Commercial space launch cognoscenti estimate that, since 2009, nearly a billion dollars in new orders has gone to a single U.S. company developing a rocket that has yet to be flown.

The catch: while it is common in the commercial aviation world to get a backlog of orders for platforms that have yet to fly, it is unheard of in the space launch industry.

Over the past two years, Hawthorne, Calif.-based Space Exploration Technologies (SpaceX) has proved its Falcon 9 medium-lift rocket can deliver unmanned payloads to low Earth orbit on at least two occasions. But a major modification to that launch vehicle is under way, including a new Merlin engine, stretched fuel tanks and a wider payload fairing. In other words, while the name is mostly the same, the actual Falcon 9 variant that SpaceX has scheduled to begin carrying telecom satellites to geostationary orbit by mid-2013 is still only in development and not expected to undergo flight trials until the second quarter of next year.

“If my estimate is correct, it's close to a billion dollars of business that SpaceX has acquired in the commercial launch market—both the LEO [low-Earth-orbit] system and the GEO system—without really a flight record,” says Frank McKenna, president of Reston, Va.-based International Launch Services (ILS), which manages commercial launches on Russian-built Proton rockets. “It's probably a first in the commercial launch industry,” he said during the World Satellite Business Week conference here Sept. 11.

Given SpaceX's advertised price of around $59 million per launch, McKenna says the rest of industry has lost about 40% of new orders it might have reaped if SpaceX was not in the game. But that was before the world's second-largest satellite fleet operator by revenue, Luxembourg-based SES, ordered three more Falcon 9 launches starting in 2015. By 2010 SES had already become the first major commercial operator to commit to SpaceX, ordering a Falcon 9 launch for the SES-8 satellite to be lofted into geostationary orbit in July 2013.

“What we have seen in the light-to-medium market is extraordinarily aggressive competition brought on by SpaceX and Falcon,” McKenna says. “I think that has resulted in fairly significant percentage drops in the prices of GEO and LEO launches [over the past three years.]”

Falcon 9 is even having an impact in Europe, where officials have pointed to SpaceX as a rationale for reviewing current launch policies and hastening plans by the European Space Agency to develop a more flexible, affordable and less-commercially dependent version of the Ariane 5.

In the meantime, the improved Falcon, dubbed Falcon 9 v1.1, is racing against the mid-2013 deadline to demonstrate flight hardware ahead of the SES launch, a mission that aims to send the Canadian science satellite Cassiope to near-polar orbit from a new launch site SpaceX is building at Vandenberg AFB, Calif.

If SpaceX is unable to make the deadline, SES has back-up options, including the Ariane 5 heavy-lift launcher managed by European consortium Arianespace of Evry, France. To date that company has already bailed out at least one former SpaceX customer. In 2009, London-based Avanti Communications swapped a Falcon 9 for an Ariane 5 to loft its Ka-band Hylas 1 broadband satellite because it could not wait for SpaceX's launcher to be qualified.

Moreover, last month Iridium announced it had rearranged launch plans for its Iridium Next constellation of 72 satellites to give SpaceX more time to prepare for the initial mission. Iridium CEO Matt Desch says the company, which signed a $492 million contract with SpaceX in 2010, is actually saving about $15 million by manifesting 10 Iridium Next satellites on seven Falcon 9 rockets, rather than nine aboard eight rockets, over the course of two years beginning in mid-2015.

Still, despite a series of lengthy delays and a few nail-biting technical issues—last-minute hardware mods on the launch pad and an engine anomaly—the Falcon 9 family is widely seen as a technological success brimming with commercial promise in an industry generally averse to innovation and risk. So far, few details of the Falcon 9 v1.1 have emerged, though qualification testing of the new rocket's more powerful Merlin engine, the Merlin 1D, is expected to wrap up by year-end.

With 50% more thrust than the Merlin 1C, the 1D will enable Falcon 9 v1.1 to loft as much as 4,850 kg (10,692 lb.) to geostationary orbit from Cape Canaveral AFS, Fla. “It's a much higher-thrust engine, but surprisingly a lot of the infrastructure of that engine is the same,” says SpaceX President Gwynne Shotwell. The chamber itself is the same. “The difference is instead of plating it, we're brazing on the jackets.”

The nozzle, too, is different: “We use a tube-wall nozzle on the 1C and we're using the milled copper for the nozzle extension,” she says.

Another change, she says, involves the rocket's nine Merlin 1D engines, which will be positioned in an octagonal configuration, rather than the “tic-tac-toe” placement on the current Falcon 9.

“You actually want the engines around the perimeter at the tank, otherwise you are carrying that load from those engines that are not on the skin,” she says. “You've got to carry them out to the skin, because that is the primary load path for the launch vehicle.”