I feel sort of obligated to comment on this, since there are few harsher critics of SLS than Your Truly (I even raised money to fight it).

The main benefit of using an HLV is that fewer individual launches are needed to get the same amount of mass into space – with fewer launches of larger payloads, a lower launch cost per unit mass is realized.

A lot of people say this, including Mike Griffin, but they never actually show their work. I’m sure I’ve written this before, but the underlying assumption here is that large vehicles have economies of scale. Well, that might be generically true, all things being equal, but all things are not equal, and in terms of economies of scale, flight rate is a much more important factor than vehicle size. In the real world, we know that each SLS flight will cost billions initially, and even with the generous assumption of two flights per year, it will still be well over a billion dollars a flight.

A few larger pieces are more easily assembled in space than are a multitude of smaller ones. The cost buy down is mitigated somewhat by the assumption of more risk, as the loss of a single HLV will more greatly impact the mission campaign than the loss of a single smaller vehicle. But the benefits of fewer launches overall and less complex on-orbit operations are usually judged to outweigh these drawbacks.

Judged by whom? The only actual analysis I’ve ever seen, performed by S&MA at JSC five years ago, indicated just the opposite, partly because (again) the higher flight rate offers reliability improvement that SLS will never see in this century. And if we fear the “complexity” of orbital operations, we might as well give up on being a space-faring nation.

He goes on to bash Falcon Heavy:

No existing commercial launch vehicle (nor any anticipated in the near future) has the launch capacity of the SLS.

Note that the need for the launch capacity of SLS has never been described or generated; there is no Design Reference Mission for it, other than (as stated in the 2010 NASA authorization) to deliver money to Huntsville, Promontory, and other places.

The largest extant commercial LV is the Delta-IV Heavy, which can put a bit less than 30 metric tons into LEO, less than half the capacity of the core SLS. Critics of SLS claim that the advent of SpaceX’s “Falcon Heavy” vehicle will render SLS unnecessary, but that launch vehicle was announced in 2011 and we have yet to see even a structural test article of it. It is stated that this vehicle will be able to put about 53 metric tons into LEO, significantly less than the 70 ton payload of the SLS core. The acceptance of this lower performance by its advocates is predicated on a proclaimed vastly lower cost, but as no Falcon Heavy has yet to fly, we have no idea of what its cost would be.

First, his numbers are out of date. The 53 MT number is from 2011, before the performance improvements to the Falcon core. In expendable mode, I’ve been told by sources at SpaceX that the performance of a densified stretched version (which is the only kind that will be flying) is more like 60 MT. As for its “cost,” no, know one knows what it will be except SpaceX. Moreover, no one cares, because we don’t pay its cost. We pay its price, which SpaceX lists at its web site. The current list price is $90M, but I suspect that’s for a reusable flight. Double the price for an expendable, and it’s still a small fraction of the cost of an SLS flight. Two of them will deliver almost the same tonnage as the 130 MT version of SLS (and NASA still has no idea how they’re going to get there), and still be a small fraction of the cost. As for the fact that “no structural test article exists of it,” why would it? What would one be “structurally” testing with such a thing? I expect the first FH we will see will be the one they plan to launch, on the pad, currently scheduled for half a year from now. It gets sillier from there.

Moreover, there are good reasons to question the technical viability of the Falcon Heavy. Released design details show that it consists of 3 Falcon 9 rockets, strapped together and burning simultaneously. Such a configuration would consist of 27 engines, all of which must burn for the same duration and thrust level. The Soviet Union once had a launch vehicle (the N-1) that had 30 rocket engines; it flew four times, each flight ending in a catastrophic fireball, largely as a result destabilizations following an engine-out condition.

Remind me, how many times did a nine-engine version of the N1 fly?

Oh, that’s right. Never.

As opposed to twenty-something for the Falcon 9. The Falcon Heavy isn’t a new vehicle sprung from the head of Zeus, as the N1 was. It is simply taking three rockets with a demonstrated flight history, and flying them simultaneously. The Falcon 9 has demonstrated engine-out capability, so there is no reason that three of them together won’t (there is plenty of gimbal authority to compensate for small loss of thrust on one of the side cores). The only problem with that number of engines is not reliability per se (it should be quite high) but schedule reliability if they maintain a rule of pad abort with an engine anomaly, because obviously the chances of that will increase with more engines. But they will work through this by a) continuously improving engine reliability with experience and b) changing flight rules and performance margin to allow a lift-off with engine out.

Given that there has in fact been no demonstration of actual technical need (that is, no payload or mission has been identified that can only be performed by a vehicle with the technical specification of SLS), yes, sorry Paul, but it is a jobs program, pure and simple. Or pure and complicated. But it is a roadblock to Mars or the moon, not a road to those places.

[Update a while later]

I missed this straw man the first time through:

A few larger pieces are more easily assembled in space than are a multitude of smaller ones.

This reminds me of a few years ago when I asked Mike Griffin what payload demanded an SLS and he yelled at me from the podium something like “We can’t take up every part and fastener on individual launches.”

Note the word choice: “a few” versus “multitudes.” In reality, if using a FH instead of SLS, it’s “a couple” versus four or five. Even with Vulcan, it would be “a couple” versus half a dozen at most. And of course we have no idea what Blue Origin has in mind. Even with smaller vehicles, it might be a dozen or so. Hardly “multitudes.”

[Update a couple minutes later]

OK, one more point. The same question to Paul I ask all other SLS supporters. If we can’t get beyond LEO without a rocket in this payload class, then why don’t we need two? After all, the Shuttle was down twice during its life for almost three years each, during which we had no (American) way to get astronauts to space. Why should we bet that the same thing won’t happen to this (Shuttle-derived) vehicle? If you don’t think that redundancy is important for this capability, it’s the same as thinking that getting beyond LEO isn’t important. The commercial alternatives give us resiliency; NASA-only solutions tend to be fragile. But that’s OK, because apparently the only thing that’s really important is maintaining the work force, which we can do whether we fly or not.

[OK, maybe one more point, I really should address this, because I don’t very often]

the Congress (who had twice voted their overwhelming support for the goals of lunar return, in two different authorization bills) mandated the construction of SLS, largely because NASA was dragging its feet on doing anything about it. Congress was concerned that an important national resource – the industrial and technical infrastructure (including its human resources) to build and fly HLV rockets – was being lost through neglect and attrition. They asked the agency to come up with a specific design for an HLV system but received no cooperation. So, they consulted external technical experts to derive the specifications of a general purpose HLV and mandated this design in the authorization. Its purpose was to make sure that the vehicle would be built and to assure that our national capability in this area would not be lost.

Let’s be clear what we’re talking about here. The capability to “build and fly” HLVs was not being lost, all that was being lost was the capability to do it with legacy Shuttle hardware, and its associated work force. That was the requirement that Congress built into the law. Moreover, it wasn’t being “lost” because, at least in terms of development (if that’s what one means by “build”) the capability had been lost decades before. Marshall Space Flight Center had not developed a heavy rocket since the 1970s, despite many failed attempts. Mike Griffin himself said that one of the purposes of Ares I was as a “training rocket,” so that they could learn how to develop rockets again, before they took on Ares V. If SpaceX flies FH in the next year (increasingly likely), they will have demonstrated their own capability to build a heavy rocket. There is no national need to maintain the ability to build SSMEs, Shuttle ETs and Shuttle SRBs and other obsolete hardware from forty years ago. There is only the need to maintain jobs in certain zip codes, which uncoincidentally generally exist in or near congressional districts or states of congressmen or senators on the space committees on the Hill.

Yes, Paul, it is a jobs program.

[Late-evening update]

Keith Cowing has weighed in as well, with lots of commenters.

Honestly, Paul is a smart guy. I cannot imagine what he is thinking, unless he is simply in the thrall of Apolloism.