Four Engines Didn't Light, and Starship's Schedule Got Tighter

At 5:45 p.m. CDT on July 16, at Pad 2 in Starbase, Texas, Starship Flight 13 got as far as ignition and no further. The Super Heavy booster began its start sequence, four engines failed to light, and the flight computer called an automatic abort. Elon Musk confirmed it afterward: some of the engines didn't start, and the vehicle stood down on its own.

An abort at ignition is the least bad way for a launch attempt to fail. The hold-down system did what it exists to do, nothing was lost, and both vehicles are intact. But this one is worth reading carefully, because of what was on board and because of what the calendar looks like behind it.

What was flying

Flight 13 used Booster 20 and Ship 40, both on their first flight, neither slated for recovery. The payload was 20 production Starlink Version 3 satellites — the first V3s that would have reached space.

The profile was suborbital. The satellites were to deploy, extend solar arrays and antennas, briefly connect to the operational Starlink network to prove the hardware works end to end, and then demise on reentry roughly twenty minutes after deployment. Disposable, by design. The point was never to add capacity; it was to demonstrate that a production V3 can be dispensed from Starship and come alive on orbit.

Two secondary objectives carried over from Flight 12 in May, where neither was achieved: relighting a Raptor during the coast phase, and a controlled booster landing in the Gulf of Mexico. Both are prerequisites for everything Starship is supposed to eventually do. You cannot go to orbit without a reliable in-space relight. You cannot make the economics work without recovering the booster.

So Flight 13 was carrying three separate unproven capabilities at once, which is a reasonable way to run a test campaign and a hard way to have a clean flight.

The engine question

Here is the detail worth sitting with. Telemetry showed four engines failing to ignite. Musk said the fix is to replace two Raptor engines on the booster before the next attempt, with the next try "hopefully in a few days" — most reporting points to early the following week.

Those two numbers do not obviously reconcile, and SpaceX has not published the reconciliation. The benign reading is that two engines had hardware faults and two more were commanded off by an automated sequence that aborts the whole start when it detects a problem — in which case the fault is genuinely two engines and the diagnosis is complete. A less comfortable reading is that the start sequence itself is marginal at the low end and four engines is what marginal looks like on a bad day.

Super Heavy flies 33 Raptors. Ignition is the single most demanding moment in the vehicle's life, and starting 33 of anything simultaneously is a reliability problem that compounds against you: even excellent per-engine ignition odds produce frequent aborts when you need all of them at once. SpaceX has already implemented hardware modifications aimed at engine reliability based on prior flight data. Flight 13 says that work is not finished.

None of that is alarming on its own — this is exactly the class of problem a flight test campaign exists to surface, and surfacing it on the pad rather than at altitude is the system working. But it is the second consecutive flight where the vehicle did not get through its full objective list, and the objectives that keep slipping are the ones on the critical path.

The schedule behind it

Gwynne Shotwell said in June that Flight 14 might attempt an orbital launch, depending on how Flight 13 went. SpaceX is targeting a monthly cadence. Neither of those is now on firmer ground than it was a week ago.

The binding constraint sits further out. NASA needs Starship ready by 2027 for an uncrewed demonstration proving the vehicle can perform a lunar landing, and a modified Starship V3 with a docking adaptor is manifested for Artemis 3. Between here and there sits the hardest unsolved item in the program — orbital propellant transfer, cryogenic, at scale, requiring multiple tanker flights per mission. That capability has not been demonstrated. It cannot be demonstrated until Starship reliably reaches orbit and reliably relights in space, and as of July 16 the vehicle has done neither on a flight where everything went to plan.

This is the arithmetic that makes an ignition abort more than a bad afternoon. Starship's remaining milestones are strictly sequential — orbit, then relight, then rendezvous, then transfer, then a lunar demo — and each one needs multiple flights to move from achieved once to achieved reliably. A monthly cadence with clean flights makes 2027 plausible. A monthly cadence where flights burn their attempts on the first two items does not.

What it doesn't mean

Flight 13 will fly. Two Raptors is days of work, both stages are healthy, and SpaceX aborts at ignition often enough that treating each one as a setback misreads how the company operates. The hold-down clamps and the automatic abort are not the failure — they are the reason there is a vehicle left to fly.

The honest read is narrower: the vehicle is fine, the pad is fine, and the schedule has less slack in it than it did on July 15. Starship is being asked to prove three capabilities it has not proven, on a cadence it has not sustained, against a deadline set by a lunar program that has no alternative lander. Every attempt that ends before liftoff spends margin that was already committed elsewhere.

The rocket is not the problem. The calendar is.