Starship Flight 14: first orbit, first real payload

“Starship is orbital.” Two words from mission control that the whole program has been building towards since the first stack cleared the tower.

Last week I was looking ahead to Flight 14. It delivered. On Monday 28 September, US time (10:48 pm Monday night here in Sydney), Booster 21 and Ship 41 lifted off from Pad 2 at Starbase, and about twenty-five minutes later a single Raptor lit for nineteen seconds and pushed Starship into a stable orbit around the Earth for the first time. Then it opened the payload door and put 26 Starlink V3 satellites into space, the first time Starship has carried real, operational payload to orbit. After thirteen suborbital test flights, that is the moment the whole thing stops being a test article and starts being a launch vehicle.

What Marcus House walked through

Marcus House is one of my go-to channels for the day-after breakdown, and this one is worth the full twenty-three minutes. The points that stuck with me, in the order he covered them:

• The licence landed at the last minute. The FAA released the launch licence, with the revisions needed for the new orbital profile, only hours after his previous video went up, and the window opened early at 7:15 am local time.
• Scale. He puts the full stack at 124 metres and roughly 5,500 tonnes at liftoff, with 33 Raptor 3 engines estimated at around 250 tonnes of thrust each. His framing: new viewers are always floored by the scale first.
• Ascent closely matched Flight 13. Speeds and timings lined up with the previous flight. The booster lost one engine on the way up, in a spot where he says they have lost one before, but Super Heavy is built to fly through a few engine outs.
• Hot staging looked perfect again. Ship lit its six engines in the same staggered sequence as last time, which he reads as proof they have the Version 3 staging routine nailed.
• A much better booster splashdown. A second engine dropped out for boostback, and SpaceX said the boostback was deliberately designed to burn through the booster's remaining liquid oxygen to test its limits ahead of return-to-launch-site flights. For landing, the software swapped in two outer-ring engines to make up the numbers, then stepped down to the centre three. After Flight 13, where only 8 of 13 engines lit and the booster hit the water hard, this was the soft Version 3 splashdown they needed before another tower catch.
• Ship lost a vacuum Raptor, and pressed on. One Raptor Vacuum shut down at about T+7:18, near the end of the burn and already close to 20,000 km/h, so the ship simply burned a little longer to make up for it. It then coasted on the usual safe suborbital path while the team checked the sea-level engines, before the "go for orbit" and the single-engine insertion burn of about nineteen seconds. His point: Starship has been only seconds of extra thrust away from orbit on earlier flights.
• Deployment in daylight, and the camera satellites. The door opened on time and the Starlinks rolled out one after another, the first time we have seen deployment in daylight. He notes each V3 satellite adds around a terabit per second of capacity, this was also Starlink Group 31 batch 1, and the last three satellites carried cameras to inspect the ship's exterior and heat shield. The on-board view from a satellite drifting away down the length of Starship is, in his words, the best view of deployment we have ever seen.
• The first real coast in space. No Starship had spent even a full hour in vacuum before this. He walks through why that matters for propellant: without radiators or active cooling, the tanks have to be kept from boiling off too fast, or getting so cold that methane turns to slush. SpaceX kept it simple, holding the ship heat shield down with the steel side facing space.
• Coming home early was a pre-planned option. SpaceX had built two abort gates into the plan, and at the first one, around the 90-minute mark, they chose to come home early because of the engine that shut down during ascent. Even so, he says they collected about four times more coasting data than any previous flight across almost two orbits.
• First deorbit burn, and new heat shield experiments. An 11-second single-engine burn set up reentry towards a target north of Hawaii, which he says is a first for Starship. The ship flew heat shield upgrades based on Flight 13 data, including extra tile retention and a curved tile design to close off plasma flow paths, plus two tiles reflown from Ship 40, the Flight 13 ship recovered from the ocean.
• An on-target landing burn. All three sea-level engines lit, the flaps tucked in for the flip, and the ship went vertical before splashdown. It then broke up in the water, as expected and planned for, unlike the lucky survivor on Flight 13.
• What next. Hints point to Flight 15 as soon as October, and he floats the question everyone is asking: a first Version 3 booster catch, a first ship catch, or both?

What made Flight 14 different

Plenty of Starship flights have had a headline. This one had a stack of genuine firsts, straight from SpaceX's own mission summary and the reporting from Space.com and Spaceflight Now:

• First orbit. A dedicated orbital insertion burn, one sea-level Raptor for about nineteen seconds, into an orbit of roughly 275 km.
• First operational payload. 26 Starlink V3 satellites deployed over about half an hour, and the Starlink team made contact with every one of them. On Flight 13 the V3 satellites were on a suborbital path and fell back; these ones will raise their orbits and start serving customers.
• First deorbit burn. Every previous ship came home on a trajectory that did the job for it. This time Starship had to relight in orbit to bring itself down, and it did.
• Longest Starship flight yet. About three hours from liftoff to splashdown, ending on target in the northern Pacific in one of two landing zones cleared before launch.
• Engine-out resilience, on both stages. Booster 21 relit 31 of 33 engines for boostback and 11 of 13 for landing and still made a controlled splashdown in the Gulf, then SpaceX triggered its flight termination system as a demonstration of the safety hardware. Ship 41 lost a vacuum Raptor and still reached its planned trajectory by burning the other five for longer.

The photos

The photographers at Starbase went all out for this one. Rather than copy their work here, these are embedded straight from their own posts on X, so the credit (and the full-resolution image) stays with them. All of these are from Flight 14 launch day, 28 September at Starbase.

A word on the FUD

Every Starship flight comes with a wave of headlines and hot takes, and some of them spread faster than the facts. None of this needs anyone to be rude; it just needs the details. A few I have seen this week:

• "It exploded." After an on-target landing burn and a controlled splashdown, the ship tipped over and its leftover propellant ignited, which is what normally happens with ocean landings and what Marcus described as expected and planned for. No ship was being recovered on this flight. The catch comes later, at the tower.
• "They cut it short, so it failed." The primary goals were to reach orbit and deploy the 26 Starlinks into an operational orbit, and both were done. Coming home early was one of two abort gates planned before launch, and SpaceX chose it out of caution after the ascent engine issue, landing in a zone cleared in advance. That is disciplined flight control, not a failure.
• "Engines keep going out, so it is unreliable." Engine-out capability is designed in, and this flight proved it: the booster still made a controlled splashdown with engines missing, and the ship still hit its trajectory after losing a vacuum Raptor. Every engine out is data the team feeds into the next vehicle, which is exactly how this program has moved from Flight 1 to orbit.
• "It is all hype." Twenty-six satellites that answered the Starlink team from orbit are not hype. NASA Administrator Jared Isaacman called it "Gorgeous launch, getting Ship to orbit and managing every step in a safe, responsible, and especially inspirational way."

Healthy scepticism is good. Just point it at the primary sources: the SpaceX mission summary, the webcast replay, and the people who actually stand on the beach with cameras.

Next up

Starship has been to orbit, delivered a real payload, and brought itself back with its own deorbit burn. What still lies ahead is huge: a Version 3 booster catch, the first ship catch, ship reuse, orbital refuelling, and an uncrewed Moon landing. If the October hints are right, we might not wait long for the next column of fire. I will be watching, and Marcus will no doubt have the breakdown ready the next morning.