SpaceX Super Heavy Starship makes Starlink Satellite Network Cheap

by

We do not know any details about the radical change to the SpaceX Spaceship Super Heavy (aka BFS/BFR) design. However, if the design enables SpaceX to hit the $2 billion lower-end development cost then SpaceX has the money to develop Starship Super Heavy and the Starlink satellite network. It would only cost about $3 billion if everything works out.

Every strategic and design decision from SpaceX is to speed up getting to a fully integrated and complete BFR/BFS. The recent mini-BFR is for low-cost testing on the Falcon 9. Stopping any second stage reuse tests for the Falcon 9 is not needed for the full Super Heavy.

Funds for Starship Super Heavy and Starlink

SpaceX has the following funds:
* Estimated $500-800 million from the Japanese billionaire (for the trip, co-promotion and some part of the company)
* $750 million from the loan.
* Some amount of profit from 40 – $60 million launches. $2.4 billion in revenue. I think maybe 30-40% operating margin. $800 million per year from launches. Note there are not forty Falcon 9 launches but any NASA, military or Heavy launches count as multiple based upon the higher price.

Two years of profits would mean about $3.05 billion available.

SpaceX might be able to get lucky with shoestring development costs to get to 1600 Starlink satellites and a working Starship Super Heavy. A working Starship Super Heavy (aka BFR) means full reusability and even without a huge increase in the number of launch customers boosts operating margin up over 90%. Four Spaceship Super Heavy at $350 million each would be a total cost of $1.4 billion. Four Spaceship Super Heavies could handle up to 200 launches per year by launching every week. 60 of the launches can be used for Starlink launches. The rest of the commercial satellite market would take time to scale up.

First 1600 Starlink Satellites for about $60 million in launches

The first 1600 satellites would be able to provide profitable premium data connections between New York, London, Dubai, Tokyo, Shanghai and other financial centers.

Light travels 45% faster in vacuum versus fiber optic cable. Lasers for Starlink satellites will enable fast low latency connections.

The top ten financial cities could spend $2 billion each per year for premium low latency connections. The next ten could spend $1 billion each per year. The connections would be two each of the other major financial centers.

From 2013-2018 over $500 million was spent over five years on microwave connections to reduce latency between New York and Chicago.

A 3-millisecond decrease in one-way communication time between the Chicago and New York areas was worth about $100 million per year. Chicago is the 17th ranked financial center.

It would take 60 launches of SpaceX Starship Super Heavy (aka BFS/ BFR) to launch about 12000 Starlink Satellites. Each Starship launch would deploy 240 Starlink satellites. If it costs $10 million to launch the SpaceX Starship, then it would cost $600 million to launch all the Starlink Satellite network. It would cost about $40 million for each partially reusable Falcon 9 launch for 20 Starlink Satellites per launch. This would mean 600 Falcon 9 launches at a cost of $24 billion. Completing the Starship Super Heavy would make deploying the Starlink Network 40 times cheaper.

It would only take seven launches of the Starship Super Heavy to deploy the first 1600 Starlink satellites.
This would be about $70 million in launch cost. $350 million for one Starship Super Heavy would be enough for the seven launches for 1600 initial Starlink network. The cost is less than the $3.2 billion to launch the first 1600 satellites using Falcon 9. $2 billion in development cost plus $350 million for one rocket and $70 million for seven launches. There are some estimates that mass production of small low earth orbit internet satellites could drop to $100,000 each. This would mean $160 million for all of the first satellites. Even at $400,000 each, the cost would be $640 million. $3 billion could get SpaceX the working Space Ship and the commercial viable phase 1 of the Starlink network.

The Commercially viable Starlink Network then starts generating $2 to 10 billion per year from premium low latency connections for the financial centers of the world.

Starship Super Heavy going to orbit should unlock more money from NASA and military and private investors. It should finally kill the SLS (Boeing Space Launch System) and hopefully free up $4 billion per year.

It is clear why the Starship Super Heavy is the key to enabling massive profit margins and low-cost launch of the Starlink Network.

41 thoughts on “SpaceX Super Heavy Starship makes Starlink Satellite Network Cheap”

  2. 2024 has been mentioned as the date for Starlink to be fully operational. With the Starship Super Heavy now due to eclipse the Falcon 9 does anyone have an idea when Starlink will now be operational?

  5. “Space” and “orbital space” are two different things. While the volume of space that satellites orbit in is certainly large, it is *not* infinite. Every satellite one lofts is both a potential target for, and potential source of, clouds of space junk. And because the creation of another cloud of space junk can itself lead to further collisions and the creation of yet more space junk, the potential for an exponential expansion of the quantity of space junk exists.

    The math that Dr. Kessler laid out accounts for the fact that there is a “Big Sky”, but concludes that if the amount of space junk surpasses a certain critical value, we don’t know exactly what the threshold value is, just that it exists, then a Kessler syndrome expansion of space junk will occur. This can potentially render low Earth orbit unusable until enough of the space junk decays out of orbit to bring the quantity below the threshold value.

    Pretending that this problem exists or that the capacity to tolerate additional space junk and potential targets of space junk is infinite, won’t make the problem go away. This is why I’m worried when plans speak of an additional thousands of satellites. Even if the operation of those satellites is exemplary and there is a plan to insure that each of these is de-orbited when it reaches the end of its life, the very presence of those satellites gives space junk another possible thing to hit (traveling several times the velocity of a rifle bullet).

    What is perhaps the most disheartening part of all of this is that the purpose of this system appears to be the facilitation of High Frequency Trading (that’s why the big deal about reducing latency). The Utility society as a whole gets out of HFT is dubious (and, if what Lewis suggests in “Flash Boys” is true, may even be negative). This seems to me to be a pretty poor justification for the kind of satellite constellation which could potentially cause us to lose access to orbital space via Kessler Syndrome.

  7. I’m not an expert on orbital mechanics, but have been told by others that the Texas launch site doesn’t work for targeting the orbital planes of Starlink without a lot of waste and likely in-orbit refueling. Looking at articles on the Boca Chica site, it appears to be appropriate for geosynchronous and moon and interplanetary launches with a constrained easterly launch being all that they can do without launching over land.

    Since that is the only launch facility they are working on so far, there is at least that indication that they are not planning on using the Super Heavy for any Starlink launches until late in the game when they have a Florida or other launch site developed, possibly even as late as the start of replenishments in the 2025 timeframe.

  9. The goal of Planetary Resources was to build hundreds of small uniform spacecraft with capacities not unlike Starlink (laser communications, etc.). Since PR was just bought by a cryptocurrency company, I suspect Elon (or Google or anyone else building a massive fleet of satellites) could pick up the IP for a song later and go into asteroid mining as well.
    Starship could one-up the whole Asteroid Retrieval Mission NASA abandoned and grab a much larger asteroid to bring to Earth. They would need a cargo Starship and possibly a crewed one to keep it company. Or they could just go grab a few tons of samples of one and put it in the cargo bay.

  10. At these prices, it might be time for this.

    Remove the @

    nextbigfuture.@com/2010/03/superconductor-and-antimatter-bootstrap.html

  13. Given the gradual deployment and replacement, I think the strategy is that the 1st satellites act as beta versions for the next ones. Get a minimal product up as soon as possible, then iterate in subsequent batches. So I still see step-wise iteration with Starlink too.

  14. He means the step-wise iterations of the starlink satellites that have been tossed out. It is rumored that SpaceX let some starlink executives go over disagreements on how many test satellites they would need before they went into production of the fully operational fleet.

  16. Where did you see tossing out stepwise tests? (Note that I didn’t read this article in detail, so if it’s mentioned here, I may have missed it.)

    AFAIK, the hop and orbital tests are still on. And they’ve recently announced they’ll use a modified F9 2nd stage as a scale model to test reentry etc. That doesn’t sound like tossing out anything to me.

  17. Don’t forget gamers…satellite internet has TERRIBLE latency and many people live in places where there is little to no broadband.

  19. Nice (if highly optimistic) analysis. Starship Super Heavy is also Starship Super Risky … humanity’s greatest space project for fraction of the cost SLS has many tech hurdles never even explored in aerospace. This “speed up” … tossing out stepwise tests … reduces my estimate of success from above 75% to maybe 25%. The idea of this costing just $2B (down from $5B) to get to operational placement of Starlink (also highly risky) seems un-doable and pointlessly fast. I have every hope for this … but this change is very risky. This could kill a company that might of created greatness on the old-but-still-crazy-fast-elon-musk-timeline.

  20. And they’ll need it, given the shelf life of these satellites is rather short, like 6-7 years and they’ll need yearly replacements in the thousands.

    The manufacturing of these satellites will be a permanent feature of the Starlink company, having to produce them by the thousands too. This can bring in economies of scale and production in series, resulting in the first actual mass produced satellites.

    And it can result in the network upgrades becoming much easier, almost trivial. The satellites will go down predictably and a batch of new ones will take their place, just make sure they have the latest upgrades.

    Starlink will definitely have versions and changing specs through time.

  21. I’ll be signing up as soon as it is available to lowlifes like me. I am SO looking forward to calling Verizon to cut my services and explain to them (recorded call for training purposes :)that it is a direct result of them hiring a lawyer/lobbyist like Ajit Pai and net neutrality going away. My whole family and all my friends will be joining me once I test the service, and I will push it to as many people as I can. The FCC and Ajit got something like 20 million emails protesting the death of net neutrality and they did not care. Let’s see whether they care to have 20 million paying customers follow suite.

  22. They’ll be sitting in LEO, so there’s a lot of protection from the magnetosphere during solar flare/CME events, and if the problem is big enough to hurt LEO satellites (ISS, etc), then all worldwide satellite operations out in GEO will also be impacted.
    Second, if Starlink can convince the big banks/markets that their service will be reliable and dependable, you’ll see the banks/markets move to LEO communications and use microwave and fiber for backup. Even cables get cut, so they probably have a variety of backup data paths to carry the load (if a bit slower).
    The only question I have is whether the switching from ground/satellite to satellite/ground (and possibly multiple satellites in between) will actually be faster than current cable/microwave networks.
    If Musk uses the banks/finance markets to pay for the first increment and then open Starlink to the rest of us, he’ll have a cash cow he can milk for trips to anywhere in the solar system.

  24. “Space” and “orbital space” are two different things. While the volume of space that satellites orbit in is certainly large, it is *not* infinite. Every satellite one lofts is both a potential target for, and potential source of, clouds of space junk. And because the creation of another cloud of space junk can itself lead to further collisions and the creation of yet more space junk, the potential for an exponential expansion of the quantity of space junk exists.

    The math that Dr. Kessler laid out accounts for the fact that there is a “Big Sky”, but concludes that if the amount of space junk surpasses a certain critical value, we don’t know exactly what the threshold value is, just that it exists, then a Kessler syndrome expansion of space junk will occur. This can potentially render low Earth orbit unusable until enough of the space junk decays out of orbit to bring the quantity below the threshold value.

    Pretending that this problem exists or that the capacity to tolerate additional space junk and potential targets of space junk is infinite, won’t make the problem go away. This is why I’m worried when plans speak of an additional thousands of satellites. Even if the operation of those satellites is exemplary and there is a plan to insure that each of these is de-orbited when it reaches the end of its life, the very presence of those satellites gives space junk another possible thing to hit (traveling several times the velocity of a rifle bullet).

    What is perhaps the most disheartening part of all of this is that the purpose of this system appears to be the facilitation of High Frequency Trading (that’s why the big deal about reducing latency). The Utility society as a whole gets out of HFT is dubious (and, if what Lewis suggests in “Flash Boys” is true, may even be negative). This seems to me to be a pretty poor justification for the kind of satellite constellation which could potentially cause us to lose access to orbital space via Kessler Syndrome.

  26. I’m not an expert on orbital mechanics, but have been told by others that the Texas launch site doesn’t work for targeting the orbital planes of Starlink without a lot of waste and likely in-orbit refueling. Looking at articles on the Boca Chica site, it appears to be appropriate for geosynchronous and moon and interplanetary launches with a constrained easterly launch being all that they can do without launching over land.

    Since that is the only launch facility they are working on so far, there is at least that indication that they are not planning on using the Super Heavy for any Starlink launches until late in the game when they have a Florida or other launch site developed, possibly even as late as the start of replenishments in the 2025 timeframe.

  28. The goal of Planetary Resources was to build hundreds of small uniform spacecraft with capacities not unlike Starlink (laser communications, etc.). Since PR was just bought by a cryptocurrency company, I suspect Elon (or Google or anyone else building a massive fleet of satellites) could pick up the IP for a song later and go into asteroid mining as well.
    Starship could one-up the whole Asteroid Retrieval Mission NASA abandoned and grab a much larger asteroid to bring to Earth. They would need a cargo Starship and possibly a crewed one to keep it company. Or they could just go grab a few tons of samples of one and put it in the cargo bay.

  32. Given the gradual deployment and replacement, I think the strategy is that the 1st satellites act as beta versions for the next ones. Get a minimal product up as soon as possible, then iterate in subsequent batches. So I still see step-wise iteration with Starlink too.

  33. He means the step-wise iterations of the starlink satellites that have been tossed out. It is rumored that SpaceX let some starlink executives go over disagreements on how many test satellites they would need before they went into production of the fully operational fleet.

  35. Where did you see tossing out stepwise tests? (Note that I didn’t read this article in detail, so if it’s mentioned here, I may have missed it.)

    AFAIK, the hop and orbital tests are still on. And they’ve recently announced they’ll use a modified F9 2nd stage as a scale model to test reentry etc. That doesn’t sound like tossing out anything to me.

  38. Nice (if highly optimistic) analysis. Starship Super Heavy is also Starship Super Risky … humanity’s greatest space project for fraction of the cost SLS has many tech hurdles never even explored in aerospace. This “speed up” … tossing out stepwise tests … reduces my estimate of success from above 75% to maybe 25%. The idea of this costing just $2B (down from $5B) to get to operational placement of Starlink (also highly risky) seems un-doable and pointlessly fast. I have every hope for this … but this change is very risky. This could kill a company that might of created greatness on the old-but-still-crazy-fast-elon-musk-timeline.

  39. And they’ll need it, given the shelf life of these satellites is rather short, like 6-7 years and they’ll need yearly replacements in the thousands.

    The manufacturing of these satellites will be a permanent feature of the Starlink company, having to produce them by the thousands too. This can bring in economies of scale and production in series, resulting in the first actual mass produced satellites.

    And it can result in the network upgrades becoming much easier, almost trivial. The satellites will go down predictably and a batch of new ones will take their place, just make sure they have the latest upgrades.

    Starlink will definitely have versions and changing specs through time.

  40. I’ll be signing up as soon as it is available to lowlifes like me. I am SO looking forward to calling Verizon to cut my services and explain to them (recorded call for training purposes :)that it is a direct result of them hiring a lawyer/lobbyist like Ajit Pai and net neutrality going away. My whole family and all my friends will be joining me once I test the service, and I will push it to as many people as I can. The FCC and Ajit got something like 20 million emails protesting the death of net neutrality and they did not care. Let’s see whether they care to have 20 million paying customers follow suite.

  41. They’ll be sitting in LEO, so there’s a lot of protection from the magnetosphere during solar flare/CME events, and if the problem is big enough to hurt LEO satellites (ISS, etc), then all worldwide satellite operations out in GEO will also be impacted.
    Second, if Starlink can convince the big banks/markets that their service will be reliable and dependable, you’ll see the banks/markets move to LEO communications and use microwave and fiber for backup. Even cables get cut, so they probably have a variety of backup data paths to carry the load (if a bit slower).
    The only question I have is whether the switching from ground/satellite to satellite/ground (and possibly multiple satellites in between) will actually be faster than current cable/microwave networks.
    If Musk uses the banks/finance markets to pay for the first increment and then open Starlink to the rest of us, he’ll have a cash cow he can milk for trips to anywhere in the solar system.

Comments are closed.