Here's the part that will make many a coffee be spat over screens in Denver etc.
Once the rationale for this particular dual-launch EOR scenario is understood, the next question is, logically, “why don’t we use the existing EELV fleet for the smaller launch?” I’m sure you will understand when I tell you that I get this question all the time. And frankly, it’s a logical question. I started with that premise myself, some years back. To cut to the chase, it will work – as long as you are willing to define “Orion” as that vehicle which can fit on top of an EELV. Unfortunately, we can’t do that.
The adoption of the shuttle-derived approach of Ares I, with a new lox/hydrogen upper stage on a reusable solid rocket booster (RSRB) first stage, has been one of our more controversial decisions. The Ares V heavy-lift design, with its external-tank-derived core stage augmented by two RSRBs and a new Earth departure stage (EDS), has been less controversial, but still not without its detractors. So let me go into a bit of detail concerning our rationale for the Shuttle-derived approach.
The principal factors we considered were the desired lift capacity, the comparative reliability, and the development and life-cycle costs of competing approaches. Performance, risk, and cost – I’m sure you are shocked.
The Ares I lift requirement is 20.3 mT for the ISS mission and 23.3 mT for the lunar mission. EELV lift capacity for both the Delta IV and Atlas V are insufficient, so a new RL-10 powered upper stage would be required, similar to the J-2X based upper stage for Ares I. We considered using additional strap-on solid rocket boosters to increase EELV performance, but such clustering lowers overall reliability.
It is also important to consider the growth path to heavy lift capability which results from the choice of a particular launch vehicle family. Again, we are designing an architecture, not a point solution for access to LEO. To grow significantly beyond today’s EELV family for lunar missions requires essentially a “clean sheet of paper” design, whereas the Ares V design makes extensive use of existing elements, or straightforward modifications of existing elements, which are also common to Ares I.
Next up for consideration are mission reliability and crew risk. EELVs were not originally designed to carry astronauts, and various human-rating improvements are required to do so. Significant upgrades to the Atlas V core stage are necessary, and abort from the Delta IV exceeds allowable g-loads. In the end, the probabilistic risk assessment (PRA) derived during ESAS indicated that the Shuttle-derived Ares I was almost twice as safe as that of a human-rated EELV.
Finally, we considered both development and full life cycle costs. I cannot go into the details of this analysis in a speech, and in any case much of it involves proprietary data. We have shared the complete analysis with the DoD, various White House staff offices, CBO, GAO, and our Congressional oversight committees. Our analysis showed that for the combined crew and heavy-lift launch vehicles, the development cost of an EELV-derived architecture is almost 25% higher than that of the Shuttle-derived approach. The recurring cost of the heavy-lift Ares V is substantially less than competing approaches, and the recurring cost of an EELV upgraded to meet CEV requirements is, at best, comparable to that for Ares I. All independent cost analyses have been in agreement with these conclusions.
So, while we might wish that “off the shelf” EELVs could be easily and cheaply modified to meet NASA’s human spaceflight requirements, the data says otherwise. Careful analysis showed EELV-derived solutions meeting our performance requirements to be less safe, less reliable, and more costly than the Shuttle-derived Ares I and Ares V.
Now is a good time to recall that all of the trades discussed above assumed the use of a production version of the Space Shuttle Main Engine (SSME). But, returning to a point I made earlier, we continued our system analysis following the architecture definition of ESAS, looking for refinements to enhance performance and reduce risk and cost. We decided for Ares I to make an early transition to the 5-segment RSRB, and to eliminate the SSME in favor of the J-2X on the upper stage. Similarly, elimination of the SSME in favor of an upgraded version of the USAF-developed RS-68 engine for the Ares V core stage, with the EDS powered by the J-2X, offered numerous benefits. These changes yielded several billion dollars in life-cycle cost savings over our earlier estimates, and foster the use of a common RS-68 core engine line for DoD, civil, and commercial users.
Praise is tough to come by in Washington, so I was particularly pleased with the comment about our decision on the 5-segment RSRB and J-2X engine in the recent GAO review: “NASA has taken steps toward making sound investment decisions for Ares I.” Just for balance, of course, the GAO also provided some other comments. So, for the record, let me acknowledge on behalf of the entire Constellation team that, yes, we do realize that there remain “challenging knowledge gaps”, as the GAO so quaintly phrased it, between system concepts today and hardware on the pad tomorrow. Really. We do.