Blue Origin Thread (merged)

[tater]

http://forum.nasaspaceflight.com/index.php?topic=41634.0

Quote

SpaceX filed on 11/15/2016 to the FCC the application for their massive LEO constellation of 4425 satellites in order to provide low latency satellitar Internet.

 

That's a lot of launches.

They are not cubesats, BTW, each is almost 400kg.

Edited November 17, 2016 by tater

[TheEpicSquared]

  Right, time for some math:

  The Falcon 9* has a LEO capacity of 22.8 tons according to spacex.com. Let's call it 20 tons, since I don't know how much payload capacity is reduced because of reusability. If each satellite weighs 0.4 tons, then theoretically each launch should be able to loft  20 / 0.4 = 50 of these satellites**.

  So, 4425 / 50 = 88.5 launches -- let's call it 100 launches*** to complete the constellation. I'm not sure how many times an F9 1st stage can be reused, but I think it was somewhere around 10****. That means that theoretically 10 F9 rockets could be enough for the whole constellation. SpaceX has successfuly launched 27 F9s, so it should be entirely possible.

  Now, prices. One brand new F9 costs roughly $62 million to build and launch. Since the constellation would theoretically require 10 F9s, the total cost of 10 F9s are 62 million time 10, or $620 million. Reusing a first stage would lead to a launch cost of $42.8 million (source). Let's call it $43 million. So 90 launches (the other 10 launches would be with the 10 brand new F9 1st stages) equals a cost of 43 million times 90, or $3870 million, aka $3.87 billion. 3.87 billion + 0.62 billion equals a grand total of 4.49 billion dollars to launch the entire constellation. For simplicity, let's call it 4.5 billion dollars. For comparison, NASA's 2016 budget was $18.5 dollars (source).

  Finally, the amount of time this would take. In 2016, SpaceX has launched F9 successfully 8 times with plans for 4 more (Echostar 23, "a couple" Iridium launches, I'm assuming 2, and Formosat-5/SHERPA. Here's the source) for a total of 12 launches in 2016. This equals 1 launch a month. Therefore 100 launches will take 100 months or 8.3 years. However, in early 2016, SpaceX said they want to increase launch rate to 18 launches a year (source). This means the constellation will only take 100 / 18 = 5.5 years. Let's take a number between those two figures - 7 years. So it should take SpaceX roughly 7 years to complete their satellite constellation.

 

So after the maths, we can conclude that the entire constellation would take around 100 launches, cost about 4.5 billion dollars and take roughly 7 years to complete.

 

*I'm only using Falcon 9 for this, not Falcon Heavy.

**In fact, SpaceX will probably launch less than 50 satellites per mission.

*** I rounded up the number of launches because SpaceX almost certainly won't launch 50 satellites at once.

**** I vaguely remember Musk saying that it was 10, however I'm not sure. Does anyone know how many times an F9 1st stage can be reused?

 

Spoiler

I'm done with math for today. Time for a nice cup of tea! 

 

[Nibb31]

20 minutes ago, TheEpicSquared said:

  Right, time for some math:

  The Falcon 9* has a LEO capacity of 22.8 tons according to spacex.com. Let's call it 20 tons, since I don't know how much payload capacity is reduced because of reusability. If each satellite weighs 0.4 tons, then theoretically each launch should be able to loft  20 / 0.4 = 50 of these satellites**.

  So, 4425 / 50 = 88.5 launches -- let's call it 100 launches*** to complete the constellation. I'm not sure how many times an F9 1st stage can be reused, but I think it was somewhere around 10****. That means that theoretically 10 F9 rockets could be enough for the whole constellation. SpaceX has successfuly launched 27 F9s, so it should be entirely possible.

  Now, prices. One brand new F9 costs roughly $62 million to build and launch. Since the constellation would theoretically require 10 F9s, the total cost of 10 F9s are 62 million time 10, or $620 million. Reusing a first stage would lead to a launch cost of $42.8 million (source). Let's call it $43 million. So 90 launches (the other 10 launches would be with the 10 brand new F9 1st stages) equals a cost of 43 million times 90, or $3870 million, aka $3.87 billion. 3.87 billion + 0.62 billion equals a grand total of 4.49 billion dollars to launch the entire constellation. For simplicity, let's call it 4.5 billion dollars. For comparison, NASA's 2016 budget was $18.5 dollars (source).

  Finally, the amount of time this would take. In 2016, SpaceX has launched F9 successfully 8 times with plans for 4 more (Echostar 23, "a couple" Iridium launches, I'm assuming 2, and Formosat-5/SHERPA. Here's the source) for a total of 12 launches in 2016. This equals 1 launch a month. Therefore 100 launches will take 100 months or 8.3 years. However, in early 2016, SpaceX said they want to increase launch rate to 18 launches a year (source). This means the constellation will only take 100 / 18 = 5.5 years. Let's take a number between those two figures - 7 years. So it should take SpaceX roughly 7 years to complete their satellite constellation.

 

So after the maths, we can conclude that the entire constellation would take around 100 launches, cost about 4.5 billion dollars and take roughly 7 years to complete.

Assuming that:

• All 50 satellites go the same inclination/orbital period
• SpaceX dumps paying commercial and government launch market to focus 100% on its own launches.

Neither of which are realistic.

 

[StarStreak2109]

Well seeing what some investment projects in other industries cost over a much shorter period of time, it really is not that much money.

Once again, I'd like to see Elon's business models here and there...

[TheEpicSquared]

15 minutes ago, Nibb31 said:

Assuming that:

• All 50 satellites go the same inclination/orbital period
• SpaceX dumps paying commercial and government launch market to focus 100% on its own launches.

Neither of which are realistic.

 

I know those aren't realistic, I was just making some (unrealistic, as you said) guesstimates. My guess is that it'll cost more and take more launches and time to finish the constellation, I was going for the most ideal circumstances.

[Northstar1989]

Whoa, whoa with the time estimates- isn't SpaceX working on an entirely new launch-facility or two at alternate sites?  They might be able to attain a much higher launch-rate if they double their launch-sites from 2 to 4...

 

Regards,

Northstar 

[kerbiloid]

11 hours ago, StarStreak2109 said:

I'd like to see Elon's business models here and there...

4400 sats, a triple of the existing total. Looks like the first orbital squatting. Then they can sell orbits.

6 hours ago, Northstar1989 said:

isn't SpaceX working on an entirely new launch-facility or two at alternate sites?

For such amount looks more like a whole space elevator.

[Nibb31]

10 hours ago, Northstar1989 said:

Whoa, whoa with the time estimates- isn't SpaceX working on an entirely new launch-facility or two at alternate sites?  They might be able to attain a much higher launch-rate if they double their launch-sites from 2 to 4...

 

Regards,

Northstar 

They are building Boca Chica, Texas, which will give them a total of 4 launch sites (KSC, CCAFS, and Vandenberg), but Boca Chica is very constrained to low inclination (equatorial or GTO) launches. Not all launch sites can launch into any orbit, for example, KSC can only launch from something like 28° to 57° inclinations. The problem is that if you want global coverage from LEO, you need a polar orbit, which is only possible from Vandenberg. Low inclination launches are useless for this application.

Edited November 18, 2016 by Nibb31

[wumpus]

8 hours ago, Nibb31 said:

They are building Boca Chica, Texas, which will give them a total of 4 launch sites (KSC, CCAFS, and Vandenberg), but Boca Chica is very constrained to low inclination (equatorial or GTO) launches. Not all launch sites can launch into any orbit, for example, KSC can only launch from something like 28° to 57° inclinations. The problem is that if you want global coverage from LEO, you need a polar orbit, which is only possible from Vandenberg. Low inclination launches are useless for this application.

Irridium has an 86 degree inclination and appears to be pretty low for "global coverage" constellations, so expect Vandenburg launches.

While any such constellations should require as many satellites for each plane as a facon heavy (or possibly even the big boy should they ever build it) can carry, does it really cost much delta-v to change planes in polar orbit?  I keep thinking that if you increase/decrease your orbit the satellite will rotate with respect to the rest of the constellation.

To a certain degree I suspect that this is designed to give the Falcon Heavy a mission.  I'm curious to see if they decide to recover the center stage or not.

[Aethon]

 

A test version of the Interim Cryogenic Propulsion Stage (ICPS) being installed on the test stand at the Marshall Spaceflight Center ahead of a big test series.

" The ICPS is the liquid oxygen/liquid hydrogen-based propulsion stage that will give NASA's Orion spacecraft the in-space push needed to fly beyond the moon before it returns to Earth on the first flight of SLS and Orion in 2018. "

https://www.nasa.gov/exploration/systems/sls/SLS-ICPS-in-marshall-test-stand

 

[Tullius]

3 hours ago, wumpus said:

Irridium has an 86 degree inclination and appears to be pretty low for "global coverage" constellations, so expect Vandenburg launches.

While any such constellations should require as many satellites for each plane as a facon heavy (or possibly even the big boy should they ever build it) can carry, does it really cost much delta-v to change planes in polar orbit?  I keep thinking that if you increase/decrease your orbit the satellite will rotate with respect to the rest of the constellation.

To a certain degree I suspect that this is designed to give the Falcon Heavy a mission.  I'm curious to see if they decide to recover the center stage or not.

It's just as expensive to change the plane in polar orbit than in equatorial orbit (think of KSP).

I rather think the idea behind this huge constellation of relatively small satellites is not to launch them separately (4400 satellites is a lot more than is needed for global coverage, for comparison Iridium uses 66 satellites. So why use multiple satellites, if you just could use a larger one?), but for hitching rides on Falcon 9 launches, where the main payload does not reach the maximum weight. This would allow smaller satellites to be launched more cheaply on Falcon 9 (or even Heavy), since the unused capacity would be used by SpaceX for one or multiple of their internet satellites, covering part of the launch costs. The huge number of satellites combined maybe with some extra launches with only internet satellites would probably easily allow for global coverage.

[Elthy]

I think the intention with that many satellites is similar to the Merlin engines, massproduction to lower costs...

[Clipperride]

An article on BBC News reporting on  SpaceX plans to launch global broadband internet coverage via 4425 satellites in 83 different orbital planes.  The level of investment required to set up a constellation of that size could, I think, be the stumbling block.  It's going to take a long time to recoup the estimated $10 billion (US) it will take to build!

[Nibb31]

4 hours ago, Elthy said:

I think the intention with that many satellites is similar to the Merlin engines, massproduction to lower costs...

Of course, it lowers the unit cost of each individual bird, but that's rather pointless if the entire project costs an order of magnitude more than any other telecom project.

The whole point of this constellation was originally to raise money for the Mars project, but even if it does become profitable, I don't see how it will bring in more billions than it will cost. In other words, if they are spending $10 billion with a super optimistic return on investment of $5 billion to fund Mars, then they might as well have just injected $10 billion into Mars.

And this is all considering that the global telecom giants (AT&T, Verizon, Vodafone, etc...) are going to sit there with their massive investments in fiber and radio networks and watch Musk pick at their market share. Good luck with that.

Edited November 19, 2016 by Nibb31

[kerbiloid]

8 hours ago, Clipperride said:

4425 satellites in 83 different orbital planes

Looks like they are going to build a Dyson sphere around the Earth.

[magnemoe]

7 hours ago, Nibb31 said:

Of course, it lowers the unit cost of each individual bird, but that's rather pointless if the entire project costs an order of magnitude more than any other telecom project.

The whole point of this constellation was originally to raise money for the Mars project, but even if it does become profitable, I don't see how it will bring in more billions than it will cost. In other words, if they are spending $10 billion with a super optimistic return on investment of $5 billion to fund Mars, then they might as well have just injected $10 billion into Mars.

And this is all considering that the global telecom giants (AT&T, Verizon, Vodafone, etc...) are going to sit there with their massive investments in fiber and radio networks and watch Musk pick at their market share. Good luck with that.

An satellite network like this will not compete with cable, it would rather complement mobile networks. 
Not sure about the economic of this however its an increasing demand to have high bandwidth internet everywhere and expensive and hard to deliver this everywhere. 
An question is if this can be effective before finished, you will file for the full network but build over time if possible. 

[Nibb31]

4 hours ago, magnemoe said:

An satellite network like this will not compete with cable, it would rather complement mobile networks. 

Of course. But you know that mobile networks are built using a heavy infrastructure of microwave and fiberoptic links. 

4 hours ago, magnemoe said:

Not sure about the economic of this however its an increasing demand to have high bandwidth internet everywhere and expensive and hard to deliver this everywhere. 

 Next year, mobile operators are going to start deploying 5G.

They won't let their huge investment in infrastructure go down the drain. If necessary, they will probably undercut anything SpaceX does by cutting prices. SpaceX will be left with the "rural" and "developing country" niches that are hard to monetize.

 

[YNM]

A few thousand 400 kg satellite constellation ? Equally broken-up ISS pieces, anyone ?

[magnemoe]

45 minutes ago, Nibb31 said:

Of course. But you know that mobile networks are built using a heavy infrastructure of microwave and fiberoptic links. 

 Next year, mobile operators are going to start deploying 5G.

They won't let their huge investment in infrastructure go down the drain. If necessary, they will probably undercut anything SpaceX does by cutting prices. SpaceX will be left with the "rural" and "developing country" niches that are hard to monetize.

 

Obvious its an cabled backbone, however if you have cable or wifi to cable this is better than mobile nets, faster and cheaper. 

An low attitude satelite like this have an fixed bandwidth for the area it fly over, satellite don't car if its central Europe or the Pacific it fly over, maximum bandwidth is the same anywhere.
5G has an totally different purpose, it is an very short range very high speed system for used in crowded areas. 
Wonder how the phone companies could stop this even if they wanted and they probably don't, if the satelite reciver can be put in any phone they would love it as they don't have to cover low populated places. Enough deals here would make this profitable, the phone companies would loose the lucrative roaming income but that is under hard pressure anyway. 

You also have a lot of companies and organisations who would pay well for an worldwide high bandwidth service. Military is an obvious one, but also airliners and boats. That this is not linked to the ground net, has huge benefits regarding emergency operations. 
 

[kerbiloid]

They should release sticky threads behind the satellites.
When they get sticked to each other, this will be literally a network, a world wide web 3.0.
Also then no need in antennas, only opticables.

[Frozen_Heart]

19 minutes ago, kerbiloid said:

They should release sticky threads behind the satellites.
When they get sticked to each other, this will be literally a network, a world wide web 3.0.
Also then no need in antennas, only opticables.

I... I don't think orbital mechanics works quite like that...

[StrandedonEarth]

29 minutes ago, kerbiloid said:

opticables

Did you just coin a new word there? I like it!

[Nibb31]

1 hour ago, magnemoe said:

Obvious its an cabled backbone, however if you have cable or wifi to cable this is better than mobile nets, faster and cheaper. 

Why cheaper ? If the network costs $10 billion, it won't be cheap.

In France, you can get 20GB 4G plans with unlimited text and phone for under 10€/month. That will be hard to beat.

1 hour ago, magnemoe said:

An low attitude satelite like this have an fixed bandwidth for the area it fly over, satellite don't car if its central Europe or the Pacific it fly over, maximum bandwidth is the same anywhere.
5G has an totally different purpose, it is an very short range very high speed system for used in crowded areas. 
Wonder how the phone companies could stop this even if they wanted and they probably don't, if the satelite reciver can be put in any phone they would love it as they don't have to cover low populated places. Enough deals here would make this profitable, the phone companies would loose the lucrative roaming income but that is under hard pressure anyway.

I'm not saying that sat communication doesn't have benefits. It all depends on how much people will be willing to pay for those benefits. If it ends up being cheaper, then existing operators will align pricing. They won't go down without a fight.

[magnemoe]

1 hour ago, Nibb31 said:

Why cheaper ? If the network costs $10 billion, it won't be cheap.

In France, you can get 20GB 4G plans with unlimited text and phone for under 10€/month. That will be hard to beat.

I'm not saying that sat communication doesn't have benefits. It all depends on how much people will be willing to pay for those benefits. If it ends up being cheaper, then existing operators will align pricing. They won't go down without a fight.

Far more expensive in Norway, lower and more spread out population and higher cost level, also more greedy phone companies. 
Still an cable is unlimited bandwidth up to an high faction of the bandwidth you pay for. 
An mobile network has an fixed bandwidth for its cell, this bandwidth is shared with all users, notice this while commuting with train even loading this forum is slow. 
This is why 5g is interesting, however you will not get 5g outside if densly populated areas as range is short. 

[Northstar1989]

3 hours ago, Nibb31 said:

Why cheaper ? If the network costs $10 billion, it won't be cheap.

In France, you can get 20GB 4G plans with unlimited text and phone for under 10€/month. That will be hard to beat.

Do you KNOW how much the mobile networks in France costed to build Nibb?  (Clearly not, or you wouldn't make that argument)

There were an estimated 50.1 million mobile phone users in France on 2016.  Each of them pays over 250 Euros a year for mobile phone coverage (10/month is not realistic for the majority of mobile users, you know that- those are only special deals- most users pay over 20 Euros/month for unlimited data in France- in fact Free Mobile is releasing an unlimited 3G LTE plan for 20/month in France and that's considered astonishing, and probably not scaleable for the entire country...)  That means that mobile companies would need to charge, at a minimum, 12.525 BILLION Euros a year for unlimited mobile data.

Now, let's make a conservative assumption that only 12% of revenue is going towards paying for the actual network, and that they amortize those costs over 10 years.  That still means that the network of France alone would cost over 15.03 Billion Euros.  That's already more than the $10 billion in American Dollars for a global satellite network, and that's just for *ONE* relatively technologically-advanced European country with a relatively favorable climate and population-density for construction of ground-based networks (countries like Norway or Canada are much more difficult).

Even if France's network only costed one TWENTIETH what I estimated above, a global high-bandwith sat network costing $10 billion USD could *EASILY* pay for itself due to its much larger coverage area and ability to efficiently cover remote or inhospitable areas...

 

Regards,

Northstar

Edited November 19, 2016 by Northstar1989