ESAS Issues Part One

The more I look at Section 6.4.4 of ESAS (Chapter 6 can be found here on the NASA website), the more I realize that the entire architecture pretty much hinges on the assumptions in this section. The reasoning in this section is pretty much *the* case for why we have to use a Heavy Lift Vehicle, and why using existing or near-term available launch capabilities isn't adequate for the lunar architecture. The problem is that in addition to my previous gripes, there are a lot of hidden, incorrect, assumptions in this part of the study that I think taint the results of the rest of the study.

Return of the Highlander Fallacy
The single biggest hidden assumption is a form of the "Highlander Fallacy". The assumption is that in their analysis, there is only one companies launcher used, launching out of one facility, that has a single pad. You can see very clearly from figure 6.19 (which I yoinked from their report) that they are envisioning a very sequential process:

Launch number two can't happen until after Launch number one has had it's pad prep, gone through all it's scrubs, and is succesfully off. This simplistic analysis breaks down really fast if you assume two or three launch providers, or two or three pads, or two or three sites (with the best being a combination of the above). Take the IMLEO requirements from ESAS, which come out to about 340klb (290klb from Ares V and 50klb from Ares I). If you tried to launch them all on one EELV, it really would be very difficult. If you could manage to break all the payloads up into stuff that could be launched on single-stick launchers (Delta IV Medium or Atlas V 401), it would take about 17 flights, which is way too many compared to the manufacturing and pad availability for either booster. But if you split it between the two, it comes back down to about 8-9 flights each. If you use a few Delta IV Heavy launches for the big pieces (the EDS stage, the Lander, and the CEV), it gets down into an even more reasonable range of launches.

There are a couple of issues that come up:

• It is a really bad thing that pretty much all of our nation's launch capacity into non-polar orbits is located within a few miles of each other in a state like Florida where weather related scrubs are more common than almost anywhere else in the country. If you have a weather related scrub for one vehicle, you'll probably have to scrub the other too.


• Any non-polar commercial satellite that wants to be launched by either of these two boosters will have to fight for a position in the queue.


• Each factory was designed with the capacity to produce something like 20+ EELV cores per year. Whether that can be increased by adding extra shifts (or whether they were already planning on double shifting) remains to be seen.


However, there are a couple of additional insights that are interesting:

• The entire cost for fielding both EELVs, including building the factories, and one or two pads each, as well as doing a cleansheet design, was less than $2.5B total. Adding an extra pad, if the demand was there, or adding extra factory capacity, could likely be handled at some small fraction of that cost (maybe something like $1B).


• If SpaceX works out, their Kwajelein and planned Cape Canaveral sites would make the whole situation even easier. With three providers, two sites, and at least one pad per provider per site, the odds of "Loss of Mission" go way, way down.


• At the kind of flights rates possibly envisioned here, the DoD could drop it's subsidy of the two boosters, and the prices for all of them would go way, way down.

Anyhow, I have to run to work now, but when I get a chance, I'll talk some more about some of the problems with this troublesome section.