Tesla vs Edison

[kevm14]

I just watched a 2 hour special on PBS on Edison, part of a series called American Experience.

http://www.pbs.org/wgbh/americanexperie ... ms/edison/

Side note, they did not mention Tesla once. It seems Edison vs Tesla is more of a revisionist history competition. The real AC/DC argument was between Edison and Westinghouse. So please excuse the attention-getting headline. Also I did not realize that Edison General Electric became General Electric with the merging of the Thomas-Houston Electric Company.

Fun fact on AC vs DC for transmission lines (more complex and contrary to what we learned in middle school science class):

http://theenergycollective.com/rogerret ... powerlines

Takeaways:
- AC has a major practical advantage of being easy to step up and down with transformers, which may be the #1 reason for its use in transmission lines
- Skin effect begins to add transmission resistance with AC and this does not affect DC at all (large diameter conductors)
- At high voltages, dielectric losses begin to enter the picture with AC but not with DC (RLC). Stated more accurately, the voltage limit for dielectric break over is much higher with DC. Perhaps more importantly, it's basically binary with DC whereas with AC losses increase as voltage increases. So since DC can use higher voltages, that translates to more power transmission.
- DC has another advantage: no need to synchronize the phases when coupling plants
- If you believe in the issues caused by the low frequency magnetic fields emitted by transmission lines (there are some documented medical effects), then DC completely eliminates those
- Solar power is natively DC.....
- Let me be clear: DC has LESS loss over distance than AC when you factor in the above characteristics. Mind = blown?

I only mention this because in some ways, the Tesla vs Edison thing (which was not really a thing anyway) is always ended with “well, AC is the superior transmission line power type, therefore Tesla wins because his was better.” That’s a serious misrepresentation of the underlying technologies as well as history itself. I think the big reason Edison lost in the electric race was not so much the AC vs DC thing, but the architecture of his power grid. Westinghouse (and other competitors) understood the advantages of having a single and modular power generating station whereas Edison (admittedly because he started literally from ground zero) had dynamo plants spaced evenly across the grid. Even in basements of buildings. He also lost because his ego was too big to really take good ideas from others and continue to evolve them.

But I do find it amusing that, aside from the transformer thing, DC is actually "superior" for long distance power transmission. But wait, you say, I learned in science class that AC goes for far longer distances without loss than DC. So, that's wrong. But I know why that rumor started. Edison's power grid was DC, yes, but much more importantly, was LOW VOLTAGE. Somehow, the characteristic of his low voltage became conflated with being DC power. That low voltage is probably why he had to have multiple dynamo plants per block. Coincidentally, being DC also supported having multiple generating stations in the first place because of the ease of coupling them. In that regard, the false conflation is understandable but still wrong.

Isn't that interesting?

Going back to the special I watched, I would say that Edison absolutely does deserve the notoriety that he has garnered through his life and posthumously. Why? He wasn't just a great inventor (by the way, he definitely had his limits - frankly he didn't really understand AC power), but the most important thing he invented was not any one system, machine or object, but the PROCESS of modern invention. The idea of taking a brilliant thought through experimentation, development, testing and eventually to market, with great success, was really his great contribution.

[kevm14]

Fun fact:

Con Edison’s electric business also dates back to 1882, when Thomas Edison’s Edison Illuminating Company of New York began supplying electricity to 59 customers in a square-mile area in lower Manhattan.

Con Ed is short for Consolidated Edison, and his start in Manhattan is exactly related to the name of the major electrical supplier in that area today. However, Con-Ed deals in not just electricity, but gas, steam and other non-utility revenues.

[kevm14]

Another fun fact:
http://en.wikipedia.org/wiki/Arc_lamp

These were in actual urban use before the incandescent bulb.

Also interesting:

Arc lamps were superseded by filament lamps in most roles, remaining in only certain niche applications such as cinema projection, followspots, and searchlights. Even in these applications conventional carbon arc lamps are being pushed into obsolescence by xenon arc lamps.

[kevm14]

http://theenergycollective.com/rogerret ... powerlines

This link also does a good job with explaining the virtues of the long distance DC power transmission line:
http://en.wikipedia.org/wiki/High-volta ... ct_current

[kevm14]

Related: http://www.technologyreview.com/news/42 ... -dc-power/

Fun prognostication: In a future of electric cars, I predict a shift to DC power transmission on our grid. That is, if this EV future requires very high current charging at home.

Feasibility detail: since we've standardized on AC, we'll probably just convert back to AC to run "legacy" devices and systems. So you'll have a DC feed to your house from the DC grid, but local conversion to AC for whatever.

[kevm14]

There is a new standard for datacenter power, at 380V DC.

http://www-03.ibm.com/procurement/prowe ... posium.pdf

http://www.i2sl.org/elibrary/documents/ ... schudi.pdf

This last one examines residential applications:

http://ecmweb.com/contractor/dc-house

[kevm14]

One challenge with DC is mechanical current interruption becomes difficult due to arcing. Solid state is probably the savior here.

With AC, the arc "burns out" when the current crosses the zero axis so arcing is limited in both distance and time.

[Adam]

There is a new standard for datacenter power, at 380V DC.

http://www-03.ibm.com/procurement/prowe ... posium.pdf

http://www.i2sl.org/elibrary/documents/ ... schudi.pdf

This last one examines residential applications:

http://ecmweb.com/contractor/dc-house

DC really does make sense for large data centers.

Consider the following use case:
Some server rack UPSs can be configured to act as a power conditioner in the event facility power is not stable or clean. When this system is in use the power path is:
- Facility AC to the UPS where it charges the battery pack
- DC battery power in the UPS is converted back to conditioned AC power that feeds the PDUs
- The server power supplies attach to the AC PDU and convert back to DC to power the server components.

Power delivery components in data centers are engineered for low conversion loss, but if you scale the above process over a 50,000+ rack data center, those losses really add up.