[ExI] solar is looking better all the time: was RE: Efficiency of wind power

[Eugen Leitl]

On Tue, Apr 19, 2011 at 11:33:28AM +0930, Emlyn wrote:

> Well, yeah. Inverters, mentioned earlier, are only necessary if you're
> going to use AC. Inside a dwelling or business, if you convert to DC

I like DC a lot, but due to infrastructure lock-in changing it
will be tough. Due to ohmic losses you want to go to a higher
voltage, e.g. the new automotive 42 V 
http://en.wikipedia.org/wiki/42-volt_electrical_system
which is still reasonably safe (or 110 or 230 V DC to
make use of existing wiring while increasing shock hazard) 
and higher DC still if you want to go residential microgrid.

In general I'm nervous about lots of galvanically linked
power electronics which default failure mode is expensive
smoke. Black swans are rare, but you can count on them landing
eventually. E.g. next solar superstorm would tough
to harden against.

I would be much, much happer about using hydrogen or
methane for galvanic separation of invididual installations.

> throughout (9V?), solar becomes a lot simpler. Appliances/electronics
> get much easier (just chuck the power pack and plug it straight in,
> they'll disappear eventually). Washing machines and such are tougher,
> but solvable.

There are DC motors and there are smart DC motors which use
power electronics driven by microcontrollers.
 
> So then it (solar) looks like a technology which will make the most
> sense on buildings to power themselves, to begin with, with industrial

I agree.

> scale stuff on the old system for a while. The grid then needs to turn

Right now the industry does profit from grid-tied small scale
installations. Most of these 17 GW peak are residential small
plants.

> inside out; rather than a system for broadcasting mass power from
> central places, and have conversion losses at the point of

In a network of sources and sinks most of the power doesn't
travel far. As power is fungible, you can think of it as
power peering. It doesn't matter for a network segment if 
ingress and egress are roughly balanced. We already have
realtime power markets, the interesting part is making the
markets sufficiently small scale that individual households
and even appliances can participate in bids. This is easy,
as even small embeddeds can carry the full stack and enough
CPU power for running bidding algorithms. You can have a large
local buffer using http://en.wikipedia.org/wiki/Vanadium_redox_battery
trading cheap peak juice with expensive nightime juice, 
load-leveling in the process.

> consumption, you need a system for sharing locally produced power,
> with conversion losses in the grid here and there where you want to
> bump the power up to a distance transmittable form; eg: a DC local
> network, with suburban subs providing inverters to jump it up to high
> voltage AC for shipping elsewhere, where appropriate.

Already long-distance are HVDC. I'm not sure you can go up
from 100 V to high kV and MV without some losses. You might
have to take an electrochemical converter road.
 
> I guess what'll happen is we'll have both grids for a while - the high
> voltage "broadcast" one and an emerging inside-out local power DC
> network, which can be a lot more ad hoc because it's just about
> shipping around surplus where it exists.

There is some work being done (unfortunately, kraut)
for deal with load difference for large scale
http://www.heise.de/tp/artikel/34/34475/1.html

-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
______________________________________________________________
ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE