Plasma Tunnels Show How Satellites Really Burn Up
28 Jul 2026
The problem: nobody fully knows what happens to a satellite when it dies
Every day, according to the European Space Agency's Space Environment Report, three or more large satellites or spent rocket stages burn up in Earth's atmosphere. Multiply that out and hundreds of tons of human-made hardware evaporate in the sky annually. With roughly 18,000 operational and defunct satellites currently in orbit as of June — and most satellites designed for a five-year replacement cycle — the volume of objects reentering the atmosphere is only set to grow.
The assumption behind most satellite designs is simple: build it so it fully melts on the way down. But a team of researchers in Germany is now testing that assumption directly, using high-powered plasma wind tunnels to recreate the extreme conditions of atmospheric reentry and observe, piece by piece, what actually survives.
How the test works
Reentering satellites plunge through the thickening upper atmosphere at around 8 km/s, generating friction heat that can exceed 1,600°C. Most of this destructive process happens between 60 and 80 kilometers of altitude.
To simulate it, the Stuttgart team's plasma wind tunnel runs at up to 6 megawatts of electrical power, with current reaching 2,000 amperes and plasma temperatures between 5,000°C and 8,000°C — hot enough to approximate real reentry conditions in a controlled lab setting.
In one test, a 100-gram cylinder of aluminum alloy 7075 melted away in large drops after about 6.5 minutes in the plasma flow, reaching 600°C before disintegrating. The results give researchers hard data on melt rates and failure points instead of relying purely on models.
What's actually at stake
The findings point to two distinct risk categories:
- What's left behind matters, not just what disappears. Aluminum oxide formed when satellites incinerate is known to contribute to ozone depletion and can reflect sunlight, potentially altering the thermal balance of Earth's upper atmosphere. As one researcher put it: "When a satellite demises, it does not disappear. It's still there in the atmosphere in the form of small particles of aluminum, and we don't know what these particles actually do in the atmosphere."
- Some parts don't melt at all. Components like titanium tanks and reaction wheels are unlikely to fully melt during reentry, meaning they can survive descent and pose a physical risk on the ground. This isn't hypothetical: in March 2024, a piece of a battery pallet jettisoned from the International Space Station pierced the roof of a Florida house.
Regulators have already set a bar for this risk. In Europe, satellite operators must design for no more than a 1-in-10,000 probability that any piece of a satellite survives reentry and reaches the ground. SpaceX, for its part, claims its Starlink satellites burn up completely — a claim plasma-tunnel testing could help independently verify.
Why founders should care
For founders building in satellite manufacturing, launch services, or space-adjacent hardware, this research carries several likely implications:
- Compliance testing may become more rigorous. If plasma wind tunnel data becomes a reference standard, satellite makers could face pressure to prove — not just model — that their hardware meets survival-probability thresholds like Europe's 1-in-10,000 rule.
- Materials choices could face new scrutiny. Components historically treated as "acceptable risk" — titanium tanks, reaction wheels — may need redesign if regulators tighten demisability requirements based on this kind of empirical testing.
- Atmospheric impact could become a regulatory front. The open question around aluminum oxide's effect on the ozone layer and atmospheric thermal balance suggests environmental impact, not just ground-strike risk, could plausibly enter future satellite design and licensing requirements.
- Liability exposure is a live issue. The 2024 Florida incident shows that surviving debris is not a purely theoretical risk — it's already resulted in property damage, and companies building large constellations should reasonably factor this into risk planning.
Given the scale — hundreds of tons of material reentering annually, with roughly 44 tonnes of space rock also entering the atmosphere per day, and thousands more satellites likely being launched — the gap between assumed and actual burn-up behavior is a problem that will probably compound rather than fade.
The caveat
This reporting relies on a single source, and the findings have not yet been corroborated elsewhere. Founders and operators should treat the specific melt-rate figures and atmospheric-impact claims as early-stage evidence rather than settled science until independent studies confirm them.