Elon Musk has pushed back hard against a NASA safety panel claim that the NASA Moon 22 launches estimate is right. The SpaceX chief, however, says the real number is far lower. His reply landed on X within hours of the panel meeting.
The clash, furthermore, broke out at the quarterly public meeting of NASA’s Aerospace Safety Advisory Panel on October 22. Panel member Charlie Precourt, a former astronaut, specifically named Starship refueling as a major risk for the crewed lunar landing. He said the last estimate stood at around 22 flights.
Musk answered directly on X. “Actual number will be more like 6 to 8,” he wrote. The gap between the two figures in the NASA Moon 22 launches debate is huge. It comes down, in fact, to how much propellant each tanker delivers to orbit and how much boils off on the way.

Each tanker flight is, in other words, a full Starship launch in its own right. The tankers dock with an orbiting depot and hand over leftover propellant. This lander launches afterwards, fills up, and departs for the Moon. The campaign needs a long chain of launches before a single landing.
The NASA Moon 22 launches figure did not come from nowhere. Precourt based it, nevertheless, on the last internal estimate available to the panel. That estimate assumed a certain tanker payload as well as boiloff rate. Change those assumptions, and the number swings wildly in either direction.
SpaceX has been testing orbital refueling concepts for years. The company plans a ship to ship propellant transfer demo as a key milestone. That test, in fact, will provide the first real data on transfer efficiency. Until it flies, both sides are doing math on paper.
Musk has a history of beating skeptical estimates. SpaceX reused boosters when experts called it impossible. The company, moreover, iterates faster than traditional aerospace. Past performance, however, does not guarantee future results. The stakes could not be higher.
Eric Berger of Ars Technica ran the numbers. He estimated that even at two Starship launches per week by late 2027, filling a depot at the panel’s NASA Moon 22 launches figure would take about three months. That schedule, moreover, leaves little room for delays before the crewed flight.

NASA’s own rehearsal plan covers a smaller slice of the chain. The Artemis III rehearsal calls for three launches and docking tests in Earth orbit. Those operations can verify spacecraft interfaces without repeating a full lunar fueling campaign. The orbital fuel transfer demo, meanwhile, is targeted for early 2027.
Neither side supplied a shared calculation. The panel as well as Musk gave no common breakdown of tanker payloads, reserve margins, or storage losses. The counting boundary is unclear. Depot deployment as well as lander launches must be counted apart from fuel deliveries. Both figures stay provisional.
The panel now plans a formal recommendation in the NASA Moon 22 launches dispute. It wants NASA to rethink Artemis IV goals and possibly its architecture. Precourt asked for detailed test and data plans covering ground and flight demos. The worry is not new. NASA’s inspector general had previously flagged cryogenic fluid handling among lander risks in a March audit.
The Artemis program cannot afford a fueling fiasco. Astronauts will one day ride the system that these estimates describe. Getting the number right, consequently, matters more than winning the argument. The NASA Moon 22 launches debate, ultimately, is about safety margins written in rocket fuel.
The next real test is the orbital transfer demonstration. Its results will show how much usable fuel each flight can deliver. Until then, the NASA Moon 22 launches debate pits estimates against estimates. The Moon will wait for no one. Engineers, meanwhile, keep refining the numbers on both sides.
