Author Topic: $200,000 to $300,000 per flight in fuel and oxygen versus a $60 million rocket  (Read 40179 times)

Offline Dave G

  • Senior Member
  • *****
  • Posts: 3231
  • Liked: 2127
  • Likes Given: 2021

Also, the barrier to entry. Early airplanes were pretty simple and made out of cheap materials.

Actually -- not trolling here -- aren't rockets pretty simple and made out of cheap materials, too?

Combining decades of experience with today's modeling, fab technology, and materials science, anyone with a spare Billion or so could make them.

Yeah, but the wright brothers started out in a bicycle shop IIRC. I don't think it's really comparable.

The Wright brothers manufactured bicycles during the bicycle craze of the 1890s.  Back then, bicycles were a mainstream form of transportation.  The only real alternative involved horse manure.  In fact, the first paved roads in the U.S. were for bicycles.

And at that time, chain driven bicycles and air filled tires were cutting edge new technology.  So the Wright brothers were thriving high tech entrepreneurs.  They were far from poor.  It was nothing like a bike shop today.

Also, by 1908, the U.S. and French governments were funding the Wright brother's airplane development.

By comparison, Elon Musk started SpaceX with $100 million.  That's a lot more than the Wright brothers had, but its nowhere near a $billion, as some have speculated.
« Last Edit: 03/11/2015 09:11 am by Dave G »

Offline Alexsander

  • Full Member
  • *
  • Posts: 127
  • Liked: 17
  • Likes Given: 5
Low price isn't all you need to corner the market.. you also need volume.

Bigelow is flying BEAM this year (CRS-8). They are hiring a lot lately. Soon they will be able to launch a private space station (aka Space Hotel). Dragon is becoming human-rated probably next year. Now imagine a couple F9 per month (maybe a weekly trip?) carrying a sold-out Dragon (6 guests + 1 Bigelow employee for rotation). There's a HUGE market for cheap tickets to LEO.

Offline macpacheco

  • Full Member
  • ****
  • Posts: 891
  • Vitoria-ES-Brazil
  • Liked: 368
  • Likes Given: 3041
It isn't just reducing the price to get something into orbit, it is a matter of seriously reducing the price of what goes into orbit as well. The 35%, 50% and even 75% reductions you speak of are in effect only 10, 15, and 20% reductions in the costs of traditional commercial satellites. However, if you coupled the 35%, 50% and 75% reductions in launch costs with radical new manufacturing options and teams of people experienced with these options that reduced the cost of the spacecraft being launched to $10,000 to $20,000 a kilo from $50,000 to $100,000 which these constellations will do, then you will have a sea change in demand.  The link from another thread that suggests that even private aircraft might benefit from mobile satellite data with OneWeb, the possibility of specialized ground monitoring applications, specialized micro gravity research affordable at cube sat prices, asset or logistics tracking applications.  All of this though is predicated on the idea that lower launch costs spur people to make larger numbers of payloads that cost less to make.
That's what SpaceX satellites is trying to do. Building its own space internet constellation is just one application of the lets make a simpler satellites with state of the art, not yet space qualified hardware, let's launch one or two prototype satellites, and improve the design until it works, and we don't need birds qualified to fly for 15 years (ending up in 20-30 years service life), 4 years is fine (ending up in about  decade life), cause by then we have much better hardware to build much better birds.
More birds at lower altitudes replaced much more frequently ~ an order of magnitude more demand in launches.
Once SpaceX makes one basic successful design with this technique, the general common sat stuff will be qualified for other designs, much like a Boeing 702SP platform.
I really hope the rest of satellite mfgs can get their act together so they wont be left behind. Of course there's no guarantee SpaceX will succeed here, but their track record so far is excellent.
Hopefully by the time SpaceX sats has a reliable platform, we'll have first gen Raptor fully reuse rockets which will probably cost much more to build but should offer 100 all stage reuses, without significant refurb costs.
« Last Edit: 03/11/2015 03:23 pm by macpacheco »
Looking for companies doing great things for much more than money

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 10351
  • Everyplaceelse
  • Liked: 2430
  • Likes Given: 13606
Still not sure what the question on this thread is but I'll make a few points.

1) The big market is the communications satellite market in GEO. Reusability for that will not be seriously addressed by SX until the BFR.

2) Satellite mfgs already know how to build cheaper satellites. It involves building with higher margins (make the structures heavier) and reducing redundancy, going from 3 string redundancy to single string (single point failure kills the string and the satellite).

This only works if you can either a)Launch two satellites for the price of one (not necessarily on the same LV) and keep one on orbit in standby or b)Have a launch provider who can take your satellite from ground storage to working on orbit in a very short space of time, during which either other satellites on your constellation (if you have a constellation) take up the load) or you live without the capability.

Depending on the task "short" might have to be less than a week to a few hours.

So if you can't cut the price in half you're going to have to cut the capacity of the satellites you launch. :(
MCT ITS BFR SS. The worlds first Methane fueled FFSC engined CFRP SS structure A380 sized aerospaceplane tail sitter capable of Earth & Mars atmospheric flight.First flight to Mars by end of 2022 TBC. T&C apply. Trust nothing. Run your own #s "Extraordinary claims require extraordinary proof" R. Simberg."Competitve" means cheaper ¬cheap SCramjet proposed 1956. First +ve thrust 2004. US R&D spend to date > $10Bn. #deployed designs. Zero.

Offline meekGee

  • Senior Member
  • *****
  • Posts: 14158
  • N. California
  • Liked: 14046
  • Likes Given: 1392
1) The big market is the communications satellite market in GEO. Reusability for that will not be seriously addressed by SX until the BFR.

2) ...

Well, the biggest market is telecommunications.  How this is addressed is technology dependent.  GEO satellites, for example, are only suitable for broadcast TV, not internet, traffic.

That said, FH reusable can launch plenty of payloads to GEO.

So I stopped at (1)...
ABCD - Always Be Counting Down

Offline Adaptation

  • Full Member
  • *
  • Posts: 160
  • Liked: 40
  • Likes Given: 38
Power collected in space for terrestrial use could represent a tremendously huge demand of launch services.  Japan has recently and successfully demonstrated key aspects of the technology.  There are plenty of places on earth that burn diesel fuel for power.  This market would also be extremely elastic.  There are customers that would pay a great deal to be able to deploy an antenna array from a couple shipping containers and get a megawatt of energy anywhere on earth for extended periods of time. 

This is not a comsat constellation, its not millionaires wanting vacation destinations.  Demand will come from places we don't even consider.  I wonder how many people have started saving up for the half million dollar transfer to a martian life. 

Offline ChrisWilson68

  • Senior Member
  • *****
  • Posts: 5266
  • Sunnyvale, CA
  • Liked: 4992
  • Likes Given: 6459
Still not sure what the question on this thread is but I'll make a few points.

Most threads here aren't questions.  They are topics for discussion.

The topic for this thread is that a launch currently costs around $60 million, but the lower-bound on the cost of the launch is the prop cost of $200,000 to $300,000.  The fact that there's a huge difference between the current cost and the lower bound is interesting.  It brings up all kinds of questions, such as "How close can we get to the lower bound?", "When can we get there?", and "What needs to happen to get there?".

1) The big market is the communications satellite market in GEO. Reusability for that will not be seriously addressed by SX until the BFR.

Looking backward, GEO is the bigger market.  Looking forward, there's good reason to think the market for launches to lower orbits is likely to grow much, much more.

2) Satellite mfgs already know how to build cheaper satellites. It involves building with higher margins (make the structures heavier) and reducing redundancy, going from 3 string redundancy to single string (single point failure kills the string and the satellite).

This only works if you can either a)Launch two satellites for the price of one (not necessarily on the same LV) and keep one on orbit in standby or b)Have a launch provider who can take your satellite from ground storage to working on orbit in a very short space of time, during which either other satellites on your constellation (if you have a constellation) take up the load) or you live without the capability.

Depending on the task "short" might have to be less than a week to a few hours.

So if you can't cut the price in half you're going to have to cut the capacity of the satellites you launch. :(

Not true.  If you have a large constellation, you can get redundancy without having a spare for every bird.  If one fails, other nearby satellites take up the slack.

Offline dorkmo

  • Full Member
  • ****
  • Posts: 710
  • Liked: 338
  • Likes Given: 848
I wonder how many people have started saving up for the half million dollar transfer to a martian life.

whos got two thumbs and wants to go to mars, this guy

i know we joke about polls, but this might be a fun one

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 10351
  • Everyplaceelse
  • Liked: 2430
  • Likes Given: 13606
Most threads here aren't questions.  They are topics for discussion.

The topic for this thread is that a launch currently costs around $60 million, but the lower-bound on the cost of the launch is the prop cost of $200,000 to $300,000.  The fact that there's a huge difference between the current cost and the lower bound is interesting. 
How fortunate you are on hand to interpret the OP's meaning. That will make this much easier.
Quote
It brings up all kinds of questions, such as "How close can we get to the lower bound?",
That would depend on how well SX can managed to mfg an upper stage for a few $100K. I'll leave others who are much more familiar with production engineering to discuss how plausible that is.
Quote
"When can we get there?",
While the upper stage is disposable and costs 10s of $m I'd say "never."
Quote
and "What needs to happen to get there?".
Recover the upper stage or make it for less than a few $100k. Of course at that point you'll have a 1st stage for about $400-600K as well.
Quote
Looking backward, GEO is the bigger market.  Looking forward, there's good reason to think the market for launches to lower orbits is likely to grow much, much more.
Historically there's good reason to think most of those constellations won't get off the ground and the rest will go bankrupt and be sold for cents on the dollar. IIRC Iridium spent $5Bn on satellite launches for 77 satellites.

And then they went bankrupt and their assets were bought for something like 20-30 cents on the $.  :(

Google will be a captive customer and make no change to the greater market. They will presumably want (and get) the same sort of special treatment from SX they get from Intel due to their volume,
Quote
Not true.  If you have a large constellation, you can get redundancy without having a spare for every bird.  If one fails, other nearby satellites take up the slack.
You seem to have missed the point.

"b)Have a launch provider who can take your satellite from ground storage to working on orbit in a very short space of time, during which either other satellites on your constellation (if you have a constellation) take up the load) or you live without the capability."

Extra emphasis added.

If other satellites in the constellation can carry out the failed satellites functions continuously you would appear to have launched an excess satellite for no reason, which is expensive and pointless.

Why would you do that? One way or another if it's there in the first place when it fails it's going to need replacing.

There is no magic to making satellites cheaper. It's not a problem provided you have sufficiently fast acting contingency plans, either by positioning on orbit spares (also single string) or have a system that supports launch on an appropriate timescale. Trouble is that timescale is likely to be much shorter than any current launch provider can deliver.

But you've launched 2 satellites  so to stand still your launch costs have to halve, other wise you run with less than 2x the number of working satellites needed spread across the necessary orbits and hope you have one close enough to change orbit to replace the failed ones coverage without much loss of service to the customers. With a few spares you'll need lots of propellant (so do they have full capacity ?)  to do the plane changes.  More spares means more launches but with "full bandwidth" capacity satellites.

This lowers the amount you have to lower the launch costs by the fraction of "excess" satellites you launch. How much by is the big question.

Historically AFAIK the 1st generation constellation services company that's done best has been OrbComm, who offered the least IE store-and-forward non real time SMS.

Google seems to be talking intra-satellite real time broadband internet routing, which implies things like VoIP. Which is more like the Iridium end of the market.

The software for this will be very demanding and the hardware will be much more complex than most (all ?) existing commm sat onboard hardware.

Iridium has not exactly been a license to print money.  :(

Current operators of course do it by simply having multiple hardware strings and ground or on board computer controlled "cross strapping" to bypass failed units.
« Last Edit: 03/12/2015 01:05 pm by john smith 19 »
MCT ITS BFR SS. The worlds first Methane fueled FFSC engined CFRP SS structure A380 sized aerospaceplane tail sitter capable of Earth & Mars atmospheric flight.First flight to Mars by end of 2022 TBC. T&C apply. Trust nothing. Run your own #s "Extraordinary claims require extraordinary proof" R. Simberg."Competitve" means cheaper ¬cheap SCramjet proposed 1956. First +ve thrust 2004. US R&D spend to date > $10Bn. #deployed designs. Zero.

Offline meekGee

  • Senior Member
  • *****
  • Posts: 14158
  • N. California
  • Liked: 14046
  • Likes Given: 1392
JS, tens of millions of $ for the second stage?

And read through the spare strategy...  All constellations have spares, they are much cheaper then 100% redundancy, and the larger the constellation, the less spares you need, as a fraction.

You then argue that the future of constellations is bleak because Iridium sucked.  20 years ago. On a voice-centric to-the-handset model.  It  is a good data point to consider, as an example of what this new constellation is not.

ABCD - Always Be Counting Down

Offline Mader Levap

  • Full Member
  • ****
  • Posts: 976
  • Liked: 447
  • Likes Given: 561
I wonder how many people have started saving up for the half million dollar transfer to a martian life.
I will take wild guess and say number is somewhere around zero.
Be successful.  Then tell the haters to (BLEEP) off. - deruch
...and if you have failure, tell it anyway.

Offline JamesH

  • Full Member
  • ****
  • Posts: 525
  • United Kingdom
  • Liked: 284
  • Likes Given: 7
Most threads here aren't questions.  They are topics for discussion.

The topic for this thread is that a launch currently costs around $60 million, but the lower-bound on the cost of the launch is the prop cost of $200,000 to $300,000.  The fact that there's a huge difference between the current cost and the lower bound is interesting. 
How fortunate you are on hand to interpret the OP's meaning. That will make this much easier.
Quote
It brings up all kinds of questions, such as "How close can we get to the lower bound?",
That would depend on how well SX can managed to mfg an upper stage for a few $100K. I'll leave others who are much more familiar with production engineering to discuss how plausible that is.
Quote
"When can we get there?",
While the upper stage is disposable and costs 10s of $m I'd say "never."
Quote
and "What needs to happen to get there?".
Recover the upper stage or make it for less than a few $100k. Of course at that point you'll have a 1st stage for about $400-600K as well.
Quote
Looking backward, GEO is the bigger market.  Looking forward, there's good reason to think the market for launches to lower orbits is likely to grow much, much more.
Historically there's good reason to think most of those constellations won't get off the ground and the rest will go bankrupt and be sold for cents on the dollar. IIRC Iridium spent $5Bn on satellite launches for 77 satellites.

And then they went bankrupt and their assets were bought for something like 20-30 cents on the $.  :(

Google will be a captive customer and make no change to the greater market. They will presumably want (and get) the same sort of special treatment from SX they get from Intel due to their volume,
Quote
Not true.  If you have a large constellation, you can get redundancy without having a spare for every bird.  If one fails, other nearby satellites take up the slack.
You seem to have missed the point.

"b)Have a launch provider who can take your satellite from ground storage to working on orbit in a very short space of time, during which either other satellites on your constellation (if you have a constellation) take up the load) or you live without the capability."

Extra emphasis added.

If other satellites in the constellation can carry out the failed satellites functions continuously you would appear to have launched an excess satellite for no reason, which is expensive and pointless.

Why would you do that? One way or another if it's there in the first place when it fails it's going to need replacing.

There is no magic to making satellites cheaper. It's not a problem provided you have sufficiently fast acting contingency plans, either by positioning on orbit spares (also single string) or have a system that supports launch on an appropriate timescale. Trouble is that timescale is likely to be much shorter than any current launch provider can deliver.

But you've launched 2 satellites  so to stand still your launch costs have to halve, other wise you run with less than 2x the number of working satellites needed spread across the necessary orbits and hope you have one close enough to change orbit to replace the failed ones coverage without much loss of service to the customers. With a few spares you'll need lots of propellant (so do they have full capacity ?)  to do the plane changes.  More spares means more launches but with "full bandwidth" capacity satellites.

This lowers the amount you have to lower the launch costs by the fraction of "excess" satellites you launch. How much by is the big question.

Historically AFAIK the 1st generation constellation services company that's done best has been OrbComm, who offered the least IE store-and-forward non real time SMS.

Google seems to be talking intra-satellite real time broadband internet routing, which implies things like VoIP. Which is more like the Iridium end of the market.

The software for this will be very demanding and the hardware will be much more complex than most (all ?) existing commm sat onboard hardware.

Iridium has not exactly been a license to print money.  :(

Current operators of course do it by simply having multiple hardware strings and ground or on board computer controlled "cross strapping" to bypass failed units.

That's a great crystal ball you have there. A lot has changed in the last 10 years since Iridium. Launch prices much lower (and going lower all the time), satellite costs much lower, technology to form a grid network in space, much more advanced.

I really don't think you can compare a 7000 satellite constellation put up by the guy who owns the launch company, with a 70 satellite constellation put up by launch provider 10 years ago.

Offline meekGee

  • Senior Member
  • *****
  • Posts: 14158
  • N. California
  • Liked: 14046
  • Likes Given: 1392
Most threads here aren't questions.  They are topics for discussion.

The topic for this thread is that a launch currently costs around $60 million, but the lower-bound on the cost of the launch is the prop cost of $200,000 to $300,000.  The fact that there's a huge difference between the current cost and the lower bound is interesting. 
How fortunate you are on hand to interpret the OP's meaning. That will make this much easier.
Quote
It brings up all kinds of questions, such as "How close can we get to the lower bound?",
That would depend on how well SX can managed to mfg an upper stage for a few $100K. I'll leave others who are much more familiar with production engineering to discuss how plausible that is.
Quote
"When can we get there?",
While the upper stage is disposable and costs 10s of $m I'd say "never."
Quote
and "What needs to happen to get there?".
Recover the upper stage or make it for less than a few $100k. Of course at that point you'll have a 1st stage for about $400-600K as well.
Quote
Looking backward, GEO is the bigger market.  Looking forward, there's good reason to think the market for launches to lower orbits is likely to grow much, much more.
Historically there's good reason to think most of those constellations won't get off the ground and the rest will go bankrupt and be sold for cents on the dollar. IIRC Iridium spent $5Bn on satellite launches for 77 satellites.

And then they went bankrupt and their assets were bought for something like 20-30 cents on the $.  :(

Google will be a captive customer and make no change to the greater market. They will presumably want (and get) the same sort of special treatment from SX they get from Intel due to their volume,
Quote
Not true.  If you have a large constellation, you can get redundancy without having a spare for every bird.  If one fails, other nearby satellites take up the slack.
You seem to have missed the point.

"b)Have a launch provider who can take your satellite from ground storage to working on orbit in a very short space of time, during which either other satellites on your constellation (if you have a constellation) take up the load) or you live without the capability."

Extra emphasis added.

If other satellites in the constellation can carry out the failed satellites functions continuously you would appear to have launched an excess satellite for no reason, which is expensive and pointless.

Why would you do that? One way or another if it's there in the first place when it fails it's going to need replacing.

There is no magic to making satellites cheaper. It's not a problem provided you have sufficiently fast acting contingency plans, either by positioning on orbit spares (also single string) or have a system that supports launch on an appropriate timescale. Trouble is that timescale is likely to be much shorter than any current launch provider can deliver.

But you've launched 2 satellites  so to stand still your launch costs have to halve, other wise you run with less than 2x the number of working satellites needed spread across the necessary orbits and hope you have one close enough to change orbit to replace the failed ones coverage without much loss of service to the customers. With a few spares you'll need lots of propellant (so do they have full capacity ?)  to do the plane changes.  More spares means more launches but with "full bandwidth" capacity satellites.

This lowers the amount you have to lower the launch costs by the fraction of "excess" satellites you launch. How much by is the big question.

Historically AFAIK the 1st generation constellation services company that's done best has been OrbComm, who offered the least IE store-and-forward non real time SMS.

Google seems to be talking intra-satellite real time broadband internet routing, which implies things like VoIP. Which is more like the Iridium end of the market.

The software for this will be very demanding and the hardware will be much more complex than most (all ?) existing commm sat onboard hardware.

Iridium has not exactly been a license to print money.  :(

Current operators of course do it by simply having multiple hardware strings and ground or on board computer controlled "cross strapping" to bypass failed units.

That's a great crystal ball you have there. A lot has changed in the last 10 years since Iridium. Launch prices much lower (and going lower all the time), satellite costs much lower, technology to form a grid network in space, much more advanced.

I really don't think you can compare a 7000 satellite constellation put up by the guy who owns the launch company, with a 70 satellite constellation put up by launch provider 10 years ago.
All of that, plus the demand side bears no resemblance to what it used to be.
ABCD - Always Be Counting Down

Offline Dudely

  • Full Member
  • ***
  • Posts: 312
  • Canada
  • Liked: 109
  • Likes Given: 92
Every year, we spend 4 trillion dollars on upgrading, replacing, and improving our telecommunication systems. Even just 10 years ago this number was at least an order of magnitude lower.

The money exists to do what they propose (and, really, to do mostly anything- 4 trillion is an insane amount of money). The real questions are more economical.

A 15 billion dollar constellation is within financial reach. The only question is how much mileage you get for what you paid.


Actually, this was the exact problem Iridium had, wasn't it? . . . their constellation worked great, it just didn't do it any better than much cheaper terrestrial solutions that were developed at the same time. No one would pay the subscription price necessary for them to pay back the credit+interest on that 5 billion.

But again, we come back to economics. Elon says the only way to make his idea for a constellation work is if the constellation performs much of the backbone packet routing currently performed by long distance fiber optic cables. I tend to agree- if they plan to offer a subscription for end users to access a certain # of bps they will most likely fail- how could they ever keep up with a faster ground based solution? If they instead sell the massive bps directly to providers, by convincing them that eliminating the 100s of routers the packet currently has to go through to get across a continent, I think they stand to make a LOT of money. Like, we're talking revenue with 11 zeros on the end of it.
« Last Edit: 03/12/2015 02:15 pm by Dudely »

Offline Alexsander

  • Full Member
  • *
  • Posts: 127
  • Liked: 17
  • Likes Given: 5
JS, tens of millions of $ for the second stage?

Considering 1st Stage Reuse, we have a wishful thinking price range between:
- US$ 6 M (Shotwell's statement) and
- US$ 18 M (30% of current official price of US$ 61.2 M).

Of course, SpaceX could choose to pocket the difference and keep the price for a launch with a reused 1st Stage closer to US$ 60 M. Some say they will charge somewhere between US$ 30 M and US$ 40 M. However, assuming they are NOT pricing below cost, they COULD lower the price a bit more. Maybe US$ 20 M as a nice round number for a launch with a reused 1st Stage.

We also have a good estimate that suggest the cost of the 2nd Stage is around 16% of the rocket. So we can calculate this estimate for some guesses of the actual cost of the full rocket.
- 16% of US$ 60 M = US$ 9.6 M
- 16% of US$ 50 M = US$ 8.0 M
- 16% of US$ 40 M = US$ 6.4 M
- 16% of US$ 30 M = US$ 4.8 M

There's also the Dragon configuration, where you can reuse part of the upper stage if the Dragon lands in the pad as SpaceX has said it will. And that's the space tourism configuration. A price of US$ 20 M for a launch with reused 1st Stage + Dragon (from a CRS mission, maybe?) could mean less than US$ 3 M per seat.

Offline meekGee

  • Senior Member
  • *****
  • Posts: 14158
  • N. California
  • Liked: 14046
  • Likes Given: 1392
The numbers, especially the 30% one, are derived from partial statements, and since we don't even know the cost of the rocket....

SpaceX is vertically integrated.  This means they invested a lot in fixed costs, son that the marginal costs, per rocket, can be low.

When we talk "cost", do we include cost of fabrication line, amortized over several years worth of production?

I think the marginal cost cost, per rocket, is surprisingly low, and upper stages will be a couple of millions or less, especially when the production line is mostly doing upper stages.

But, like everyone else here, I have no knowledge to their costs.
ABCD - Always Be Counting Down

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 10351
  • Everyplaceelse
  • Liked: 2430
  • Likes Given: 13606
That's a great crystal ball you have there.
Not really. When I look into the past I use a history book
Quote
A lot has changed in the last 10 years since Iridium. Launch prices much lower (and going lower all the time), satellite costs much lower, technology to form a grid network in space, much more advanced.
Not as much as you seem to think.
Quote
I really don't think you can compare a 7000 satellite constellation put up by the guy who owns the launch company, with a 70 satellite constellation put up by launch provider 10 years ago.
That's going to be put up, future, not current tense. I'm not sure how many years away this is from happening.

What you're missing is that this Google network is exponentially more complex due to a)Intra satellite connectivity and b) The scale of the problem. BTW I'm not sure if they now have a global frequency allocation for this system. If not then the system has to switch frequencies while the satellites (and their handsets) are in violent  relative motion. This would be better avoided if possible.

Actually, this was the exact problem Iridium had, wasn't it? . . . their constellation worked great, it just didn't do it any better than much cheaper terrestrial solutions that were developed at the same time. No one would pay the subscription price necessary for them to pay back the credit+interest on that 5 billion.
Exactly
Quote
But again, we come back to economics. Elon says the only way to make his idea for a constellation work is if the constellation performs much of the backbone packet routing currently performed by long distance fiber optic cables. I tend to agree- if they plan to offer a subscription for end users to access a certain # of bps they will most likely fail- how could they ever keep up with a faster ground based solution? If they instead sell the massive bps directly to providers, by convincing them that eliminating the 100s of routers the packet currently has to go through to get across a continent, I think they stand to make a LOT of money. Like, we're talking revenue with 11 zeros on the end of it.
That reduces the number of ground stations and raises their technology level. Handy given complexity of this problem.

SpaceX is vertically integrated.  This means they invested a lot in fixed costs, son that the marginal costs, per rocket, can be low.

When we talk "cost", do we include cost of fabrication line, amortized over several years worth of production?
No. The phrase you're looking for is "cost of sales" and it's the parts and raw materials required directly to build the product. The cost of a production line is a capital expenditure and is depreciated over a period of time (its projected life time) with the write off amount per year depending on various standard formulae.
Quote
I think the marginal cost cost, per rocket, is surprisingly low, and upper stages will be a couple of millions or less, especially when the production line is mostly doing upper stages.
HMX said SX would supply him with an unmodified Merlin 1c for $5m.

Now the new version should (presumably) be cheaper and we can expect that SX were planning to make a profit on that deal

So $1m for the stage and $1m for the engine?

[EDIT The term "learning  curve" comes from aircraft production, where it's the cost reduction per doubling of production numbers. Historically it's been about 15%. However as mfg became more automated (such as in the F15) the curve became shallower.

If the current price for an upper stage is actually about $17m you might like to work out what the production volume needs to be before it drops to about 1/8 of that level ]

All of that, plus the demand side bears no resemblance to what it used to be.
Indeed. The aggregate bandwidth requirements are into the terra bits.  :)

« Last Edit: 03/14/2015 12:44 pm by john smith 19 »
MCT ITS BFR SS. The worlds first Methane fueled FFSC engined CFRP SS structure A380 sized aerospaceplane tail sitter capable of Earth & Mars atmospheric flight.First flight to Mars by end of 2022 TBC. T&C apply. Trust nothing. Run your own #s "Extraordinary claims require extraordinary proof" R. Simberg."Competitve" means cheaper ¬cheap SCramjet proposed 1956. First +ve thrust 2004. US R&D spend to date > $10Bn. #deployed designs. Zero.

Offline Dudely

  • Full Member
  • ***
  • Posts: 312
  • Canada
  • Liked: 109
  • Likes Given: 92
Actually, this was the exact problem Iridium had, wasn't it? . . . their constellation worked great, it just didn't do it any better than much cheaper terrestrial solutions that were developed at the same time. No one would pay the subscription price necessary for them to pay back the credit+interest on that 5 billion.
Exactly
Quote
But again, we come back to economics. Elon says the only way to make his idea for a constellation work is if the constellation performs much of the backbone packet routing currently performed by long distance fiber optic cables. I tend to agree- if they plan to offer a subscription for end users to access a certain # of bps they will most likely fail- how could they ever keep up with a faster ground based solution? If they instead sell the massive bps directly to providers, by convincing them that eliminating the 100s of routers the packet currently has to go through to get across a continent, I think they stand to make a LOT of money. Like, we're talking revenue with 11 zeros on the end of it.
That reduces the number of ground stations and raises their technology level. Handy given complexity of this problem.

Yes, it allows them to gain a significant fraction of potential revenue from the constellation before being required to have full ground coverage. It is a clever technical solution overlooked by most commentators.

If you can provide a huge, direct, low latency pipe between two servers on opposite ends of a continent you can sell that pipe at a premium because it is a much lower latency connection than the "regular" connection which is routed through hundreds of switches and routers. This sort of transaction is constantly happening between ISPs and tech companies all over the world. The company I work for is a great example. We buy a direct link from our main servers in Ontario to our satellite office on the East coast where we do development. This link is jaw-droppingly enormous (full-bore you can send an HD movie in one second) and I can only imagine how much they charge.

The market for this is 10x the end-user market. 4G is chump change.


A satellite constellation could, in theory, provide a link between San Francisco and New York without having to provide coverage for any of the physical area in between, and could charge the same rate ISPs currently charge for the low latency pipes that already exist between those two places. (In theory, you could make it even better than direct fiber since the speed of light is 40% slower in fiber optics compared to vacuum, or air).


. . . Let's see, are there any companies with lots of servers all over the world that currently spends hundreds of millions of dollars a year trying to figure out how to better keep the data between their many servers up to date? Say, maybe a company that recently invested a large amount of money in SpaceX?

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 10351
  • Everyplaceelse
  • Liked: 2430
  • Likes Given: 13606
A satellite constellation could, in theory, provide a link between San Francisco and New York without having to provide coverage for any of the physical area in between, and could charge the same rate ISPs currently charge for the low latency pipes that already exist between those two places. (In theory, you could make it even better than direct fiber since the speed of light is 40% slower in fiber optics compared to vacuum, or air).
That's interesting.

It's also the very definition of the "high frequency trading"  used by some financial institutions. They'd certainly pay for it but wheather the consequences would be a good idea is another matter.  :(

Either offers a good justification for the investment however.  This is starting to move OT.
« Last Edit: 03/13/2015 01:52 pm by john smith 19 »
MCT ITS BFR SS. The worlds first Methane fueled FFSC engined CFRP SS structure A380 sized aerospaceplane tail sitter capable of Earth & Mars atmospheric flight.First flight to Mars by end of 2022 TBC. T&C apply. Trust nothing. Run your own #s "Extraordinary claims require extraordinary proof" R. Simberg."Competitve" means cheaper ¬cheap SCramjet proposed 1956. First +ve thrust 2004. US R&D spend to date > $10Bn. #deployed designs. Zero.

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
1