The Crowded Sky: Orbital Debris and the Economics of Low Earth Orbit

An object in low Earth orbit travels at roughly 7.5 kilometres per second. Two objects in different orbital planes can meet at relative speeds above 14 kilometres per second — around forty times the speed of a rifle bullet. At those velocities, mass matters less than you would expect and speed matters far more, because kinetic energy scales with the square of velocity.

A one-centimetre aluminium fragment, something you could hold between finger and thumb, carries kinetic energy in a hypervelocity impact comparable to a small car travelling at motorway speed. It will not dent a spacecraft; it will pass through it, and the impact vaporises material at the contact point. This single fact is the foundation of everything that follows.

Why orbits do not clear themselves

Whether debris persists depends almost entirely on altitude, because the only reliable removal mechanism is atmospheric drag.

The atmosphere does not stop at a boundary; its density declines roughly exponentially with height. Even at 400 kilometres, where crewed stations operate, enough residual gas remains to steadily rob orbital energy. Objects there re-enter within months to a few years unless actively reboosted. This is a genuine advantage of very low orbits: they are self-cleaning, and a failed satellite becomes a temporary problem rather than a permanent one.

Higher up, the picture changes sharply. Around 600 kilometres, decay takes years to decades. At 800 kilometres, centuries. Above 1,000 kilometres, drag is effectively negligible on any human timescale, and a fragment placed there stays for millennia. Unfortunately, the 700-to-900-kilometre band is highly desirable for Earth observation and certain communication constellations, because sun-synchronous orbits — which cross the equator at the same local solar time each pass, giving consistent lighting for imaging — sit in this range. The most useful shell is also the least forgiving.

Solar activity complicates prediction. During solar maximum, increased extreme-ultraviolet flux heats and expands the upper atmosphere, raising density at a given altitude and accelerating decay. Operators have learned this the hard way: satellites deployed at low altitude during a period of unexpectedly high solar activity have re-entered before reaching operational orbit.

Debris is not evenly distributed. It concentrates in specific altitude and inclination bands, and it is precisely the most commercially valuable bands that accumulate the most persistent junk.

The cascade argument

In 1978, Donald Kessler and Burton Cour-Palais published an analysis arguing that above a certain object density, collisions between existing objects would generate fragments faster than atmospheric drag removed them. Each collision produces thousands of trackable pieces and vastly more untrackable ones, each of which becomes a collision hazard, driving a self-sustaining increase in debris population.

Two points about this argument are routinely misrepresented.

First, it is not a sudden event. It is a change in the sign of a rate — debris generation exceeding debris removal — playing out over decades or centuries. There is no moment at which orbit "closes".

Second, it does not require new launches to proceed. Once the object population in a given shell is dense enough, the cascade continues from what is already there. That is why the relevant question is not only how many satellites are launched but how many derelict massive objects — spent upper stages, dead satellites — remain in high-persistence orbits. A handful of large intact objects contribute disproportionately, because a collision involving a multi-tonne rocket body produces far more fragments than one involving a small satellite.

Two events dominate the current debris environment. A 2007 anti-satellite weapon test destroyed a satellite at roughly 860 kilometres, generating over three thousand trackable fragments in one of the worst possible altitude bands. In 2009, a defunct communications satellite and an active one collided at about 790 kilometres, producing a comparable quantity. Between them, these two events account for a substantial share of all catalogued debris — and both occurred in orbits where the fragments will remain for centuries.

Tracking, and the limits of knowing

Space surveillance networks using ground radar and optical telescopes maintain catalogues of objects in orbit. Tens of thousands of objects larger than about ten centimetres are tracked. Estimates of the population between one and ten centimetres — large enough to destroy a satellite, too small to catalogue reliably — run to the high hundreds of thousands. Objects above one millimetre, capable of degrading solar panels, sensors and seals, number in the hundreds of millions.

Tracking accuracy is the operational bottleneck. A conjunction warning is a probabilistic statement: given positional uncertainty in both objects, what is the chance they pass within a lethal distance? Positional uncertainty for a catalogued object in low Earth orbit is often measured in hundreds of metres to kilometres along the direction of travel, driven by imperfect knowledge of atmospheric density, solar radiation pressure, and object attitude.

The consequence is a genuine dilemma. Operators typically manoeuvre when collision probability exceeds a threshold — commonly one in ten thousand. Because uncertainty is large, most manoeuvres are performed against encounters that would have missed anyway. Each manoeuvre consumes propellant, which is the finite resource determining mission lifetime, and interrupts service. Better tracking would reduce unnecessary manoeuvres far more effectively than better propulsion.

Coordination is the other half. When both objects are active and controlled by different operators, both must know who is moving and in which direction. There is no universal binding mechanism for this. Voluntary data sharing and commercial conjunction-assessment services fill part of the gap, but with tens of thousands of active satellites the volume of warnings is already large enough to require automation, and automated systems from different operators must agree on protocol or risk manoeuvring into each other.

What has actually improved

The situation is not uniformly deteriorating, and several genuine improvements deserve acknowledgement.

Post-mission disposal has become standard practice among responsible operators. Spacecraft reserve propellant to lower their orbit at end of life, targeting re-entry within a defined period. Regulators in several jurisdictions have shortened the expected window from twenty-five years to five, which materially reduces the time a derelict spends as a hazard.

Upper-stage passivation — venting residual propellant and discharging batteries after a stage's work is done — addresses a major historical source of debris. Many early fragmentations were not collisions but explosions of pressurised tanks or overcharged batteries in derelict stages, sometimes decades after launch.

Design for demise engineers spacecraft to burn up more completely during re-entry, reducing ground casualty risk. This is in tension with an emerging atmospheric-science concern: material vaporised during re-entry, including aluminium oxides, is being deposited in the stratosphere in measurable quantities, and the long-term effects on stratospheric chemistry are not well characterised. With re-entry rates rising steeply, this has moved from a theoretical question to one requiring measurement.

Active debris removal has moved from concept to demonstration. Magnetic capture, harpoons, nets, and robotic arms have all been tested in orbit or on representative targets. The technical difficulty is that most targets are tumbling, were never designed to be grabbed, and may have degraded surfaces. The economic difficulty is more fundamental: removal missions cost roughly what a satellite launch costs, and nobody is obliged to pay. Proposals to attach standardised capture interfaces to new spacecraft would make future removal far cheaper, and cost almost nothing to include at design time.

The commons structure

Strip away the engineering and this is a textbook common-pool resource problem.

The benefits of launching a satellite accrue to its operator. The costs of the debris and congestion it contributes are distributed across every present and future operator, including those who have not yet launched. Under those incentives, individually rational decisions aggregate into collectively poor outcomes — the standard result.

Existing governance is thin relative to the problem. The 1967 Outer Space Treaty establishes that states bear international responsibility for national activities in space and are liable for damage caused by their space objects, and that orbit cannot be appropriated by any state. Debris mitigation guidelines developed through international coordination bodies are technically sound but non-binding; they take effect only through national licensing regimes, which vary in stringency and create incentives for regulatory arbitrage.

Radio spectrum and orbital slot coordination happens through the International Telecommunication Union, which has a functioning allocation mechanism for geostationary positions but was not designed to manage physical congestion from tens of thousands of low-orbit satellites.

Proposals to correct the incentives include orbital-use fees scaled to the collision risk and persistence a spacecraft imposes, mandatory insurance or removal bonds, and stricter conditions on constellation licences. Economic modelling suggests risk-scaled fees could substantially raise the value of the orbital environment by discouraging the most damaging behaviour — but any such scheme requires coordination across launching states, and no forum currently has the authority to impose one.

What to watch

Several indicators track the environment's real trajectory better than headline satellite counts.

• Post-mission disposal compliance rates, published by space agencies. Compliance for large constellation operators in low orbits has improved; compliance for smaller operators and for upper stages remains uneven.
• The number of large derelict objects in the 700-to-900-kilometre band. These dominate cascade risk, and reducing them is the highest-leverage removal target.
• Conjunction manoeuvre frequency per active satellite. A rising rate signals congestion becoming operationally expensive even without collisions.
• Any further deliberate fragmentation event. A single destructive anti-satellite test can undo a decade of mitigation progress, which is why a growing number of states have declared moratoria on destructive testing.

Low Earth orbit is not full, and the framing of imminent closure is not supported. But it is a finite, shared, and slowly self-cleaning resource in which the most valuable regions clean themselves the slowest, and in which the consequences of carelessness are borne by everyone. That is a manageable situation — provided the management is designed before the physics decides for us.