[ExI] Asteroidal mining was Nukes was less expensive > energy

[Keith Henson]

On Sat, Sep 24, 2011 at 5:00 AM,  Adrian Tymes <atymes at gmail.com> wrote:

>
> On Tue, Sep 20, 2011 at 4:35 PM, Keith Henson <hkeithhenson at gmail.com> wrote:
>> On Tue, Sep 20, 2011 at 3:24 PM, ?Adrian Tymes <atymes at gmail.com> wrote:
>>> On Mon, Sep 19, 2011 at 7:56 PM, Keith Henson <hkeithhenson at gmail.com> wrote:
>>>> Have you ever worked the engineering numbers? ?Dr. Eric Drexler and I
>>>> have done some of them, as has Dr John Lewis of the University of
>>>> Arizona.
>>>
>>> Care to share them, then?
>>
>> Sure. ?Getting rid of waste heat by radiation is one of the
>> fundamental problems in space.
>
>> Moving, shaping metal is another one.
>
> ...that's not the numbers.  That's just citing what the problem is.

Did you read the references?  They are full of design detail, scaling
laws, specific masses, etc.

# Henson, H.K., and K.E. Drexler: Vapor-phase Fabrication of Massive
Structures in Space, Space Manufacturing AIAA 1977
# Henson, H.K., and K.E. Drexler: Gas Entrained Solids: A Heat
Transfer Fluid for Use in Space Space Manufacturing AIAA 1979

The second one was reprinted in the L5 News.  The take home number
from the first one is that a metal vaporizer can move its own mass in
metal in less than a day.
>
>> 50,000 tons was based on best estimate of how long it would take a
>> plant to make its own mass in product and the demand for nickel (1000
>> tons per day) for a relatively mature power sat industry.
>
> Thing is, we do not (yet) have a relatively mature power sat industry.
> Discussion of mining is being done in today's context - in which there
> only seems to be one destination for the goods: down the gravity well
> on Earth.
>
>> 5-10 GW of
>> power was based on the energy to melt, roll into thin ribbon, move it
>> into 750 psi warm CO and sort out the carbonals.
>
> Where do you get these numbers from?

Dr. Lewis.  But it isn't hard to work out.  1000 t/day of NI takes
processing 10,000 tons of asteroid (at 10% Ni).  That's 10 million
kg/day or 400,000 kg/h.  To heat and melt iron takes 54 kJ/mol.  There
are 17.9 mol in a kg so it takes around 1000 kJ/kg to heat and melt a
kg or around 1/4th a kWh/kg.  So just the first step would be drawing
100,000 kW or 0.1 GW.  What with radiation loses (high at this temp)
and the rest of the process, a good fraction of a GW is a reasonable
estimate.  So a 5 GW power sat might be a little oversized, but then
you need to consider that 1986 DA swings out beyond Mars.  That cuts
the output down by a factor of 4, so one rated at 5 GW in earth orbit
would not be oversized after all.

>> Besides, you want
>> to just use a power satellite off the production line. ?The power sat
>> alone is 25,000 tons.
>
> The second statement is why you don't want to just use a power satellite
> off the production line.

At 1000t/day, the asteroid processor repays its mass cost in 50 days.
The nickel (or Invar) it makes is worth at least the $100/kg transport
cost up to GEO.  Or around 37 B/year.  Assuming, of course, that you
are making power satellites in the TW/year numbers.

>> Maybe. ?It's not really a mine as much as a smelter and a good sized
>> one at that. ?I figured 100 people actually working in the processing
>> plant and the rest being support and dependents. ?These people are
>> isolated, with transit times as bad as to Mars. ?They have no choice
>> about growing at least some of their food.
>
> 1) Oil platform style operations.  No dependents.  Rotate people.

That would depend on some really high delta V transport.  Minimum
energy orbits take years, not 30 minutes to fly out to the platform.

> (Yes,
> this absolutely requires a drop in the cost per kg to launch.  OTOH, the
> kind of traffic this would generate would, itself, cause much of that drop.)

For power satellites to make sense at all the cost to GEO has to get
down to $100/kg or less.

> 2) 100 people to work the processing plant?  Stuff like that can be fully
> automated - and even if light speed lag prevents that, automation can
> still cut that to less than 10, maybe no more than 5, mainly as
> troubleshooters.

Perhaps.  I have worked in oil refineries, which have about the lowest
ratio of people to product in all of industry and the most automation.

They still have several hundred people in a typical refinery.

>> You might be correct that that is a better approach. ?I don't think it
>> would work with the one I analyzed,
>> http://en.wikipedia.org/wiki/1986_DA though.
>
> That would be one of the larger asteroids.  Save it for later, after you have
> a robust asteroid moving capability.  Start with (much) smaller rocks.

Why?  The point is to make the demand for product.  The processing
plant gets sized for that rather than the size of the asteroid.

>>?Of course if anything jams up the automation, you will
>> lose a couple of years of production before a human can be on site.
>
> Unless you have redundant automation.  Say, some of those Personal
> Satellite Assistants, deliberately isolated from the main automation so
> they can fix stuff when the main automation goes FUBAR.

You have more trust in automation than I do.  I once saw a process
upset in an oil refinery cascade into sinking the floating roof on a
tank full of 700 deg F thick oil.

>> What I have done on this topic is just slightly beyond the "back of
>> the envelope" stage while matching the scale of the project to the
>> demand for Invar for a modest 200 GW per year power sat project.
>> Perhaps I am too hard on people to ask that they back up "good ideas"
>> with a little physics and chemistry.
>
> Then let's talk physics.
>
> You ran the numbers for 1986 DA?  Let's take a rock 1/1,000,000th
> of it.  That's about 2*10^7 kg, and 23 meters across.
>
> If a 50,000 ton processing plant can do 1986 DA in acceptable time,

Estimated mass 2 x 10^10 tonnes/(10,000 t/d x 356 day/yr) or about 5,500 years.

I picked it because it was the best known mostly metal composition,
not because I could use it up in a few decades.

> then a 50 ton plant can do this in acceptable time.  That's a handful
> of people, for the entire station.  (Decreased operational requirements
> aside, good luck fitting 500 people plus life support in 50 tons.)  Of
> course, park it in lunar orbit and you don't even have to have people
> permanently on site.
>
> Assuming a similar distribution to 1986 DA - that's a trillion dollars
> for the platinum, plus relatively ignorable amounts of other metals.
> Scaling that down, that's a million for this mini-rock.
>
> Now, here's the thing.  Having launched all that mass, if you're
> bringing the small rock into lunar orbit or some other fixed location,

Lots of luck trying to get permission to move big chunks anywhere near
earth orbit.

> you can use iron slag & solar energy (if, say, you're using an ion
> engine: that's ionizables & electricity) to refuel the rock mover, and
> bring in another one while you're processing the first, without having
> to launch anything more to move it.  So you get another million.
> And another, and another.  Sure, the price of platinum eventually
> tapers off - but you can do a 50 ton mission for tens of millions
> these days.

Falcon Heavy is 50 tons to LEO for $100 M IIRC.  Half that to GEO.
Typical one way missions to asteroids are around 7 km/s beyond LEO (or
was that GEO?).  You won't have a lot of choice since the metal type
asteroids are rare and I have dibs on `986 DA.  :-)

> It won't taper off fast enough to prevent you from
> recovering investment plus interest.
>
> Also, smaller bodies have an advantage in radiating heat: more
> surface area per mass.  One could do (far) worse than to simply
> ape the ISS, straight up.  That's certainly not cooking its crew,
> and I believe the ISS is somewhere around 500 tons at this time.
>
> Ore refining can be done in graphite centrifuges, with large lenses
> to focus solar radiation in.  Melt the ore in contact with the walls,
> spin it up, spin-gravity-sort the output.  Granted, we're talking big
> lenses, tiny centrifuges, but it gets the job done.
>
> How's that for starters?

Reflectors are much lighter than lenses.

Keith