Comparing Astra and Rocket Lab business models

[PM3]

edit/gongora:  Split off from the Astra thread as the numbers in the examples have started diverging greatly from what Astra is intending to do.

Satellite insurance risks and therefore costs are directly linked to the reliability of the launcher.

Exactly. And on the price of the launch which is insured. Let's do an example:

We want to send 100 kg of satellite(s) to SSO. We payed $ 3 Mio. for the satellite(s) and now want to buy a dedicated launch.

Rocket Lab:

- launch costs $ 6 M
- let's assume 99 % launcher reliability (one failure in 100 launches)
- insurance must cover the 1 % failures + management costs and earn some money, so insurance premium may be about 3 % of total costs (check: Ariane 5 insurance premiums are around 3 %)

On the bottom line, you will pay ($ 3+6 M) * 1.03 = $ 9.27 M total

Astra:

- launch costs $ 2.5 M
- let's assume a launcher reliability of 80 % (one failure in 5 launches)
- 19 % less reliability => ~ 19 % more insurance costs, i.e. ~ 22 % of total costs

On the bottom line, you will pay ($ 3+2.5 M) * 1.22 = $ 6.71 M total.

So you save lots of the money when launching with Astra. No matter if it is a single satellite or 20 nanosats, and including full insurance.

The cheaper your satellites are, the more you will save with Astra (in % of total costs). So Astra will get the high volume constellation contracts and achieve the high launch cadence, which is necessary to earn money with cheap launches. Rocket Lab will get the low-volume contracts for rather expensive custom satellites and stay with a lower launch cadence.

This all presuming that Astra manages 80 % reliability and scaling production to at least ~ 2 rockets per month.


Btw, Astra is not tied to Kodiak, e.g. the DARPA launches will go from two different sites. Launch sites would compete for high volume Astra launches, which gives negotiation power to Astra. Don't believe that they would pay $ 0.5 M per launch for a high launch frequency at Kodiak.

[envy887]

iif the launcher is unproven or consistently unreliable you may not even be able to get insurance on your payload, or on the flight costs. You are playing with Gambler's Ruin i.e. just because the last one failed doesn't mean the next one will fly and vice versa. ...

If the vehicle reliably fails at a 20% clip, insurers will gladly charge 30% of the insured value to cover it. That's not gambling - it's a guaranteed profit as long as they can cover enough launches.

[edzieba]

To use the earlier Fedex analogy:

If you have a Priceless Ming Vase you want to ship, you send it via a white-glove courier service and insure it up the wazoo. You only have one, and have no real way to replace it. You play white-glove-courier prices and maximum-gouge insurance costs due to that risk.

If you want to ship 10,000 moulded porcelain vases that all need to go to slightly different places, then you Fedex them in individual boxes or small packs. Some may break in transit, so you make some extras. Each vase may cost a few bucks, and each shipping charge may be close to the same. It's inconvenient to have them break and need to send another shipment, but it may well be cheaper to skip insurance altogether if the vases and shipping cost little enough that you can afford to lose a few.

[gongora]

Astra is targeting reliabilty far higher than 80%. 

[Lars-J]

I'm still not sure that targeting a 95% reliability (which I assume they are) is going to be enough.

Payloads are still important to smaller customers. And whereas large constellations can handles such losses more easily, I remain skeptical that smallsat launches will capture much of that market, when Soyuz and F9 can launch constellations more effectively and cheaper.

[TrevorMonty]

RL prices aren't set in concrete, plus they are moving to a RLV.

[gmbnz]

If Astra need 100 flights per year to be profitable I hope they have a lot of runway in their financing... Both because they won't have the desired launch rate for a while but also because I bet those early rockets cost a lot more than the later ones.

All companies (SpaceX, RL, VO) have missed both launch and then scaling targets by miles. The fact that they are still going suggests that whoever is holding onto the purse strings at those companies is more realistic than the PR machines.

Not really related to Astra but here's one of my all time favourite hypes - if the business plan actually required this then it would be bad:
https://twitter.com/jeff_foust/status/964214407306928128

RL prices aren't set in concrete, plus they are moving to a RLV.

I haven't seen any pricing the a reused stage, has RL ever said anything? It's be interesting to know if savings are a similar proportion to SpaceX.

[ringsider]

iif the launcher is unproven or consistently unreliable you may not even be able to get insurance on your payload, or on the flight costs. You are playing with Gambler's Ruin i.e. just because the last one failed doesn't mean the next one will fly and vice versa. ...

If the vehicle reliably fails at a 20% clip, insurers will gladly charge 30% of the insured value to cover it. That's not gambling - it's a guaranteed profit as long as they can cover enough launches.

I would reply explaining why that is incorrect but I don't like that his thread was split off from the Astra discussion.

[TrevorMonty]

If Astra need 100 flights per year to be profitable I hope they have a lot of runway in their financing... Both because they won't have the desired launch rate for a while but also because I bet those early rockets cost a lot more than the later ones.

All companies (SpaceX, RL, VO) have missed both launch and then scaling targets by miles. The fact that they are still going suggests that whoever is holding onto the purse strings at those companies is more realistic than the PR machines.

Not really related to Astra but here's one of my all time favourite hypes - if the business plan actually required this then it would be bad:
https://twitter.com/jeff_foust/status/964214407306928128

RL prices aren't set in concrete, plus they are moving to a RLV.

I haven't seen any pricing the a reused stage, has RL ever said anything? It's be interesting to know if savings are a similar proportion to SpaceX.

They need to recover and refly a
boosters before deciding on pricing. I'm sure there will be significant discounts for first 1 or 2 reflown boosters. Planet would be likely first customer, willing to take risk for cheap price, especially as they can handle loss of few cubesats.

[high road]

If Astra need 100 flights per year to be profitable I hope they have a lot of runway in their financing... Both because they won't have the desired launch rate for a while but also because I bet those early rockets cost a lot more than the later ones.

All companies (SpaceX, RL, VO) have missed both launch and then scaling targets by miles. The fact that they are still going suggests that whoever is holding onto the purse strings at those companies is more realistic than the PR machines.

Not really related to Astra but here's one of my all time favourite hypes - if the business plan actually required this then it would be bad:
https://twitter.com/jeff_foust/status/964214407306928128

RL prices aren't set in concrete, plus they are moving to a RLV.

I haven't seen any pricing the a reused stage, has RL ever said anything? It's be interesting to know if savings are a similar proportion to SpaceX.

RL say reusability is only about keeping up with customer demand, as they can't produce Electrons fast enough. So no price reduction.

OTOH, this is coming from the company that keeps saying they're launching once a month, which is a considerable overestimate. So YMMV

[john smith 19]

Just a note.

Historical ELV reliability figure is about 5% failure IE 1 in 20. Ariane 5 and Atlas V have managed better

BTW this is not the first time people have asked what happens if you trade reliability for cost.

OTRAG in Germany did it in Germany in the 70's and  Loral were planning to do it for bulk re-supply to the ISS (1 tonne of low value easily replaceable stuff).

And TBH it's not unreasonable strategy if the payloads are low value (like the proverbial ISS toilet paper and bottle water run) or (like) astra planning to launch smallish clumps of much bigger constellations at each launch so any one launch failure is not that big a deal..

However that raises the question when a launch fails (and your business model expects failure) and how the MIB and FAA react.  SOP is to stop all launches till a cause is found but it seems astra want more of a "Let's keep launching in the mean time while we investigate and change the design of the next batch accordingly"

Time will tell how well that sits with the FAA.
Time will tell how

[TrevorMonty]

Just a note.

Historical ELV reliability figure is about 5% failure IE 1 in 20. Ariane 5 and Atlas V have managed better

BTW this is not the first time people have asked what happens if you trade reliability for cost.

OTRAG in Germany did it in Germany in the 70's and  Loral were planning to do it for bulk re-supply to the ISS (1 tonne of low value easily replaceable stuff).

And TBH it's not unreasonable strategy if the payloads are low value (like the proverbial ISS toilet paper and bottle water run) or (like) astra planning to launch smallish clumps of much bigger constellations at each launch so any one launch failure is not that big a deal..

However that raises the question when a launch fails (and your business model expects failure) and how the MIB and FAA react.  SOP is to stop all launches till a cause is found but it seems astra want more of a "Let's keep launching in the mean time while we investigate and change the design of the next batch accordingly"

Time will tell how well that sits with the FAA.
Time will tell how

That downtime while failure are investigated, means no cashflow and customers switching LVs as they don't want to wait. A failure means cost of that LV needs to be recovered from next few launches profit margin.

Better to start with reliable and more expensive LV and work towards reducing launch costs. Customers will always come back tomorrow to reliable LV if price has been reduced to competitive level.

[john smith 19]

That downtime while failure are investigated, means no cashflow and customers switching LVs as they don't want to wait. A failure means cost of that LV needs to be recovered from next few launches profit margin.

That's how it normally happens. Which means Astra need either a) A much expedited investigation process b) FAA allows them to continue launching during the investigation.

Better to start with reliable and more expensive LV and work towards reducing launch costs. Customers will always come back tomorrow to reliable LV if price has been reduced to competitive level.

Well that's the question isn't it? 

Is the increased up front engineering analysis, presumably higher quality parts (and possible multiple redundancy) worth the (projected) increased business?

Astra seem to be saying that it is not.

[PM3]

Astra target a rocket production price that is closer to $ 250.000 than 500.000. What is Electron production cost?

Electron can launch only from dedicated, expensive pads which need a year to build and then people to maintain. "Astra says its Rocket 3.0 and all associated launch infrastructure fits within four semi-trailers."

Astra is going for mind-boggling low cost and quick iterative development. Just like SpaceX. Both Californian Companies.

And I wonder how much Rocket Lab's reusability initiative is driven by Astra, forcing them to lower cost (on the long run).

https://twitter.com/SciGuySpace/status/1230899624015212544

[jongoff]

Time will tell how well that sits with the FAA.
Time will tell how

That downtime while failure are investigated, means no cashflow and customers switching LVs as they don't want to wait. A failure means cost of that LV needs to be recovered from next few launches profit margin.

Better to start with reliable and more expensive LV and work towards reducing launch costs. Customers will always come back tomorrow to reliable LV if price has been reduced to competitive level.

While I'll admit I'm a skeptic of the wisdom of a mass-produced lower reliability rocket, I think there are some misconceptions here. When it comes to orbital launches, the FAA only is in charge of regulating 3rd party safety. In most cases when a rocket fails the stand-down isn't imposed by the FAA, it's driven by the launch provider wanting to know what's going on. Unless the failure happened in a way that specifically implicated the launch provider's ability to ensure 3rd party safety, I'm not sure if the FAA can actually force a stand-down. If you've really drunk the koolaid on this being a deliberately less than 100% reliable vehicle, you could in theory keep launching while you do the failure investigation in parallel (or even not do a failure investigation). I'm not a fan of that approach, and being too cavalier about that might jeapordize your ability to get future launch licenses and 3rd party liability insurance (which IIRC is required), but I don't think there's actually a requirement for Astra to stand down after a failure.

Not saying that's how I would do things, just trying to question assumptions.

~Jon

[john smith 19]

While I'll admit I'm a skeptic of the wisdom of a mass-produced lower reliability rocket, I think there are some misconceptions here. When it comes to orbital launches, the FAA only is in charge of regulating 3rd party safety. In most cases when a rocket fails the stand-down isn't imposed by the FAA, it's driven by the launch provider wanting to know what's going on. Unless the failure happened in a way that specifically implicated the launch provider's ability to ensure 3rd party safety, I'm not sure if the FAA can actually force a stand-down. If you've really drunk the koolaid on this being a deliberately less than 100% reliable vehicle, you could in theory keep launching while you do the failure investigation in parallel (or even not do a failure investigation). I'm not a fan of that approach, and being too cavalier about that might jeapordize your ability to get future launch licenses and 3rd party liability insurance (which IIRC is required), but I don't think there's actually a requirement for Astra to stand down after a failure.

Not saying that's how I would do things, just trying to question assumptions.

~Jon

I did not know this. I had assumed they would be grounded, in the same way aircraft are when it's believed there may be a type wide issue.

So, in principle, with the right customer launches could continue following a launch failure.  Interesting.

[PM3]

Here we go: Astra's no-frills low-cost business model beats Rocket Lab's high-quality high-cost model. NASA contract went to unproven rocket because auf much lower price. For sure they did an overall calculation including risk hedging, and Astra hat the better offer on the bottom line.

The competition came down to Astra and Rocket Lab, whose proposal had several strengths, but also a “significantly higher” price than Astra. NASA concluded that “after reviewing the benefits associated with Rocket Lab’s proposal and Astra’s assessed risk in combination with their significantly lower price, the technical benefits do not offset the significant difference in price” and selected Astra for launching the TROPICS mission

https://spacenews.com/spacex-bid-on-launch-of-nasa-cubesat-mission/

[Davidthefat]

IMO, these discussions really come to no or the wrong conclusion as they are typically done with out of date information (you have no idea what actual customers are actually paying for missions). Each mission isn't priced the same either. A lot of the information and numbers that people attach to things are typically not public and can be very off the mark.

[trimeta]

I kind of wonder to what extent Astra is moving away from "our rockets fail more!" and more into "we carry less payload, but if that's all you need we have lower launch costs" as a selling point. Although if they're really charging like 1/3 of Rocket Lab, then even if they're able to carry as much as 50% of Electron's payload, that's actually better from a cost-per-kg standpoint. (Not that we can say much about their payload capacity, since it seems like they won't get very much payload until they're flying Rocket 4; even the TROPICS mission only requires 6U of payload per launch.)

That said, another factor is profit per launch: if Rocket Lab is selling launches for $7.5 million but they only cost $4 million internally (and with first-stage reuse, they should be able to get below that), that's a profit of $3.5 million per launch. Even if they end up booking half as many launches as Astra, they could still end up with more profit.

[PM3]

Astra business model confirmed, as discussed above. Chris Kemp on the upcoming Tropics launches:

“I know the team will do everything we can to make sure all three launches and all your satellites are deployed, but it’s good to know that the price point of three launches allows NASA to enable a mission where even if only two are successful […] it is nice to know that even NASA is designing constellations so that the overall constellation performance is the end goal, not thinking about every single satellite, every single rocket launch.”

“If two out of the three [launches] are successful, it’s not mission failure,” he said. “It’s just a lower refresh rate for the constellation.”

Article including Rocket Lab price comparison:

https://techcrunch.com/2022/05/13/astras-playing-the-long-game/amp/