507 comments

[ 1.7 ms ] story [ 250 ms ] thread
Rough luck. Sounds like no real risk or huge loss though.

Imagine any other company on earth losing 40 satellites at a time and shrugging it off. That would bankrupt most space companies (except for maybe Planet).

SpaceX is really working at different orders of magnitude here.

I think the correct comparison is losing one launch.

Most companies could handle that fine. That's what insurance is for.

Huh, I've never thought about insurance for stuff like this.

Do you think this was covered?

It would be tricky and expensive to insure so many of their own launches so I'd guess no.
My favorite story of satellite insurance is HGS-1. https://en.wikipedia.org/wiki/PAS-22

> However, a failure of the Blok DM-2M fourth stage left it stranded in a highly inclined (51.6°) and elliptical orbit, although still fully functional. It was declared a total loss by its insurers.

> The satellite was transferred to Hughes Global Services Inc., which was then a subsidiary of Hughes Space and Communications, with an agreement to share any profits with the consortium of 27 insurers.

> Edward Belbruno and Rex Ridenoure heard about the problem and proposed a 3–5 month low-energy transfer trajectory that would swing past the Moon and leave the satellite in geostationary orbit around the Earth. Hughes had no ability to track the satellite at such a distance and considered this trajectory concept unworkable. Instead, Hughes used an Apollo-style free-return trajectory that required only a few days to complete, a trajectory designed and subsequently patented by Hughes Chief Technologist Jerry Salvatore. This maneuver removed only 40° of orbital inclination and left the satellite in a geosynchronous orbit, whereas the Belbruno maneuver would have removed all 51° of inclination and left it in geostationary orbit.

> ...

> These operations consumed most of the satellite's propellant, but still much less than it would take to remove the inclination without the Moon-assist manoeuvres. With the remaining fuel, the satellite could be controlled as a geosynchronous satellite, with half the life of a normal satellite – a huge gain, considering that it had been declared a total loss. The satellite was then maneuvered to geosynchronous orbit at 158° West.

This ended up being the first lunar mission involving a communications satellite and the first lunar mission from a non-government entity.

Watching the orbit on the wikipedia page is wild.

Both apparently good for the future of space debris, and a publicity opportunity... Despite losing probably millions of dollars... Good job SpaceX.
The press release specifically states the satellites will de-orbit and burn up by design.
yes, they turned a loss into publicity... while not harming future space operations... They turned a loss into a win-win, I think.
The press release states the same safety plan SpaceX has been stating. We'll get to see if it works (which I expect it to), but shouldn't assume it already has.
The only reason they are lost is because they fell too far down. How could they possibly fail to deorbit?
I fully expect they will. But I want to actually see it happen instead of just blindly regurgitating corporate statements that it won't be a worry in the near future.
> blindly regurgitating corporate statements

Unlike a lot of their claims, this is basic physics. Why would you need to see it happen? Like, you didn't answer my question. It wasn't about what you expect. I asked how anything else could happen.

Their statement wasn't "this time it will deorbit" It was "as always, this and all starlink satellites were designed to deorbit". There is no reason anything else will happen (I don't know enough to say if there is a 1/1000000 chance that some atmospheric disturbance could add extra energy and push them up). However, as a rule, I don't want to base anything on Starlinks standard disclaimer.
Physics decides that. There's no need to trust SpaceX at all. The satellites will deorbit within a few days. These things aren't magic. If a corporation says gravity pulls you toward the Earth, you shouldn't suddenly start doubting that gravity pulls you downwards until it does.
Do we know what is the expected lifespan of a Starlink satellite? I wonder how well this event factored into their cost calculations, like is this a "once-a-decade" event that actually happens every couple years, or was it in-line with their actuaries?
The current generation has a 5-7 year lifespan and after that they are to either naturally decay or manually deorbit. The next gen that is supposed to be deployed with starship provided things get off the ground is intended to stay in orbit for a decent bit longer since they are larger, more expensive satellites.

In the grand scheme of things, this initial set is more a prototype than anything else so I doubt the loss will have any serious permanence. Now getting starship to production on the other hand will really be what makes or breaks the starlink program (as it won't even be possible to deploy v2 if they were to dedicate every possible falcon 9 launch towards just deploying sats).

I think this loss it a bit more painful than you imply. Sure, they're prototypes, but these are the latest prototypes. They were probably equipped with laser links, and while 40 satellites is only a small fraction of the entire constellation, it would be a much more substantial fraction of the laser-equipped ones.
This didn't affect satellites that were in their final LEO orbits; it affected part of a batch of newly launched satellites that were lower down, on their way up toward their final orbit, and therefore susceptible to the temporary thickening of the upper atmosphere. The effect on the lifespan of Starlink satellites is just to add slightly more "infant mortality" compared to a hypothetical world without geomagnetic storms -- 40 satellites is about 2% of the ones they've launched.
Of all press releases in recent memory, this is the one that reads the most like hard science fiction.
what's cool is, it doesn't even feel futuristic anymore.
Someday potentially we could buy/lease/rent our own satelites over satelites.com use it for personal research, photography etc.,
You can already make your own and get it launched if you have enough money. Cubesat components are pretty much off the shelf except the payload. And launch costs are to the tune of 30k per kg or so. I'm sure a lot of overhead / integration is added but it's financially possible for the well-off :)
Launching a satellite is already sub $100k. Technology both miniaturizes and industrializes at an exponential rate. Think about how small and cheap computers have become in the span of a hundred years, from "occupies a building and costs a fortune" to "fits in your pocket and everyone can afford one". Sputnik 1, the first satellite ever launched cost 33 billion 1985 US dollars. 45 years later, you can get a cubesat into orbit for $100k.
I really don't think Sputnik cost more than the entire Apollo program, nor 8x the cost of the JWST. I can only find a couple of sites saying "33.000 million" for Sputnik's cost, unsourced.
That could be an EU/US issue. In the EU it is common to separate thousands (or millions) with dots, and decimals with commas. In the US it is the other way around. So those sources could be saying (or were read by GP) as 33 thousand million (33 billion) or it could mean 33 million and 0 thousands.

The billion sounds linguistically more correct, but I have no idea if those are inflation adjusted dollars or whatever. It could be an error that telephones around the internet as 33.0 million by the US system then converted to someone in the EU to 33.000 because they misread the number.

I definitely interpreted it as 33 thousand million. Whoops :D
Amateur groups have been launching satellites for decades now...
Somewhere there is a scientist trying to convince his peers this wasn't a geomagnetic storm but instead a alien spacecraft but they don't believe him.
Oh my god I think you're right.

Somewhere this morning a low-level SpaceX engineer was apologetically skipping out of the house. A full breakfast—uneaten, except for the toast he grabbed—laid out on the kitchen table and now the focal point of his wife's muttering. In the background a TV has the morning news on, briefly mentioning the geomagnetic storm. "I promise I'll be back in time for Kaitlyn's play. ...Yes I know how much this means to her."

But he won't be, will he? Because by the end of the day, the entire world will have been transformed in ways nobody could have seen that morning. Except for the perennially-ignored news anchor on in the background.

EDIT: Yeah, https://xkcd.com/1387/

I think the press release is lacking in detail. For example, they completely fail to disclose whether recovery was attempted by rerouting power from life support to the structural integrity field.
I believe a geomagnetic storm calls for rerouting power to the subspace accelerators backing your deflector dish, and of course sufficient impulse power to maintain an appropriate heading, not rerouting to the SIF. Please get your spaceship right. It's a technical forum!
This is why I read HN. You just can’t find experts like these anywhere else.
Sorry, was assuming SpaceX had quantum slipstream already. But it appears I was mistaken. In the case of a slipstream drive, integrity for entry and exit of the slipstream would be paramount. But if they’re still using warp, then your comment of course is accurate.

Impulse, by the way, is covered by the press release [1] and, I agree, would need to be a priority no matter which FTL drive is used to get the satellites to their final destination.

[1] > The Starlink team commanded the satellites into a safe-mode where they would fly edge-on (like a sheet of paper) to minimize drag—to effectively “take cover from the storm”

Well, "continued to work closely with the Space Force’s 18th Space Control Squadron" was a bit on the nose.
We live in the future for quite some time now.
I think most people live in the present, yet if they dig deeper, realize the future is here somewhere.
Even though the future seems far away, it is actually beginning right now.
The cost breakdown:

  40 x $250k       = $10m
  1 falcon9 launch = $30m
  ------------------------
                     $40m


https://www.nextbigfuture.com/2019/12/spacex-starlink-satell...
The Falcon 9 should be pro-rated since some of the mission payload survived.

  40 x $250k                   = $10m
  1 falcon9 launch * (40 / 49) = $24.5m
  -------------------------------------
                                 $34.5m
Isn’t that retail cost of launch? Since they own the spaceships it should be 1/2 or less of the retail price.
That's the money they would have made and since I assume they don't any marketing that'd be the total realized cost.
That would work if they had sold all of their capacity, but I think there's plenty of unfilled Falcon 9 capacity.

I don't think every Starlink launch could be replaced by a paying customer.

I don't think any would, it's all spare capacity.
There might be the opportunity cost of a doing a retail launch instead, but I would guess for the Starlink launches, if they weren't launching them they wouldn't be launching anything at all, so you're probably right.
Another way to think about is that the cost to launch the remaining 9 went up to $3.3m/satellite instead of 612k/satellite.
That’s a lot of money… but also, it’s forty internet sattelites, so maybe it’s actually shockingly cheap!
(comment deleted)
> SpaceX could already have $250,000 per Starlink satellite cost and $15 million per launch cost.
(comment deleted)
Very interesting development, mainly that it posed no threat of causing a cascade of space debris since it's in such a low orbit. That feels like an important aspect for considering the acceptability of so many objects from one company/entity in orbit, especially considering the existing cost to astronomy.

So we've heard from SpaceX about their LEO sats, but what about other ones further out?

I believe the faster orbital decay was part of the reason for putting them so low (on top of the obvious latency advantages).
The thing that's not being well conveyed to people here is the storm didn't cause any satellite failure. The storm caused an expansion of the atmosphere which prevented re-orientation and orbit raising. The satellites are still functioning fine.
What's crazy to me is that 50 years ago, losing 40 satellites would have been a significant portion of all satellites ever launched. Today, it's a temporary setback for a company that launched nearly one thousand satellites last year.
That's part of the reason there's just so much junk flying around in orbit.

https://en.wikipedia.org/wiki/Kessler_syndrome

If you read the article, SpaceX specifically designed the process to avoid orbiting junk.
Yup, spacex stuff is all LEO. Without input it all crashes into the atmosphere in a few years (3 I believe)
And we inhale it all. (Yes, I know it's insignificant; all things start insignificantly.)
Just wait until you learn just how much random space stuff enters Earth’s atmosphere in any 24 hour period
When you consider the century-long global effort burning 2-3 billion gallons of oil every day to raise CO2 from 320ppm to 420ppm, you get a sense of the scale involved.
Yea ~0.033% to 0.040% took ~879,000,000,000,000 kg of carbon with the oceans sucking up another 590 Gt, and land taking 528 Gt from 1750 to 2012. We added another 0.004% in the last 10 years.
this argument is pretty general; you're a priori attempting to prove that "acknowledged insignificant things are actually significant". This feels too strong!
About 15,000 tonnes of material enter the Earth's atmosphere each year from meteors. I know the makeup of satellites vs. random meteors isn't exactly the same, but it's probably not concerning in terms of pollution sources, and it's really hard for me to imagine it getting there.
Someone already mentioned burning oil, I'd like to also mention coal which makes oil look very clean by comparison. Coal is a few ppm uranium, arsenic, and other unpleasant metals (this is from memory, but it has lot of impurities). We burned literally billions of tons of it, putting thousands or tons of these metals into the atmosphere in the process.

Also, leaded gasoline

If you read the article, it's not just that they operate at LEO, but the launch process begins at even lower altitudes so that they can quickly deorbit any problem satellites.
Would spacex be liable if one of their satellites hits your car or house?
If there’s so little atmosphere that a Starlink doesn’t burn up before it hits your car, the ICE engine would have trouble running anyway.
How modified would a Tesla Cybertruck need to be to run on the moon?
If I had to guess, probably quite a bit. I doubt the battery cells or the electronics would hold up under hard vacuum.

I recall a story posted here that if you put an iPhone in helium it will break the MEMS components inside of it and it will fail to boot, and that's still at 1 atmosphere.

The battery is already sealed, and 1 atmosphere isn't a lot of pressure to contain. It might even be able to be pressurized to quite a bit less. Regarding the electronics, engineering-wise that's a solved problem. And it would only need the drivetrain and charging electronics.

I suppose that a wider track would be useful in 1/3 G, but that is a mod that off-roaders have been doing at home for decades. Ditto ride height and wheel size.

I would not be surprised to see a Cybertruck offloading from a Starship someday, maybe even on a private mission after Artemis finishes.

Yes. But they're designed to burn up on entry. I think it's been mentioned on HN that the delay in their laser link network was due to the need to redesign the laser component to ensure it would break up.
When people say something will "burn up on reentry", a phrase that I've heard almost all of my life but never really given much thought too, how burnt are we talking? Surely it doesn't just completely atomize?
I don't know specifically but as a layman I'd imagine the satellite breaking up into smaller and smaller pieces due to the rapid heating and the "wind resistance"(?) from making contact with the atmosphere at high speeds.

I'd also imagine that at some point in this breaking-up-process the pieces are small enough to sublimate from the extreme heat, i.e. "burning up".

Do you remember, back when SpaceX used aluminum grid fins, when you'd see a Stage 1 re-entering, and you could watch the grid fins erode away in realtime video?

Think of that, but for, well, everything.

You know how racecars spectacularly fly to pieces when they crash? That is by design - as it rapidly sheds energy. Same thing applies to things that "burn up on reentry", so long as you prevent the thing from coming down as one solid lump of glowing hot metal - the surface area to mass ratio renders the fragments harmless.

I wonder if anyone has ever tested the whole thing about field mice surviving a drop from any height due to their low terminal velocity... those bat bombs in WWII don't count.

Small satellites like Starlink burn up in athmosphere. But if you're curious about liability laws for falling satellites, here's the relevant UN convention: http://www.unoosa.org/pdf/gares/ARES_26_2777E.pdf
That convention makes states (i.e. national governments) liable, but doesn't by itself make SpaceX liable. I assume SpaceX would be liable under ordinary tort law in the U.S., irrespective of the convention. At least for claims against U.S. nationals, the convention is probably mostly irrelevant; it's purpose is to act as a backstop in case compensation can't be found using normal legal procedures--e.g. courts won't otherwise hear a claim (AFAIU, the U.S. is relatively liberal in permitting foreign entities to sue in domestic courts for all manner of situations, and to do so on equal footing), or the entity is unable to pay compensation, in which cases the state signatory effectively acts as a guarantor for all the entities within its jurisdiction.
> That convention makes states (i.e. national governments) liable, but doesn't by itself make SpaceX liable.

The US is liable for any damage from space SpaceX causes in a different country. If it hits a house in the US, that's a completely different question.

Probably? But I imagine that won't/can't happen because the satellite is likely designed to burn upon reentry. It would seriously surprise me if no regulatory body thought of this before they sent thousands into LEO.
Yeah, I mean to be specific your homeowners' insurance would pay you directly for "falling debris", but then they would go sue SpaceX to get their money back.
This reminds me of an episode of Jim Henson's TV series Dinosaurs where a space rock hits the Sinclairs' house, and the insurance refuses to pay out. Because they had "meteor" insurance, so their house would only be covered if it was floating in the upper atmosphere at the time of the strike. Instead, the insurance said their house was hit with a meteorite (a meteor that reached the ground), which wasn't covered under their policy.
They should have argued that it hadn't yet hit the ground when it hit their house.
It wasn't a meteorite when it hit the house, only when it hit Earth. I know, hit me with the downvotes.
I think their house was made of earth (rock).
If you can build your house, or drive your car, at that altitude I rather think you'd be the one being liable! :D

(based on the fact these satellites are designed to burn up completely upon re-entry into Earth's atmosphere and so no 'chunks' reach the surface)

I don’t think this is true. Looking at this: https://i.stack.imgur.com/pMchk.jpg from UN, objects above 500km take ~25 years to decay from orbit. They reportedly start at 550km, so I wonder how they hope to achieve 3 years of deorbit without active control.
That infographic is a very simplified view of a complex subject. Here's a more detailed overview: https://www.spaceacademy.net.au/watch/debris/orblife.htm

Bear in mind that Starlink is much lighter and "draggier" than typical artificial satellites. Based on publicly available data, I get a ballpark figure of roughly 20 kg/m^2 for the ballistic coefficient of a tumbling Starlink satellite.

Just eyeballing the graphs in the link I provided, that corresponds to a lifetime of something like 2-9 years, depending on solar conditions.

It depends on density and cross section (and solar flux); a 500km orbit can be 2-30 years, depending. That's all basically still in the "self-cleaning" domain. Being rather dense and flat, I'd expect Starlink to be on the lower end of that range.

http://www.spaceacademy.net.au/watch/debris/orblife.htm

The Starlink "0.9" batch was launched in May 2019. Most reached operational altitude; those that did not decayed quite early, as expected. Those that remained operational have by now been deliberately dropped, but some 5 seem to have become unresponsive at a more-or-less operational altitude.

See dashboard at https://forum.nasaspaceflight.com/index.php?topic=49936.200

Starlink-43 is the first of those to fully decay, reentering around 26 January. https://in-the-sky.org/spacecraft.php?id=44257

Starlink-24 is probably going to be the slowest of that cohort. It might have a couple years left. https://in-the-sky.org/spacecraft.php?id=44257

Natural decay in <5 years is really quite fast as these things go.

TFA says space junk shouldn't be a problem with the launch method they used

>SpaceX deploys its satellites into these lower obits so that in the very rare case any satellite does not pass initial system checkouts it will quickly be deorbited by atmospheric drag. While the low deployment altitude requires more capable satellites at a considerable cost to us, it’s the right thing to do to maintain a sustainable space environment.

Need third party analysis of that, personally
The search term "Gabbard diagram" should tell you all you need to know. Especially if you find some animated ones. Here's a particularly nice one:

https://www.youtube.com/watch?v=5ijKirDjwOU

The Gabbard diagram is a way of expressing the orbital locations and decay of the particles produced as a result of an orbital collision. Note how higher objects tend to stay high for a long time, but low objects (below about 600km) drift lower at an appreciable timescale, and once they're below about 500km they fall the rest of the way very very quickly.

Starlink satellites deploy to 550-km orbits.

Edit: Oh, these were launched into a 210-km orbit, which combined with the higher atmospheric density due to the geomagnetic storm, basically dragged them back down before they could raise to their normal orbit. Oof.

Even more strongly, SpaceX launches to a ~200km orbit and has the satellites raise themselves. This way any satellites without attitude control deorbit immediately.

Also, the vast majority of SpaceX satellites will be at 350km.

Your numbers are a bit on the low side.

They start off at a 350 km ( https://in-the-sky.org/spacecraft.php?id=50803 was launched earlier this year).

They then raise them to about a 550 km orbit for its operational lifetime ( https://in-the-sky.org/spacecraft.php?id=45102 )

And then they lower it a bit, and it decays more rapidly ( https://in-the-sky.org/spacecraft.php?id=44249 )

Another one that didn't go operational - https://in-the-sky.org/spacecraft.php?id=44282 - you can see it go up, fail, get lowered, and decay.

The VLEO orbit for Gen2 is 340km, and will comprise the bulk of the satellites.

https://en.wikipedia.org/wiki/Starlink#cite_note-sn20170917-...

> SpaceX has two constellation applications pending FCC approval — one for a 4,425 satellite constellation operating in Ka- and Ku-band from around 1,200 kilometers, and another for 7,518 V-band satellites flying between 335 and 345 kilometers.

I am not aware of any Starlink satellites currently in either of those two orbits.

Yes, I used future tense in my comments.
Many third parties have analyzed it. FCC, Nasa, their competitors, and others.
To the people that provided third party analysis, thank you, to everyone else, be like those people.
> 50 years ago, losing 40 satellites would have been a significant portion

made me think "100 years ago, it would have been (40/0)*0.01 = infinity percent", which probably means i need to lay off the coffee a bit.

Is it bad that my thought in response was, well if you consider suborbital rockets a fractional satellite, then you can analytically continue the function of satellites over time all the way back to the invention of blackpowder? 1 satellite = ability to orbit at least once = ~150 km apogee. So a 150m launch would be 1 millisatellite...

I too probably should lay off the coffee.

What's kind of crazy is the expected lifespan of an individual satellite in the 'full' 12k constellation is supposed to be 5 years so 6-7 satellites a day will reenter or just go offline and need to be replaced. Unless that 12000 number includes some in longer parking orbits they bring down into the ELEO service orbit that's a dedicated Starlink launch about once every week.
Yeah, it is crazy! Crazy more expensive and wasteful than installing land lines.

People in the developing world are not going to pay a high enough price to float this business. The vast majority of people in rich countries have cheaper and better internet already installed.

There are some things where satellite internet makes sense but I don't think Starlink [0] makes sense for a lot of home users. Industry and transportation it might make a good case for but it'll have to prop any consumer market up on the prices charged to business users.

[0] at the stated numbers at least, the final constellation will likely be much smaller than originally quoted Musk is familiar with embellishing and overpromising timelines.

40 sats makes up about 2% of the current fleet - so a rounding error in cost and time.

The important part of this news is how those 40 sats deorbit and leave no junk either in orbit or on earth.

SpaceX should build a satellite picker satellite to collect their dysfunctional Starlink satellites and bring them back on Earth. They could study their vulnerabilities and apply the lessons learned to shield future Starlink satellites against solar storms.
They are building that - it’s called Starship and one of the configurations will be a cargo ship with a bay and an arm that can pick stuff up and bring them back.
It is not that the satellites were insufficiently shielded against solar storms, it's that the atmosphere was impacted and thus their tactic of low Earth orbit didn't work at this precise moment.
The press release implies that the issue is known. They entered a safe mode just after launch, and had not yet reached an orbit from which they were able to successfully leave safe mode.
A complete waste of resources. Cheaper just to send more of them up there. They know what caused these to come down.
You apparently did not read the article as the solar storm did not in fact knock out any satellites.
An entire launch worth of satellites is not a "rounding error."

The carbon impact of just one launch is pretty obscene.

In context of 1700 starlink satellites already in orbit and 30+ launches that put them there, it’s a rounding error.

Same goes for carbon impact, etc.

1700 - 40 = 1660. Rounded it to the nearest integer.
In scientific notation (orders of magnitude is what matters here):

1.700e3 rounded -> 2e3

1.660e3 rounded -> 2e3

Put another way: if your bill for lunch is $16.60, wouldn't you just as easily call it $17? Or even $20, if you're feeling a bit imprecise?

Yes, where I live, a lot of businesses charge you credit card surcharge of 2-3%. That’s $17 for a $16.60 priced item.
(comment deleted)
A falcon 9 uses about as much fuel as three transatlantic flights, doesn't it?
> The carbon impact of just one launch is pretty obscene.

It really isn't. It's what, 15 aeroplane flights' worth? For the sake of bringing internet to hundreds of thousands of people? By letting people shop online or stream entertainment they've probably paid for themselves in reduced car journeys already.

You're citing the cost of 1% of the network against the entire network's function. It's more like 750 "flights worth." To provide internet service to ~250,000 people - about one quarter the population of a major east coast city.

Also, these are consumable - much more so than cable plants.

Starlink is horrifically bad for the environment compared to ground-based infrastructure. As witnessed by the very high cost of Starlink service, which is likely losing mountains of cash despite that.

> For the sake of bringing internet to hundreds of thousands of people? By letting people shop online or stream entertainment they've probably paid for themselves in reduced car journeys already.

That's not how that works.

Individual consumers aren't the only customers of Starlink. It also will bring high-bandwidth low-latency connectivity to the entire globe once enough laser-linked satellites are up. Industry and militaries will be using it. Whether this is a net positive for the environment or otherwise remains to be seen.

Also, the point of Starlink is to fund SpaceX's goal of a methane rocket. Methane was chosen because it can be generated from atmospheric gases and solar power. The logistics around scaling it up remain to be solved, but once it is solved, rocket launches will be basically zero carbon.

If you can extract methane from the atmosphere at scale you don't need rockets to solve the carbon problem.
The idea would be to synthesize methane from the CO2 in the atmosphere of Mars, which isn't quite as hard as on Earth: CO2 falls as snow near the poles there.
> Starlink is horrifically bad for the environment compared to ground-based infrastructure.

So what should people who dont have access to the ground-based infrastructure do?

> You're citing the cost of 1% of the network against the entire network's function. It's more like 750 "flights worth."

Well you opened up by talking about "just one launch".

> To provide internet service to ~250,000 people - about one quarter the population of a major east coast city.

But people who are currently very hard to supply internet to and extremely car-dependent. Building housing in cities would probably be a better solution, and I'm all for efforts to make that legal, but realistically it's going to be extremely difficult to overcome upper-middle-class opposition to that and so other ways to mitigate the impact of those people's lifestyles are very much worthwhile.

> Starlink is horrifically bad for the environment compared to ground-based infrastructure. As witnessed by the very high cost of Starlink service, which is likely losing mountains of cash despite that.

On the contrary, running cable to many of these places would be overwhelmingly environmentally damaging (in many cases you'd have to build a road through a forest or blast through a mountain), and that's reflected in the high (sometimes infinite) cost of internet services. People are buying Starlink because it's cheaper (in the places they live) than ground-based alternatives, and as you've observed, the environmental cost is generally closely correlated to the monetary cost.

> That's not how that works.

That's exactly how it works. You've got to look at the net impact. For people who don't travel internationally, reducing car usage is the most effective single lifestyle change they can make to reduce carbon impact; if Starlink reduces trips to the extent that e.g. people shift from a two-car to a one-car household, that's huge. (And note that if it's reducing travel into towns or cities then that's benefitting the rest of us as well, by reducing congestion and parking usage, which in turn lets us turn more land over to productive uses).

>On the contrary, running cable to many of these places would be overwhelmingly environmentally damaging (in many cases you'd have to build a road through a forest or blast through a mountain)

How so? How do this people landed in this place without a road and electricity? Rural homes must have roads to reach the city to buy fuel and food. There are also wireless terrestrial ways to transmit internet too.

> How do this people landed in this place without a road and electricity? Rural homes must have roads to reach the city to buy fuel and food.

They might rely on generators for electricity, and access might involve a boat, or a quad bike trail that would need enlarging before you could bring construction equipment in and run a cable along it.

> There are also wireless terrestrial ways to transmit internet too.

I mean ultimately the proof is in the pudding; if it were practical to run internet some other way they would already get it that way.

>They might rely on generators for electricity, and access might involve a boat, or a quad bike trail that would need enlarging before you could bring construction equipment in and run a cable along it.

Who are this people? ARe they poor farmers or rich dudes on islands?

>I mean ultimately the proof is in the pudding; if it were practical to run internet some other way they would already get it that way.

It might be practical but not obscenely profitable. I double that Elon sells his stuff only to the above people, the people living in isolated environments with no roads and electricity are too few. In many places it is legal to put cables on the electricity polls and connect a village 25Km away in 1 single day, from what I read this is not allowed in US so you have no internet or a monopoly.

> Who are this people? ARe they poor farmers or rich dudes on islands?

You know, there are poor dudes that live on islands too.

Like one of my friends. It's a river island, it's only about five kilometers to the nearest cellular tower, and as you might guess running fiber across a major waterway is out of question.

So we built a 20-meter tower with a 1.5-meter parabolic dish on top just to get 3G-ish speeds on the site.

Quit the condescension, please.

There are a LOT of people living in such environments. Hell, when we were doing rural cellular in Africa, it was like three days over non-existing roads to get to the installation site. No electricity, only satcom, and that with 1-second rtt. Starlink is the best thing that could've ever happened for those applications.

Why I am not convinced that Elon and all HJN-ers here are thinking at poor people but instead find this extreme cases to shamelessly use them to promote the technology to work around the fact they don't like their mobile provider.
Because you're looking at your own prejudices rather than the facts? I'm no fan of Musk and Starlink will never be any use to me (I live in the biggest urban area in the world), but knee-jerk reactionism disguised as environmentalism is one of the biggest barriers to actually fixing environmental issues.
> the very high cost of Starlink service

It's $99/mo for 100+ megabit service. Sure, it's moderately expensive if you're comparing to cable service in cities. It's downright cheap compared to the cost of connecting the extremely rural customers Starlink targets with physical cables. You don't think running a new set of physical cables to every rural residence in the world (or even cell towers to cover the same) would be expensive or have any environmental impact? Starlink wins on both and not by a small margin.

And then of course there's the prospect of bringing service to the remote mountains and jungles and deserts and polar regions and oceans where nobody lives and which will never (and should never) have ground-based infrastructure. And a bit of sorely needed competition to keep the ground-based monopoly service providers in check. And service for disaster areas or war zones where infrastructure is destroyed.

Plus $500 sign up fee ;)

This might all seem reasonable in rural America but it’s hugely expensive for internet in Europe.

That pays for the antenna. The antennas are expensive now because they are new technology, never manufactured on this scale before, but that is changing fast. The fee will get smaller soon enough. Ground-based infrastructure will not be experiencing any analogous rapid cost reduction. And in rural areas it costs a lot more than $500 per residence already. What infrastructure exists is heavily subsidized.
I know it pays for the antenna but the fact remains the cost of Starlink isn’t just $99/month like you’ve claimed.
I think when I got my first cable broadband the modem cost ~$300, probably more than $500 now after inflation.
> Plus $500 sign up fee ;)

That's going to average out to less than $10 a month pretty quickly.

Let's first see if you can even keep using the same make of antenna before replacing it with a new model for 5 years
The problem with that logic is that it’s not paid across 50 months. It’s a sign up lump sum to pay for the antenna rather than a fixed monthly cost added to your contract duration.

Also your post doesn’t take into account the life span of the antennae, that could be less than 5 years. It doesn’t take into account any competing services that you might want to switch to (eg some internet contracts in the U.K. typically run for 2 years and you’re encouraged to shop around at that point). Nor does it take into account any other personal circumstances that might lead one to cancel Starlink.

I had to pay hundreds for my first cable modem, so the early stage costs are comparable.
My point wasn’t whether the cost is reasonable. My point was just that the first month is a $600 bill rather than $100 like the GP suggested.

For some that cost is still reasonable for what they’re getting. And that’s fine. But from a purely mathematical standpoint the GP was significantly off in terms of the price they quoted.

You can all argue about whether that matters or not but that’s really just a subjective matter. Some folk cannot afford a $500 sign up cost. Whereas for others it will be an absolute bargain for what they need. It really depends on their income/savings, what other internet providers are available to them and how dependent they are on the internet. So the answer to the point you’re discussing is going to vary depending on the individual.

Thanks for clarifying, it's a fair point. Especially since lots of folks who don't have the upfront costs will finance it on a credit card and end up paying even more in the long run. (Hoping they wouldn't also do that for the monthly charge too, but lots of people finance normal living expenses every month, beyond their monthly income, until their credit breaks under the load)
If we're using price as a proxy for environmental cost, it seems irrelevant whether it's paid up front or over time.

Lifespan certainly matters, but the financial structure doesn't.

You do realise not everyone has $500 to spare? For some people the distinction matters a lot.

It’s the same reason some people with high end phones pay a larger monthly phone bill. They want a phone they cannot afford to buy so they pay in instalments as part of their mobile contract.

So yes, it absolutely does matter to a lot of people.

This is a beta with no subsidies yet and no promotions because they have too many people interested.

So you're worrying about that way too early.

You'll be able to get starlink without a big upfront fee later, I assure you.

It's an issue but it's a separate issue from "high cost of service".

> This is a beta with no subsidies yet and no promotions because they have too many people interested

Not entirely true. The hardware is sold at a huge loss. If I recall correctly the dishes cost a grand to make. Musk expects those costs to come down but the fact remains he’s losing money with every sign up.

My point wasn’t about whether the cost is worth it though. It’s that the service costs 6 (!!!) times more than the GP claimed.

It’s really not a complicated point I’m making here either. I’m not judging the service, just correcting an error in someone’s post.

> So you're worrying about that way too early.

I’m not worried. Why would I be worried about how much an ISP I don’t need costs?

> You'll be able to get starlink without a big upfront fee later, I assure you.

I could afford it now if I wanted it. But I’m also socially aware enough to see that most people aren’t as privileged as we are.

> It's an issue but it's a separate issue from "high cost of service".

I’m skeptical this will ever become economical for low income families or lesser developed counties. The latter being a group that could really benefit from this.

Regarding cost of service, we moved on from the environmental impact a while ago and now just discussing the financial cost of service.

> I could afford it now if I wanted it. But I’m also socially aware enough to see that most people aren’t as privileged as we are.

You and everyone else.

Is that implication not obvious?

You're so locked on to the privilege issue and misreading the word "you" that you're completely ignoring my point, which is that the privilege of paying a big upfront fee will not be necessary.

Just like dozens of other services, you'll be able to sign up for a contract and the equipment fee will become either a small monthly amount or get waived entirely.

Eventually you might be right. But you’re now talking about theoretical future rather than what the literal monetary cost of the service is right now.

I say that because my point wasn’t to say that Starlink is uneconomical for the poor and always will be. My point was just that the GPs post was factually incorrect because he left off a hugely significant sum (relative to the first month) for the cost of Starlink right now.

The point about privilege only came about after others handwaved that $500 as loose change and I only really made that point to illustrate how ridiculous it was to say “it’s fine to get a payment wrong by half a grand because we’re wealthy”. It just smacks of being out of touch with reality (or perhaps they were just Starlink/Musk fanboys and defending factually incorrect posts if it looks favourably on Starlink?).

My personal opinions of Starlink is that it looks really promising. In fact I nearly signed up 9 months ago myself when I moved house. The kicker was I needed it a available within a week and didn’t really want to run two service providers. I’m guess given that Starlink is still beta I was asking a little much there. The problem is my current ISP locks me into a 2 year service and as much as I can afford Starlink there’s other things I’d rather spend my money on than paying for two ISPs just for curiosities sake. I’ll probably review that decision when the contract for the current ISP has > 12 months remaining.

I’m sure I’m 5 years time the cost of Starlink will become a lot more attractive to lower income families in the west. Or at least those that might need to depend on it might find it easier to gain access to the service. But I’m skeptical we’ll see the service affordable for most rural communities in Africa, India or South America, to name but a few large regions that could really benefit from such a service. Maybe Musk does have a plan here. He certainly doesn’t owe these people any charity; however it does feel a missed opportunity if some arrangement couldn’t eventually be found.

Anyway, that’s my 2c worth of opinions :)

It makes a huge difference to whether people buy it. It does not make a difference as a measure of "how much carbon does it release when they buy it", which is the question being debated in this thread.
> which is the question being debated in this thread.

That was discussed elsewhere in this thread. This discussion in this fork is very specifically about monitory costs.

Hence all the dollar signs. ;)

It's crazy how many options it opens. With combination of remote work, many areas that were effectively shielded off from digital nomads, now have a chance to attract those tourists. I remember as well rejecting lots of vacation places because of terrible internet. Poor local people stayed poor and I spent my money in popular tourist destinations that already had high internet speeds. Overall its one of the factors that can transform the society and possibly partially break down the big cities.
Isn't it $99/mo for service? Afaik there is no guaranteed bandwidth at all at the moment as all users are beta testers? All users are on a "best effort all you can get" plan. I'm sure there will be different tiers as the network matures.
I hate to break it to you, but you don't get guaranteed bandwidth on your wired consumer ISP either. Reported speeds have been more than reasonable for a consumer service. 100 Mbps isn't even the peak speed, that's more like 250 Mbps.
I'm not convinced that starlink is cheaper or better for than ground based cellular infrastructure - its closest technical analog.

It does however possess two chief advantages - the appearance of being novel and speed of construction. To build a cellular network to cover the areas that starlink does would take a decade - where it can be done at all. The aforementioned novel appearance also gets it largely free spectrum, which I believe starlink didn't have to go thru auction for.

To be clear - it's not novel, it's more or less a cellular network where the cell sites move about instead of the subscriber - which is something iridium did. I'm still not sure it can make money however.

> I'm not convinced that starlink is cheaper or better for than ground based cellular infrastructure - its closest technical analog.

Starlink is a miracle for people with shitty or no cellular infrastructure. There are places in Australia that have exactly one cell tower for signal and they get 3G speeds at best. And if that tower has a fault, they'll be out of access for days or weeks.

$99 per month for Starlink is a no-brainer.

I don't think it can be profitable at that price is the point.
Current satellite broadband (in the US) is in that same pricing ballpark, I don't see why Starlink couldn't be profitable at that price-- or even lower-- once it scales up.
Different Opex and technical requirements -

The opex of keeping a constellation that large in orbit, has not been proven to be cheaper than operating a ground based wireless network (and Starlinks opex is certainly more than a convention geosynchronous satellite system). Starlink operates with an assumption, that has not yet been proven to be true - that the cost per kg to launch something into LEO will go down as time goes on. If this prediction does not prove to be true, then it has no hope of making money, much less being self-sustaining.

Beyond this, the shorter lifetime of their sat fleet compared to a geosynchronous sat, means they must be replaced sooner, but its not obvious that the sats are significantly cheaper - this on top of the inherent launch costs, and the much much much larger fleet, implies that there may be a disconnect in pricing.

> To be clear - it's not novel

Except no-one has run constellation of thousand satellites.

Except no-one is trying establish sat-to-sat communication links (I may be wrong here, someone should provide proof)

Except no-one can provide you with low latency, high speed satellite internet. If you have alternatives (I do where I live) - Starlink is probably not for me.

You seem to paint as if starlink is something trivial, as some years ago people were thinking CRAZY - they can't possibly put that much satellites up there. Not to mention that people were really surprised they could put 60 satellites on a single rocket. That they established mass production of satellites. Etc.

This kind of comment reminds me of the Dropbox introductionary post [1]: "you can already build such a system yourself quite trivially by getting an FTP account, mounting it locally with curlftpfs, and then using SVN or CVS on the mounted filesystem"

[1] https://news.ycombinator.com/item?id=8863

It's not exactly equivalent, but I think the TDRS provides communication between different objects in space.
> Except no-one is trying establish sat-to-sat communication links (I may be wrong here, someone should provide proof)

Iridium has been running sat-to-sat communication links since 1998?

Taking what Iridium did and scaling it up 500x does not make it novel.

Iridium did all of the things Starlink does - it just did them with 2G bandwidths, instead of 5G bandwidths.

Furthermore the opex of keeping a constellation that large in orbit, has not been proven to be cheaper than operating a ground based wireless network. Starlink operates with an assumption, that has not yet been proven to be true - that the cost per kg to launch something into LEO will go down as time goes on. If this prediction does not prove to be true, then it has no hope of making money, much less being self-sustaining.

Beyond this, the shorter lifetime of their sat fleet compared to a geosynchronous sat, means they must be replaced sooner, but its not obvious that the sats are significantly cheaper.

The appearance of novelness is what allowed SpaceX to get the spectrum to operate what is effectively another cellular network - well a hybrid between fixed wireless and cellular.

> It's what, 15 aeroplane flights' worth?

Try a single aeroplane?

For just the fuel, yeah, but the expendable second stage adds a lot of impact.
How so? https://en.wikipedia.org/wiki/Micrometeorite#Introduction

> An estimated 40,000 ± 20,000 tonnes per year (t/yr) of cosmic dust enters the upper atmosphere each year of which less than 10% (2700 ± 1400 t/yr) is estimated to reach the surface as particles.

The thread is about CO2 emissions.

The entire total carbon emissions of producing and flying the second stage is non-trivial.

> The thread is about CO2 emissions.

I agree, but the post I was responding wasn't talking about CO2 emissions.

A Falcon 9 caries just a bit more fuel than a fully loaded 777. Which is big, to be sure, but still a drop in the bucket just in terms of transportation.
This isn't quite a fair comparison, though, because it doesn't consider the energy requirements to refine lox.
How significant are those? 2x 777 flights worth? 10? There are thousands of airplane flights every day, but only a couple of these per month.
It takes about .7 MJ to refine a kg of lox, compared to 43 MJ released by a kg of kerosene. So that's not really a significant fraction.
It's estimated that about 260g of CO2 is emitted per kg of liquid oxygen delivered to end users.

And the oxidizer to fuel ratio is 2.5, so basically this doubles the impact.

(And let's not get started on the rest of energy intensive things involved in space launch).

Yes, space launch isn't ridiculously energy intensive, but it does put the 777 to shame.

Ok, so with each kg of kerosene burned emitting 3.6 kg of CO2 about 85% of a launch is due to the kerosene. Whether 15% is significant or not is in the eye of the beholder but I don't think my contention was unreasonable. There are, of course, ground crews and radars and tarmacs and so forth associated with a rocket launch but the same is true with a 777 flight as well. Most 777 flights aren't to the plane's max range, this is more of a Pacific crossing fuel consumption but I still think its broadly comparable.

The big difference, I guess, is that a Falcon 9 might only make it through 10 or 20 flights while a 777 might do thousands. That is a big difference in energy usage but I don't feel competent to try to quantify it.

First of all, the last straw that breaks the camel's back. These flights are additional and more are coming.

Second rockets reach higher parts of the atmosphere and the impact is higher compared to the lower flying planes.

The falcon 9 first stage has been tested to be reused 11 times so far. A boeing airframe on the other hand lasts 30 years. There is therefore the cost of bringing these rare earth materials found perhaps around the world together in one place to launch a finished falcon 9 rocket, independent of any fuel considerations, that should also be thought about.
(comment deleted)
These launches are getting SpaceX the money to advance to methane-powered rockets: they will literally make rocket fuel out of air and sunlight, making them absolutely carbon neutral.
Is the manufacture of biofuels actually carbon neutral? What about the massive deforestation of Indonesia - caused by Germany's biofuel mandate?
or is it not a biofuel? are they using the Sabatier reaction so they just need energy input (could be solar)?
Yes, the stated goal is the sabatier reaction from solar energy.

The apparent purpose of this is pretty much a tech-demo for mars. In terms of minimizing environmental impact on earth it would pretty clearly be better just to send the solar energy to the grid and divert natural gas that would have gone to generating electricity to launching rockets.

Does shipping a rocket to Mars with enough propellant to make the trip and land when it gets there affect the balance?

Every rocket to Mars carries hydrocarbon that will never come back.

In context - using solar energy to extract atmospheric CO2 and manufacture methane - yes.

We're not discussing "massive deforestation" here. Germany's biofuel mandate does not affect rockets in Texas.

ISRU (what you're describing) does require a large amount of energy to do at scale. Also, ISRU has never been at the scale required to fully fuel a rocket.

There is a decent writeup on wikipedia[1] on how difficult this would actually be to do ISRU on mars. The TL;DNR is 56,200 meters squared of thin film solar panels. Granted, solar power is roughly half as effective on mars as it is on earth, but still. For reference, an American Football field is 7,140 meters squared.

[1] https://en.wikipedia.org/wiki/SpaceX_Mars_program#Mars_prope...

(Note from the link: that's the requirement to provide fuel for one starship launch every 26 months.)
I think it is worth having a conversation about the actual numbers for carbon cost for launches, vs cost of other internet deployment technologies. Starlink is really cool but I want to know if I'm a fanatic or making rational decisions.
From the press release:

"up to 40 of the satellites will reenter or already have reentered the Earth’s atmosphere. The deorbiting satellites pose zero collision risk with other satellites and by design demise upon atmospheric reentry—meaning no orbital debris is created and no satellite parts hit the ground."

The satellites are deliberately inserted into a low orbit so that any that fail to initialize properly will get pulled down by atmospheric drag. After successful initialization, they travel to a higher orbit (where presumably deorbiting under failure is more of an issue).

I guess I'm being pedantic, but as a percentage of a whole, isn't 2% not a rounding error by definition?
It is if you're rounding to the nearest 5%.
it’s a fraction …. as long as we are using whole numbers
(comment deleted)
Those satellites are small but not infinitesimally small. So yeah, I'd say them going down in whole number units is a safe bet.
Someone driving a venture, that is trying to become profitable will disagree.
With the same logic you could say that covid was a rounding error. Human lives are more valuable then satellites though
Lets at least try to keep this offtopic, inflamatory subject at bay. Please?
Depends who you ask. The insurance company would reimburse your family, for your death, for rather less than the replacement cost of the bird.
This level of pedantry is one of the things I appreciate about HN
HN"."
That was a verytrivial correction.
"very trivial"
The parent username is verytrivial, so I think we can accept the humour and let it go without splitting, don't you agree?
What makes you think that my comment wasn't humorous?
I was just pushing the pedantry forward ;)
Thanks. I could tell it was a joke but didn't get it.
I guess also pedantic but say 50.4 rounding to 50 is a 2% change hence it can be a rounding error.
Depends on the order of magnitude you're working with.
If you earned $100k, that would be $2000.

Would you write it off as a rounding error?

Yeah. I mean, if I was doing something risky.
Hanging out on WallStreetBets? Yes

Paying for lunch? No

I already write off 43k due to taxes, so sure, why not .
Maybe rounding error isn't the best term, but it seems below the threshold to worry about. Here in California, probably at least 30K of it would go to taxes anyway, and I wouldn't know how much within $2000 of accuracy unless I paid a tax accountant $300 to type the numbers into their software. Notably, the IRS requires you to be within 90% of your tax liability over the year to avoid penalties. 0.1*30k is $3000, so the IRS seemingly agrees that $2000 on $100K is below the threshold of being substantial.
(comment deleted)
2% of a big number is still a big number.

62M to launch. 250-500k per satellite.

That's $72M to $82M down the toilet. Given median per-person income is around $36k, that represents over 2,000 man-years of loss.

There are less than 30,000 people out of 7,900,000,000 in the entire world worth more than 100M.

You could buy around 20 windmills and power 15,000 entire homes per year with that money (or around 40,000 homes with natural gas).

What Bezos is paying for his boat could pay for all of that 6 times over.
It might not be the full $62M though. SpaceX often launches other payloads at the same time. I recall one that sent off a couple of BlackSky says at the same time. I don't know what % of the payload that constitutes though, and as I write this out I realize that I'm probably splitting hairs...
https://spaceflightnow.com/2022/01/06/spacex-deploys-49-more...

This had 48 satellites, so definitely a dedicated launch.

The launch I referenced with BlackSky also had 48 satelites: https://www.satellitetoday.com/launch/2021/12/02/spacex-laun...

I didn't find a quick # of lift capacity of Falcon 9 to LEO in reusable mode, but SpaceX has stated >= 60% capacity in reusable mode. LEO in non-reusable mode has about 23,000kg, so there could be a fair bit of headroom for extra mass in LEO reusable mode even with 48 starlink sats. Admittedly, probably not enough to really change the magnitude of the GGP's calculations.

The bigger question I would be asking right now is if it happened this time with 40, who is to say it won't happen with 100 next time and what is being done to prevent it in the future?
If we end up losing 2% of the population due to covid, I look forward to the news briefing that this loss is just a rounding error.
Global death is 1% in a year. And you won’t hear about it, because it’s rounding error.
> SpaceX deploys its satellites into these lower obits

Unintentional pun on "orbits", perhaps?

It's incredible to me that launching satellites is so easy now that a failure of 40 satellites can be tolerated. On the bright side, I guess this gave SpaceX a chance to talk about their debris-limiting strategy!

I wonder who'll write their orbituary.
New York Times

Capital City, Earth – Satellite Starlink XB-4590 has shut down yesterday, February 8, 2132, following a geomagnetic storm that irreparably crippled its circuits. XB-4590 was 5 months old. It is survived by the 32,008 other active Starlink satellites orbiting the Earth. SpaceX CEO Elon Musk IV declined to comment for this article.

My main takeaway here is that Elon Musk's grandson, "∑∫π²∞", decided he was tired of the escalating unicode name game and just named his kid Elon Musk IV.
It would be Elon IV Musk ... if they cared to follow the norm.
I think you misread that. Elon Musk IV is the given name. Her full name is Elon Musk IV Musk. The same way Mario's and Luigi's full names are Mario Mario and Luigi Mario.

Her friends call her IV.

> ∑∫π²∞

This reads: Sigma Integral Pi-Squared Infinitiy.

There are worse names ¯\_(ツ)_/¯

> There are worse names

The last name: Musk.

And he tries to clean it.

https://www.quora.com/Is-it-true-that-Elon-Musks-family-got-...

Dunno - quote from Musk

>‘My dad will have a carefully thought-out plan of evil. He will plan evil. Almost every crime you can possibly think of, he has done. Almost every evil thing you could possibly think of, he has done. It’s so terrible, you can’t believe it.’

Doesn't really sound like he's trying to say all the Musks were wonderful.

In Rolling Stone? Wow. Did not know that.
A high altitude aircraft will take a sample of the atmosphere where XB-4590 burned up on reentry so that it's remains may be interred at the local e-waste processing facility.
Any loss of a mass production sat can be tolerated. Or at least tolerated better than the loss of a bespoke one off sat.

So the loss of a batch of starlink or GPS or glonass is not a tragedy like the loss of Webb would have been.

It's not the mass production that's the issue. It's the mass launch.
I have to wonder if anybody else lost satellites. Starlink is at an especially low orbit, so either they were really unlucky, or they have poor radiation shielding.

[EDIT] OK I made the wrong assumption that this was due to radiation, but it was due to the atmosphere being 50% more dense than usual due to the storm, which led to failure of the satellites. This makes good sense. Next time I'll read TFA before I comment!

It was due to increased atmospheric drag, not a direction result of radiation.
The Starlink satellites are launched into a very low orbit, and raise to their final orbit.

These satellites had just launched, and had not yet raised their orbit, so they were lower than the rest of the constellation.

> SpaceX deploys its satellites into these lower obits so that in the very rare case any satellite does not pass initial system checkouts it will quickly be deorbited by atmospheric drag. ...

> Unfortunately, the satellites deployed on Thursday were significantly impacted by a geomagnetic storm on Friday.

So, while Starlink's final orbit is lower than (almost?) all other satellites, these satellites were even lower than that. I would be very surprised if any other satellites were effected.

This method of parking in a lower orbit is also why the trains of satellites are highly visible for a few weeks after launch. They extend their solar panels horizontally to minimize drag, which causes them to reflect the sun just after dusk and just before dawn. Once they reach their higher orbit they reorient their panels perpendicular to the earth, greatly lowering their apparent brightness.
The page says that they are deployed out of the ship at very low altitudes before they climb to their final orbit. So I think this only applies very early on in the deployment process and wouldn’t impact anyone else, or even existing deployed starlink satellites. They even describe why they start lower than normal in the article ;)
Say what you want about Elon Musk but SpaceX is one of the most inspiring thing in the world to me. I recently drove down to Boca Chica beach to see the spaceships in person: you can get really close to them it's pretty wild, and honestly I think it was the most amazing man-made thing I've ever seen in my life
Is it possible to predict a geomagnetic storm, before launching?
It is, was, they were warned and launched anyway. I assume they calculated they could get through it and didn't. Live and learn. My guess is this was in budget an 'acceptable losses' to the constellation.

Fun site to track this stuff: https://www.spaceweather.com/

(comment deleted)
Woah, I wonder how much atmospheric heating of earth these storms are causing.

(A quick google of NASA suggests: it can be several gigawatt-hours of power from one event, even accounting for the bulk of it being radiated back into space.)

Wonder if there's any way to capture that power?
I thought that it was due to some kind of interference from the title but it is interesting that it turns out to be increased friction from the atmosphere.
I love how they have turned a failure into a success! So cool!
You mean into a PR success?
They are so responsible, aren't you glad NASA is gone now?
Who's gonna clean after SpaceX' now? It's amazing that I'll get a $1,000 fine if I drop a small piece of garbage here in Irvine, but SpaceX can go away polluting the sky indefinitely without any responsibilities!
The constellation is self cleaning to some large degree because of their relatively low orbits. Satellites that die on final orbit have ~5 years before they've reentered on their own, and before they orbit raise right after launch, it can be as low as a couple weeks.

These orbits were an intentional choice by SpaceX. One of intentional effects benefits is to not contribute to space junk concerns.

That still leads to pollution though - and I'm not even mentioning the pollution from their launching!
SpaceX is using their satellites to accomplish useful things for a lot of stakeholders. They also get special permission to allow their satellites to descend to the earth at EOL. These satellites stop orbiting quite quickly after they're no longer functional - they don't orbit indefinitely.

Your litter in comparison would be a totally different kind of problem, and evidently, it is a problem at human scale. That's why there are fines.

Well, I'm talking about small litter like a banana peel, which lots of other Earth inhabitants will actually appreciate. Even if these satellites burn into the atmosphere, they still pollute with the fumes. Not to mention the pollution from the launch itself!
By the way, regarding the littering - I've been doing something I call "distributed composting." For example, if I eat an apple outside of my home, I just throw the remains somewhere it won't be an eyesore, but where it will decompose in no time or better yet - feed some animals or insects. But if a city employees catches me throwing "the litter" into the bushes, they will give me a fine for "littering".
You better try and fine the universe for dumping tens of thousands of micrometeroid debris into the atmosphere every day too!
Well, that's outside of our control - unlike Starlink's junk.
How are you (or anyone) harmed by this junk in ways that the universe doesn't already harm you?
Asteroids enrich our planet with elements and the satellites are already made from existing ones. I'm never gonna use Starlink and not only because it's super pricey, but because the idea is just stupid.
> Asteroids enrich our planet with elements and the satellites are already made from existing ones.

Well the elements from both arrive the same way.

> I'm never gonna use Starlink and not only because it's super pricey, but because the idea is just stupid.

It's cheaper and better than the competitors. If HughesNet and Viasat can serve customers, why can't Starlink and do it better and cheaper than either.

Not cheaper, no - SpaceX is doing a classical dumping and I'm not sure why the authorities are letting this fly! In general, Elon Musk got a lot of privileges from the government while acting as a Libertarian im words, not deeds! He is no hero!
I have a genuine question, are we handing SpaceX a monopoly by allowing them to launch these satellites, or will competing companies be able to launch similar services?

Will space be too cluttered for competition? (I'm guessing space is really big and that we can have _a lot_ of small satellites up there.)

Competing companies are able to launch their own satellites. Amazon is launching Project Kuiper satellites as soon as Q4 2022.
Monopolies are legal, anti-competitive practices performed to sustain monopolies are not (give or take that they are unenforced.)
There's already multiple competitors in late stages doing this style of satellite internet. SpaceX just got there first.
SpaceX granted themselves the monopoly on these large scale constellations by being the first (and so far only) organization to develop (mostly from scratch!) reusable orbital rocket stages. Anyone else (without Falcon 9) launching the same number of launches/satellites would spend several large multiples of the amount SpaceX did, which makes it economically infeasible.

There will be other reusable orbital launch systems (although it looks like #2 will also probably be SpaceX, heh) but until that time everyone else is priced out of doing what SpaceX is doing.

It's got some analogs with TSMC and the latest process size microprocessors. Nobody else is prevented from doing it as well... it's just that they figured it out first and nobody else has developed equivalent technology yet.

We didn't hand them anything. SpaceX took the risk of being a first mover in a very difficult field and looks to have had their bet pay off.

There's room for more satellites. Space is pretty big, and with some careful automated coordination we can avoid collisions. But someone else needs to first figure out how to cheaply launch a competing fleet, and I doubt anyone will come close for years.

I think its a fair question. Handing out unlimited well water rights in the 1860s was also probably seen as not giving away anything. There's plenty of water.
The difference is SpaceX don’t have these orbits forever. The satellites will fall back to earth and congress can enact regulations that limit the number of them launched in the future.
Amazon's Project Kuiper

Behind, but it's there

(comment deleted)
Launching on what, exactly?

Right now they have 9 launches planned using Atlas 5. That's not going to get them a usable fleet of satellites. If they want to launch anything beyond that, Blue Origin must finish the BE-4: a rocket engine years behind schedule. (Either using them for Vulcan rockets or for the New Glenn).

From Amazon's perspective, those 9 launches are money wasted if BO doesn't come through. I'd be ecstatic if they did, but my money is on yet another delay from them.

If I had to guess, by the time Kuiper is launching enough satellites to matter, SpaceX will be doing the same thing using Starship at an order of magnitude lower cost. And I don't even think Starship will be doing real launches for years.

Amazon can launch Kuiper satellites on any launch vehicle, not just New Glenn.
Sure, but who has capacity to take on any serious number of launches? And who can do it at a competitive price to make Kuiper anything but a financial disaster?

The answer is: nobody.

What made Starlink financially possible was SpaceX's lower launch costs and high capacity.

I agree that space can accommodate these satellites, but what about the used up resources required to build the satellites? The shortage of natural resources, as of now, will only worsen and the losses are more costly than the price - its always possible to print more money but its not possible to print more resources.
2000 satellites, each a few hundred kg?

What's the typical daily production of cars in the USA? And how much materials do they use?

There is plenty of room for additional constellations.

The fact that "near misses" are too frequent is largely a result of inaccurate tracking and poor communication. (Next gen space traffic control will allow vastly higher densities.) And the fact that actual collisions are occuring too frequently is due to old uncontrolled junk, not a literal lack of room or inability to maneuver by modern sats.

> Next gen space traffic control

Do you have more info on that? What can the people who track satellites do that they aren't already doing?

It's still up in the air right now. There is a Space Force squadron that does one-way tracking and issues collision alerts, but no transponder standard or centralized traffic control. Companies, once alerted, coordinate individual collision avoidance maneuvers over email. It's ludicrous.

The govt still hasn't decided what agency will take the helm or what the system will look like. It hasn't been worth it to invest more until recently.

https://spacenews.com/from-space-traffic-awareness-to-space-...

Proposal for GPS-based sat transponders, analogous to radio-based aircraft transponders:

https://aerospace.org/sites/default/files/2018-05/GPS_Transp...

Did we hand Tesla a monopoly when we allowed them to build a bunch of electric cars when no one else was willing to take them seriously?
Not but you handed them a monopoly by allowing them to install charging stations with a proprietary connector throughout the country.
What was the non-proprietary charging option circa 2012 when they started building the charging network?

Also, who else were denied license to build the charging stations at that time to allow for Tesla monopoly? Or was it that everyone else amusingly laughed at the cute effort by an electric mobility upstart who didn’t know better?

There are a bunch of small companies building demonstrators right now. Even before SpaceX started Starlink, there were companies trying to do the same thing. SpaceX seems to be the most successful so far but there is still a lot of work to do to build out a network as reliable as GEO comms or something like Iridium.

There are a bunch of different kinds of services offered or soon to be offered by other companies. Some only handle small text messages at like 4800 baud (like Orbcomm) and some others are planning commercial cell towers in space (like AST). Everyone has different needs so not everyone is going to require a blistering fast, low latency, high availability link. So while I applaud SpaceX going for the gold to deliver such service, it's not for most people and is overkill for most applications required a satellite uplink.

Competitors were actively licensed at the same time as SpaceX, the British government (and probably some other governments) is also actively funding competitors, and the US military has made statements that leads me believe that they plan to as well. Rather than handing them a monopoly, it seems like the governments of the world are doing everything reasonably possible to create a competitive market.

I'm not sure it will succeed, it's a "natural monopoly" type of market (high fixed cost, low marginal cost), and SpaceX has an incredible advantage in being really frickin good at launching rockets (though they do offer to sell those services).

There is plenty of room for active satellites, if we manage to clean up defunct satellites and space debris some time in the next decade. The more constrained resource is frequency spectrum.

I think SpaceX might run into trouble if they would refuse to launch sats from a competitor, given their dominant position in space launch. But they would happily launch sats for e.g. Project Kuiper by Amazon/Bezos or OneWeb.

I don't think there will be a monopoly for SpaceX, at least not a global one.

It's a matter of nation sovereignty / media control: governments need local service providers that they can controlled. There's no way China will allow its citizens to freely use SpaceX, just like how Americans probably won't use satellite internet from a Chinese company.

StarLink is probably a failed financial model before it is even completed. They aren't the first-mover, either. HughesNet provides better service at a lower price, and they only have to keep 3 satellites operating at outer orbit.

StarLink is trying to reduce ping by flying thousands of satellites at low orbit. The problem? No matter what you do, the ping will be worse, and cost more money, than broadband.

The developing world will never be able to pay for StarLink at a price high enough to replace thousands of satellites every 5 years. SpaceX will depend on users from rich countries to pay exorbitant fees, and users in those countries already have access to cheaper, faster broadband.

So, it may be that StarLink will go the way of SolarCity and need to be bailed out by another Musk company.

Better service?!?! HugesNet geostationary satellite internet has 600-800ms ping, and maybe 20mbps down; it's barely usable. And it only works in lower latitudes.

Cheaper, faster broadband is only available if Comcast and the other loathsome telecoms can get their acts together, and they have been failing to do so for 30 years.

A constellation of 4,400 satellites, every 5 years, each costing $250k in parts and $50k in launch costs, totals to $264,000,000 per year of opex dedicated to keep the constellation in orbit. A poor country with about ten million subscribers would have to charge roughly $2/month to keep up with satellite replacement costs, but unlike terrestrial utilities, the satellites conveniently circle the whole globe.

You obviously never ever seen rtt graphs for low-orbit and GSO satcom links.

Also, the current prices of either.

> HughesNet provides better service at a lower price

You're either a troll or you have absolutely no idea what you're talking about. As a former Hughesnet customer, now on Starlink I can tell you first hand that literal 56k dialup would have been better than Hughes on many days. Starlink on the other hand is basically indistinguishable from the cable internet I had when I lived in what might be called civilization.

Queue the Joe Rogan episode where we find out what "really" happened.
So…we have the Space Weather Prediction Center [0] and raw data from NOAA for this exact purpose. Does SpaceX not review the data / warnings before launch?

There was a G1 warning posted until 21:00 UTC on the launch date:

Space Weather Message Code: WARK05

Serial Number: 1635

Issue Time: 2022 Feb 03 1428 UTC

EXTENDED WARNING: Geomagnetic K-index of 5 expected

Extension to Serial Number: 1634

Valid From: 2022 Feb 03 0555 UTC

Now Valid Until: 2022 Feb 03 2100 UTC

Warning Condition: Persistence

[0]: https://www.swpc.noaa.gov/homepage

(Warning citation: search for the serial on the warnings and watches page: https://www.swpc.noaa.gov/products/alerts-watches-and-warnin...)

40 of 49 launched were lost, about 80% loss. Perhaps they intend to learn about geomagnetic storms and deal with them better next time.
As far as I can tell, they didn't lose 40 because of the storm directly, but because they entered safe mode before reaching their orbit, and were then unable to exit safe mode because they were too close to earth. So the data, other than "don't launch during a geomagnetic storm" is likely to be minimal.
I'm not so sure. They had the satellites in their "flat" mode and adjusted them to be edge on into the worst of the drag induced by the storm. They effectively tested that profile and have measurement data on how much drag they have to deal with under those situations. That data may allow them to limit damage to the entire fleet with larger storms, or make more informed decisions about future launches.

That data probably isn't worth the cost of the satellites they lost though...

> Preliminary analysis show the increased drag at the low altitudes prevented the satellites from leaving safe-mode to begin orbit raising maneuvers

The increased drag was caused by the storm.

(comment deleted)
That was a minor event to everybody else. I checked the PJM power grid alerts, and they didn't issue a geomagnetic disturbance warning. Their last warning was on 11.04.2021, and it didn't reach a level that affected operations.
> That was a minor event to everybody else. I checked the PJM power grid alerts, and they didn't issue a geomagnetic disturbance warning. Their last warning was on 11.04.2021, and it didn't reach a level that affected operations.

I think it's entirely possible that this type of warning would _only_ be relevant to SpaceX, because their satellites are so low and so small. They're going to notice atmospheric drag to a much higher degree by virtue of size and orbital insertion point, compared to other interested Space Weather parties. G1 is relatively minor for most space weather customers, but there are obviously times when G1 conditions aren't observed that they can launch during too.

"Does SpaceX not review the data / warnings before launch?" is peak Hacker News pettiness. Try not to assume the worst, especially when it's pretty silly.
> Try not to assume the worst, especially when it's pretty silly.

I'm just asking questions -- I would be genuinely curious if SpaceX has a system for determining and evaluating risk when it comes to a situation like this. It does seem absurd that they wouldn't review the active warnings prior to launches. What is indisputable here is that they've announced the loss of 40 satellites during an active warning. I think it's valid to wonder what the criteria is for a launch, given that the cost of the launch + payloads is quite high, and the cost of delaying the launch is (relatively) low.

Certain people high up in the org get irritable over launch delays. Sometimes others think the risk isn't worth the yelling and go ahead in less than perfect conditions.
You're putting an emotional ("get irritated") label on it, but there are no doubt risk vs return evaluations made for these situations. They go beyond just raw "irritation" I'm sure.
I think you’re unduly discounting the influence of emotions on leaders. IMO there are many executives who are for more driven by their personal feelings than we would like to believe.
Presumably you haven’t seen any of the “exited” emails. Or seen some of the petty reasons people get fired for at that company.
I wonder if there was additional unannounced payload. SpaceX frequently seems to launch spy satellites. It may be that other cargo timing took a higher priority…

Also curious if atmospheric warming is a predictable side effect of one of these storms or if we’re just learning about it.

Space weather is much-studied but not always predictable (although I think particular situation wasn't too surprising).
You could have phrased it: "how does SpaceX review...?"
"So…we have the Space Weather Prediction Center [0] and raw data from NOAA for this exact purpose. Does SpaceX not review the data / warnings before launch?" is not just asking questions; it's full-on passive-aggressive.

Being pseudo-anonymous and safely behind a screen depersonalizes the experience enough that we get tempted to take swipes for some reason (I do too), and it diminishes the experience on places like HN especially where we aspire to do better than reddit or /.

The best test for this kind of thing is "If I were speaking to the responsible person face-to-face, would I say it this way?" Because then the question becomes "Do I want to hurt this person?"

spacex isn't a person.
The US government says otherwise

> Who Is A United States Person?

>United States Person. United States person means United States citizens (including minor children); United States residents; entities, including but not limited to, corporations, partnerships, or limited liability companies created or organized in the United States or under the laws of the United States; and trusts or estates formed under the laws of the United States.

https://www.fincen.gov/who-united-states-person

Fine, I'll be direct. Why does it appear that SpaceX failed at risk management in the presence of readily available data to suggest delaying launch?
You don't know what SpaceX's risk models show. They likely have to balance launch delay costs with launch risk costs, among other variables.
Let me refer you to the rational wiki on the phrase "just asking questions" https://rationalwiki.org/wiki/Just_asking_questions

The way you framed your question comes off, to paraphrase the wiki, as an attempt to make a wild accusation socially acceptable. Given your response here, I don't get the impression that this was intentional, but given just your original post I absolutely did get the impression that it was intended as an accusation.

---

To attempt to answer the question, the launch weather report [1] does take into account solar activity, and rated it as a low risk. SpaceX takes a more detailed look at the weather than what's in the public report too. However that report almost certainly only counted solar weather during the launch itself - I don't know of public documentation either way towards them looking at post-launch weather as a launch criteria.

[1] https://www.patrick.spaceforce.mil/Portals/14/Weather/Falcon...

  > I think it's valid to wonder what the criteria is for a launch.
SpaceX operates on the very edge of what is possible. It seems very likely that they saw this as an opportunity to check how viable a Starlink launch is during these conditions. And better to do it with a 40-sat launch on Falcon than on a 300-sat launch on Starship. This storm was an opportunity to learn and to refine the launch criteria.
It is really up there with the HN classic "Does anyone else think the entire multi billion dollar targeted advertising industry is a scam that doesn't work?"
I don’t necessarily think it’s a scam but am rather skeptical of the way that the results get measured and reported.
Thank you! This HN trope is so tiring. It's almost impossible to convince developers as a group, but to convince them one by one is easy.
I still believe this, but am willing to learn. Honest question: Would you mind trying to convince me?
Sure. It's a lot easier when I can swing open a Facebook/Google Ad (etc) account and show you the actual data. I have tried having this conversation online before and people literally say things like 'you're just a shill how can we trust anything you say', as if I would post this sort of thing in bad faith?

It's so weird. Anyway, let me have a go. Initial points:

- Lots of marketing is full of rubbish and smoke and mirrors. This does not mean that ALL marketing is full of rubbish and smoke and mirrors.

- Many companies have marketing 'as a function', but other companies have marketing as quite literally their only source of revenue. These companies would know that digital marketing doesn't work as they wouldn't have revenue. It is that simple.

- I have never worked for a CEO that has accepted smoke and mirrors as answers, and I would never work for a CEO like that. The CEO of my company has a Stanford PhD in Physics. He was one of those child geniuses you used to see in the newspaper, before you stopped reading newspapers. This guy is not an idiot.

Some marketing - a lot of marketing - cannot be directly attributed to revenue, but we know it works. How? Because we can, and do run isolated tests of that exact tactic and test if it works in isolation.

Take the example of banner ads. I have run them in isolation (ie - I have run campaigns where literally the ONLY way you could have known this information was to see a banner ad) and have achieved spectacular results despite noone ever clicking on the ad. (People emailed and called asking to take up the offer.)

I have also have achieved horrifically bad results. Same tactics, but everything else was different - audience, market, placement, product being sold etc.

Every product has a better or worse way of being marketed. It takes a TON of experience to learn how to adapt; many junior marketers think they're experts because they crush it in one job, and then go to the next job and have the exact same tactics fail. Why? Because business is hard. Marketing is about so much more than just running ads.

I run marketing right now for a very young startup. A year ago, we were adding around $5-15k in ARR each month (we only bill annually.) Now, we are adding around $80-100k in new ARR each month. Our marketing spend, inc salaries, is around $40k/m and our sales are growing wayy faster than our marketing spend. The results would be even better if I could convince my CEO to raise our prices.

(I can't complain - at least I got an unsolicited salary increase a few months ago.)

Marketing is the only form of lead generation for us. So much so that I had to plead with the CEO and head of sales to stop relying on my function and to go out and actually set up proper outbound sales processes.

Just because Uber's head of marketing was dropping - and wasting - hundreds of millions on programmatic advertising, doesn't make us all idiots.

/endrant.

(oh. One last thing. This is really the cherry on top: our market is devs. We're literally marketing to devs.)

Thank you for taking the time and attention to write this up. I find your insights valuable, and I tend to agree.

I think my hesitance towards the ad world is coming from my personal experience. I’ve worked at marketing agencies and as digital strategist for a company, so I saw both sides of the chain. My guess is there are a lot of companies that don’t know how to get the highest value out of online ads, and these can be fooled easily by marketing agencies or ad sellers (that a lot of time also don’t really know how to optimize).

I guess as always: don’t blame the tool.

There have been more expensive scams.
Like most advertising before targeting. Not so much a scam though, more a theater performance.
Like you ask this poster to assume the best of the company, I think you should assume the best of the poster.

In my opinion it's pretty clear the jist of his message is "how did this happen when there was a freely available warning"

"Be Curious, Not Judgmental" - Ted Lasso, and (not?) Walt Whitman.

To ask why "Does SpaceX not review the data / warnings before launch?" instead of asking "why did Space-X choose to launch when there was a warning" are two drastically different questions. The first has an assumption that checking for the warning is the problem while the second does not make an assumption of the problem.

I think it's a pretty poor assumption that an organization like Space-X would not check the warnings before launch. They likely have an NOAA rep advising their decision. It's silly to assume a highly regulated multi-million dollar discipline that has little tolerance for failure (space flight) would be so haphazard. Even pilots of small airplanes are required to check government notifications before each flight. So I'm curious, why did they launch when that warning existed?

Based on Tesla’s history with FSD, I don’t think giving any Musk company the benefit of the doubt is warranted.

There are two options:

1) They don’t monitor the warnings.

2) The warning doesn’t matter. Move fast and breaking things.

Both options are bad.

3) The warning was missed due to <fill in the blank process failure>

4) The severity of the situation was misjudged

5) They decided to try a real-world experiment to determine whether their expectations were correct

I suspect some combination of 4 and 5.

Seeing how the satellites were taken down not by the storm itself but the atmosphere slightly expanding to make the low-orbit to high-orbit move physically impossible, I wonder if this is something that hadn’t been encountered before, or is a known phenomenon but with unknown consequences.

Known phenomenon with unknown consequences tends to be the kind of stuff SpaceX likes to hurl headfirst into to see what breaks.

I would be astounded if SpaceX asks NOAA's advice for how to position/maneuver satellites.
(comment deleted)
Your comment qualifies as a thought terminating cliche! At least I come to this forum wanting to think about the kind of questions OP has asked.
I have a big collection of “why don’t they just” comments from HN. It’s pretty amusing to look at them in aggregate.
Would you mind sharing it?
Maybe someday. I wouldn’t want to share it without finding time to write a few paragraphs. This community is really great and I don’t want to criticize without context. HN also over-represents one engineering discipline and one region.

I don’t know how to summarize my feelings, but I work in aerospace, primarily in hardware, in Asia, often as a manager. I see a lot of comments that simplify and often criticize those topics based on the simplification. I’ve also said “just do this” to software engineers so I want to emphasize that this is not a dig at certain types of engineers or people but maybe most of us can do better at recognizing where our expertise starts and stops.

I'd almost agree if not for the fact that the warning was ignored. And that ESA had to move their satellites because Elon Must thinks his satellites are more important than others.

So, what were the reasons? Hubris could be a reason.

They literally lost 80% of the satellites in the last launch. Given that Elon was telling everyone to work over the holidays, this event reeks of gross incompetence.
It was a minor storm. For some reason the atmosphere puffed up more than expected or with a bad timing.
I think they just rolled the dice on that one.

Spacex will probably IPO soon and needs Starlink to look like it could actually make money at some indefinite time in the future.