Author Topic: Reusability effect on costs  (Read 147060 times)

Offline gospacex

  • Senior Member
  • *****
  • Posts: 3028
  • Liked: 535
  • Likes Given: 604
Re: Reusability effect on costs
« Reply #480 on: 05/21/2017 08:13 PM »
Well, two out of two reused boosters have taken several months to refurbish so far. And I'd imagine that SpaceX is going to reduce those times gradually, not suddenly.

Don't confuse time to re-fly with time to refurbish. The booster can't re-fly until there's a customer willing to re-use.

At some point, SpaceX might simply stop giving customers a choice.

The contract is to deliver payload to a specified orbit. Such factors as paint color on the rocket, composition of the tank wall alloy, prop load mass, etc, is up to SpaceX and not something customer particularly cares about. Whether this booster is new or reused (and how many times reused) may become similar.

Offline guckyfan

  • Senior Member
  • *****
  • Posts: 6441
  • Germany
  • Liked: 1620
  • Likes Given: 1484
Re: Reusability effect on costs
« Reply #481 on: 05/21/2017 08:17 PM »
Block 3/4 boosters sound like candidates for expendable launches. The marginal cost of a couple of weeks refurbishment of a single core would very low, even if they have been written off for regular relaunches. Of course block 5 FH refurbishment costs would presumably be even lower, but using an expendable block 3/4 core would save one flight's portion of lifetime on three FH cores.

I see your argument about expendable flights. But those will not be many.

Assume only 10 block 5 cores built before reuse becomes standard. Assume only 10 reflights per core. Assume there will still be a few customers like NASA and DoD who want new cores for a while. They will soon have 20 cores worth 200 flights. That is enough to last them well into beginning to launch the constellation in 2019.

Online M.E.T.

  • Full Member
  • ****
  • Posts: 471
  • Liked: 199
  • Likes Given: 13
Re: Reusability effect on costs
« Reply #482 on: 05/21/2017 08:26 PM »
Block 3/4 boosters sound like candidates for expendable launches. The marginal cost of a couple of weeks refurbishment of a single core would very low, even if they have been written off for regular relaunches. Of course block 5 FH refurbishment costs would presumably be even lower, but using an expendable block 3/4 core would save one flight's portion of lifetime on three FH cores.

I see your argument about expendable flights. But those will not be many.

Assume only 10 block 5 cores built before reuse becomes standard. Assume only 10 reflights per core. Assume there will still be a few customers like NASA and DoD who want new cores for a while. They will soon have 20 cores worth 200 flights. That is enough to last them well into beginning to launch the constellation in 2019.

There is always the possibility that the 100 flight target proves optimistic. Look, even 10 reflights mean a revolution in the cost of access to space, so that would not be a disaster. But I can see a scenario where cores get reflown 10 times, or maybe 20 o 30 times, rather than 100 times.

That still is a huge achievement. But would require the manufacture of more cores than just 10.

Offline guckyfan

  • Senior Member
  • *****
  • Posts: 6441
  • Germany
  • Liked: 1620
  • Likes Given: 1484
Re: Reusability effect on costs
« Reply #483 on: 05/21/2017 08:49 PM »
I was talking about 10 reflights only per core. The most conservative of all guesses. Very soon 10 cores with 10 flights each makes 100 available flights.

Offline gospacex

  • Senior Member
  • *****
  • Posts: 3028
  • Liked: 535
  • Likes Given: 604
Re: Reusability effect on costs
« Reply #484 on: 05/21/2017 09:25 PM »
There is always the possibility that the 100 flight target proves optimistic. Look, even 10 reflights mean a revolution in the cost of access to space, so that would not be a disaster. But I can see a scenario where cores get reflown 10 times, or maybe 20 o 30 times, rather than 100 times.

After you learn which part of the booster wears off faster than others after 10 flights, you redesign and strengthen it (or have a planned R&R for it). Thus, you gradually make boosters have more reuses.

Online macpacheco

  • Full Member
  • ****
  • Posts: 828
  • Vitoria-ES-Brazil
  • Liked: 332
  • Likes Given: 2498
Re: Reusability effect on costs
« Reply #485 on: 05/21/2017 10:09 PM »
This is someting I've been pondering on. Specifically, the difficult decision of when to stop reusing Block 3 and 4 boosters because Block 5 is available. On the one hand, just throwing away a perfectly reusable booster seems a waste of tens of millions of dollars. On the other hand, they cost more and take longer to refurbish than the Block 5 with its optimized reusability features.

So come next year, they will have a dozen or so Block 5's in operation, but also have maybe a dozen or more Block 3 and 4 landed cores still sitting in storage, some of which have been reflown once, and some which haven't been reflown at all.

So how to decide when to stop using pre-Block 5 cores that are still available?
Remanufacture those into FH side boosters. Its been done already. At the factory they should be able to fully upgrade to the latest specs being produced.
But several of those will instead be used for max performance Inmarsat like expendable launches.
Want a new booster, unless you pay a lot more, its limited to ASDS recovery performance.
A used booster is ok, you can give up your reused discount for an expendable ride with max performance.
Several dozen new F9 boosters will still be made. The ones in the barn right now are just for immediate needs.
In the short term, SpaceX priority is to sign as many reuse customers as possible, even if the reflight is in itself expendable.

The move from M1C to M1D wasn't just about performance. SpaceX managed to get more performance out of that upgrade, while making the engine less expensive.
SpaceX ended up with a beast of a rocket with quite low cost.
Reuse just once or twice is already a economic great deal for a F9 Block II/III/IV. Because they are so cheap to make to begin with.
The big deal about reuse is to free up limited Hawthorne production capability and the ability of McGregor to test new stages. Refurb booster stages apparently are only going to be static fired. That is already expected to take place with the upcoming BulgariaSat launch.
Before a new F9 booster is test fired as a complete unit, each engine must be tested individually. A F9/FH US only needs a single M1Dvac to be tested. Are M1Dvac tested outside their stages ?
« Last Edit: 05/22/2017 03:26 AM by macpacheco »
Looking for companies doing great things for much more than money

Online john smith 19

  • Senior Member
  • *****
  • Posts: 6402
  • Everyplaceelse
  • Liked: 870
  • Likes Given: 5498
Re: Reusability effect on costs
« Reply #486 on: 05/21/2017 10:27 PM »
No Human operator. It was a big computer sitting on the ground. Something that was way to heavy for a rocket to lift but was fast enough to do high accuracy navigation steering and engine cutoff timing. It could have been sufficient for a computer controlled landing on a landing pad. This accuracy was why it was still in use up to the 1990's.
That's odd. I thought the only ICBM system that used a guidance computer based on the ground was the Titan 1,after which digital IC's meant you could build them small enough to put on board.

Quote from: oldAtlas_Eguy
The ICBM version was designed such that the X band tracker up-link would capture the Atlas in-flight during the sustainer phase and issue SECO that would then result in the warhead hitting the specified target. The Atlas's were launched at ~1min intervals to be able to give the shared guidance station the ability to acquire and guide each of several missiles.

The main reason that Atlas was not made into a reusable LV was that it was a throw away weapon system. That after they were launch there was no reason to recover and use then again. Possibly even the probable no facilities left standing either.

One of the first engines the F-1 was so good that it could have been reused. Also the RL-10 was so good as well that it could have done a dozen restarts in-space if they could figure out how to get a dozen restart cartridges on the engine. The RL-10 actually has 4 cartridge start ports for use and has been used for 4 in-space starts. So there were engines in the mid to late 1960's that could have been used to create fully reusable LV's. In fact they were almost used on shuttle. But because development funding was limited the cost of operations was sacrificed for lower development costs.

Added:
Now back from history to the effect of reusability on costs. Reusability has been the goal since the 1960's but has been only partially successful if even that until now. SpaceX tells us that their solution will lower their costs and has offered customers a lower price for use of used boosters. But only once their use of used booster exceeds that of new ones can it be said that their reusability solution has become successful.
Odd I though the RL10 had gone to some kind of augmented spark ignitor, but that's more recently.  IIRC the J2-S went to a cartridge start system from the (complex) tank head start tanks. I didn't think it was ever deployed. IIRC they had gone further and worked out how to spin up the turbines without starter cartridges, leading to unlimited starts.
"Solids are a branch of fireworks, not rocketry. :-) :-) ", Henry Spencer 1/28/11  Averse to bold? You must be in marketing."It's all in the sequencing" K. Mattingly.  STS-Keeping most of the stakeholders happy most of the time.
So you're going to Mars to seek a better life. What does that mean to you? Always spot a fanbois by how they react to their idols failures.

Offline WmThomas

  • Full Member
  • *
  • Posts: 119
  • An objective space fan
  • Liked: 49
  • Likes Given: 892
Re: Reusability effect on costs
« Reply #487 on: 05/21/2017 11:43 PM »
The Atlas D,E,F,G, and H all used pressure fed vernier engines for steering. ,,,

The Atlas vernier engines produced 2,000lbs thrust each. And had the option for its own small shared turbo pump for independent operation after SECO (sustainer engine shutdown).

A modified Atlas design with 4 verniers landing legs and some areosurface controls could have in the 1960's done a VTVL. ...

Thanks for the information. I suppose the other companies could develop some kind of strap-on landing rockets. That's a good idea.

But would the mass penalty be too great? The strap-ons could have worse ISP than Merlin1-D and they would add weight all on their own, meaning their mass penalty would be greater than  SpaceX's for the same recovery profile.

Also, using weak engines could allow hover, but wouldn't they waste more fuel on descent than the powerful Merlin1-D hoverslam?

I'm just wondering why we haven't seen any proposals of this sort from the rocket companies.

Online Lar

  • Fan boy at large
  • Global Moderator
  • Senior Member
  • *****
  • Posts: 9070
  • Saw Gemini live on TV
  • A large LEGO storage facility ... in Michigan
  • Liked: 5761
  • Likes Given: 3843
Re: Reusability effect on costs
« Reply #488 on: 05/22/2017 02:28 AM »
There seems to be a certain amount of "reusability hasn't been proven to work until Thing X happens" with a shift over time of what X is. While that's perhaps interesting to some folks, I'm not sure it's on topic here... not completely. If you can't tie back to costs, it's not.
"I think it would be great to be born on Earth and to die on Mars. Just hopefully not at the point of impact." -Elon Musk
"We're a little bit like the dog who caught the bus" - Musk after CRS-8 S1 successfully landed on ASDS OCISLY

Offline oldAtlas_Eguy

  • Senior Member
  • *****
  • Posts: 3243
  • Florida
  • Liked: 1628
  • Likes Given: 179
Re: Reusability effect on costs
« Reply #489 on: 05/23/2017 03:14 PM »
The following post on the Bulgariasat Update thread implies great reduction in cost of refurbishment.

http://forum.nasaspaceflight.com/index.php?topic=42913.msg1681755#msg1681755

This is that between these two that the refurbishment costs reduced by ~factor of 4. If the SES-10 refurbishment cost 50% that of a new booster, the Bulgariasat-1 cost ~12% that of a new booster. Even if a new booster cost $30M (this is actually the highest estimate for what a new booster cost to manufacture) then the refurbishment cost is <$4M. If the cost of a new booster is $20M (this is the low end for the estimated cost to manufacture) then the refurbishment cost would be <$3M. This puts the savings of at least $17M up to as much as $26M for this launch.

Even if the reduction factor is only ~3 then the savings become from $16M to $25M. Only a variation of $1M from the higher reduction factor.
« Last Edit: 05/23/2017 03:17 PM by oldAtlas_Eguy »

Re: Reusability effect on costs
« Reply #490 on: 05/23/2017 03:40 PM »
The following post on the Bulgariasat Update thread implies great reduction in cost of refurbishment.

http://forum.nasaspaceflight.com/index.php?topic=42913.msg1681755#msg1681755

This is that between these two that the refurbishment costs reduced by ~factor of 4. If the SES-10 refurbishment cost 50% that of a new booster, the Bulgariasat-1 cost ~12% that of a new booster. Even if a new booster cost $30M (this is actually the highest estimate for what a new booster cost to manufacture) then the refurbishment cost is <$4M. If the cost of a new booster is $20M (this is the low end for the estimated cost to manufacture) then the refurbishment cost would be <$3M. This puts the savings of at least $17M up to as much as $26M for this launch.

Even if the reduction factor is only ~3 then the savings become from $16M to $25M. Only a variation of $1M from the higher reduction factor.

I thought that SES-10 was supposed to fly in October before the Amos mishap? That would make it around five months from landing and relaunch, or 4 months to refurbish the core (according to Gwynne Shotwel). Since this launch would be roughly five months from first flight, would it be reasonable to assume that they're not yet at the point of rapid reusability, and that there hasn't been such a change in cost savings compared to the last reused core? Don't get me wrong, they've saved a lot of money already, but I don't think they're quite there yet.   

Online Lar

  • Fan boy at large
  • Global Moderator
  • Senior Member
  • *****
  • Posts: 9070
  • Saw Gemini live on TV
  • A large LEGO storage facility ... in Michigan
  • Liked: 5761
  • Likes Given: 3843
Re: Reusability effect on costs
« Reply #491 on: 05/23/2017 03:48 PM »
Don't confuse elapsed time with effort required.... that 4 months go by between flight and reflight doesn't imply that there was 4 months of continuous effort. (much less if it was one person or a standing army of 1000... )

When we see SpaceX say "we can't reuse on this flight, we don't have a booster ready" then we'd know that they don't have effort down low yet....
"I think it would be great to be born on Earth and to die on Mars. Just hopefully not at the point of impact." -Elon Musk
"We're a little bit like the dog who caught the bus" - Musk after CRS-8 S1 successfully landed on ASDS OCISLY

Re: Reusability effect on costs
« Reply #492 on: 05/23/2017 04:02 PM »
Don't confuse elapsed time with effort required.... that 4 months go by between flight and reflight doesn't imply that there was 4 months of continuous effort. (much less if it was one person or a standing army of 1000... )

When we see SpaceX say "we can't reuse on this flight, we don't have a booster ready" then we'd know that they don't have effort down low yet....

I guess I didn't think of it that way. Although, I'm assuming that similarly with airplanes, they technically "lose" (not really) money by having the core on a hangar and not helping to deliver a payload into orbit, which is what I was thinking about. So at this point, they could have greatly reduced how much it costs to refurbish a core because it takes less effort do so presumably, and now they're going to aim to maximise the revenue that they get from each core. This will done by reducing the turn around time by making refurbishment quick and cheap (or eliminating it altogether), and by ensuring the core has high durability.

Cue in block V...

Online envy887

  • Senior Member
  • *****
  • Posts: 2945
  • Liked: 1372
  • Likes Given: 915
Re: Reusability effect on costs
« Reply #493 on: 05/23/2017 04:11 PM »
Don't confuse elapsed time with effort required.... that 4 months go by between flight and reflight doesn't imply that there was 4 months of continuous effort. (much less if it was one person or a standing army of 1000... )

When we see SpaceX say "we can't reuse on this flight, we don't have a booster ready" then we'd know that they don't have effort down low yet....

I guess I didn't think of it that way. Although, I'm assuming that similarly with airplanes, they technically "lose" (not really) money by having the core on a hangar and not helping to deliver a payload into orbit, which is what I was thinking about. So at this point, they could have greatly reduced how much it costs to refurbish a core because it takes less effort do so presumably, and now they're going to aim to maximise the revenue that they get from each core. This will done by reducing the turn around time by making refurbishment quick and cheap (or eliminating it altogether), and by ensuring the core has high durability.

Cue in block V...

Airplanes are built and operated assuming they will spend more time flying and earning money than on the ground.

Rockets are (until now) built and operated assuming they will be used once and that's it. You can't use the same financial model for both and expect sensible results.

Re: Reusability effect on costs
« Reply #494 on: 05/23/2017 04:20 PM »
Don't confuse elapsed time with effort required.... that 4 months go by between flight and reflight doesn't imply that there was 4 months of continuous effort. (much less if it was one person or a standing army of 1000... )

When we see SpaceX say "we can't reuse on this flight, we don't have a booster ready" then we'd know that they don't have effort down low yet....

I guess I didn't think of it that way. Although, I'm assuming that similarly with airplanes, they technically "lose" (not really) money by having the core on a hangar and not helping to deliver a payload into orbit, which is what I was thinking about. So at this point, they could have greatly reduced how much it costs to refurbish a core because it takes less effort do so presumably, and now they're going to aim to maximise the revenue that they get from each core. This will done by reducing the turn around time by making refurbishment quick and cheap (or eliminating it altogether), and by ensuring the core has high durability.

Cue in block V...

Airplanes are built and operated assuming they will spend more time flying and earning money than on the ground.

Rockets are (until now) built and operated assuming they will be used once and that's it. You can't use the same financial model for both and expect sensible results.

I agree, there is much change needed before this model even comes close to being realistic for this market. Would one of the main drivers of rapid reusability then simply be that (assuming that labour costs are the main costs for SpaceX) you are spreading those costs over dozens flights instead of just a couple? You'll still be paying roughly the same (as in wages don't decrease), but since you're getting more revenue for the same costs, you're better off right?

Offline oldAtlas_Eguy

  • Senior Member
  • *****
  • Posts: 3243
  • Florida
  • Liked: 1628
  • Likes Given: 179
Re: Reusability effect on costs
« Reply #495 on: 05/23/2017 05:55 PM »
Launch cadence.

One of the biggest impacts to per launch costs is the crew at the launch site. If there is not a launch they do not have much work to do. Their time (from the standpoint of the company) is being wasted. Also for a flow not everyone is involved with every step of the process. So unless there is an assembly line setup to LV processing at the launch site, again wasted manpower. If SpaceX with 2 pads can launch a little less than every 2 weeks, they can share personnel between pads too as long as the launches are staggered (about one a week) between the two pads. So at a rate of 1 a week (once every 2 weeks per pad) SpaceX would be spending almost the same in labor as they would be to launch off of 1 pad every 2 weeks.

But this higher cadence and assembly line like launch processing is a result of rapid reusability. If spaceX can get there. There will be several costs savings in various other places than just in refurbishment costs. Generally the cost of launch (even for a new booster) will decrease because of higher cadence.

Now as far as my estimates on the relative costs of refurbish ment between SES-10 and Bugariasat-1 is that SpaceX stated the cost of refurbishment for the SES-10 booster was 50% that of a new booster. The next item is statements that refurbishment on later vehicles would be 1/8 the the cost of the SES-10. What I was showing is that it seems that Bulgarisat-1 shows that SpaceX has yet to reach that 1/8 the cost of SES-10. But as Lar states we do not know if they are simply counting time (a mixture of storage and active work) or the work of a certain size crew or crews for a period.

But the costs have gone down. The exact amount is not known but even just cutting the costs in half from that of SES-10 is significant ($5M or more in additional savings).


Re: Reusability effect on costs
« Reply #496 on: 05/23/2017 06:10 PM »
Launch cadence.

One of the biggest impacts to per launch costs is the crew at the launch site. If there is not a launch they do not have much work to do. Their time (from the standpoint of the company) is being wasted. Also for a flow not everyone is involved with every step of the process. So unless there is an assembly line setup to LV processing at the launch site, again wasted manpower. If SpaceX with 2 pads can launch a little less than every 2 weeks, they can share personnel between pads too as long as the launches are staggered (about one a week) between the two pads. So at a rate of 1 a week (once every 2 weeks per pad) SpaceX would be spending almost the same in labor as they would be to launch off of 1 pad every 2 weeks.

But this higher cadence and assembly line like launch processing is a result of rapid reusability. If spaceX can get there. There will be several costs savings in various other places than just in refurbishment costs. Generally the cost of launch (even for a new booster) will decrease because of higher cadence.

Now as far as my estimates on the relative costs of refurbish ment between SES-10 and Bugariasat-1 is that SpaceX stated the cost of refurbishment for the SES-10 booster was 50% that of a new booster. The next item is statements that refurbishment on later vehicles would be 1/8 the the cost of the SES-10. What I was showing is that it seems that Bulgarisat-1 shows that SpaceX has yet to reach that 1/8 the cost of SES-10. But as Lar states we do not know if they are simply counting time (a mixture of storage and active work) or the work of a certain size crew or crews for a period.

But the costs have gone down. The exact amount is not known but even just cutting the costs in half from that of SES-10 is significant ($5M or more in additional savings).

So they're getting a better bang for their buck by effectively increasing productivity. Would they move the work force responsible for manufacturing new first stages to other projects to keep costs low and productivity high? Like you said, it is in their best interests to keep the crew busy and not have them standing around and doing nothing.

(Slightly OT) 

Online Lar

  • Fan boy at large
  • Global Moderator
  • Senior Member
  • *****
  • Posts: 9070
  • Saw Gemini live on TV
  • A large LEGO storage facility ... in Michigan
  • Liked: 5761
  • Likes Given: 3843
Re: Reusability effect on costs
« Reply #497 on: 05/23/2017 06:17 PM »
Airplanes are built and operated assuming they will spend more time flying and earning money than on the ground.

Rockets are (until now) built and operated assuming they will be used once and that's it. You can't use the same financial model for both and expect sensible results.

Yep.

... not yet.... maybe not ever. BUT... Get the cadence to once every week across two pads and you're definitely moving in that direction. It is a baby step but it's moving from custom to scheduled/assembly line... And then (as long as the payloads are there[1]) you can keep driving process improvements. SpaceX is all about those. Contrast the time for the first ASDS unload-trip to hangar to the time for the latest.

So they're getting a better bang for their buck by effectively increasing productivity. Would they move the work force responsible for manufacturing new first stages to other projects to keep costs low and productivity high? Like you said, it is in their best interests to keep the crew busy and not have them standing around and doing nothing.

Obviously they would. First thing they'll be doing? Making more second stages than 1:1 with first stages. They already have made at least one more, presumably, and are about to have made 2 more (discounting test articles)


1 - CommsX makes sure they will be. Talk about eating your own dogfood...
"I think it would be great to be born on Earth and to die on Mars. Just hopefully not at the point of impact." -Elon Musk
"We're a little bit like the dog who caught the bus" - Musk after CRS-8 S1 successfully landed on ASDS OCISLY

Re: Reusability effect on costs
« Reply #498 on: 05/23/2017 06:26 PM »
Obviously they would. First thing they'll be doing? Making more second stages than 1:1 with first stages. They already have made at least one more, presumably, and are about to have made 2 more (discounting test articles)

Is it true that the Mvac is significantly more complex than its sea level counterpart? There's going to be people who are probably going to know how to manufacture and assemble most version, but do we have any idea on how easy or hard it is to train those people to build second stages? I know that they've made a couple of design choices specifically to make both as cheap as possible by using similar engines, same diameter, etc.

(Lar, I realise I'm getting slightly off topic, so move this comment to an appropriate thread if I'm too far off.)   

Edit/Lar: fix quotes. And fine so far I think...
« Last Edit: 05/25/2017 03:44 AM by Lar »

Online macpacheco

  • Full Member
  • ****
  • Posts: 828
  • Vitoria-ES-Brazil
  • Liked: 332
  • Likes Given: 2498
Re: Reusability effect on costs
« Reply #499 on: 05/23/2017 06:39 PM »
From now to the 3rd Block IV refurb it should drop to about 2 weeks refurb. This is likely to happen still in 2017.
Not trying to start a discussion about this opinion. Just staking my claim and we'll see in Dec/17 or Jan/18 what happens, when I think at least the 6th total refurb and 3rd Block IV refurb will take place.
Reuse is already economical for SpaceX, even with 2 months refurb.
Bringing it down to 2 weeks will make it very economical.

With 3 refurbs, the workers taking care of it will likely have a lot of practice. From then on its eliminating replacement of parts that were being replaced just in case and that with more data can be ruled out as safe to reuse. Plus Block IV will make more parts robust enough to aim for at least a few reflights without replacement. Block IV might still not be ready for 10 reflights with modest refurb. Even if it can only do 3 or 4 with a 1-2 week refurb will be already ultra economical even compared with a 2 month refurb.
Looking for companies doing great things for much more than money

Tags: