Starship is doing orbital refueling tests and

There’s a step in the Mars plan that gets less attention than the rocket itself. It’s not the launch. It’s not the landing. It’s what happens in between.

Orbital refueling.

Starship can carry about 100 tons to low Earth orbit. That’s enormous. But to get to Mars with enough fuel to land, you need much more than 100 tons in orbit. You need a full tank, which means launching a tanker Starship, docking with the Mars-bound Starship, and transferring cryogenic propellant in microgravity.

SpaceX just tested the fuel transfer mechanism. And it worked.

Why this is the hardest part

Launching rockets is hard but understood. Landing rockets is hard but SpaceX has demonstrated it. Orbital refueling in microgravity with cryogenic fluids is something nobody has done at this scale.

Cryogenic propellant (liquid methane and liquid oxygen) boils. In orbit, without gravity to keep the liquid settled at the bottom of the tank, the propellant floats. Bubbles form. The transfer port needs to handle a mixture of liquid and gas. The thermal management needs to keep the propellant cold enough through the transfer.

None of these are unsolvable problems. All of them are engineering challenges that required actual testing in actual orbit.

What the test showed

The details SpaceX has shared are limited, but the test apparently demonstrated controlled transfer of fluid between tanks in microgravity. The thermal management worked. The settling technique (using small thruster firings to push the liquid toward the transfer port) worked.

Not perfectly. Not at full scale. But the mechanism is sound.

If SpaceX can refuel a Starship in orbit, the Mars architecture works. Launch a ship. Launch 5-8 tankers. Transfer the fuel. Send the fully loaded ship to Mars during the transfer window.

Every piece of this chain has now been tested individually. The rocket launches. The rocket lands. The ship survives reentry. The fuel transfers in orbit. What remains is putting the chain together end to end. And doing it reliably enough to send cargo, and eventually people, on a one-way trip lasting 6-9 months.

The timeline I believe

SpaceX says uncrewed Starship to Mars in 2026. I think 2028-2029 is more realistic. Crewed missions in the 2030s.

But those dates are less important than the fact that every technical element has been demonstrated. Orbital refueling was the last major question mark. The question mark is now a data point.

The path to Mars is, for the first time, technically demonstrated end to end. Not hypothetical. Tested.

That’s worth sitting with for a moment. Every element of the technology needed to send humans to another planet has been tested in some form. The combination hasn’t happened yet. But the pieces are on the table.

I’ve been watching SpaceX build this ship for seven years. The water tower that everyone laughed at. The explosions. The belly flop. The catch. And now the fuel transfer.

It’s getting close enough to taste.

Related thinking: