Four years ago, unable to find a comprehensive summary of the ongoing abject failure known as the NASA SLS (Space Launch System), I wrote one. If you’re unfamiliar with the topic, you should read it first. 

It is hard to believe four years have gone by, but in all that time, the SLS has launched only once. Time flies when the rocket doesn’t. I don’t often write blogs in a sharply critical tone, so as always, the usual disclaimers apply. I write in my personal capacity as some Guy with an Opinion on the Internet. 

Four years ago, I wrote that the SLS was a cripplingly embarrassing national failure and a tragedy waiting to happen. That remains true, of course, but now I will go further and underscore that by continuing to humor this monstrosity, NASA has squandered its technical integrity and credibility. 

In ways that are far clearer to me now, after three years of running my own budget- and schedule-constrained hardware technology development program, I have more insight into the structural issues that evidently underlie every part of the SLS program. 

NASA managers routinely complain of difficulties in hiring and retention – difficulties they never faced 20 years ago, before the SLS and before the private space companies that, unlike NASA, are able to offer some combination of market-rate compensation, a career track that rewards ambition and competence, and a workplace that swiftly departs underperformers. 

Just imagine the mental agility required to actually want to work for an agency that continues to insist on technical doctrine no less absurd than “2+2=5” from top to bottom, from onboarding documentation all the way up to press releases, bilateral agreements and policy papers. Everyone at NASA knows the SLS is a looming catastrophe, but no-one can say it. Officially, it’s still the most powerful rocket ever built (except for Starship) and our official vehicle to the Moon and Mars! In reality, it’s insanely expensive, dangerous, and underpowered and can barely lift a reasonable payload to LEO.

Four years ago, I wrote that the best time to cancel the SLS was 20 years before, and the second best time was then. Four years on, the program has consumed another $20b with nothing to show for it. $20b, bringing total development cost to over $100b. This program burns $12m per day! 

In the meantime, NASA has abandoned all pretense of caring about or delivering cost control on any major project, with scope, schedule, and budget blowouts affecting practically every major program and forcing the cancellation of many of them. This is symptomatic of an agency who, compromising their technical integrity on their flagship program, subsequently lost the ability to maintain technical integrity anywhere else. 

Litany of other canceled and delayed projects

Mars Perseverance, a supposedly cheaper built-to-print replica of the Curiosity Mars rover from 2012 then required extensive re-engineering costing $2.4b, the same as the previous rover. At least it didn’t cost more! Meanwhile, much of its instrument payload was consumed by a sample handling system for a sample return mission that, at this rate, will never occur.

Mars Sample Return’s budget grew from $1.6b to $11b while the schedule slipped multiple years to the right. Somewhat beyond the scope of this post (I wrote another already) but worth pointing out that various elements within NASA assumed that it had grown too important to fail and acted accordingly, and while it was staffed up with dozens of strategy interfacing managers across a bunch of centers, they still hadn’t closed basic architectural trades – on a project that had been studied since the 1970s. This fundamental failure of project discipline is reflective of the broader problem – an agency tasked to execute the most challenging technical projects has lost its way in the wake of the ongoing SLS debacle. As of September 2024, MSR has already spent $3b before being “paused” indefinitely while mostly the same people who caused this problem try to figure out how to solve it. 

VERITAS (mission to Venus) indefinitely postponed due to budget constraints. SLS, like JWST before it, eats everything in its path, meanwhile NASA hasn’t sent a single mission to Venus since 1989.

HWO/HabEx/Luvoir and the new (bi)decadal survey. JWST, originally budgeted in 1999 for $1b to launch in 2007, ultimately launched in 2021 after spending $10b, so horrendously late and over budget that instead of re-imposing any kind of programmatic discipline or inventing a contract structure that doesn’t reward Northrop Grumman for wasting money, NASA instead has decided to delay the next big space telescope (currently termed the Habitable Worlds Observatory) into the 2050s. This may as well be forever, given that every human currently alive who knows anything about building space telescopes will be retired by then.

Dragonfly – a super cool nuclear powered robotic octocopter to explore Titan, was originally budgeted at $850m and is now pushing well beyond flagship status at $3.35b. How is it possible to miss this badly? Is budget estimation just some figleaf used at NASA to push the favored mission over the selection line, and then another set of books is used to actually build and run the mission? Are any of the program managers personally incentivized to innovate, improve productivity, or control cost?

VIPER – a robotic moon rover at the south pole. Originally conceived as the Resource Prospector rover from NASA Ames, then canceled in 2018 after spending $100m on development, the concept returned as VIPER. Its budget grew from $250m to $450m and most recently to $685m, exceeding a critical cost cap leading to cancellation despite being a supposedly essential part of Project Artemis. It has completed assembly and environmental testing, but will now be scrapped instead of launching on the new Griffin commercial lunar lander, which will instead fly with a mass simulator. 

Psyche – a mission to a metal asteroid. It missed its launch window due to a software issue discovered during final check out, growing its budget from $1b to $1.2b and pushing VERITAS into limbo. How did this happen? They completed flight hardware and shipped it to the cape but had skipped an integrated test bed test prior to assembly. When this was performed it was discovered that an aspect of the hardware design prevented the navigation software from working properly. Engineering a workaround took months, adding $200m to the budget.

NEO Surveyor. In 1998 Congress mandated that NASA map 90% of near Earth objects larger than 1 km – asteroids capable of destroying our entire civilization. Now, 30 years later, the mission has been pushed back another two years despite an increased budget, now expected to top $1.2b. The official line is that despite this delay overall costing more money, it saves money from the budget in the short term. Meanwhile, NASA’s mission to study existential risks to all of humanity? Shrug. 

Europa Clipper. Budget grew from $2b in 2013 to $5.2b. One instrument, a superconducting magnetometer, was canceled due to poor progress. This instrument’s technology had been built and flown before but the necessary expertise was lost due to retirements and an achingly low flight rate. During final testing before launch in a few weeks, a known problem with radiation-sensitive MOSFETs was rediscovered. Known, as in this problem was flagged as a blocker at about five different design reviews over the last decade, but for whatever reason (perhaps sunk costs) was ignored and pushed to the next phase. Well now the mission is integrated with Falcon Heavy and ready to launch, I sure hope it works properly when it gets to Jupiter. 

Ingenuity – developed for less than $25m, because JPL was spending its own money. The little helicopter was, like Sojourner, a business development tech demonstrator and, in the event, a formidable asset for Perseverance until it crashed on its 72nd flight. But instead of learning something about NASA centers rediscovering cost discipline when it’s their own money on the line, it’s now 5 years later and much of the talent that built it has moved on, with follow-on Mars helicopters stuck in endless studies.

CASIS – ISS National Lab. In an effort to defray costs and prolong the life of the station, NASA spent years building CASIS (an independent non-profit) and the ISS National Lab to find private customers. At just $4m per person per day, perhaps we shouldn’t be surprised that industry didn’t overwhelm them with attention, spending $75m to stimulate $150m of revenue (against an ISS annual budget of $3b). Its record was further tarnished by Charles Resnick, the chief economist on the project, expensing prostitutes. 

Chandra X-ray observatory, already launched and operating in space, is being defunded with 10 years of operational life remaining, because of budgetary pressures. 

Block 1B and the Exploration Upper Stage. Yet another configuration of SLS with a beefier upper stage, but only a marginal increase of launch capacity. Costs billions, a decade behind schedule, and provides no additional capability not already obtainable from commercial launch providers. 

Artemis space suit provider Collins backs out. For a program that’s been in development for almost a quarter of a century, it is a constant unpleasant surprise to find that critical complex pieces of machinery, in this case, space suits, are left to the last minute then farmed out to contractors. Collins, who once built space suits for Shuttle and Apollo, won a contract after NASA’s internal suit development program burned through a billion dollars with nothing to show for it. After spending more than $100m, Collins realized they had no path to success and asked to abandon the contract – NASA complied. NASA wisely also commissioned space suits from another commercial contractor, Axiom Space, except…

Artemis space suit provider Axiom Space in trouble. Axiom, one of several commercial entities attempting to build space stations, boasted an all star team drawn from NASA’s Johnson Space Center, plus a bunch of plump contracts for development. Starting in 2016, the CEO former NASA official Mike Suffredini staffed up 800 engineers and by 2023 was struggling to meet payroll. What’s the plan now? Land on the moon but don’t go outside? At the same time, SpaceX has developed their own suits, including a recent EVA-capable version used on Polaris Dawn. Does anyone else see which way the wind is blowing?

I haven’t forgotten Gateway and Starliner, they get a full treatment below. All these projects are pretty cool. All these projects should have been built and launched years ago. All these projects inexplicably take far longer and cost far more than they should, and no-one ever seems to drill into the why and how and let’s fix it!

NASA’s leadership has not risen to the challenge, and why would they?

While NASA HQ has thoroughly mismanaged Mars Sample Return (according to the independent review board) they are yet to return to in person office work post COVID, resulting in much of their DC office building being leased to other tenants. I don’t know about you, but if my space agency officials were so unable to do their jobs that half their missions were canceled and the other half delayed, I don’t think I’d assume that staying remote would fix that problem. 

Four years ago, I wrote about fixing this situation: “The required organizational transition is equivalent in scope and difficulty to the Apollo program, and will be made or broken during the tenure of the next NASA administrator.” This obviously has not occurred. 

As we inch toward the end of the Biden Administration, and without any particular insights into how Trump or Harris would change space policy, it pains me to say that Administrator Senator Nelson has not risen to this challenge. Aside from pushing Kathy Lueders out after she awarded HLS to SpaceX, we have little to show for four years of work across the agency. 

After I published my previous post on SLS, a number of NASA insiders called me over the following months to express profound pessimism that any kind of organizational transition or improvement could possibly occur. At the time I was optimistic, now I am not. 

Indeed, if we ignore SpaceX’s continuing achievements in launch, cargo and crew Dragon, Starship, and Starlink, NASA’s ability to run its programs have never been weaker. But for SpaceX, we would still depend on Russia for dodgy Soyuz launches to station, even as Russia continues its lawless rampage in Europe. But for SpaceX, both the US and Europe would lack adequate strategic access to space. But for SpaceX, Artemis would stand zero chance of success.

Let’s take a look at the SLS

Not only has no-one responsible for the SLS side of the Artemis program introduced the barest modicum of accountability, if anything the situation has gotten far worse. 

NASA has spent $20b on SLS and related programs in the last four years, so let’s tick off updates since my previous blog on this topic. $20b should buy a lot of progress, but if anything, this program is even further from any semblance of functionality than it was back then. 

NASA finally worked out how much each launch actually costs, and it’s even more than we thought

Four years ago, I wrote:

“Remember, even if it actually was fast and safe to reuse Shuttle hardware, even if the program was well managed, it would still only manage an SLS flight every year or two and cost between $2b and $3b per flight. Maybe more. Actually, no-one knows for sure. Can you stop asking? […] Even the people whose only job is to know exactly how much the SLS costs apparently do not know.”

By March 2022, we learned during a House Science Committee hearing from NASA OIG Paul Martin that the marginal launch cost for each of the first four Artemis SLS launches will be $4.1b. This doesn’t include any development costs, which will total $93b by 2025. Incredible! 

Since this report came out, the SLS launch schedule has slipped roughly two to one, so it’s safe to assume that the marginal launch cost has doubled (?) since then. How could we know?

It was after this embarrassing reveal that NASA tried and failed to sell the DoD on a cost sharing program, because …

The DoD hates SLS even more than I do

Four years ago, I wrote:

“I’m 100% okay with Congress cutting generous checks to politically important constituencies to develop hardware for human space exploration, but maybe, just maybe we should have worked out where we were going and what we were going to do there first? Can we now act surprised that SLS is, at best, an incandescently expensive turkey that’s not much use to anyone?”

Last year, we got an update from Joey Roulette at Reuters with brutal quotes from the DoD after NASA floated cost sharing to help reduce SLS fixed costs. Shuttle itself collaborated with DoD back in the 1970s to garner enough funding to complete development, resulting in a bunch of design compromises that significantly decreased its safety and utility. Still, perhaps SLS managers retained some hope… 

“It’s a capability right now that we, the DoD, don’t need,” Colonel Douglas Pentecost, a senior rocket acquisition official with the U.S. military’s Space Force, said in an interview. “We have the capability that we need at the affordability price that we have, so we’re not that interested in some partnership with NASA on the SLS system.”

“I don’t see the cost going down at this point to be competitive, just given the history and how challenging of a rocket it is to build,” said Cristina Chaplain, former assistant director of the Government Accountability Office (GAO), the investigative arm of the U.S. Congress.

“Even when they stabilize production, I don’t see them having quite the factory line that you need to get for that kind of thing,” added Chaplain, who led GAO audits of SLS.

“If they can sell it, they (Boeing and Northrop) get more flow through the factory, that drives down our recurring engineering cost,” Jim Free, head of NASA’s space exploration wing overseeing the Artemis program, said in an interview. “We’ve thought of a map to get there, we need them to engage, to agree with that map.”

Convincing about 400 suppliers across 46 U.S. states, already struggling with rising labor costs, to increase production and staffing would be another issue, according to Amit Kshatriya, the head of NASA’s new Moon to Mars office, formed in March to manage the agency’s Artemis and SLS strategy.

Even if Boeing and Northrup do not achieve NASA’s cost-cutting goals, the agency still plans to pressure them on reducing costs, Kshatriya said.

“We can’t just wait and hope that all these answers come as a part of this one procurement,” Kshatriya added.

Just go ask GPT-4 to script the conversation between NASA’s bloated and hapless program management offices trying to figure out how to shave cost from the SLS after two decades of their predecessors aggressively locking in the highest possible costs and lowest possible production rate. Jim Free, faced with this improbably catastrophic program, responded by cobbling together a few Powerpoint slides begging Boeing to charge less, yet strangely they’re unreceptive – probably too busy panicking after yet another excruciatingly embarrassing Starliner failure. (It wasn’t the last!)

Why doesn’t the DoD want the SLS? It’s 30 times too expensive, totally unavailable, and by the way, the solid boosters shake the crap out of any payload you put on it. What’s not to love?

I’m impressed by the degree of wishful thinking required to convince yourself that a) anyone wants SLS voluntarily even at half price, b) doubling production rate would halve fixed costs, c) $1b/year of fixed costs is even comprehensible, and d) Boeing is going to have a cushy sinecure for you in a few years, too. 

As of 2024, Jim is still the Associate Administrator at NASA. At the time of Roulette’s article, Kshatriya had just been promoted from “Assistant Deputy Associate Administrator for Exploration Systems Development” to “Deputy Associate Administrator for Moon to Mars Program” becoming Jim’s number two. They’ve been working this problem for three years now – surely long enough to see that success is not in the set of possible outcomes. At what point will they have the fortitude to report this obvious truth up the chain?

Aerojet managed to “earn” 10 times the purchase price for the SLS engines, which NASA already owns

As of 2020, Aerojet officially earned $146m per SSME engine that NASA already paid to develop and build and already had in a warehouse leftover from the Shuttle program. For reference, that’s more than the entire purchase price of a Falcon Heavy. Per engine. 

In 2023, we learned that despite Aerojet being paid $2.1b to recondition 16 of these engines for SLS, by the end of the contract in 2020 they had delivered only five. NASA’s inability to get a refund for these nonsense “services” already bought and paid for brought the taxpayer’s cost to re-purchase SLS engines to over $420m per engine. Once again, these are engines that NASA already owned – and that cost only $40m each to build in the first place. Not that that’s a good price, SpaceX currently builds the far more advanced Raptor engine for under $1m each, and launches the entire Falcon 9 rocket for less than $20m. 

Why is NASA giving $2.1b to a private corporation to (fail to) perform unneeded services on inappropriate engines for a rocket no-one wants? 

Just to hammer this point home, we’ll calculate the size of Aerojet’s nonsense grift in terms of Falcon 9 launches. For the price of one SSME, built new back in 2000 in 2024 dollars, SpaceX can launch 3 Falcons. For the newly updated reconditioning price, SpaceX can launch 30 Falcons, which is enough to re-launch the entire International Space Station from scratch. For the entire value of Aerojet’s $2.1b contract, SpaceX can launch 150 Falcons, which is roughly two year’s worth of launches as of 2024 flight rates. If NASA is in the business of launching missions to space, why is it giving 10% of its entire annual budget to a contractor only for engines that it already owns?

The SLS’s launch tower now costs far more than the world’s tallest building

In 2019, NASA awarded Bechtel a contract to deliver a launch tower – a glorified steel truss far simpler than the booster catching towers SpaceX assembles in weeks – by March 2023 for a total cost of $383m. 

As of today, the OIG reports that the tower will cost $2.7b and is to be finished by September 2027, but more likely 2029. For reference, the Burj Khalifa is seven times taller, contains paying tenants, hotels, and shops, and was built in five years for just $1.5b. 

If you had $2.7b in 27 million $100 notes, and you piled them up, they would be so much taller than Bechtel’s non-existent launch tower that you’d need not one, not two, but 23 separate piles to exhaust the supply. Whoever wrote Bechtel’s side of the contract certainly earned their bonus. Whoever wrote NASA’s side should be made to paint the entire structure with a toothbrush – but I expect they’ve long since been on Bechtel’s payroll in some kind of advisory no-show job. 

The Orion space capsule’s heat shield doesn’t work

The Orion capsule, which has also burned through tens of billions of dollars over two decades of development and flown in flight configuration exactly one (1) time, uses an old and extremely expensive heat shield technology, due to its supposed reliability. 

After the test flight in November 2022 (almost two years after the botched SLS Green Run test), you’ll never guess what happened! The heat shield failed in scary and unpredicted ways. NASA is a publicly funded agency and its information is meant to be publicly available. Despite multiple FOIA requests NASA successfully hid this embarrassing information from the public for nearly another two years, before eventually some photos of the damaged heat shield leaked out. It should be noted that this failure was not unexpected, and had been predicted since the material selection in 2009, but NASA management continued, and continues, to kick the can down the road.