Low Earth Orbit Atmospheric Scoops

[MatterBeam]

2 hours ago, DerekL1963 said:

Not really, no.  Lunar mining (for example) is dismissed in a few lines as being "too expensive".  Boosting from Earth is equally curtly dismissed ("requires too big [expensive] of a rocket").   Comparisons of technology are noticeable by their absence.
Yet, you dismiss all technologies that aren't your favored solution as "too expensive" without providing any justification for the claim.  

You can't have it both ways.  You can't claim to be comparing the "objective" merits of various technologies when you only discuss one technology.  You can't bring in costs for solutions that aren't your favored one and then dodge the question of the costs of that solution.  (And your readers in this case are the denizens of the KSP Science & Spaceflight sub-forum.  We're interested in the details, not so much in high level handwaving.)

Lunar mining is a big project where large installations are installed on the lunar surface to either work autonomously or keep humans alive for years on end. It will cost a lot to set up and operate. 
Sending propellant up from the lunar surface and down to LEO costs at least 5670m/s. This is less than bringing it up from Earth's surface (9500m/s) or from an asteroid (10km/s+), but it simply cannot compare to the 0m/s cost of collecting propellant gasses from the upper atmosphere. 

Also, the lunar surface is notably devoid of hydrogen and extracting oxygen from oxidized minerals incurs a hefty cost in the form of melting and electrolysis. It is on the order of ~70MJ per kg of O2 before efficiency losses. Compare this to the ~1MJ/kg energy cost of simply cooling down the oxygen until it becomes liquid.  

Putting all these elements together, I can safely say that a lunar colony will not be able to produce propellants anywhere as easily as a scoop does, and certainly cannot start producing thousands of tons per year with a single Falcon 9 launch. 

My last point is that this blog post, like other technology-specific posts before, are all about detailing and analysing a specific concept. Extended comparisons with other technologies using hundreds of price assumptions is far beyond the scope of what I am trying to do, which is familiarize readers with the technology itself. If a reader decides it is very interesting, they are going to tweak the specifics of their own works or settings to push the technology into pro-eminence. I have no control over that, so why close up the number of options a reader has by asserting that one option or another is 'better.'

1 hour ago, Terwin said:

I thought @MatterBeam was dismissing development costs, not operational costs.

The operational costs that were indicated as follows:

refueling from earth: $$$ per ton (see per-ton launch costs elsewhere for exact numbers)  

Atmospheric scoop:  1 falcon-9 launch to LEO to put the craft into space providing ### tons of liquid nitrogen and ### tons of liquid oxygen per year of operation over and above the needs of self-fueling(operational life-span TBD but expected to be measured in years or decades)

Lunar or asteroid refining: large/expensive mining and refining infrastructure that will take a long time to set up and will need a way to operate long-term without human presence

(admittedly the moon/asteroid argument is not terribly clear and should probably include references to the cost to deliver replacement parts due to the high degree of wear and tear on mining equipment for a good apples-to-apples comparison)

 

Developmental costs are highly subjective. Operational costs do have historical, political, economical, financial, industrial and cultural variations, but they do correlate more strongly with the science and maths behind the technology. For example, if it costs 70MJ/kg to produce oxygen on the Moon and 1MJ/kg to liquefy it in low orbit, I can at least assert that the equipment and power requirements on the Moon will be much higher. Maybe not exactly 70 times higher, but 'significantly higher'. 

Just now, p1t1o said:

There is also the matter that if a moon based mining plant suffers a failure, it wont destroy itself completely.

I do think the *quoted* figures sound good, but apart from the minutae of its operation (everything seems to be assumed to work at 100% efficiency, with infinite mean-time-to-failure for every component), my main concern is reliability. The machinery required to process atmospheric gaws is non-trivial, and not only does it need to be 100% automated but also operated without mechanical intervention for years.

Nothing about human achievements says to me that that will be anywhere near as easy as it sounds, and it doesnt sound easy.

The long range space probes are probably more stable than an atmospheric processing plant in terms of reliability (things mostly being solid state, no cryogenic plumbing operating continuously, not in an unstable orbit etc) and have broadly the same "level" of complexity, and more importantly, are expected to operate independantly for comparable terms. And they do fail catastrophically with alarming regularity, considering the time, money and effort that goes into preventing that.Replacements are rarely, if ever, launched. All those cool stories about missions being saved by clever reprogramming usually involve severely curtailed mission capability.

What this whole debate boils down to, in my opinion, is that the scientific principles, and rough modes of operation, ARE in fact rather trivial to work out. There's no maths on this thread that a highschool student couldnt follow (another thing which makes me suspicious).

What IS difficult, is the economics and engineering. Will anyone insure it? Can you garuantee customers willing to use your service (not necessarily a given)? What *exactly* are these "development" and "operating" costs? Those are not insignificant questions, the entire success of the venture might hinge on the answer to just one of them. Absolutely nobody will invest based simply on a statement that "It will be cheaper I think" If anyone has ever worked on budgeting complex projects, one knows that it can rapidly match the complexity of any scientific or engineering study.

Price, not engineering principles or technological limits, is what has kept us away from the Moon, from Mars, for this long.

If the worked example scoop failed immediately after being released from the Falcon 9, it would be a roughly 22 ton mass with a 26604m^2 airbrake creating 252 newtons of drag. If it takes 50m/s for it to fall out of orbit, then it can wait for repairs for about 22000*7800/252*50= 3.4 hours. More likely it won't deploy its massive scoop and just wait with a stowed cross-section area of about 10m^2, so it will survive for 419 days. 

If the scoop works for a few months and collects 100 tons, then breaks down, it will survive for 19 days. When full at 1250 tons, it will stay in orbit with the scoop deployed for 9 days. With the funnel retracted, it will stay in orbit for 65 years.

It becomes apparent that if anything goes wrong with the scoop, all it has to do is retract or eject the funnel and it will stay in orbit for a very long time. During this time, it will wait for repairs just a hop away from any launch site. 

Also, mechanical intervention can happen every time a cargo hauler docks with it to offload propellant. This can be dozens of times a year. 

Long range probes are exposed to the full radiation of the sun and space, and multiple passes through the Van Allen belts if they're bringing back propellants using reusable vehicles. 

The technological and matheatical comparisons are what the ToughSF blog focuses on. Other less technical comparisons are harder to pin down and work out without a fully detailed list of the assumptions an author is using for their setting. An example of this is aversion to nuclear technology. If nuclear thermal rockets are in use, then a high-Isp scoop that collects nitrogen and oxygen is best. If they are not in use, a lower Isp scoop that consumes all of its nitrogen and retains the oxygen is best. That's a massive change, based on something which even in our own reality has oscillated throughout history!

[p1t1o]

@MatterBeam

It doesnt matter if it stays in orbit indefinitely, you still have to send another ship up, which is exactly what you wanted to avoid.

13 minutes ago, MatterBeam said:

Long range probes are exposed to the full radiation of the sun and space, and multiple passes through the Van Allen belts if they're bringing back propellants using reusable vehicles.

Radiation environment is a good point, but do I need to start listing probes that have failed in LEO or due to human error?

Galileo's main antenna got fouled up because they stored it wrongly on the ground!

4 minutes ago, MatterBeam said:

The technological and matheatical comparisons are what the ToughSF blog focuses on. Other less technical comparisons are harder to pin down and work out without a fully detailed list of the assumptions an author is using for their setting. An example of this is aversion to nuclear technology. If nuclear thermal rockets are in use, then a high-Isp scoop that collects nitrogen and oxygen is best. If they are not in use, a lower Isp scoop that consumes all of its nitrogen and retains the oxygen is best. That's a massive change, based on something which even in our own reality has oscillated throughout history!

Totally get it, its a great topic for a blog and its not really fair to expect it to come as a fully-realised project proposal, this would of course take one person years.

But, to take it to a serious discussion, one cant just guesstimate away huge facets of a problem, unless you are specifically recognise that you are making success an assumption for the sake of argument. This is perfectly valid, but it means that no meaningful comparisons can be drawn versus other solutions.

 

In my opinion, if it were to work perfectly, according to all your hypotheses, then yeah, these scoops are great. But also in my opinion, it'd have to be attached to something like the ISS, which means its way too expensive and Im back to boosting propellant up from the surface.

Mass produced, cheap, heavy lifters are a deceptively tough solution to beat.

The scoops can wait until we are post-scarcity, and then we'll have the tech to allow them to take care of themselves.

[MatterBeam]

@p1t1o: Oh definitely. If two-stage reusable launch vehicles become available without major refurbishment between flights, we could send propellants into orbit by the tons! It would pull the rug out on atmospheric gas scooping concepts. 

Edited September 13, 2017 by MatterBeam

[DerekL1963]

1 hour ago, MatterBeam said:

Putting all these elements together, I can safely say that a lunar colony will not be able to produce propellants anywhere as easily as a scoop does, and certainly cannot start producing thousands of tons per year with a single Falcon 9 launch. 

Since I didn't question either how easily either could be done, or claimed some other solution would be able to start producing so easily, I completely fail to see your point.  No offense, but while you quoted what I wrote - but did you actually read it?
 

1 hour ago, MatterBeam said:

My last point is that this blog post, like other technology-specific posts before, are all about detailing and analysing a specific concept. Extended comparisons with other technologies using hundreds of price assumptions is far beyond the scope of what I am trying to do

You're the one that introduced costs and used them to dismiss solutions other than your preferred one - while not actually producing any cost estimates for your preferred solutions.  You cannot have it both ways.  You can't claim your solution is cheaper - and then fail to justify that conclusion.

 

1 hour ago, MatterBeam said:

The technological and matheatical comparisons are what the ToughSF blog focuses on.

Since you made no technological or mathematical comparisons... You just rejected other solutions and went into deep detail on your preferred solution.  Again, what you claim to want to do and what you actually do are at odds.

Edited September 13, 2017 by DerekL1963

[YNM]

@MatterBeam

1 hour ago, DerekL1963 said:

Since you made no technological or mathematical comparisons... You just rejected other solutions and went into deep detail on your preferred solution.  Again, what you claim to want to do and what you actually do are at odds.

True. You don't even try to be sceptical about your own idea.

There's a good reason why you shouldn't believe anything unless proven otherwise when it comes to deducing is something viable or not.

 

[FleshJeb]

11 hours ago, MatterBeam said:

This isn't really something they'll teach you unless you're an aerospace student. I just learnt it from reading up on hypersonic spaceplane designs. 

I see your point now about storing sunlight. Thankfully it will not be necessary. You can absorb sunlight effectively with a black surface, and discharge the heat through a shadowed surface, or a surface angled edge-on to the sunlight. You can capture the energy across the temperature differential from the heated to cooled sides.

The equilibrium temperature in direct sunlight is about 394K. A surface in perfect shadow will cool down to 3K. Across this temperature difference, you can run a heat engine continuously while in sunlight, and on residual heat over the night side of the planet. 

Radiatively cooled liquefaction of the collected gasses is viable option, but it will require massive surface areas to remove heat as the temperature drops to cryogenic levels. At 394K, you are radiating 1300kW/m^2. At 77K, it drops down to 64W/m^2. To bring nitrogen down below boiling point, you need even lower temperatures. This implies very heavy equipment due to their sheer size, when a relatively lightweight heat pump can do the same thing in a compact structure.

I WAS an aerospace student. I quit after my second year and went back to being a Land Surveyor. Thanks for the info.

From the article I linked in my first post: "However, the typical range of temperatures was found to be from -170˚C [103K] to 123˚C for LEO satellites while -250˚C [23K] to 300˚C could be experienced in other orbits." It may be interesting to plug these numbers into your modeling.

My advocacy for radiators is also based on reducing the number of moving parts. Operations and Maintenance costs are going to be a major factor in viability. I understand that's not what the thread is about, but reducing points of failure is well-worth considering in the space environment. You could also spin your craft about the long axis to introduce a "thermal pulse rate", and reduce the need for moving parts. This might be a way to keep the scoop unfurled, and reduce the mass of its structural components. Difficult to do solar power, though.

After a few minutes on Wikipedia, it appears that radiative efficiency scales with (deltaT)^4, while heat engine efficiency has a linear relationship with deltaT.

I bet you could design a radiative precooler that does better than the 400W/kg you quoted in your first post. You could save mass on the refrigerator, even with the efficiency loss. In addition, you'd save mass in power generation.

Another question just popped into my head: What's the refresh rate on our gaseous resource? Could we eventually impact the density of the Thermosphere? Wikipedia says that the Thermosphere above 85km is 0.002% of the 5*10^18 kg of Earth's atmosphere. So, probably not, but it's always good to check the environmental concerns.

Thanks for the thread. Yours are always good for intellectual play.

 

[KSK]

OK, first things first - I love the concept. I grew up playing Elite and one of the upgrades to your spaceship in that game was a fuel scoop. The idea of real life fuel scoops tickles every sci-fi nerd bone in my body.

However, even if we assume @MatterBeam's numbers add up, the practicalities of deploying a large enough scoop to generate useful amounts of propellant per year seem daunting. The few experiments that have been carried out with tethers for example, have been challenging and a tether looks positively straightforward compared to deploying and controlling several hundred square metres of scoop.

This also strikes me as a scaling problem waiting to happen. For example, given the size of the structure involved, even small increases in mass per square meter of scoop material are going to add up quickly, to the point where they knock the rest of your calculations out of kilter.

We've also got a lot of demanding material requirements here. The scoop needs to be able to withstand intense UV radiation and reactive oxygen species for years at a time whilst remaining very lightweight and being capable of being folded up into a rocket fairing and then smoothly deployed in space.

I do also wonder whether this is a solution in search of a problem. Having thousands of tons of scooped propellant on-orbit sounds great but by the time we're generating sufficient traffic in space to need that propellant, I'm thinking the improvements in launch vehicle technology to lift that traffic to orbit are going to make hauling propellant up from Earth a very competitive option to set against a technically finicky atmospheric scoop.

Edited September 14, 2017 by KSK