Credits: Public domain / SpaceX, CC0 1.0.
SpaceX automatically aborted Starship flight 13 on Thursday, July 16, at Starbase, Texas. Several Raptor engines on the Super Heavy booster failed to ignite. The launcher did not leave the ground and did not explode. Two engines must be replaced before another attempt, the date of which remains undetermined. This update remains critical for Starlink and the Artemis lunar program.
An Automatic Shutdown at T-0, With No Liftoff or Explosion
The countdown had reached zero. The ignition sequence for Super Heavy’s 33 engines had just begun. Onboard computers then ordered the shutdown. SpaceX confirmed the attempt was called off on July 16 at 10:47 p.m. UTC. Teams then drained the liquid methane and oxygen, the two propellants that power the launcher.
The telemetry shown during the broadcast displayed four engines that were not producing thrust at the moment of shutdown. That observation, however, does not amount to a diagnosis. An engine may fail to appear lit because it is faulty. It may also have been disabled by the start of the automatic shutdown sequence. SpaceX has not publicly confirmed that four Raptors were out of service.
Elon Musk first said that some engines had failed to ignite. He then announced that two Raptors would be removed and replaced. The two figures therefore describe different realities: four non-ignitions visible on screen, but two engines designated for replacement. The underlying technical cause has not been made public.
The SpaceX chief mentioned the beginning of the week of July 20 as the most likely window. This is not a confirmed launch date. Engine replacement, vehicle inspections, and, if necessary, another ground test could still change that schedule.
The T-0 shutdown is neither a crash nor an explosion. The safety system prevented liftoff when the available thrust was not within spec. This does not make the attempt a success, but it did prevent an ignition anomaly from becoming an in-flight incident.
What Flight 13 Still Had to Validate
Flight 13 was supposed to be the second test of Starship V3, a larger and heavily modified version of the system. Its Super Heavy booster uses third-generation Raptors. In the long run, this architecture is meant to enable reuse of both stages, a goal far more ambitious than a simple suborbital launch.
On flight 12, on May 22, the launcher had lifted off, but several anomalies remained unresolved. The booster had failed to complete a controlled splashdown in the Gulf of Mexico after re-ignition problems. An upper-stage engine had also shut down prematurely, forcing SpaceX to give up on relighting a Raptor in space.
The U.S. Federal Aviation Administration closed its investigation into that previous test on July 13. It cited, among other things, thermal effects on propulsion components and incorrect settings in the engine warning system. SpaceX changed hardware, software, and the ignition sequence before flight 13.
The new mission was supposed to verify those fixes, attempt a relight in space, and test changes made to the heat shield. It also carried 20 operational Starlink V3 satellites. Placed on the same suborbital trajectory as Starship, they were to deploy their antennas and solar panels. They were then supposed to briefly communicate with ground stations and other satellites before reentering the atmosphere.
These goals explain why the next Starship launch is not just about replacing two engines. SpaceX still has to show that the flight 12 fixes work together under the real conditions of liftoff and a full flight.
SpaceX’s Approach to the Bet on Full Reusability
SpaceX promotes a highly iterative process: build, fly, measure, fix, and then repeat quickly. The flight hardware itself serves as a test instrument. This method accepts more visible anomalies during development, as long as each attempt produces usable data and the safeguards work.
That approach should not be reduced to a simple fail-fast doctrine. A ground abort reveals an anomaly, but also the system’s ability to detect it. Conversely, piling up tests guarantees neither future reliability nor schedule compliance. The outcome is measured by the gradual closure of technical issues.
The comparison with Ariane 6 highlights this difference in logic. The European launcher was developed under the European Space Agency and is operated by Arianespace. It was qualified before providing predictable institutional and commercial service. Starship, by contrast, remains a prototype program that SpaceX rapidly changes between flights. Both lines test intensely, but they do not expose the same risks at the same time.
The bar set by SpaceX is also exceptionally high. Starship is 124 meters tall, combines 39 engines across its two stages, and aims for full reusability. For lunar missions, it will also have to master rendezvous and propellant transfer in orbit. The number of tests therefore is not enough to judge the program’s maturity without looking at what each one was meant to demonstrate.
Artemis Depends On A Starship Still In Development
NASA has selected a lunar version of Starship as the Human Landing System. This vehicle is meant to carry astronauts between lunar orbit and the surface. The program cannot settle for a launcher that merely reaches space. It requires an uncrewed lunar landing, complex orbital operations, and a level of safety compatible with human spaceflight.
A NASA Office of Inspector General audit was released in March 2026. It estimated that the Starship intended for Artemis III was already at least two years behind schedule. Further delays remained possible. In practice, this dependency places the lunar lander among the milestones that determine the schedule for the U.S. return to the Moon.
That does not mean a new Artemis delay can be blamed solely on the July 16 shutdown. Flight 13 is a suborbital test of the base system, not a rehearsal for a lunar mission. Still, each lingering difficulty with propulsion, relighting, or reusability reduces the margin available for the complex demonstrations NASA expects.
Europe Has Assets, But Is Racing To Catch Up On Reusability
The contrast also feeds the European debate. In an assessment published in 2025, the European Commission noted that the continent’s companies lag behind on reusable launchers. That gap also affected the United States and China. SpaceX is not the only one moving ahead. Blue Origin is also developing launch capabilities and a lunar lander for NASA. Several Chinese players are working on stage recovery, too.
Europe is not starting from scratch, however. Ariane 6 and Vega-C have restored its independent access to space. It has an industrial base, scientific expertise, and the Guiana Space Center. This spaceport is about 500 kilometers north of the equator. It benefits from Earth’s rotation to improve eastward launch performance. It also offers clear trajectories over the ocean.
The strategy approved at the end of 2025 by the European Space Agency funds upgrades to Ariane 6 and Vega-C. It also supports the European Launcher Challenge and next-generation technologies. The Themis demonstrator is meant, in particular, to prepare for reusing a European main stage. These projects show that the catch-up effort is underway, but also that this capability is not yet operational.
Flight 13’s shutdown sums up the tension inherent in Starship. SpaceX’s rapid-test method allows it to confront an unusual architecture with reality early on. It also lays bare everything that still separates the prototype from a routine reusable launcher. Starlink and, indirectly, the U.S. lunar schedule depend on that transformation.