To actually do the refuelling, though, requires feats of technology. Picture a satellite hurtling through space at 11,000 kilometres an hour. The gas station would have to spot it, manoeuvre alongside, and connect to it by a probe or robotic arm. Then the robotic arm would have to open up the fuel valve to pump in hydrazine - that is, presuming the door isn't seared shut after a decade or more in space.
"It's a complex operation," Mr. Friedmann said. "It's not welding car doors."
Refuelling a satellite in space has been done only twice before, in tightly controlled experiments, both times using new satellites launched as part of the test, not ones that had spent years in orbit. The Japanese space agency accomplished a space refuelling in the late 1990s. In 2007, the Pentagon hired Boeing Co. - with MDA aboard as a subcontractor -and repeated the deed.
No one has ever captured a satellite already in orbit. Assuming it is technically possible, the next question is whether it is economically attractive. The possibility of a collision between the refuelling station and a satellite could wipe out any potential profits. Then there's the broader question of whether it's worth refuelling old satellites, given the possibility that other equipment might fail a year or two later.
"Satellite servicing is very much one of a kind," said Dominique Rora, a senior space underwriter at Axa Corporate Solutions, a specialty insurer in Paris. "We are all very much interested." But he cautions that making the technology work will require deep pockets and persistence.
MDA's great advantage is its mix of proven technologies. In addition to its long experience with satellites, it owns the robotics technology made famous in the Canadarm, as well as systems that allow a mobile service station to track and approach a target satellite. Putting these pieces together in a full satellite-servicing system is the trick.
According to analyst Paul Steep of Scotia Capital, MDA will soon make a decision on whether to proceed with the project. What it needs is a lead customer, probably in the communications satellite business, to help back the costs of a first mission. The all-in bill is likely to be several hundred million dollars.
One encouraging sign for the project is the frequent and enthusiastic mentions it earns from the normally reserved Mr. Friedmann. "It's a project that if you don't have a good shot, you probably don't talk about it," said Mr. Li, of Raymond James. "The fact Dan does talk about it, and spends a lot of time on it, tells me it's moving forward."
MDA will have to move quickly if it wants to stay ahead of state-backed competitors. Earlier this year, the German space agency awarded contracts to companies led by OHB Technology AG for preliminary work on a project similar to MDA's, with an estimated cost of about $260-million.
MDA hasn't yet talked about how much it might spend. Without any backing from the Canadian government - though using technology developed with government support - it would be the first public company to ever embark on such an ambitious mission.
For Mr. Friedmann, limited government support is just one more obstacle to overcome. Two years after being rejected in his bid to exit the space business, he is embracing its new possibilities. "Having had [the sale] blocked, my job is to do the best within the constraints - and we've done a good job of that."
A space service manual
Refuelling satellites already in orbit could extend their working lives by years. Here are the key numbers on MDA's proposal for service stations in space:
Estimated time the docking manoeuvre would take a communications satellite out of service.
Amount, in kilograms, of hydrazine that the system might carry, enough to provide five years worth of fuel to 10 satellites.
Estimated replacement cost of 65 satellites that ran out of fuel or failed in the past decade.
Cost to replace 136 satellites that are expected to be decommissioned early, from 2012-20, mostly because they run out of fuel.