A Hydraulic Pin, a $75 Billion IPO, and the Biggest Rocket Ever Built: Inside SpaceX's Starship V3 Scrub

At 6:40 p.m. CDT on May 21, 2026, SpaceX's launch commentator Dan Huot announced the scrub of Starship Flight 12 — the debut of the company's third-generation Starship, the tallest and most powerful rocket ever assembled. The cause was not a Raptor engine fault, not a structural failure, not an avionics glitch. It was a hydraulic pin on the launch tower arm that refused to retract [1][6].

The anticlimax was fitting for a program that has oscillated between spectacular explosions and remarkable recoveries over 12 flight attempts in three years. But the timing of this particular scrub — three weeks before what is expected to be the largest initial public offering in history — adds a financial dimension that no previous Starship delay has carried [4][8].

What Went Wrong at T-Minus 40 Seconds

The automated countdown sequence reached its built-in hold at T-40 seconds and stalled. SpaceX recycled the count multiple times, troubleshooting issues with both the water deluge system beneath the launch mount and the tower arm's hydraulic retraction mechanism [6][5]. The water diverter — the system that floods the launch pad with thousands of gallons per second to suppress acoustic and thermal damage — triggered an initial hold. After that was resolved, the hydraulic pin that locks the launch tower's mechanical arm in place failed to disengage [1][6].

Elon Musk posted on social media that the pin mechanism needed overnight repair. SpaceX converted the aborted countdown into a wet dress rehearsal, successfully loading the full 11.5 million pounds (roughly 5,000 metric tons) of liquid methane and liquid oxygen propellant into the V3 stack — itself a milestone for the new vehicle [1][5]. A second attempt was scheduled for 5:30 p.m. CDT on May 22 [5].

The scrub was caused by ground support equipment, not the rocket itself. That distinction matters. The vehicle completed its propellant loading without anomaly. The Raptor 3 engines were not implicated. But ground infrastructure failures are not trivial — the launch tower and its "chopstick" arms are integral to SpaceX's reusability architecture, and any unreliability in those systems could limit launch cadence as much as a vehicle defect would [6].

How This Compares to Prior Starship Setbacks

Starship's flight test record is a study in rapid-iteration engineering, with dramatic failures punctuating genuine progress.

Flight 9, launched May 27, 2025, ended with the loss of both the Super Heavy booster and the upper stage. SpaceX traced the upper-stage failure to a damaged pressurization system diffuser that was not caught during preflight analysis, which caused rising methane levels in the nosecone and cascading attitude errors [9]. The booster descended at an intentionally steep angle to gather performance data, but one of 13 center engines shut down during the landing burn, and the vehicle disintegrated at roughly one kilometer altitude due to extreme stress on internal fuel transfer tubes [9].

Flight 10, after an initial scrub on August 24, 2025, and a second abort on August 25, finally launched successfully on August 26. It achieved SpaceX's first payload deployment from Starship — eight Starlink simulators — and completed a soft splashdown in the Indian Ocean [10]. Flight 11, on October 13, 2025, was the final V2 flight, completing aggressive reentry banking maneuvers and a 13-engine landing burn demonstration [11].

The Flight 12 scrub is categorically different from the Flight 9 failure. A ground equipment hydraulic issue does not indicate a vehicle design flaw. But it does underscore the complexity of the ground systems that V3's larger, heavier vehicle places new demands on.

Source: SpaceX / Spaceflight Now

Data as of May 21, 2026CSV

What Makes V3 Different

Starship V3 is not an incremental update. It is a comprehensive redesign across both stages — Super Heavy booster and the Starship upper stage — incorporating SpaceX's new Raptor 3 engine and significant structural changes [2][12].

The key engineering changes:

Thrust. V3's total liftoff thrust reaches approximately 10,000 tons-force (roughly 18 million pounds), a 21% increase over V2. This comes from Raptor 3 engines generating about 280 tonnes of thrust each at sea level, compared to roughly 230 tonnes for the Raptor 2 [2][12].

Propellant and structure. The Super Heavy booster's propellant capacity increased from 3,650 to 4,050 metric tons, and the upper stage from 1,500 to 1,600 metric tons. A redesigned fuel transfer tube — now "roughly the size of a Falcon 9 first stage," according to SpaceX — enables faster simultaneous ignition of all 33 Raptor 3 engines [2]. The full stack stands 407-408 feet (124 meters) tall, about 5 feet taller than V2 [5][12].

Reusability architecture. V3 is designed to carry up to 100 metric tons to low Earth orbit with full reusability — nearly triple the payload of V2 in expendable mode. The vehicle incorporates lessons from Flight 10 and 11's splashdown tests, with reinforced heat shielding and landing hardware designed for tower catch operations [2][12].

Independent aerospace analysis has been cautiously positive but not uncritical. Aging Aircraft Solutions, an aerospace consulting firm, noted that "the repetition of similar upper-stage failures" in earlier flights raises questions about whether corrective measures have been thorough enough [13]. SpaceX's own testing encountered a setback in November 2025, when a V3 Super Heavy booster suffered a sidewall rupture during gas system pressure testing — a failure attributed to a potentially malfunctioning valve or structural weakness [13]. The company has not publicly disclosed whether that specific failure mode was fully resolved before clearing Flight 12 for launch.

The $15 Billion Rocket and the $75 Billion IPO

SpaceX has invested over $15 billion in Starship development, including $3 billion in 2025 alone and an additional $930 million in the first quarter of 2026 [4][12]. The NASA Human Landing System contract, worth $4 billion in its current form ($2.89 billion for Artemis III plus $1.15 billion for Artemis IV), has paid out approximately $2.6 billion to date — about 65% of the award [14][15]. SpaceX says it is self-funding "over 90% of system costs" [14].

Source: NPR / SpaceX S-1 filings

Data as of May 21, 2026CSV

The financial stakes around Flight 12 are unlike any previous Starship test. SpaceX filed a confidential draft registration statement with the SEC on April 1, 2026, for what is expected to be a $75 billion raise at a $1.75 trillion valuation — the largest IPO in history [3][8]. The prospectus could become public as early as this week, with the investor roadshow beginning around June 4, pricing as early as June 11, and trading potentially starting June 12 on the Nasdaq under the ticker SPCX [3][8].

SpaceX made $18.6 billion in revenue in 2025, a 33% increase from the prior year, driven overwhelmingly by Starlink [4]. But it reported a $4.9 billion loss as capital expenditures nearly doubled to $20.7 billion [4]. Its launch business alone operated at a $662 million loss in Q1 2026 [4].

Franco Granda, a PitchBook researcher, told NPR the Starship test is "super important for the IPO," warning that a failed launch could cause investor enthusiasm to "diminish quite dramatically" [4]. SpaceX acknowledged in its own filings that "any failure or delay in the development of Starship...would delay or limit our ability to execute our growth strategy" [4]. Analyst Tim Farrar cautioned that justifying valuations above $1 trillion requires believing Musk will achieve "something much bigger" than current operations [4].

The timeline overlap between launch attempts and capital-raise activities is documented but not inherently problematic from a securities perspective. SpaceX's S-1 filing must include risk factors covering launch cadence, regulatory risk, and the Starship program's contingencies [8]. The SEC requires a public S-1 filing at least 15 days before the roadshow begins [8]. Whether the proximity of a high-profile launch to investor marketing constitutes a disclosure concern depends on what the S-1 says about the V3 program's readiness — and whether the scrub itself is treated as a material event. No former SEC officials or securities lawyers have publicly flagged specific concerns about this timeline as of May 22, 2026, though the S-1 has not yet been made public [8].

Launch Windows and Schedule Pressure

SpaceX's FCC communications license for Flight 12 is valid through October 2026, providing roughly 131 days of remaining window [16]. The company has already filed Flight 13 authorization with a May 29 launch window, suggesting confidence in rapid turnaround [16]. The Flight 12 scrub, caused by ground equipment rather than a vehicle defect, should not require an extended stand-down.

However, the broader schedule is governed by more than just hardware readiness. NASA's Artemis program depends on Starship achieving orbital refueling capability — a technology that has never been demonstrated at the scale required for lunar missions. Each delay compounds pressure on an already-slipping timeline.

How Starship's Record Compares

Measuring Starship against historical rocket development programs requires acknowledging fundamentally different approaches to testing.

Saturn V flew 13 times between 1967 and 1973 with zero total failures — but its first flight came after years of component-level testing on Saturn I and IB, and each Saturn V mission cost billions in today's dollars [17]. Ariane 5's first flight in 1996 self-destructed 37 seconds after launch due to a software error inherited from Ariane 4; its second flight in 1997 was also a partial failure; the third flight in 1998 finally succeeded [18]. Blue Origin's New Glenn reached orbit on its first attempt in January 2025, though its first-stage landing failed; the second flight in November 2025 achieved a successful booster landing [19].

Starship reached what SpaceX considered a full success on its fifth flight in October 2024, after two total failures and two partial successes [10][11]. By that measure, it took more flights than Saturn V or New Glenn but fewer than Ariane 5 (adjusting for the fact that Ariane 5's failures destroyed payloads while early Starship flights were uncrewed tests by design).

Source: NASA / ESA / SpaceX / Blue Origin

Data as of May 21, 2026CSV

SpaceX's "test, fail, fix, fly" philosophy makes direct comparison difficult. The company accepts — even plans for — vehicle losses in pursuit of faster iteration. That approach has produced 11 flights in three years, a cadence that traditional programs could not match. But it has also meant that Flights 8 and 9 in 2025 both ended in vehicle destruction, coming after a stretch of three consecutive successes [9][10].

NASA's Artemis Fallback

NASA's Artemis III mission has been delayed repeatedly. As of February 2026, NASA administrator Jared Isaacman confirmed a revised plan: Artemis III will now be a crewed test in Earth orbit, docking the Orion capsule with one or both lunar landers from SpaceX and Blue Origin, rather than a crewed lunar landing [20]. The first crewed lunar landing has been pushed to Artemis IV, tentatively targeting 2028 [20].

This restructuring gives SpaceX more time but does not eliminate the pressure. The $4 billion HLS contract is a fixed-price award, meaning SpaceX bears cost overruns — NASA does not pay more if Starship development takes longer [14][15]. NASA has also opened competition for an alternative Artemis III lander, with Blue Origin's Blue Moon Mark 1 identified as a frontrunner. Blue Moon is a simpler, smaller vehicle that does not require orbital refueling, reducing its technical risk at the cost of payload capacity [20].

If Starship fails to achieve full and rapid reusability on NASA's timeline, the contractual fallback is straightforward: SpaceX absorbs the development cost, and NASA shifts landing missions to whichever lander is ready first. The agency's Office of Inspector General issued a report in March 2026 on HLS contract management, scrutinizing milestone payments and schedule risks [14].

Boca Chica, the FAA, and the Neighbors

SpaceX's regulatory treatment at Boca Chica has been a sustained point of contention. The FAA completed a Programmatic Environmental Assessment (PEA) rather than a full Environmental Impact Statement (EIS) for Starship operations — a decision that environmental groups challenged in court. A federal judge dismissed the lawsuit in September 2025, ruling that the FAA's approach met legal requirements [21].

The distinction matters: a PEA is a shorter, less exhaustive review than an EIS. Critics argue that the scale of Starship operations — including the world's largest rocket launching from a site adjacent to the Boca Chica Wildlife Refuge and state parkland — warranted the more thorough process [21][22]. Defenders of the FAA's approach point out that the PEA included mitigation measures that are binding conditions of SpaceX's launch license, and that the agency retained the authority to require an EIS if conditions change [22].

Whether this differs from historical precedent is debatable. The FAA has used programmatic environmental assessments for other novel launch vehicles, though none at Starship's scale. The 2024 decision to begin preparing an EIS for evaluating up to 44 launches per year at Kennedy Space Center's LC-39A suggests the agency itself recognizes that higher cadence may require more rigorous review [7].

On property compensation, SpaceX's buyout offers to Boca Chica residents have been a source of friction since the company began acquiring land. Some homeowners reported receiving offers they considered well below fair market value — one cited a $51,900 offer for a coastal property [23]. Residents, many of them retirees on fixed incomes, argued the offers were insufficient to cover relocation costs [23]. SpaceX reportedly considered but ultimately did not pursue eminent domain to acquire remaining properties [23]. Whether these offers constituted "legally mandated compensation at fair-market rates" is contested — the residents say no; SpaceX has not publicly addressed the specific valuations.

A Worker's Death and a Broader Safety Record

The Flight 12 scrub came just days after a more somber event at Starbase. On May 16, 2026, Jose Luis Bautista, a 25-year-old contracted worker from Donna, Texas, died after falling eight feet from scaffolding at the facility. He was pronounced dead at 5:15 a.m. from blunt force trauma [24]. OSHA confirmed it is investigating the incident [24].

The death followed a pattern of safety concerns. In 2025, OSHA issued SpaceX seven "serious" safety citations, including allegations that an improperly inspected crane was put back into service, resulting in $115,850 in penalties [24]. The Scientific American reported that the V3 launch was initially postponed from its earlier target in part because of the investigation [12].

What Happens Next

If overnight repairs to the hydraulic pin mechanism succeed, Flight 12 could launch on May 22. The mission profile calls for Booster 19 to splash down in the Gulf of Mexico roughly seven minutes after liftoff — no tower catch attempt for this first V3 flight. Ship 39 would continue to orbit, deploying 20 Starlink simulators and two operational modified Starlinks equipped to gather heat shield imaging data, before attempting a controlled splashdown in the Indian Ocean more than 60 minutes later [5].

A Raptor engine relight is planned at approximately 39 minutes into flight — a key demonstration for the orbital refueling capability that NASA's lunar missions require [5].

The stakes are concrete. For SpaceX's engineers, Flight 12 validates or exposes the V3 redesign. For NASA, it signals whether Starship is on track for Artemis milestones. For investors weighing a $1.75 trillion valuation, it is the most visible data point available before June's roadshow. And for the residents of Boca Chica and the workers at Starbase, it is a reminder that the world's most ambitious rocket program operates in their backyard — with all the risks and disruptions that entails.