[nordsieck]

It is particularly bad for a satellite in geostationary orbit to break up or fail. Satellites are packed as tightly as possible into that orbit due to its economic importance (it's very useful for a satellite, particularly communications satellites, to always be over the same part of the Earth), so there is a higher than normal likelihood that this could be seriously disruptive.

[perihelions]

- "Satellites are packed as tightly as possible into that orbit due to its economic importance"

Note that that's in the sense of angular separation, as viewed from the ground. They're physically hundreds of kilometers apart.

edit: (Geostationary orbits are ~42,000 km from the Earth center-of-mass; each degree of angle is an arc of ~700 km).

[naikrovek]

> They're physically hundreds of kilometers apart.

That’s pretty close when your neighbor just exploded and there’s almost exactly zero air resistance to prevent debris from reaching you.

[ben_w]

Yes there's no air resistance, but also most of the fragments aren't going your way.

If you have a 25 m^2 cross section in the direction of the explosion, at that distance you have a roughly 1 in 246 billion chance of any given bit of debris hitting you.

[anjel]

It might not place neighbors at appreciable risk but wouldn't debris still prevent replacing the failed satellite with another one at the same precious original address?

[ethbr1]

Wouldn't debris at the same address (after some time) therefore have zero relative motion?

[user32489318]

No, because the debris = tiny pieces of aluminum, will be pushed around by solar radiation. Also, there’re tiny meteorites, and other pieces of debris colliding with it, which adds energy to the system, if you like. TLEs are not maintained for small debris, so you can’t really predict conclusively. But my hunch is that eventually, the orbit will become a bit more parabolic, precession of which could put it into a trajectory of a S/C and cause a collision.

[verzali]

Things in geo orbit aren't perfectly stable, since the Earth's gravitational field is not perfectly uniform. Without active station keeping things tend to gather at two particular longitudes over the equator. So eventually that debris will probably end up at those points.

[throw4950sh06]

What is the chance of getting hit by further broken pieces of that satellite and other satellites?

When calculating risk, you have to take into account how many are there and what is the chance that any will be hit. Then you have to calculate what's the chance this will happen again, etc - and only then you can calculate the risk to your own satellite.

It's true that the chance of getting hit by one broken satellite is small. But that assumes there are exactly 2 things on the orbit.

[ethbr1]

Aka Kessler syndrome [0] or neutron flux / cross-section (and associated equations, if you want to model it that way) [1].

[0] https://en.m.wikipedia.org/wiki/Kessler_syndrome [1] https://en.m.wikipedia.org/wiki/Neutron_flux

[throw4950sh06]

That's if the risk comes out at 100%, but there's some space below that.

[randmeerkat]

> If you have a 25 m^2 cross section in the direction of the explosion, at that distance you have a roughly 1 in 246 billion chance of any given bit of debris hitting you.

Source?

[a1369209993]

  $ units
  You have: 25m2 / 2tau(700km)^2
  You want: /billion
    * 0.0040600751
    / 246.30086

[Tepix]

Getting hit by debris that flies away directly from an explosion would very bad luck indeed. Just think about how well you would have to aim to hit someone 10km away.

But some debris (in particular slower pieces) will probably oscillate around the geostationary orbit giving it countless chances of hitting other satellites.

Has someone modelled this for example in Kerbal space program?

[lupusreal]

> But some debris (in particular slower pieces) will probably oscillate around the geostationary orbit giving it countless chances of hitting other satellites.

Almost all of the debris will have orbits which intersect their orbit of origin.

[blkhawk]

well while that is true only debris that has the right speed will stay there. the issue is that debris will probably have gotten some impulse that turns it orbit elliptical. that means it will slowly ratchet along geo touching it every day or so at a different place. so you get a changes every day for every piece to hit something. And even a big piece that might just have a delta of 30km/h might damage your panels on your working Sats.

[naikrovek]

Yes, exactly.

It almost certainly won’t hit you directly, but that stuff is in orbit with you now, and it is uncontrolled.

[furyofantares]

That doesn't seem very close in terms of the area traced out by each object in relation to the area of the sphere? (And less if you consider volume since they won't be at exactly the same altitude.)

[bryanlarsen]

It can reach you, sure. But the chances of hitting are miniscule. If you could throw a basketball a few hundred kilometres you're still likely to miss the net.

[accrual]

Indeed, Intelsat 33e has a couple nearby neighbors in similar orbits and inclinations.

    ID      Name                    Orbit   Incl.
    98050A  ASTRA 2A                57.2    4.93
    09017A  WGS F2 (USA 204)        57.5    0.01
    14023B  KAZSAT-3                58.5    0.02
    12008A  BEIDOU-2 G5             58.7    2.10
    16053B  INTELSAT 33E (IS-33E)   60.0    0.04 <-- 20+ debris components
    19014A  WGS 10 (USA 291)        60.3    0.01
    04007A  ABS-4 (MOBISAT-1)       61.0    3.86
    10008A  EWS-G2 (GOES 15)        61.5    0.04
    19049B  INTELSAT 39 (IS-39)     62.0    0.02

https://www.satsig.net/sslist.htm

[nordsieck]

Looks like WGS-10 is the closest.

From Wikipedia, it looks like it's a USSF satellite launched in 2019 with a service life of 14 years. It provides wideband communications to DoD customers.

https://en.wikipedia.org/wiki/USA-291

[bryanlarsen]

It's 0.3 degrees ahead in orbit. Which means the debris needs to speed up to collide. It's possible if the break-up was explosive, but most, if not all of the debris is more likely to stay at the same velocity or slow down.

[Aachen]

Speeding up does not work, that'll just put it in a higher orbit. My understanding of these issues is that introduced eccentricity gets you: if it takes the "inside corner" (is that the English word for binnenbocht/innenkurve?) around the earth, it could then then meet you on the other side where the orbits intersect. If it sped up (flying out of the corner, in this race analogy) and thereby took a longer way around, it'd rather be a danger to those immediately behind (with each orbit, progressively further back along the circular orbital path)

Disclaimer: this comes from playing a self-made orbital mechanics game, I have no training whatsoever let alone professional experience with this

[sangnoir]

> Speeding up does not work, that'll just put it in a higher orbit

Speeding up doesn't raise the orbit; it makes it (more) elliptical while still intersecting with the old orbit (shared with neighbouring satellites)- you need at least 2 maneuvers to raise an orbit. You're also assuming a perfectly pro-grade acceleration. In an explosion, different pieces go in different directions, I suspect there is a vector that results in faster speed in the same orbit, but I'm no rocket scientist.

[sangnoir]

Correcting an inaccuracy in the last sentence I introduced in an edit: Kepler's laws means you cant have different velocities in the same orbit - bit its possible for an explosion to cause a projectile to intercept a satellite that was ahead of it in a new orbit.

[panki27]

My KSP experience supports your view :)

Parts that were sped up have their opposite orbital side raised, parts that got slowed will have their opposite orbital side lowered.

[pavel_lishin]

> this comes from playing a self-made orbital mechanics game

My interest is piqued!

[Aachen]

Thanks, but don't expect too much! (In particular on mobile where you can't zoom by scroll wheel or use arrow keys to fly.) I mentioned it in an earlier comment, probably easiest to refer back to that for some context/hints: https://news.ycombinator.com/item?id=35763506

[bryanlarsen]

Any velocity changes will only happen once so will result in an eccentric orbit where only one side of the orbit is raised or lowered -- the other side will still be in SSO.

[lupusreal]

Most of the debris will have eccentric orbits with varied periods which intersect geostationary orbits. Every satellite in a geostationary orbit is at risk, not just the ones near it.

[Reubachi]

How does one tell the dynamics of orbital objects/debris relative to forces acting on them? Is there a name for this type of field?

[visviva]

Astrodynamics

[simne]

Ballistics.

[taftster]

So does that mean the other satellites "behind" it are in more danger?

[bryanlarsen]

The one behind it is 1000km away vs 200km away for the one in front.

[matrix2003]

Not to mention debris can be in GEO for a long, long time. People worry about LEO constellations causing Kessler syndrome, but the reality is that LEO debris deorbits in the order of months/years. GEO is much, much longer.

[nordsieck]

> Not to mention debris can be in GEO for a long, long time.

On human timescales, it's basically forever. Hopefully we'll develop the tech to clean up debris in space, but it's extra challenging to do it in geostationary orbit since it's so far away from Earth, both in terms of actual distance, and delta-V.

> People worry about LEO constellations causing Kessler syndrome, but the reality is that LEO debris deorbits in the order of months/years.

It's a little more complicated than that. The time to spontaneously deorbit is based on orbital height. Starlink can deorbit on its own in 5-10 years because it's orbiting so low. But any OneWeb satellites that malfunction[1] will take 1000+ years to deorbit because they're up at 1000+ km.

---

1. Like this one

https://spacenews.com/oneweb-mulls-debris-removal-service-fo...

[matrix2003]

Yep! That's a great point! I forgot that LEO encompasses quite a bit of difference as well. Starlink has been in the news lately, so that's mostly where my mind was. I believe the newly announced Starlink shells are even lower, so that's good news from a failure standpoint.

[Aachen]

> Hopefully we'll develop the tech to clean up debris in space

Rendezvousing is pretty established tech so long as you know a precise and stable orbit of your target, afaik? Which, for geo, we would I think. So taking up some grabbing mechanism probably does it, then use ion engines, burning retrograde (avoiding the need for heavy fuel) until you get it to a low LEO orbit, let loose, and let the problem solve itself within a few ~weeks. Then move on to the next piece, so you don't need a launch to orbit for every individual piece of debris. You also don't have to circularise your orbit to just rendezvous and grab it. And you probably also don't have to go out of plane even if the target object is, if I'm visualising this correctly, because there's always a node where the planes intersect and you can just start the path up to geo at the right point in your LEO orbit

Grossly simplified, devil in the details, but this seems very possible with today's technology and potentially less expensive in terms of delta V than it may seem at first glance

[tomp]

no

if it remains in GEO orbit (same speed vector), it will remain in same "place" relative to other satellites, and won't ever hit them

if it changes speed vector, it's no longer in GEO orbit

[bryanlarsen]

if it changes speed vector, it'll be in an eccentric orbit with one of either perigee or apogee at GEO.

[exitb]

The most important contributor to a Kessler-like scenario is extremely high relative speed of items traveling on crossing orbits. It’s not very relevant to the situation in a single geostationary orbit shared by all the objects.

[Tepix]

Note that for every 1 km at the earths surface, you get 6.61 km at geostationary orbit. So there's quite a bit of room (264,924 km circumference vs 40,075km at ground level).

[idunnoman1222]

No, you cannot shift orbit with a single burn maneuver so whatever explosion unless it exploded the other way later cannot shift orbit if the pieces accelerated relative to earth they’re going into a higher orbit if they decelerated they go into a lower orbit Transverse thrust would cause a procession which should be very unlikely to hit another Geo stationary satellite in the future

[accrual]

Without having details of the explosion, I imagine some parts will slow down and some will speed up. They'll all be clustered together around the original orbit, and will take many years to drift any considerable distance unless it was a very high velocity explosion.

[dylan604]

I remember this taking a second to click in high school physics class. The initial thought is that after an explosion all of the pieces fly away from a stationary point. The satellite is not stationary, so the pieces "flying away" only have their speed subtracted from the speed of the satellite.

The fact that orbital speeds are faster than explosions took a bit to sink in.

[idunnoman1222]

That’s tangential to you not being able to fly in a straight line from one Geo stationary orbit to another

[dylan604]

How is it tangential? It's directly related. The satellite is moving in a straight line at a specific speed. The explosion does not negate that speed, rather it just applies a momentary bit of acceleration/deceleration along a new vector. So unless an explosion can provide more acceleration that can null out the original object's speed, the pieces will remain in orbit. That's just how fast orbital speeds are, and it took a bit for that to click

[idunnoman1222]

But we’re worried about the pieces hitting another satellite in goe stationary orbit further along the orbit. A single burn can’t do that in a direct way because there is no direct path between two satellites at the same orbital distance (speed)

[UltraSane]

It isn't great but the diameter of geostationary orbit is 84,328 kilometers so there is a lot of room.