I’ve had several people in several places ask me if I was going to do a point-by-point rebuttal of Mike Griffin’s comments to the STA this week (for reference the text of his comments is available here). While I don’t have the time to go into every single disagreement I have with what he said, I think there are a couple of key points I would like to point out. In other words, I’ve come to discuss Griffin, not to Fisk him.
Missing the Vision
Dr Griffin starts his defense of the chosen Constellation architecture by framing it “in the
context of policy and law that dictate NASA’s missions.” As he said on page 2:
Any system architecture must be evaluated first against the tasks which it is
supposed to accomplish. Only afterwards can we consider whether it accomplishes
them efficiently, or presents other advantages which distinguish it from competing
He then went on to discuss President Bush’s original announcement of the Vision for Space Exploration, and the NASA Authorization Act of 2005. I agree that it is important to make sure you know up-front what yardstick your program is going to be measured by. However, I think one thing becomes quickly obvious as you read Dr Griffin’s quotes from those documents–he entirely focuses on the technical implementation details, and never once mentions the actual policy goals!
Goal and Objectives
The fundamental goal of this vision is to advance U.S. scientific, security, and economic interests through a robust space exploration program.
These goals are the yardstick by which any VSE implementation needs to be judged. The rest of the technical details of how the space exploration program is carried out needs to be viewed in the light of these three areas of US interests. It doesn’t matter if a proposed implementation hits all of the other technical details, if it doesn’t really further US scientific, security, and economic interests, it isn’t really compliant with the goals of the president’s Vision.
Going into a little more detail on these goals, the Renewed Spirit of Discovery document continues (emphasis mine):
In support of this goal, the United States will:
• Implement a sustained and affordable human and robotic program to explore the solar system and beyond;
• Extend human presence across the solar system, starting with a human return to the Moon by the year 2020, in preparation for human exploration of Mars and other destinations;
• Develop the innovative technologies, knowledge, and infrastructures both to explore and to support decisions about the destinations for human exploration; and
• Promote international and commercial participation in exploration to further U.S. scientific, security, and economic interests.
Once again, all of the specific technical details like the CEV, retiring Shuttle in 2010, etc. are all pursuant to these goals.
Lastly, the NASA Authorization Act of 2005 (available here) states, once again with my emphasis:
The Administrator shall establish a program to develop a sustained human presence on the Moon, including a robust precursor program, to promote exploration, science, commerce, and United States preeminence in space, and as a stepping-stone to future exploration of Mars and other destinations.
Once again, you will notice that the key goals of this Vision, elucidated by both the President and Congress include not only science, but commerce, and in the president’s case security.
I could go on about how Dr Griffin’s focus on the parts of the Authorization Act that talk about heavy lift and shuttle derived ignored other sections in the act that talk about “encouraging the commercial use and development of space to the greatest extent practicable” (see Section 101.a.2. parts B-C). But I think the fundamental issue is that by focusing exclusively on just the technical side of the requirements, and not on the underlying goals, Griffin is missing the Vision.
On page 7, Dr. Griffin starts making his case for the Constellation architecture with this somewhat ironic statement about the Space Shuttle:
Once before, an earlier generation of U.S. policymakers approved a spaceflight architecture intended to optimize access to LEO. It was expected – or maybe “hoped” is the better word – that, with this capability in hand, the tools to resume deep space exploration would follow. It didn’t happen, and with the funding which has been allocated to the U.S. civil space program since the late 1960s, it cannot happen. Even though from an engineering perspective it would be highly desirable to have transportation systems separately optimized for LEO and deep space, NASA’s budget will not support it. We get one system; it must be capable of serving in multiple roles, and it must be designed for the more difficult of those roles from the outset.
And then Dr Griffin goes on to try and justify an architecture based on building a duplicative LEO capable only launch vehicle first, and hoping that when that vehicle is finally done, that there will be funding for developing “the tools to resume deep space exploration”…
After that auspicious start, Dr. Griffin then reminds us that “the new system will and should be in use for many decades.” Of course some of the historical analogies he draws could lead one to different solutions than it led him. For instance, he mentions that “In space, derivatives of Atlas and Delta and Soyuz are flying a half-century and more after their initial development.” An interesting thing to note about Atlas and Delta is that the only reason why vehicles with the name Atlas and Delta are “still flying” a half-century after their initial development, is precisely because they are only derivatives of the original. In fact, the current EELVs have very little in common with the vehicles that originally bore their names.
On pages 8 and 9, Dr. Griffin concludes that (emphasis mine):
The implications of this are profound. We are designing today the systems that our grandchildren will use as building blocks, not just for lunar return, but for missions to Mars, to the near-Earth asteroids, to service great observatories at Sun-Earth L1, and for other purposes we have not yet even considered. We need a system with inherent capability for growth.
While I disagree with the direction Dr. Griffin is going, I do agree with his point in that last sentence. We do need a transportation architecture that has inherent capability for growth. I just don’t think that the Constellation architecture really fits that bill.
The Promise of Commercial Space
Now, lest you think I’m going to spend yet another post hammering on Dr. Griffin, I’d like to quote a part of his speech that I really agreed with:
Further application of common sense also requires us to acknowledge that now is the time, this is the juncture, and we are the people to make provisions for the contributions of the commercial space sector to our nation’s overall space enterprise. The development and exploitation of space has, so far, been accomplished in a fashion that can be described as “all government, all the time”. That’s not the way the American frontier was developed, it’s not the way this nation developed aviation, it’s not the way the rest of our economy works, and it ought not to be good enough for space, either. So, proactively and as a matter of deliberate policy, we need to make provisions for the first step on the stairway to space to be occupied by commercial entrepreneurs – whether they reside in big companies or small ones.
I have to say that for all my disagreements with Griffin, he at least talks a good talk when it comes to commercial space. I full-heartedly agree with his point in this paragraph. When you think about it, even assuming everything works out according to his plan, Constellation is never going to be capable of supporting more than a dozen people off-planet at any time. While that may be a lot more than we have now, Ed Wright has a point when he says that that is a round-off error, not an exploration program. Basically, the only way we’re going to see large numbers of people off planet, and the only way we’re going to see the large-scale manned exploration and settlement of our solar system in our life times, is if the private sector can eventually play a much more expansive role in space transportation. As it is right now, so long as the commercial industry continues to play second fiddle to parochial interests and NASA-centricism, we’re not really going to go much of anywhere.
So, the fact that NASA is at least doing something to help promote that day is a sign that they at least partially get it. A successful and thriving entrepreneurial space transportation industry is going to help them actually achieve their goal of extending human life throughout the solar system in a robust program of space exploration.
Griffin continues with more good comments in his next paragraph:
If designed for the Moon, the use of the CEV in LEO will inevitably be more expensive than a system designed for the much easier requirement of LEO access and no more. This lesser requirement is one that, in my judgment, can be met today by a bold commercial developer, operating without the close oversight of the U.S. government, with the goal of offering transportation for cargo and crew to LEO on a fee-for-service basis.
But here is where the conversation takes a dangerous turn:
Now again, common sense dictates that we cannot hold the ISS hostage to fortune; we cannot gamble the fate of a multi-tens-of-billions-of-dollar facility on the success of a commercial operation, so the CEV must be able to operate efficiently in LEO if necessary. But we can create a clear financial incentive for commercial success, based on the financial disincentive of using government transportation to LEO at what will be an inherently higher price.
To this end, as I have noted many times, we must be willing to defer the use of government systems in favor of commercial services, as and when they reach maturity. When commercial capability comes on line, we will reduce the level of our own LEO operations with Ares/Orion to that which is minimally necessary to preserve capability, and to qualify the system for lunar flight.
While I agree that the government not only is the government being “willing to defer in favor of commercial services” is a really good idea, I think that this approach (of hedging their bets by coming up with a competing in-house launcher) is fraught with risk. Also, while on first blush, it may appear to be common sense to not “hold the ISS hostage to fortune”, it is my contention that this line of reasoning not only doesn’t hold as much water as it seems.
First off, as has been pointed out on numerous occasions, including in Griffin’s statements above, a commercial solution to ISS crew/cargo is going to be a lot more affordable than the in-house Ares-1/Orion solution. It has been mentioned before by people high up at NASA, that they really need COTS to succeed, because if they have to fly all the ISS missions themselves (especially if ISS doesn’t get retired in 2016, which Dr. Griffin mentioned in this speech as a possibility), there really won’t be anywhere near enough money to develop the lunar portions of the proposed Constellation architecture in time for the 2020 lunar return goal. You could say in a way that the existing Constellation architecture holds the rest of the Vision hostage to the fortune of COTS. If COTS doesn’t succeed, there’s no way NASA is going to be able to afford executing on the rest of the vision. If the supposed “backup plan” for ISS resupply won’t produce acceptable results anyway if COTS doesn’t turn out, NASA shouldn’t be trying to make it a backup plan at all–they should invest more heavily in making sure that there are multiple COTS competitors and that they have enough resources to succeed. One of the single biggest execution risks for any COTS company is financing risks. And having a NASA “backup plan” that could potentially compete with them is one of the single biggest obstacles to be overcome in raising money for a COTS team.
Which brings me to my other concern. The danger of having NASA in-house launch vehicles and space access capabilities that can serve as a backup to COTS also allows them to directly compete with COTS if the budgetary situation goes sour. Think about it. If Ares-1 finally gets built and working, but Ares-V doesn’t get funded, there’s nothing for Ares-1 to do but service ISS. With how hard the esteemed congressmen from Florida, Utah, and Alabama are fighting to maintain the Shuttle workforce and infrastructure (even to the point of suggesting continuing to fly the Shuttle!), does anyone really think that they would just “stand down” at that point, even if there was a clearly superior commercial alternative? Not very likely. I’m sure they would come up with some technical reason why Ares-I was superior (after all, our probabilistic risk assessment says that Ares-I has a 1:2106.5923 chance of killing a crew, while our numbers show that they have a 1:500 chance–who do you want flying our brave astronauts?) and find a way to not actually stand down. The frustrating thing is that by setting things up the way NASA is doing, the NASA people don’t even have to be malicious for such a result to happen–it’s a natural and likely consequence of the perverse incentives that NASA and Congress are setting up.
So, while I personally think that Dr. Griffin really and emphatically believes in and supports commercial space development, I’m afraid that there’s a high chance that some of his well-intended choices could end up coming back to haunt us.
Moon, MARS!!!! and Beyond
The last item I’d like to point out in Dr. Griffin’s speech is one of the justifications he used for the “1.5 launch” architecture they selected. Dr. Griffin made the point that while he feels that Constellation needs to be backward compatible with ISS as a backup plan, it also needs to be forward compatible with Mars, because sometime in the 2030s, we’re going to be going there. Now, I’m of the opinion that trying to guess what the best technical approach will be for a problem 30 years from now is somewhat of a fools errand. But that’s just me I guess.
So, starting on page 16 he begins to layout his case:
On the other end of the scale, we must judge any proposed architecture against the requirements for Mars. We aren’t going there now, but one day we will, and it will be within the expected operating lifetime of the system we are designing today. We know already that, when we go, we are going to need a Mars ship with a LEO mass equivalent of about a million pounds, give or take a bit. I’m trying for one-significant-digit accuracy here, but think “Space Station”, in terms of mass.
Now, I’m not going to go into the fact that there are probably plenty of other approaches to Mars exploration that can change the equation entirely. That’s a post for another day. For now, let’s just run with that premise.
He then repeats the “everyone knows that ISS taught us that using 20 ton vehicles to build something big is a bad idea” catechism, but that’s not what I’d like to discuss. The real gem is in this paragraph on page 17 (emphasis mine):
But if we split the EOR lunar architecture into two equal but smaller vehicles, we will need ten or more launches to obtain the same Mars-bound payload in LEO, and that is without assuming any loss of packaging efficiency for the launch of smaller payloads. When we consider that maybe half the Mars mission mass in LEO is liquid hydrogen, and if we understand that the control of hydrogen boiloff in space is one of the key limiting technologies for deep space exploration, the need to conduct fewer rather than more launches to LEO for early Mars missions becomes glaringly apparent.
It is true that one can draw that inference–that hydrogen boiloff means you should build as big of an HLV as possible. However, the conclusion I would draw is that if cryogenic propellant storage technologies are “key limiting technologies for deep space explortion”, then the right answer is to stop trying to kludge around the problem–develop them! Don’t use the existing state of the art in propellant handling and problems that are still 20 years down the road drive multi-billion dollar development projects today.
There are current technologies under development that could yield very low to zero boiloff of cryogenic propellants. There are multiple groups (ULA, Boeing, groups working with Glenn Research Center, etc.) pursuing multiple approaches to solving these problems. There are passive cooling and active cooling techniques. This isn’t some high-risk technology like nuclear fusion. The technologies needed for cryogenic fluid management in space are mostly low-risk extensions of 40 years worth of research and development. More to the point, many if not all of these technologies need to be developed to make Constellation work for lunar trips anyway, and would still be needed for Mars trips.
Is 2030 really so close that we can’t afford to do this right and actually develop the technologies we need instead of trying to kludge by with existing technologies?
Once you have the boiloff issue reduced or solved, that ~500klb of hydrogen ceases to be a headache, and begins to be an opportunity. That’s a lot of demand for propellant in orbit, and it can be supplied commercially. You’re already going to need propellant transfer technologies anyway if you have to launch the hydrogen in multiple launches, so what’s to stop launching it in even smaller launches?
I guess my point is that if one of the key arguments for the 1.5 launch architecture over a more commercial one, or a less expensive shuttle derived one like DIRECT is hydrogen boiloff, I think their kludge around the issue isn’t the right approach, and that they’d be better off just doing it the right way. Also, part of the reason why we have a federally funded aerospace program is to help prove out the technologies necessary for enabling the commercial exploitation of space, and actually solving problems like these would be much a much more responsible use of public funds than developing a kludge around point design like Ares V that doesn’t advance the state of the art for the commercial benefit of the country.
I guess overall while there were some good points, there was also a lot of issues with Dr. Griffin’s latest defense of Constellation. As discussed, I think that an a myopic focus on the technical details while ignoring the overall goals of the VSE has led to an architecture that isn’t responsive to the key policy goals laid out by the president and reiterated by Congress (particularly with respect to promoting the commercial and security interests of the United States). I think that in spite of Griffin recognizing the need for growth and flexibility in any architecture, that he chose a rather brittle and inflexible one. I also think that while he showed that he does recognize the potential of commercial space, and the importance of NASA trying to promote it, I think that the way he’s running COTS and Constellation will likely end up being highly counterproductive. Lastly, I think that in many cases, when confronted with a solvable engineering problem, Constellation has instead decided to kludge around the problem instead of properly solving it.
There are plenty of other issues I could’ve raised, but I figured these were some of the more obvious ones that I felt needed discussion.
Latest posts by Jonathan Goff (see all)
- Random Thoughts/Rocket Legos: Masten Xephr as a Vulcan SRB Replacement? - May 5, 2017
- Random Thoughts: RLVs and Megaconstellations - April 14, 2017
- Random Thoughts: First Pass Analysis of a White Dragon/Xeus Lunar Sortie Mission - February 28, 2017