SLS “Affordability”

This new “study” may be the beginning of the end for the program. This graf stuck out, though: “McConnaughey is leading the study for Kathy Lueders, NASA’s chief of human spaceflight. Even before the study’s initiation, McConnaughey had been pushing for the SLS program to become more cost-effective. One goal of this analysis is to find ways for the large NASA rocket to compete effectively with privately developed rockets as part of the agency’s Artemis Moon program.”

No one seems to ask the question: Why should NASA even be attempting to compete with private industry? This is not a proper role of a government agency, but we’ve been stuck in this mode since Shuttle.

38 thoughts on “SLS “Affordability””

  1. Here is the sentence that grabbed me hardest: “One source said the Biden White House may seek to fly SLS only a handful of times, halt work on the Exploration Upper Stage, and plan the future of Artemis around commercial launch vehicles.” Because this strikes me as the most likely (i.e., politically viable) path now to wind the program down, if that’s what Biden’s people decide they want to do.

    Homer Hickam tweeted a similar idea this morning. “SLS should be used for cargo only and limited to 4 flights. #obsolete . . . 3 or 4 are in flow. Use them up & get Huntsville going on advanced propulsion research and demos & out of the chemical Rocket business. . . Not looking to save money because it’s too late. FY2021 funds have already been obligated. To abruptly shut down the SLS contract would cost almost as much to continue it. Best to wind it down. Cargo should be lunar base infrastructure. See my 1993 study.”

    Sadly, he’s probably right that it would likely cost as much money to kill it this afternoon as it would to just launch the stuff in flow.

          1. There are bound to be enormous contract termination costs that you’d have to deal with. So, we either pay billions a year and maybe one day will get something (no matter how bad) for it, or we terminate the contract and have to pay billions to get absolutely nothing. Damned if you do, damned if you don’t.

          2. Hello Larry,

            Just so.

            That’s why, even though I’d want to cancel SLS *today*, I could live with Homer Hickam’s proposal. Because in all likelihood, it’s the best case scenario, politically, for terminating the program.

            I’m certainly anything but a fan of Joe Biden, but if his administration can make even that much happen, I think I’d tip my cap. It won’t be easy.

          3. “There are bound to be enormous contract termination costs that you’d have to deal with.”

            That’s been the big government aerospace and defense business model for a generation, now. Get a hugely expensive program going, take it all the way to CDR, then have the government cancel it due to cost overruns and dim prospects for success – and reap the term liability. The contractor never has to actually deliver anything, but gets paid almost as much as he would have if he had. It’s splendid “work”, if you can get it.

            Or, as Dilbert once said to his pointy-haired boss: “I admired your ability to get paid for this.”

            Debating whether to keep it going, and what to do with it if it is kept going, is arguing over how many angels can dance on the head of a pin. We will never see any angels dancing, and can’t stop the dance committee’s activities, so just give it up. It’s bigger than any of us.

          4. We will never see any angels dancing, and can’t stop the dance committee’s activities, so just give it up. It’s bigger than any of us.

            The self-licking ice cream cone gets liked.

            I still prefer cutting losses now. Cancellation fees are also sunk costs, even if they come later. Anything else is simply rewarding the model for building in the poison pill / golden parachute that protects it.

    1. Sadly, he’s probably right that it would likely cost as much money to kill it this afternoon as it would to just launch the stuff in flow.

      Doubtful. I have yet to run into the white elephant that was more costly to kill than to keep alive.

  2. If SpaceX succeeds in getting Starship into orbit, there would be little technical justification for continuing government subsidization of the less capable SLS booster, which is expendable and costs much, much more.

    Proponents of the SLS rocket are not blind to this. Some believe SpaceX will not succeed with its Starship program, and indeed myriad technical challenges remain.

    SpaceX’s Starship program, relative to the SLS, cannot fail, because even expendable Starships on expendable super-heavy boosters would be more capable and vastly cheaper than SLS, and obviously capable of far more than one flight per year due to the faster build speed.

    Others think NASA could find ways of making the SLS rocket more competitive, and that is one point of this study.

    That is likely impossible because the SLS is a design dead-end. The four RS-25’s are $150 million each, and even if they got the SLS SRB’s down to the Shuttle programs cost (4 segments, re-usable, higher-rate production, older dollars), they’d tack on another $50 million. Then the eventual four RL-10’s for the upper stage add another $70 million. So that’s $720 million per mission, minimum, just for the engines. They’d still have to build a rocket to mount them to.

    And as we all know, there’s absolutely no way to make it re-usable. At best they could re-use the SRB’s, which NASA now argues is actually more expensive than throwing them away, and my above engine prices assumes they were re-using those anyway.

    1. So that’s $720 million per mission, minimum, just for the engines.

      Meanwhile, you’ve got SpaceX cranking out Raptors for under a million bucks a copy, trying hard to shave down to $250,000.

      Aerojet has made noises about future reductions via economy of scale, but that really depends on increasing cadence – which requires more funding – and perhaps through nozzle improvements and 3D printed powerheads, which might shave 20% off the cost (be still my beating heart)… And we have not even got into the replacement boosters for the SRB’s, once the first 9 SLS launches have exhausted Northrop’s remaining component stores.

      So yes, if the engines are a fixed cost of this scale, I don’t see how production costs can be shaved in a major way. It’s going to be a hugely expensive rocket, period. And it is hard to imagine McConnaughey concluding otherwise.

  3. At best they could re-use the SRB’s, which NASA now argues is actually more expensive than throwing them away…

    The SRB’s were never really “reusable”; more like “rebuildable from scratch”.

    1. Yeah, but that’s what, 350,000 lbs of high grade steel (similar to 4340) that they didn’t have to buy for each launch?

      I wonder what it would cost to fish the two boosters out with a big electromagnet so you could sell them to a metals recycling center?

      And of course the four SSME’s, even if banged up badly by the impact, would still be worth quite a bit on Ebay, especially the ones that have flown on Shuttle missions.

      Sometimes I think we should all spring to put up a metal building at Boca Chica and tell Elon to just toss his scrapped test articles in there. They would make excellent museum piece someday, perhaps more significant than a lot of the early Apollo hardware.

      1. Apparently it is a popular tourist activity to walk around the launch site and collect bits of sploded Starships.

  4. The “affordability study” concept made me laugh. Why? Because, absent major design changes (such as different engines, which would require a different thrust structure, etc) it’s utterly pointless and cannot work.

    And that’s actually a pretty good description of SLS in general.

    Musk’s goal (and we know those usually slip) is for Starship to do an orbital launch in July (July 1st). That’s July of this year, BTW. That’s highly unlikely to happen IMHO, but it does make me confident that we’ll probably see Starship in orbit this year.

    And as others have mentioned, even going fully expendable on Starship/SH is vastly cheaper, and has a higher flight rate, than SLS. It also has more payload capacity (in mass and volume) through TLI. So, Starship seeing orbit before SLS does IMHO should be a further nail in the coffin of SLS. I concur with others who say that the first public sign of this will be the cancellation of the new upper stage.

    A prediction I’ve made before and will again; they won’t cancel SLS, exactly, they’ll simply do a provider change. To facilitate this, the Starship/Superheavy stack will be named the Starship Launch System (SLS).

  5. The only way to significantly drive down the cost of SLS would be to have an absurdly high flight rate, north of a dozen flights a year where they could mass produce the critical parts. As it is, it’s basically hand made.

    1. Pretty much everything that NASA has flown, attempted to fly, or blown up on the pad was hand made.

    2. “The only way to significantly drive down the cost of SLS would be to have an absurdly high flight rate, north of a dozen flights a year where they could mass produce the critical parts. As it is, it’s basically hand made.”

      There is no way to produce any new SSME’s. Once the existing stock is depleted, it’s game over. We’re paying $110 million an engine not to build a new engine, but to get an existing engine to work for one more flight. That’s precisely because each SSME was a one-off; no two of them are alike, and they were essentially hand-made. A program to produce new SSMEs would undoubtedly cost as much as has already been spent on the entire SLS program – and would take even longer to produce anything – if it ever did (see previous comment).

        1. I am not saying that they should do this, only that the capability exists and they are not limited to the stock on hand. What their current max throughput is? Only AJR knows.

      1. My recollection is, Aerojet Rocketdyne received a contract last May to revive the RS-25D production line, with a $1.7 billion payment up front, and an order to produce 18 new engines, representing 2 test engines and 16 flight engines (enough for 4 additional SLS cores). I think the current limiting factor is when the existing supply of steel SRB casings runs out, and there are enough on hand for seven flight sets, taking the program through Artemis VII. The program plan of record is through Artemis IX, accounting for the existing 4 RS-25D flights sets, the four new flight sets on order, and, I think, one flight set they can cobble together from refurbished test engines and spare parts. I don’t know if there’s a plan to come up with two more SRB flight sets, but 2029 is a little ways off.

        1. A big part of that $1.7 billion, if I recall, was for upgrades and improvements to the production process to make the engines cheaper and “more producible”, giving us the expendable RS-25E version, which costs more than the RS-25D because it’s cheaper, as some have snarked.

          2020 Space News article: AJR defends SLS contract costs

          The Aerojet contracts come to 24 engines at a cost of about $150 million each. Off the top of my head, I can think of 11 Raptors destroyed in flight tests or static fires, which would’ve been $1.65 billion and probably several years worth of production if they’d been RS-25’s.

  6. It is easy to compete when protected from market signals but also easy to lose. It is a good thing that government didn’t step in at the behest of its traditional contractors to prevent the competition.

  7. Given the issues that keep cropping up on the technical side with SLS, such as the aborted Green Run, and reports of software issues (Boeing software – and I note Starliner still hasn’t reflown) etc, and the inherent risks on any first launch, I wonder what will happen if Artemis 1 RUDs?

    That would be a very interesting situation for the powers that be, especially if Starship is flying successfully by that time.

    1. I don’t think you’ll see that situation. SLS will never launch, just like Ares V never launched. They’ll just change the name and pretend it’s a new program.

  8. The Starship Share Loot System (SSLS): Launches in expemdable mode, uncrewed to LEO. the Aretemis Stack, consisting, top to bottom, of Orion (LM), National Team transfer element (Northrop Grumman), Habitation Module (Thales Alenia) and Dynetics HLS, . Once in LEO, the crew is brought up by Starliner (Boeing). The Artemis Stack is put through TLI. After transposition and docking of the Orion/TE to HM/HLS, it goes into LLO using the transfer element. The HLS separates with two crew aboard and lands on the Moon. After a one week stay, it ascends and docks to the orbiting elements, refuels from the TE, exchanges crews and lands a second time. Depending one fuel supply aboard the TE, it may make more landings. The extended life support supplies for the expedition are carried aboard the habitation module. Which almost justifies the fact that the cost of the Artemis Stack is around 10x the cost of the ependable Starship LV, and fully justified by the fact that the budgetary largesse counts as a Congressional Blow Job.

  9. I want to make a small edit to your rational mission plan.

    “The HLS separates with two crew aboard and lands on the Moon. After a one week stay [ed. at the Elon’s Moon-tastic Lunar Casino and Resort, located in the tourist zone that’s at the midpoint of SpaceX’s industrial and commercial zones], it ascends and docks to the orbiting elements, refuels from the TE [ed. SpaceX duty-free Gas-and-Go], exchanges crews and lands a second time.”

    That’s assuming, of course, that Elon doesn’t have a few schedule slips.

  10. This is promising; it’s what the beginning stages of a climb-down from SLS would look like. And there has to be a climb-down sooner or later; eventually everyone is going to notice there’s no progress and no hope of any.

    OTOH, it’s a NASA study; they say whatever the front office tells them to say. I remind everyone that it was a NASA “study” that got us into this mess; it showed that Ares / SLS was the bestest, most affordable rocket evar. They could still screw this up, and that would delay a climbdown by another couple years. Which some people would dearly love.

  11. Meanwhile, today, we got a prominent defense of SLS in the New York Times. (Not paywalled so far.) It’s by David Brown, author of that new book on Europa Clipper.

    Link: NASA’s Last Rocket: The United States is unlikely to build anything like the Space Launch System ever again. But it’s still good that NASA did.

    It’s a target rich environment for you, Rand, but you’ve expended so much ammunition at this firing range I could hardly blame you for bothering at this point.

    “No other American rocket can send astronauts to the moon in a single launch. The Falcon Heavy, a large rocket built by SpaceX that has flown three times, is not certified to launch humans.”

    1. That is a target rich environment. The SLS can’t send astronauts to the moon on a single launch either, unless they decide to use lithobraking.

      And I could think of absolutely no other rocket that could be “man rated” faster than Falcon Heavy, since the capsule, upper stage, and core stage are already flying NASA astronauts, and the two side boosters are just two more first stage Falcons, with the modifications to make the whole stack work as intended.

      The reason the $100 million rocket, with almost the capabilities of the $2 billion SLS, isn’t being man rated is that Elon thinks that $100 million is too expensive compared to Starship/Super Heavy.

      1. The only big roadblock to man-rating FH than I can think of is that it hasn’t done 5 flights yet. (and yes, SLS is slated to carry crew on flight 2).

        And you’re absolutely right; the only way SLS can single-launch deliver astronauts to the lunar surface is at several kilometers per second. Seeing as the whole stated justification for SLS is the need to do missions in a single launch I can only thus assume that lithobraking is now an approved method.

  12. The Times article is about as good a defense as anybody can make at this point. What is striking is its defensive tone and minimal ambitions. SLS defenders are no longer claiming that it’s the key to America’s future in space. Rather, it’s now an ultra-cautious, extremely expensive hedging bet in case Musk doesn’t deliver Starship soon, and SLS does keep its much-delayed schedule this time.

    If that’s the best case its defenders can make for it, it’s not going to last too much longer. Although a Nelson-led NASA will certainly stretch out the agony as log as possible.

Comments are closed.