Used upper stages seem to be a big contributor to space debris. Wouldn't it be fairly easy and cheap to instead have an upper stage somehow physically lock onto a big piece of space debris and push it down into the atmosphere, or to the graveyard orbit outside of GEO? Sooner or later, the big pieces of junk in orbit will collide and fragment into thousands of dangerous pieces, so removing a small number of the most massive ones should be very helpful in the long run.

Launchers can obviously not always carry a payload which makes use of all their capacity. So there's often place for extra fuel for a small secondary mission after the payload has been inserted in its orbit or trajectory. That's why there's a market for secondary payloads. Even Curiosity's sky crane on Mars had 140 kg of unused fuel after landing. And Chang'e 5T is touring cis-lunar space after having delivered its return capsule to Earth. Upper stages, of course, often end their primary missions in the most commonly used orbits, where it is most important to remove debris. Large maneuvers would generally not be necessary. This concept could be applied opportunistically for launches with spare capacity and in orbits similar to a specific big piece of debris.

What are the problems with this? Is the cost of filling up a rocket to the top prohibitive? Is it dangerous to approach and grab a dead spinning debris object, maybe it actually risks causing a collision and a new field of debris? Or is it an economic problem of the tragedy of the commons which prevents even small investments in removing space debris?

  • 2
    $\begingroup$ Upper stage designers can aid in countering orbital debris - by deorbiting the stages. It is an historical fact that hydrazine-fueled stages tended to explode in space after some time. By purging the tanks and/or deorbiting, designers can do more than any hunt for a piece of junk can achieve. $\endgroup$ – Deer Hunter Dec 18 '14 at 16:17
  • 2
    $\begingroup$ The big pieces of junk aren't a big problem, mapped out and possible to dodge. It's the pebble-sized debris, big enough to pierce right through a craft but too small to show up on a radar - and far more numerous than the big ones. $\endgroup$ – SF. Sep 26 '17 at 22:39
  • $\begingroup$ @SF. In the long run, the big pieces will hit something and be fragmented into debris streams. By taking down the few biggest ones, in the long run most of the debris will be avoided. That's the argument anyway, but maybe it's proposed just because that's what's doable. $\endgroup$ – LocalFluff Sep 27 '17 at 3:10
  • 2
    $\begingroup$ @LocalFluff: Most of these in low orbit will deorbit all by themselves long before. These in higher orbits won't be reachable by 2nd stage. $\endgroup$ – SF. Sep 27 '17 at 8:20

Rendezvous and docking is not a trivial problem — you're talking about the addition of essentially an additional complex payload to the upper stage. Sensors, attitude control, comms, and some means to actually grapple and hold on to the likely tumbling target once you've gotten there. You may find a target of opportunity that doesn't need much plane change to approach, but even small plane changes are expensive in terms of $\Delta V$.

It's a challenging problem even for a purpose-built spacecraft — not sure it's feasible as an add-on to an upper stage.

  • $\begingroup$ But a hundred plus dockings have taken place successfully, I know only of one which failed, a Progress-MIR docking in the 1980's. And the sense/ctrl/com et cetera functions required are quickly getting smaller cheaper better from day to day now. I would think that an upper stage has, or would make good use of, such facilities anyway. $\endgroup$ – LocalFluff Dec 18 '14 at 16:09
  • 1
    $\begingroup$ Maybe - my expertise isn't in the spacecraft design end of things. But note also that most of those successful dockings involve the most expensive endeavours undertaken by mankind, many had a human in the loop and all involved stable, cooperative targets designed to be docked with (with the possible exception of Hubble). It's an interesting idea - somebody with more experience in this area should weigh in. $\endgroup$ – CoAstroGeek Dec 18 '14 at 16:25

Many satellites have very specific narrow launch windows, of some minutes / day when all constraints are met for reaching the target orbit. Such constraints will usually not match well to a rendezvous soon after launch to an existing specific debris. If launch slips (a common event), the new launch window for primary payload will not in general still be compatible with a rendezvous with the same debris target. (The launch windows to ISS, for instance, are a few minutes wide.) A low fuel cost rendezvous at some date possibly many months after launch might be possible, but that would mean a need for the stage to be operable for that long (which means many changes to the stage; most stages operate for a few hours, not days or weeks).

And as already said, almost all rendezvous have been with controlled targets designed with rendezvous in mind, not inert unpowered tumbling vehicles.


As others have said pretty much every piece of this ends up being challenging.

The fuel reserves needed for both managing to accomplish the rendezvous initially and then for moving it are far from trivial. It's the reason GEO sats are pushed out a little bit to a graveyard orbit instead of trying to deorbit them.

Intercepting a quickly moving uncontrolled object is also far from trivial.

Grappling a quickly moving and potentially tumbling object that wasn't designed specifically to be grappled is far from trivial.

Modifying a second stage for extended operation in space is far from trivial. You have to worry about fuel storage assuming cryogenic fuels are used. A simple battery may no longer be enough which now means solar arrays.

That's all ignoring the fact that most launches are to GTO (a transfer orbit) and not directly to GEO so the second stage isn't even making it to GEO in the first place which further increases the fuel requirements.

  • $\begingroup$ Yeah. And an upper stage is not designed to function like a spacecraft other than for delivering payload. Need solar panels instead of batteries and communication with Earth and on and on. But maybe a satellite deorbiting payload could remain attached to the upper stage and use its engine? $\endgroup$ – LocalFluff Sep 27 '17 at 12:08

There's a very big problem here: Rendezvous. Sure, it's a pretty routine maneuver with the ISS but that's because the rocket is launched into the correct orbit in the first place. If the rocket has to match orbits with it's target that gets very expensive (in terms of delta-v) very fast.

Furthermore, suppose there was something in the right place they could match orbits with it. What are you going to do? Politely ask it to come along for a ride down to the fire? You don't have any way to couple the booster to the target. If you align it to the exact center of mass (is it even known?) and light your booster how do you ensure it doesn't fall off? (Take a baseball bat and balance it on your hand--vertically. Hard, isn't it? Now take another bat and sit it on top of the bat you're balancing. Finally, give it a good shove up into the air--without the second bat falling off.)

Now, for a dedicated de-orbit tug you could do something like harpoon the target and drag it along. Said tug would have a very small engine which would keep the force down to something for which this would work. However, you're talking about the upper stage. Lets take the Falcon 9 upper stage since I just had occasion to do some math on it a few days ago. It's engine puts out 210,000 pounds of thrust. Assuming it harpooned something it's own size that means the tether is going to experience 105,000 pounds of force--that tether had better be anchored awfully well! And what's the geometry? If the booster is pushing the target the coupling must be solid enough to withstand that kind of force without shifting (not merely a harpoon!) If the booster is pulling that avoids the big balance problem but means the tether cables go awfully close to the rocket exhaust. Can you say "melt"?

  • $\begingroup$ If an upper stage "responsibly" deorbits itself after having delivered its payload, then it would intersect a number of orbits on the way down, where it could halt and take care of a debris. Hypothetically. There are NIAC level ideas about tethers and bags and bubble gums to dock with no-cooperative objects. But I have, since I asked this, realized that this is not an early way upwards. $\endgroup$ – LocalFluff Sep 30 '17 at 18:40
  • $\begingroup$ @LocalFluff Assuming a very low thrust engine I have no problem with docking with a non-cooperative object. My problem with it is doing so while applying a hundred thousand pounds of force to the connection. And why do you assume there's something in the exact same orbital plane but lower down? $\endgroup$ – Loren Pechtel Sep 30 '17 at 21:41
  • $\begingroup$ Geostationary orbit is more or less one and the same location, and one of the most poplar. When an upper stage gets itself out of the way, it might pick up some other decommissioned satellite to bring with from there. $\endgroup$ – LocalFluff Oct 1 '17 at 4:27
  • $\begingroup$ @LocalFluff upperstages usually only go into GTOs (Geostationary Transfer Orbits) to "pick up" another satelite it would need to get itself into Geostationary orbit and then deorbit from there, with the mass of the decomissioned satelite. this is significantly more dv than required tio deorbit just the US from GTO, the fuel for which would eat into the primary payload mass budget. $\endgroup$ – JCRM Sep 23 '20 at 5:51

Assuming that the debris is large enough to pick up in a meaningful fashion, which is far from always the case (orbital-velocity flakes of paint anyone?), and besides the problems with launch windows, delta-v requirements both for orbital adjustments for rendezvous (and for deorbiting), the rendezvous itself, docking and grappling which have already been mentioned (I'll just handwave those away here, deferring to the existing answers for discussion on how the previous factors are difficult), there is an additional complication. Getting the thing back down to Earth in a safe manner.

Now, you could aim for a big pond of water and just tell people to stay out of the way through appropriate channels. (For aircraft there are NOTAMs, and I'm certain there is something similar for ocean vessels.) I'm sure more than one spacecraft has been deliberately sent into the Atlantic or the Pacific. But even something like that adds complexity as well. I doubt that the state of a particular piece of space debris is perfectly known, but mass makes a big difference for rocket handling in a freefall/microgravity environment. Being off by a few percent of mass could easily throw your deorbit burn off completely, causing you to hit the Earth in some very different location than you planned. And like someone pointed out in some other answers to another question, about the only certainty is that once you perform a deorbit burn, you are going to be on an orbit that intersects ground in some fashion, somewhere. Best hope there isn't anyone in the way where that happens.

  • $\begingroup$ Not many satellites are large enough that any part of it will survive reentry. $\endgroup$ – Hobbes Dec 19 '14 at 11:03
  • $\begingroup$ @Hobbes The upper stage of the rocket might be large enough, though. I don't know exactly what the limit is (and it probably depends on the reentry angle as well). Which brings us back to knowledge of mass for deorbit burn calculations. $\endgroup$ – user Dec 19 '14 at 11:48
  • $\begingroup$ Rocket stages have a low mass-to-volume ratio (empty tank), so they experience a lot of drag and tend to burn up. $\endgroup$ – Hobbes Dec 19 '14 at 15:18
  • $\begingroup$ I've never heard that impacts from re-entering upper stages or small sats was that much of a concern. Can you provide a cite? $\endgroup$ – HopDavid Dec 19 '14 at 15:34
  • $\begingroup$ Good point. The best cite I have is Aquarius' RTG which admittedly is quite dated. $\endgroup$ – user Dec 19 '14 at 18:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.