A ship came home. Now it has to fly again.
Last week a spacecraft returned from orbit and was caught by the tower it launched from. That is the first half of full reuse. This issue is about the second half.
Both halves of a rocket should fly again, like an aircraft.
Boosters have been landing and reflying for years. Upper stages have not. They go all the way to orbital speed, so they come back about eight kilometres per second faster and have to survive re-entry. Until last week, every orbital upper stage ever flown was thrown away after one use.
This matters beyond cost per launch. Plans for the Moon and Mars depend on refuelling a ship in orbit from separate tanker flights, and that only becomes affordable if the tankers fly again and again.
Fourteen orbits, then two steel arms.
Starship Flight 15 lifted off from Starbase, Texas, on 13 November 2026 and released 26 Starlink satellites. Three of them carried cameras and photographed the ship’s heat shield in orbit. With the shield judged healthy, controllers skipped the backup ocean splashdown. About 22.5 hours and 14 orbits after launch, Ship 42 flew back to its launch site and the tower caught it. It is the first orbital upper stage ever recovered.
Only half of the planned double catch worked. The Super Heavy booster diverted away from the tower after a low-propellant alarm cut its return burn short. SpaceX put it down to slightly underperforming engines and listed two fixes: more consistent engines, and shutting most of them down about a second earlier on the way up to keep more propellant for the trip back.
One is ahead. Three are worth watching.
- SpaceX
- 15 Starship flights. Three boosters caught and two reused. One ship caught, none reflown yet. Its president has said the first ship reflight will be attempted early in 2027.
- Stoke Space
- Building Nova with both stages designed for reuse and a metal upper stage cooled by its own hydrogen. Has raised about $2.3 billion. Has not flown to orbit; the first launch is targeted for early 2027.
- Blue Origin
- Landed a New Glenn booster in November 2025 and flew it again in April 2026. The rocket has been grounded since a pad explosion in May, with the next launch targeted for January 2027. Its upper stage is expendable.
- LandSpace
- Landed and recovered a Zhuque-3 first stage on 19 August 2026, eight months after its first attempt exploded, and has since qualified the recovered engines for reuse. Booster reuse only.
Catching it was the easy half.
- Turnaround
- No ship has flown twice. Nobody yet knows how many heat-shield tiles need work after one re-entry, or how many days that takes. A former NASA heat-shield researcher has called the tile design a “dead end” for rapid reuse.
- Booster margin
- The new booster version has not yet returned with propellant to spare.
- Refuelling
- Moving propellant from one ship to another in orbit has never been done. The test was planned for 2026 and has slipped to early 2027.
- Flight count
- A Moon mission needs many tanker launches for one departure. Every one of them has to work.
How these could be solved.
This part is XLight’s own analysis, not reporting.
- Measure, don’t assume
- Flight 15 inspected its own heat shield from orbit before committing to come home. Make that standard and the output becomes a tile-by-tile repair list that exists before the ship lands. Turnaround then depends on how fast tiles can be swapped, which is a factory problem.
- Buy margin with payload
- Cutting the engines a second early costs a little payload and buys a booster that comes back. Until the engines are consistent, that is the cheap trade.
- Bigger tankers
- The next ship version is planned at about 70 m, with over 60% more propellant. SpaceX’s own estimate is over 200 tonnes delivered per tanker, which would cut a lunar refill to five or six launches.
- Keep a second design alive
- If tiles never turn around fast enough, the alternative is a metal shield that is actively cooled, which is the route Stoke is taking. Two different answers to the same problem make it more likely that one works.
What to watch next: the first ship to fly twice, the first propellant transfer, and the number of days between a catch and the next launch.
XLight magazine is independent. It is not affiliated with, sponsored by or endorsed by any company it covers.