Amazon Project Kuiper Broadband Constellation

[Eric Hedman]

The main purpose of this lawsuit is to make money for the lawyers.  Everything else is secondary.

[imprezive]

I find it interesting how many people think ramping up F9 to accommodate Kuiper would be a Herculean tasks but 3 independent launchers ramping from zero launches to what would be their companies highest cadence ever is somehow more plausible.

[ThatOldJanxSpirit]

The main purpose of this lawsuit is to make money for the lawyers.  Everything else is secondary.

A tad too cynical.

Bezos companies track record in executing space projects is, let’s face it, less than impressive in terms of expenditure of time and money v tangible results. Kuiper is facing a formidable competitor that is already established, with a structural advantage in launch cadence and costs that is only likely to grow wider.

Amazon shareholders are more than entitled to question if pursuing Kuiper as intended (or at all) is actually in their interests. There also do seem to be grounds to question if the Amazon board fulfilled their duties in scrutinising this project.

[AmigaClone]

[from alexphysics13:]

The only way SpaceX would be able to compensate for the surge of customer flights would be to spend more money on Falcon 9 R&D and on ramping up cadence perhaps with more droneships etc.

R&D takes time, so it cannot in general respond to a "surge". As of right now, Starship R&D is much more likely to speed up Starlink than F9 R&D and has higher ROI.

Falcon Heavy launches disrupt the schedule at LC-39A, because the pad must be reconfigured. SpaceX can perhaps increase launch cadence there if they can move Falcon Heavy launches to LC-37B, but this cannot happen unless they can negotiate for the use of that pad and then convert it. Conversion cannot really start until after the last Delta IV Heavy launches in March. SpaceX needs to provide a vertical payload integration service for NSSL in any event as part of their NSSL Phase 2 service, and I think LC-37B is a good candidate, similar to SLC-6 at VSFB.

My understanding is that LC39A also undergoes at least some minor reconfiguration between launches with a single stick Falcon 9 topped with a fairing and one topped with a Dragon 2.

It might be possible to make some changes to LC39A that would allow that pad to be reconfigured faster between a Falcon Heavy and a Falcon 9. One possibility would be adding a second Transporter-Erector permanently configured for the Falcon Heavy. If SpaceX were to acquire SLC37, that Falcon Heavy transporter-erector could be transported to that launch site.

[ThatOldJanxSpirit]

[from alexphysics13:]

The only way SpaceX would be able to compensate for the surge of customer flights would be to spend more money on Falcon 9 R&D and on ramping up cadence perhaps with more droneships etc.

R&D takes time, so it cannot in general respond to a "surge". As of right now, Starship R&D is much more likely to speed up Starlink than F9 R&D and has higher ROI.

Falcon Heavy launches disrupt the schedule at LC-39A, because the pad must be reconfigured. SpaceX can perhaps increase launch cadence there if they can move Falcon Heavy launches to LC-37B, but this cannot happen unless they can negotiate for the use of that pad and then convert it. Conversion cannot really start until after the last Delta IV Heavy launches in March. SpaceX needs to provide a vertical payload integration service for NSSL in any event as part of their NSSL Phase 2 service, and I think LC-37B is a good candidate, similar to SLC-6 at VSFB.

My understanding is that LC39A also undergoes at least some minor reconfiguration between launches with a single stick Falcon 9 topped with a fairing and one topped with a Dragon 2.

It might be possible to make some changes to LC39A that would allow that pad to be reconfigured faster between a Falcon Heavy and a Falcon 9. One possibility would be adding a second Transporter-Erector permanently configured for the Falcon Heavy. If SpaceX were to acquire SLC37, that Falcon Heavy transporter-erector could be transported to that launch site.

I’d suggest that detailed discussion of how SpaceX ramps cadence should be in a SpaceX thread. Suffice it to say that SpaceX appears confident of ramping to 140 per year.

https://twitter.com/elonmusk/status/1698045247547212093

That, plus switching to Starlink V2 on Starship should be sufficient for launching Kuiper.

[freddo411]

[from alexphysics13:]

The only way SpaceX would be able to compensate for the surge of customer flights would be to spend more money on Falcon 9 R&D and on ramping up cadence perhaps with more droneships etc.

R&D takes time, so it cannot in general respond to a "surge". As of right now, Starship R&D is much more likely to speed up Starlink than F9 R&D and has higher ROI.

Falcon Heavy launches disrupt the schedule at LC-39A, because the pad must be reconfigured. SpaceX can perhaps increase launch cadence there if they can move Falcon Heavy launches to LC-37B, but this cannot happen unless they can negotiate for the use of that pad and then convert it. Conversion cannot really start until after the last Delta IV Heavy launches in March. SpaceX needs to provide a vertical payload integration service for NSSL in any event as part of their NSSL Phase 2 service, and I think LC-37B is a good candidate, similar to SLC-6 at VSFB.

Excellent foreshadowing the future of LC-37B.   

If ULA pays for 37B by the year, it must certainly get it off it's books if I doesn't plan to use it.   Seems like Vulcan already has the needed pad infrastructure, although there may be an argument for them to setup a second pad for redundancy and increased cadence

[freddo411]

If all the competitors were to order rides on Falcon 9, SpaceX would not be able to launch Starlinks as frequently as they do now. The less customers SpaceX gets, the more Starlinks they can launch, the more competitive they are with other folks.

We don't actually know how many starlink sats SX is able to build per unit time.   it seems like they can build enough for 6 or 8 launches per month.   This is just an outsiders view; not grounded in data.   Growth of the starlink system depends on things other than just launching satellites; we don't know how quickly other parts of the starlink system can scale.

We also don't know how many kuiper sats are available to launch per unit time.   We can make some bounding assumptions that kuiper needs about 100 launches in the next three years based upon guesses about satellites per rocket, and their goal to meet the FCC deadline for constellation launch.   So that's approximately an extra 33 F9 flights per year.   

The maximum cadence of Falcon 9 flights is easier to estimate, based on limits on recovery, pad recycling, and such.   It's also clear that the maximum cadence could scale upward by bringing online additional copies of pads/drone ships.

So one can't say definitively that a bunch of additional customers would exceed SX's ability to launch all customers and all starlink satellites in a timely manner.

[DanClemmensen]

[from alexphysics13:]

The only way SpaceX would be able to compensate for the surge of customer flights would be to spend more money on Falcon 9 R&D and on ramping up cadence perhaps with more droneships etc.

R&D takes time, so it cannot in general respond to a "surge". As of right now, Starship R&D is much more likely to speed up Starlink than F9 R&D and has higher ROI.

Falcon Heavy launches disrupt the schedule at LC-39A, because the pad must be reconfigured. SpaceX can perhaps increase launch cadence there if they can move Falcon Heavy launches to LC-37B, but this cannot happen unless they can negotiate for the use of that pad and then convert it. Conversion cannot really start until after the last Delta IV Heavy launches in March. SpaceX needs to provide a vertical payload integration service for NSSL in any event as part of their NSSL Phase 2 service, and I think LC-37B is a good candidate, similar to SLC-6 at VSFB.

Excellent foreshadowing the future of LC-37B.   

If ULA pays for 37B by the year, it must certainly get it off it's books if I doesn't plan to use it.   Seems like Vulcan already has the needed pad infrastructure, although there may be an argument for them to setup a second pad for redundancy and increased cadence

Kuiper intends to launch initially on Atlas V (8 launches) and then move to Vulcan (38 launches), Arianne 6, and New Glenn. I do not know how quickly ULA can launch Atlas Vs: the most they have ever done in a calendar year is 9. Vulcan and Atlas share CCSFS SLC-41  but use separate stacking facilities, so there is unlikely to be any contention for the pad, and a good chance that all the Atlas Vs will have launched before the first Vulcan Kuiper launch. (They will also share VSFB SLV-3E).

The current record for reaching the tenth launch of a new large rocket after the first launch is held by F9, which took four years. Maybe one of the three new rockets will beat this. Vulcan also has at least 13 non-Kuiper launches on its manifest before the end of 2025.

I do not think ULA will need LC-37B for Vulcan for the foreseeable future.

I am not hostile to Kuiper or to ULA: I wish for them to succeed. However, I have become deeply skeptical of all projected dates in the entire space industry, not just ULA and BO but also NASA and SpaceX.

[Jim]

It might be possible to make some changes to LC39A that would allow that pad to be reconfigured faster between a Falcon Heavy and a Falcon 9. One possibility would be adding a second Transporter-Erector permanently configured for the Falcon Heavy. If SpaceX were to acquire SLC37, that Falcon Heavy transporter-erector could be transported to that launch site.

no, the  transporter-erector uses rails at the pad

[TrevorMonty]

[from alexphysics13:]

The only way SpaceX would be able to compensate for the surge of customer flights would be to spend more money on Falcon 9 R&D and on ramping up cadence perhaps with more droneships etc.

R&D takes time, so it cannot in general respond to a "surge". As of right now, Starship R&D is much more likely to speed up Starlink than F9 R&D and has higher ROI.

Falcon Heavy launches disrupt the schedule at LC-39A, because the pad must be reconfigured. SpaceX can perhaps increase launch cadence there if they can move Falcon Heavy launches to LC-37B, but this cannot happen unless they can negotiate for the use of that pad and then convert it. Conversion cannot really start until after the last Delta IV Heavy launches in March. SpaceX needs to provide a vertical payload integration service for NSSL in any event as part of their NSSL Phase 2 service, and I think LC-37B is a good candidate, similar to SLC-6 at VSFB.

Excellent foreshadowing the future of LC-37B.   

If ULA pays for 37B by the year, it must certainly get it off it's books if I doesn't plan to use it.   Seems like Vulcan already has the needed pad infrastructure, although there may be an argument for them to setup a second pad for redundancy and increased cadence

Kuiper intends to launch initially on Atlas V (8 launches) and then move to Vulcan (38 launches), Arianne 6, and New Glenn. I do not know how quickly ULA can launch Atlas Vs: the most they have ever done in a calendar year is 9. Vulcan and Atlas share CCSFS SLC-41  but use separate stacking facilities, so there is unlikely to be any contention for the pad, and a good chance that all the Atlas Vs will have launched before the first Vulcan Kuiper launch. (They will also share VSFB SLV-3E).

The current record for reaching the tenth launch of a new large rocket after the first launch is held by F9, which took four years. Maybe one of the three new rockets will beat this. Vulcan also has at least 13 non-Kuiper launches on its manifest before the end of 2025.

I do not think ULA will need LC-37B for Vulcan for the foreseeable future.

I am not hostile to Kuiper or to ULA: I wish for them to succeed. However, I have become deeply skeptical of all projected dates in the entire space industry, not just ULA and BO but also NASA and SpaceX.

Flightrate is determined by demand, production capability, range availability and company's resources. ULA has always had capability to support dozen or more launches are year. They never had demand uptil now. Both SpaceX and RL needed to build up production capabilities from scratch, which is case for lot of new LVs. A6 will use lot of A5's production factories.
Assuming first few flights go well ULA should be able to ramp Vulcan flightrate quickly.

[DanClemmensen]

Flightrate is determined by demand, production capability, range availability and company's resources. ULA has always had capability to support dozen or more launches are year. They never had demand uptil now. Both SpaceX and RL needed to build up production capabilities from scratch, which is case for lot of new LVs. A6 will use lot of A5's production factories.
Assuming first few flights go well ULA should be able to ramp Vulcan flightrate quickly.

However, the initial several launches of a new rocket are part of the development process. Rate is therefore also constrained by the rate at which lessons learned can be applied to later rockets. Until the design stabilizes, I think a company would be reluctant to ship the second rocket from the factory until the first flight has been evaluated, etc. What's the travel time from Decatur to the Cape? Toulon to Kourou? Sure, ULA and Arianespace may eventually ramp up to a high cadence and I hope they do, but for now, I feel that the tenth launch for each will occur at least four years after the first. This is probably true for Starship also, but that's off-topic for this thread.

I'm just an armchair spectator, and I'm just naively looking at history. What is your guesstimate for the number of Vulcan flights in the first year after initial flight? What about Ariane 6? New Glenn?

[Zed_Noir]

It might be possible to make some changes to LC39A that would allow that pad to be reconfigured faster between a Falcon Heavy and a Falcon 9. One possibility would be adding a second Transporter-Erector permanently configured for the Falcon Heavy. If SpaceX were to acquire SLC37, that Falcon Heavy transporter-erector could be transported to that launch site.

no, the  transporter-erector uses rails at the pad

SpaceX can always build new transporter-erectors with SPMT mobility modules with separate variants for Falcon 9 and Falcon Heavy.

New transporter-erectors will be needed for Vandenberg and if pad SLC-37B is pickup by SpaceX for Falcon Heavy launches. Might as well build new transporter-erectors for pad SLC-40 and LC39A. Specifically have a pool of mobile transporter-erectors to share between the various SpaceX Florida pads.

Shouldn't be any harder than assembling a Starship integration & retrieval tower.

[meekGee]

The main purpose of this lawsuit is to make money for the lawyers.  Everything else is secondary.

A tad too cynical.

Bezos companies track record in executing space projects is, let’s face it, less than impressive in terms of expenditure of time and money v tangible results. Kuiper is facing a formidable competitor that is already established, with a structural advantage in launch cadence and costs that is only likely to grow wider.

Amazon shareholders are more than entitled to question if pursuing Kuiper as intended (or at all) is actually in their interests. There also do seem to be grounds to question if the Amazon board fulfilled their duties in scrutinising this project.

Well put. Lawyer make money, that's for sure, but this suit has been a long time coming.

Kuiper should have been done under BO, and then Amazon would have been free to partner them, since that's not a straight purchase like launch is.

[TrevorMonty]

The main purpose of this lawsuit is to make money for the lawyers.  Everything else is secondary.

A tad too cynical.

Bezos companies track record in executing space projects is, let’s face it, less than impressive in terms of expenditure of time and money v tangible results. Kuiper is facing a formidable competitor that is already established, with a structural advantage in launch cadence and costs that is only likely to grow wider.

Amazon shareholders are more than entitled to question if pursuing Kuiper as intended (or at all) is actually in their interests. There also do seem to be grounds to question if the Amazon board fulfilled their duties in scrutinising this project.

 If there is enough customer demand to make Starlink profitable then market is big enough to support a competitor or two. At around $2000 year only need a couple million rural subscribers to make $10B outlay worthwhile.
Kuiper will already come with AWS customer base so they don't need anywhere near 2 million rural customers. 

[Jim]

Specifically have a pool of mobile transporter-erectors to share between the various SpaceX Florida pads.

nope.  Too large to move between facilities.   And they are specific to each pad.

[Eric Hedman]

The main purpose of this lawsuit is to make money for the lawyers.  Everything else is secondary.

A tad too cynical.

Bezos companies track record in executing space projects is, let’s face it, less than impressive in terms of expenditure of time and money v tangible results. Kuiper is facing a formidable competitor that is already established, with a structural advantage in launch cadence and costs that is only likely to grow wider.

Amazon shareholders are more than entitled to question if pursuing Kuiper as intended (or at all) is actually in their interests. There also do seem to be grounds to question if the Amazon board fulfilled their duties in scrutinising this project.

Have you ever met and had to deal with an attorney?

As the comedian Steven Wright once said, "99 percent of attorneys give the rest a bad name."

Now on a more serious note, here is the board of directors that people think shirked their responsibility:

https://ir.aboutamazon.com/officers-and-directors/default.aspx

I think this group, considering their backgrounds, probably understands the risks businesses take when venturing into new arenas than most people.

[meekGee]

The main purpose of this lawsuit is to make money for the lawyers.  Everything else is secondary.

A tad too cynical.

Bezos companies track record in executing space projects is, let’s face it, less than impressive in terms of expenditure of time and money v tangible results. Kuiper is facing a formidable competitor that is already established, with a structural advantage in launch cadence and costs that is only likely to grow wider.

Amazon shareholders are more than entitled to question if pursuing Kuiper as intended (or at all) is actually in their interests. There also do seem to be grounds to question if the Amazon board fulfilled their duties in scrutinising this project.

 If there is enough customer demand to make Starlink profitable then market is big enough to support a competitor or two. At around $2000 year only need a couple million rural subscribers to make $10B outlay worthwhile.
Kuiper will already come with AWS customer base so they don't need anywhere near 2 million rural customers.

The question is not whether AWS needs Kuiper - that's the CEO's decision. (And is reasonable)

The question is why pay more Amazon dollars (and lots of them) to avoid a launch provider he doesn't like and instead use another that he happens to own.

Basically, is Amazon an independent company or is it JB's plaything.

[DanClemmensen]

If there is enough customer demand to make Starlink profitable then market is big enough to support a competitor or two. At around $2000 year only need a couple million rural subscribers to make $10B outlay worthwhile.
Kuiper will already come with AWS customer base so they don't need anywhere near 2 million rural customers.

Fine, but the Amazon shareholders are asking why they are paying three providers $10 B for 92 launches using unproven rockets when they could instead pay SpaceX maybe $5 B for the same service.

[Zed_Noir]

Specifically have a pool of mobile transporter-erectors to share between the various SpaceX Florida pads.

nope.  Too large to move between facilities.   And they are specific to each pad.

As with historical ordnance units (e.g. large & siege artillery) that breaks down for transportation. The erectors and reaction tables could be move as separate components before being joined at the launch complex before a launch campaign. Also the pads should be reworked slightly for the new transporter-erectors to handle a common set of hardware.

[M.E.T.]

The repeated assertions that excessive customer demand can force SpaceX to slow Starlink deployment against their will are non-sensical.

If SpaceX sees a customer launch as worth more than the opportunity cost of not launching another Starlink batch, then they will WILLINGLY forego the Starlink launch. If they don’t, they will not accept the customer request, or offer to launch it at a time of their choosing.

In each case, they will make the decision based on what makes business sense for SpaceX.

The idea of “forced” customer launches is absurd. People seem to love invoking “anti-monopoly” laws at the drop of a hat, but choosing to launch their own Starlink network over customer payloads is SpaceX’s right. They have no obligation to be at anyone else’s beck and call, beyond what has contractually been agreed to.