The three disposal paths
How Satellites Are Disposed Of: Deorbit, Graveyard, or Passivation
Down, up, or dead in place. Every satellite ends one of these three ways.
Every satellite that reaches end of life — whether from propellant exhaustion, component failure, or mission completion — must be disposed of to avoid becoming a collision hazard. The three principal strategies are controlled deorbit (lower the orbit until atmospheric drag causes reentry and destruction), graveyard orbit transfer (raise the orbit above the operational band to a disposal region), and passivation in place (remove all stored energy to prevent explosions, then leave the satellite where it is). The choice depends on altitude, propellant reserves, and regulatory requirements.
At LEO altitudes below ~600 km, atmospheric drag naturally deorbits a satellite within 5–25 years depending on altitude and ballistic coefficient. Above ~700 km, drag is negligible and debris persists for centuries to millennia. GEO satellites at 35,786 km cannot practically deorbit (the delta-v required is ~1,500 m/s); they are instead raised ~300 km to the graveyard belt. The disposal choice is dictated by physics: the atmosphere is the only free disposal mechanism, and it only works at low altitudes.
Controlled deorbit
Controlled Deorbit and Atmospheric Reentry
Lower the orbit, let drag do the work, burn up on the way down.
How it works. A satellite fires its thrusters to lower its perigee into the denser atmosphere (typically below 200 km). Aerodynamic drag then decelerates the satellite, converting orbital velocity into heat. Most satellite structures — aluminum, composites, solar panels, electronics — disintegrate between 70 and 90 km altitude. Some components (titanium tanks, reaction wheels, dense metal fittings) can survive to the surface. Controlled deorbit targets an uninhabited ocean area (typically the South Pacific Oceanic Uninhabited Area, or “spacecraft cemetery”) to minimize ground risk. Uncontrolled deorbit — where the satellite simply decays naturally — results in reentry at an unpredictable location.
Starlink’s approach. SpaceX operates its Starlink constellation below 600 km specifically so that failed satellites deorbit naturally within a few years. Active satellites use onboard krypton ion thrusters to maintain orbit; when decommissioned, the thrusters lower the orbit to accelerate reentry. SpaceX deorbits roughly 5–7 Starlink satellites per day as part of normal fleet management (replacing aging satellites with newer versions). The low operating altitude is a deliberate design choice: it makes the constellation self-cleaning, at the cost of higher drag and more frequent station-keeping maneuvers.
The ISS. The International Space Station requires periodic reboost to counteract drag at ~400 km. When the ISS is finally retired (planned 2031), it will undergo a controlled deorbit into the South Pacific, the largest controlled reentry in history. SpaceX holds a $843 million contract to build the US Deorbit Vehicle for this purpose.
Graveyard orbit
Graveyard Orbit: Disposal Above the Operational Belt
From geostationary orbit, the only practical disposal is up, not down.
A GEO satellite at 35,786 km cannot deorbit economically — the delta-v to lower perigee into the atmosphere is approximately 1,500 m/s, far more propellant than any GEO satellite carries. The alternative is a graveyard (or disposal) orbit roughly 300 km above the GEO belt (~36,086 km). The maneuver requires only ~11 m/s — about three months’ worth of station-keeping fuel. The IADC formula accounts for solar radiation pressure effects and the satellite’s area-to-mass ratio to ensure the retired satellite will not drift back into the operational belt.
After the orbit raise, the satellite is passivated: remaining propellant vented, pressure vessels relieved, batteries discharged and disconnected, transmitters shut down. The goal is to make the satellite as inert as possible so it cannot later explode and scatter debris. Hundreds of defunct satellites now occupy graveyard orbits. They will remain there essentially forever — at GEO altitude there is no atmospheric drag. This is the accepted trade-off: the graveyard region is vast and sparsely populated, so collision risk is negligible, but the debris accumulates permanently.
Compliance is imperfect. Not all GEO operators perform the full disposal sequence. Some run out of propellant before reaching the graveyard altitude. Some fail before the maneuver can be executed. Every GEO satellite that remains in the operational belt after end of life becomes a permanent uncontrolled hazard to the slot it occupies and its neighbors.
Passivation
Passivation: Making a Dead Satellite Safe
A dead satellite with stored energy is a bomb waiting to go off.
Passivation is the process of removing all stored energy from a spacecraft at end of life. It is required in addition to — not instead of — deorbit or graveyard transfer. The steps: vent remaining propellant and oxidizer through dedicated passivation valves; discharge batteries to a safe level (below ~50% state of charge) to prevent thermal runaway; release pressurant gas (helium, nitrogen) from tanks; and disable all transmitters to avoid spectrum interference.
The reason is historical. Dozens of satellites and upper stages have fragmented on orbit decades after launch due to residual propellant ignition, battery explosion, or pressure vessel rupture. These accidental breakups are a major contributor to the debris catalog — ESA documents over 660 fragmentation events. A passivated satellite is inert: it cannot explode, it generates no new debris (absent a collision), and it does not interfere with active spectrum assignments. An unpassivated satellite is an indefinite source of risk.
Active debris removal
Active Debris Removal: Cleaning Up What Was Left Behind
The only way to reduce debris in the 700–900 km band. No operational system exists.
NASA debris evolution models project that preventing a Kessler cascade in the sun-synchronous band (~800 km) requires actively removing 5–10 large objects per year — defunct satellites and spent rocket bodies with the highest collision probability. Natural decay at these altitudes takes centuries. The only solution is to go get them.
Astroscale (Japan/UK). The leading ADR company. ADRAS-J (Active Debris Removal by Astroscale-Japan) launched July 2024, rendezvoused with a defunct H-2A upper stage at 15 m distance, and began decommissioning in early 2026. ADRAS-J2 will return to grapple the same object and deorbit it (FY2028). Astroscale is also developing COSMIC — a UK Space Agency-funded mission to remove two inactive British satellites. Refueling demonstration planned for mid-2026 at GEO.
ClearSpace (Switzerland/ESA). ClearSpace-1 is an ESA-funded mission to capture and deorbit the PROBA-1 satellite using a four-armed robotic chaser. Planned launch 2028 on Vega C. Intended as a technology demonstration for commercial ADR services.
The economics problem. Active debris removal costs tens to hundreds of millions of dollars per object removed. There is no established business model — the debris has no owner willing to pay, and no international liability framework compels removal. Government contracts (ESA, JAXA, UK Space Agency) fund the demonstrations, but scaling to 5–10 removals per year at current costs would require hundreds of millions annually from public funding. An alternative model: “debris removal as insurance” — constellation operators pay into a removal fund proportional to their collision risk contribution.
The atmospheric problem
Satellite Reentry and Stratospheric Metal Contamination
Deorbiting a satellite does not make it disappear. It converts orbital aluminum into stratospheric aluminum oxide.
When a satellite reenters the atmosphere, it does not vanish. Its aluminum structure, solar panels, batteries, and electronics are vaporized at 70–90 km altitude, converting solid metal into aerosol nanoparticles (1–100 nm) of aluminum oxide (Al&sub2;O&sub3;), lithium compounds, copper, and over 20 other metallic elements. These particles are deposited in the mesosphere (50–85 km) and take up to 30 years to drift down to stratospheric altitudes (15–50 km), where 90% of Earth’s ozone resides.
The measurements. NOAA flights in 2022–2023, using laser mass spectrometry from a NASA WB-57 aircraft over Alaska at 19 km altitude, found that 10% of stratospheric sulfuric acid aerosol particles contained aluminum and other metals from spacecraft reentry. Over 20 elements were detected, including lithium, copper, lead, magnesium, titanium, beryllium, chromium, nickel, and zinc. Niobium and hafnium appeared in ratios matching the C-103 alloy used in rocket nozzle extensions. In 2022, reentering satellites deposited 41.7 metric tons of aluminum into the atmosphere — 29.5% more than the natural background from meteoritic material.
The chemistry. Aluminum oxide nanoparticles catalyze the destruction of stratospheric ozone. They are not consumed by the reaction — they trigger destructive chlorine-ozone reactions and persist for decades, continuing to destroy ozone molecule after molecule as they drift through the stratosphere. This is the same catalytic mechanism that made chlorofluorocarbons (CFCs) so destructive, and the reason the Montreal Protocol banned them. The difference: no treaty governs aluminum oxide from satellite reentry.
The February 2025 detection. Researchers at the Leibniz Institute for Atmospheric Physics in Kühlungsborn, Germany, published a landmark study in Communications Earth & Environment (February 2026) documenting the first direct observation of a reentry pollution plume — a lithium vapor spike detected by resonance fluorescence lidar on the night of February 20, 2025, coinciding with the uncontrolled reentry of a Falcon 9 upper stage. This was the first time a specific reentry event was tied to a detectable atmospheric contamination signature.
The million-satellite reentry scenario
What Happens When Tens of Thousands of Satellites Reenter Per Year
The disposal mechanism that solved the debris problem creates an atmospheric one.
Satellite operators chose low orbits and short lifetimes specifically to avoid the debris problem — if a satellite deorbits in five years, it cannot contribute to Kessler syndrome. This design philosophy drives the FCC’s 5-year rule, SpaceX’s sub-600 km operating altitude, and the entire LEO constellation architecture. It works for debris. It creates a different problem for the atmosphere.
Current scale. With existing planned constellations (Starlink, Amazon Leo, OneWeb, Guowang, Qianfan, and others totaling ~50,000 satellites), researchers estimate 912 metric tons of aluminum will reenter Earth’s atmosphere per year once full deployment and replacement cycles are running — equivalent to about 150 Space Shuttles vaporizing annually. A NOAA-CIRES study (2025) modeled 10,000 metric tons per year of alumina by 2040, consistent with mega-constellation growth projections. At that rate, a satellite burns up every one to two days from each major constellation.
The million-satellite scenario. SpaceX’s Starmind filing alone is for one million satellites. The AI1 satellite carries 110 m² of deployed liquid radiator panels and 75 m of solar array wingspan — significantly more aluminum per satellite than a Starlink V2. If the AI1 constellation operates with a 5–7 year replacement cycle, it would deorbit roughly 400–550 satellites per day, every day, indefinitely. Each reentry deposits its aluminum structure, radiator panels, and compute hardware into the upper atmosphere. At 2,100 kg per AI1 satellite (much of which is aluminum structure, solar panels, and radiators), a daily disposal rate of 400 satellites deposits roughly 840 metric tons of material per day into the mesosphere — over 300,000 metric tons per year. For comparison, the total annual meteoritic mass influx to Earth is approximately 50–100 metric tons per day.
The unknown. No one has modeled the atmospheric consequences of this scale of metallic aerosol injection. The NOAA 2025 study modeled 10,000 metric tons/year of alumina and found 1.5°C mesospheric heating, 10% polar vortex weakening, and ozone impacts. The million-satellite scenario is 30× larger. The effects on stratospheric chemistry, mesospheric temperature, polar vortex dynamics, noctilucent cloud formation, and long-term ozone trends are not characterized. The Leibniz Institute detection demonstrated that individual reentry events are now measurable; a continuous stream of hundreds per day would represent a sustained, anthropogenic modification of the upper atmosphere without precedent outside of volcanic eruptions.
The regulatory vacuum. No national or international regulatory framework governs the atmospheric effects of satellite reentry. The FCC regulates spectrum. The FAA regulates launch safety. The EPA does not have jurisdiction over the stratosphere. The Montreal Protocol governs manufactured ozone-depleting chemicals but was not designed for metallic aerosol catalysis from orbital disposal. The atmospheric contamination from satellite reentry falls in a governance gap between space regulation, atmospheric regulation, and environmental law — a gap that did not matter when a few satellites reentered per year and may matter a great deal when the number is hundreds per day.