Second in a series of a detailed examination of the PR that is produced by United Launch Alliance.
Launch Vehicle Recover and Reuse Paper
www.ulalaunch.com/docs/defaul...
Business case and spreadsheet post
forum.nasaspaceflight.com/ind...
@Eager_Space on Twitter
Triabolical_ on Reddit
/ eagernetwork
/ eager-space-1038430522...
Any day with a spreadsheet is a good day!
Nice one, I'm enjoying these. Facts & logic.
New favorite space channel. Well thought out, explained, and presented ideas that dont smell of bias. Examples given are understandable for both aerospace nerds and the general public. Keep up the good work!
Thanks. I'm trying to keep my opinion tamped down but it's hard to do on these videos.
@@EagerSpace I get it, ULA isn't doing themselves any favors here, it's pretty cringe worthy. The saddest part is that you have to imagine there's senators and government folk who eat this stuff up.
Good video. Cost per kg really only makes sense as a measure where your payload is approaching the limits of the rocket. Most flights are not at the rocket maximum payload so this has no barring on the actual costs in the real world. Elon uses the term rapidly reusable, this hints at ULAs real problem, it’s that their cost to refurbish is far too high. SpaceX have been laser focused on reducing costs which gives them a very profitable reuse model in comparison. I think what the graph from ULA shows is that they cannot engineer a reusable vehicle that makes economical sense.
Do we have any information on whether SpaceX uses investor funds to subsidize the Falcon 9? SpaceX appears to have substantial capital, largely due to Elon Musk's influence, so it's possible they could use this capital to subsidize their own product in order to push competitors out of the market.
When Falcon 9 was developed, NASA put in about half of the money and SpaceX put in the rest, and that was outside investor money. I don't think they were taking in significant investor money during the development of reuse. They've certainly taken in money to pay for starlink and starship. I don't think there's any evidence that SpaceX undercut their prices to try to push competitors out of the market. There really wasn't any need to - Proton had crappy reliability and you had to deal with the Russians to launch it, Ariane 5 was quite a bit pricier and they were already launching as fast as they could, and ULA had no real commercial business. The prices that ULA was charging for DoD payloads were very high (Atlas V and Delta IV) and astronomical (Delta IV Heavy) - even the cheap SpaceX prices were much, much cheaper. They have since bumped up their prices and ULA has reduced their prices so it's more even now.
@@xponen As far as anyone outside SpaceX knows, a Falcon 9 with booster landing has an internal cost of something like $20 million. So for each launch they sell at something like $55 million, the company adds $30 million in gross profit. That sounds great, but it's still not enough. They are only selling something like 30 launches per year (with the remaining launches being for Starlink), which is a gross profit of less than a billion dollars. Maybe roughly a billion dollars when accounting for Dragon and Falcon Heavy. When they are spending something like $2 billion per year on Starship, they still need more money, and that is where investor funds come into the picture.
@@xponen SpaceX was "just" the first in the industry to invent a *virtuous* self-licking ice cream cone, creating an internal customer that drove launch rate high enough for re-use to make sense.
NASA funds 85% of starship
In the spectrum of lies, damned lies, and statistics, this definitely falls under statistics. Not a blatant lie like in part one, but definitely picking certain numbers to arrive at a desired conclusion. As explained in the video, Falcon 9 likely does better than the graph's estimate. As for SMART reuse, I do genuinely hope it can reach that level of cost savings, but it still remains to be seen.
I’ve been watching all of your videos, pausing to view new uploads when they come out. I feel like I understand everything about the space industry thanks to you.
Your videos are always great, thank you!
Thanks.
One benefit of reuse that’s very specific for SpaceX is that it enables a tremendous launch cadence for Starlink. Their ability to quickly launch constellations with a large number of satellites makes the network become operational and therefore potentially profitable much faster. I don’t think they could sustain a similar launch cadence if they had to produce a new booster for every launch.
It probably would be possible, but it would require a much larger factory. At the moment, they are building close to three upper stages per week. Doing the same thing for boosters is concievable. But once they are scaled for 100+ expendable launches per year, going back down to a smaller number would mean that the factory is way too big, and they would have far too many employees. It can cause some operational challenges. Reuse is more flexible. It allows for scaling up/down operations without massive hiring sprees and layoffs. And additionally, reuse allows for a similar sort of "dial-a-rocket" flexibility as ULA has with their solid rocket boosters. If you need full performance, you expend the vehicle, while if you don't need a lot of performance, you do return to launch site.
ULA has been very careful to not make that comparison, because the SpaceX approach doesn't require you to be buying expensive solid rocket motors from another company.
They could, but reuse has simplified their logistics considerably.
if reuse with falcon 9 wasn't making Space X bank they would have quietly stopped doing it, that they haven't tells you it's cheap..
SpaceX has done 31 or 32 rounds of fund raising so far. That's how they cover operational costs, not through revenue. If reuse were secretly more expensive than they were letting on abandoning it could really cripple their primary source of operational funds.
@@opcn18 Space X deals with institutional investors, you can't buy anything, so unless you are claiming these people are stupid it's making bank, as detailed at a few million to refurb and they even skip the repaint they are going to be doing well.
@@opcn18 Fundraising is how they cover the cost of new programs, until the new programs start making money, where the two new major programs in recent years have been Starlink and Starship. By all accounts, the launch business is providing substantial (but insufficient) free cash flow for the new programs. And the same also seems to be the case for Starlink, at this point. The current money drain is Starship, which will continue for at least a while. I'd guess that we'll be closer to 2030 before Starship is fully self-sustaining. Not including any costs related to internal Mars or Moon missions.
@@TheShorterboy Tesla's rise to become the world's most valuable car company is mainly attributed to the impact of Elon Musk. Despite selling fewer cars than its competitors, Tesla's value soared to 1 trillion USD at one point, showcasing investor trust in Musk's vision and leadership.
I've always viewed that as the "existence proof" for reuse. SpaceX figured out how to do it and what they needed, expended all the block 4 boosters that may not have been economical, and then started flying the hell out of the block 5 ones. Either they're really stupid or reuse is a lot cheaper.
19:12 that's a nice print
Nobody has commented on what they print is yet...
This was great, well done. Given all the satellite constellation, resupply, tourism, and refueling systems in planning (plus ridesharing), I would imagine the market being dominated by payloads at or close to launcher capacity going forward. Are you going to look at the economics of ULA's SMART reuse or whatever they are proposing these days?
I've thought about looking at SMART. The hard part is that ULA is quite opaque compared to SpaceX and SMART is so much more complicated. I don't know what the reentry hardware will cost and they need to pull it out of the water, ship it back to the factory, somehow revalidate the engines (how much will Blue charge for that?), and then reintegrate it back into their first stage flow.
@@EagerSpace Maybe still worthwhile to bound it? They did provide that nice breakdown of their rocket cost. I was surprised that the structure cost was actually significant. Their reuse plan might not look that good even under "best case" assumptions. I read about the splashdown idea, but what happened to the mid-air catch?
@@marksinclair701 if mid-air catch was viable, rocket lab would still be trying it - they couldn't make it work even with a much smaller rocket, and these things get much harder the bigger the rocket is.
@@andytroo Didn't they do this with film cannisters returned from old spy satellites? I know they're smaller but the rendevous problem is the same. Surely with modern heavy-lift-drones/tracking/etc. it could be accomplished today? I don't think RocketLab had the resources to make it work, not that it can't be done.
@@EagerSpace I thought another thought on this. Can you just come at it from the reverse direction? So determine target recovery and refurbishment costs for a given level of reuseability payback - defined as threshold number of flights.
Love the video. Great analysis. Besides ULA obviously slanting the analysis, I think it also shows a difference in the thinking between ULA and SpaceX. It seems to me that ULA thinking was the market for launches would stay about the same: roughly 12 launches a year. Based on needing 10 flights to break even and multiple boosters to handle all the launches (I'm sure they assumed longer than a month to refurbish), it would take a long time to a break even on operational costs, to say nothing of R&D costs. SpaceX bet (correctly) that if they lowered cost, the market would get bigger, thus justifying the R&D to create a reusable booster. Some minor nitpicks on the presentation: - At 14:16, you say Atlas IV while showing an Atlas V. - The graph that starts at 14:46 of all the ULA launches looks a little odd with dots not lining up with the names of the rockets on the x-axis. I can see you needed the space for when you compare it to the Falcon9, but it definitely felt wrong at first. - Two dots for double digits is strange looking. Maybe a rounded rectangle that's big enough to hold two digits would work better?
I'm not actually sure that SpaceX was in it for the cost savings. The widely reported number for developing reuse is $1 billion, but my guess is that that includes not only all the reuse hardware development but the upgrade and refinement to get to block 5 plus the recovery assets. If that number is right, if they save $10 million per launch that's 100 launches to break even, which is a ton when you are looking at their flight rate in 2015. I suspect that Starlink may have been an idea back then...
@@EagerSpace Also, I assume reuse was important to Musk based on his stated goal of colonizing Mars. You can't get to $13/kg without a lot of reuse of the expensive parts of Starship. Presumably Musk saw Falcon 9 booster research as valuable for learning how to reuse a rocket while at least part of the bill was pad for by customers.
The Market did not get bigger because of SpaceX's launch costs though, it got bigger because of SpaceX self-consumption. That is a very different proposition from the 'build it and they will come' which people have imagined would happen and which has been the downfall of all high volume/low cost launch concept.
@@kennethferland5579 Looking at the numbers on Wikipedia (assuming it can be trusted), it looks like both things are happening: SpaceX has increased the size of the US market and a lot of the increase in flights is also through self-consumption. I looked at the number of ULA and SpaceX flights from 2010 to 2018. I picked this range because it's before Starlink launches and while Vulcan was announced in this time period, ULA was still selling Atlas V and Delta IV. ULA goes from 8 launches up to 14 and then back down to 8. SpaceX goes from 2 to 21 launches. So it looks like the US market got bigger, at least enough to justify two companies doing launches. Eyeballing the graphs of R-7, Proton and Ariane launches, they seem relatively stable from 2010-2018. SpaceX was probably stealing some flights from other rockets, but I think they were also increasing the total market. The crazy growth in Falcon 9 flights in the past few years is definitely Starlink. Their non-Starlink flight have grown a bit, but the record breaking number of launches is all due to Starlink. Since SpaceX is a private company, it's unclear if their costs and prices were sustainable without Starlink. There's a TechCrunch article that claims to have SpaceX financial data from 2018 and 2019. They lost significant money in both those years, but they also spent more than R&D than their net loss. So maybe if they weren't owned by a billionaire who wants to live on Mars, they could have slowed down the R&D work and gotten profitable on Falcon 9.
17:00 Well, here's the whole crux of the discussion, which is quite convincing. But as to the PR/Lie question, I have to toss in the observation that when ULA was writing up the 2015 paper (doing work in 2014?) the F9 performance figures were pretty low. You refer to this by saying you used figures for the 2017 F9 and then subtracted a bit of the performance. But was that enough to make it a fair comparison to what ULA was looking at in 2014-15? So closer to PR than lying?
The real crux of the discussion is that $/kg is a stupid way to evaluate the competitiveness of a launcher and ULA didn't try to make a real engineering argument around their point. Falcon 9 V1.1 first flew in 2013 and had flown 14 times by my count when the paper was published. Sowers shared the excel file on the forum in April of 2015, considerably before the paper was published. If ULA had made a market coverage argument with V1.1, that would have been interesting as the coverage wouldn't have been great, though given that SpaceX had already upsized their launcher a bunch in V1.1 you would have to wonder if they weren't planning more. I could see ULA missing this as they don't do engine development. I don't think this one is lies as much as misleading and poorly done, and it appears from Sowers arguments that they simply didn't think of doing a market-based analysis.
I wonder if ULA could also pursue full booster reuse (with propulsive landing), and what the minimum amount or reengineering required would be. Could a vulcan booster survive reentry and aerobrake? Maybe an engine skirt, like what the original atlas missile had that launched mercury, could accomodate a small engine required for a hover-slam landing.
They have the wrong architecture - too few engines to do the landing and they stage very late so the first stage probably can't survive reentry and the payload cost will be higher.
@@EagerSpace late staging also makes RTLS much harder, and that is where the true wins are - drone ships are expensive, and reuse cost is critical for ROI ...
Very interesting analysis. It's obviously not really relevant today since it's almost a decade old, but it nicely illustrates how the booster flyback concept was seen at the time. Some of these inaccuracies might be more obvious with the hindsight of almost ten years of operation and 300 F9 booster landings, but it's also possible that a slightly better analysis at the time lead ULA to pursue a similar reuse concept to F9. Instead it looks like they were so convinced of the SMART reuse that they were willing to take on the competition with F9 and used these inaccurate studies as PR material for it.
The reason ULA did this was because they can't do reuse the way SpaceX does. You need a cluster of engines - at least 7 - to do that or the thrust is too high to land. ULA buys their engines and that many engines would be costly if there was an engine that made sense to use, and there isn't one other than the Merlin that SpaceX makes. They also have a weak upper stage which means they stage at a higher altitude and velocity, too high to be able to get back in the atmosphere at a reasonable speed. That's why they came up with the SMART approach to get the engines back. They haven't done SMART yet because it doesn't make business sense - the money they would save is not worth the development cost. They've been talking about it more because they have a big contract to launch Kuiper satellites for Amazon and in that world it may make sense, but Kuiper looks like it's stalled right now and ULA's biggest focus is getting Vulcan flying so that they can start getting paid for the space force launches that are their bread and butter.
If I may expand on Eager Space's answer: These are the biggest reasons ULA couldn't go for full reuse - they'd have to design a rocket from scratch that was optimized for reuse and get an engine manufacturer to build an engine optimized for full reuse. There's no way they or any stockholder-owned company would do that unless the DoD paid them to. F9 works economically for SpaceX because they sacrificed upper stage isp in order to use the same fabrication techniques and engine for both stages. F9's upper stage is quite oversized compared to other rockets - exactly because F9 has to stage lower. It'll be interesting to see how Neutron's reuse works out. It uses a small upper stage and is planned to stage a lot higher for this reason. Their Electron has made high speed reentries without doing a reentry burn - higher speeds than F9 does. This has apparently given them the data to calculate the highest altitude they can stage Neutron.
@@donjones4719 ULA wouldn't have to switch from the BE-4. It is intended to be reusable for New Glenn. But it would require a different vehicle than Vulcan Centaur.
@@SpaceAdvocate As Eager Space points out in his reply in this string, the BE-4 is too large an engine to land with for Vulcan or anything close to a Vulcan-sized rocket. It has too much thrust to be throttled down for landing. Only the much larger New Glenn can use it this way. So yes, ULA would have to have planned a different vehicle than Vulcan Centaur - a much different vehicle. ULA would realistically only have considered a Vulcan-sized or F9-sized rocket to succeed Atlas V, that is the sweet spot for the great majority of lucrative launches. For that they'd need a set of smaller engines.
@@donjones4719 They wouldn't have to switch from BE-4, but they would need a landing engine of some sort. In theory it could be something relatively low complexity like a hybrid engine using the same LOX supply as the BE-4s. Maybe with an electric LOX pump for fine throttle capability.
You're forgetting a huge part of the costs per launch, are related to operate and maintain the launching infrastructure including Staff. If reusing the F9 boosters were so economically advantageous, SpaceX prices to LEO should be drastically lowered by now.
Costs go down, but that doesn't necessarily mean lower prices. Prices are set according to the pricing of the competitors. The SpaceX optimal pricing scheme is something like $1 under the next cheapest competitor, all else being equal. (Realistically, all else is not equal. For one thing, SpaceX has unparalleled reliability. So they can actually charge higher prices than the competition.) SpaceX is likely making something like $30 million in gross profit for every commercial launch. That's about a billion per year for ~30 commercial launches, and that covers something like half the annual cost of the Starship development program.
@@SpaceAdvocate Your source for those profit margins per launch? If Falcon 9 was so profitable to operate, SpaceX wouldn't be in a rush to certify Starship, and would opt for developing a bigger fairing to Falcon Heavy. As other companies would be in a rush to reuse their boosters too.
@@RogerM88 Since SpaceX isn't a public company, we can only make reasonable estimates for profit margins. But there have been some leaks and announcements about profitability and costs to base the estimates on. And no, a bigger fairing for Falcon Heavy has a limited market. How many additional launches per year would this option give SpaceX? One? Two? SpaceX has bigger fish to fry. And other companies are in a rush to reuse their boosters. New Glenn, Electron, Neutron, Tianlong-3, Long March 9, Terran R, Ariane Next, etc.
My goal was not to create a realistic model of Falcon 9 costs, merely to explore what we can learn about those costs from simple models. A real model would be a thesis-level project and it will be very challenging to tease apart which spacex costs to assign to each program. A bedrock principle of commercial businesses is "don't leave money on the table", and this is especially true in a business where it costs so much to do research and development. SpaceX initially priced their products at a price that they thought was appropriate for a new entrant with unproven capabilities, and that mostly served them well though they did not make much money on the first CRS contract. Since they've become well-known, they have increased the government prices so that they don't leave money on the table. You can see this in the second CRS contract and you can especially see it on the current NSSL contract, where with their increased price and ULA's reduction the two companies are roughly comparable. That's just a rational decision there - there was no way the contract was going to go to anybody other than ULA and SpaceX and if you have a duopoly you want to charge what the market can bear. NSSL 3 is taking a slightly different model and it looks like there's a move to push some of the launches to the competitive per launch model. There's simply no reason to lower prices without anybody trying to take business away from you.
Starship is driven by Mars. Yes, if they get full reusability it will likely be cheaper in cost per kg - which they care about for Starlink to move up in satellite mass - but that's not why they are investing so much money there. With the exception of NSSL launch - where DoD is paying for an outside company to build a larger fairing - there is precisely one possible reason for a bigger fairing on FH, and that is starlink. To do that, they'd need to do the bigger fairing, design and build a new payload adapter and strengthen the second stage to take the added force. They'd presumably fly RTLS with the boosters and drone ship with the center core, and the payload there is probably in the 30 ton range. It's not clear whether FH is cheaper in that scenario, but SpaceX has clearly chosen not to spend the money and engineering time on a scenario without commercial application and instead has chosen to fly the hell out of Falcon 9. As for other companies, this reply is already too long, but ULA can't do booster reuse because their architecture is wrong and they would need a much higher flight rate to make the investment worthwhile. There are good reasons why other companies aren't there.
Even with cooking the books like ULA does to make them look competitive with SpaceX, there is just no reasonable argument that makes ULA competitive from a cost perspective. The only differentiation that ULA has is their launchers reach orbits that a Falcon 9 cant. But now that the market is developing kickstages like what Impulse is doing to get payloads to GEO and beyond from a starship, ULA’s competitiveness just vanishes. So however they present their numbers, they are dinosaurs and I give them less than 10 years until they scrap launchers altogether and pivot to space services.
I feel like this comment is ignoring the existence of Falcon Heavy? There's nothing ULA could do with Delta IV Heavy or can do with Vulcan that Falcon Heavy cannot.
Falcon heavy left the chat
@@lazarus2691 except payload size. That is falcon heavy's main weakness.
@@Smiles10130 Yarp, totally forgot about that. OP was specifically talking about orbits so I was only thinking mass. With that said, the extended fairing is coming soon(TM), which should close out the last of that capability gap.
Until present date, ULA disposed in one Vulcan Centaur launch 2 BE-4 engines... meanwhile SpaceX with 3 Starship launches, dispose hundreds of engines. By that metric, ULA still has plenty of engines to use with commercial launches, before achieving those numbers.
So what? Since when does the number of engines expended matter? If you're going to pick a metric, something like dollar value of hardware expended is a better metric. For now, a fully expended Vulcan is probably roughly equal to a fully expended Starship, in hardware dollar value. I'd guess the first flight of Vulcan expended something like $14 million in boosters, a core stage worth $25 million ($16 million in BE-4s), an upper stage worth $40 million ($30 million in RL10s) and a fairing worth $15 million, or $94 million in total. I'd guess the Starships are clocking in at around $100 million ($40 million in Raptors).
@@SpaceAdvocate Vulcan Centaur is already in operation collecting flight data. Starship is still far from being fully operational, and already disposed of hundreds of engines, the most expensive parts of the rocket. Also each Starship launch, was rumoured to be around 2bil. And that's for an empty prototype vessel. So I'd go easy on the hype around Starship.
@@RogerM88 No, there are no rumors that a Starship launch costs $2 billion. That was conjecture from some critics, where it is known SpaceX is spending around $2 billion per year on Starship, and at one point they had only done one launch. So they divided $2 billion by one. You couldn't get a worse analysis. Even in this simplistic model, the ceiling for cost would be around $400 million. They are doing around 5 launches this year at a cost of around $2 billion in total. But we know the actual cost is way less, as SpaceX is spending a lot of money on Starship infrastructure.
@@SpaceAdvocate Even if you take 400mil per launch, Falcon Heavy with a bigger fairing is almost as capable rocket, in cargo figures to LEO. And for the HLS contract, if you need at least 12 refiling missions (12x400mil), plus no less than 2bill for the Lunar Starship Crew rated, all of the sudden, SLS costs look more reasonable.
@@RogerM88 That's assuming no reuse. Plus a very expensive HLS Starship. I think SpaceX will be reusing at least the Super Heavy by the time they start doing launches for HLS, dropping the launch cost to probably in the area of $50 million. And $500 million for the HLS Starship seems more reasonable than $2 billion. And that is also reuseable, so it could do multiple landings, dropping the cost. I think it should be fairly easy to get below $1 billion cost per moon landing for HLS Starship. That would be half the cost of SLS Orion, for completing a much more difficult task.
I'm not sure how you know that the model from 7:35 is the same as the one from the white paper ULA created. The author of the post doesn't share a name with any of the paper authors and the post you showed doesn't seem to suggest he had a role in the paper either, rather being a regular user.
George Sowers was Vice President and Chief Scientist of ULA. It's pretty easy to look this up. At the very least these papers are likely related.
@@noname117spore Got it.
I probably could have made that clearer. The paper does say that Sowers came up with the reuse ratio and I'm assuming that the model is his since is the one who shared it, but he could be passing on the work that the authors did. T
Is it just me or no word on launch capacity & number of flights overall... if SpaceX can launch much more frequently because of reusability- than the overall profit margin should (in theory at least) be higher. and if you have more profit that means you can also lower the launch cost per flight. or am I missing something? It's kinda like ordering 1 custom product vs 100- obviousley the cost per unit should be lower the more you order.
ULA had a term for that in their model, but it was a negative one because they assumed you would build fewer things with reuse and therefore those things would cost more. It wasn't a major cost factor IIRC. I toyed with looking at launch capacity and number of flights but it's really hard to model and I don't know where I would get fair numbers to plug into the model. So, yes, you could do a better model that would consider launch rate and fixed costs into the model, but I think it would have gotten in the way of the main message and made things less understandable. I prefer models that are *obviously* overly simplistic because you can still draw useful conclusions and people don't start quoting your numbers.
Still going with lies
What is the true mass to LEO of F9 or FH expendable? I learned recently that my beloved SpaceX may have issued figures that are at best PR. In an online forum exchange someone who's expertise I trust told me the SpaceX user's guide states the actual max payload is limited by what the structure of the second stage can bear, which is 18t. Thus the F9 22.8t and FH 63.8t figures are merely about how much thrust the engines can produce, not the actual mass that can be launched. The Jupiter 3 is the heaviest payload FH has launched and it's only ~half the reported 18t limit . AFAIK the most massive F9 flights were some Starlinks which were purportedly near the 18t limit. Idk how to dive into the payload guide and would probably misread it if I did. Has SpaceX been posting a big PR lie for years with the 63.8t figure?
The figure is not a lie. But it does require a custom payload adapter. But there just isn't a market for 63 tons to LEO on Falcon Heavy. The payload fairing doesn't have the space for any rational payload with that mass. It would have to be something like a large space station module, but that definitely wouldn't fit into the payload fairing. Some sort of fully fueled propellant depot or space tug might fit, but no one has wanted to launch one yet. If there was a market for compact 63 ton payloads to LEO, SpaceX would have been all over that.
@@SpaceAdvocate So it's the payload adapter that can handle only 18t, the upper stage itself can? That's very different. Yeah, 63.8t to LEO isn't very useful - except as a simple metric indicating how powerful the rocket is overall, i.e. that it can send a *lot* of mass to GTO or GEO or TLI. Listing all those figures only works for space fans and the 63.8t figure is meant for slightly more general consumption, I presume. (Btw, the mass is almost certainly more than 63.8t now, it hasn't been updated for years and more performance has been tweaked out of F9 in that time.)
@@donjones4719 Yes, it's the payload adapter that's the current limitation. Payload to LEO is the primary yardstick by which the performance of rockets is measured. If Falcon Heavy only had figures for GTO, you wouldn't even be able to tell if it's a heavy lift launch vehicle, because that's defined to LEO. It's really the same for SLS, as an example. SLS Block 1 is specified to do 95 tons to LEO, but if you actually showed up with a 95 ton space station module and wanted to put it on the rocket, it would require a lot of work. It doesn't even have a fairing.
63.8 tons on FH or 22.8 on F9 would require a different payload adapter and the FH payload would likely require a modified second stage. But it's absolutely not a lie, if you want to launch that much payload to LEO, you can go to SpaceX and they will tell you how much it's going to cost to do that. It's going to be a expensive because it's custom engineering and it isn't current useful in the market because nobody wants to cost that much. And that's a fully expended FH so the price is going to be pretty high without mods.
Falcon Heavy is in the "super heavy" launch vehicle class by the US standard of 50 tons, but I use the russian standard of 100 tons because I think it's a better dividing line.
SMART has not been planned for any launch in the future. Is like "clean coal" plants, they are mentioned a lot, but none are ever built.
It doesn't make sense at ULA's current launch rate - the development cost would be greater than the cost savings. It might make sense if Kuiper launches happen at a high rate, so maybe we see it then.
I truly think ULA's mentions over the years that they'll be developing SMART is that they are embarrassed that they can't manage to build a reusable rocket - they can't afford it. No one will pay them to do it, and even if the DoD wanted to pay them the price would be way too high - ULA knows how uneconomical a company they are. Corporations may not have egos but the people running them do. They (and everyone else but China) don't want to admit how they simply can't compete with SpaceX.
Really sad to see that someone with a "Dr" to their name, cant imagine such alternative way of analyzing something.
I'm not sure whether that's him not knowing about market-based analysis or him understanding very well what result they would get if they used it.
Their graph shows that reusing the whole rocket is always worse than reusing only engines lol
For instance, ULA buys BE-4 engine from Blue Origin. ULA can only manufacture the disposable rocket, while having to purchase the engine from someone else.
In ULA's world, the engines are the most expensive part of the rocket and their architecture means they can't bring back the whole booster...
Objection! Cost per (useful) kg to space _is_ probably the single most important metric. Your counterargument is weak on two fronts: first, the Falcon Heavy cost per kg is basically fake, and second, Electron flies frequently but it neither flies a lot of stuff nor makes much money doing so. The market for Electron is hugely dominated by SpaceX's rideshare, because SpaceX's rideshare is cheaper per useful kg to orbit. In some very real sense it is absolutely surprising GAzelle launched on Electron, because very few payloads of this class do.
I suggested the second part of this to him on Twitter not too long ago with a hint that there's a fundamental flaw involved. He didn't quite find it (although I guess, since it's a combination of 3 assumptions leading to a false result, undermining any 1 of those assumptions fixes it (plus another 4th possibility that was paid lip service in this vid), and undermining the importance of $/kg would work. It's just not my preferred method that I was hoping he'd discover).
I'd agree that it's the single most important metric. It's not the only important metric, though. An analysis looking only at cost per kg can miss the forest for the trees.
Yet even though you raise these objections, the launch market has clearly and obviously chosen Falcon 9 as its work horse. It is the cheapest launch vehicle with the capabilities it has, the reliability at this point is unmatched, and I'd wager they can launch with a quick turnaround time given their cadence. Cost/kg is not the core metric. It's a combination of cost, and payload capacity. The goal for the customer is finding a booster that meets the payload capacity they need, at the lowest possible cost. This perfectly explains why Falcon 9 is popular: it's both cheap, and capable enough for most payloads.
Over ruled! Cost per kg by launch is the better metric, as it doesn't matter if you can launch at $1000 per kilogram if your payload is only 100 kg and the launcher carries 20 tons. You're still paying $20 million to launch your payload, rather than $100,000, unless, as with SpacX's rideshare, you can find a whole lot of other payloads going to the same place..
It's all about markets and coverage. There are several markets in play... 1) Payloads that fly electron because they want a specific orbit/schedule/whatever. 2) Payloads that could fly either electron or rideshare but choose rideshare because it's cheaper 3) Payloads that can only fly rideshare because electron costs too much. A lot of what rideshare flies is market #3. There are certainly some payloads in market #2 where SpaceX is taking business away from Electron. There also may be some customers who are doing development using rideshare - in either market #2 or #3 - who will need a specific orbit when they go into deployment, so in some cases this may be a good thing. Market #1 is the one that only Electron can cover, and it doesn't matter that their cost/kg is high because rideshare isn't an option to get where they want to do, and on a per-flight basis Electron is a lot cheaper than Falcon 9. Customers buy payloads to specific orbits first, then they look at price.
Good video, and I’m glad you covered the forum thread and analysis Sowers posted that I linked you on Twitter… But you still didn’t find the fatal flaw I was hinting at. I’ll say this: you can fully abide by the $/kg analysis and use the numbers they inputted for Falcon 9 and still find a major issue with the paper that would unfortunately wind up changing the entire equation on the spot.
Why are you being coy? If you have novel insight into a discussion, share it. You're not a villain in a mystery novel.
@@rhannan011 Eh, I should be. Was just wondering if I'd get a more in-depth analysis of the article if I gave the hint, not the answer. Anyways, the analysis: A: Assumed $/kg is the metric to judge by B: Assumed a launch cadence of 10 flights per year C: Assumed it was best to do the data analysis based on yearly operational cost per launch instead of marginal cost. Or to put it another way: They assessed value by $/kg, but normalized based on launch cadence, and factored in fixed costs as part of the analysis. The marginal cost savings from reuse were being mostly offset by the fixed cost/kg lifted to orbit since the reusable rocket is only doing 70% as much in a year as the expendable. They should've instead normalized on equal kg to orbit in a year.
I already had the forum thread before you posted - Part II was almost done before I posted part I. There were a lot of things I didn't like about the equation and approach, but I decided I didn't want to dig too deeply into it as I would lose people in the complexity, and talking about the fixed cost didn't make the cut. The model I did originally compared F9 expendable and drone ship on an equivalent payload to orbit basis, but it was really hanky and just didn't make the point I wanted to, so I did the cost per kg one instead. The fatal flaw isn't in the model, the fatal flaw is that cost per kilogram is a stupid way to evaluate reuse.
@@EagerSpace oh cool. Didn’t realize you had already gotten into it. But I still think the equation being stuck on an equal flight rate is more the flaw than $/kg, although the latter is still a flaw.
fizz…
4 buzz fizz 7 8 fizz buzz 11 fizz 13 14 fizzbuzz 16 17 fizz 19 buzz fizz 22 23
It wasn't clear why you made part 1, and it's even less clear why you made this part 2. This stuff is nearly a decade old, and you are viewing it entirely through the lens of what happened after. ULA did the workup on their own to guide their decision about future developments. They aren't a scientific body and didn't release this paper to teach people how to make the decision. They were just communicating that there was a process guiding their decision, and laying an outline of what that was. In retrospect SpaceX seems to have made the right choice and it is working well for them, but ULA making a different choice before that information was known isn't a slight on SpaceX.
He isn't really looking at it through the lens of what happened after. For launches, he used ULAs launches in the period 2000-2009. Even at the time of the analysis, it was clearly flawed. And while I agree it is ancient history, it is still relevant. ULA is still planning on SMART reuse. It's not like what happened after has caused ULA to change course. And at this point, they can't really change course. They've engineered their way into a corner, where landing the first stage is extremely difficult, if not impossible. (Due to staging very high and fast.) The time for changing course was before they settled on the current design of Vulcan Centaur. I'm not sure if ULA cares, though. They are probably just going to try to survive long enough to get bought out. And if they are bought by Blue Origin, you can expect launches to get moved over to New Glenn and Vulcan Centaur to get axed. We'll see - if they aren't bought out by Blue Origin, I would expect Vulcan to keep flying for as long as they have juicy government contracts. A lack of commercial viability doesn't have to be a show-stopper.
@@SpaceAdvocate He repeatedly compares they to SpaceX reuse, something that did not exist at the time. That's what I was referring to. We do not currently know that smart reuse isn't a better idea. We will see after they start doing it if they were right.
@@opcn18 At the time, both booster reuse and SMART reuse existed on paper. What's happened since then is that booster reuse has been made a reality - but that hasn't really added any new information to the analysis. Booster reuse was implemented in the way it was expected to be implemented. At this point, it's not really possible to see how SMART reuse could be a better idea. If you look at Falcon 9, the first stage is thought to cost around $20 million of which the engines is less than $9 million. SMART reuse would throw away something like $11 million in hardware that SpaceX currently reuses. As the chart at 17:05 shows, there's basically zero payload hit from booster reuse, relative to the actual market, so that wouldn't affect the conclusion. I will say that SMART reuse might make more sense for ULA than it does to SpaceX, because ULA doesn't make their own engines, which means they are expensive. ULA will be paying at least $16 million to Blue Origin for the engines, and the thrust structure and plumbing likely pushes the cost of the hardware they would be recovering to something like $25 million. So, they'd be recovering a greater dollar value in hardware with SMART reuse than SpaceX recovers with booster landings! This may be offset by what they will need to pay Blue Origin for engine refurbishment. Blue Origin will of course be trying to make as high a profit as possible.
@@SpaceAdvocate confirming your projections is ABSOLUTELY new information. Most of Musks projections have been confirmed false, that was a rare one that was confirmed accurate. There is a big energy hit from sending up all the fuel you need to come down safely. It is not as if F9 with full first stage reuse outperforms Vulcan Centaur across all launch profiles. ULA had the opportunity to pursue propulsive reuse with vulcan, they put their money where their mouths where by following the SMART reuse strategy. Also the tank that gets thrown away with SMART reuse does not cost 55% of the cost of the first stage. ULA is looking to recover ~70%-80% of the first stage value. Vulcan was designed for a high energy first stage, the staging happens at too great an altitude to do full first stage recovery. The die is cast. For them to go back and follow a different strategy at this stage in the game would take billions of dollars of rework which would take a great many launches to make back. ULA made a projection. They followed it up. If it was a lie they were the only ones that that lie could possibly hurt. We have new information since that time but that new information can't propagate back to change the past. We should wait to see how ULAs vehicle pans out before we draw a conclusion.
@@opcn18 I will grant you that I don't think ULA were lying when they said SMART reuse was better. I think they sincerely believed it.