Saturday, June 3, 2023

Another SLS Story to Make Us Sick

The story entitled “A new report finds NASA has spent an obscene amount of money on SLS propulsion” was in Ars Technica last week, but like a bad night at Taco Bell, it has been sitting in my craw and making me sick since then.  I knew it was bad, but not all of how bad it is.  An independent report from NASA's Inspector General (IG), Paul Martin, (pdf warning) warns costs are so poorly controlled, it could jeopardize the entire Artemis plans to return to the Moon.

The 50-page report analyzed contracts that NASA has given to Aerojet Rocketdyne, for RS-25 main engines, and to Northrop Grumman, for solid-rocket boosters. The engines and boosters power the first stage of the Space Launch System rocket, which made a successful debut flight in November 2022. The rocket will launch astronauts for the Artemis missions to the Moon.

I'll start with RS-25 liquid engines, also referred to by their old name, Space Shuttle Main Engines or SSMEs.  I've been reading and reporting about cost overruns on these 1970s-vintage engines for years now.  In May of 2020, I 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.  Engines that will litter the floor of the Atlantic after one use.  Worse, that $584 million doesn't count the solid rocket boosters, the upper stage(s) or anything else. 

In the report, IG Martin points out that through some sort of creative paperwork (moving some costs elsewhere), the claimed price for the RS-25 is closer to $100 million.  "When calculating the total cost of the new RS-25 engines, NASA and Aerojet are only including material, engineering support, and touch labor (hands-on labor effort), while project management and overhead costs are excluded."  NASA and Aerojet are trying to achieve a 30 percent cost savings by the end of this decade, bringing the cost down to $70 million each.  

As I've said many times, the RS-25 engines are rated at 512,000 pounds of thrust and that's not a level that nobody else can provide. Both Blue Origin's BE-4 engines and SpaceX's Raptor 2 (and the newer Raptor 3) are in the same class.  Blue Origin sells the BE-4 for less than $20 million.  The Raptor's design price point is under $1 million.  I understand that price is a goal and the current crop of various earlier versions were likely more than that price, but I'm fairly sure they're not close to $146 or even $100 million either.  Note that they simply can't buy the BE-4 or Raptor engines.  Those are both methane-oxygen engines while the RS-25 engines burn liquid hydrogen and lox. 

But it's not just the RS-25s; the solid rocket boosters manufactured by Northrup Grumman went up in cost as well.  From the IG report:

The most significant increase was related to Northrop Grumman’s Boosters contract, which grew from $1.8 billion to $2.5 billion under Constellation and then from $2.5 billion to $4.4 billion under Artemis. Booster-related work also increased the contract’s schedule by 5 years beyond the original December 2017 launch readiness date.

I don't have any Constellation program documents to compare, so the 39%  increase from $1.8 to $2.5 billion is kind of meaningless to me.  The 76% increase from $2.5 to $4.4 billion on Artemis, though, is stunning.  These engines are built in spare SRB body segments left over from the Space Shuttle era.  Yes, they added another segment (now five, were four).  The IG report doesn't mention buying new SRB segments. 

In comparison, the RS-25 engines were always expected to be expensive, yet Aerojet’s contract grew from $1.1 billion to $1.5 billion under Constellation and then to $2.1 billion under Artemis.  With the same disclaimer about Constellation, $1.1 to $1.5 billion is 36% (about the same as the SRBs) and $1.5 to $2.1 billion on Artemis is a 40% increase (much less than the SRB increase).

The cost overruns are not without impact, and one of IG Martin's points is that the program has become so expensive that it jeopardizes the chances the return to the moon will ever happen.  Just the cost overruns for the propulsion system of the SLS rocket alone are costing NASA about as much as it will spend on developing two reusable lunar landers—SpaceX's Starship and Blue Origin's Blue Moon.  That's only a portion of all the other cost overruns in the Artemis program, from the Mobile Launch Umbilical Tower to the SLS Exploration Upper Stage.  Has anything associated with Artemis come in on time and on budget?  

The source article and IG report are worth a read if you're having a slow, rainy weekend.  It's a good overview of why government agencies and politicians shouldn't be allowed near contracts for this sort of work. 

NASA conducts a test of an RS-25 rocket engine on the A-1 Test Stand at Stennis Space Center.  NASA photo.



10 comments:

  1. And the cost of the F1-B (the modern version of the F1, with CNC, 3D printing and modern manufacturing and materials) was figured out at less than a million each even with development costs added in. That's the F1 variant where the engineers scanned multiple surviving F1 engines and then CADed it up.

    Even a plain-jane J2 (the precursor of the SSME/RS25) or an enhanced version, which Aerojet-Rocketdyne had kept current, would have been much less if built in a modern way.

    But, no, stupidity has occurred.

    NASA could even have reactivated any of the various J2 based aerojet engines, either the ring version or the linear version, for far less than tossing RS25s away.

    When NASA was looking at the next-gen Saturn vehicles, they concluded that strap on Hydrogen/Oxygen boosters using F1 engines would have been more cost effective and safer to use than solid rockets, and vastly more controllable.

    And J2-based H2/O2 boosters were also looked at, and some designs had 2-4 J2s.

    All far cheaper than the modern uncontrollable solid boosters or the overly engineered RS25s, which were designed to handle all aspects of flight, be relightable, reusable, useful in both atmospheric and vacuum environments.

    Sad part is... all of the issues with using solid boosters and the cost overruns, and using RS25s and the cost overruns, were known at the beginning of the Constellation and Artemis programs.

    Both NASA and Aerojet-Rocketdyne have looked into using a dumber version of the RS25s, which would have been easier to manufacture, had less parts, been far cheaper to build, and been disposable. But, no, gotta toss away $100M engines like they are cheap party favors.

    All of this could have been stopped early on. All of it. All of it was known and talked about from the very beginning. But, no, gotta keep pushing overly expensive and legacy products in order to keep the money flowing.

    Right now, if I was Nelson, I'd tell all legacy aerospace to go self-copulate and just go with SpaceX and Sierra Nevada and others that can produce inexpensive and actually functional hardware that exceeds what Boondoggles-R-Us can make.

    ReplyDelete
    Replies
    1. And, yeah, I know, I keep bringing up old engines and what was studied before, but, gee, isn't that what NASA is doing with the hardware for Artemis?

      Delete
    2. But those old engines were invented by pale stale White male Christians, and therefor were surely no good.

      Delete
    3. So was the RS25, which, if you scratch the surface, is a reworked J2 from the Saturn era.

      Delete
  2. https://www.inverse.com/science/nasa-sls-launch-cost

    ReplyDelete
  3. They should just hire SpaceX to put a base on the moon.

    ReplyDelete
    Replies
    1. There's no inherent limit to how much a company can do. The argument that "we don't want to be distracted" really doesn't hold water as long as they ensure that the cash flow remains positive for all the parts. I think SpaceX should diversify their efforts and put together "Team Mars", "Team Luna", and "Core Technologies".

      I understand that there might be a problem finding enough skilled engineers, but that's a matter of incentive and a good HR department. Elon needs to stop insisting that he interview all new hires, and find some people he can trust to whom he can delegate that job.

      Delete
    2. They have done that. But 'Team Mars' and 'Team Luna' are side divisions of 'Core Tech' as the emphasis right now is getting Starship working and into space.

      Though there are signs that Team Luna is working on the Lunar Starship. Which Team Mars will be riffing off of to do Mars Starship.

      You know, with the ever-changing versions of Starship and Booster, it must be pissing the Chicoms off when they get yet another outdated copy of plans and blueprints and modeling. I wonder how much the CCP has spent to spy on SpaceX?

      Delete
  4. People fundamentally don't understand how technology works and where it lives. It lives in people's heads who are doing the actual productive work. We no more have the technology to build the Saturn V than the medieval English did. The people that had that knowledge walked out into retirement decades ago, leaving dead paper for the hoarders thereof.

    ReplyDelete
  5. Why not just stack dollar bills to get to orbit? Horrible.

    ReplyDelete