"Cheating" On the Way to Orbit

John Jurist, Sam Dinkin, and David Livingston have a good paper on the limitations to a reduction in the cost per pound to low earth orbit.  In the paper, the authors detail what it costs in broad terms to put 1,000 pound payload into orbit under various scenarios.  The conclusion is that it is difficult to get costs under $2,000 per pound on a 1,000 pound package, if only because range fees and insurance equal over $1,000 per pound.

Their conclusions on the costs of these systems seem broadly reasonable, although I would emphasize that there are several ways to "cheat" to minimize development costs and maximize revenues on a reusable launch system, of which SpaceX likely is taking some advantage.

(1) Spreading R&D over several vehicle models.  Engines and other components may be shared across vehicle models, depending on design.  SpaceX is sharing many components across at least two models:  the Falcon I and Falcon V.

(2) Getting the government to pay for one or more demonstration flights.  The market risk and therefore the discount rate decline rather steeply when the government is promising money for demonstrations.  SpaceX is getting about $6 million apiece from DARPA for its first two launches against about $50 million - $75 million in total R&D for the Falcon I.  The article mentions an 18% discount rate as an assumption for a start-up rocket company, but I think it rightly should be much higher than that -- more like 25% per annum if you don't have any customers signed.

(3) Getting customers to pay for other things beside launch.  SpaceX may get a contract with the Air Force for "responsive launch."  By this, it means that SpaceX provides the capability to launch on 48 hours notice and do quick turnaround to launch again.  NASA, Bigelow Aerospace, and others may find it worthwhile to pay a modest amount of "insurance" for responsive launch service guarantees in case of emergency.  In the start-up phase, DARPA is paying some small millions to SpaceX for a study on responsive launch, further reducing the risk in the business proposition, if DARPA doesn't require much paperwork.

(4) Opportunistic R&D.  SpaceX has an expendable system that they are designing for reusability, when it is cheap to add to the design.  If the reusability doesn't pan out, then nothing is lost.  If the reusability turns out to work, then they have all of the cost reduction of reusability without the assumed large associated R&D costs.

Anyway, good luck to SpaceX and here's to the hope that it can climb the mountain of R&D one step at a time.

Comments

[Kevin Parkin]

One other thing that occurs to me about this paper is that range costs and insurance might be specific to the US.

If lower cost technology is introduced there may be nothing to stop other countries going way down through the price floor quoted (in fact China is supposedly achieving $4k/kg just with existing technology, whereas we get about $16k/kg).

[Daniel Schmelzer]

As I understand it, range fees vary from location to location in the US (I don't know how much variation there is--you would think this pricing information would be readily available, but oddly it doesn't seem to be). If a large percentage of the launch costs are going to range fees, maybe customers would spend more time being savvy buyers of range services. Even if the services are provided by the government, I would expect some market responsiveness regarding price.

SpaceX doesn't look like they are spending a fortune on their pads and are already looking at 4 pads at 3 locations. Maybe adding a couple more locations would help start the virtuous competitive cycle.

With regard to the price floor, what I am taking away from all of these discussions is that we would do well to expect the R&D mountain for chemical reusables to be staggeringly high. However, it's encouraging to see some prospect of incremental improvements coming to fruition, however small.

[Juan Suros]

The paper hints pretty broadly that building your own private spaceport will lower fuel and range cost significantly, which seems reasonable. I think they may have missed the boat on insurance costs, however. I didn't catch it if they said it themselves, but is the insurance burden directly a result of the high launch costs, or the high value of the payloads?

Modern satellites are very expensive. They are designed to operate for years with no maintenance, and their designs put a very high premium on small size and mass. Space probes are even more expensive. It seems a reasonable assumption that a vastly higher spacelift traffic volume would lower the value of the average payload. Wouldn't the insurance costs on a load of girders and Kobe beef be lower than that on a communications satellite?

I wish the authors had done the numbers on the Falcon V launcher, where commercial rates are available to compare with their analysis.