How Do They Synchronize Power Stations With The Grid?

There are probably times in every Hackaday reader’s life at which you see something and realise that the technology behind it is something you have always taken for granted but have never considered quite how it works. Where this is being written there was such a moment at the weekend, an acquaintance on an amateur radio field day posted a picture of three portable gas-powered alternators connected together and running in synchronization. In this case the alternators in question were fancy new ones with automatic electronic synchronization built-in, but it left the question: how do they do that? How do they connect a new power station to the grid, and bring it into synchronization with the line? There followed a casual web search, which in turn led to the video below the break of a bench-top demonstration.

If two AC sources are to be connected together to form a grid, they must match each other exactly in frequency, phase, and voltage. To not do so would be to risk excessive currents between the sources, which could damage them and the grid infrastructure. The video below from [BTCInstrumentation] demonstrates in the simplest form how the frequencies of two alternators can be matched, by measuring the frequency difference between them and adjusting their speed and thus frequency until they can be connected. In the video he uses neon bulbs which flash at the difference frequency between the two alternators, and demonstrates adjusting the speed of one until the bulbs are extinguished. The two alternators can then be connected, and will then act together to keep themselves in synchronization. There are further videos in which he shows us the same process using a strobe light, then demonstrates the alternators keeping themselves synchronized, and phase deviation between them.

Of course, utility employees probably do not spend their time gazing at flashing neon bulbs to sync their power stations. The same measurements are not performed by eye but by electromechanical or electronic systems with automatic control of the contactors, just as they are in the fancy electronic alternator mentioned earlier. But most of us have probably never had to think about synchronizing a set of alternators, so to see it demonstrated in such a simple manner should fill a knowledge gap even if it’s one only of idle curiosity.

58 thoughts on “How Do They Synchronize Power Stations With The Grid?

  1. Often a syncroscope is used in a power plant setting to get the geberating equipment like the turboalternators in the same phase. A really neat example is in the battersea power station auxillary control room.

    1. The most basic approach is a “zero cross” detector. Once the circuit sense the zero cross, it triggers the oscillator for the next cycle (normally 16.6666 ms before in the case of 60Hz). Of course it keeps sensing permanently to ensure the synchronicity, but also to avoid “islanding” and disconnect the inverter in case of a blackout (normally if it looses 2 cycles, 33ms).

    2. I think that can be made comparing the grid and the inverter signals using a phase locked loop. The frequency generated by the PLL will be the reference for the inverter. In the case of the grid power off, the feedback loop in the PLL can be disabled and system can run at a constant frequency based on the local grid (50Hz or 60Hz).

      Another interesting point is the control of the generated power, if the solar panel and the inverter are connected to supply the excess of energy to grid, the system need to control it to not overload the inverter.

      1. The amount of power pushed to the grid is controlled by the phase of the generated voltage waveform. If it tries to advance the grid frequency, current is forced into the grid. Same thing as with rotary generators: the generator tries to advance the grid and so power flows out, which puts a load on the generator and keeps the frequency stable – only very minutely shifted in phase relative to the grid.

    1. I was in the control room for a startup one time about 15 years ago.

      Big red button – they watched the frequencies on an old school seven segment display and when it was time they banged the button and hoped the turbine didn’t explode into a million pieces (spoiler alert…it didn’t).

      They said some plants had fully automated systems back then…suppose most do now, although changing anything requires a lot of red-tape so it wouldn’t surprise me if there are still some doing it the old way.

    2. Boy I have! Once with a 16MW IC turbine synced incorrectly with a synchroscope. The generator section broke all the bolts holding it to the skid as well as the shaft and all wiring and conduit. The generator continued on its path out of the metal building it was housed in and slammed into another IC turbine in an adjacent building. I and the stupid operator were about 40 feet away at the control panel. Much noise flame shrapnel,smoke and arcing ensued.
      No one was injured and it’s fortunate I didn’t have to run away like the operator did, cause my whole body was locked up watching this debacle. Wished I had a smartphone in 1982.

      This event is rated by me as the third out of six dangerous events I have witnessed working at coal-fired and nuclear power plants. And second most damaging event related to grid syncing. I saw a step up transformer turn to plasma.
      From a distance, the top level of the unit. I didn’t hear the explosion because I was somewhat near the safety relief valves.
      I lost a bunch of dbs of hearing about then. Lots of J/cm. That unit no longer exists not but I bet the marks from my hand holding on to the guard railing were still there when they demoed the plant.

        1. There most certainly are breakers and/or fuses, but yes, it’s also too quick.

          I know of a small (<10MW) generator that was turned on at roughly 120 degrees out of phase – the worst possible scenario. What happens is the grid (which is some GW, connected at a rather low impedance) tries to force your generator into sync, instantly. Unfortunately inertia exists, and steel shafts are not infinitely strong. In other words, the shaft sheared itself off. From memory this shaft was something like 300mm/12" of solid steel…

          There should have been an interlock to prevent this, but for whatever reason it didn't work.

      1. I don’t know if this is an urban legend or not, but here it is anyway.
        Our local electrical generator, Eskom, had a small coal fired power station, which was primarily used for training purposes. The station gets stripped down and rebuilt periodically, as part of their training. This includes removing and reinstalling the taps on the transformers that feed the frequency gauges.
        This particular occasion, the taps were installed backwards, resulting in a reading that was 180 degrees out of phase. So when the dial speed spinning, indicating that they were in sync with the grid, and they closed the breaker, the grid effectively said Nope!
        Since the grid was backed by thousands of megawatts, and this was something like a 30 MW turbine, the grid won. The turbine stopped turning instantly, only to start turning again 1/60th of a second later. This resulted in a multiple ton turbine being launched out of the building and over the river several hundred meters away.
        The facility was closed down, and never used again.
        According to my recollection of the story, this happened in the 1950’s or thereabouts, which explains why i have not been able to verify the story.

        1. I guess it is possible, but in the plants I have been in (mostly smaller plants… not an operator, but the work put me in contact with the operators), even with automatic controls, much less synchroscope, there is still a lamp (as well as a meter) across each switch phase. If the lamp isn’t out (or the meters don’t show near zero), the phase isn’t right.

    3. A Synchroscope is a common instrument in power systems with more than one rotating machine (generator). The design of a Synchroscope can be rather complex (using transformers) or fairly simple using a capacitor (one input phase-shifted by 90 degrees then compared in Quadrature). The main disadvantage of a capacitor based Synchroscope is limited bandwidth, which at AC line frequencies is not a big deal. Two capacitors of different values provide complete DC isolation between the two inputs. Then there is the issue of differing voltages between the two inputs (do the math for two sinusoids in quadrature with differing phase and/or amplitude. Or just simulate it in SPICE). The use of coupled auto-transformers can help in this case. Legacy high-end Synchroscopes have two galvanic indicators (meters with needles for the “Makers” out there) in one device, a top indicator for frequency difference (independent of voltage difference) and a bottom indicator for voltage difference (independent of frequency difference). In modern times the likes of Synchroscopes are realized using analog/digital data acquisition & display (e.g., a data-acquisition system and a PC – hopefully one NOT running Windows and NOT irresponsibly connected to the Internet!) I would really like to see some detailed tear-downs of legacy analog Synchroscopes on the Web. Anyone with links?

  2. Brings back memories. Had to sync alternators to each other and to the line in an electrical lab in college. It may be one of the reasons I went into computers. Way too much power for me!

  3. I cringe when I see crock clamps shunting current transformers. Personally I like to solder a shunt, like 0,1 Ohm right at the current transformer, then use a voltmeter to measure the output. If a current passes through an un shunted current transformer, the output voltage raises to, in theory, an infinite voltage. In practice, a damaging, potentially dangerous voltage.

    Nicely set up demo though.

  4. My dad worked at a wastewater treatment plant with massive backup generators. The whole plant would go down on power loss, the main breakers would automatically open, then the generators would start up and processes would resume.

    To go back to mains power, there were three light bulbs. Once they all went dark you would reset the breakers, then disconnect and shut down the generator.

    I’m not aware of any incidents, but I would imagine things would not be good if you hit that button at the wrong time.

    1. That system would be called a “make-before-break” system. That is so the system could be transferred back to the grid while live before shutting down the generator. I also work in water treatment. I wish we had one at my current location…

  5. I vaguely remember my power systems professor saying that there is a way to run a generator onto the grid by allowing the frequency difference between the grid and the new generator (with almost no load on it) drive the synchronization. Basically any difference between the two frequencies speeds up or slows down the generator.

    1. In the 90s I knew a guy in the US who’s family owned a small lake and who had assembled some sort of dam. He said that if they were careful they could run the pump turbine in a way which would energize it and sync it to the grid and according to him could be configured to backflow power sufficient to run the whole area yet bring the utility meter near to a stop while draining the lake into the turbine flow. Not sure how true this is but I also doubted his owning several machine guns including an old WW-II Sten Mk-II which he demonstrated for me the one time I ever visited the property but I didn’t get to see the damn dam configuration though.

    2. A generator and a motor are (substantially) the same thing. Some smaller hydro units (Mechnicsville in CT, if it is still there, for example) use slightly modified induction motors as the generator. Self synching.

    1. So, with the grid, all the generators are synced to each other, with the ones that originally started the timing no longer existing. However, there are plants designated as black start plants, meaning that they can be used to bring the grid up from cold, generally these are nuclear or hydro plants as they can be set up to be very stable in frequency and have extra capacity to deal with synchronization. These plants are generally figured as being for restarting isolated grid segments after a natural disaster or large scale blackout, though they could also be used to bring the national grid(or north american grid) up from a cold start.

      So really, it’s all synced to a memory of a generator that hasn’t existed in decades held by every synchronized generator. So it’s a hivemind, with certain plants being designed such that they can dictate phase and frequency from a cold start.

      1. They also have huge generators required to start up the power plant again, and all of the other plants pull power from the black start plants. These facilities take a lot of power to operate, and even more to start.

  6. I used to do this regularly as a ship’s engineer. 4 generators and shore power (plus a grossly deficient automatic system) meant that I had to do it by hand on a regular basis.

    We had an LED synchroscope and manual control of the breakers. It was easy enough with a bit of practice, although I did manage to do it once blindfolded. There is was a check relay that energised at the right point to prevent you from engaging breakers with the generators out of phase. By listening for the click I gauged when to hit the button.

    1. When I was on a nuclear sub in the 1970s, I was Electrical Officer for a time. Synchronization wasn’t automatic for us. We would adjust the frequency of our generators in order to have them running very slightly faster than shore power for the synchronization, and make the connection just as the phases approached matching. The idea was to compensate for lag time in the switch closing, and that the hardware would automatically compensate for small differences in phase.

      I recall that one time, the technician was a bit impatient, and hadn’t adjusted the frequency to be close enough. He made the connection when the phases matched (roughly), but the phase meter was moving more quickly than was safe, so he hit the switch just past the match, and we were several degrees out of sync when the connection was made. I’ll admit to not remembering exactly what the results were, apart from the Officer of the Deck calling to express his … disappointment.

  7. As was told to me the procedure is: you get the apprentice to stand next to the generator and then throw the breaker when it’s slightly out of sync, causing a large angry diesel engine to try and jump off its mountings and the apprentice to require a change of underwear.

    Thankfully by my time it was all automatically syncing, so they just got me to stand near the generator and throw the switch without warning me the thing was about to roar into life (with a loud warning siren too). My underwear survived.

    Aaaah, happy days.

  8. I used to do a fair bit of temporary site power for music festivals. For any decent sized stage we use redundant generators (very clever, they talk over RS485 to optimise fuel efficiency including autostart/stop). They have a built in synchroscope and do it all automagically, but you can see the process and the big contactor firing.

  9. uhh….noob question… if you have a generator powered up, online, and supplying the load – why would you want to connect it to the utility grid ? Unless you’re “selling” back electricity. My backup house generator,has a transfer switch specifically designed to prevent such a scenario (backfeeding to the utility).

    1. Load sharing among multiple generators (alternators). If you have a turbine or solar installation on your house, you sync it up to reduce your consumption of the grid power. If you operate a grid, you sync up multiple plants to carry the total load.
      Your transfer switch is for the safety of electric company linemen who might be working on lines that should be dead, but might be backfed from your generator.

  10. In theory and practice, the AC Line (Mains) frequency when fed by rotating machines with rotational mass (e.g. shaft-driven AC generators) is a good measure of load. Sampling the line frequency of interconnected (or even isolated) systems (grids) allows time-based mathematical modeling (i.e., simulation) of the system(s) via (e.g.) Jacobian Matrices of partial derivatives. Sometimes parts of the grid are intentionally isolated from other parts of the grid to see how they individually behave in terms of load over time (peak hour vs. minimum hour load). Repeatedly in the clueless STE[\A]M or Maker community I see simple attempts to use the AC Line frequency as a timing reference. The information above should be explanation-enough why this is a BAD idea (at-least unless significant efforts are made to statistically process the data).

    1. AC power has been used for timing for years, and has proven to be reliable. IIRC they also maintain the the frequency such that any differences from nominal cancel out overtime.

    2. The clocks I have which use the AC grid for a time base don’t drift noticeably over a period of months. The grid may be an imperfect timebase (especially when there’s an outage) but it’s pretty good for a cheap and widely available timebase.

  11. I’ve connected to the main Edison grid thousands of times.. Everything from a 1 Mw Konesberg gas turbine to a 50 Mw Fuji steam turbine. All the same.. Sync scope rotates to “0” (Frequency connect point) and I snap the breaker and we were on line. What you can’t do, is “Push” Edison, as the system temp goes sky high. By “Push” I mean, go faster than the Edison grid frequency. Since a “Generator” (USA) runs at 60 Hertz, it’s the “Zero” on the snyc scope. Any faster, (61 Hertz) is disaster..
    Things go bad in a hurry if you actually are stupid enough to try to push the grid. I always thought 59.9999 was good, and then ramp up the fuel to increase power to the grid. We made a buck or two..

  12. Wow, i never thought about this. Amazing to think that all the generators are spinning sincronized in a grid. Is like waves in the sea i guess. There is big mases moving a certain speed and you try to match the speed as close as you can to that masses and then it will push or pull you to phase

  13. For a large grid, say the size of a continent, doesn’t the speed of light cause unavoidable synchronization problems? The grid isn’t just a linear set of nodes; it’s highly interconnected. If the perceived phase of two nodes adjacent to me differs, how can I synchronize?

  14. I am curious about synchronization. For grid-tied solar inverters, each inverter is a high impedance current source and dumps power in the grid in sync. It is not allowed to regulated voltage, just protect to prevent under or over voltage and frequency/phase drift. Say there are multiple power generators which are perfectly synchronized but have slight voltage difference, will the generator at lower voltage sink reverse current from the higher voltage grid? I am a DC power guy and without necessity of synchronization, we are not able to achieve power sharing between parallel voltage sources without active intervention. Then how will multiple AC voltage sources (ac power generators) share the power proportional to their capacity even if in perfect frequency and phase sync?

Leave a Reply to nuclearCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.