Amazon Kuiper places largest commercial launch order with ULA, Arianespace, Blue

[Star One]

As it says the big loser in all of this seems to be the small launchers. I wonder if that’s why RL have decided to already go bigger with Neutron.

[GWH]

Interesting how the most capable and most reusable vehicle is launching the least.

    18 contracted launches for Ariane 6

    12 contracted launches for New Glenn (up to 15 more)

    38 contracted launches for Vulcan

    Ariane 6 can put up 30-40 satellites per launch

    Vulcan can put up 45 satellites per launch

    New Glenn is capable of 61

Source: https://twitter.com/jeff_foust/status/1511340865834430464?s=20&t=hDNqe9fQDU_qHaJUvg6zuA and: https://arstechnica.com/science/2022/04/amazon-signs-blockbuster-launch-deal-for-its-satellite-megaconstellation/

68 launches total (not including the additional NG flights), 3072 satellites if you assume the median of 35 for Ariane 6.

If the numbers were re-arranged to be an 18+18 split for Ariane 6 & Vulcan that would mean New Glenn would need to launch 1632 of them - roughly half - requiring 27 launches to do so for a total of 63 launches.

If New Glenn were ready and capable to do those 27 launches it would save 5 launches in total, and at least $0.5B in savings without even assuming that reusable booster New Glenn costs less than Vulcan.

...

[Lars-J]

So since neither of these launch vehicles are operational (yet) - even assuming if they are operational without delays, how close is Kuiper going to come to the deadline for the # of satellites to launch by a certain date?

[JayWee]

So since neither of these launch vehicles are operational (yet) - even assuming if they are operational without delays, how close is Kuiper going to come to the deadline for the # of satellites to launch by a certain date?

Look at the other thread https://forum.nasaspaceflight.com/index.php?topic=47811
I suggest merge.

[GreenShrike]

I think the forum should bring in a sweepstakes to pick how quickly someone will mention Starship in a thread as a solution to a particular launch contract or in fact anything else.

Why shouldn't Starship get mentioned? It will be one of many LVs offering commercial services in the relevant timeframe. Seems to me that discussing the competition and alternatives would be highly relevant in a thread about the "largest commercial launch order"...

For example, will there also be an award for mentioning Japan's H3 or ISRO's GSLV Mk3? Because Amazon's apparent strategy of "all of the above" for making certain they have sufficient and timely launches available from *somebody* seemingly only lacks contracts with those two. (China and Russia being out for obvious reasons, of course.)

Of the two, H3 is more of a surprise to me, as JAXA/Mitsubishi are really trying to design it with an eye towards commercialization. ISRO doesn't seem to have much excess capacity for their GSLVs to be of much interest.

And what about an award for mentioning OneWeb? Their Gen2 constellation of ~6300 sats is supposed to start launching mid-decade in the same timeframe as Kuiper's launches, but Amazon calling dibs on 68 flights worldwide has to suck a lot of the oxygen out of the available launch capacity.

ULA, in particular, with their 35 pre-existing launches and now 38 more, is looking at launching an average of ~12 Vulcans a year if those are all launched between 2023 and mid-2029 (i.e. when Kuiper needs to be complete by), and I don't think ULA is going to hit a 12 per year pace for Vulcan right out of the gate. Add in the remaining Atlas 5 launches and you're looking at something approaching SpaceX's (them, again!) launch rates. It should really keep Canaveral's range operators busy for the foreseeable future -- between SpaceX, ULA and Blue, will there be time for anything else?

Ariane 6's 18 launches is 4.5 per year if ending in 2026, or 2.5 if by 2029. Add in their GTO and European government launches, and I'm not certain how many more they could sell. And A64 really is kinda small for LEO -- at 21t it's only about the same capacity as an expendable Falcon 9 -- so even if OneWeb managed to get a half dozen A64 launches a year, that's not going put much of dent in their requirements.

Amazon didn't tap any of the upcoming medium launchers -- Neutron, Terran R, Beta -- so maybe they'll have capacity mid-decade. They'll certainly want to sell OneWeb a bunch of flights, at any rate.

I think SpaceX is out for Kuiper launches, though.  Blue, ULA and ArianeGroup would need to all seriously mess up their respective launchers' debuts and/or fail to quickly ramp their launch cadence -- and even then, H3 should have capacity, and ISRO might actually get their SCE-200 powered launcher ready. On the other hand, if things do go pear shaped (highly unlikely, given ULA's and ArianeGroup's reputations), then Amazon might need to pull a OneWeb and buy from SpaceX to meet their 2026 deadline. In the end, all the money in the world won't help if there's nothing else to buy.

And then there's New Glenn, but will competitors consider Blue to be as unfavourable as SpaceX? Blue is one step removed from Amazon/Kuiper, after all, and Telesat is (hopefully) launching on NG, so maybe not.

On the other hand, OneWeb has already held their nose and bought SpaceX launches -- and OneWeb doesn't have Amazon's mountain of cash to throw at all the high-cost vehicles. I'm thinking that OneWeb might have to seriously start getting their own launchers under contract for their Gen2 constellation, or be caught short when they want to start launching.

I wonder how much OneWeb's Soyuz situation situation spurred Amazon to spread the love, or just added emphasis as to why not going sole-source is a good idea. And I wonder if OneWeb will necessarily be following suit for Gen2.


All-in-all, this launch contract is certainly momentous, for both the launch market, as well as other megaconstellation operators.

[TrevorMonty]

I don't really consider Neutron not winning anything meaningful, as they are at least as far out/risky as New Glenn and of course Amazon isn't going to pick them over Blue's launcher.

Amazon will have more shells to build out after this one, after all...

There is likely to be a few of RL's components on each of these satellites.

Sent from my SM-T733 using Tapatalk

[XRZ.YZ]

Roaring Twenties for Rockets.

[Robotbeat]

Spending $5-8bn (at a guess) on launches is also quite exciting.

LOL!!  Amazon has ~$90B in cash to spend.  And that doesn't include Bezos' personal wealth.  He/they (Amazon) don't give a rats ass about spending ~$10B on some rockets.  No more than Musk cares about spending ~$3B on a twitter conniption (he's already made ~$1.5B on that transaction - in three days!).

The amount of cash these two men hold (not mention what Amazon, Apple, FB, etc, has) is obscene...

They don’t hold that much actual cash. Once all the fees are due, about half that.

This isn’t a personal Bezos project, it comes out of Amazon’s cash reserves. $10B (just for launch!) out of $80B in cash for a side project not directly related to their core businesses is non-trivial. Add another $10B for the satellites.

$10B is more than SpaceX has spent or will spend on Starship before its first crewed launch. If they operated like SpaceX, Amazon could develop and test their own reusable launch vehicle and launch the whole constellation for less than what they’re paying these 3 largely expendable launch providers.

It’s a really impressive bonfire of cash! Lots of expendable stages in the ocean.

[Robert_the_Doll]

Amazon have signed Kuiper launch contracts with ULA, Blue Origin, and Arianespace. A lot of launch contracts.
38 Vulcan launches, 18 Ariane 6 launches, and 12 New Glenn launches (with an option for an additional 15).

It's odd that Amazon reports having contracted for 38 Vulcan launches when Tory Bruno reports that Vulcan has 35 contracted launches in total so far.

If you are referring to the Off-Nominal podcast interview, Tory Bruno stated he had sold "over 35" Vulcan launches.

[ZachF]

Interesting how the most capable and most reusable vehicle is launching the least.

    18 contracted launches for Ariane 6

    12 contracted launches for New Glenn (up to 15 more)

    38 contracted launches for Vulcan

    Ariane 6 can put up 30-40 satellites per launch

    Vulcan can put up 45 satellites per launch

    New Glenn is capable of 61

Source: https://twitter.com/jeff_foust/status/1511340865834430464?s=20&t=hDNqe9fQDU_qHaJUvg6zuA and: https://arstechnica.com/science/2022/04/amazon-signs-blockbuster-launch-deal-for-its-satellite-megaconstellation/

68 launches total (not including the additional NG flights), 3072 satellites if you assume the median of 35 for Ariane 6.

If the numbers were re-arranged to be an 18+18 split for Ariane 6 & Vulcan that would mean New Glenn would need to launch 1632 of them - roughly half - requiring 27 launches to do so for a total of 63 launches.

If New Glenn were ready and capable to do those 27 launches it would save 5 launches in total, and at least $0.5B in savings without even assuming that reusable booster New Glenn costs less than Vulcan.

...

Looking at those numbers then it appears that Kuiper sats will likely be ~500-600kg each.

Don’t forget the 9 Atlas V launches too. That’s probably ~300 satellites total.

[Robotbeat]

From what I can tell, Ariane 64 is about $6000/kg IMLEO. Vulcan is probably similar, although who knows what New Glenn is. If Kuiper were to compete with the 40,000 satellite Starlink constellation at 600kg apiece, that'd cost them $144B to launch them. Almost double the total cash reserves of Amazon. (and that doesn't count the satellites.(

Starship, at ~$100/kg (100-150 tonnes at $10-15 million a launch, which seems pretty reasonable to me), that's just $2.4B, maybe a quarter of what Amazon just committed to. Starship + Starlink will start launching before any of the new batch does. And probably even before the Atlas V launches the first Kuiper satellites.

To go toe to toe with SpaceX without an operational fully reusable launch vehicle will mean Amazon runs out of money. They're going to need the Jarvis-ified New Glenn, even if they're prepared to bonfire cash to stay in the game.

...ooo, this is getting good!

[M.E.T.]

From what I can tell, Ariane 64 is about $6000/kg IMLEO. Vulcan is probably similar, although who knows what New Glenn is. If Kuiper were to compete with the 40,000 satellite Starlink constellation at 600kg apiece, that'd cost them $144B to launch them. Almost double the total cash reserves of Amazon. (and that doesn't count the satellites.(

Starship, at ~$100/kg (100-150 tonnes at $10-15 million a launch, which seems pretty reasonable to me), that's just $2.4B, maybe a quarter of what Amazon just committed to. Starship + Starlink will start launching before any of the new batch does. And probably even before the Atlas V launches the first Kuiper satellites.

To go toe to toe with SpaceX without an operational fully reusable launch vehicle will mean Amazon runs out of money. They're going to need the Jarvis-ified New Glenn, even if they're prepared to bonfire cash to stay in the game.

...ooo, this is getting good!

Indeed. And those were just the financial implications of increased launch cost - and mostly focused on the initial constellation (which will have to be continuously replenished). We haven’t even touched on the impact of slower launch cadence and resultant schedule lengthening.

Historically, BO, ULA and Ariane don’t move particularly fast. Can they ramp up their cadence to match SpaceX’s already ludicrous F9 launch pace - not to mention the hockey stick curve Starship’s introduction will bring to the SpaceX capability graph?

Highly unlikely.

[Robotbeat]

I actually would be surprised if 5 years from now, Ariane 6, Vulcan, and New Glenn COMBINED have reached the one-a-week rate Falcon 9 is doing right now.

The most Atlas V ever launched was 9 times to orbit a year. The most Ariane 5 ever launched was 6 times per year (even Ariane 4 only ever flew up to 12 times per year and the Ariane family altogether only ever reached 12 per year maximum). The most Blue Origin launched was 0 times to orbit per year (and New Shepard, a much tinier vehicle with far fewer constraints than an orbital launcher, only flew at most 6 times per year suborbitally.) The most optimistic numbers here for history would point to 27 launches per year altogether.

It'll be a huge challenge to even do 12 launches per year altogether on average through 2026, considering they're all starting at zero launches so far this year.

[M.E.T.]

I actually would be surprised if 5 years from now, Ariane 6, Vulcan, and New Glenn COMBINED have reached the one-a-week rate Falcon 9 is doing right now.

Now, I am the first to emphasise that launch revenue is fairly insignificant in the context of the huge potential of the LEO internet market. But as a cost element to the provision of your competitive internet service, an extra few billion in annual launch costs, coupled with a slower roll out cycle will make you bleed cash every year while trying to compete with a rival who does not face those disadvantages. Especially if that rival is 5 years ahead of you schedule wise, and already in the net positive earnings phase while you are still trying to get a service up and running and attract customers.

Not to mention that the rival will not be standing still, but instead will be constantly innovating to try and increase its already substantial lead.

Frankly, this has the potential to be a huge cash drain for Amazon over many years if they don’t play it smart.

Edit

Also, I admit my confusion as to which of the two threads to post in. Just trying to engage with Robotbeat’s valid points raised.

[TrevorMonty]

I wouldn't underestimate ULA's ability to launch at high cadence. They've not had reason to in past as most launches are one off special missions. Kuiper launches will all be same which helps speed things up especially with payload preparation.

Sent from my SM-G570Y using Tapatalk

[Robotbeat]

I wouldn't underestimate ULA's ability to launch at high cadence. They've not had reason to in past as most launches are one off special missions. Kuiper launches will all be same which helps speed things up especially with payload preparation.

Sent from my SM-G570Y using Tapatalk

That's true, and if SMART reuse works, that should help with launch rate. HOWEVER... this is going to take a while to ramp up launch rate. I doubt SMART reuse will enter into the equation until maybe the very end.

But this is a huge overall global launch rate, between Starlink, OneWeb, and Kuiper. Last year, we got 144 orbital launches globally, eclipsing finally the Cold War peak of 139 in 1967.

We are, in the very least, in the midst of a new golden age of spaceflight.

[Asteroza]

I wonder how close this will put eastern range ops to their limits? We know there was refurbishment and other other work done to increase launch cadence, but I suspect that was assuming the driver would be SpaceX and not so much from any others. With Atlas5/NG/Vulcan all pushing up their cadence, can the eastern range accommodate that with the equipment and people they have now, or are they going to need another upgrade and added personnel?

For that matter, how much pressure will this put on Vandenberg as well, as they haven't had the same level of upgrades, just as SpaceX was increasing their cadence to speed up Starlink, while now also taking on the OneWeb buildout?

[Robotbeat]

I suspect SpaceX will still dominate the launch rate in terms of range resources.

[su27k]

Remember the previous "largest commercial launch order in history", and how "well" that worked out for the company buying the launches? (OneWeb Just Placed What It’s Calling the Largest Commercial Launch Order in History)

Yeah, I'm seeing a high probability history will repeat itself: If you make decisions that doesn't make sense economically, sooner or later it'll come home to roost and bite you in the behind. And this time don't count on Musk being magnanimous and lend a helping hand, he may be willing to help a non-competitive competitor who has already fired Greg Wyler, but it's doubtful there's the willingness to pull Bezos out of the mess of his own making.

[Robotbeat]

However, Amazon actually has money and is spreading the risk among multiple providers. But I agree. Not picking SpaceX for ANY of the launches is probably not helping their credibility. If I were an Amazon shareholder, I'd be asking why they're burning billions more of their cash pile than is necessary to launch Project Kuiper. And doing so on completely unproven rockets, even while multiple potentially cheaper launch vehicles in the same class are under development and planned to be completed soon (Neutron, Terran-R, etc).

(Realize that at the end of 2020, ULA was only worth about $1.5 Billion. Amazon could've BOUGHT ULA.)