SpaceX Starship 40: Why Ocean Recovery Proved Harder Than Landing

What Happened

SpaceX's 13th integrated Starship flight test launched from Starbase in South Texas on July 24, after two earlier scrubbed attempts — one an engine abort at T-0 when four of Super Heavy's 33 Raptor engines failed to start, and one a weather-related scrub caused by cloud cover obscuring views of the heat shield during ascent 211. When the rocket finally flew, the upper stage, known as Ship 40, performed close to flawlessly: it deployed 20 Starlink V3 satellites, completed an in-space engine relight, survived reentry, and executed a controlled flip-and-burn landing in the Indian Ocean about an hour after liftoff 91119. The Super Heavy booster, Booster 20, did not fare as well — only 10 of 13 intended engines ignited for its landing burn, two more shut down, and the booster hit the Gulf of Mexico at high speed and was destroyed 911.

What made Ship 40 remarkable was not just that it landed, but that it stayed in one piece. SpaceX called it the program's "softest" splashdown yet, and a company spokesperson described it on the livestream as a "dream scenario" 91213. Rather than breaking apart or sinking as every prior prototype had, Ship 40 tipped over gently and kept floating, airtight, transmitting telemetry and video for more than 100 hours as it drifted over 400 kilometers across the ocean 1415.

The Recovery Nobody Planned For

SpaceX had not designed Flight 13 around retrieving hardware — prior vehicles were expected to be destroyed on impact, so the mission was built to gather data, not salvage a spacecraft 14. Ship 40's intact, buoyant landing changed that calculus. On July 28, Elon Musk confirmed SpaceX would send a ship to recover the vehicle, launching what USA Today called the company's largest rescue operation to date 8912.

The operation quickly became a multi-vessel affair. SpaceX's support ship Go Australis was joined by Norwegian tugboats Normand Ranger and Skimmer Tide, and satellite imagery from geospatial firm Vantor showed the fleet working to attach tow lines to the 171-foot spacecraft 12. Tracking data cited by CNN showed the tugs moving at roughly one knot while broadcasting restricted maneuverability — a sign of just how unwieldy the tow had become 12.

Conditions worsened from there. On August 7, SpaceX said its team faced "challenging conditions and increasingly rough seas," and Musk admitted publicly that "recovery is not looking good right now" 91320. Multiple outlets treated that statement as the low point of the saga: the Caller-Times framed the episode as the ship possibly being "lost to the sea," and Fox 35 Orlando quoted a Florida Tech spaceport expert warning that if the vehicle couldn't be brought ashore, SpaceX might have to sink it to keep sensitive hardware out of reach of competitors 1320.

Despite that, the team pushed through. After what SpaceX described as 24 days at sea, Ship 40 reached calmer waters off Christmas Island, an Australian territory roughly 840 nautical miles northwest of the mainland, on August 18 6916. Space.com characterized the episode as a 24-day "ocean ordeal," a framing that captures both the achievement and the difficulty involved 16.

From Floating Hulk to Cargo

Getting Ship 40 out of the water required equipment more associated with heavy marine salvage than spaceflight. The Normand Ranger held the spacecraft in position until the 710-foot semi-submersible transport vessel Forte arrived, partially submerged its deck, and allowed the rocket to float into position using guide ropes before being refloated with its cargo secured 101718. Engineers used the calmer water near Christmas Island to photograph the heat shield and engines up close and, according to later reporting, collected physical heat-shield tile samples for analysis 17.

By late August, Ship 40 was loaded aboard Forte for the journey back to Starbase — a trip SpaceX and multiple outlets put at roughly 12,400 miles, expected to take several months given the transport vessel's cargo-carrying speed 561718. The vehicle is considered unlikely to fly again given its prolonged saltwater exposure; its value now is forensic, feeding data into future heat-shield and structural improvements 91317.

Where the reporting agrees

Across outlets, the factual skeleton of the story is remarkably consistent. All confirm the July 24 launch date, the 13th-flight-test designation, Ship 40's intact splashdown as a first for the program, and the sequence from Musk's July 28 recovery announcement through the August 7 warning that things looked grim, to the August 18 arrival at Christmas Island 68912131620. The 171-foot length of the upper stage, the involvement of Go Australis, Normand Ranger and Skimmer Tide, and the eventual use of the semi-submersible Forte for transport are corroborated across CNN, NASASpaceflight, BGR, itechpost and basenor 1012151718. There's also broad agreement on why this matters: SpaceX's ultimate goal is a tower catch of the upper stage, not an ocean rescue, and Ship 40's value lies in what its heat shield and engines can teach engineers rather than in the vehicle flying again 4791012.

Where it doesn't

The divergences are mostly matters of framing and specificity rather than contradiction. USA Today and the Caller-Times both lean into the drama of Musk's "not looking good" comment, with the Caller-Times headline going as far as suggesting the ship "may be lost to the sea" — a characterization that, in hindsight, oversold the risk relative to the eventual outcome 913. Fox 35 Orlando is the only outlet to raise the possibility that SpaceX might have deliberately sunk the vehicle to prevent it falling into competitors' hands, a claim attributed to a Florida Tech spaceport expert rather than to SpaceX itself 20. NASASpaceflight is alone in offering a granular, engineering-level account of the launch attempts — the specific engine-abort and turbopump-moisture issues on the first try, and the exact engine counts during Booster 20's failed landing burn — detail no general-audience outlet matches 11. BGR and itechpost diverge slightly on distance framing: BGR emphasizes Ship 40's roughly 1,000-mile drift to Christmas Island and a 12,400-mile total return voyage, while itechpost frames the same 12,400-mile figure around the Forte's cargo speed, without meaningfully conflicting on the numbers themselves 1018. Financial-outlet coverage (Barron's, Seeking Alpha, Finbold) sits apart from the recovery narrative entirely, treating Starship reusability as a stock-moving engineering bottleneck for SpaceX's private valuation rather than as a story about one spacecraft's ocean journey 347.

The reading the evidence supports

Taken together, the coverage supports a single, coherent conclusion rather than two competing ones: Ship 40's flight was an unambiguous technical success, and its recovery was an unplanned, genuinely difficult logistical improvisation — not a crisis that nearly failed, but not a smooth victory lap either. The moment-to-moment reporting that treated Musk's August 7 comment as near-catastrophic reflects an accurate real-time snapshot of the operation's uncertainty, not an error, but the eventual 24-day resolution shows that uncertainty was more about the unpredictability of open-ocean salvage than about any flaw in the spacecraft itself. NASASpaceflight's more technical framing — that Ship 40 succeeded while Booster 20's Block 3 design still needs work — is the most useful lens for understanding why SpaceX is treating this recovery as a data-gathering bonus rather than a step toward reflying this particular vehicle. The throughline every outlet converges on, whether financial analysts assessing reusability's effect on launch costs or space reporters tracking tugboats by satellite, is that surviving splashdown and being economically recoverable are two different engineering problems, and Ship 40 has only solved the first one.