[ExI] Serious topic

[Damien Sullivan]

On Mon, Feb 28, 2011 at 03:15:09PM +0100, Eugen Leitl wrote:
> On Sun, Feb 27, 2011 at 07:06:00PM -0800, spike wrote:
> 
> > Keith, I like to imagine the kinds of transitions that can be made
> > quickly if necessary, should these kinds of scenarios play out.  We
> > have three areas in which energy use can be reduced: home lighting
> > and heating, food, and
> 
> Home lighting takes electricity, and is the easiest to fix, though
> many lack the means of buying solid state or metal halide lightings,
> nevermind refitting their home electric infrastructure.
> 
> Changing heating is far more expensive, and to what? I heat with
> locally sourced wood from renewably managed forests/my other place
> is deep geothermal, but that's not an option for many, especially 
> overnight.

I long thought heating was the big barrier, since burning fuel is a very
efficient way of making heat.  Electricity turns into heat all too
readily, but making the electricity... OTOH, heat pumps!  Instead of
turning a joule of electricity into a joule of heat, you can do that
while bringing in 3 joules of heat from outside.  This kind of makes up
for the fact that you're using electricity for heat.

> In general people don't seem to see what these >15 TW total mean, and what 
> doubling and tripling electrification to substitute for missing
> fossil liquids and gases mean (1 TW/year conversion rate, for
> the next 20 years, and photovoltaic surface doesn't fabricate
> itself, put itself up, and connects to the grid, while rebuilding
> it in the process, and adding energy buffering capacity).

OTOH, looking at just the US, we're talking 3TW of power use.  Say 1/3
electricity from coal, 1/3 gasoline for cars, 1/3 gas for heat.  (This
is a caricature of reality, but I think good enough for now.)  We'd want
to build nukes to replace the coal, more nukes to provide equivalent
transportation capacity but not necessarily in the same manner (electric
trains and trolleys, vs synthesizing gasoline from air), and more nukes
to drive heat pumps.  We're talking a TW of gasoline and a TW of gas...
but the gasoline is doing more like 200 GW of actual work, because IC
engines suck.  And per above, we'd need says 300 GW of electricity to
run the heat pumps for 1 TW of heating.

Oh, and the coal is 1 TW of coal that turns into 300 GW of electricity.

So summing up we need 800, call it 900, GW of new nuclear plants.  Cost
estimates of nukes vary very widely, AFAICT, depending in large part on
the interest rate -- which means this is something governments are
better at, simply because they can borrow more cheaply.  (Even more true
of renewables.  Anything capital heavy has a bias to being done by he
who has the cheapest capital.)  At the low end, $1.5/watt, at the high,
$5 or more.  So $1.35 trillion to $4.5 trillion.  Over 20 years, say,
$60 billion/year to $225 billion/year, out of $14,000 billion GDP/year.

I call that affordable.

Mind you, this doesn't account for the oil to fertilizer cycle, or the
costs of rebuilding public transport.  (Then again, if we were buying
fewer cars, there'd be idle capacity to be retooled for trains and
rails.)

-xx- Damien X-)