I was wondering why it happened again... this refers to the events of early February.

Here's a thread from back then about it

https://news.ycombinator.com/item?id=30267587

Maybe it would be good to append "in February" to the topic? The current title is confusing - I see I'm not the only person who thought, "wtf, again?".

The actual title (of paper and linked page) suffixes 'in February 2022'. So it should just be that by the usual submit as-is guideline, unless that was too long.

I believe this was a risk spacex were deliberately taking to get more satellites launched on a single rocket.

By launching to a lower altitude initially, then more of the propulsion effort is done by the satellites themselves, and less is done by the rocket. All other things being equal, this simply means put more fuel in the satellites, and less in the rocket. Total mass is the same.

But all other things are not equal. Specifically, the satellites have 'engines' which have over 10x the specific Impulse - ie. Thrust per kilogram of fuel used.

So it makes huge sense to launch the satellites to a super low orbit and use their far more efficient engines to boost themselves up.

With the downside being the satellites ion drive engines take a long time to do the job, and if they aren't strong enough to fight against the headwind of atmospheric drag during that time, then the mission will fail.

In this case, spacex just pushed a little far.

IIR, SpaceX talks up the initial low orbits as an important anti-space-junk measure. Starting low, any Starlink satellites which fail to power up & orient & respond to ground control commands will be fail-safe de-orbit by atmospheric drag. And they will have spent zero time at more-viable altitudes, where they might be a hazard to operational satellites.

Space junk isn't really a concern at any of the orbits spacex uses. They only talk about it because this anti space junk launch plan happens to be the cheapest, and it distracts from other environmental issues they don't look so good in.

It's a concern, it is just that it may only knock out space access for a few years instead of hundreds, and it may destabilize nuclear détente, orienting things towards theater missiles for a while.

I think that you have a great argument for what they were doing. But one thing to consider too is that the space weather events they launched into were known 2-3 days out. So if it was a concern they would have scrubbed the launch. I think they saw it as a good opportunity for a test and launched even though they knew they would lose some satellites. What was different about the ones successfully launched versus the ones that didnt make it?

Note: this happened in February.

They seem to be losing a lot. Could become a real issue once they get up to serious amounts of satellites.

That's what like 2% in just this incident. Not catastrophic but not insignificant either if that rate holds

Around that when it happened in February, but under 2% now. 2,335 launched and 2,112 are still in orbit.

They are in low orbits to be safely disposable without risking orbital debris.

I had not realized they were going with such an incredible number of satellites. Absolutely incredible.

It's the tradeoff of having such low orbits that a StarLink customer might be closer to a satellite than a terrestrial ISP's nearest node. Lower orbit means less surface coverage, so you need more satellites. That's why old-school satellite internet providers put them 20,000+ miles out and have pings pushing or exceeding one second.

The 22,236 mile figure comes from the desire for a geosynchronous orbit. Not a desire for increased coverage.

Many satellites at that altitude use specialized antennas to limit their coverage to specific areas of the earth. There is no sense in wasting the little available energy on a satellite in sending RF to places where you don't intend to provide coverage.

It is just the tip of the iceberg. They have FCC approval and plans for nearly 12,000 (as of 2019), and are considering a possible later extension to 42,000.

If you have many satellites, you lose many in case of electromagnetic storms. Remember that about 25% of all satellites that currently are orbiting Earth are Starlink satellites, and that they are designed to be easily disposable and in very low orbits.

No, you misunderstand.

They need so many satellites that using higher orbits would quickly guarantee Kessler syndrome. Instead they orbit low enough to require significant fuel simply to maintain orbit. These low orbits also allow individual Starlink satellites to be less than 1/10th the mass of normal communication satellites at ~260kg.

In comparison at the opposite side of the spectrum, the largest communication satellites are Telstar 19V launched in 2018 @ 7,076 kg, and TerreStar-1 launched in 2009 @ 6,910 kg.

> They need so many satellites that using higher orbits would quickly guarantee Kessler syndrome.

I know there are good reasons to be in LEO, but wouldn’t higher orbits mean a lower likelihood of collision simply due to the increased distances?

Probably, yeah. But higher orbits means that dead satellites will stay in orbit for much longer, and that the debris from colliding satellites will stay in orbit for much longer. When the satellites are in such a low orbit, dead satellites will quickly deorbit due to atmospheric drag, and same with the debris produced by collisions.

Sure, but space is big, and satellites are small. I bet all the Starlink satellites in orbit fit comfortably inside the area of an American football field.

Satellite visualizations always make space appear more crowded than it really is.

That’s a misleading was of thinking about it. A sofa doesn’t need much space, a sofa traveling 17,000 mph for decades needs a lot of space.

Worse a sofa hit with another sofa at 5,000+ mph creates a cloud of debris that then endangers stuff across quite a lot of space.

Actually no, because junk stays in the space longer altitudes and velocity. Collision risk = density X junk time in orbit X velocity