The Myth of the Low Cost HLV

I commented on this over at hobbyspace, but felt this deserved its own quick little blog post.

Every now and again, people will trot out a tired argument that HLVs are proven to be much cheaper per pound than smaller vehicles. Or that expendable HLVs are much cheaper than reusable ones like Shuttle. The people then trot out the marginal cost of the Saturn V: ~$431M, and say “look, it only cost about $1700/lb, that’s cheaper than anything we have today”, and usually go on to conclusions about how “if we hadn’t retired the Saturn V, we would have been able to accomplish so much more”.

There’s a couple simple problems with this analysis:

  1. You can’t compare costs in 1967 dollars to 2006 dollars without factoring in inflation. If you use the inflation calculator, that $431M in 1967 is supposedly equivalent to nearly $2.5B in 2006 dollars. That’s nearly $9k/lb, which is three times worse than existing EELVs. Even if you disagree with the methodology of that inflation calculator, some inflation has occured. Here’s a link to a NASA site with several different inflation calculators. Only the most optimistic two would place the marginal cost of a Saturn V at less than $2B each, and the worst was nearly $3B each.
  2. You can’t compare just the marginal cost of a government program to a commercial program that also has to cover its fixed costs. A fair “apples-to-apples” comparison involves factoring in both fixed and marginal costs. When you choose to keep a vehicle around or to use a vehicle, you have to pay for those fixed costs. They don’t come for free, so just handwaiving them away doesn’t work. According to the Wikepedia site, about $6.5B was budgetted for Saturn V over the period of 1964-1973, with the peak being $1.2B in 1966. Adjusting for inflation (and assuming naively that you can just divide the total program cost by the number of years to get an average since more detailed info isn’t easily available), you get the equivalence of somewhere between $3B-4.5B per year. The maximum flight rate for Saturn Vs was 4 per year. However, you spend that amount whether you launch one or four missions. Sure, you could probably cut back a bit if you were going into a one-flight per year mode, but the end result is you’re talking at best $3.5B per flight of Saturn V (assuming 4 flights a year), and at worst something in the $4.5B per flight for a one-flight per year tempo (assuming you could halve fixed costs and still keep a flight per year in the system) That comes out to nearly $17k/lb to orbit, which is actually more expensive per pound than Shuttle at 4-5 flights per year!
  3. Even keeping one-two Saturn V flights per year going would have cost $4.5-7B (in 2006 dollars), which would’ve been something like 50-70% of NASA’s total budget.

Just food for thought next time anyone pines about the loss of the Saturn V. With the amount the NASA budget was slashed by Nixon, had they kept Saturn V, they could’ve afforded almost nothing else. There would’ve been no money for big space stations, there would’ve been no money for a continuing moon program, there would’ve been no money for Voyager and all the rest.

And the sad thing is that we’re making the same mistakes again. We’re tying ourselves down to expensive vehicles that cost billions of dollars not-to-fly, and billions more to fly occasionally. Dumb idea. I wish NASA’s inability to learn from the past wasn’t dooming us to have to repeat it with them. Here’s to hoping that this time around we have a solid alternative. Then NASA may end up facing what we have to out here in the real world. You either learn and adapt, or you die.

Anyhow, that’s enough for now. It’s off to work now.

[Update: I was looking through the numbers I referenced, and some of them may be at least slightly incorrect. Apparently the $431M per launch number actually came by dividing the other number (the $6.5B) by the total number of Saturn V’s flown. So that may have included development costs and fixed costs as well as marginal costs. However, there are a bunch of other funding categories listed that might also be “fixed cost” related that weren’t included in the $6.5B number. If the total cost per flight was really “just” $431M in 1967 dollars, then the numbers aren’t quite so bad. It’s only 70% as expensive as the shuttle…which is to say still more expensive than most other launch vehicles. But it’s hard to tell 100% what the fixed costs were, and therefore how much it would have cost to keep Saturn Vs flying had it not been for Nixon “being so shortsighted”. Does anyone have any better data they can bring to the table? Stuff that more clearly separates out what percentage of the other categories are actually Saturn V operations related?]

The following two tabs change content below.
Jonathan Goff

Jonathan Goff

President/CEO at Altius Space Machines
Jonathan Goff is a space technologist, inventor, and serial space entrepreneur who created the Selenian Boondocks blog. Jon was a co-founder of Masten Space Systems, and is the founder and CEO of Altius Space Machines, a space robotics startup in Broomfield, CO. His family includes his wife, Tiffany, and five boys: Jarom (deceased), Jonathan, James, Peter, and Andrew. Jon has a BS in Manufacturing Engineering (1999) and an MS in Mechanical Engineering (2007) from Brigham Young University, and served an LDS proselytizing mission in Olongapo, Philippines from 2000-2002.
This entry was posted in Uncategorized. Bookmark the permalink.

11 Responses to The Myth of the Low Cost HLV

  1. Anonymous says:

    When I was working at Northrop in El Segundo, I checked out this interesting study of the Big Dumb Booster Concept that was done before the Moon Launch when the company was still TRW. One point the author was harped on was that Big was not cheaper than medium or even small launchers. What mattered most was a high flight rate. If you can launch once a week for a year, you are probably going to make money.

  2. tankmodeler says:

    >>What mattered most was a high flight rate.

    The Holy Grail.

  3. Lee Valentine says:

    Assuming that the Saturn V had been kept in production at a rate of four per year since 1970, we would predict a decreasing cost per unit because of the learning curve.

    After roughly five doublings, you find, assuming a cost decrement of 15% per doubling, that the cost has fallen to about one billion dollars per Saturn V.

  4. Karl Hallowell says:

    Lee,

    What’s the justification for that sort of cost decrease? I mean why that particular rate? I’ve never seen that rule of thumb before.

  5. Lee Valentine says:

    Karl, Clayton Christiansen cites it in,IIRC,” The Innovator’s Dilemma”. I’m not sure of the original provenance, but he says it’s been shown to be a good number over many different manufactured items. The specific number is a cost decrement of 15% for each successive doubling of the number of units produced. Obviously, there’s an asymptote.

  6. tankmodeler says:

    >>After roughly five doublings, you find, assuming a cost decrement of 15% per doubling, that the cost has fallen to about one billion dollars per Saturn V.

    Lee, I don’t doubt the reduction of costs with increased production nor, necessarily, the number you quote, however, to get to the “five doublings” that’s 20 Saturn launches and from a start of , what, 3.5 Billion each in 2006 money? That’s a LOT of cash to get down to a less expensive heavy lifter. Even the Ares V is looking cheaper than that. And there’s nowhere near enough upmass needed to require that many big boosters, at least not at those initial costs.

    It’s very much a chicken & egg thing at those costs. No-one will put together a program that will eventually get you down to 1 billion per launch if you have to spend 45 billion to get there. You need a program that starts at 1 billion a mission (with several launches per mission) and then build a program around that. You’ll start to see your cost savings again, but the arcitecture has to be formed around smaller payloads.

    Lets face it, if you break out the various payloads of the existing plan into their individual bits, what do you really need a huge booster for, anyway? To go to Mars even NASA admits they are going to have to do assembly in space, so lets leverage the ISS experience and make it so, right now. Launch it small and build it in orbit and everyone wins. Everyone. Except, of course, the blue-collars at the NASA plants and the white-collar hangers on.

  7. Lee Valentine says:

    Sorry, I was not clear about the meaning of successive doubling. What I intended was the series 8,16,31,64,128. The point was that the cost could be expected to drop from learning curve effects.It’s quite expensive to get those per unit cost reductions.

    In addition, there were plans to increase the performance and to parachute the first stage into the Atlantic for recovery and reuse.

    It seems to me that the cheaper way to do things now is with in space assembly and smaller launch vehicles.

    The point was not to argue for the Saturn V, just to point out an effect that Jon didn’t mention.

  8. tankmodeler says:

    Ahhh! Gotcha.

    I agree. 🙂

  9. Mike Puckett says:

    I also wonder what effect uprated engines, modern electronics and lightweight tanks would have had on Saturn V performance. I would impagine that by modern standards, there is alot of fat that could have ben trimmed.

    Perhaps you could have added 50 tons or so to its lofting capacity and further reduced the cost per pound per launch.

    Not saying reproducing the Saturn V is the way to go, just wondering what might have been if instead of directing NASA to build the shuttles, they were directed to develop a lower cost, common element, partially reusable Saturn family of vehicles.

    Basically what some are thinking about today with the EELV’s but going from big to small instead of small to big.

    I remember reading about a Stage and a Half(think original Atlas) partially reusable derived from the S-IC and F-1 concept that could have orbited around 30 tons.

    Also, automation and reducing the standing workforce to shuttle levels or less could have made big strides towards cost reductions.

    A lot could have been done with the 75 billion dollars or so dumped into developing,, operating and fixing and re-fixing the shuttle these past 34 years.

    The road not taken…..

  10. Jon Goff says:

    Mark,
    I also wonder what effect uprated engines, modern electronics and lightweight tanks would have had on Saturn V performance. I would impagine that by modern standards, there is alot of fat that could have ben trimmed.

    Probably correct. Using a friction stir welded monocoque or semi monocoque structural design, modern electronics, engines using modern controls and manufacturing processes, etc could probably reduce the manufacturing cost while upping the payload a bit. Maybe even as much as 20-30% or more. I’m not sure.

    Perhaps you could have added 50 tons or so to its lofting capacity and further reduced the cost per pound per launch.

    But alas, that doesn’t really get you that much. The problem is that at the point we were at at the end of Apollo, keeping the Saturn V flying at all just took up too much of NASA’s budget. Even if the price per pound goes down, if the price for keeping the program alive and flying one time a year are too high, it wouldn’t matter if you could triple its payload to LEO.

    On the low and the high end (ie for very small launchers and very big ones), mission costs tend to be more imporant than cost per pound on orbit. XCOR and several others (ourselves included) think that you could do a relatively low-cost upper stage on a suborbital RLV. The costs per pound would be heinous, but because the total mission price is low enough, customers who wouldn’t normally be able to afford a space mission now can. On the high side, if an HLV costs so much in fixed costs that it takes up most of NASA’s budget, it doesn’t matter if it only cost $500/lb to orbit, it’d be too expensive.

    Not saying reproducing the Saturn V is the way to go, just wondering what might have been if instead of directing NASA to build the shuttles, they were directed to develop a lower cost, common element, partially reusable Saturn family of vehicles.

    Not sure. My guess is not that much better than where we are now.

    ~Jon

  11. templerman says:

    Why can’t NASA build and operate a relativly responsible and cost-effective space program? Simple. They have switched to operating NASA like other buracratic agancies. Spread program goals and costs out over yearly budget increses, and throughout every congressional district. Exactly in the same manner that the DoD does.
    When I was in university, I remember a paper done on the space shuttle and space station programs. This was in the late 80′-mid 90’s. One number that stood out was that the cost to build the shuttle could have bought 150 plus Saturn-V missions launched to the moon.
    If we can’t change the way the Nation does appropreations and program cost over-runs we might as well just stay here on the “good Earth” and let other nations rule space.

Leave a Reply

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