Why NASA Keeps Using Notoriously Leaky Hydrogen for Artemis II

Lead: In early February, a planned wet dress rehearsal at Kennedy Space Center for Artemis II was halted after teams detected repeated leaks of super-chilled liquid hydrogen, forcing engineers to stop fueling several times and leaving NASA unable to complete the test. The delay pushed back a mission that would carry four astronauts on a roughly 10-day lunar flyby and prompted more than a week of inspections and repairs. The leak pattern echoed problems seen during the 2022 Artemis I campaign and earlier Space Shuttle operations, renewing scrutiny of why NASA relies on a fuel that routinely challenges ground crews. Agency engineers and outside experts say the answer mixes performance advantages, legacy decisions and difficult material limits.

Key takeaways

• Artemis II: The crewed lunar flyby originally planned for early 2026 was interrupted during an early-February wet dress rehearsal when liquid hydrogen leaks forced multiple halts in tanking.
• Mission profile: The planned Artemis II flight would carry four astronauts on about a 10-day circumlunar trajectory; the agency could not finish the full pre-launch fueling test and extended inspections for over a week.
• Leak metrics and hardware: NASA requires leak rates below 16% during fueling to proceed safely; recent leaks were traced to the Tail Service Mast Umbilical (TSMU), and technicians replaced seals around two propellant lines.
• Historic pattern: Similar hydrogen seepage delayed Artemis I in 2022 and has been a recurring challenge throughout the Space Shuttle program (1981–2011).
• Why hydrogen: Liquid hydrogen delivers the highest specific impulse of common chemical propellants — giving superior efficiency and payload capability, especially in vacuum — but is the lightest element and prone to seepage.
• Material limits: Cryogenic temperatures (about −423°F) create thermal stresses; seals (typically PTFE/Teflon) and adjoining structures change shape and can open microscopic leak paths.
• Policy roots: Part of SLS’s hydrogens usage stems from a congressional directive to leverage Shuttle-era hardware, workforces and contractors rather than redesign from scratch, a decision that shapes current performance and operations.
• Recent progress: A partial-fill test later in February showed fewer of the earlier leaks, indicating some fixes and operational techniques (seal warming, hardware replacements) are helping.

Background

Liquid hydrogen has been the high-performance propellant of choice for many deep-space stages since the mid-20th century and was used on the Apollo upper stages and the Space Shuttle’s main engines. Its appeal is simple and technical: hydrogen’s low molecular weight yields a very high specific impulse (Isp), meaning more thrust for each unit of propellant mass and better payload efficiency for missions beyond low Earth orbit. That efficiency becomes especially valuable for lunar missions where every kilogram saved translates into greater mission capability.

But hydrogen’s physical properties — it is roughly 14 times lighter than air and must be kept at roughly −423°F (≈20 K) — make it difficult to contain. Thermal contraction of pipes, seals and structural interfaces under repeated cooling cycles introduces mechanical complexity. Historically, the Shuttle era and many subsequent programs logged recurring work to locate and mitigate seepage, shaping institutional expertise but also persistent operational headaches.

Main event

During the early-February wet dress rehearsal for Artemis II, launch controllers observed repeated hydrogen leaks within hours of beginning tanking and halted the flow several times for safety. The anomalies centered on the Tail Service Mast Umbilical (TSMU), the three-story interface that routes propellant and services between the ground and the rocket. Because hydrogen is highly flammable in concentrated pockets, safety rules require conservative limits and multiple pausing points during fueling.

Technicians conducted inspections and removed and replaced seals around two propellant lines in the TSMU following the rehearsal. Engineers also used procedural remedies during tanking: briefly warming lines so seals could relax into position before being cooled again, and carefully monitoring leak-rate telemetry. NASA officials said the vehicle in question had not previously been exposed to cryogens, so teams were characterizing how that particular assembly ‘breathes’ and vents as temperatures drop.

A partial hydrogen-fill test later in February produced less of the earlier leakage behavior, agency leaders reported, suggesting some combination of hardware swaps and improved procedures reduced risk. Still, NASA emphasized the SLS remains an experimental, early-use vehicle and that crews should not yet regard it as ‘operational’ in the way recurring commercial launchers are defined.

Analysis & implications

Technically, the choice to use liquid hydrogen is a classic trade-off: exceptional in-space performance versus greater ground-level handling complexity. For missions that must maximize payload mass to deep space, hydrogen’s high specific impulse is often decisive. Replacing hydrogen with denser propellants (RP-1 kerosene or methane) on upper stages would reduce in-space efficiency and could require substantially larger booster designs or more launches to meet the same mission goals.

Policy and procurement choices amplified the operational consequences. A congressional requirement to reuse Shuttle-era hardware and sustain legacy workforces nudged SLS toward hydrogen architecture that closely mirrors Shuttle systems. That approach retained industrial capability and jobs but also preserved known design constraints and tooling tolerances that are susceptible to cryogenic leakage.

From a program-cost and schedule perspective, the reliance on legacy interfaces increases recurring troubleshooting and ground campaign complexity. An SLS that flies infrequently — with long gaps between flights — slows the engineering learning cycle and can keep subtle issues from being fully ironed out in flight operations. That leaves the agency dependent on methodical test campaigns and incremental fixes rather than rapid iteration.

Materials science remains a longer-term limiter. Current seal materials like PTFE are chosen because they historically perform best under extreme cold and hydrogen exposure, but for very large interfaces and repeated thermal cycling, the margin is small. New polymers, composite interfaces or active thermal management could reduce leaks, but implementing such changes mid-program is costly and time-consuming.

Comparison & data

Putting those numbers in context: an engine that achieves a higher Isp uses propellant more efficiently, which can reduce the mass of fuel required for the same delta-v. That efficiency advantage explains hydrogen’s continued use on upper stages for deep-space missions despite the handling burden at the pad.

Reactions & quotes

“Hydrogen tends to find its way out of things you want to try to contain it in.”

Adam Swanger, NASA cryogenics research engineer

Swanger framed the technical trade-off plainly: hydrogen’s low density drives performance but makes leakage more likely and harder to eliminate across large interfaces.

“Using Shuttle-era hardware shifted consequences and cost when it comes to trying to operate the rocket.”

Casey Dreier, Planetary Society (space policy expert)

Dreier emphasized the policy origins of the SLS architecture and argued that preserving legacy elements increased operational friction compared with a ground-up redesign.

“It’s an experimental vehicle…nobody sitting in one of these chairs needs to be calling any of these vehicles operational.”

Amit Kshatriya, NASA associate administrator

Kshatriya’s remark underlined NASA’s stance that early flights are part of a learning phase where anomalies are expected and studied.

Unconfirmed

• Whether the rocket rollout stresses definitively caused the most recent TSMU seal failures remains unproven; NASA has said it is one possible factor but has not confirmed a root cause.
• It is not yet confirmed that material changes alone could eliminate leaks for a rocket as large as the SLS without major redesign; that remains an engineering hypothesis needing testing.
• Projections of how many SLS flights will be required to fully characterize and stabilize leak behavior are still estimates and depend on future campaign cadence and data.

Bottom line

Liquid hydrogen will likely remain a core propellant for deep-space stages because its performance advantage is difficult to match with denser fuels. For Artemis-class missions that need maximum mass efficiency to reach lunar trajectories, hydrogen’s high specific impulse offers real mission value even when balanced against handling complexity.

At the same time, the operational friction around hydrogen — recurring leaks, labor-intensive mitigation and sensitivity to legacy hardware choices — means NASA will continue to invest in inspection, procedure refinement and selective hardware fixes. Whether those steps are sufficient to make SLS fueling routine will depend on how quickly the agency can accumulate operational experience and whether material or design changes become practical within program constraints.

Sources

• CNN — Artemis II coverage (news report)
• NASA (official agency statements and press materials)
• The Planetary Society (nonprofit analysis and space policy commentary)