DIRECT and Other ESAS Alternatives

I’ve been following the discussions over on over the past few days regarding DIRECT and some other ESAS alternatives. One of these days I’d love to get access to their “Level 2” subscription service, but what with my $2k dental mistake, I doubt that’s going to happen anytime soon.

Anyhow, there were five alternative technical approaches discussed, with varying levels of detail given. The current ESAS plan is bloated, high-cost, low-return, and already running into technical issues. The Shaft is still having some pretty serious performance shortfalls, with the estimated payload down to 21 metric tonnes down from a previous number of around 27. The vehicles are getting less and less Shuttle Derived as time goes on. 4-segment SRBs are morphing into new 5-segment SRBs (but still miraculously keeping all the paper-reliability of the ones that have actually flown). SSMEs are pretty much out. Ares V no longer really has any part of it that is really Shuttle Derived in any sense of the term (except maybe the paint scheme for its SRBs and LOX/LH2 tank). Budgets are slipping, as are the delivery dates. More and more infrastructure modifications are appearing necessary.

In other words those of us “armchair engineers” and “internet rocketeer” have slowly been proven right, over and over again. ESAS is supposedly infallible, yet time and time again, the technical issues we picked with it are turning out to be right. But don’t you worry, those technical decisions are still being studied. The design isn’t final yet. We have Top Men working on it. Top Men…

Anyhow, everyone with any sort of engineering clue can smell the blood in the water, so all sorts of pet architectures are coming out of the woodwork now. I’ve been purposefully avoiding trying to waste a bunch of time, brain power, and innocent electrons putting together my own detailed plan, precisely because it just isn’t worth it. More to the point, any centrally planned lunar architecture is very likely going to be obsolete by the time it has a chance to get implemented. But that doesn’t mean that there aren’t useful things to learn from these other attempts. So here’s a few short comments on each of the five ideas presented on NASASpaceFlight.

The guys who put together the DIRECT proposal, and the site did a fairly thorough job of presenting their case. Here’s how I see the pro’s and con’s:


  • By using only one vehicle instead of two, the yearly fixed cost for the launch vehicles is about half of that for both Ares I and Ares V. That’s still $1B per year regardless of flight rate, but it’s better than $2.2B per year fixed costs that ESAS entails. Lower fixed costs mean that there is more money every year available for either more missions, or to fund parallel projects such as propellant depots, COTS, or man-rating EELVs. It also means there’s actual money available to buy on-orbit propellant if companies can figure out how to deliver it. The current ESAS plan really doesn’t have the money available unless it closes down either Ares I or Ares V. And if you really think that we’re going to abandon one of those vehicles after they actually exist, just because of commercial capabilities, you’re fooling yourself. Look how hard NASA has been fighting to keep the Shuttle workforce around. Do you think those political “realities” are going to change even when the vehicles are provably obsolete?
  • By only developing one vehicle, and by actually trying as hard as possible to reuse Shuttle hardware and infrastructure, the development costs for DIRECT would likely be as low or lower than even Ares I, but with the benefit of having all of your lunar launch vehicles developed and fielded by the early part of the next decade instead of the end.
  • This plan is the most feasible from a political standpoint, since it keeps at least some of the pork gravy flowing into the right congressional troughs.
  • DIRECT has the capability of bringing lunar missions at a much earlier date than 2018.
  • Having regen cooled versions of the RS-68 would be cool. Bumping Isp up to 435s from 420s, and bumping thrust up by 5% would have a noticeable effect on the performance of the Delta-IV. Might even be enough to allow the single-stick Delta-IVM to be able to fly manned missions with a sufficiently depressed trajectory to not have any blackout zones, as the Lockheed guys are claiming they can do for an Atlas V 401. I’ve always felt that regen engines, while supposedly a bit more complicated to make are really the way to go if you’re serious about rocketry.
  • With the lower fixed cost of DIRECT, you could actually fly 2-3 lunar sorties per year before you’ve even spent as much as the fixed cost of just keeping the Ares I and Ares V lines open. In other words, you could at least double the mission tempo without increasing NASA’s budget, or raiding other programs.
  • By not needing as much money as soon, due to for instance not needing to rush a 5-segment SRB into existance, that puts less of a squeeze on the aeronautics and science sides of NASA. Now that we’re looking at a Democratic congress, and possibly even a Democratic Senate (depending on how things settle out after a recount), that might be a more politically survivable approach.


  • It’s still a shuttle derived vehicle. While NASA’s fixed costs for running earth-to-orbit transportation will drop considerably compared to Shuttle or the predicted numbers for ESAS, it’s still $1B going every year to pay contractors to provide more redundant launch capacity.
  • It still another example of NASA having its own booster instead of buying earth-to-orbit transportation. Contra-Mark Whittington, Ares I, Ares V, and DIRECT do not send crew and cargo to the moon. They send a little bit of crew and cargo to LEO along with a lot of propellants, and a little bit of transfer stage. Launching crew, cargo, and propellants to LEO is not something that only NASA can do, and in fact is something that NASA has demonstrated that it can’t do in anything resembling a safe or cost-effective manner.
  • It still only really helps encourage and enable commercial space capabilities if the supposed cost savings over ESAS are actually used for…encouraging and enabling commercial space capabilities. There’s no guarantee that that money wouldn’t just be funneled down some other congressional rat-hole.
  • Like Ares V, it still doesn’t have many good upgrade paths that could actually allow it to take advantage of advancements in commercial space capabilities as they occur. Even if commercial propellant depots on-orbit become a reality, DIRECT is still stuck with maintaining what would then be a completely redundant launch infrastructure (with about 80-90% of your IMLEO being propellant that can now be launched on smaller, cheaper, commercial vehicles, why do you still need an HLV?).

While DIRECT may possibly be the most rational and economically affordable and sustainable SDLV-based architecture I’ve seen to date…that ain’t sayin’ much. It probably has the most chance of happening, and would be a large improvement over the current ESAS debacle, but is still far from ideal.

I don’t have quite as much time to go over the other concepts, but I want to give at least a few short thoughts. This Italian plan has some technical points that I strongly agree with, but also suffers from some serious issues:


  • I like reusable lunar landers.
  • I like on-orbit propellant transfer and reusable lunar tugs. If NASA licenses the on-orbit propellant transfer technologies, that allows for commercial earth-to-orbit transportation companies to get involved, which would likely drop the transportation costs by a huge margin. The amount of propellant needed for even a meager NASA lunar program will likely rival near-term orbital tourism launch demand. And having that technology on-the-shelf for the private sector makes it easier for them to use it for their own commercial ventures.
  • I like the idea of eventually having transportation and refueling nodes in either LUNO or L1/L2.
  • I like the idea of not having to develop an HLV


  • Not Invented Here Syndrome (heck this wasn’t even invented in the US).
  • Even though a new HLV isn’t developed, I can’t see NASA not horribly botching the design, development, and implementation of a fully-reusable tug, lander system, lunar space station, propellant depots, and everything else. The technological heart is in the right place, but run as a centrally-planned NASA run and operated project, it would likely end up as bad of a debacle as ISS and Shuttle have been.
  • While the recurring costs of lunar sorties would likely go down, the fixed cost of a NASA run and operated lunar transportation system like this would be astronomical.

Lockheed’s Plans
I think Lockheed’s plans are the closest to my opinions, and I’ve talked about a lot of the elements on my blog previously, but here’s a recap:


  • No need for launcher development. That alone saves $20-25B, which is likely enough to buy something like 360 Atlas V 401 launches, or a combination of Atlas V, Delta IV, Falcon IX, and Kistler K-1 flights. That also means that you can start development of your actual lunar hardware right now.
  • The several billion per year freed up over the next several years could likely mean both having multiple independent commercial manned ISS service providers sooner than Ares I could be fielded, but it also means that most of the NASA lunar hardware could be ready to fly within the 2010-2014 timeframe, possibly allowing much sooner lunar sorties.
  • NASA would have no money tied up every year in the fixed costs of operating its own earth-to-orbit transportation system. That comes to over $2.2B in savings per year that can go towards paying for more sorties, paying for a much more sophisticated exploration program, and possibly supporting multiple exploration bases.
  • Lockheed’s on-orbit propellant transfer concepts are based on existing flight hardware with decades of experience, and could probably be brought all the way to commercial viability very quickly, even if they have to develop it on their own dime.
  • The EDS could likely be derived fairly strongly from Lockheed’s venerable Centaur upper stages. This would likely drop development and operations costs.
  • The lunar architecture would be launcher independent. This provides a lot of robustness to schedule slips, technical issues, and contractors trying to milk money out of NASA when they have them over a barrel. It also makes it so that the architecture can immediately benefit from lower-cost or safer launch vehicles as they come into existance. The large increase in global flight demand in a market that would mostly be open to only US launchers would draw a lot of commercial investment money back into the US launch market. The flight rates involved are even getting into the range that true orbital RLVs might actually get sufficient funding to have a chance.
  • The current ESAS plan has now ability to re-top-off the tanks on the LSAM and the EDS once launched, which means that if the CLV is late, the mission will be scrubbed. The Lockheed plan however would be able to handle schedule slips like that rather trivially–just launch another tanker and fill ‘er up.
  • This plan helps move us toward becoming a truly spacefaring society. Not just a society with a government agency that every once in a while sends a few employees to visit space.
  • Actual NASA employees could spend more of their time developing hardware for use on the moon, like reusable landers, base equipment, ISRU hardware etc. The current ESAS plan spends over 2/3 of its money on just developing and operating the earth-to-orbit part of the transportation system. Once you’ve also paid for the development of the LEO to LUNO transportation systems, and the landers, almost none of the money NASA plans to spend over the next decade is actually slated towards developing tools and technologies for use on the moon. Almost none of it is slated towards developing the components that would be needed for even a spartan base. At the rate we’re looking at, such things couldn’t even begin to be developed until after Ares I, Ares V, EDS, and LSAM were all fielded and operations–which means that actual base development wouldn’t even start until post-2020.
  • Most of the “Shuttle contractors” that ESAS is trying to keep together are Boeing and Lockheed employees (as well as employees of ATK and a couple other large contractor companies), not NASA employees. If NASA were buying dozens of commercial launches per year from Boeing and Lockheed, and if they and the COTS teams were providing the manned launches as well, many of those “shuttle employees” would actually be retained. Not the ones working in dead-end technologies like the technicians refurbishing large, segmented SRMs, or those doing tile repair, or overhauling SSMEs mind you. But the ones that have skills that are actually relevant to a truly robust space transportation architecture would be retained. Losing a skilled workforce in an obsolete technology is not always a bad thing. How many people lamented the skilled wagon makers that were put out of a job by Ford?


  • Unlikely to actually happen. Makes too much sense, and doesn’t “keep the team together” as well as the Shuttle Derived options do. Doesn’t allow NASA to play around with big rockets either.
  • EELVs are still depressingly expensive and unreliable compared to where space transportation needs to get to. A lot of care would need to be taken to make sure that we wouldn’t be replacing one launcher-specific architecture for another one.
  • A commercial LEO launcher based architecture would benefit from more flexibility on NASA’s part regarding minimum mission crew size. If NASA allowed for smaller crews, they could get far more missions, greater flexibility, and greater cost efficiencies of scale. Alas, NASA seems to feel that anything less than 4 people in a mission would somehow represent a step backwards. Even if it allowed them to fly more people to the moon per year, made the architecture more flexible, allowed for more diverse exploration, and generally made more sense.

There have to be some other Con’s, but I can’t think of many at this point.

CBO Alternatives
While I commend the CBO on noticing that the ESAS architecture isn’t the most cost effective way of doing things, they still swallow whole some of the same false assumptions that ESAS did. While I’m sure an Atlas Derived launch architecture that used brand new UberAtlases could still be cheaper than Ares I and Ares V, I think that you lose almost all the benefit of going with a commercial vehicle once you start insisting on basically completely redesigning it again.


  • Marginally cheaper than Ares I and Ares V.


  • You still end up with NASA paying large amounts of development costs for a new launcher, then having to pay all the fixed costs for the launcher as well as marginal costs. It has all the same problems as the current plan, just with slightly different packaging.

I’m not even sure where to start here. While it is an interesting technical approach, it seems utterly out of contact with reality. While ambitious, it also is I don’t think very realistic. It has all the drawbacks of the MoonLight plan, while also having all the drawbacks of SDLV plans. High development costs, high fixed costs, unrealistic assumptions about how easy ISRU is going to be or how effective robotic precursors can be. I just really don’t see anything in thise scheme even worth giving it more thought.

Anyhow, it’s getting on towards 1am, and I have to be on the road to work in about 7 hours, so I think I’m going to call it a night. There’s my take on things. In summary, I think that of these alternatives, that DIRECT is the most likely to get NASA to go along with it, but Lockheed’s approach is the most in the right direction technically and economically. The other three while having some interesting ideas are all unrealistic or misguided in my opinion.

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 the founder and CEO of Altius Space Machines, a space robotics startup that he sold to Voyager Space in 2019. Jonathan is currently the Product Strategy Lead for the space station startup Gravitics. 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.
Jonathan Goff

About Jonathan Goff

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 the founder and CEO of Altius Space Machines, a space robotics startup that he sold to Voyager Space in 2019. Jonathan is currently the Product Strategy Lead for the space station startup Gravitics. 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.

16 Responses to DIRECT and Other ESAS Alternatives

  1. mz says:

    Excellent post! You’re thinking clearly and with a wide scope.

    One thing detracts from a lot from the whole value though:
    “Doesn’t allow NASA to play around with big rockets and kill more astronauts.”
    That’s cheap. 🙁
    Certainly claiming NASA wants to kill more astronauts is not raising the expectations of your intellectual abilities regarding the other points in the post.

  2. Jon Goff says:

    Yeah, that was a bit over the top. I just toned it down a bit. It wasn’t made as a factual observation, but a piece of unnecessary snark. It’s gone now.

  3. Anonymous says:


    If Lockheed can use on-orbit refueling to do moon missions, why can’t Boeing and the Delta IV “uber-heavy” and SpaceX and RP-Kistler do the same thing??

  4. Lee Valentine says:

    Jon, in the 80’s, a number of AI labs attempted to build special purpose hardware to achieve computer vision. That centrally planned ad hoc approach could not keep up with the progress of general purpose commercial computers. It appears that this is a relevant analogy with ESAS. True market and competition driven commercial space seems about to begin. If it is successful, then ANY ad hoc NASA VSE plans may be overtaken by commercial efforts, just as those government commissioned special purpose computers were outclassed in the free market.

    In the seventh bullet point under Lockheed’s plans, you say “now” when you mean “no”. Cognitive dissonance makes my head ring.

    Great post, thanks.

  5. Jon Goff says:

    Good catch on that “now” vs “no”. As for the centrally planned ad hoc systems vs continually evolving general purpose systems, I think the analogy sounds pretty solid.


  6. Jon Goff says:

    If Lockheed demonstrates on-orbit cryogenic transfer succesfully, everyone else will know it’s possible, and I wouldn’t be surprised if there were several competitors in that market niche within a short period of time. Lockheed even seems to prefer that from what I got out of their paper on the topic. Even if they can’t compete as well on propellant launches, I think they expect they can make a lot more on launching other hardware more often if the propellant on-orbit becomes cheap enough that NASA (and others) can fly more lunar sorties.


  7. Josh Gigantino says:


    One of the designers of the Lockheed Lunar architecture was posting on for a while. He indicated exactly what you suggest, but more so. Lockmart, in this context, would ideally be buying propellant from others to fill their depots. If they have to do some of the tanker flights themselves, they will, but would rather be building infrastructure and extending boundaries. They really seem to want to play ball with companies and the emerging market. There are certain things that only megacorps and they realize that and are looking to play from their strengths in the new market, unlike certain other defense-oriented companies.

    You’re best quote, and it relates to the “machine vision computer” above:

    “More to the point, any centrally planned lunar architecture is very likely going to be obsolete by the time it has a chance to get implemented.”

    The market is an ecosystem, it is niches and routing around inefficiency.


  8. tankmodeler says:

    What’s really scary is that NASA seems to thnk (at least publicly) that they are doing things in the most sensible and least expensive manner. I’ve just seen the lastest Ares Status Report and it goes on and on about how the Ares program is a cost effective way to get to the moon and how the program is making the best use of time and money. I’m not entirely sure what spatial continuum you have to be in to warp reality enough to actually believe that, but it worries me that NASA can beleive that it is conforming to the intent of the October CBO report while actually doing nothing to stop the systems bloat that is the hallmark of the Stick and Stack programs.

    I agree that the DIRECT system will probably be the best compromise between the disparate political and engineering points of view and would be the more likely candidate if Ares is “re-scoped” by Congress. Its one real selling point is that, for a change, it truthfully _is_ Shuttle-derived equipment. The current Ares is Shuttle-derived as much as I’m lemur-derived. It’s true, but has no effective bearing on the system under discussion.


  9. Stefanos says:

    Very nice thought-out, but I have to disagree in your comments about Teamvision. I think they grasped exactly what the VSE should have been.
    It is long range, complex and expensive, but it strongly reminds me of the motos’ of open source software: Release early and release often. And the proposal could always benefit from newly available technology (COTS). And provides a viable way for long-term space habitats and their purpose (refuelling). I was thinking along the same lines, and came up with very similar conclusions, because moon is, exactly, the first step.

    Excellent vision.

  10. Ed says:

    Lee Valentine, I’ve seen you make a bunch of really intelligent comments on blogs now (on this one above, on your refutation of Bell etc)… dude, it’s time for you to start a blog.

  11. Jon Goff says:

    He actually does blog over at


  12. libs0n says:

    I prefer the Shuttle B concept over Direct, and it’s definately sexier mated with that X-38 descendant. Quoted below is the post from which I first learned of it, in the same thread that introduced me to the Direct launcher.

    gunsandrockets wrote:

    This link describes the basics of an updated Shuttle C design. (Greg describes a multi-purpose manned/unmanned vehicle, I differ in that I think it should only be used as a cargo launcher)

    This configuration would provide the fastest and probably cheapest development and lowest infrastructure cost to achieving a Heavy Lift Cargo Launcher. The only difference between the existing Shuttle system and this Shuttle derived vehicle is the substitution of the Orbiter. The SSME and OMS would be gone. All the other components of the Shuttle would remain, the ET, SRB etc.

    Unlike the Direct launcher, no new engine development projects would be neccessary. This updated Shuttle C design would use existing Shuttle SRB for boosters, 2 x RS-68 for the main engines and 2 x RL-10b for the final stage. Unlike the Direct launcher, no extensive redesign and altered manufacture of the ET would be neccessary. Unlike the Direct launcher, this vehicle is a pure cargo launcher and can avoid development requirements for manned flight.

    There are two crucial differences in this design compared to the original Shuttle C. The original Shuttle C was semi-reusable. The 3 engine SSME + 2 engine OMS cluster was podded and supposed to be recovered from orbit for reuse. That meant the Shuttle C would drag 60+ tonnes of deadweight mass up to orbital velocity (ET mass + engine pod mass) reducing potential payload. But the updated design can stage before final orbital velocity since the liquid propellant engines are now completely expendable. Orbital velocity is now acheived with an upper stage using twin RL-10b engines. In fact the existing 30 tonne Delta IV upper stage can be used for this purpose. The vacuum ISP of the RS-68 engine is 8% lower than SSME, but the other advantages of this updated design, such as the 465s ISP of the RL-10b, should more than make up for that.

  13. Monte Davis says:

    Fine survey.

    In the interest (again) of minimizing snark, I’d suggest removing the scare quotes from “realities” in the first bullet point under Direct->Pro. That they’re realities you regret doesn’t make them less real.

  14. SMetch says:

    Yes, by all means get some sleep. The TeamVision plan from Era 1 to 3 is the “Direct” plan with the addition of ELV’s being used for testing and potentially fielding the CEV to ISS as well. As Ross has indicated while the ideas he described in greater detail are not new, ourselves included, they are in fact significantly more effective than the current plan. I wish Ross all the luck in the world in getting NASA to move towards this direction.

    Concern Era 3-6 of our AIAA paper.
    +20 years from now.

    Found here:

    We attempted, unlike all other VSE proposals including ESAS I might add, to incorporate both Mars and Beyond into the proposal. Failing that all we will really have is Apollo on steroids or quaaludes take your pick. If what we came up with seems a little daunting that’s because once you get beyond the art work and into engineering a manned mission to Mars you will realize just how daunting it really is. The aero braking assumptions NASA is currently using, particular the Mars Descent portion, are completely wrong. We still have no-go areas on Mars for the comparatively feather weight landers due to lack of atmosphere to slow them down without significant amounts of descent propellant. Translation, manned Mars surface mission are going to need a lot more TMI mass then currently planned. Translation its time to stop playing around and get serious about both HLV and ISRU. The Jupiter III is 80% finished by weight as I write add 20% and a little ambition/courage and we are there. The key thing now is to preserve the option of doing this 20 years from now and that is what Direct and Era 1-3 of our proposals is all about.

  15. Anonymous says:

    Thanks Jon.

  16. HaloJonesFan says:

    tankmodeler: Well, if you start out with the assumptions that NASA will design and build every part of the program and that you must at least pay lip-service to heritage design, then Area is the most cost-effective way to do it…in the sense that it’s the hoary old management cliche of a “bid” requirement written so specifically that the only way to fill it is to do exactly what you wanted.

    I think that Jon has a very good point about the launcher, though. Part of the reason that nobody has done much manned spaceflight is that, for a long time, everyone believed that every part in a MSF mission needed to have a NASA part number. If the government is making explicit statements that it’s willing to allow spaceflight by private entities…well, then there’s no reason not to develop for it!

Leave a Reply

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