Impact of Internet access due to new SpaceX re-usability achievement

[farzyness]

HI all,

Wanted to share with you all an article that outlines the potential impact of SpaceX's recent achievement in re-usability, and how it could enable fast internet everywhere due to much lower costs of rocketry into LEO.

http://www.degeneratestalk.com/posts/2015/12/22/fast-internet-everywhere-is-now-possible-thanks-to-spacex

[Nibb31]

We still don't know if recovering a stage actually allows economically viable reusability.

Reusing and refurbishing the first stage might allow some cost savings compared to building a new one, but there are many more factors involved in the cost of orbital launch. The manufacturing cost of the first stage is actually only a small part of the total cost of launching a rocket, maybe only 30% (optimistically). The rest of the cost is mainly the workforce, planning, infrastructure, logistics, R&D, transport, administrative overhead, etc... And there's transport, integration, launch services, and a lot of stuff that isn't recoverable (upper stage, fairing...). 

The payroll of the workforce is the biggest part of the launch cost, and reusability doesn't magically reduce the workforce. SpaceX is just about as lean as a launch provider can be and has already slashed prices as much as they could.  Even with reusability, they will still need a factory and lots of engineers. Even if you need a few less people on the production lines to build less boosters, you need people to refurbish, prepare, and transport the recovered stage, which wasn't needed before. 

And this assumes that the first stage is actually free, which it isn't. It's designed to fly maybe 20 times, but not indefinitely. You could assume that by spreading the manufacturing cost over 20 flights reduces the cost per flight by 95% of the above-mentioned 30% figure, but it isn't even that simple.

Disposable rockets (especially the Falcon 9) are actually designed to be (relatively) cheap, partly because they are produced in numbers. Manufacturing costs diminish with volume, meaning that as launch volume increases, the unit cost of each booster decreases.If you have 50 launches per year, with a disposable model, you need to mass produce 500 Merlin engines and 50 first stages. With 100% first stage reusability, the same factory has to build only 2.5 first stages and 72.5 engines. The result is that due to lower procurement volumes and higher fixed costs, those reusable stages are going to cost a lot more than the disposable ones. Enough to seriously cut into the reduction induced by reusing the stages in the first place. Instead of saving 95% on the manufacturing cost of the first stage, for the same amount of flights, the real cost reduction might only be 50%.

So in the end, what sounded like a 95% (of 30%) reduction of launch cost might only turn out to be a 15% reduction, which has the potential to bring the cost of a Falcon 9 launch from $60 million down to $45 million. It's a nice perk to pass on to your customers, but it's not a game changer.

Now, from the customer's point of view, the actual launch is only a small part of the total cost of a typical project. Maybe, again, 20%. The rest is the satellite itself (the biggest part of the budget), the ground stations, the insurance, and the actual operations. This means that in the grand scheme of things, the total saving that a customer can expect when they put a satellite in service is 15% of 20%, which is only 3%. On a $200 million comsat project, that's a whopping $6 million saving on their total expenses. Again, it's a nice saving, but it's not a revolution.

Note: The above figures are educated guesses, but I believe the orders of magnitude are pretty close.

Edited December 23, 2015 by Nibb31

[Guest]

The way the people in the control room were cheering, you would think they had just won the lottery, not just a mere 4%.

I would also think that the barriers to faster internet have more to do with greedy communications companies, controlling governments, and copyright law than lowering the cost of a satellite launch. We've had plenty of time to expand networks, build towers, run cable, etc., but yet it hasn't been done. Add to that the threat of internet providers being held liable for the actions of their customers. You can no longer provide "free" or "cheap" wi-fi to anyone, because if they do something online that a copyright troll doesn't like, you can be disconnected and/or held financially responsible.

[peadar1987]

Well it's not all about the money. It's a great engineering achievement, and pretty awesome in its own right

[Elthy]

Faster internet could be quite cheap. Someone calculated to build fibre networks with almost unlimited bandwidth for whole germany you would need about 90.000.000.000€, a bit more than 1000€ per citizen. The reason this doesnt happen is that its better for companys to keep you on a slow connection. E.g. if you have slow cable they build LTE in your area, but you can only use like 60GB per month for a high price, if you want more you have to pay even more. For the 400GB usage per months (no filesharing!) of my family we would have to pay like 200€, but we are lucky that we have fast internet over TV cable...

[magnemoe]

2 hours ago, Nibb31 said:

increases, the unit cost of each booster decreases.If you have 50 launches per year, with a disposable model, you need to mass produce 500 Merlin engines and 50 first stages. With 100% first stage reusability, the same factory has to build only 2.5 first stages and 72.5 engines. The result is that due to lower procurement volumes and higher fixed costs, those reusable stages are going to cost a lot more than the disposable ones. Enough to seriously cut into the reduction induced by reusing the stages in the first place

Problem with that argument is that 500 is not mass production its two engines each day.
SpaceX has an large backlog with reuseable rockets they can cut down on this wile not expanding production.
Yes the fixed cost stay the same without reuseablity they would have to ramp up production 
First and second stages share a lot of parts on the falcon 9 too.

After getting some experience an used stage would be safer than an new as its has been used.
Error rate always fall after some use, then start rising as product start to age.

[Nibb31]

22 minutes ago, magnemoe said:

Problem with that argument is that 500 is not mass production its two engines each day.

In the aerospace industry, two units a day is mass production. Sure, it's not comparable to the volumes for cars or cell phones, but compared to the production levels of other rocket engines or even jet engines, SpaceX benefits immensely from economies of scale on two main levels: sharing the fixed costs over more units and discounts from procuring larger volumes.

Those benefits have been the major factor in allowing SpaceX to cut costs so much. Engine commonality in particular was a genius idea, because it is much cheaper to a have a production line build 10 identical engines for each rocket than to have two production lines for first and upper stage engines. 

When you start reusing boosters on a large scale, your production volumes decrease significantly, and with that decrease you lose most of those scaling benefits. They will undoubtedly still be profitable, but I don't think that the actual savings when comparing the two models are as clear-cut as many people seem to think.

Edited December 23, 2015 by Nibb31

[Frozen_Heart]

I don't think that they plan to slow rocket production down even if they do start reusing them. I wouldn't be surprised if they want several going at once to get through the launch manifest quicker, which they are massively behind on. I doubt that they will be reusing stages soon enough to help with that though.

Even so I'm still not convinced that the 1st stage booster is much of the launch cost as a whole. Though even getting $10million dollars off the launch would be pretty huge.

Also is SpaceX still working on the space internet idea? I heard somewhere that it has taken a back seat as wasn't really being worked on anymore.

[cantab]

Satellite internet will never be able to match even ground cellular networks, never mind ground wired/fibre networks, for delivering high speeds to large populations. You've got a finite radio bandwidth to use and a large area to cover with it.

It's useful for sparsely populated areas where that large area still doesn't mean many people, it's useful for poor countries where people don't expect as much, but it's not gonna cut it when it comes to rich cities.

[Nibb31]

10 minutes ago, Frozen_Heart said:

Even so I'm still not convinced that the 1st stage booster is much of the launch cost as a whole. Though even getting $10million dollars off the launch would be pretty huge.

Yes, as I said, it's a nice incentive, but it won't be a game changer in that won't stimulate a substantial growth in the launch market. The difference between $60 million and $50 million isn't big enough to create new business models and a huge demand for space launches.

10 minutes ago, Frozen_Heart said:

Also is SpaceX still working on the space internet idea? I heard somewhere that it has taken a back seat as wasn't really being worked on anymore.

I think it's on the long list of stuff that Elon has tweeted about, the fanbois have got their knickers in a twist about, and then has been put on the back burner because it simply isn't such a great idea. SpaceX seems to be concentrating on RTF, Raptor, MCT/BFR these days, which is already quite a lot for a lean company.

I don't think the LEO internet constellation was ever as much as a cash cow as some people seemed to think. The telecom industry is highly competitive, with low margins, and rapid evolution. 4G offers more than most people really need these days with a wide coverage, and 5G is on the way. Catering to the folks who don't have access to 4G is becoming more and more of an unprofitable niche. It doesn't necessarily make sense to make a huge investment in a niche technology with a long lead-time when the rest of the telecom world is advancing at a quick pace.

[YNM]

Space internet ? While you will have fiber optic cables anywhere soon ? Not that sound IMO...

[Frozen_Heart]

19 minutes ago, YNM said:

Space internet ? While you will have fiber optic cables anywhere soon ? Not that sound IMO...

And then there is me with 400kb/s internet that hasn't been updated since it was first installed.  The companies are less and less interested in supplying us so if this breaks it might not even be replaced.

[YNM]

Wow, really, in the UK ? Where are you, Scotland ? The last thing I saw near my home here was a promotion booth for new cable optic installation... At least two national provider have installed LTE networks at least in the island. Sure you may ask "what about the other island" but sadly it seems there haven't. Ah... Remoteness I suppose ? But guess that actually we do want to have less and less transmitters. Any satellite would serve only as major gateways (like, a band of satellite for a country or so), not for a single people like what it was trying.

[wizzlebippi]

I think spacex will see substantial cost reductions eventually, but it won't be for years. They're going to need to recover several boosters, examine them, test for residual strength, and implement their findings in the design multiple times until they can recover a booster with no structural damage. If they launch a recycled booster in the next 2 years I'd be surprised. Even more surprised if it carries a customer's payload.

[Frozen_Heart]

19 minutes ago, YNM said:

Wow, really, in the UK ? Where are you, Scotland ? The last thing I saw near my home here was a promotion booth for new cable optic installation... At least two national provider have installed LTE networks at least in the island. Sure you may ask "what about the other island" but sadly it seems there haven't. Ah... Remoteness I suppose ? But guess that actually we do want to have less and less transmitters. Any satellite would serve only as major gateways (like, a band of satellite for a country or so), not for a single people like what it was trying.

Down in Devon. On the moors we have very little infrastructure. Even our water has to be pumped up from a nearby stream.

[RainDreamer]

Google with their wifi drone plan is probably more plausible here.

[Apexazimuth]

5 hours ago, cantab said:

Satellite internet will never be able to match even ground cellular networks, never mind ground wired/fibre networks, for delivering high speeds to large populations. You've got a finite radio bandwidth to use and a large area to cover with it.

It's useful for sparsely populated areas where that large area still doesn't mean many people, it's useful for poor countries where people don't expect as much, but it's not gonna cut it when it comes to rich cities.

I think you're referring to commsats in geostationary orbit.  The majority of satellite ISPs use relays in geostationary orbit, which is why latency sucks.

The future of satellite ISP's are not through satellites in GEO, though.  There are constellations of satellites in MEO and even LEO, that can provide broadband connections with latencies under 10ms.  With the combination of cellular and low-orbit relay constellations, broadband internet could be provided globally with no geographical limitation.

Doing this requires dozens of small relays in orbit, which is why bringing the cost of launches down is key to making this happen.

[PB666]

14 hours ago, farzyness said:

HI all,

Wanted to share with you all an article that outlines the potential impact of SpaceX's recent achievement in re-usability, and how it could enable fast internet everywhere due to much lower costs of rocketry into LEO.

http://www.degeneratestalk.com/posts/2015/12/22/fast-internet-everywhere-is-now-possible-thanks-to-spacex

I saw this very early this morning but decided not to thread the topic, the problem is the publication is rather speculative about what can be achieved at what price. The article is one of those just so presentation The UK is already in process of placing a wider array of satellites that would provide cell phone and internet service to remote areas, so this is also not novel.

Space X return launch vehicle however for placing supplies and equipment in space for deep space missions such as a future mars landing mission

Edited December 23, 2015 by PB666

[cantab]

3 hours ago, Apexazimuth said:

I think you're referring to commsats in geostationary orbit.  The majority of satellite ISPs use relays in geostationary orbit, which is why latency sucks.

The future of satellite ISP's are not through satellites in GEO, though.  There are constellations of satellites in MEO and even LEO, that can provide broadband connections with latencies under 10ms.  With the combination of cellular and low-orbit relay constellations, broadband internet could be provided globally with no geographical limitation.

Doing this requires dozens of small relays in orbit, which is why bringing the cost of launches down is key to making this happen.

Lower orbits help the latency, but there are still reasonable limits to the spectral bandwidth and the area on the ground that has to share that bandwidth linking up to the satellite.

https://en.wikipedia.org/wiki/ViaSat-1 can serve as an example. It quotes a total capacity of 140 Gbps, spread over 72 "spot beams". Sounds impressive. But then consider that if we assume each beam has 2 Gbps, if one of those beams is serving New York City then each New Yorker gets an internet connection speed of about 20 bits per second. A lower orbit isn't going to turn bits-per-second into megabits-per-second. Satellites are hopeless for serving dense urban areas.

[AngelLestat]

1 hour ago, cantab said:

Lower orbits help the latency, but there are still reasonable limits to the spectral bandwidth and the area on the ground that has to share that bandwidth linking up to the satellite.

https://en.wikipedia.org/wiki/ViaSat-1 can serve as an example. It quotes a total capacity of 140 Gbps, spread over 72 "spot beams". Sounds impressive. But then consider that if we assume each beam has 2 Gbps, if one of those beams is serving New York City then each New Yorker gets an internet connection speed of about 20 bits per second. A lower orbit isn't going to turn bits-per-second into megabits-per-second. Satellites are hopeless for serving dense urban areas.

This is the way you analyze a business plan?

They use a different frequency, also you would not sale internet to each person in the world, not even mention your worst example of each person on the most denser city.

Take into account you have more than 4000 sats, this mean more than 1 sat for populated areas, because farm areas would not require much and you still can get connection of those sats.

Then as last, to calculate a business plan you need to calculate the cost, and how much clients you can have (there is a limit).  So you divide the revenue for those clients, the lifetime of the service and the overall cost plus the technology value that you can sale to other companies trying the same thing.. Then there are many economic and business strategies that can be applied to increase the revenue or reduce the overall cost.

As final point.. a lot of the biggest companies in the world are interested, like samsung, google,  WorldVu, etc.
You really believe than none of these companies add 2+2 to know if it worth or not?

---------------------------------------------------------------------------

Spacex prices will go down at the same time their confidence in return their stages rise, also more often they launch, is easier for them to deal with ground operation, planning, etc.
The prices will go down faster than everybody here thinks, and the business applications will rise at the same time.
In other words.. the same thing I am saying since I am in this forum.

Edited December 24, 2015 by AngelLestat

[Bill Phil]

It's like this: 

Throw away paper plates are easy to use and cheap. Heck, you can use them as fuel... *chuckles* 

Compare that to plates you wash. You need to either wash them yourself in the sink, taking time, as well as other resources, or you can use a dishwasher machine, which requires electricity, and has an upfront cost. 

Dishwashers are only "economical" because they're a convenience and electricity has been cheap for quite a while in developed nations (for the most part).

Reusability requires a different infrastructure entirely than throwaways, a completely new one. Is it economical for rockets? That depends. Plates aren't turning thousands of RPM, aren't at enormous pressures and aren't at enormous temperatures. But over time it could be made economical, with lots of investment.

[PB666]

43 minutes ago, Bill Phil said:

It's like this: 

Throw away paper plates are easy to use and cheap. Heck, you can use them as fuel... *chuckles* 

Compare that to plates you wash. You need to either wash them yourself in the sink, taking time, as well as other resources, or you can use a dishwasher machine, which requires electricity, and has an upfront cost. 

Dishwashers are only "economical" because they're a convenience and electricity has been cheap for quite a while in developed nations (for the most part).

Reusability requires a different infrastructure entirely than throwaways, a completely new one. Is it economical for rockets? That depends. Plates aren't turning thousands of RPM, aren't at enormous pressures and aren't at enormous temperatures. But over time it could be made economical, with lots of investment.

Are you comparing an SFRB with a paper plate? Just remember paper plates cavitate,mleak spill, etc. Lets not overstate the case like the OPs link does but lets also give the reuseabikity folks have thier day in the sunshine, im a big reusability fan, i minch up old legal documents and veg waste and compost them, I hate waste, erks me to no end, The other thing is that returnable vessels help to keep the density orbital bands clean. 

I see alot of promise with this vessel for potentiall building a mid orbital factory, because it looks like there is the potential of having an assembly line of fast turn over vehicles, if NASA can have an in orbit assembly platform then we can seriously talk about manned missions within the inner solar system. 

[Bill Phil]

1 minute ago, PB666 said:

Are you comparing an SFRB with a paper plate? Just remember paper plates cavitate,mleak spill, etc. Lets not overstate the case like the OPs link does but lets also give the reuseabikity folks have thier day in the sunshine, im a big reusability fan, i minch up old legal documents and veg waste and compost them, I hate waste, erks me to no end, The other thing is that returnable vessels help to keep the density orbital bands clean. 

I see alot of promise with this vessel for potentiall building a mid orbital factory, because it looks like there is the potential of having an assembly line of fast turn over vehicles, if NASA can have an in orbit assembly platform then we can seriously talk about manned missions within the inner solar system. 

I'm pointing out that a new infrastructure IS required for reusability. One which may not help for rockets. If it does, then that's great! But it might not.

Edited December 24, 2015 by Bill Phil

[PB666]

2 hours ago, Bill Phil said:

I'm pointing out that a new infrastructure IS required for reusability. One which may not help for rockets. If it does, then that's great! But it might not.

Sure, we need to see what the line looks like once they start cycling, but it could be years before that is in an equilibrium state and you can estimate the variable cost per cycle.   This stuff doesn't startup from a lemonade stand, you need to sit on a pretty heavy cap inv before you are going to have a consistent revenue flowing in, if the fed were smart it would return th interest lost on the investment (return on investment if place elsewhere) cause its sure not going to be properly expensed in depreciation. Basically the only way they would ever get that back if they sell they whole thing or go public and sell stock. You just dont have lines of people banging on your door wanting too launch cargo into low earth orbit. The other thing is that there is nothing stopping the chinese from copying thier rocket and offering an asian verion of a space truck, of course a national gov can afford to throw many if they think its for the overall benefit of thier economy. 

What i am saying barring a sudden unexpected demand for mod/high end payloads, the don't have enough clients to reach an equilibrium state where they can basically spread there startup cost over hundreds of cycles and 10s of launchers. The true cost may not be known for quite a while. 

 

 

 

Edited December 24, 2015 by PB666

[RuBisCO]

Some of the things world wide internet coverage could do

- People in poor countries can get online, from the middle of nowhere.

- People in Orwelling nightmares could get online, North Korea is going to have trouble once sat-modems start sneaking their way in the country.