Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Clocks in town mysteriously changing by 10 minutes

  1. Jun 19, 2004 #1

    Ivan Seeking

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    http://www.saultstar.com/webapp/sitepages/content.asp?contentID=70613&catname=Local+News [Broken]

    Just in time, North Shore fast-clock mystery solved
    http://www.saultstar.com/webapp/sitepages/content.asp?contentID=70740&catname=Local+News [Broken]
     
    Last edited by a moderator: May 1, 2017
  2. jcsd
  3. Jun 19, 2004 #2
    If they traced it to a change in the power line frequency, it's not so mysterious, is it?
     
  4. Jun 19, 2004 #3
    It'd be a nice April fools joke one could play on a small town. Rev up the frequency at the genreator as high as it'll go before things plugged into the grid start breaking. Do that the night before. Have the whole town get up at 4am instead of 8am let's say. :)
     
  5. Jun 19, 2004 #4
    What needs explaining to me is why any power station isn't running at 60hz as a matter of course. The whole point of the choice of 60 hz was for the accuracy of electric clocks. The normal customers of that station will always be running fast if they are always operating a couple herz in excess of 60.
     
  6. Jun 19, 2004 #5

    chroot

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I agree, I do find it a bit weird that a generating station would run anywhere from 60.5 to 62 Hz... power stations are very sensitive to the power factor on their lines (they don't like wasting power charging and discharging inductance, for example), and power factor depends directly on frequency.

    The only thing I can think of is that adjusting the frequency is the poor man's technique of managing power factor -- versus actually building the capacitor yards that are needed to adjust it properly.

    - Warren
     
  7. Jun 19, 2004 #6
    Really I belived it just evolved that way into a very inefficient power transmission system!
    Im not sure thw above quote is correct.
     
  8. Jun 19, 2004 #7
    Yeah, it seems to me that if they aren't running at 60 hz the only possible explanation for it boils down to: for some reason they can't.

    I have very little idea about how power stations actually operate. I've seen pictures of the huge turbines, but don't know what all needs to be done to control the voltage, herz and everything else that needs controling.

    How do they utilize "capacitor yards" in this?
     
  9. Jun 19, 2004 #8

    chroot

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Quoting from "The Art of Electronics" by Horowitz and Hill:

    As you can see, the power company wants to keep the power factor as close as possible to 1.0. It can adjust the power factor by adjusting [itex]\omega[/itex], the frequency. Even a 0.5 Hz change can result in a large savings of energy in a large system. This is the only reason I can imagine for wanting to have an adjustable frequency like this -- but I'm no power distribution expert.

    - Warren
     
  10. Jun 19, 2004 #9
    This explanation makes perfect sense: an economic reason.

    I find that fascinating that such a small difference in herz makes such a big difference in cost.

    Thanks for digging that quote up.

    I still wonder what goes on with the people normally fed by this station. Maybe they all just adapted to the fact they couldn't rely on certain clocks.

    -Zooby
     
  11. Jun 19, 2004 #10

    chroot

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Keep in mind I'm not an authority on this particular generating station or its operating principles. :smile: It's just the only reason I can see (as an EE) for wanting to vary the frequency. It might also have some kind of mechanical rationale -- perhaps letting the generators run a few hertz faster or slower allows them to better manage the water flow through the dam, or deal with waste heat better, or something. I'm just speculating -- it'd be fun to find out the real reason why this plant has such variability.

    - Warren
     
  12. Jun 19, 2004 #11
    Yeah, it would be interesting to find out. I've been under the impression that 60 hz was pretty strictly observed so that things were the same all over. I wonder if you dug into it, you'd find out there's more stations like this one that fudge the 60 hz for various reason, both economic and the other sorts of reasons you suggested.
     
  13. Jun 20, 2004 #12
    But is it a matter of wanting to vary the frequency, or just not being able to control it precisely.
     
  14. Jul 29, 2004 #13
    losing or gaining frequency

    What it sounds like to me is that if Hydro One is disconnected from the Ontario Hydro grid is that Hydro One is acting like an "island."
    The turbines in most hydro plants usually spin the rotor of the generator at around 350 rpm. If you do the math, you could figure out how many poles are required in the generator to give you a output frequency of 60 hz (3600 rpm).
    I guess in layman's terms, when Hydro One is an island independent from the electric power grid, it is hard for the utility to regulate their frequency. If the load demand is greater than the current generation, the generators are going to bog down and your frequency is going to drop. If, in the other case, your load demand is less than your current generation, your generators are going to speed up. There are valves that must be controlled to adjust the flow of water into the turbine to keep it at around 350 rpm. There was a hydro in northern Wisconsin that would disconnect from the grid for line outages. The hydro output enough power to power a couple small towns. The changing load demand made frequency control difficult. The operator there used to control these water regulating valves, and therefore adjust the frequency, by comparing his wristwatch with an electric-powered wall clock. If the wall clock lost time (frequency was too slow), he would open the valves a bit more, and if it gained time, he would close them off a little bit.
     
  15. Jul 29, 2004 #14
    That's quite an amazing story: keeping it in time with his wristwatch.

    So, in the grid situation it seems you would still have peak demand times vs slower times (night). Do these usually just average out, or are they constantly making adjustments for the whole grid somehow?

    -Zooby
     
  16. Jul 30, 2004 #15

    Ivan Seeking

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    As for the watch, even a 2% error would be easily detectable in one minute. That is a very cool story, IMO.

    As for the grid, once a generating station is on the grid, the grid determines the frequency; the price for any errors in the generating frequency is efficiency. I guess the grid frequency is determined by an average of all contributing stations.
     
  17. Jul 30, 2004 #16

    russ_watters

    User Avatar

    Staff: Mentor

    Its even more important than that. Being off by small fractions of a hz for an extended period leads to all sorts of problems with the grid. For example, if one plant is at the top of its sine wave and the next plant over is at the bottom, the voltages cancel out to zero and the grid collapses.

    edit: actually, they'd be in parallel, so its more complicated than that. Gotta think about it...

    My guess is that that plant isn't capable of accurately regulating its frequency - iirc, the large sections of the grid itself have their frequency attached to an atomic clock at some monitoring station and all of the plants synchronize to it.
     
  18. Aug 3, 2004 #17
    I was a utility worker for over 14 years. I held a Senior Reactor Operators license at a nuclear power plant. Striker actually had the answer a couple posts back. If a power plant is connected to the grid, the grid will determine frequency. However, if you somehow get disconnected from the grid, the speed of rotation of your turbine will determine frequency and this is sensitive to load changes, steam/water pressure (for hydro) changes, etc.
     
  19. Aug 3, 2004 #18
    If you're connected to the grid, you can't fudge frequency - it will be rock solid at the grid frequency no matter what you try. All you can really affect is the amount of reactive power you generate and its sign (in or out as they say in the utility biz).
     
  20. Aug 3, 2004 #19
    There is no atomic clock anywhere that regulates grid frequency. It's just the fact that there are a LARGE number of generators connected to the grid all running at about 60 Hz. If you should vary much from that, you get feedback effects from this essentially infinite power source that drag you back to 60 Hz.
     
  21. Aug 4, 2004 #20

    russ_watters

    User Avatar

    Staff: Mentor

    That explains why a single unit can't change the frequency, but not how the frequency stays at 60hz. Why doesn't it fluctuate more with load? Something has to regulate it.
     
  22. Aug 4, 2004 #21

    Averagesupernova

    User Avatar
    Science Advisor
    Gold Member

    Look at it the same way you look at thousands of people clapping in unison to a beat. No one single individual regulates the group, but everyone individually strives to hit the beat right on. If you are off, you will know it. If the generator is off, it will know it also. However, one generator that is significantly large enough relative to the others could tend to lead (not the lead/lag sense) the others wherever it wants to.
     
  23. Aug 5, 2004 #22

    russ_watters

    User Avatar

    Staff: Mentor

    I understand how an individual has little impact over the group, but that doesn't explain how the whole group stays pegged at 60hz. Power draw from the grid varies widely and that will tend to push the frequency around, yet from one day to the next the grid won't have so much as a single extra cycle.

    What if tomorrow, someone decided to run the grid at 50hz (like Europe). How would that be accomplished?

    I'm looking for info on the web and there are a lot of sites that say its tied to atomic clocks, but I can't find any specifics on how it is regulated. Still looking...
     
  24. Aug 7, 2004 #23

    Averagesupernova

    User Avatar
    Science Advisor
    Gold Member

    I've never worked for the power company but I know a little bit about something called tracebility. Tracebility deals with how an end product or service is tied back to the National Bureau of Standards. Most likely a number of power plants, if not all of them, are tuned into WWV out of Ft. Collins CO in order to obtain a frequency standard. They then use this timebase to measure the frequency of or RPM of the generators. However, it probably doesn't do much good for error correction since if one generator measures a frequency error, all the plant can do is report it to the rest of the grid. I believe they also count the TOTAL number of cycles in a given period of time (possibly a day or more) and actually simultaneously correct for it. Probably over a period of a couple of minutes or something, maybe more I am not sure. This was told to me by an electrical engineer who had power distribution classes in college.

    Someone also mentioned power factor correction by capacitor 'yards'. I was always under the impression that power factor correction was compensated for AT THE LOAD. Every subscriber is allowed a certain amount of reactive load. If the subscriber has a highly reactive load (usually inductive) then there is a fair amount of current flowing in the wires feeding the subscriber. If you multiply this by the voltage you get a product referred to VA or more commonly KVA. NOT WATTS. The meter actually measures in watts and will ignore reactive power. But the power company doesn't like highly reactive loads because their wires still have to carry the current. If you have a high enough of a reactive load they can require you to correct for this with capacitors in parallel with the inductive load. This forms a quasi parallel-tuned circuit. Impedance of a parallel resonant circuit goes up causing the current demand to drop. You don't pay any more or less for the power, but your line losses will be less. However, current between the inductive load and the capacitor will still be high. This is not to say that a power plant cannot 'see' a reactive load when hooked to the grid thus causing a high power factor. I would think they would adjust the phase of the generator to compensate for this. Seems more likely than capacitor 'yards'.
     
    Last edited: Aug 7, 2004
  25. Aug 7, 2004 #24
    10 seconds would be a large error. 10 minutes? Hard to believe.
     
  26. Aug 8, 2004 #25

    Averagesupernova

    User Avatar
    Science Advisor
    Gold Member

    10 minutes in a day would be 60.41666 hertz instead of 60 hertz.
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook