Reuse business case

[GWH]

A presentation by RUAG from 2016 on fairing reuse: https://slideplayer.com/slide/15726077/

Then this in 2021 from Tory:
https://twitter.com/torybruno/status/1407765967640399872?s=20&t=qQRqzq2TUL5N4oDMVC89iQ

But then 70+ launches vs 35+ can really change things.

[TrevorMonty]

A presentation by RUAG from 2016 on fairing reuse: https://slideplayer.com/slide/15726077/

Then this in 2021 from Tory:
https://twitter.com/torybruno/status/1407765967640399872?s=20&t=qQRqzq2TUL5N4oDMVC89iQ

But then 70+ launches vs 35+ can really change things.

MAR has best outcome for fairing but needs two helicopters operating off same vessel. Unlike booster reentry location isn't as defined reducing recovery window. Water immersion adds extra costs to refurbishment and build. Seems to be working for SpaceX so maybe worth go by ULA or RUAG.

Sent from my SM-T733 using Tapatalk

[edzieba]

A presentation by RUAG from 2016 on fairing reuse: https://slideplayer.com/slide/15726077/

Then this in 2021 from Tory:
https://twitter.com/torybruno/status/1407765967640399872?s=20&t=qQRqzq2TUL5N4oDMVC89iQ

But then 70+ launches vs 35+ can really change things.

ROAG's new fairing is out-of-autoclave, which removes the big bottleneck in production rate: the autoclave. It is far cheaper to vacuum-bag a large part and lay heaters over it (or vacuum bag it and roll into a large atmospheric pressure oven) than to build a sufficiently large heated pressure vessel to cure a part. Not only does that make parts cheaper, but it means production can scale more easily without needing to build another autoclave.
SpaceX went for fairing re-use because their fairings are autoclaved, and they could not increase production rate to meet demand without another autoclave or without a complete fairing redesign for OOA manufacture. Even zero cost savings would have been worth it to avoid that expenditure, as long as the R&D costs for reuse were lower than the bill for the autoclave (and at the time, I suspect some of the R&D costs for CFC entry behaviour would have fallen under the heading of then-BFR development) minus the lost revenue of missing launches waiting on fairing manufacture.
But for ROAG's OOA fairings, the cost of scaling production for a spike in demand is much lower. It becomes a lot harder to justify the cost of R&D for reuse, especially with a lower upfront cost. And in order to offer a 'fairing as a service' RUAG would also need to operate a fleet of recovery aircraft and/or vessels at every company launch site, eating into any cost saving per fairing.

[GWH]

Back when ULA published pricing the price difference between an Atlas 401 and 501 was $11 Million. 

Granted the 5m fairing is much more complicated as it fully encases Centaur and provides support near the payload interface.

Assume the 4m fairing is only $1M, to ball park a $12M customer cost to the 5m fairing,  take off $4M for added complexity to get to $8M?
That would still put the 50% cheaper fairing at $4M price to customer.... maybe $3M cost?

[baldusi]

SpaceX fairings are a lot sturdier than ULA/Ariane since they are designed for the extra forces of horizontal integration. It weights like twice as much (same size) as the others. So that might also make it easier for them to recover. But yes, Ariane and ULA's 5m fairing used to be 5/6M a decade ago. I don't think it could have changed too much (+/-2M)

[SoftwareDude]

Indeed, Tory Bruno likely bases calculations on a much lower launch cadence than SpaceX for fairings or whatever else needs can be reused.  It is hard to make a case for reuse when one has few launches because there is no near-term return on investment (ROI) on reuse development. ULA has to change how it processes payloads and stacks rockets to get to a higher launch cadence, making reuse economics make sense. That's a change in the business model that has to come from ULA's owners.
I watched a video where Sandy Munro pointed out that Tesla saved $60K a quarter by eliminating just one bolt from the Model Y.  One has to have the volume to realize savings.

[TrevorMonty]

The LOFTID test is scheduled for Nov. SMART engine pod recovery is relying on this technology demonstrator.

[FutureSpaceTourist]

[…]

twitter.com/scubajotaro/status/1427977249903595533

SMART or full stage recovery?

https://twitter.com/torybruno/status/1428085474325897222

SMART makes more sense for us because we specialize in complex,  high energy orbits

Is Tory saying that recovering the whole booster is tricky because ULA typically need to stage at a point (high and/or fast) that makes full stage recovery difficult?

But aren’t SpaceX, through a combination of F9 and FH, soon going to be achieving those same orbits with booster reuse? The rocket may be ‘over-sized’ for a mission, to enable recovery, but with multiple reuses still works out economic / cheaper?

Here’s a new post from Tory with a chart I don’t recall seeing before (but I may have missed it):

twitter.com/petey_nebby/status/1543342951601938437

Happy to hear that its taken this long to get 2 flight BE4’s and would love to see pictures! Bottom line is we need more options to get yo space but also please stop dropping rockets in the ocean as its not theory anymore

https://twitter.com/torybruno/status/1543345280761962497

It hasn’t been theoretical for a couple of decades. It’s about what mission set the rocket is optimized for.Low energy commercial orbits are tolerant of large propellant reserves for return flights. High energy orbits drive a different architecture, which is why we’ll use SMART

[edzieba]

Corrected graph (don't start relative charts at nonzero values!)

[LouScheffer]

This chart  strikes me as quite misleading.

Longest coast, and time to separation, are indeed engineering challenges.  But in no way is making a second stage that can coast for six hours six times harder than creating a stage that can coast for one hour.  Likewise time to separation.

Similarly, increasing the number of burns does not scale in difficulty with the number of burns required.  Furthermore, it's not clear that one burn should be the baseline.  According to usage, there should be an additional column at the left, labelled "leo with booster recovery".  This should have 3 burns as the baseline.  Then only the direct-injection missions are as hard; all the others are *easier* by that metric, requiring only one or two burns.

[JayWee]

Also, don't forget the disposal burn.

[deadman1204]

I Agree, he is obfuscating. What Tory means is that full reuse doesnt work with ULA rockets. Falcon 9 lands after a geo or interplanetary launch, so its wrong to claim reuse can't work work with higher energy orbits.

Smart will be a million times better than nothing, but this is still ULA reacting to a world where they are not used to competing with anyone.

[LouScheffer]

Also, don't forget the disposal burn.

And the space debris created.  Precisely because they have no disposal burn (too expensive) the direct injection to higher orbits leaves garbage in orbit.  And sometimes these stages explode (3 Centaurs have) creating even more debris.

[meekGee]

Corrected graph (don't start relative charts at nonzero values!)

Pathetic.
10 (sorry) 7 years after the infamous spreadsheet by Sowers that proved reuse was a fool's errand, almost a full decade, and the combined might of Boeing and LMCO will field, maybe this year maybe not, a new expendable with some, fingers crossed, plans for partial reuse. And only after being arm-twisted into the effort.

But wait!  it's on purpose!  we're not losers, we intended for it to be this way because 10-year-old talking points.

Such BS.

Everyone knows about how US mass penalty relates to high-energy orbits. This has NOTHING to do with booster reuse. And actually NOTHING to do with second stage reuse either, because you can always choose to expend it, but that's a separate topic...

[DanClemmensen]

Corrected graph (don't start relative charts at nonzero values!)

Pathetic.
10 years after the infamous spreadsheet by Sowers that proved reuse was a fool's errand, a full decade, and the combined might of Boeing and LMCO will field, maybe this year maybe not, a new expendable with some, fingers crossed, plans for partial reuse. And only after being arm-twisted into the effort.

But wait!  it's on purpose!  we're not losers, we intended for it to be this way because 10-year-old talking points.

Such BS.

Everyone knows about how US mass penalty relates to high-energy orbits.   That has NOTHING to do with booster reuse.  And actually NOTHING to do with second stage reuse either, because you can always choose to expend it, but that's a separate topic...

Clearly reuse is not viable. It'll never catch on. No company could make money with such a stupid idea. There is no company named S....X that has been launching more than two-thirds of the world's payload mass to orbit on reused boosters for the last two years.

[deadman1204]

Also, don't forget the disposal burn.

And the space debris created.  Precisely because they have no disposal burn (too expensive) the direct injection to higher orbits leaves garbage in orbit.  And sometimes these stages explode (3 Centaurs have) creating even more debris.

I guess I thought that modern centaurs were designed to not have explosion issues anymore. Don't they vent all their volatiles if it won't be able to come back to earth for disposal?

As well, doesn't ULA make an effort to put the centaur in a less risky/used orbit if it can't be disposed of? Like above the geo belt or whatnot?

[Robotbeat]

I Agree, he is obfuscating. What Tory means is that full reuse doesnt work with ULA rockets. Falcon 9 lands after a geo or interplanetary launch, so its wrong to claim reuse can't work work with higher energy orbits.

Smart will be a million times better than nothing, but this is still ULA reacting to a world where they are not used to competing with anyone.

One should mention that the infamous Sowers “spreadsheet” assumed return to launch site, not droneship, which the VAST majority of SpaceX landings have been.

Droneship landing is really enabling for Falcon 9 reuse. and I suspect Starship will eventually go in that direction over time, as will Neutron, Terran-R, etc. It’s just too much of a performance improvement to ignore, IMHO. (Although Super Heavy is currently optimized for RTLS, not downrange recovery, but if they stretch it some more, it’ll be more optimized for down-range landing.)

[DanClemmensen]

Droneship landing is really enabling for Falcon 9 reuse. and I suspect Starship will eventually go in that direction over time, as will Neutron, Terran-R, etc. It’s just too much of a performance improvement to ignore, IMHO. (Although Super Heavy is currently optimized for RTLS, not downrange recovery, but if they stretch it some more, it’ll be more optimized for down-range landing.)

Droneship is only useful when the payload mass exceeds the maximum that can be handled in an RTLS landing. There are currently no such payloads, so it's unclear that maintaining the whole Droneship support infrastructure is worthwhile. It's likely to be more cost-effect for those rare huge payloads to launch to a low orbit and then refuel the SS to reach the required orbit.

[meekGee]

Droneship landing is really enabling for Falcon 9 reuse. and I suspect Starship will eventually go in that direction over time, as will Neutron, Terran-R, etc. It’s just too much of a performance improvement to ignore, IMHO. (Although Super Heavy is currently optimized for RTLS, not downrange recovery, but if they stretch it some more, it’ll be more optimized for down-range landing.)

Droneship is only useful when the payload mass exceeds the maximum that can be handled in an RTLS landing. There are currently no such payloads, so it's unclear that maintaining the whole Droneship support infrastructure is worthwhile. It's likely to be more cost-effect for those rare huge payloads to launch to a low orbit and then refuel the SS to reach the required orbit.

Exactly.  It's almost like SpaceX can dial-a-payload, if you know what I'm talking about.

Starlink is the ultimate optimization exercise.  They can load fewer and RTLS, but clearly the barge is cheap enough.

But normal LEO satellites can get a cheaper launch if they can RTLS, and clearly that still works.

And still tho - Tory's statement is all about high energy orbits, and that pertains to upper stage reuse, not booster reuse.

It is just mixing up the talking points in order to "justify" the introduction of an expendable maybe-SMART booster in 2023.

[DanClemmensen]

Droneship landing is really enabling for Falcon 9 reuse. and I suspect Starship will eventually go in that direction over time, as will Neutron, Terran-R, etc. It’s just too much of a performance improvement to ignore, IMHO. (Although Super Heavy is currently optimized for RTLS, not downrange recovery, but if they stretch it some more, it’ll be more optimized for down-range landing.)

Droneship is only useful when the payload mass exceeds the maximum that can be handled in an RTLS landing. There are currently no such payloads, so it's unclear that maintaining the whole Droneship support infrastructure is worthwhile. It's likely to be more cost-effect for those rare huge payloads to launch to a low orbit and then refuel the SS to reach the required orbit.

Exactly.  It's almost like SpaceX can dial-a-payload, if you know what I'm talking about.
Starlink is the ultimate optimization exercise.  They can load fewer and RTLS, but clearly the barge is cheap enough.
But normal LEO satellites can get a cheaper launch if they can RTLS, and clearly that still works.

You are describing F9 here, where droneships are optimal for certain payload masses. I was describing Starship, where I believe droneships will never be optimal for any payload mass. With F9, the only ways to increase max payload mass are droneship, expended booster, or Falcon Heavy, and launches all expend the second stage. With Starship, you can increase max payload mass by refuelling, or by expending either or both stages for really crazy and rare payloads that cannot be split. You can expend several SH per year for less money that maintaining a droneship infrastructure.  By contrast with F9, dividing a payload between two launches is cheap, so aggregating payloads and then using a droneship makes no sense.