Eutelsat OneWeb: Constellation - General Thread

[ncb1397]

We're looking at somewhere north of $2.2b just in launch costs. If I were their bank, I'd ask seriously why they're doing 21 Soyuz launches for at least $50 million apiece at ~6 tons per launch when they could do half as many Falcon 9 launches (since reusable F9 payload capacity and payload fairing capacity is about twice Soyuz) for about the same price per launch. And I'd question what the point of the LauncherOne launches are, since proper spare management could eliminate the need for so many smallsat launches.

I think it would still be 18 Falcon 9 launches because there are 18 orbital planes. Playing with GMAT, it appears to require about 1.3 km/s to change the angle of the orbital plane(10 degrees) from one to another at 1200 km altitude so populating multiple orbital planes may not be possible with one launch or would deplete the fuel capacity of the satellites to unacceptable levels. And I think the initial constellation is 40 satellites per orbital plane, so we are talking 1 Soyuz launch of 36 satellites and 2 Virgin Orbit launches of 2 satellite each or 18 launches of Soyuz and 36 launches of Launcher One.

36 * $10 million = $360 million
18 * $50 million = $900 million
Total = $1260 million.

62 million[1] * 18 = $1116 million

Difference = $114 million

[1] https://www.spacex.com/about/capabilities

Add in that Branson was willing to invest and SpaceX probably wasn't...and you get history.

[GreenShrike]

The initial deployment is on Soyuz, not LauncherOne.  The year that contract was signed (which was only 3 years ago) SpaceX had 6 successful launches, and the year after that SpaceX had 8 successful launches with a big backlog left to work through.

Ironic, then, that SpaceX's mis-steps sucked OneWeb so far out of (financial) position that they're now in some difficulty regarding overall costs, which can only be beneficial for Starlink's prospects.

On the other hand, if SpaceX had executed according to plan 2015 forward, F9's value proposition for OneWeb launches may have been much greater, leading OneWeb to spend significantly less on launch than what they're currently contractually obligated to, leaving Starlink with a financially stronger competitor at the present time.

[gongora]

Iridium shifts between planes and I doubt they're using that much propellant to do it. There are 21 Soyuz launches (the first only has 10 sats) with options for more Soyuz and Ariane 6 launches. They also cost more than $50M each.

[Mardlamock]

Iridium shifts between planes and I doubt they're using that much propellant to do it. There are 21 Soyuz launches (the first only has 10 sats) with options for more Soyuz and Ariane 6 launches. They also cost more than $50M each.

The slower option would be to park them into a lower orbit and have them boost up after the planes are aligned, as the nodal precession varies with apogee. What sort of propulsion do Oneweb sats have, are there any specs or is it still tightly lipped stuff?

[gongora]

They will normally be dropped off somewhere around 500km and be responsible for raising themselves to operational orbit and then back down low enough to deorbit in a reasonable amount of time.

[meekGee]

The take-away is that if you're designing both launcher and constellation, you have a huge advantage no matter what.

You have advance knowledge of the development path of both, you can optimize them together, and yeah, you pay cost, not price.

Vertical integration - it works.

Even if there wasn't a history between them, there would not have been much Wyler could have done.



-----
ABCD: Always Be Counting Down

[Robotbeat]

We're looking at somewhere north of $2.2b just in launch costs. If I were their bank, I'd ask seriously why they're doing 21 Soyuz launches for at least $50 million apiece at ~6 tons per launch when they could do half as many Falcon 9 launches (since reusable F9 payload capacity and payload fairing capacity is about twice Soyuz) for about the same price per launch. And I'd question what the point of the LauncherOne launches are, since proper spare management could eliminate the need for so many smallsat launches.

I think it would still be 18 Falcon 9 launches because there are 18 orbital planes. Playing with GMAT, it appears to require about 1.3 km/s to change the angle of the orbital plane(10 degrees) from one to another at 1200 km altitude so populating multiple orbital planes may not be possible with one launch or would deplete the fuel capacity of the satellites to unacceptable levels. ...

...you don't need to do it propulsively. As gongora points out, usually they are dropped off at a lower altitude and insert themselves. But if they're dropped off at a lower altitude, you can just let nodal precession due to Earth's equatorial bulge do the work for you. Just have to time it right.


...and the satellites have enough propellant to put themselves in the right orbit, you may be able to do this in just 9 Falcon 9 launches.

[Robotbeat]

The initial deployment is on Soyuz, not LauncherOne.  The year that contract was signed (which was only 3 years ago) SpaceX had 6 successful launches, and the year after that SpaceX had 8 successful launches with a big backlog left to work through.

Ironic, then, that SpaceX's mis-steps sucked OneWeb so far out of (financial) position that they're now in some difficulty regarding overall costs, which can only be beneficial for Starlink's prospects.

On the other hand, if SpaceX had executed according to plan 2015 forward, F9's value proposition for OneWeb launches may have been much greater, leading OneWeb to spend significantly less on launch than what they're currently contractually obligated to, leaving Starlink with a financially stronger competitor at the present time.

SpaceX is doing well enough now that I don't think it has to worry about OneWeb taking Starlink's business. Starlink is a much more powerful capability and may have different kind of customers as well. And half a billion in F9 revenue may help Starlink quite a bit.

[su27k]

The initial deployment is on Soyuz, not LauncherOne.  The year that contract was signed (which was only 3 years ago) SpaceX had 6 successful launches, and the year after that SpaceX had 8 successful launches with a big backlog left to work through.

I doubt this is a major consideration, Iridium signed up with SpaceX when the latter only had one successful Falcon 9 launch in total.

[jongoff]

The initial deployment is on Soyuz, not LauncherOne.  The year that contract was signed (which was only 3 years ago) SpaceX had 6 successful launches, and the year after that SpaceX had 8 successful launches with a big backlog left to work through.

I doubt this is a major consideration, Iridium signed up with SpaceX when the latter only had one successful Falcon 9 launch in total.

I think the main reason why OneWeb didn't go with SpaceX is pretty obvious isn't it? I mean if you had a launch provider you had been working with on a constellation project that stabbed you in the back, and announced a competing constellation, would you be in a rush to give them a big juicy launch contract to help fund their constellation?

~Jon

[Oli]

Consider that the recent estimates (from OneWeb themselves!) put OneWeb at $6 billion, that's cost growth of a factor of 3 to 4.

And considering that most of that article's estimated costs are actually launch or insurance and that the bank is now questioning their plan based on high costs, it definitely seems that reusable launch is enabling in this case.

I seriously doubt the cost increases are due to increased launch costs or increased launch insurance. So ultimately it's still ~$1.4bn for initial deployment vs. up to potentially $6bn total cost. I'm guessing LauncherOne is not used for regular deployment, that wouldn't make much sense (unless they got a huge discount).

...you don't need to do it propulsively. As gongora points out, usually they are dropped off at a lower altitude and insert themselves. But if they're dropped off at a lower altitude, you can just let nodal precession due to Earth's equatorial bulge do the work for you. Just have to time it right.


...and the satellites have enough propellant to put themselves in the right orbit, you may be able to do this in just 9 Falcon 9 launches.

And how long will that take? The problem with any slow deployment method is that service is impossible or limited during that time. Even more relevant when satellite lifetime isn't 15 years.

The take-away is that if you're designing both launcher and constellation, you have a huge advantage no matter what.

Actually the takeaway is quite the opposite. Launch doesn't seem to be the problem.

I think the main reason why OneWeb didn't go with SpaceX is pretty obvious isn't it? I mean if you had a launch provider you had been working with on a constellation project that stabbed you in the back, and announced a competing constellation, would you be in a rush to give them a big juicy launch contract to help fund their constellation?

That suggests OneWeb's launch contract would have had a significant influence on SpaceX' decision to do Starlink or not, which I find a bit silly. The contract was signed in 2015, back then SpaceX had a huge backlog and reliability issues.

[Lar]

Contracts can be renegotiated or exited via penalty clause.

Seems like the choice at this point is between not getting funded further (and the launch providers get only the money they got so far for advance payments, plus whatever they can fight for in bankruptcy court) or getting funded but having a large amount of money go to a firm that is developing and deploying a more capable and cheaper competitor (and the current launch providers get what they got so far plus whatever penalty payments apply.. this number may be big enough that it sways the decision). Would you rather come in second, or not finish at all?

I'd pick second. Except I have a big ego and it would chap me to enable my competitor to win (sooner than otherwise).

[Robotbeat]

The initial deployment is on Soyuz, not LauncherOne.  The year that contract was signed (which was only 3 years ago) SpaceX had 6 successful launches, and the year after that SpaceX had 8 successful launches with a big backlog left to work through.

I doubt this is a major consideration, Iridium signed up with SpaceX when the latter only had one successful Falcon 9 launch in total.

I think the main reason why OneWeb didn't go with SpaceX is pretty obvious isn't it? I mean if you had a launch provider you had been working with on a constellation project that stabbed you in the back, and announced a competing constellation, would you be in a rush to give them a big juicy launch contract to help fund their constellation?

~Jon

”stab you in the back” is not accurate, Jon. There was disagreement over the scope of the constellation, and OneWeb left SpaceX, not the other way around. But sure, I can imagine OneWeb making such a decision for such reasons.

But we’re discussing the bank’s decision. The bank does not care about that ego clashing. They care about “So you expect us to give you like an extra billion and sink it in the ocean because you can’t hold your ego in check?? Sorry, but you’re going to have to find someone else to fund your dreams!”

[Robotbeat]

Oh, and I don’t necessarily think soyuz was such a bad choice in 2015, but it certainly looks expensive now!

[ncb1397]

The initial deployment is on Soyuz, not LauncherOne.  The year that contract was signed (which was only 3 years ago) SpaceX had 6 successful launches, and the year after that SpaceX had 8 successful launches with a big backlog left to work through.

I doubt this is a major consideration, Iridium signed up with SpaceX when the latter only had one successful Falcon 9 launch in total.

I think the main reason why OneWeb didn't go with SpaceX is pretty obvious isn't it? I mean if you had a launch provider you had been working with on a constellation project that stabbed you in the back, and announced a competing constellation, would you be in a rush to give them a big juicy launch contract to help fund their constellation?

~Jon

”stab you in the back” is not accurate, Jon. There was disagreement over the scope of the constellation, and OneWeb left SpaceX, not the other way around. But sure, I can imagine OneWeb making such a decision for such reasons.

But we’re discussing the bank’s decision. The bank does not care about that ego clashing. They care about “So you expect us to give you like an extra billion and sink it in the ocean because you can’t hold your ego in check?? Sorry, but you’re going to have to find someone else to fund your dreams!”

That's not their gripe. They don't think there is enough French content. Removing Arianespace doesn't help there.

[meekGee]

The take-away is that if you're designing both launcher and constellation, you have a huge advantage no matter what.

Actually the takeaway is quite the opposite. Launch doesn't seem to be the problem.

"Seem" to whom?  Most people here are pointing out just how large the difference is, especially when considering that the competitor (Starlink) will only be paying cost.

OneWeb has other problems, for sure, but cost of launch is certainly a major one.

-----
ABCD: Always Be Counting Down

[jongoff]

That suggests OneWeb's launch contract would have had a significant influence on SpaceX' decision to do Starlink or not, which I find a bit silly. The contract was signed in 2015, back then SpaceX had a huge backlog and reliability issues.

I'm not sure how you came to that conclusion. SpaceX's decision to part ways with OneWeb and do a rival constellation was probably based on seeing how potentially lucrative it was, and probably had little to do with the launch purchase. Oneweb's precursor, Google, and SpaceX were working a lot closer prior to the breakup than just Oneweb's precursor shopping around for launch providers. At least that's my understanding.

~Jon

[Robotbeat]

That suggests OneWeb's launch contract would have had a significant influence on SpaceX' decision to do Starlink or not, which I find a bit silly. The contract was signed in 2015, back then SpaceX had a huge backlog and reliability issues.

I'm not sure how you came to that conclusion. SpaceX's decision to part ways with OneWeb and do a rival constellation was probably based on seeing how potentially lucrative it was, and probably had little to do with the launch purchase. Oneweb's precursor, Google, and SpaceX were working a lot closer prior to the breakup than just Oneweb's precursor shopping around for launch providers. At least that's my understanding.

~Jon

No, it had to do with SpaceX wanting something much more ambitious than just a "modest" bent-pipe constellation of smallsats. Greg Wyler refused, so took his ball and went home. Don't just automatically assume it's SpaceX being shady; it's not like Greg Wyler doesn't have ego to spare; I've watched him try to stifle some competition in this space by waving around frequency rights and attempting to develop new artificial monopolies on LEO altitudes as yet another barrier to competition (in the name of orbital debris mitigation). (And you're right that they were working more closely together than just SpaceX being the launch provider.)

A megaconstellation is an obvious application of a reusable rocket. I had one sketched out to use F9R before I heard about OneWeb/SpaceX's constellation plans, and I considered it one of the few business justifications for full reuse. But that ONLY works well to justify an RLV if your constellation's total mass is large enough. The 650-900 smallsat (125kg) OneWeb constellation was likely too small for that, and no one should fault SpaceX for wanting to ensure development of a constellation large enough (12,000 times 500kg at full size, replaced up to once every four years) to justify F9R or even BFR.

I mean, the OneWeb constellation could technically be launched by a single BFR launch and may last 5-10 years. The SpaceX full constellation would require 10 BFR launches every year (100 Falcon 9s) even if the satellites themselves don't grow (as they likely would).

A OneWeb constellation would only slightly benefit F9R's viability and would have virtually no impact to BFR. But SpaceX's constellation is big enough to take advantage of BFR and justify its development to investors. That's an enormous difference, WHETHER OR NOT OneWeb/Starlink is wildly profitable or merely break-even.

[ZachF]

The take-away is that if you're designing both launcher and constellation, you have a huge advantage no matter what.

Actually the takeaway is quite the opposite. Launch doesn't seem to be the problem.

"Seem" to whom?  Most people here are pointing out just how large the difference is, especially when considering that the competitor (Starlink) will only be paying cost.

OneWeb has other problems, for sure, but cost of launch is certainly a major one.

-----
ABCD: Always Be Counting Down

Reading between the lines, It seems like Musk wanted to wait a bit in order to leverage lower launch costs into producing a more capable constellation, and Wyler just wanted to get there first, damn the torpedoes... It looks like history is leaning towards Musk.

Just to color the difference in cost the two constellations will be paying, It will cost OneWeb ~$2.2 billion to put up a ~150 tonne constellation... It will probably cost SpaceX $4.5 billion to put up their ~2,000 tonne initial constellation.

[TrevorMonty]

Is it worth having insurance for these launches? I'd thought it was cheaper to self insure for this number of launches. Satellites are easily replaced and
Soyzu has excellent record especially when launched by Arianespace. The occasional LauncherOne failure wouldn't matter.