Friday, October 29, 2021

NASA: Cut SLS Costs by 50% and We'll Use It 'Till 2050

According to a document found by Ars Technica, NASA has asked the industry for ideas on how to go about "maximizing the long-term efficiency and sustainability" of the Space Launch System (SLS) - both the rocket and its ground systems.  This was not directed just at prime contractor Boeing, but at any company in the industry.

The request, which appears on the website in the form of a Request For Information (RFI), opens with the following explanation. 

The National Aeronautics and Space Administration (NASA) invites industry to submit responses to this Request for Information (RFI) to assist NASA in maximizing the long term efficiency and sustainability of the Exploration Systems Development (ESD) programs, including the Space Launch System (SLS),  Exploration Ground Systems (EGS) and Cross-Program Systems Integration (CSI) office by minimizing production, operations, and maintenance costs.  NASA will use the information received from this RFI on a non-attribution basis for informing future acquisition(s) to address this challenge.

NASA says it sees itself as the "anchor tenant" of the system, and says they see themselves procuring one manned flight per year, which doesn't strike me as much of a pace of operations.  They then say industry will "market" the large launch vehicle to other customers, including the science community and other government and non-government entities.  

NASA says it wants to transition ownership of rocket production and ground services to the private industry. In return, this private contractor should build and launch the SLS at a substantial savings of 50 percent or more off of the current industry "baseline per flight cost."

The problem for us observers is that NASA has never published what that baseline cost is.  We have seen some figures that imply that the cost would have to be over $1 billion per launch and possibly well over.  In May of 2020, we found out that the first stage engines would cost $146 million per engine, so nearly $600 million ($584 m) for just the four engines of the booster core.  That doesn't count the solid rocket boosters, the upper stage(s) or anything else.  I don't think they could put the rest together for under $1 billion.

Ars asked the NASA communications office on Tuesday for this figure, but as of Wednesday morning there has been no response. In 2019, the White House Office of Management and Budget estimated the cost of one SLS launch a year at "over $2 billion." Subsequently NASA did not deny that figure, but it has not been transparent with taxpayers about the rocket's expected costs.

There's an ironic story here.  The SLS was first started as a make-work program in 2011 for companies from the Shuttle era, partly architected and pushed by Florida Senator (at the time) Bill Nelson.  He proudly announced the SLS would be delivered on time and under budget.

“This rocket is coming in at the cost of... not only what we estimated in the NASA Authorization act, but less,” Nelson said at the time. “The cost of the rocket over a five- to six-year period in the NASA authorization bill was to be no more than $11.5 billion. This costs $10 billion for the rocket.” Later, he went further, saying, "If we can't do a rocket for $11.5 billion, we ought to close up shop."

As most of you know, a decade latter Bill Nelson is the NASA administrator, he's still pushing the SLS, the rocket still hasn't flown, its first flight has slipped yet again to 2022, and while they've gotten farther in the last year than at any time before now, the cost is over $30 billion so far.  If not being able to "do a rocket for $11.5 billion" is a clue for being time to close up shop, not being able to do it for $30 billion should lead to the doors being nailed shut and the shop razed.

It's arguable that the SLS is a relic of the Apollo age; put it all up in one giant stack.  It makes me wonder about the prospects of using SLS until 2050 as parts go obsolete and impossible to get.  Virtually all of the new space companies are moving in the direction of reusability.  SpaceX has obviously led this effort, but other launch companies, including Blue Origin, United Launch Alliance, Rocket Lab, Relativity Space, and others, are all working on reusable hardware. The European, Chinese, Japanese, and Russian space industries are as well. After reusable boosters come refuelable upper stages, tug "boats" in space, and in-space assembly.  All of those can be launched into orbit on smaller and more affordable rockets.  

NASA Artist's Concept of SLS.

NASA will hold a virtual "industry day" on its SLS request on November 10 and desires responses from industry by January 27, 2022.



  1. Thing is, the ARES system (part of the Constellation program which gave us the Orion capsule) did everything that the SLS does, except better.

    A light launcher (kind of a Saturn 1B equivalent) using a tall-stacked recoverable SRB to launch crews using Orion or for light loads to LEO.

    A medium launcher for medium lift loads to LEO.

    And a heavy, configurable by the number of SRBs attached, to go from Heavy to Super Heavy.

    Killed by Obama, under Nelson's urgings. Why? Because Nelson wasn't getting any under-the-table money as far as I know.

    Dammit. We had all of this. Even back in the Saturn days, we had all of this. But.. noooooooo.... We have to reinvent the wheel every 8 years or so.

    And who's bright idea was it to use engines designed to be reused over and over again as a throw-away item? It would be one thing if they were cheap to build like SpaceX's engines (relatively speaking, their engines are cheap-cheap-cheap) but at $145million (or far more) per motor?


    1. Don't get too maudlin about ARES (or was it Constellation? This program has been resurrecting like a zombie since 2000!): That solid-first stage was an uncontrollable death-trap. Unpredictable startup side-thrusts would have flipped the thing over on the pad. The spring-mass inter-stage designed to "fix" the unsurvivable vibrations was a rube-goldberg bandaid on a sucking chest wound: Instead of unsurvivable vibrations, the entire vehicle was turned into a wobbly non-rigid pogo stick.


  2. SLS was never anything but a jobs program, and a major gift to the contract engineering community. I was on a contract for an engine manufacturer when the ramp up in hiring for it began, and the offered rates were astronomical(pun intended). I passed, because I wanted to stay in the Northeast.
    Why do you think that most engineering contractors support President Biden? He's been the best president for the contract engineering community since Johnson, according to some of the real old timers. I'm just a kid, with only about 40 years in the business.

    1. Absolutely. As transparent as DC gets is that it seems like all NASA contracts are giveaways to the "Cost Plus Contractors." Plus, I've got to believe Nelson was getting kickbacks on that all along the way, too.

      I only worked on a few NASA programs from a subcontractor, but I heard many times that the most remarkable thing about places like the JPL was that the workers still managed to get stuff done.

  3. It strikes me that, once again, government (NASA, in this case) is the problem rather than the solution. Sure, they still occasionally get it right, but I'm beginning to think those are the subversive works of people inside the agency.

  4. SLS is a jobs program like most everything else the government gets involved in. It seems the government is just there to make sure the job doesn't go away and all the politicians get to brag about the gravy that gets spilled in their state.

  5. "Welfare for rocket scientists" might make some sort of sense if we were actually allowed to be rocket scientists. (eg: these fields aren't profitable, and industry is a mess, but we need these kind of skills for military/other non-economic reasons...) Same with nuclear bomb physicists.

    Instead it's more like welfare for powerpoint monkeys. I worked for a time adjacent to the project, and no one was *doing* anything. No one was building and testing engines. No propellant chemists were at work on learning/improving the in-space propellant chemistry of the N2H4/MON propellant. (Quite complex: Read Ignition sometime: Those guys were all long gone and retired, and now it's all just blind flailing.) We were all shuffling powerpoint around and doing "analysis" on computer models. The few tests that existed were intended to proceed flawlessly so that people could check a box, because they certainly didn't have time/budget/spare parts for correcting discovered problems, or even just shaking out and characterizing a complex system. Pieces in crates from the old shuttle program needed to just slot together like legos because Congress decreed it to be so.

    Everyone was lying to everyone else about the status of their components: Deep problems with things like valves and diffusers were handwaved away because "it would put us behind schedule, and where else are we going to get them?!"

    It's a soul sucking mess, and I cannot express how glad I am that the first flight will be unmanned.


    1. It's a soul sucking mess, and I cannot express how glad I am that the first flight will be unmanned.

      As I understand it, this is exactly the thinking that condemned Challenger and Columbia. In Columbia's case, chunks of foam had hit shuttles before, but it never took out the vehicle and crew. Until it did. NASA seemed to have developed the attitude that if they played Russian Roulette with a six-shooter and didn't die after spinning the cylinder once or twice then it must be safe.

      The way I read this is that if the unmanned vehicle gets lucky the next flight is manned. That might get lucky once or twice, too. At some point, the luck runs out.

    2. Read Ignition sometime: Those guys were all long gone and retired...

      I'm seeing a couple of books that could be that one. Is that the one by John D. Clark and Isaac Asimov?

    3. That's the one.

  6. A rocket program, a real one, is really a manufacturing research program in disguise. Like so many other things, our computerized, powerpointized Toffler-lauded post-industrial society looks a lot like a pre-industrial cargo-cult.

    Elon Musk's people can build rockets because they are building rocket parts. (And testing, and launching, and learning, etc...)


    1. I'm sure you must have seen Elon Musk's insistence that manufacturing is the most important part of what they're doing. The money spent on design optimization is a rounding error next to what they'll spend when they build the numbers of Starships they're planning on.