Author Topic: Reuse business case  (Read 136147 times)

Online LouScheffer

  • Full Member
  • ****
  • Posts: 1908
  • Liked: 2498
  • Likes Given: 279
Re: Reuse business case
« Reply #540 on: 06/29/2017 01:28 PM »
From the very first post in this thread:
The figure of merit I chose to evaluate is pure $/kg of the reusable system compared to $/kg of the same system used in expendable mode.  [...]

An example of where this pure case applies directly is the launch of a commodity like propellant to fill an orbiting stage or depot.  In that case every launcher would load as much as it could carry and $/kg is all that matters.
This, I think, is the fundamental flaw.  The customers that exist today are not optimizing $/kg -they are searching for a minimum-cost solution to launching their particular payload, whose mass is already determined.   

IF you are filling a propellent depot, the application George points out, then re-usablility is much tougher to justify (70% of the cost, for 70% of the capacity, and re-use never pays off).  But if you want to launch a single payload, and it fits within that 70%, then re-use pays off immediately.

I think a more realistic (but more complex) spreadsheet would start with an assumed distribution of demand (so many 4000kg launches, 4500 kg launches, 5000 kg launches, etc) and look at the cost per payload.  You could compare re-usable, dial-a-rocket, paired launches, and other approaches.

Offline gospacex

  • Senior Member
  • *****
  • Posts: 3028
  • Liked: 537
  • Likes Given: 604
Re: Reuse business case
« Reply #541 on: 06/29/2017 01:33 PM »
Sowers suggests that this might be 0.4 for SpaceX, leading to a need for a 10-reflight breakeven.  But SpaceX says that this variable is more like 0.7, which leads to a 3-reflight breakeven.

SpaceX has said that the first stage represents about 70% of the cost of the rocket, not 70% of the launch service price.

It was addressed before. All fixed components of launch are irrelevant when you compare reused core and expendable core, since they do not change. That is an another mistake in the comparison.

The correct question is, "after how many reflights reuse of a core breaks even compared to expending a new core each time", not "compared to *the cost of entire launch*". Therefore other costs (example: cost of the second stage) are irrelevant.

Offline RedLineTrain

  • Full Member
  • ****
  • Posts: 740
  • Liked: 452
  • Likes Given: 475
Re: Reuse business case
« Reply #542 on: 06/29/2017 01:48 PM »
Sowers suggests that this might be 0.4 for SpaceX, leading to a need for a 10-reflight breakeven.  But SpaceX says that this variable is more like 0.7, which leads to a 3-reflight breakeven.

SpaceX has said that the first stage represents about 70% of the cost of the rocket, not 70% of the launch service price.  Not positive how that difference plays into Dr. Sowers' model, but I don't think you can just substitute 0.7 for the k value.

Price is irrelevant to the model.  k is defined as "the fraction of recovered HW production cost to total launch service cost."  The total launch service cost would be the cost of producing the rocket plus operations to get it launched.

Given that ULA's customer is the government, the operations costs are a lot higher.  But for SpaceX, many/most of its customers are commercial.  So I stand by the k = 0.7 for SpaceX.

Offline Paul451

  • Full Member
  • ****
  • Posts: 1489
  • Australia
  • Liked: 749
  • Likes Given: 614
Re: Reuse business case
« Reply #543 on: 06/29/2017 01:52 PM »
I looked at that as a marketing initiative from ULA, not a true engineering analysis. It will be interesting to see if Dr. Sowers talks about that again now that he no longer works for ULA.

Given his distinct lack of interest in improving his model, I'd say your first assumption is correct.

Per meekGee's original post, the payload penalty assumptions are likewise perhaps high (it's not clear that they were figuring in sea landings).
What "payload penalties"?

The model treats revenue as if each launcher has a fixed, linear $/kg income rate. Hence if reusability costs you 30% of your payload, the model assumes you lose 30% of your revenue. (If you, for example, assume the original price had a 100% mark-up on unit-costs, then even saving 60% on the unit-costs through reusability is completely eaten by that phony reduced revenue. And that's before you add any increase in fixed costs for developing reusability.)

While $/kg is a quick'n'dirty way to compare across launch providers, this model shows the limits of doing so, especially for comparing launchers with variants of themselves.

If the top 30% of your launcher's capacity is only required for 5% of your potential market, but revenue-per-payload is roughly the same, then you don't have to save much from reusability before you are better off. (Even if those top 5% of payloads typical pay double the usual market rate, that's still less than 10% of your revenue, not 30%.)

It's not that the model is "simple", as its creator claims, it's that the model is completely misleading on that measure alone.

But in addition...

IF you are filling a propellent depot, the application George points out, then re-usablility is much tougher to justify

The model effectively assumes a fixed number of launches. It assumes that you aren't going to increase your market share via lower prices, to make up for giving up that potential top 5% of the market. Nor will you lose market share if you don't add reusability. That adds another artificial distortion.

It makes no sense to assume fungible payloads ($/kg) and a fixed number of launches. If you are really are modelling a future fuel-depot market, that market will be measured in number of tonnes/yr, not number of launches/yr. Reduce the capacity of each launch and you need to increase the number of launches, reducing the share of fixed costs in each launch as well as the unit-cost via reusability.

Offline TrevorMonty

Re: Reuse business case
« Reply #544 on: 06/29/2017 02:41 PM »
Reuse does rely on high launch rate to recover R&D in reasonable time. Current launch market is finite with only so many missions avaliable to compete for. With lot of customers deliberately spreading their missions between current launch providers to spread their risk and to stop a monopoly.

What SpaceX and Blue especially are banking on is a new markets opening up due to lower price. The likes of space tourism as example could dramatically increase demand which goes up as launch price comes down. The existing  launch providers can't justify $Bs on R&D for non exist market that may or may not appear. These old space companies have been burnt before by large satellite markets that didn't appear.
SpaceX also to certain extent with F1, why can it once flying if there was large demand.

Where reuse does shine is in short lead time between booking and flying. Possible greater reliability of reused stage, yet to be proved. Being able inspect a flown stage should allow for  engineering of more reliable LV.
Both of these features are hard to put a price on but may well be more important than lower price.
« Last Edit: 06/29/2017 02:42 PM by TrevorMonty »

Online saliva_sweet

  • Full Member
  • ****
  • Posts: 560
  • Liked: 431
  • Likes Given: 1379
Re: Reuse business case
« Reply #545 on: 06/29/2017 03:32 PM »
It's been obvious for quite some time that SpaceX has been aiming toward refurb costs way lower than the ULA spreadsheet was assuming.

The spreadsheet isn't assuming anything. You can plug in any value you like. That's why I like it.

Offline envy887

  • Senior Member
  • *****
  • Posts: 4707
  • Liked: 2616
  • Likes Given: 1407
Re: Reuse business case
« Reply #546 on: 06/29/2017 03:46 PM »
It's been obvious for quite some time that SpaceX has been aiming toward refurb costs way lower than the ULA spreadsheet was assuming.

The spreadsheet isn't assuming anything. You can plug in any value you like. That's why I like it.

Many things are not modeled and therefore assumed constant... but improving the model makes it a lot more complex an increases the already high uncertainty.

Online saliva_sweet

  • Full Member
  • ****
  • Posts: 560
  • Liked: 431
  • Likes Given: 1379
Re: Reuse business case
« Reply #547 on: 06/29/2017 03:55 PM »
Has anyone reexamined the original spreadsheet in view of what SpaceX has been observed to be doing?

I've been doing that more or less since Sowers posted it (and getting a lot of flak for it). Everything Elon Musk, ULA and Arianespace have been saying is entirely concordant with this spreadsheet. You can use it to see what conditions need to be met for reuse to truly make sense over the best theoretical expendable. It requires high launch rates and low overhead costs. These are the conditions Spacex needs to force. If they succeed, the savings can heavily snowball and reusable system will easily beat the best expendable. The spreadsheet shows and quantifies that. It also shows that SMART and Adeline (or whatever the ariane equivalent was) are dead ends and will never fly.

People often mistake Sowers' work for some sort of current market analysis. It is not.

Offline GWH

Re: Reuse business case
« Reply #548 on: 06/29/2017 10:32 PM »
Has anyone reexamined the original spreadsheet in view of what SpaceX has been observed to be doing?

I've been doing that more or less since Sowers posted it (and getting a lot of flak for it). Everything Elon Musk, ULA and Arianespace have been saying is entirely concordant with this spreadsheet. You can use it to see what conditions need to be met for reuse to truly make sense over the best theoretical expendable. It requires high launch rates and low overhead costs. These are the conditions Spacex needs to force. If they succeed, the savings can heavily snowball and reusable system will easily beat the best expendable. The spreadsheet shows and quantifies that. It also shows that SMART and Adeline (or whatever the ariane equivalent was) are dead ends and will never fly.

People often mistake Sowers' work for some sort of current market analysis. It is not.

Yeah its a pretty weak spreadsheet since it doesn't breakdown fixed costs vs. launch rate, allow for FH and corresponding uptick in core production, there is no production sharing for common hardware like F9 booster & US architecture, refurbishment is assumed on a per flight basis and not intermittent etc.

Offline oldAtlas_Eguy

  • Senior Member
  • *****
  • Posts: 3494
  • Florida
  • Liked: 1897
  • Likes Given: 226
Re: Reuse business case
« Reply #549 on: 06/30/2017 01:21 AM »
In the SpaceX case there is a few complicating events. One was that SpaceX increased their expendable payload capability of the F9 from the original v1.0 to the Block 5 by 145% while only increase the price by 15%. So the effect is that even by loosing 30% of that payload capability increase means that the reusable version has increased payload capability by 71% over that of the original while reducing the price by 7.5%. So the actual occurrence was that payload increased and price decreased.

Another item is that the 2 reuse flights this year represents 5% of all orbital launches this year. A BTW SpaceX has launched 22% of all orbital launches in the world so far for the year. If they do 70% of their launches reuse launches for 2018 that would represent 15% of all launches world wide.

Tags: