An entire launch worth of satellites is not a "rounding error."
The carbon impact of just one launch is pretty obscene.
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.
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.
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.
So what should people who dont have access to the ground-based infrastructure do?
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).
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.
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.
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.
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.
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.
This might all seem reasonable in rural America but it’s hugely expensive for internet in Europe.
That's going to average out to less than $10 a month pretty quickly.
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.
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.
Lifespan certainly matters, but the financial structure doesn't.
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.
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".
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.
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.
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 :)
That was discussed elsewhere in this thread. This discussion in this fork is very specifically about monitory costs.
Hence all the dollar signs. ;)
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.
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.
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.
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"
Iridium has been running sat-to-sat communication links since 1998?
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.
Try a single aeroplane?
> 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 entire total carbon emissions of producing and flying the second stage is non-trivial.
I agree, but the post I was responding wasn't talking about CO2 emissions.
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.
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.
Second rockets reach higher parts of the atmosphere and the impact is higher compared to the lower flying planes.
It isn’t really. Checkout Everyday Astronaut’s take on it: https://youtu.be/C4VHfmiwuv4
Same goes for carbon impact, etc.
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?
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.
Every rocket to Mars carries hydrocarbon that will never come back.
We're not discussing "massive deforestation" here. Germany's biofuel mandate does not affect rockets in Texas.
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...