Author Topic: Elon Musk eyeing partnership to launch 700 internet satellites  (Read 69649 times)

Offline Danderman

  • Extreme Veteran
  • Senior Member
  • *****
  • Posts: 9657
  • Liked: 385
  • Likes Given: 468
Elon Musk eyeing partnership to launch 700 internet satellites

http://online.wsj.com/articles/elon-musks-next-mission-internet-satellites-1415390062

(Always link the source article, not some rewrite from a blog site - Chris).
« Last Edit: 11/08/2014 03:10 pm by Chris Bergin »

Online philw1776

  • Full Member
  • ****
  • Posts: 1301
  • Seacoast NH
  • Liked: 1010
  • Likes Given: 413
I hope that on Elon's part this is just a partnership where he's seeking to expand the SpaceX launch market.  He's already got 2 1/2 companies demanding substantial amounts of his attention, Tesla, SpaceX and Solar City.
Lori, you have got to tell your friend Elon he can't do that.(FH) He's in our lane! You made us get out of low-Earth orbit, so we've given him that lane, but this is our lane!  We build the big rockets!

Online Owlon

  • Mechanical Engineer
  • Full Member
  • ***
  • Posts: 300
  • Denton, Texas
  • Liked: 145
  • Likes Given: 112
I hope that on Elon's part this is just a partnership where he's seeking to expand the SpaceX launch market.  He's already got 2 1/2 companies demanding substantial amounts of his attention, Tesla, SpaceX and Solar City.

I'm sure it is. He's said before that he doesn't really have time to do anything else, even though he has some ideas he'd like to explore if he could.

Offline NaN

  • ... a free man
  • Full Member
  • *
  • Posts: 146
  • Liked: 80
  • Likes Given: 121

Offline nadreck

Yes this is like Teledesic, and I hope that they consider putting the GPS notion in here too. Then after making a go of it in Earth orbit, I think a smaller (72 satellite) version could be sent to Mars to support exploration and settlement operations there.

The 20 - 30 F9 launches required for this would be a great anchor customer for used F9R first stages.  Also the manufacturing cost per 'bird' here will end up being a tiny fraction of the cost of a classic comms satellite. IF Elon has any part of this, it will likely be a demonstration of new, not quite as space hardened, but more state of the art electronics than current satellites have. This is a potential zero to hero move that leapfrogs competition. However that does not mean it is without risk, and I am willing to bet that in this 'partnership' Elon is basically only going to be out the cost of the launch vehicles, but in for a good share of the potential return.

I am going to hazard a guess that the first flown bird (and R&D leading up to it) will be about about $100,000,000 and that subsequent ones will be built for under $200,000
It is all well and good to quote those things that made it past your confirmation bias that other people wrote, but this is a discussion board damnit! Let us know what you think! And why!

Offline Nomadd

  • Senior Member
  • *****
  • Posts: 4037
  • Boca Chica, Texas
  • Liked: 13377
  • Likes Given: 437
 Building, testing and launching the satellites would be simple compared to the one item they dismissed as a detail. Acquiring and coordinating the spectrum usage across the planet would be a real bear. Teledesic actually made great progress in that and finally went under when various governments and agencies started taking back the spectrum rights.
Those who danced were thought to be quite insane by those who couldn't hear the music.

Offline sanman

  • Senior Member
  • *****
  • Posts: 4603
  • Liked: 717
  • Likes Given: 8
So Musk is continuing to find new ways to go vertical, eh?

How many launches could SpaceX be able to reap from this? I assume that each SpaceX launch would be able to deploy multiple satellites at a time.

What are the chances of existing satellite networks being disrupted by this new one? (ie. made superfluous or redundant)

Are there any anti-trust legal issues when you're involved in being space launch provider and also trying to become a consumer of space launch services?

Offline nadreck

Building, testing and launching the satellites would be simple compared to the one item they dismissed as a detail. Acquiring and coordinating the spectrum usage across the planet would be a real bear. Teledesic actually made great progress in that and finally went under when various governments and agencies started taking back the spectrum rights.

Actually  Greg Wyler has the spectrum already.
It is all well and good to quote those things that made it past your confirmation bias that other people wrote, but this is a discussion board damnit! Let us know what you think! And why!

Offline Ludus

  • Full Member
  • ****
  • Posts: 1254
  • Liked: 664
  • Likes Given: 326
As a side business the constellation could also provide a supersize version Planet Labs whole earth 24/7 real time imaging.
http://en.m.wikipedia.org/wiki/Planet_Labs

Offline AncientU

  • Senior Member
  • *****
  • Posts: 6252
  • Liked: 4123
  • Likes Given: 5668
So Musk is continuing to find new ways to go vertical, eh?

How many launches could SpaceX be able to reap from this? I assume that each SpaceX launch would be able to deploy multiple satellites at a time.

What are the chances of existing satellite networks being disrupted by this new one? (ie. made superfluous or redundant)

Are there any anti-trust legal issues when you're involved in being space launch provider and also trying to become a consumer of space launch services?

A single F9R launch might carry 20 sats plus deployment hardware, based on the arrangement used for Orbcom.  That would mean that a launch per month to give operator time to maneuver, test, activate each new set of 20, spread over three years, woul deploy all 700.  Cannot imagine an operator dealing with much larger numbers arriving on orbit faster than that.  Other thread suggested a handful of FH flights, but think that's unworkable and probably also makes redistribution into orbital planes impossible.
« Last Edit: 11/09/2014 12:08 pm by AncientU »
"If we shared everything [we are working on] people would think we are insane!"
-- SpaceX friend of mlindner

Offline oldAtlas_Eguy

  • Senior Member
  • *****
  • Posts: 3608
  • Florida
  • Liked: 2070
  • Likes Given: 247
So Musk is continuing to find new ways to go vertical, eh?

How many launches could SpaceX be able to reap from this? I assume that each SpaceX launch would be able to deploy multiple satellites at a time.

What are the chances of existing satellite networks being disrupted by this new one? (ie. made superfluous or redundant)

Are there any anti-trust legal issues when you're involved in being space launch provider and also trying to become a consumer of space launch services?

... Other thread suggested a handful of FH flights, but think that's unworkable and probably also makes redistribution into orbital planes impossible.

Since like GPS or Irridium use of a single inclination but phased planes, a single launch with a properly shapped eliptical transfer orbit can reach any orbital plane in the constelation.

Offline MP99



So Musk is continuing to find new ways to go vertical, eh?

How many launches could SpaceX be able to reap from this? I assume that each SpaceX launch would be able to deploy multiple satellites at a time.

What are the chances of existing satellite networks being disrupted by this new one? (ie. made superfluous or redundant)

Are there any anti-trust legal issues when you're involved in being space launch provider and also trying to become a consumer of space launch services?

A single F9R launch might carry 20 sats plus deployment hardware, based on the arrangement used for Orbcom.  That would mean that a launch per month to give operator time to maneuver, test, activate each new set of 20, spread over three years, woul deploy all 700.  Cannot imagine an operator dealing with much larger numbers arriving on orbit faster than that.  Other thread suggested a handful of FH flights, but think that's unworkable and probably also makes redistribution into orbital planes impossible.

What's the likely volume of 20 sats, given the antennae they need?

Would that fit in an F9 fairing?

Cheers, Martin

Offline watermod

  • Full Member
  • ****
  • Posts: 434
  • Liked: 131
  • Likes Given: 127
Please don't make the mistake Iridium did and put physically steerable antennas on it.   Use beam forming phased arrays if you want to aim at the other sats for radio comm purposes.

Offline oldAtlas_Eguy

  • Senior Member
  • *****
  • Posts: 3608
  • Florida
  • Liked: 2070
  • Likes Given: 247
A gravity gradient stabilized sat using a small weight and tether extending away from Earth with the tether being actually a high gain antenna that is used for sat to sat comm at >60GHz frequencies (uncontrolled frequencies because of atmospheric absorption) then a earth pointing array on the sat itself, no maneuvering fuel or very little is needed to remain oriented correctly. After the tether is deployed no additional mechanical movement of any kind would be needed.

An optical sat to sat or sat to ground link would require some sort of mechanical pointer system.  a mechanical pointer system would then require a set of torque gyros for stabilization and then a fuel RCS system or a magnetic field driver to maintain orientation and to despin the gyros. A tether could still be used to reduce the number of gyros from 3 to just 1 and reduce fuel or power consumption to 1/3 to despin the single gyro. Optical would give the sat a tremendous bandwidth possibility for a lower power consumption even considering the extra power to maintain stability. Multi-giga bit space qualified optical link tech is available.

Looking at the Business case.
- The sat would need to be able to provide >50Mbits/S up and down link capability per sat for the subscriber link which is time shared between subscribers (~300-1000 subscribers using 1 sat simultaneously) an average bandwidth per user of 100KBits with a monthly usage bits cap of 260Gbits. Question what is Wylers frequency alocation license bandwidth? This determins the max per sat bandwidth and the business case and hardware would be designed around the license.
- At $40/month per subscriber and 600,000 subscribers thats $288M per year of revenue with a sat lifetime of just 5 years the business breaks even.  The number of subscribers (or equvelent connection usage such as a multiple subscriber connect for an aircraft or ship) and the price per subscriber are tradeoffs.

Offline su27k

  • Full Member
  • ****
  • Posts: 1482
  • Liked: 1364
  • Likes Given: 129
What are the chances of existing satellite networks being disrupted by this new one? (ie. made superfluous or redundant)

Are there any anti-trust legal issues when you're involved in being space launch provider and also trying to become a consumer of space launch services?

Yeah, isn't this a direct competitor to Iridium? Anti-trust issues aside, it would make some awkward customer relations if SpaceX or Elon Musk is to be involved with satellite manufacturing/operation business.

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 28853
  • Minnesota
  • Liked: 9045
  • Likes Given: 5798
It makes a lot more sense for Musk to be just arranging the deal so it's an anchor customer for F9R. It's need 2-6 flights per year consistently with the same basic payload, so could allow streamlining. A negotiated block deal could allow SpaceX to reduce the price of launch quite a bit.

If SpaceX actually is to build the satellites, I'd be surprised... Although it's possible because SpaceX is building up expertise with Dragon. The satellites would probably be shorter lived than GSO sats, so cheaper solar cells like Dragon uses may make more sense.
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline MikeAtkinson

  • Full Member
  • ****
  • Posts: 1918
  • Bracknell, England
  • Liked: 694
  • Likes Given: 95
- The sat would need to be able to provide >50Mbits/S up and down link capability per sat for the subscriber link which is time shared between subscribers (~300-1000 subscribers using 1 sat simultaneously) an average bandwidth per user of 100KBits with a monthly usage bits cap of 260Gbits. Question what is Wylers frequency alocation license bandwidth? This determins the max per sat bandwidth and the business case and hardware would be designed around the license.

No, it determines the per beam bandwidth, a satellite might have several beams.

If I were designing this I would look at 20-100 beams per satellite, fixed phase array, with a time division within a beam and an allocation process so that different types of users (mainly downlink, mainly uplink and 2-way, fixed or bursty bandwidth, etc.) can be given bandwidth on what they pay for. Beams that connect to the Internet connected ground stations need several satellites of bandwidth.

I agree with gravity stabilization, but this needs a beam with a rigid mount to the satellite does it not.

I think that directed antennae are needed for the sat-sat comms. With 35 sats per plane they are ~1200 km apart which is several times their height, they need to carry several times a single satellites up/down bandwidth. I think optical comms is the way to go here as long as the pointing requirements can be met.

Using something like this each satellite would be able to provide 300-1000Mbit/s. Each satellite spends about 20% at near max capacity (the rest of the time over oceans, deserts, etc), so the 700 satellite constellation has up to 3,500,000 active subscribers. With 50% of income paying for the ground segment, advertising, customer relations, corporate overhead, etc. I reckon about $15/month for a 1Mbit/s downlink, 20kbit/s uplink average (10% duty cycle), 250Gbit/month.

I think optimisation of subscribers per kg of satellite would lead to larger satellites, perhaps 250-500kg. However it is hard to get subscribers until there is full coverage which means they must be launched quickly so lots of small satellites per launch. So a 2 generation approach might be in order, 1st generation small, light and not very capable but cheap to build and launch and a 2nd generation that is much more capable but more expensive.

As a final note, I think that this puts 2nd stage reuse on the agenda: "I don't expect the Falcon 9 to have a reusable upper stage, just because the - with a kerosene-based system, the specific impulse isn't really high enough to do that, and a lot of the missions we do for commercial satellite deployment are geostationary missions. So, we're really going very far out. These are high delta-velocity missions, so to try to get something back from that is really difficult." - Musk, 20 LEO missions is about half what is needed to make a reusable 2nd stage economic. Including allowance for dispenser, margin and high inclination orbit, one plane per launch with a reusable 2nd stage seems eminently doable.


Offline Ludus

  • Full Member
  • ****
  • Posts: 1254
  • Liked: 664
  • Likes Given: 326
- The sat would need to be able to provide >50Mbits/S up and down link capability per sat for the subscriber link which is time shared between subscribers (~300-1000 subscribers using 1 sat simultaneously) an average bandwidth per user of 100KBits with a monthly usage bits cap of 260Gbits. Question what is Wylers frequency alocation license bandwidth? This determins the max per sat bandwidth and the business case and hardware would be designed around the license.

No, it determines the per beam bandwidth, a satellite might have several beams.

If I were designing this I would look at 20-100 beams per satellite, fixed phase array, with a time division within a beam and an allocation process so that different types of users (mainly downlink, mainly uplink and 2-way, fixed or bursty bandwidth, etc.) can be given bandwidth on what they pay for. Beams that connect to the Internet connected ground stations need several satellites of bandwidth.

I agree with gravity stabilization, but this needs a beam with a rigid mount to the satellite does it not.

I think that directed antennae are needed for the sat-sat comms. With 35 sats per plane they are ~1200 km apart which is several times their height, they need to carry several times a single satellites up/down bandwidth. I think optical comms is the way to go here as long as the pointing requirements can be met.

Using something like this each satellite would be able to provide 300-1000Mbit/s. Each satellite spends about 20% at near max capacity (the rest of the time over oceans, deserts, etc), so the 700 satellite constellation has up to 3,500,000 active subscribers. With 50% of income paying for the ground segment, advertising, customer relations, corporate overhead, etc. I reckon about $15/month for a 1Mbit/s downlink, 20kbit/s uplink average (10% duty cycle), 250Gbit/month.

I think optimisation of subscribers per kg of satellite would lead to larger satellites, perhaps 250-500kg. However it is hard to get subscribers until there is full coverage which means they must be launched quickly so lots of small satellites per launch. So a 2 generation approach might be in order, 1st generation small, light and not very capable but cheap to build and launch and a 2nd generation that is much more capable but more expensive.

As a final note, I think that this puts 2nd stage reuse on the agenda: "I don't expect the Falcon 9 to have a reusable upper stage, just because the - with a kerosene-based system, the specific impulse isn't really high enough to do that, and a lot of the missions we do for commercial satellite deployment are geostationary missions. So, we're really going very far out. These are high delta-velocity missions, so to try to get something back from that is really difficult." - Musk, 20 LEO missions is about half what is needed to make a reusable 2nd stage economic. Including allowance for dispenser, margin and high inclination orbit, one plane per launch with a reusable 2nd stage seems eminently doable.

So you'd estimate $630M annual revenue based on base use 3.5M x $15 monthly fee. You also note that there would be a wide range of use classes and a lot of the system capacity would be underutilized. In remote ocean and arctic locations there would be both lots of available capacity and some users with deep pockets and no better alternative. I wonder how aircraft and ships that are in places with gaps in current coverage fit into the model? Doesn't a satellite like this provide the foundation for an earth hi res imaging system without adding much weight? (I'm assuming perhaps incorrectly that most of the mass of an imaging satellite is PV panels, antenna, and digital comm gear that's already present and the camera/lens mass is a tiny fraction). Could this be a significant revenue stream?

Offline MikeAtkinson

  • Full Member
  • ****
  • Posts: 1918
  • Bracknell, England
  • Liked: 694
  • Likes Given: 95
So you'd estimate $630M annual revenue based on base use 3.5M x $15 monthly fee. You also note that there would be a wide range of use classes and a lot of the system capacity would be underutilized. In remote ocean and arctic locations there would be both lots of available capacity and some users with deep pockets and no better alternative. I wonder how aircraft and ships that are in places with gaps in current coverage fit into the model? Doesn't a satellite like this provide the foundation for an earth hi res imaging system without adding much weight? (I'm assuming perhaps incorrectly that most of the mass of an imaging satellite is PV panels, antenna, and digital comm gear that's already present and the camera/lens mass is a tiny fraction). Could this be a significant revenue stream?

Ships are already covered by Inmarsat and various other comms systems. Although this would be cheaper, I don't see ship owners replacing what they already have installed outside of a major refit. So in the long term ships might be a major market, but in the short term the marine market is likely to be small.

The same with aircraft comms. Systems are being installed now in many airliners, by 2020 most will probably have some form of broadband. Replacing an incumbent is always difficult and for airliners could easily take another decade. So again in the long term aircraft could be a large market, but this is more something for a 2nd or 3rd generation satellite to address.

The market for satellite imagery of Earth is getting really crowded, lot of suppliers. A 700 satellite constellation could offer very frequent returns over the same place, which has some utility, but I doubt that it will be worth the extra mass on the satellites. The variety of sun angles will make many traditional uses of imagery difficult or impossible, but could open up new uses.

Offline FinalFrontier

  • Senior Member
  • *****
  • Posts: 4138
  • Space Watcher
  • Liked: 753
  • Likes Given: 163
Glad to see a thread about this. Meant to post one a few days ago but I see dandermen beat me to it. Seems like a pretty neat idea all around, its not really a direct competitor to existing sats because hes target internet specifically nothing else like telecomms ect so I think it wouldn't be too big of an issue. Not totally sure though. I hope this actually happens however I would like to see Spacex doing more than just LV's/dragon in the near future.
3-30-2017: The start of a great future
"Live Long and Prosper"

Tags: