Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: power-grid physics



On 08/22/2003 03:43 PM, Frohne, Vickie wrote:

> Edison was the DC guy, and his company actually did put generating
> stations on nearly every city block. Tesla pioneered AC at Niagara
> Falls, but was financially ruined by Edison. (Lots of facinating
> history omitted)

Yup. The Edison versus Westinghouse wars are
also fascinating.

> Probably, because a DC setup requires a generating station on every
> city block, making a county-wide distribution grid very unlikely if
> not impossible. With DC, transmission line power losses are too high
> for long distances. I'll have to recheck my 1908 engineering book for
> exact numbers, but I think 300-500 feet from the generator was about
> the maximum,

A lot has changed since 1908. At any given voltage,
DC has always been *more* efficient than AC. In 1908
they could deal with high-voltage AC more easily than
high-voltage DC.

There are still many arguments in favor of AC. They
include
-- transformers to connect juice to/from the
high-voltage low(er)-current long-distance feeders.
-- saturable transformers (one for every house
or two) to regulate the voltage customers see, in
the face of greatly changing loads.
-- zero-point switching: close a switch when
there's no voltage across it, open a switch when
there's no current through it.

The last is nothing to be sneezed at; the big
circuit breakers at a substation are synchronous
so as to minimize the arc when they open. These
things are a highly amusing combination of cleverness
and brute force: they have a syncrhonous motor
that stays in phase with the current. The motor
drives a disk with a dog; the dog is normally
retracted, but when it comes time to trip the
breaker the dog sticks out and knocks out the
three bars that carry the current, bing, bing,
bing. A view looking down the axis is:

Z

O----
Y X

where the dog is about to whack the X phase. And
even so it takes lots of cleverness to extinguish
the bit of arc that occurs anyway.

I suspect that at the time the system was built,
they really had no choice but to go with AC.

OTOH if we were going to redo it from scratch,
we might do a lot of things differently.

Already some long-distance tie-lines are high-voltage
DC. I'm not 100% sure what they use for getting
power on/off these lines, but I imagine it's
a big piece of rotating machinery plus some
tricky and large bits of electronics.

The existence of semiconductors changes things
quite a lot. But this isn't a panacea; you
don't want your semiconductors to be blown out
by a lightning strike.

The current practice of making the system
synchronous over wide areas, and using phase
and frequency excursions for load-balancing
and load-regulation ... it boggles the mind.
I'm not surprised that the system is flaky;
it's a miracle that it works at all.

I don't think the question is going to be settled
by appeal to basic physics principles; I think
there are lots of devilish details that will
require detailed engineering work.