Comment by wongarsu
5 days ago
> Or any satellite malfunction and failure to de-orbit in time
Last year they had one "dead as a doornail" Starlink satellite in space. [1] It's v1.5, so deployed sometime between 2021 and 2023. It should be naturally deorbited from atmospheric drag by now.
There was also the other Starlink satellite with a tank rupture last December [2]
A low number of dead satellites isn't an issue as the other satellites can steer around it. Their orbit also quickly decays to a level where it's below the orbital plane of the other satellites. The real danger is if a large enough number malfunction that they start colliding with each other at high speeds
1: https://starlink.com/public-files/Starlink_Approach_to_Satel...
2: https://www.space.com/space-exploration/satellites/a-spacex-...
A satellite of that magnitude can have many failure modes, onboard computer will do their best to de-orbit if it encounters unrecoverable failure: it can use thrusters, it can change attitude to increase surface area towards the sun or the atmosphere, rearrange solar panels etc etc. Assisted de-orbit it can be done very quickly. Unassisted de-orbit will take some time. This is a known and solved problem, we have been relying on this for many years. This is what you’re referring to with your links.
What I’m trying to communicate is that if s/c fails in a non-recoverable way, thruster stuck on, pierced propellant tank, adcs failing in a specific way (e.g. you can get unlucky and get particular bit-flipped that pass checksum etc etc), it is theoretically possible to end up in a non-recoverable state. For example: accelerate into an elliptical orbit or due to orbit perturbation catch-up with your neighbors (all of which will need to do orbital maneuvers and waste propellants). This stuff happens. I’m no longer in this field, but my team lost university satellite because of this. Everyone hopes for a nice decay orbit but it can get funky, and very very hard to model.
Lastly, there’re cases where satellites have been destroyed on purpose. Look up Chinese and Russian tests. The debris field produces for this is hard to model, it will propagate and react to solar winds, upper atmosphere disturbances, neighboring objects,.. Small particles pierce through everything. You will not see them, you will not be able to track them.