SLS Costs

For years, NASA has been providing a BS number of a billion dollars a flight (with no basis). Now, at least the White House is admitting that it’s at least two billion.

And here’s an update this morning. Yes, five billion a flight is a low estimate.

22 thoughts on “SLS Costs”

  1. I ask a question for information only – not arguing:

    As I understand it, the SLS contract is cost plus fixed fee. An argument against that is that it provides incentive to stretch the program and make it cost more.

    Question: What payment structure would you suggest instead?

    1. SAA = Space Act Agreement
      Fixed Price
      Milestone based
      Company has to pay for each milestone on their own dime and they do not get paid until NASA is satisfied with the work.

  2. For what it’s worth, the exact wording of the recent White House mention of SLS cost per flight being $2 billion was “..estimated cost of over $2 billion per launch for the SLS once development is complete…”.

    My guess is that this is simply adding the recent $900 millionish per-copy estimate for the not-yet-developed EUS Exploration Upper Stage to the $1 billionish per-copy NASA cost claim for each SLS main stage flight. I think it’s EUS development completion (and thus the higher-lift SLS 1b) that’s being reffed by “once development is complete”.

    The $2 billion per SLS flight cost is still complete nonsense, of course. Even ignoring sunk development costs, the way to bet on such NASA pork programs is that annual flight-program budget will stay about the same as during development. Everything will still be hand-built by the same bloated cost-plus organizations, and major layoffs won’t be politically acceptable (even if they might be technically practical.)

    Given the SLS infrastructure will support perhaps one flight per year, estimating cost per flight is relatively easy. ~$2 billion a year for the SLS main stage establishment, another ~half billion a year for the launch support structure establishment, I’m guessing another ~$1 billion a year for the EUS development establishment, plus an additional ~$1 billion a year for the Orion establishment (whether an Orion is on the mission or not, they’ll still be on the NASA budget.)

    So, likely something on the rough order of $4.5 billion per SLS 1b mission, WITHOUT even trying to allocate development costs.

    1. annual flight-program budget will stay about the same as during development.

      What? You mean there won’t be a mass layoff in Alabama once the project development ends and construction moves to Louisiana with operations moving to Florida? Why would you make such a bet?

  3. The really sad part is, I’m not sure even *this* will stir any change in Congress, given that the FY 2020 appropriations bills of both houses are proposing nearly a billion dollars in extra funding for SLS over the administration request.

    Senior Senator from Alabama:”$2 billion, huh? Clearly, we need to increase funding for this program.”

  4. Surely someone will look at those numbers, look at Falcon 9 Heavy and Starship, and start calculating how many Navy hulls or military aircraft it could be spent on instead.

    1. They have, which is why I’ve been reading article after article about how USS Ford CVN will cost more than expected and operate less than the Nimitz. See how much we are wasting on carrier development for the Navy when they can just build more Nimitz carriers? And lets not get started on the F-35 [because Lockheed is our prime for Orion too].

  5. The payload of 95 tons of $50 bills is about what each launch will cost. Compared to Falcon Heavy’s 70 tons for $90 million, (it will carry almost enough $1 bills to pay for the mission) that’s about forty times as much per ton for SLS.

  6. Space.com: Elon says Starship may cost $2 million a mission

    The Starship system, which consists of a reusable 100-passenger spaceship stacked atop a huge reusable rocket known as Super Heavy, will use just $900,000 worth of propellant to get off Earth and into orbit, Elon Musk said here today (Nov. 5) at the first U.S. Air Force Space Pitch Day.

    “If you consider operational costs, maybe it’ll be like $2 million” out of SpaceX’s pocket each time, Musk said during a conversation with Lt. Gen. John Thompson, commander of the Space and Missile Systems Center, Air Force Space Command, at Los Angeles Air Force Base.

    I think Elon is overly optimistic, but even with lots of padding, his prices are still in a different universe from the SLS. Knocking three zeros off is like trading a Ford F-350 for a Lockheed Martin F-35.

    1. I too think Elon is being very overoptimistic on that 2 million per flight figure. That’s also just op costs, no R&D, no development. Actually, it’s not even that – the cost of carrying the entire support infrastructure (personnel, pads, transport, etc, etc, etc) is not included.

      So, let’s assume Elon is off by an order of magnitude. And then let’s double that. So, 40 million per flight. And that’s for a system that’s inherently far more capable than SLS is every metric save one; high speed probe launches. For that, SpaceX would need to make an expendable, stripped-down “Starkicker” (A stripped-down non-reusable Starship), refuel it in a high energy orbit, and expend it (you’d end up with more Delta/v than SLS could manage, and that’s for a probe several times the mass SLS could manage to the same Delta/V). That’d cost a lot – probably close to a hundred million (most of it R&D) to make the first one. So, let’s call it 100 million per launch (And ignore the fact that the R&D isn’t done for every single launch). That’s 5% of even nominal production cost for a SLS.
      Oh, and with “Starkicker”, you might also have the option of retaining some of that delta/V for the probe’s arrival, such as entering orbit at the destination, a very handy thing IMHO)

      I don’t have the math skill to prove it in a way a congresscritter could comprehend, but I’m of the opinion that 100 million is less than whatever number of billions an SLS flight would entail.

      1. It’s far worse than that. Space X could fly hundreds of Starship missions to LEO to assemble a construction shack, then fly hundreds of more missions to keep hundreds of workers supplied as they build a deep-space stage, by hand, and it would still come out cheaper than a single SLS deep-space launch.

        At $5 billion a launch, with 5 astronauts, an Orion seat would cost $1 billion – per flight. At $2 million, with 100 passengers, a Starship seat would cost $20,000. So for the same cost as putting one butt in a seat on the SLS, Elon could put 50,000 people into LEO. It’s like buying a Virginia class nuclear attack submarine for the cost of a single Mk-48 torpedo.

        1. Cost isn’t price but even with a healthy markup, that’s still an impressive difference.

          I want to know how Musk determines his cost. How many reuses before the per flight cost is that low or is that the one off cost? How many refueling launches to support a passenger ship?

          The last time I looked at this, I was wondering how many reuses would it take to get a ticket price into the low hundreds of thousands of dollars. But now, it looks like it could be under $200k for a trip to Mars. That opens a lot of doors.

          Or, if someone made an in-space transit service, how much would it cost to go from LEO to anywhere? Could it be done cheaper than using SpaceX?

  7. SLS need skin in the game. Announce that “ready or not” SLS with top brass procrastinators on board will launch in 12 months. Go.

  8. Until Elon gets this thing to the moon or mars, NASA is still the only game in town. And it is still immoral to allow suckers to keep their money, so SLS will eat as much money as we let them throw at it. Since it hasn’t flown yet either, Starship cost estimates are just as accurate as SLS. NASA will not be flying Starship, so a retired NASA employee assigning numbers to this is of doubtful reliability. I like Wayne Hale’s work, but on this subject, I don’t believe he is in a good position to know.

  9. The cost of a fully refueled Starship in LEO will bottom out at $12mln due to tanker requirements. I think the $2mln prediction is like the $7mln prediction once bandied about for a fully reusable Falcon 9.

    I will note Musk seems to have come up with ways to sharply reduce the infrastructure costs by using cheap hangars and Sprung structures in place of things like VAB. Also the SuperHeavy launch mount going in as LC-39A Annex doesn’t have a flame trench lined like a blast furnace. Instead, it has a water cooled elevated steel structure. That’s going to be a lot cheaper than pounding concrete pilings into a barrier island, a lesson SpaceX learned from Boca Chica.

Leave a Reply

Your email address will not be published. Required fields are marked *