Retired Firefighter    ■    Power Dispatcher    ■    Husband    ■    Daddy    ■    Grandpa    ■    Crazy Man

Monday, August 17, 2009

Tutorial 2 - AC Supply and Demand

Welcome back to another boring ramble. You seem to enjoy this punishment a little too much.

When we last left off, I mentioned how you cannot store AC power. There is no such thing as a battery for AC power, since the basis of the chemical process used to push DC electrons from your handheld gadget and automotive batteries in one direction only is wholly incapable of producing the precise 60Hz, constantly-reversing flow cycle utilized by the North American AC power grid.

Therefore, the power used by AC systems is generated at the same moment that it is consumed.

If you tried to hook up a single power plant to a city with no external connections today, it would be very difficult to control the system's stability and very hard on the power plant trying to do it. When AC systems first came into being a century or so ago, a precise frequency of 60Hz on the nose was not as important to maintain as it is today. For sure, a higher or lower frequency might result in a typical turbine (spinning at 3600 RPM) having funny vibrations and tripping to prevent damage, but relatively wide frequency swings could be tolerated because there really wasn't any sensitive electrical equipment in use by consumers 100 years ago that would burn up if exposed to, say, 59.43Hz or 60.78Hz. By comparison, at today's standards, anything outside of 59.90-60.10 is extremely unusual and would get our attention in a big hurry. 59.95 or lower, in fact, is usually a good indication of a noteworthy system disturbance.

As an amusing aside: Analog clocks on the North American grid, on 120 volt AC plug-in power supplies, are engineered to count one second for every 60 cycles of AC. If the frequency were to theoretically, say, drop to 59.00Hz for one hour, the clock would only advance 59 minutes during that hour. So, even minor fluctuations in frequency, over time, can throw these old clocks off. Accumulated frequency deviations are recorded as Time Error, and the operators of power companies are from time to time ordered to control to a slightly different frequency (59.98Hz or 60.02Hz as needed) in order to keep this accumulated deviation minimized. Heaven forbid an old clock lose two hundredths of a second over the course of a few days.

And yeah, if you plug in a clock built for the European-style 50Hz system into the North American grid, it will spin fast to the tune of 72 minutes for every hour of actual time.

Back to topic.

If you have a perfect match of mechanical-force-moving-a-generator to instantaneous-electrical-demand, the frequency will be 60Hz and all is well. As soon as something gets turned on or off (light switch, blender, table saw, whatever), the balance is disturbed and frequency will change. If there is too much mechanical input compared to demand, the frequency will rise until the balance is restored. This is accomplished by, among many other options, perhaps backing off the water fed to a hydro turbine, or by slightly manipulating the valves at a thermal (coal/gas/oil) power station to reduce the amount of steam reaching the turbine. Likewise, adding demand will reduce frequency until more supply is provided. Once balance is restored, frequency returns to 60Hz. If you get roughly an equal amount of low and high imbalances, the accumulated deviation tends to cancel itself out most of the time.

Now, as I said at the beginning of this post, if you had one power station managing an isolated city, it would be very hard for this power plant to keep up. Imagine at 10:20PM when a huge number of people turn off the 10:00 news and the lights, and hit the sack. The plant would probably not be able to back down fast enough to match the fast-dropping demand. Conversely, at 6:25AM when furnaces, coffee pots, water heaters, and lights all get rapidly turned on, it would be very very difficult for one power plant to keep up and prevent a huge frequency decay and system collapse. 100 years ago, these huge demand swings did not exist at this scale, so single-plant systems that wouldn't survive today were not so stressed out then.

Back in the 30's and 40's, and going forward, power companies began to tie their systems together to increase stability and reliability. By having a power line between you and a neighbor, you could schedule one of your power plants for an outage to repair or upgrade something, without having to black out some or all of your system, because you could fill in the shortfall by buying power from your neighbor, and vice-versa. In addition, if you have huge demand swings on the intertied systems, there are more total power plants exposed to the swings and able to respond. As the number of customers in a single intertied system increases, the individual impacts of things being turned on and off, combined with the averaging effect of having such a huge number of users on the same system, decreases the impact of individual demand changes in proportion to the whole, making load fluctuations much smoother. This in turn reduces wear and tear on power plants trying to chase demand and maintain 60Hz.

An analogy of the above. If you have shopping basket and keep randomly adding or removing oranges, the weight will change noticeably every time something goes in or out. Your arm will get tired of the changes. If you instead have a dump truck and have twenty people randomly adding or removing oranges, even if you do it kind of quickly, the dump truck really doesn't notice. Isolated AC power systems are shopping baskets. AC grid interconnections are fleets of river barges (bigger than dump trucks!) with crowds and crowds of people rapidly adding and removing oranges, and despite the chaos there is never enough sudden coordinated increase or decrease in volume to do much more than make even a slight ripple in the water.

That is probably the lamest analogy I have come up with for a while, but it will have to do.

Since frequency is sychronized on AC equipment tied together, the frequency is for all practical purposes exactly the same on any given point in an interconnection. There are two major interconnections in North America (east and west). Thus the frequency in Maine is exactly the same as that in Louisiana, and the frequency in British Columbia is the same as that in New Mexico... barring system disturbances and 'islanding', of course. If something seriously goes kaboom in Florida and whacks the frequency, dispatchers in Saskatchewan see it within four seconds on their frequency charts, and very likely look at each other and say 'I wonder who just got whacked, eh?'

I can't go on. Half of you are asleep already anyway.

What we learned: (1) AC power is generated almost precisely at the moment it is consumed by controlling output based on demand and frequency. (2) Large AC interconnections made up of bunches of power companies - their generators and loads all tied together - tends to blunt the impact of load changes, which minimizes imbalances in load and demand. (3) There are two huge interconnections in North America. The East is gargantuan, while the West is merely huge (Texas and Quebec are their own AC grids, weakly tied to the big boys with special high voltage DC power lines - more on that some other time). (4) Power companies actually spend time trying to make sure your analog clocks stay precise even though you can usually get the exact time from your cell phone or the Internet in less time than it takes to check or set your AC wall clock, and this of course assumes that you have ever attempted to set your AC wall clock to the precise exact second in the first place.

(Click this link to see all posts tagged "tutorial")

No comments:

Post a Comment