Important enough

A fourth Falcon 1 stood on Omelek, a few acres of uninhabited reef on Kwajalein Atoll. The Pacific is most of the picture. Orbit was not yet a fact.

SpaceX was incorporated in 2002 to build, not to buy. Later recollection (Jim Cantrell, Ashlee Vance, Musk on 60 Minutes) says the preceding plan had been a greenhouse on Mars, and that buying a refurbished missile in Russia did not survive the price. That office, and no other: launch cost became the problem, so building replaced buying. It is not a 2002 press kit.

The object is a small two-stage kerosene-and-oxygen rocket, about seventy feet. Merlin on the first stage, Kestrel on the second. Tom Mueller, first employee and vice president of propulsion, is why those engines exist. In March 2008 he said the next Falcon 1 would fly the new Merlin 1C from Kwajalein. This was not Falcon 9. First they had to prove a small one could finish the job. In 2008 a private company reaching orbit with a fully liquid-fueled rocket it had designed and built itself had not been done.

The four flights are not three failures and then courage. They are failure, finding, a change that entered the next vehicle, and a new unknown.

Flight 1, 24 March 2006: an aluminum B-nut on the fuel-pump inlet cracked from corrosion not visible in inspection. Kerosene leaked. Fire at about thirty seconds. A DARPA board found the locking wire still in place. It was not a loose nut. Stainless fittings and tighter inspection went into the next vehicle. Humidity treated as a cause.

Flight 2, 21 March 2007: first stage flew. At staging the interstage bumped the second-stage nozzle. LOX slosh. The engine shut down around T+7:30. About 289 km, about 5.1 km/s. Baffles and a control-logic change went into the next vehicle. The next vehicle also flew a new engine, Merlin 1C.

The surviving fourth-flight record is a webcast and a same-day report. Spaceflight Now: a smooth countdown; Merlin 1C; first-stage burn about two and a half minutes; stage separation five seconds after burnout (the Flight 3 correction); Kestrel; fairing; a second-stage restart to circularize. Employees in the Hawthorne hangar cheered when Kestrel lit. Max Vozoff is named as mission manager. The pad is not invented.

What failure would have meant, as far as the record will go. By June 2008 SpaceX said nearly five hundred people. Ratsat was a dummy; Flight 3 had already lost payloads, including NASA’s. On 4 August 2008, after that failure, company history records a further Founders Fund investment of $20 million. That is 2008 capital. It is not proof that a fifth rocket was waiting. In 2012 Musk said he had put on the order of a hundred million dollars of his own money into the work, and that a failed fourth would have left no resources for a fifth. Those figures are 2012 speech. They are not a 2008 ledger line we have in hand. We do not have a payroll, a visa list, or a rent book, and we will not invent one.

The vehicle reached orbit with a dummy payload. A happy ending does not prove the gamble was wise. It proves the vehicle worked that time. Learning had entered the hardware. Whether continuing had been wise is not the same fact.

On 23 December 2008 NASA announced Commercial Resupply Services awards. SpaceX’s share was about $1.6 billion for twelve missions. That paper is almost three months after the pad. It is not “days later,” and it was not a fact on 28 September. COTS is not CRS. The 2006 demonstration agreement is a different paper again.