[ExI] Skylon as first stage.

Keith Henson hkeithhenson at gmail.com
Thu Apr 21 15:25:47 UTC 2011


On Thu, Apr 21, 2011 at 3:44 AM, Eugen Leitl <eugen at leitl.org> wrote:
> On Wed, Apr 20, 2011 at 09:46:34AM -0700, Keith Henson wrote:
>
>> I looked at this in some detail from the perspective of a moving cable
>> (loop) space elevator.  That's the gold standard for lifting stuff to
>> GEO, being slightly more than 100% efficient.  (Ask if you can't
>> figure out why.)
>
> This is all very well, but we don't have such elevators, and
> the number of launches required to build such an elevator, even
> if we had theoretical-strength SWNT ribbons (which we don't)
> would seem enough to wreck major havoc on this planet (look
> at the environmental footprint of the Shuttle, and Russians
> are only slightly better).

_If_ we had the material (which we don't) you only have to lift a
small thread to state the project.  Then you lift larger and larger
cables, doubling the capacity of the cable about 3 times a year.

"In the background was the vibration of the 31-foot diameter driver
wheels turning at 900 rpm, the whip-cracking sound of the supersonic
space elevator cable, reminding Marc of a flag flapping in a strong
wind, and the occasional run down and run up of the variable speed
cable.  The elevator was mostly lifting parts now, but over its life,
more than ninety percent of the capacity of the elevator had raised
more cable and counterweight.  Four more doublings would take it from
its current capacity of 125 tons a day to its design capacity of 2,000
tons per day."

(From UpLift, the 4th chapter in the saga that includes "the clinic seed.")

> Meanwhile, on the Moon you only need commercial aramide.

Spectra.  Density is very important when designing space elevators.

>> Takes (round numbers) 15 kWh per kg.  Reasonable number for power sat
>> mass is 5 kg/kW.  A good way to look at energy return on energy
>> invested is how long it takes to get a payback.  For ground solar or
>> wind it's measured in years.
>
> EPBT for current CIGS or CdTe is under a year (EROEI is 40:1 at the
> moment). The trend is is of course towards better, though it will
> turn asymptotic at some point.
>
>> For a power satellite made with stuff brought up by elevator, 5 kg
>> will take 75 kWh to lift it.  The 5 kg makes 1 kw, so the payback time
>> is just over *3 days.*
>>
>> Chemical rockets are around 2.5% efficient so the payback time is 40
>> times that long or about 120 days.
>
> Chemical rockets currently means kerosene/LOX or maybe liquid
> methane/LOX or even liquid hydrogen/LOX. Synfuels will screw with
> your energy, and e.g. you waste half of energy in liquid hydrogen
> to liquify it.

Takes around 50 kWh/kg to make H2 from water and 20 kWh/kg to liquefy it.

> Including rocket construction, supply chain, the
> EPBT for chemical rocket-launched SPS is probably never. EROEI
> needs to be better than 5:1 to bother,

For oil, coal, oil sands, etc, EROEI is a good metric.  It's not so on
sun or wind renewable energy.  There, the more useful metric is time
to repay the energy invested in energy out.  For single use rockets
you may be right, we could work it out given that rockets are mostly
aluminum which takes considerable energy to reduce.  But reuseable
rockets are a different story because the energy cost of the materials
is spread over hundreds to thousands of flights, and you can recycle
the worn out ones back into metals.

> so I think chemical rocket
> SPS is stone cold dead. The only reason you want to do it is
> to supply global wireless power for military applications, where
> prices are less relevant.

I have tried to make a case for military SBSP based on current or even
projected rockets and just can't do it.

>> For the laser part, it draws around a GW to send 60 t/h to GEO.
>> (Starting from a sub orbital boost by the Skylon.)  1 M kW/60,000kg is
>> 17 kWh/kg
>
> Yes, but Skylons don't yet exist. It is not obvious we can
> make scramjets to work, though of course I hope we will.

The Skylon is a lot closer to existing than a lot of stuff we talk
about on this list.

>> The Skylon phase burns 66807 kg of hydrogen per launch.  The energy
>> content for three per hour would be 14029470 kWh (at 70 kW/kg), or 233
>> kWh/kg.
>>
>> Together, 250 kWh/kg, (6% efficient) so material for a kW of
>> production would take 1250 kWh to lift--which gives an energy payback
>> time of around 52 days.
>>
>> By renewable energy standards, that's amazingly good.
>
> I think we'll be at EROEI of 100:1 and payback times of few
> months for terrestrial PV within 20 years or less.

How are you calculating EROEI for wind or solar?

> I think that
> would be pretty good. Of course, long-term is self-replicating
> machine-phase photovoltaics both on Earth and in space.

And you complain Skylon doesn't exist yet?

>> See any problems with the logic or math?
>>
>> > When we're looking at kW/kg, we shouldn't forget that the
>> > kg is at ~Mach 25, and it came from the bottom of the gravity
>> > well which taxes your EROEI -- unless it came from lunar material.
>>
>> Later in power sat production it's worth going after lunar material
>> just so you don't need to be flying so often.  But the way to go after
>> lunar material is with a moving cable elevator out through L1.  It
>
> I think we'll first get chemical rockets from in-situ synfuels
> from lunar cryotrap water, and then maglev launches. The advantage
> of maglev is that it scales, and can produce effectively continous
> stream of material, and is self-amplifying since you can beam
> down more power from Earth-Moon space SPS as you run out of
> available flux.

Hmm.  Do you have any idea of how hard maglev is?  Do you realize
there is just one place on the moon where maglev makes any sense?
Ever heard of achromatic orbits?  Google achromatic orbit heppenheimer

Any idea of the difficulty of catching a stream of packages of lunar
dirt with the mass and velocity of an 18th century cannon ball?  Don't
forget they have shotgun pattern and some will flat out miss the
"catcher."  A *lot* of thought went into mass drivers back in the 70s.
 It wasn't easy then and it isn't now.

>> takes an investment of around 100,000 tons and pays back the
>> investment at around a 1000 tons per day (pays back in mass in 100
>> days).
>>
>> But you really need big lift capacity from the earth before you go
>> after lunar materials.
>
> The whole point of boostrap with ISRU is that you need minimal
> amount of material until you go near self-rep closure of unity,
> and none after you're above. The ultimate free lunch, long-term.

With full blown nanotechnology a coke can sized payload is all you
need to convert the moon to whatever you want.  But for the current
state of the art, the only study I know about in recent years came in
at $2 T and at least 20 years.

If you can put numbers on how to do it and support them, I am eager to listen.

Keith




More information about the extropy-chat mailing list