[elpachuco]

>>Lest you think I’m taking sides already, let me stress that it’s not at all clear that AC is actually better than DC

No, it has been settled for almost a hundred years that AC, for long distance transmission, is the best. Too bad he included something like this. It diminishes his credibility and colors the entire article.

Edit: So I seem to be quite wrong. Well, at least now I know. I stand corrected.

[kwantam]

No it hasn't.

AC transmission is nice for a few reasons, but the biggest reason for its use historically historically has been that you can step down from high to low voltage with a passive element (a transformer).

(Why step up and down? You want to transmit at the highest voltage you can to reduce conduction losses (higher V -> lower I for the same P, and thus lower I^2 * R losses).)

But long distance AC transmission is ugly in a bunch of ways. For one, over very long distances a power line becomes a reasonably efficient 60 Hz antenna, which means radiated losses in addition to conduction losses. For another, AC generators want to produce 3-phase power; transmitting this directly means more wire, balancing phases, et cetera. Further, AC transmission suffers additional conduction losses when reactively loaded because of circulating currents (this is why people care about "power factor"). In practice, the complexity of managing the reactive loading of HVAC transmision wires is substantial: transformer nonidealities turn a resistive load on the secondary into a slightly reactive load seen at the primary, which introduces circulating losses. And let's not forget that an AC grid requires that all the generators feeding the grid are precisely synchronized, while a DC grid has no such synchronization issues, which simplifies the task of bringing power plants up and down.

By comparison, DC transmission requires complexity at the sender and receiver: generation is for the most part a fundamentally AC process (i.e., electrical generators), and of course homes are built for AC power, so the receiver must switch DC back to AC. But as power devices become better and cheaper, and the power grid's load increases (without commensurate increase in copper to handle higher currents), long distance DC transmission is becoming more commonplace. (And since modern AC transmission requires some power factor management anyway, the choice really isn't between "dumb transformer" and "massive inverter" anyhow.)

For example, when I lived in Boston it received a substantial amount of its power as HVDC from somewhere in Canada. Three big (like, BIG) inverting stations around town were responsible for turning this DC into AC for local transmission. This setup ostensibly saved substantially on conducted and radiated transmission losses, according to the folks who gave us a tour of one of the inversion stations.

Wikipedia has more on this topic. http://en.wikipedia.org/wiki/High-voltage_direct_current

[tormeh]

DC is better over long distances than AC. The problem with DC is that generators and many engines operate on AC and that was more important at the beginning of electricity use.

[wlesieutre]

The other big caveat being that for both AC and DC, you want to do long distance transmission at very high voltages. DC transmission has been put to real world use (converted back to AC into the grid), and it tends to be on the order of 100,000 to 1,000,000 volts.

Historically, AC was much more feasible to do this with because you can change the voltage with transformers, which are basically paired coils of copper wire. These days we have the capability to convert between AC and high voltage DC, but it's still requires comparatively complicated equipment and isn't as reliable.