The U.S. Air Force is in the early stages of determining how it might certify previously used rockets — such as the Falcon 9 first stages piling up Florida — to launch military satellites.

The Air Force awarded SpaceX its first big military launch contract in April, an $82.7 million award to lift a GPS 3 satellite in 2018.

Less than two weeks later, SpaceX landed the first stage of its Falcon 9 rocket on a drone ship in the Atlantic Ocean, marking the third time in seven attempts the Hawthorne, California-based company has successfully landed a booster after launch.

The timing, naturally, leads to questions about when — or if — the Air Force will consider reusing rockets to loft national security payloads into orbit.

“A certification or re-certification for the re-use of a rocket that has been previously launched has not been developed at this time,” the Space and Missile Systems Center, or SMC, said via email. “The Air Force is working to address this topic.”

In a response to questions from SpaceNews, SMC said the Air Force’s Evolved Expendable Launch Vehicle Program, through which the Defense Department contracts for most of its launches, “has not had discussions with industry about how we would buy or use launch services from a rocket that has already launched.”

However, the Air Force is studying how it would approve the use of a rocket that has previously flown. The primary goal of such a certification effort would be to reach the same expectation of success the Air Force demands of a flight-certified expendable rocket. Equivalent reliability is no small consideration given that military and intelligence satellites can cost $1 billion or more to build.

“In general, the Air Force would require the same level of mission assurance and reliability on a reusable system as expected of our certified expendable launchers today,” an SMC spokeswoman said via email. “The approach to achieve that level of confidence is just beginning to be assessed and will require active partnership with our launch service providers to ensure continued mission success in launching our National Security Space payloads.”

Reusability is once again a hot topic following SpaceX’s multiple landing demonstrations, not to mention Blue Origin’s recent suborbital launch-and -anding feats. United Launch Alliance, which now finds itself competing with SpaceX for national security missions after years of having that market to itself, is developing its next-generation Vulcan rocket with reusability in mind and has already begun working with the Air Force on the initial steps of that rocket’s certification. Sometime in the 2020s, the Denver-based company plans to re-fly Vulcan’s main engines after using a helicopter to catch them in midair as they descend beneath a parachute.

SpaceX and the ULA are the Air Force’s only providers currently certified to launch EELV-class payloads. The Air Force occasionally uses non-EELV-certified providers, such as Orbital ATK, to launch smaller or experimental payloads such as the ORS-3 mission that launched in 2013 aboard a Minotaur rocket.

The immediate appeal of reusability is the potential for big cost savings. SpaceX President Gwynne Shotwell told an audience of satellite executives in March that reusing Falcon 9’s first stage could result in a 30-percent price drop.

But the challenge for any launch provider whenever the topic resurfaces has been showing it can establish a sufficient flight rate to realize those savings.

“There are really good reasons the business isn’t doing this today,” said George Sowers, ULA’s vice president of advanced programs, who’s often referred to as the company’s “chief mad scientist.”

Executives at ULA and Arianespace have said a rocket would have to fly 10 times, at a minimum, just to recover the costs associated with making a rocket reusable in the first place.

A new certification process

Certification to launch national security satellites is already a cumbersome, carefully outlined affair. The Air Force took nearly two years to sign off on SpaceX’s Falcon 9 rocket and the delay caused Air Force Secretary Deborah Lee James to ask for independent reviews of how to streamline the process.

New standards for a rocket that’s already launched would complicate the certification — or re-certification — procedures.

What would those new steps entail?

They could include processes to show the Air Force how a company recovered a booster or engine, stored the hardware and what kind of refurbishment work might be necessary, Sowers said.

In other words, “how do I prove that re-used hardware has another launch left in it,” he said.

Another approach may be increased ground tests, he said. For example, a recovered rocket engine may be compared to a rocket engine has endured 10 full-duration burns on a test stand to “prove by inference” its flight worthiness and level of wear and tear.

Finally, Sowers said, the Air Force may turn to NASA for some of the techniques the agency used to certify the space shuttle, which flew 130 missions with a fleet of five orbiters equipped with reusable main engines and recoverable strap-on boosters that were fished out of the ocean, refilled with solid propellant and used again.

The Air Force is no stranger to reusable spacecraft, having invested in a number of its own concepts over the years. The X-37B spaceplanes it inherited from NASA have made return visits to orbit (the one that’s up there now is a year into its second tour), although they must rely on expendable rockets to get there in the first place.

If past is prologue, the Air Force is unlikely to be the first deeppocketed customer to ride a Falcon 9 that’s already flown. A safer bet is satellite fleet operator SES, which has said it’s willing to go first on a reused Falcon — provided it gets a steep discount. SES raised its hand to be SpaceX’s first customer for the Falcon 9 Full Thrust rocket, but the enhanced launcher ultimately debuted late last year with 11 Orbcomm satellites onboard instead.

“We actually asked them: If we do recover it, can we use it again and get a good price discount?,” SES Chief Technical Officer Martin Halliwell said last June, when the launch of its SES-9 satellite looked to be only a few months out. “We’re still in discussions.”

As it happened, SES-9 didn’t launch until March thanks to Falcon 9’s first failure just a couple of weeks after Halliwell called dibs.

When Falcon 9 returned to flight Dec. 22, it was with Orbcomm on board. While Falcon’s second stage was still carrying its 11-satellite payload to its drop-off point, first stage touched down safely near its Cape Canaveral launch pad for the first time.

When SES-9 finally launched in early March, SpaceX skipped a return-to-launch-site landing attempt in order to give the late-to-launch satellite a bigger, time-saving boost. SpaceX instead tried to land Falcon’s first stage on the deck of a drone ship floating in the Atlantic. And for a third time in a row, SpaceX failed to stick a water landing.

However, SpaceX’s next two drone ship landing attempts — on April 8 following a NASA cargo run to low Earth orbit and again May 6 following a launch to geostationary transfer orbit for Japan’s JSAT Corp. — were picture perfect.

Maybe SES — or the Air Force — can buy a ride on one of those.