Cracks in the International Space Station are causing air leaks – how much longer can it remain habitable?

On June 5, 2026, Nasa ordered five astronauts aboard the International Space Station (ISS) to shelter in a docked spacecraft and prepare to abandon ship. The reason was a longstanding, but worsening, air leak in the Russian part of the station.

An hour and a half later, the alert was lifted, allowing the crew to resume their work. But the episode reminds us that humanity’s most expensive science experiment – and a remarkable symbol of international cooperation – is showing its age.

The ISS was born from a thaw in relations between Washington and Moscow immediately following the cold war. In the early 1990s, the countries decided to merge separate space station projects that each was working on.

The air leak, in an older section of the ISS known as the Zvezda PrK transfer tunnel, is a dose of engineering reality. It is caused by fine cracks in the structure of the tunnel, which provides access to a spacecraft docking port. There is no permanent fix and the cracks have already been patched and repatched using a sealant.

Nasa and Russia’s space agency Roscosmos dispute the seriousness of the problem. Roscosmos says the slow leak poses no danger, while the US space agency views it as an elevated safety risk. Not only is the structure under strain, but also the fabric of the agreement that keeps the ISS running.

In 2024, the ISS advisory committee’s chair Bob Cabana said: “Nasa has expressed concerns about the structural integrity of the PrK and the possibility of a catastrophic failure.”

In early June 2026, new cracks appeared and leak rates rose. This prompted Roscosmos to propose a fix. According to a report in the technology magazine Ars Technica, the Russians wanted to carry out repairs to the hull using a drill.

Nasa balked at the plan and Roscosmos subsequently dropped it, only to come up with a new one. Under this second proposal, Russian cosmonauts would use a saw to remove a load-bearing bracket in the tunnel.

When Nasa heard this, agency officials ordered the five astronauts to shelter in SpaceX’s Crew Dragon capsule, ready to detach from the space station should an accident occur.

Roscosmos ditched the second plan too, prompting Nasa to call the astronauts back onboard. Since the incident, Roscosmos has told Nasa that it will close off the PrK tunnel from the rest of the station, in a bid to draw a line under the problem.

A commercial future

The ISS was never meant to be immortal. The plan was always to let a competitive commercial market take over with privately built space stations. Nasa would then send its astronauts to these new orbiting outposts, which would also be frequented by private space travellers.

The current frontrunner to reach low Earth orbit is Vast’s Haven-1. This has roughly an eighth of the ISS’s living space and is built for short stays of up to a month. It also relies heavily on a docked SpaceX capsule for air and power. Haven-1 is a testbed, not a like-for-like replacement, and despite optimistic assessments of launching in 2026, it is now unlikely to fly before 2027.

The stations that could fully replace the ISS are further away. Vast’s modular Haven-2 is aiming for a first module in 2028 and is targeted for completion in 2032, the very year the ISS is now due to leave service. This leaves no margin for the delays these programmes routinely suffer.

Two other companies, Axiom and Starlab Space, also plan large orbiting outposts. However, these are still in development, and Axiom recently encountered financial trouble.

The problem is that the companies, and those who finance them, are not ready to take the step up. The extent to which ISS replacements will be market-led has been consistently overstated. Government space agencies are going to be the main tenant and the paymaster, which makes this a commercial sector that the US is paying to bring into being.

Faced with less-than-ideal private sector engagement and delays, US lawmakers have extended the life of the ISS. Nasa had been due to operate the ISS until 2030. Recent legislation, which is awaiting approval, would delay ISS decommissioning until 2032.

This is, in part, a response to Nasa’s own stalled procurement of commercial replacements for the ISS. The recent legislation recognises this, tying any retirement of the ISS to replacements being ready and warning against a scenario where China is the only country with a continued human presence in low Earth orbit.

How the ISS ends

And there is a messy legal landscape hanging over the ISS as it approaches the end of its life. In order to decommission the outpost, Nasa will push the 420-tonne space station into Earth’s atmosphere using an adapted SpaceX Dragon vehicle – at a cost of roughly US$840 million (£638 million).

This controlled re-entry will take place over Point Nemo, the remotest part of the Pacific Ocean. This greatly reduces the risk of debris landing on populated areas.

Yet the space station will be the biggest orbiting object ever sent through the atmosphere and pieces as large as a small family car could survive the descent. If something were to go wrong, who would be liable?

Under the United Nations Liability Convention, a treaty from 1972, the nation that launches a space object is liable for any damage it causes. But the ISS was built from modules launched by more than one country, principally the US and Russia.

Where two or more states launch together, they are jointly and severally liable, the latter term meaning that any one of them can be pursued for the whole of the damage. The ISS partners comprise the US, Russia, Japan, Canada and participating member states of the European Space Agency.

If a piece of the station were to land where it shouldn’t, causing damage, the liability is absolute – no fault need be proved. If the de-orbiting process were to damage a satellite in orbit, liability is dependent on fault. Proving fault in a scenario as complex as ISS decommissioning could be extremely difficult.

The ISS air leak is a reminder that space stations will require continual maintenance. Private companies will not be able to get away with a “sell and forget” mindset when they launch their outposts.

With no obvious paying customers beyond space agencies lining up, investors are understandably reluctant to rush into expensive commitments. Extending the life of the ISS provides a temporary patch, but it does not remove the fundamental problem of who pays to replace it.