“A long term strategy and corresponding plans must also be developed . . . a set of notional milestones, launches, and hardware developments that are sufficiently defined so as to allow a cost estimate” – NASA Advisory Council finding April 2015
There is a story, perhaps apocryphal, that in the backseat of the car from NASA Headquarters to the White House to brief President Kennedy on the possibility of a moon landing, the legendary NASA Administrator James Webb decided to double the estimated cost of the program. Whether that part is true or not, the Webb estimate delivered that day in the spring of 1961 was significantly lower than the actual Apollo program.
Norman Augustine’s famous book of “Laws” concerning government acquisition states that all program cost estimates are subject to a correction factor of [1+ 0.52/(1+8t3)] where t is the percent of the procurement period completed. Or as he finishes the chapter with Law XXIV: “The most unsuccessful three years in the education of cost estimators appears to be fifth-grade arithmetic.”
During the so-called Space Exploration Initiative (SEI) days of the late 1980’s, the ’90-day-study’ came up with a very detailed plan to go to Mars . . . and the cost estimate made that plan dead on arrival at Congress. This lesson has not been lost on the NASA leadership.
A historical example may be in order. Look at the Apollo program hardware, specifically the Service Module and its rocket system the SPS (Service Propulsion System). That rocket engine is tremendously more powerful than the subsequent lunar landing flights needed. Why was such a large rocket engine installed on the Apollo SM? In 1961 when the first real plans for lunar landing were baselined, Direct Lunar Ascent was the designated mode. Some sort of huge lander would drop the entire CM/SM stack onto the lunar surface and the SPS had to be big enough to lift the astronauts, the Command Module, and the Service Module off the lunar surface and put them on a trajectory for the Earth.
To put that big stack – the CM/SM and the Landing Stage on a trajectory to the moon, the puny Saturn V was not big enough. Developing a much larger rocket was required – they called it Nova. Nova would have twice the number of F-1 engines as the Saturn V, tanks twice the diameter, much taller, more stages, etc., etc., etc. Exactly how the Nova rocket would be built was never figured out – it would be too big to fit under the ceiling of the factory at Michoud where the Saturn V first stage was made. The notions of how to transport that rocket to the launch pad were . . . notional.
Then along came some bright boys at Langley headed by John Houbolt who advocated an operationally more complex idea called Lunar Orbit Rendezvous – which only needed the Saturn V already under development.
The Nova rocket, the 100 foot tall Lunar Descent Stage, all went in the dust bin of history were never developed. But the contract for the SPS engine had already been let. Any real need to downsize that engine? No, but much less propellant would be carried in the tanks. If the Apollo CM/SM were somehow magically transported to the surface of the moon, the SPS had enough oomph to lift them off . . . but probably not enough gas onboard to get very far.
LOR was a good idea. Lots of folks are proposing ideas for future space travel. Some of them are actually pretty good. Locking a plan down means new, good ideas can’t complete.
History cries out with lessons. Some of them are subtle. Having detailed plans is generally good; believing in them too much is not. In the military they are fond of quoting the maxim: “No battle plan survives its first encounter with the enemy”. In space, the enemy is physics and chemistry . . . and finances. It may be that flexibility and leaving options open provides a better path for our long term ambitions in space. Who knows what may be invented in the next five years that could change the entire game plan?
Would we have made it to the moon if we tried to build the Nova rocket to do it? Maybe, maybe not.
The wrong plan can easily come with a forecast cost – a shock to the system – such that the program is never approved. Having a reasonable plan for the next step while keeping the goal in sight might actually be better. Waiting a little while doing some testing and development might be a good idea. Finding creative ways of controlling costs is mandatory.
Meanwhile, anybody seen Zephram Cockrane out there? Or at least the ghost of John Houbolt?