Frozen igniter fluid lines were to blame for the failure of an upper-stage restart during the first Falcon 9 v1.1 launch on Sept. 29.

Officials at Space Exploration Technologies (SpaceX), manufacturer of the rocket, did not detect the problem during ground tests because “ambient air kept the lines warm,” says Emily Shanklin, a spokeswoman for the company. It is unclear whether thermal testing was conducted. “We’ve added insulation and made sure that cold oxygen can’t impinge on the lines” in future missions, she says.

The Canadian payload reached its orbit during the Sept. 29 launch, allowing the company to declare the mission a success, but the restart attempt exposed a problem not previously seen in testing.

The upper-stage restart was a risk-reduction exercise in preparation for the Nov. 25 launch of SES-8, during which the system must work properly to deliver the payload to its geostationary orbit. This will be the first attempt to delivery payload into that high orbit, which is in demand given most communications providers operate there, as well as the U.S. Air Force.

SES-8 will be followed closely by another Falcon 9 v1.1 launch to geostationary orbit, this time for a Thai communications satellite.

The failure was the result of a “pretty straightforward error,” according to Barry Matsumori, vice president of commercial sales and business development. He spoke Nov. 14 at Aviation Week’s A&D Programs conference in Phoenix. The additional insulation did not require requalification work, he said. SES has already fueled the spacecraft, signaling confidence in readiness of the Falcon 9 v1.1 for the mission, he added. He said the Thaicom launch will follow one month after the SES mission.

Success of the upper stage is critical for SpaceX to capture the kind of business that company founder Elon Musk has been after. The company has long sought to unseat United Launch Alliance (ULA), which builds and operates the Atlas V and Delta IV family of rockets, as the sole source of launch boosters for the Pentagon, NASA and the intelligence community.

As such, the company is undergoing a rigorous process to get certified to compete with ULA for such missions. Certification, however, requires three successful Falcon 9 v1.1 launches, two of which must be consecutive. Also included is a rigorous review of the company’s technical data and engineering and technical processes.

SpaceX officials have said the Sept. 29 launch is the first of the three required for Air Force certification. The Air Force, however, has not confirmed this and officials there say they are continuing to analyze data from the mission to determine its eligibility.

Though the Air Force is on the cusp of issuing a massive, first-of-a-kind multiyear contract with ULA for up to 36 cores, 14 additional launches projected in the next five years are potentially up for grabs. “Every one [of the 14 launches] requires at least a minimum of two starts on the upper-stage engine . . . They had an anomaly [and] what we are doing with them now is we are watching the anomaly process as it plays out and looking at how they handle it,” Scott Correll, the Air Force’s program executive officer for launch, told Aviation Week earlier this month. “The upper-stage engine is critical to our ability to meet our national security space requirements.”

He surmises that SpaceX could be ready to compete as soon as 2015 in advance of a fiscal 2016 competition. Company officials, however, said they are targeting a more aggressive schedule.

Correll says he hopes to sign the 36-core deal with ULA by mid-December, when he retires from his position.