A couple weeks back, I brought up some fundamental flaws I saw in the logic that the Exploration Systems Architecture Study group used to come up with their preferred lunar transportation architecture. In a discussion on NASASpaceFlight.com, I started realizing that even using their own methodology, a dry-launch propellant-transfer architecture might actually have a lower “Loss of Mission” probability than the preferred ESAS Architecture, in spite of requiring many more launches. And that is even ignoring almost all of the flaws I pointed out previously.
In the ESAS paper, they detailed three main sources of potential Loss of Mission:
- Rendezvous and Docking Failures
- Launch Vehicle Failures
- Launch Vehicle Availability Issues
On the surface, it would seem that an EELV based mission, that might require 6-12 flights per mission would obviously lose in this analysis. After all, more flights means more chances to lose a payload, more payloads that might fail to dock correctly, and more payloads whose schedules slip, right? Not necessarily.
One of the key things to remember about the ESAS architecture is that most of the Initial Mass in LEO (IMLEO) is propellants. A frequent commenter on this site said something elsewhere to the effect that “Most of what Ares V does is launch LOX”. Without the propellants, you could pretty much launch all of the ESAS hardware on two EELVs, one standard size, and one heavy, especially if the Heavy were an Atlas-V using the Phase I Wide Body Centaur upper stage mentioned in my last post. All of the rest of the flights are delivering propellants.
Upon a little thought, you realize that the only failures that can cause a loss of mission are failures on hardware flights. A lost propellant flight doesn’t cost you the mission, since you can easily fly a replacement. Only a failure involving the docking or launch of one of the hardware flights actually eliminates something unique that you needed for the mission.
So, comparing the ESAS architecture to a dry-launch architecture, we find that:
- Both architectures only have one mission-critical docking event that of the CEV to the LSAM/EDS stack. All other launches in the dry-launch architecture are not unique, and will not result in a loss of mission if the rendezvous and docking fail. At most they will result in a delay. So, the LOM numbers for rendezvous and docking failures are identical between the two architectures.
- Both architectures only have two mission-critical launches, the hardware launch (LSAM and EDS), and the crew launch. If a crew is lost or if the cargo gets put into a “fishing orbit” as I like to call it, the mission is over. ESAS claims that their all-new launchers will be safer and more reliable than EELVs, but only slightly so. And theoretical hardware-reliability numbers rarely reflect the true reliability of the systems. Witness Shuttle and Falcon I. But you could call this as slightly in favor of the current ESAS architecture. Their odds of losing a mission-critical launch are probably about 1% lower than for Dry-Launch.
- Where the current ESAS architecture loses the worst is ironically where they claimed a multi-launch architecture suffers the most–launcher availability. Basically, the current ESAS architecture has no way of topping-off the EDS/LSAM propellants before departure, so any delays above a certain amount end up costing the mission. Without a way of transfering propellant on-orbit, all of that expensive hardware ends up becoming next to useless if the CLV is late, or if the CLV suffers a non-fatal launch failure. The dry-launch architecture however can make up for delays. Any boiloff issues don’t cost the mission, they just delay it slightly and add a tiny amount to its marginal cost. The odds of losing a mission due to CLV delays and boiloff with the current architecture are non-zero, but the odds of losing a drylaunch architecture to that are much, much lower.
To be fair, one can definitely quibble about whether the slightly higher theoretical launcher reliability is more or less of a factor than the inability to tolerate launch delays. However, what one cannot quibble about is that we’re paying a very, very high cost for what is at best a very small, and entirely theoretical advantage.
An interesting sidenote from this though is that even if you insist on building Ares I and Ares V, developing on-orbit cryogenic propellant transfer still makes sense, as it can decrease your odds of launcher availability problems causing a loss of mission.
One can also claim that on-orbit cryogenic propellant transfer is so far off in the future that comparing it with the ESAS architecture is fantasy. However, Lockheed has a very valid point when they state that every time they relight a Centaur in orbit, they’re demonstrating on-orbit cryogenic propellant transfer. They’ve only done it 200 times more than Ares I or Ares V has ever flown. While there is some development and qualification yet to be done on the concept, it’s probably closer on both a monetary, and technical standpoint to reality than the J-2X or the 5-Segment SRB.
But the main takeaway I have for this is that the arguments for Ares I/Ares V are a whole lot less solid than ESAS tries to prove.

Jonathan Goff

Latest posts by Jonathan Goff (see all)
- Random Thoughts: A Joint International Debris Remediation Effort - October 22, 2021
- Unorthodox Reusable Lunar Landers Concepts - June 12, 2021
- Goff Family 2021 Summer Sabbatical Part 1: Utah Trip - June 1, 2021
Mr. Goff,
ESAS is a fig leaf as Marsha Ivins said to us when we asked them Feynmann’s question.
ATK got what they bought.
Well said! There are a lot of assumptions in the ESAS report that don’t seem to hold up when closely looked at. Keep on championing dry-launch and orbital refueling.
Over at Curmudgeon’s Corner, Mark Whittington wrote:
“OK, so we build more infrastructure (that means pads, processing facilities, and so on) to enable us to launch the modified EELVs in a shorter period of time.
The problem is, that costs money, on top of the money it will take to modify and man rate the EELVs. I can see that cost rapidly eating up whatever savings one might get by canceling Ares 1 and Ares 5.”
Don’t these launchpads, processing facilities, and so on, already exist, in multiple locations across North America and around the world?
Jon,
One small caveat in your dry-launch mission safety numbers might be (and I have done no numbers to determine if this is real or not) that multiple docking events run the risk of damaging the mission critical hardware in a docking accident. The impact of the Progress module with Mir few years ago shows that even a very standard docking event can go wrong. This is not to say that it can’t be accounted for, but just to say that it’s probably something that would need more thought before being able to take full advantage of the dry launch scenario.
Paul
Paul,
That is a valid point. There will be some non-zero probability that a rendezvous and docking failure could risk the mission hardware. That probability would be much lower than just a mere failure (because damaging failures are only a subset of all failures), and I’m not sure what realistic numbers for that would be. Also, the more you do something (and the more frequently), typically the better you get at it.
That said, a dry-launch scenario would definitely require some careful attention paid to mitigating that concern.
The question then becomes is that slightly higher risk there, and the theoretically slightly higher risk on launching hardware stuff offset by the significantly lower risk due to delays and such?
I’m not sure, but my gut tells me it still makes more sense. Especially when we don’t really know what the Ares I/V reliability numbers are going to be in practice instead of theory. And also with the fact that we have to pay $30-50B to find out. We can afford to lose a *lot* of lunar mission hardware before an EELV system end up being more expensive than the planned architecture.
~Jon
I find it bizarre that NASA has baselined such a REALLY bad return to the moon architecture. I mean I’ve been around, and know the pork is more important then the mission, etc – and theirs lots of different ways to get the job done… but this retro Apollo done really badly SUCKS!! Other then some folks on the forum.nasaspaceflight.com/forums the one thing everyone seems to agree on is the Ares/CEV/etc has a major “you’ve got to be kidding?†factor.
Its not just the questionable cost, performance, safty numbers – or that it doesn’t set the base to expand our space – its virtualy everything about it!
Its amazingly bad.
>>Its amazingly bad.
Kelly,
Except for the people who work at NASA and will keep their jobs over the next 20+ years because of it.
I have no doubt that it can work, just that it’s not at all the most efficient(ore even a _more_ efficient) way of getting there. The active rebuttals from the NASA community regarding our (the open community) carping against the ESAS plan strike me as a serious case of “Methinks he doth protest too much”. However if it was my job on the line, I’d probably do the same thing.
Still, I think we’re only going to see efficient ways of getting to the moon when it is purely private enterprise paying the bills. Government has waaaaay too many other considerations on their minds rather than maximising benefit/cost.
Paul
Even assuming that there is a valid
point to the concerns regarding the
present workforce supporting the
Shuttle architecture – after all,
it’s hard to argue that it would
not be bad politics to put a large
number of skilled professionals out
of jobs – even assuming that we
have this “standing army” of rocket
plumbers to keep employed, can’t we
put them all to work on a less
complex and impractical vehicle
system?
>>can’t we put them all to work on a less complex and impractical vehicle system?
Dave,
The problem is not that the skilled rocket plumbers will be out of work. They won’t, at least not for long. The problem is that all the bureaucrats & hangers-on at NASA will be out of work and that is a huge proportion of the white collar workforce. The blue-collar workforce will be cut significantly as some of the main differences in the private vs. NASA arcitectures boils down to more efficient and much smaller standing armies of employees, many of whom are already working at the various booster manufacturing plants. There would be some increase in workforce to handle a higher production rate, but it is nowhere near directly proportional to the increase in launch-rate. So the blue collar guys and girls will inevitably get in the neck as will the white collar excess in the NASA system.
Lets face it, _that’s_ where the money savings comes from. The actual hardware in any high tech system is a small fraction of the manpower costs.