The line frequency of all the power plants in the USA is 60Hz.
In that there are hundreds of power plants connected to the grid, and that it is crucial to many modern electrical appliances that the line frequency is exactly 60Hz, how exactly is the line frequency of all the power plants coordinated?
Secondly, given all the solar and wind generators now feeding into the grid, and the variability of their output, how are they coordinated with the grid requirements?
Conventional generators are inherently synchronized - if one is running too fast, the grid will “suck” power from it, and cause it to slow down. Generators that are running slow will tend to draw power from the grid and will be speeded up.
AC inverters use electronic methods to make sure that the power they are generating is synchronized.
And, 60Hz is not particularly important over the short term. Over long periods, it’s nice to average to 60Hz, so clocks keep correct time, but no machine relies on the power frequency being better than 10% or so.
If you connect an AC generator to an AC electric grid, it will
A: become synchronized, or
2: asplode
To prevent the latter case, you want to try to get it in the general neighborhood of the grid frequency before connecting it. That can be done with a pair of oscilloscopes or similar apparatus connected to each side of the switch. Nowadays there are automated systems that do it.
Once a generator is connected to the grid it will stay in sync.
Is this true of inverter portable generators too? I know Honda makes a special cord to parallel their inverter units and I wonder if there’s a control line in there to keep then in sync.
You need to be careful when you bring a generator online. They have “synchroscopes” to tell you the phase angle on either side of a breaker. You issue control pulses to adjust the voltage and frequency. When the voltage and angle of the generator are within a certain percentage of the bus, you close the breaker and you’re off to the races. Once it’s online, it’s in sync and tends to stay that way.
A lot of clocks don’t rely on the 60Hz frequency anymore. Nowadays, the vast majority of them have quartz movements, which means they’re completely DC internally and have a piece of quartz which vibrates at a precise frequency due to an electrical field. Divide that frequency down a few orders of magnitude and you have once-a-second time source, acceptably accurate over the medium term with no worries about humidity, temperature, or minor earthquake.
You start with one generator. Connecting a second one to it is a bit tricky. You need to get the voltage and frequency the same, and you need to connect the two of them when their AC cycles are exactly in sync. This is actually not as difficult as it might sound. You can use a thing called a synchroscope which will tell you the phase difference between the two generators. If you get the two generators close in frequency, the phase will slowly change between them. Wait until the sychroscope reads zero, and throw the switch. Simple.
You can also use syncing lights, which are basically lights connected to each line. When there’s a voltage difference between the two lines, the lights light up. When the generators get close in frequency, the lights will slowly fade in and out as the phases go in and out of sync. When the lights are out, throw the switch. Again, simple.
Now that your two generators are connected, it actually becomes a bit more difficult to change their speeds. If you remove power from one, instead of slowing down it just ends up drawing current from the other, and becomes a motor instead of a generator. If you don’t add power to the other generator then it will slow down too, but since it will keep trying to push the other generator along the two generators will stay locked at the same speed. Similarly, if you try to increase the speed of one generator, it ends up trying to power the other generator and again the slower generator starts acting more like a motor. You can speed it up, but both generators still stay locked at the same speed so you have to provide enough power to speed both up at once.
Now you can add more generators to the line using the same procedure as above. With more generators on the line, it becomes even more difficult to change the speed of all of the generators. If you try to speed up one generator, it would have to power all of the other generators like motors in order to speed the overall frequency up, and once you add a bunch of generators onto the line, the power required to do that is more than any one generator can handle. At that point your line becomes pretty close to the equivalent of an infinite bus. Each generator is too small to affect the overall system. Each generator can add mechanical power so that it provides more electrical power to the “grid”, or it can remove mechanical power and provide less power, eventually becoming a motor that is powered by the grid if too much mechanical power is removed from it. But the generator remains locked on the grid frequency. It can’t change if it wanted to. Doing so would require huge amounts of current, far more than the generator is capable of supplying.
It is possible to change the “grid” frequency at this point, but it has to be a coordinated effort between all of the generators on the system. If just one generator tries to speed up, then the other generators will overpower it and nothing will happen. But if all of the generators speed up at the same time, then the grid frequency goes up.
Wikipedia’s article on AC synchronization:
Youtube video showing how to use sync lights:
Just guessing, but it’s probably a control line that allows them to share the load as evenly as possible. If you don’t coordinate the generators, one can try to supply all of the power while the other one does nothing, and then the first one burns out or trips its breaker.
Thanks.
Many thanks to Engineer-C-G (and others)
The explanations and links have been very informative.
I read the title as “…power pants…”.
I would have expected them to be parallelling bulk DC, but they’re doing it at 120 volt AC. No idea if the inverter has internal synchroscope functions, or if the two units communicate with X10signalling. I do know there’s more to the inverter unit than meets the eye - the thing is microprocessor-controlled.
A wiring diagram for Honda’s EU2000i is toward the bottom of this page.
X10?
Surely, you jest!
And they do that on occasion to initiate time correction.
Usually it’s done very early in the am.
While grid synchronized clocks are becoming rarer there is still a need to insure that there are the correct number of cycles averaged over time.
Typical forced frequency changes are in the 1/100 of a cycle per second range and last for an hour.
Perhaps its not important, but I used to work with electric utilities and remember being told that they tried to keep accurate to something like one cycle per 24 hours.
I used to work for MISO(Midwest Independent System Operator). We were tasked with maintaining the integrity of the grid, running the power auction each day and assigning / removing power providers from the grid. The phase conversations above covered the specifics. We would verify that a provider was “in phase” before we let them connect to the grid.
As for frequency; if you overpower the grid (i.e. produce more power than is being consumed) you drive the frequency up. Conversely, if the grid is under-powered the frequency drops (min allowed: 59.959 Hz ). Most of this fluctuation is covered by active feedback to generators. Large drops in supply (e.g. a power plant trips offline) require immediate replacement of that power. This is covered either by “spinning reserve”, generators that are spinning, but not connected to the grid (typically because their owner lost the bid for power that day, but didn’t shut them down) or by use of a “instant on” supply. In the Midwest US, that is “The Pond”, an artificial reservoir on the shores of Lake Michigan with a small hydroelectric dam. Open the pen-stocks and you have nearly instant power (something like an additional 200MW within 2 minutes).
As for wind…ah wind. No one in the power industry likes wind power. Not for evil, big oil conspiracy reasons. Just because it is way too variable. We used historical models to predictthe wind, which are surprisingly accurate, but if the wind should drop…you have to react to the drop in supply (see above). That means either keeping spinning reserve or paying big bucks for power from the pond. Both of which negate some of the savings you would get from “free” wind power.
Fascinating stuff.
They used to. The frequency tends to drop during the day because of all of the extra load on all of the generators (especially in the summer when everyone is running their air conditioning). As BubbaDog said, they would make up the difference in the wee hours of the morning.
In 2011, they decided to run a 1 year long experiment where they would relax the requirement for the long term accuracy of the 60 Hz cycle count, the idea being that very few people are still using synchronous AC clocks these days and they were wasting a lot of effort keeping something accurate when no one really needed it to be so accurate. To find out if anyone really did need it to be so accurate, they would stop regulating it and see who complained. I remember reading news stories about the experiment when it started, but never heard anything about it after that. I don’t know if they went back to regulating the long term accuracy after that or not. Since I never read about anyone being upset by the experiment, I suspect that they never bothered to restore the strict long term accuracy requirements.
I can find all kinds of links to the 2011 news stories on google, but I didn’t find anything in my (admittedly brief) google search looking for the experiment’s results.