SpaceX Launches 60 Starlink Satellites as Part of Satellite Internet Constellation
60 Starlink satellites begin to separate after deployment from the Falcon 9 second stage. (Credit: SpaceX webcast)
A SpaceX Falcon 9 rocket successfully launched 60 Starlink satellites from Cape Canaveral on Thursday evening, kicking off an ambitious program to provide global broadband services from space.
The satellites separated as a group from the second stage of the booster one hour and two minutes after the on-time launch at 10:30 p.m. EDT. Video from an on-board camera showed the satellites slowly separating from each other at an altitude of 440 km before SpaceX ended the webcast.
Starlink satellites are equipped with one solar array instead of two, minimizing potential points of failure pic.twitter.com/bJirVr67fF
— SpaceX (@SpaceX) May 24, 2019
SpaceX has received approval from the Federal Communications Commission to launch nearly 12,000 satellites to provide global communications services. Founder Elon Musk has said the constellation will be economically viable when 1,000 satellites are operational.
SpaceX designed Starlink to connect end users with low latency, high bandwidth broadband services. With a flat-panel design featuring multiple high-throughput antennas and a single solar array, each Starlink satellite weighs approximately 227 kg, allowing SpaceX to maximize mass production and take full advantage of Falcon 9’s launch capabilities.
To adjust position on orbit, maintain intended altitude, and deorbit, Starlink satellites feature Hall thrusters powered by krypton. Designed and built upon the heritage of Dragon, each spacecraft is equipped with a Startracker navigation system that allows SpaceX to point the satellites with precision.
The Falcon 9’s first stage successfully landed on SpaceX’s “Of Course I Still Love You” droneship in the Atlantic Ocean. It was the third launch for the stage, which previously supported the Telstar 18 VANTAGE mission in September 2018 and the Iridium-8 mission in January 2019.
53 responses to “SpaceX Launches 60 Starlink Satellites as Part of Satellite Internet Constellation”
Leave a Reply
You must be logged in to post a comment.
They’re phoned home…
https://twitter.com/elonmus…
Hopefully the beginning of (well, foundation for) a new revenue stream for SpaceX and global internet for the rest of us Earthlings. Good luck to the SpaceX team; so far, so good!
Congratulations! Another home run for the folks at SpaceX!
Still think Dish type broadcast is better. These 60 could be spread out all over Earth. In the 5 min passover, it dumps compressed signal to HD ,millions of receivers. Next one comes by and dumps again the way NASA does with there probes. MP-4 would work. Don’t know how many channels however. Dish has 500 channels or so. These little sats may not be able to have that many. What ever is possible. Charge $50 rather than the $120 I am paying. This has really got me thinking. I hope I got it right. Doug can patent this and sell licenses and maybe I can have the IM mission using my SRB chicken launcher. That is all I want Doug. Go for it.
WTF? If prescheduled broadcast video is so much better than internet access, why are you posting here instead of vegetating on a couch in front of a TV?
Dish has no need for the low latency of low altitude constellations or the bidirectional point to point traffic Starlink is designed for. There’s no need for it to bother with tracking antennas of any kind. Conversely, it’s almost totally useless for the applications that Starlink is built to handle.
12000 sats / 60 sats/launch = 200 launches. If they can somehow do 10 launches for just this system per year that’s 20 years to constellation completion. But wait! These sats don’t live for 20 years. So the first sats will be long dead by the time the last are deployed, so you’re going to need a lot of replenishment missions to keep the constellation populated. SX better hope SHS works… Without it, this system never makes it to completion.
*Note: This post uses real math, as opposed to fan boy math 🙂
Note: The fully operational satellites may actually be larger. The satellites just deployed are a developmental version of the Starlink satellite, specifically, a version without satellite-to-satellite comm link such as is planned for future versions.
Meanwhile, as a developmental version, the current satellites are intended to de-orbit and be replaced by a more capable iteration before extensive time passes. They were placed in lower orbit to, in part, facilitate de-orbiting after a relatively short lifespan. At his recent conference call on Starlink, Mr. Musk explained that, ideally, new technology will make older Starlink satellites obsolete in about five years, at which point SpaceX would deorbit them and launch new, more capable versions.
Also, there will be no “by the time the last are deployed”, or at least not in terms of while the constellation is operational. The idea is to have them low to minimize debris potential and to reduce latency, while continually replacing the oldest ones. For such a strategy, I t helps to have your own fleet of [re-useable] launch vehicles
Finally, regarding the “12,000” number, quoting a story reporting on Musk’s conference call on Starlink, “Musk said Starlink will be able to start partial service with around 400 satellites, and will reach “significant operational capability” at around 800 satellites. Once the constellation reaches 1,000 spacecraft, it will become “economically viable,” he said…”The constellation doesn’t need 12,000 satellites to constitute a win, he said, but that would mark a “very successful outcome.”
@Lee wrote:
You are presuming SpaceX will completed the Constellation with the Falcon family. When the Starship enters service, SpaceX will have the capability to dump from 600 to 1800 or more Starlink satcoms per launch.
AIUI the Starlink Constellation requires about 20% replenishment annually after full deployment. Which works out to 3 to 4 Starship flights.
Of course presuming the SpaceX Starship enters service around 2022.
As I said in my original post, SX better hope SHS works.
I’m sure they do. I know I do.
That’s not very many Starship flights to spread the costs of developing it across. Is it supposed to be really cheap to operate once the development costs are paid for?
That won’t be the only things Starship will be doing.
At least initially, I think it will. If SHS enters service in late 2020, I think it would be in freighter trim and used exclusively for Starlink deployment launches until late 2022 except perhaps for an unmanned development launch or two of the initial passenger version. After that, we’ll see the passenger version fly to and from Earth orbit with more and more people aboard. Sometime in 2023, Mr. Maezawa and friends will get to do his Dear Moon lunar flyby. Not long after that, I foresee initial missions to prove out Starship lunar landings and takeoffs. Then there will be freighter missions to pre-position cargo to be used by the first human missions in more than a half-century, then the first of those missions.
Not to mention the fact that it’s going to be hard to get 1800 sats to the right orbits from just one launch. But I guess the laws of phydics in general and orbital mechanics in particular don’t apply to elon…
Sure they do. But Starship should be good for some different altitude burns and perhaps even modest plane-change burns so it wont have to drop off its entire load of birds in one place.
There is also the possibility of a deployment bus carrying a few dozen sats and enough V to send them to another plane or orbit. Vac engine, tank and dispenser with deorbit on completion. If several planes or altitudes are needed on a given flight, it could save an enormous amount of propellant on a mission.
It could be developed in a short amount of time compared to most craft and could morph into a general purpose orbital tug.
Possible, I suppose, but I don’t find that too likely.
It would come down to economics, convenience, technical feasibility, and a few other variables. It’s fairly easy for me to sit back and suggest possibilities for other people.
Whether it would be feasible or desirable might come down to what the margins are in the vehicle. My thought would become more compelling if they were really stretched to the limit on a mission and needed just a little more help.The final mass of SS would be a factor which might make it desirable to avoid numerous plane/altitude changes with the whole ship.
Eventually when SpaceX have orbital refueling capability. They could send a Starship loaded with comsats to waiting orbital Starship tanker(s) to top off the propellant tanks. That increases the number of plane changes and orbital height adjustments for the Starship carrying the comsats. If required additional tanker flights could be added later.
A lot of the development costs of SHS are already paid for – especially those attributable to the Raptor engines. Remaining development expenses will not be trivial, but are well within SpaceX’s ability to pay for on an as-incurred basis.
The goal for SHS is that it will cost SpaceX less to fly it on an orbital mission than it did to fly a Falcon 1. If the variable costs of a launch are mainly the cost of propellants, that seems quite achievable.
SHS flights made for deploying/upgrading Starlink will not help defray the costs of SHS development as they will be paid for by SpaceX itself. The cost of any individual SHS Starlink deployment launch, though, will be dwarfed by the pro rata future revenue that will be made from the satellites deployed.
Both SHS and Raptor will continue to be improved after SHS reaches initial operational capability just as have the F9, FH and Merlin. SpaceX is going to want to build out a fleet of Starships, especially if lunar-directed business ramps up and/or the terrestrial point-to-point service spitballed in 2017 is inaugurated. In the latter case, quite a few new SH boosters will eventually be required as well.
If Starlink can actually start generating revenue with only 400 sats deployed, that means the money could start flowing in as soon as early next year and might grow to fairly substantial figures by next year’s end. It seems quite possible the 1,000-bird deployment level will have been reached or even surpassed by then.
SHS may well be in early service by then too. Its first two years in service will likely be spent exclusively on Starlink deployment while building experience and confidence for later more varied missions, especially those involving carriage of human crew and passengers.
The internal volume of the SpaceX fairing is about 146m3, and the internal volume of Starship is to be about 1100m3. Ignoring the impossible to calculate at this time differences between Starlink satellite stacking inside those two different volumes, and taking 1100/146 * 60 implies that SS will be volume constrained to about 450 Starlink satellites. A further complication might be that there could potentially be more than one model of satellite depending on the altitude. The “original” SpaceX plan had constellations at MEO, LEO, and VLEO, all with different working lifetimes.
Guessing launch costs for F9 to be about $20M ($5M for booster per launch, $10M for second stage, $5M for launch/recovery/refurbishment costs), the cost of launching 1000 satellites is about $340M, so about $4B for 12,000 satellites. With the system being economically viable at 1,000 satellites it is quite feasible that the whole constellation could be launched by F9.
I think your estimate of launch costs is high by roughly a factor of four so putting all 12K birds up with F9’s would run $1 billion not $4 billion.
But I also think SHS, which will cost even less per launch and put up a lot more birds at a whack, will be doing so for the first time before the end of next year. Even if Starship requires an orbital refueling to deploy a full load of birds to multiple orbital altitudes/inclinations, two SHS launches – one freighter, one tanker – would still come in below the price of a single F9-based deployment run.
Putting up all 12,000 birds with F9 would require 200 launches. In order to transition directly from initial deployment to incremental replacement, that means 40 launches per year. Even if SpaceX can use Vandy for some of the missions, it looks to me as if it still might not be doable from Canaveral soon enough. Fortunately, if SHS arrives on the schedule I think it will, that won’t be a problem.
I was primarily taking issue with the 1800 sats per launch guesstimate, and then taking a stab at a worst case scenario. With that said, your estimate of $5M per F9 launch, recovery, and expended second stage, strikes me as a tad optimistic to say the least, but somewhere in the $10-15M might well be possible. But if SpaceX can get the average cost per satellite to $250K (surely achievable and even much less once a production line is in place), that’s $3B (or much less) for the constellation hardware. Estimates of a recurring F9 launch cost (to SpaceX) notwithstanding, I am probably at least as optimistic as you regarding their initial and ongoing launch costs – I suspect Elon’s early estimate of $10B for the entire system is now looking decidedly conservative – more like half that. The earliest possible arrival of SSSH will obviously make for an even rosier outlook, and I think they have till 2026? to put the full constellation in place, so a very early arrival is not critical. All in all I see little to no room for doubt or pessimism for Starlink’s prospects.
Is you 1100m3 number for volume aboard the Starship from wikipedia? If yes, that is from the 2016 iteration of the ITS Spaceship. The 2019 Stainless Steel Starship iteration might have more internal volume, never mind the chomper satellite delivery variant of the Starship.
I got the 1100 from my confused memory, so thanks for the correction. ITS was 12m diameter for that 1100m3. The 9m Starship will not be chomping up more internal volume, it will be less at around 825m3 (source: SpaceX BFR presentation), leaving space for 825/146*60= 340ish satellites. But my main take away point remains – SpaceX do not need to rely on having SS/SH to complete Starlink, though it would certainly make the task easier and cheaper. As soon as they have enough satellites lofted to reach economic viability, which may plausibly occur before the end of 2019, the launchers for the completion and maintenance of the constellation is only a side issue. However, I am personally optimistic that we might see SS/SH launching around a 2020-21 time frame and probably before New Glenn, Vulcan-Centaur or SLS.
No matter how you look at it, Starlink never makes it to the full stated constellation size of 12,000 sats without SHS. THAT was my point. My point had nothing GB to do with when the constellation reaches economic viability…
Unless you look at it with 30-35 F9 launches a year, then it could be done even without the highly likely early arrival of SS/SH. The cost of an F9 is relatively low to SpaceX and a profitable Starlink pays that cost easily. And that was my point.
I don’t see there being that much range capacity, in addition to all the other US launches, anytime soon…
If USAF continues to run Canaveral, you’re probably right. If, say, Space FLorida takes over management of CCAFS, though, things could change a lot and in a hurry.
I don’t think it has much to do with who runs the range. One of our grads (who you would classify as a welfare recipient, since he was trained by astronomers) works in a key position on the Eastern range. One launch per week is about all the range will ever be able to handle. And you’re not gonna be able to turn around one of the SX pads that fast.
With SpaceX using their GPS based Autonomous Flight Safety System, (AFSS) the range doesn’t need days between SpaceX launches to reconfigure the radars and telemetry. ULA Vulcan is planned to use similar from 2020.
SpaceX pad turnaround might be the bottleneck – anyone know how quick it’s been done ?
Right now the bottleneck is first stage turnarounds. They still haven’t turned one around in less than something like 76 days.
Eastern range, in 2018, was aiming for “2 in 24” ie. two launches in 24 hours occasionally. They may have given up on that, but that implies there is no fundamental limit of one launch per week ?
SX have two pads at KSC so they could do two a week (since they use AFTS) ?
I’d bet sooner than 2022. Judging by Musk’s latest tweetstorm; Boca Chica TX is building Starship Mk1 and Mk2, Cocoa FL also has two occupied build stands, and Super Heavy #1 is not starting for 3 months.
This indicates there are 4 Starships under construction.
This is very encouraging. With 4 hulls, the loss of even 2 or 3 in testing is survivable as long as telemetry reveals the problems. Hoping for the best but allowing for the worst is sound development strategy. It seems more likely that they will come out the other side with operational vehicles if they are allowing for losses.
Always try to build right the first time, but realize that the real world will inform one of mistakes. That is one of the failings of several other companies that they think they are going to get perfection on the first try by spending years on design and simulations only to have a single flaw ground the project for more years. How long would the hiatus for SS2 have been if they had had at least two operational airframes?
Do we still think think there are 4 Starships in construction ?
I thought Boca Chica had Starhopper and Starship Mk1, and Cocoa FL was building Starship Mk2 (for a total of three).
These are v0.9 test satellites testing the radios, 2 methods of solar panel deployment, satellite deployment and getting a basic constellation started.
These will be followed by v1.0 satellites launched on F9 and the v2.0 satellites launched on Starship. If you think Starship will only launch 60 at a time I have this bridge….
Read my post. The part where I said they better hope SHS works…
Guess they are able to raise money after all…
https://www.cnbc.com/2019/0…
12,000 satellites and that is just SpaceX’s constellation. Does this not bother anybody? Are we going to need designated corridors to actually launch things like people and hardware off the planet?
I need to see how this is going to effect ground based observations. Astronomers are going to have to keep local ephemeris of these satellites so we can schedule our exposures to avoid them and our moving object detection to ignore them.
Or Space telescopes are going to have to get a *lot* cheaper, and higher up on the Earth/Moon gravity well. Actually, both will probably happen.
I’ve been meaning to construct a cost of running ground telescopes using the 40 years of Spacewatch operations to construct one term in the space based vs ground based argument. If maritime launch vehicles become a reality, I think that’ll put the nail in the coffin for 6 meters and below. Esp if we can come up with a common bus that would establish a standard to build to. It’s probably about 20 years out. Which would run out the current active ‘small’ ground observatories in the United States.
Sorry, but it’ll never happen. Our observatory has 6, 14, 16, 17, 18, and 32 inch scopes, all with modern Instruments, all automated. Not counting my salary (because you’d need at least one of me to run that many space telescopes), we run this facility for <$50k/year. Most of that is for upgrades and repairs to various systems, which you can’t do to a space scope. Even if launch is free, the cost of running a space scope will always far outpace a reasonably run ground facility.
I once asked a well known solar astronomer why we still spend the money to go on eclipse expeditions all over the world when we have satellites. Two reasons. First, you can image deeper into the corona than you can with a coronograph. More importantly, the cost of one space mission is more than the total cost of every total solar eclipse expedition in history…
Many predicted the death of small groundbased observatories when the Mayall came on line. It never happened, because there are too many things that need doing that are too expensive to do on a 4m or larger scope. The same will be true of space scopes.
You are certainly correct that ground-based scopes of all sizes will not be obsoleted by the advent of even relatively inexpensive and numerous space-based scopes.
Mirror size for mirror size, space-based scopes will still cost a lot more than ground-based ones but that doesn’t mean a 2- or 3-order of magnitude reduction in cost from Hubble levels isn’t possible. And someone out there will certainly design the space-based equivalent of a Dobsonian scope.
Space-based scopes can be designed to be upgradeable and repairable too. Use of telerobotics rather than human astronauts to do these jobs will keep maintenance costs affordable to any institution financially capable of fielding a space-based scope in the first place.
Telerobotic repair in space will always be more expensive than me driving 20 miles each way and spending 1-2 hours replacing a failed component.
No argument. But if steadily heavier economic use of cis-lunar space eventually renders Earth-based astronomy impractical, it’s not like anyone who still wants to do astronomy is going to have a lot of choice. Best to be thinking up ways to make that as palatable as possible. Digging in your heels and imagining you can keep space from being economically exploited in all sorts of ways is just going to get you a pair of broken ankles.
If these sat constellations render ground based astronomy impossible, small college observatories (like ours) will cease to exist. We will never be able to afford any sort of space based astronomy. Ever. That is a shame, since we have graduates who are at Gemini north, Hubble, TMT, and several other large observatories. For a small university in the mountains of NC, we punch WAY above our weight class.
And it really pisses me off that you refer to astronomers as a class of welfare recipients. Do you have any idea how many technologies critical for national defense astronomers have either invented or been key in developing? Have you ever heard of night vision, astro navigation, or precision timing (atomic clocks). You can thank astronomers for all those, and more…
Space based astronomy costs will come down to something affordable to your university if the prices also come down enough to create a problem for you. At $100.00 a pound launch price, a ton of observation platform can be placed in orbit for $200,000.00 plus the cost of the instruments and bus. Not at all out of line with a new terrestrial observation platform and possibly cheaper when land acquisition, infrastructure, and buildings for weather protection is considered.
I would be curious as to how much it costs to build a particular size of terrestrial observation capability. Clearly, building new is a hit to using existing equipment. I just don”t have any idea how much of a hit, and how much mitigation is possible with programming and cleverness.
Also, with the advent of cubesats and such, it seems likely that a fairly light observatory might be possible and relatively inexpensive. Not cubesat, but definitely sub-ton masses.
Let’s look at this. Note that there is a lot of hand waving below, but even hand waving makes clear the cost of a space based scope compared to a ground based scope. For a space based scope, you need:
1) The scope itself. The costs of building a scope that can operate in space are quite a bit higher than the cost of the equivalent ground based instrument and always will be, due to the extra equipment required to operate in space, and to actually survive in space more than a few years. So it’s already more expensive. By how much, I don’t know, but let’s say a factor of two (which is really, really conservative). A one meter telescope, ground based, costs around $600,000 -> $700,000 instrumented. So let’s say around $1.5M for a space based version with one instrument. This price does not include things like solar panels, pointing sensors, or thermal blankets.
2) Launch costs. Your one ton estimate is probably 1/2 the mass of a space qualified one meter scope, so let’s believe Musk or someone will get launch costs down to $100/lb. That means launch costs are closer to $400,000.
3) Cubesats ain’t gonna do it. Or small sats. For optical astronomy, you need aperture. Period. You can’t get around the pesky laws of physics. So you’re left with launching something that will gather enough light to do the science you need to do. For most university observatories, that’s somewhere between 30 and 40 inches of aperture.
4) We change instruments on our scope a couple of times a week. Other observatories do the same. You can’t change instruments on a space scope unless you come up with a Hubble-like instrument bus. But now you’ve probably at least doubled your launch mass, so now launch costs are up to about $800,000. You’ve also likely doubled your scope cost to $3M.
5) Instead of paying ~$100/month for broadband to ship our data back to campus, you’re going to have to find a way to ship your data back from space. I know you will say Musk will just let us use StarLink for free for this purpose, but there is no guarantee of that. His antennas will also be pointing the wrong direction to communicate with a space scope above the constellation.
6) Pointing. You’ve got to add on sensors to allow you to point the space scope precisely, and keep it pointed stably. I have no idea how much this would cost, but it ain’t free.
7) Power. I can’t just call up the local power company to connect my space telescope to the grid. Gotta add solar panels. This is also not free.
8) Other systems like thermal protection blankets. Also not free.
9) Lifetime. Our 32″ scope is 25 years young. I fully expect it to still be operating in 100 years. There are many ground based scopes that are over 100 years old that are still doing useful science. Space is a hard environment. Yes, Hubble has lasted almost 30 years to this point, but it’s also had several servicing missions. For a one meter class scope, it won’t be worth servicing it. So now, after say 10-20 years, you get to start all over at step 1) above, but everything will be more expensive the second (and third, and fourth, and nth) times. But wait! If even a small critical component fails, I can have it replaced from my spares in less than a day. If I don’t happen to have a spare, it might take a week or two. If my scope is in space, it’s game over. However, if I design my space scope with multiple redundant systems, I’ve just increased my costs by at least another factor of two, so my scope is now at $6M. And a failure could still turn the scope into a pile of junk.
So no, there is absolutely no way a space based scope will ever be remotely as affordable to smaller colleges and universities as ground based scopes are.
Also, please resist the temptation to look at my numbers for a one-meter class professional telescope and compare them to amateur grade telescopes. The latter work fine for casual observing, but they can’t hold up to the use that professional observatories require. The numbers for amateur scopes also don’t include the cost for instruments.
Finally some here seem to think that astronomers are swimming in piles of unattended cash. While that may be true for some of the bigger places, those of us at smaller colleges and universities have to work hard to be able to do what we do. We use recycled computers and buy metal from scrap yards to make instruments. There is not one new-bought computer at our research observatory. So yeah, we do rely on tax dollars, but we make every penny count.
You apparently get your hands dirty to make things happen on a limited budget. The ones doing your job in a few decades will be the same type of person, but will have figured a way around the obstacles. Either workarounds on the ground, or problem solving in space. Either way, astronomy will continue. Unless that whole community refuses to adapt which is quite unlikely. Quick checking suggests the mirror will be under a quarter ton, so there is room for improvement in the auxiliary mass.
Sounds like a job for AI, scheduling the optimal observation windows then telling human observers what they are. Or programing the telescopes to edit them out of the images.
Another fantastic show from SpaceX showing how it should be done. Yet another vertically integrated enterprise by Musk takes flight.