> > There are no lower bounds for the cost of terrestrial nuclear
> > power. Add automation and scale, and the cost modulates toward
> > zero.
> How can you seriously argue that there exist no lower bounds for
> the cost of terrestrial nuclear power?

The fuel supply is vast -- 40 trillion tons of uranium alone.
The fuel is energy-dense -- $100/bbl oil = $1 million/kg uranium.
The fuel can be mined robotically.
The fuel can be processed and assembled robotically.
The reactor units can be built, fueled, and run robotically.
The spent reactors and spent fuel are dense and contained and
therefore easy and cheap to deal with. Simply mothball them
(today, this is called SafStor for the reactors, and
onsite-storage for the spent fuel). The radioactive-flux
continuously decays. Someday, a team of robots might come along
and recycle everything. If not, there would be no loss.
http://www.google.com/search?q=SafStor+nuclear

> The Council of Foreign Relations looked at the issue of uranium
> supplies:

There has recently been turbulence in the uranium market. The CFR
article explained why. The present spot-market price is $60/lb,
down from the CFR article's quoted price of $138/lb some nine
months ago.
http://www.uxc.com/review/uxc_Prices.aspx

The long-term view is an approach toward zero-cost for uranium,
as long as demand continues to increase. Uranium and thorium
together add up to some half to one trillion years of fuel
supply, at current nuclear-electricity generation rates. The
math is pretty simply, if we allow for some rounding errors.

Current uranium consumption is 65,000 tons per year.
http://www.world-nuclear.org/info/reactors.html

Current economically-minable-with-current-technology (without
breeder reactors) uranium totals some 40 trillion tons.
http://nuclearinfo.net/Nuclearpower/UraniuamDistribution

Richard Garwin estimates that breeder reactors would extend
supplies 60-fold. That would bring consumption down to 1,000
tons per year. 40 trillion tons / 1,000 tons = a 40 billion
year fuel supply. Add 160 trillion tons of thorium, and the
fuel supply becomes 200 billion years. Double the current
nuclear-electric thermodynamic efficiency of 34% to 68%, and
the fuel supply doubles to 400 billion years. Nudge up the
efficiency another 20% to 82%, and the fuel supply becomes a
nice, even half-trillion years.

If we use, instead of Richard Garwin's 60-fold breeder-induced
fuel-supply extension figure, Bill Hannahan's...
http://www.nuclearcoal.com/energy_facts.htm

...120-fold figure, the Earth's fission-fuel supply comes to
one trillion years worth.

It would not actually last that long, of course, since the
half-lives of the fuel isotopes are:

uranium-235 700 million years
uranium-238 4.5 billion years
thorium-232 14.1 billion years

Even if the thorium supply were tapped for most of, say, 20
billion years, the earth would likely have to be moved to
make way for an expanding sun.

In reality, breeder reactors will probably not be "needed" for
many hundreds or even thousands of years, even assuming
continuous upward demand for fuel. The breakeven uranium price
is currently thought to be some $1000/kg of mined uranium.
Below that price, breeder reactors do not make economic sense.
Even the current artificially-inflated spot-market price of
$60/lb = $132/kg is not nearly high enough to justify breeder
reactors.

http://www.cfr.org/publication/14705/global_uranium_supply_and_demand.
html

Thanks for the link.

As to why all mineral prices ultimately trend toward zero,
there are some theory explanations here:
http://www.juliansimon.com/writings/Ultimate_Resource

> The fuel supply is vast -- 40 trillion tons of uranium alone.

You're making it quite clear that you're a very binary type of fellow
who has a distaste for nuance and analysis. You zoom right in on the
40 trillion tonnes and ignore information on ppm rates. So, a vein of
ore at 10,000 ppm is no more or no less useful to miners than uranium
buried in the oceanic crust at 0.1 ppm.

> The fuel can be mined robotically.

Prove it.

> The fuel can be processed and assembled robotically.

Prove it.

> The reactor units can be built, fueled, and run robotically.

Prove it.

> The spent reactors and spent fuel are dense and contained and
> therefore easy and cheap to deal with. Simply mothball them
> (today, this is called SafStor for the reactors, and
> onsite-storage for the spent fuel). The radioactive-flux
> continuously decays. Someday, a team of robots might come along
> and recycle everything. If not, there would be no loss.

Securing multiple storage sites comprised of a large assortment of
radioactive debris is a security nightmare than increases the sources
for dirty bomb material.

Your habit of simply asserting that your position is cheap to
implement is no more than bias reinforcement than does nothing to
advance your case with people who don't subscribe to your bias. So too
is your habit of asserting that robots can perform economic and mining
miracles but that their jou-jou can only be applied to favor your
fantasy but not other fantasies. Look, if magic robots can make
nuclear power cost free then why can't this deus ex machina work to
make geothermal power or solar satellites or dark matter reactors
perform to the same standard?

> The long-term view is an approach toward zero-cost for uranium,
> as long as demand continues to increase. Uranium and thorium
> together add up to some half to one trillion years of fuel
> supply, at current nuclear-electricity generation rates. The
> math is pretty simply, if we allow for some rounding errors.

Please inform us on how much energy it will take to strip mine Mt.
Everest, all of the himilayas, filter every liter of the Earth's
oceans, etc so that the magic robots can gather the meager uranium
content locked within?

> Richard Garwin estimates that breeder reactors would extend
> supplies 60-fold. That would bring consumption down to 1,000
> tons per year. 40 trillion tons / 1,000 tons = a 40 billion
> year fuel supply. Add 160 trillion tons of thorium, and the
> fuel supply becomes 200 billion years. Double the current
> nuclear-electric thermodynamic efficiency of 34% to 68%, and
> the fuel supply doubles to 400 billion years. Nudge up the
> efficiency another 20% to 82%, and the fuel supply becomes a
> nice, even half-trillion years.

Wow, you sure are fast and loose with improving the efficiency gains
of reactor processes, so if you want to argue in the realm of fantasy
why do you stop at 82% efficiency, why not go for 600% efficiency? Why
bother observing physical limitations when you're conjuring up magical
efficiency gains, strip mining the Earth, filtering every liter of
ocean water and doing this with magical robots which perform these
tasks without consuming energy?

> Even if the thorium supply were tapped for most of, say, 20
> billion years, the earth would likely have to be moved to
> make way for an expanding sun.

You argue that the Earth would likely have to be moved to make way for
an expanding sun, but you don't provide any analysis for how you
determine the likelihood. Could you please share with us your analysis
of this point.

Argument from fantasy which is bolstered by the misuse of data is
wholly unconvincing.