Thousands of satellites orbit Earth, and every one of them will eventually stop working. What happens next matters more than most people realise, because a dead satellite left in a busy orbit becomes a collision hazard for decades. Satellite operators therefore plan an end-of-life disposal as part of the mission, not as an afterthought.

For satellites in low Earth orbit, the cleanest disposal is a controlled re-entry. The operator uses the satellite's remaining fuel to lower its orbit so it enters the atmosphere over an uninhabited ocean area, where it mostly burns up. Large structures such as space stations need careful planning for this, since some fragments can survive to the surface. The International Space Station's eventual disposal, for example, is planned as a controlled deorbit into a remote stretch of the Pacific Ocean. Smaller satellites may simply be left to decay naturally, since atmospheric drag at low altitudes will bring them down within a few years.

Satellites in geostationary orbit, about 36,000 kilometres up, are too far away to bring down cheaply. Instead, operators boost them a few hundred kilometres higher into a graveyard orbit, a sparsely used band where they will not interfere with working satellites. This requires reserving enough fuel for the final manoeuvre, which is why mission planners budget disposal fuel from the start. A satellite that runs out of fuel before it can move to the graveyard becomes a permanent obstacle in a valuable orbital slot.

Not every satellite gets a tidy ending. Some fail unexpectedly and cannot be moved, others break apart in collisions or explosions, and each breakup creates thousands of fragments. Those fragments travel at orbital speeds, where even a centimetre-sized piece can damage a working satellite. This is the space junk problem, and it is self-reinforcing: more debris means more collisions, which create more debris. Tracking networks monitor tens of thousands of objects, and operators manoeuvre working satellites to avoid predicted close approaches, sometimes several times a year.

Engineers are testing active debris removal ideas, from nets and harpoons to robotic arms and drag sails that pull dead satellites down faster. Regulators are also tightening rules, increasingly requiring operators to dispose of satellites within a few years of mission end rather than decades. The long-term goal is a sustainable orbit: satellites designed to be removed, missions planned with disposal in mind, and eventually, cleanup of the worst existing debris. Space is vast, but the orbits we use most are narrow highways, and keeping them clear is now part of the job.

A dead satellite remains the property and responsibility of its owner, even decades after it stops working. International guidelines, rather than binding law in most cases, ask operators to dispose of satellites responsibly, and national regulators increasingly attach disposal requirements to launch licences. That leaves a grey area for the oldest debris, launched before modern rules existed, where no one is clearly obliged to clean up. Proposals for debris removal must therefore navigate ownership and liability questions as well as engineering ones.

The practical consequence is that prevention matters more than cleanup for now. Modern missions are designed with disposal in mind from day one, including fuel reserves for a final manoeuvre and passivation, the safe venting of leftover fuel and discharge of batteries so a dead satellite cannot explode later. Each satellite disposed of properly is one fewer collision risk for the satellites that follow, and the industry's track record is improving as these practices become standard.