated its capability, and those estimates fit within NASA's overall budget. But the estimates cannot replace official cost baselines and are "poor tools to measure cost performance over time," the GAO said in the new report.

The audit also noted that NASA hasn't updated its estimates to include increasing costs, like those due to the delay of the Artemis 4 mission from 2026 to 2028. The space agency had awarded a contract of close to $2 billion to build the core stages for the Artemis 3 and Artemis 4 missions, but GAO's audit of the contract shows "the cost to produce successive core stages is increasing over time." 

This is not the first time that SLS costs have been scrutinized. Those concerns came as early as 2014, just three years after work on the SLS began. To monitor program affordability, the GAO had asked NASA at the time to outline costs for flying the SLS Block 1, which was being built for Artemis 1, as part of the future Artemis 2 and Artemis 3 missions.

"NASA partially concurred, but has not yet implemented this recommendation," the GAO said in a summary statement, which was also released on Thursday.

In late 2021, a report by NASA's Office of Inspector General showed that NASA will likely spend a total of $93 billion on the Artemis program between 2012 and 2025, and that each SLS launch will cost about $4.1 billion. A large chunk of the budget was attributed to hiring contractors in every U.S. state and more than 20 similar partners across Europe.

The GAO report notes a few steps recently taken by NASA to save costs, including using specific contracts that place the risk on contractors rather than the government, "but it is too early to determine the effects of such strategies," the GAO said in its summary statement.

Join our Space Forums to keep talking space on the latest missions, night sky and more! And if you have a news tip, correction or comment, let us know at: community@space.com.

Sharmila Kuthunur
Space.com contributor

Sharmila Kuthunur is a Seattle-based science journalist covering astronomy, astrophysics and space exploration. Follow her on X @skuthunur.

  • billslugg
    SLS can put 130 metric tons into LEO at a cost of $4.4B, this is $34,000/kg
    Musk projects Starship will put 300 metric into LEO at a cost of $2M, or $7/kg.
    There is no way SLS is going to thrive.
    Reply
  • Unclear Engineer
    Bill, I think your conclusion about what will thrive and what will whither is correct, but it seems as though you are comparing apples to oranges with the numbers, so far. The real cost of each program is a combination of how much it cost to develop the system and how much it costs per launch once developed. The intended number of launches by the two competitors is vastly different, and clearly favors SpaceX. But, so far, Artemis has one flight and SpaceX has none that put anything into orbit. Total costs divided by successfully inserted LEO payloads would favor Artemis at this point, but that is expected to change drastically. Do we even know how much SpaceX has spent on developing Super Heavy and Starship so far?

    It has always seemed weird to me that NASA is relying on other contractors to get things like "NASA's" lunar lander to lunar orbit independently of NASA. If they can do that, why not just have them take the NASA astronauts with them to lunar orbit?

    I think Artemis is basically the U.S. staying in the game until private companies actually succeed, in case the private companies do not succeed, or do not succeed faster than other countries' competing national space programs. At this point, it looks like SpaceX is one of the 2 major contenders, along with China, for actually doing something on the Moon I expect Congress will redirect NASA to use commercial transport for future Moon science missions as soon as it is demonstrated to be reliably available.
    Reply
  • billslugg
    My analysis was on cost per launch, not including development expenses. It is an apples to apples comparison. The ratio is 5,000 to one. It's dead, Jim.
    Reply
  • fj.torres
    Unclear Engineer said:
    Bill, I think your conclusion about what will thrive and what will whither is correct, but it seems as though you are comparing apples to oranges with the numbers, so far. The real cost of each program is a combination of how much it cost to develop the system and how much it costs per launch once developed. The intended number of launches by the two competitors is vastly different, and clearly favors SpaceX. But, so far, Artemis has one flight and SpaceX has none that put anything into orbit. Total costs divided by successfully inserted LEO payloads would favor Artemis at this point, but that is expected to change drastically. Do we even know how much SpaceX has spent on developing Super Heavy and Starship so far?

    It has always seemed weird to me that NASA is relying on other contractors to get things like "NASA's" lunar lander to lunar orbit independently of NASA. If they can do that, why not just have them take the NASA astronauts with them to lunar orbit?

    I think Artemis is basically the U.S. staying in the game until private companies actually succeed, in case the private companies do not succeed, or do not succeed faster than other countries' competing national space programs. At this point, it looks like SpaceX is one of the 2 major contenders, along with China, for actually doing something on the Moon I expect Congress will redirect NASA to use commercial transport for future Moon science missions as soon as it is demonstrated to be reliably available.
    Actually, we "do* know the development cost (so far) of STARSHIP.
    From Wikipedia:

    "SpaceX Chief Financial Officer Bret Johnsen disclosed in court that SpaceX has invested more than $3 billion into the Starbase facility and Starship systems from July 2014 to May 2023. Elon Musk stated in April 2023 that SpaceX expects to spend about $2 billion on Starship development in 2023."

    Plus the $3B Nasa is paying for HLS. Maybe. That might end up paying for the three/four prototypes doing the orbital refueling and moon flight tests.

    Note that that includes the multiple boosters and Starships sitting at Starbase waiting to launch. Its not just one vehicle. As to why Starship hasn't been to orbit, don't forget the 18Month delay inflicted on them by their enemies.

    But you're right that it's not oranges to oranges; SLS is a congressionally mandated kludge of a pork barrel jobs program for old space, to develop a single vehicle, whereas Starship is a commercial program to develop and build a whole *family* of vehicles, some reusable, some disposable, and others Platforms for space telescopes and habitats, plus two factories and two launch facilities with as many as four launch towers.

    A full life cycle cost analysis of the two full programs is not going to be 5000 to 1.
    It will be much worse because of the ground facilities and the habitats. Especially because of the ROI.

    What I'm keeping an eye out for isn't how Starship performs. Rather the question is what SpaceX does with the capabilities they are unleashing. Heavy lift is the least of the potential of the Starship family.

    With SLS, the moon is its ceiling, for Starship it is the floor.
    Reply
  • Unclear Engineer
    But that is based on some assumptions about the number of launches and the manufacturing processes to make the vehicles, plus the (assumptions to some degree) reusability of the launch vehicles. Just saying that SpaceX development costs are far from negligible, maybe even compared to the SLS, which is using a lot of repurposed parts and launch facilities, while SpaceX had to built all of that from scratch. It doesn't look to me like the actual hardware is cheaper by a factor of 5000, so much as the number of vehicles made per manufacturing infrastructure and the number of uses per vehicle has provided a much larger denominator for cost per launch. Of course, there is also the efficiency factor of SpaceX vs entrenched company processes, but I still think the estimated cost per launch is reduced by building more vehicles and using them more often in the commercial industry compared to the few-and-far-between NASA missions with one-and-done vehicles.

    Whatever, I think we agree that the Artemis program hardware is not likely to be long-lived, and that NASA will soon be buying rides on SpaceX vehicles. Still, the Artemis Program is an international consortium that has political as well as economic goals, and I expect the Program to continue, even with commercial launch and other non-government infrastructure. That seems to be the actual plan.

    NASA will probably also try to develop some competition for SpaceX, to be sure that U.S. national interests are not damaged by some problems with one company. Not to mention that NASA probably does not want to be stuck with monopolistic pricing by a contractor that is not beholden to it financially. NASA seems to be hoping that Blue Origin will get its act together and provide some real competition for SpaceX.
    Reply
  • fj.torres
    Unclear Engineer said:
    But that is based on some assumptions about the number of launches and the manufacturing processes to make the vehicles, plus the (assumptions to some degree) reusability of the launch vehicles. Just saying that SpaceX development costs are far from negligible, maybe even compared to the SLS, which is using a lot of repurposed parts and launch facilities, while SpaceX had to built all of that from scratch. It doesn't look to me like the actual hardware is cheaper by a factor of 5000, so much as the number of vehicles made per manufacturing infrastructure and the number of uses per vehicle has provided a much larger denominator for cost per launch. Of course, there is also the efficiency factor of SpaceX vs entrenched company processes, but I still think the estimated cost per launch is reduced by building more vehicles and using them more often in the commercial industry compared to the few-and-far-between NASA missions with one-and-done vehicles.

    Whatever, I think we agree that the Artemis program hardware is not likely to be long-lived, and that NASA will soon be buying rides on SpaceX vehicles. Still, the Artemis Program is an international consortium that has political as well as economic goals, and I expect the Program to continue, even with commercial launch and other non-government infrastructure. That seems to be the actual plan.

    NASA will probably also try to develop some competition for SpaceX, to be sure that U.S. national interests are not damaged by some problems with one company. Not to mention that NASA probably does not want to be stuck with monopolistic pricing by a contractor that is not beholden to it financially. NASA seems to be hoping that Blue Origin will get its act together and provide some real competition for SpaceX.
    Artemis goals are 90% political and 10% scientific and only exists because of China's stated goal of a moon base and their history of claiming sovereignty over other countries' (and international) territories. See: nine dash line, new oficial China map, and years of ignoring international law.

    If China's economy were to crash, ending their manned space program, Artemis would in an instant go the way of NASA's CONSTELLATION (not to be confused with the exploration organization of the same name in the brilliant game recently released, STARFIELD). Purely political, yes.
    Reply
  • billslugg
    Ariane 6 will put 21,650 to LEO for US$120M, cost is $5,500/kg. This plan is not sustainable.
    Reply
  • Wes I
    fj.torres said:
    Artemis goals are 90% political and 10% scientific and only exists because of China's stated goal of a moon base and their history of claiming sovereignty over other countries' (and international) territories. See: nine dash line, new oficial China map, and years of ignoring international law.

    If China's economy were to crash, ending their manned space program, Artemis would in an instant go the way of NASA's CONSTELLATION (not to be confused with the exploration organization of the same name in the brilliant game recently released, STARFIELD). Purely political, yes.
    Agreed, as a lifelong, avid fan of space exploration---It's political from a space race persepecitve, similar to the 1st-Man-on-the-Moon goal of the Apollo program. This time around it's the 1st human on Mars with a Moon base. The Artemis goals are about human space exploration and basic scientific discovery, not sustainability (or profitability?) necessarily. Unless the sustainability is whether it bankrupts the US, then who cares? At only $93 billion, even with major cost overruns, the money is much better spent on Artemis and the SLS, than geopolitical military conflicts and wars. Which effort produces and which one destroys???
    Reply
  • Unclear Engineer
    I think the "sustainability" issue is not whether it bankrupts a nation, but whether the nation has other, less expensive options to reach the same goals.

    But one of the goals that is re-emerging from the conflict in Ukraine and the increasing tensions with China is that a nation needs sufficient internal capabilities to meet its goals without interference by other nations. So, space programs with peaceful goals are a selling point for the development and maintenance of in-house technological capabilities that can also have military applications.

    With that other goal in mind, I understand why the U.S. is continuing to provide support to Boeing and other industrial entities long after their cost effectiveness has been superceded by newer companies. But, that can't go on forever. We are already seeing the type of consolidations in the launch industry that we have seen in the aircraft production industry following World War II.
    Reply
  • fj.torres
    billslugg said:
    Ariane 6 will put 21,650 to LEO for US$120M, cost is $5,500/kg. This plan is not sustainable.
    Neither is Galileo nor the A380.
    But some capabilities are funded for non-economic reasons, especially dual use systems like targetting constellations, heavy lift aircraft (especially in the age of RAPID DRAGON), and launch vehicles.
    Particularly in light that with globalization fracturing and the US ever more isolationist there is no guarantee those capabilities will be available when needed.
    National security can't be outsourced to low cost bidders.
    Reply