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

Question about power generation

  1. Jan 1, 2013 #1
    Hi, it has recently occurred to me that current technology does not allow massive electricity storage, and this has made me rather confused, as this implies that power generation has to be exactly matching the demand of power at any given time or else energy is wasted because of generating too much power or power outage will happen as not enough power is generated, judging from the constant supply of electricity, I suspect that the former is much more likely, and this means that massive amount of resources are wasted just to maintain constant electricity supply, surely this cannot be true, any advice or information would be much appreciated

    thanks in advance

    ps: this question is purely out of curiosity
  2. jcsd
  3. Jan 1, 2013 #2


    User Avatar

    Staff: Mentor

    Power plants are throttled.
  4. Jan 1, 2013 #3
  5. Jan 1, 2013 #4


    User Avatar
    Science Advisor
    Homework Helper

    Hydro-electric systems store gravitational potential energy. They don't store electricity.
  6. Jan 1, 2013 #5

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    And batteries store chemical energy.

    To the electric utility dispatcher who must plan ahead which plants to run in order to meet today's demand, a hydro plant is just another source of electricity. Its fuel source is a lake instead of a coalpile, and he'd probably think of lake's inventory as megawatt-hours instead of gallons or acre-feet..

    "Stored Electricity" is a not bad way to think of a hydro plant, so long as one accepts it's a convenient shortcut and not a rigorous definition.
  7. Jan 1, 2013 #6

    thanks for the reply, however I am still a little confused, as doesn't majority types of power plant have to maintain critical temperature, so resource might be wasted when power demand is below the minimum generating requirement to keep the power plant running, and even if it is throttled how does the power plant control know how much to throttle, do they have a mechanism with instantaneous feedback, or do they crudely estimate the amount of electricity to be used?

    thanks in advance
  8. Jan 1, 2013 #7


    User Avatar

    Staff: Mentor

    Power demand is both predictable and relatively slow to change, so it isn't difficult to follow.

    Steam plants can throttle without much trouble and except for nuclear plants, without much waste. Maintaining temperature does not mean maintaining power. Think about a boiling pot of water on the stove: it boils over a wide range of burner settings, only difference being how fast.
  9. Jan 1, 2013 #8


    User Avatar
    Science Advisor
    Homework Helper

    The key thing here is that the generating plants are not indepedent of each other. They are all connected to the same national (or even international) electricity grid.

    Sure, if the demand drops drops far enough, you shut down some of the generating plants completely. That's why you have a mixture of large plants (e.g. nuclear or large coal-burning steam generators) which are efficient but can only respond slowly to changes in demand, and also smaller generators (e.g. natural gas fired) which can be stopped and started lilterally in a few minutes.

    In simple terms, you can think of this like the cruise control on a car. You set the speed you want to travel, and the system automaticalliy adjust the throttle to burn more or less gasoline, or even apply the brakes, depending on whether you are on level road or going up or down hill.

    If one electricity generator on the network tries to supply "to much" power, because all the generators are connected together what happens is that ALL the generators will tend to speed up a little bit, not just the one that is working "too hard". The control system works by keeping the frequency of the supply fixed at its correct value (50 or 60 Hz depending what country you are in). That frequency corresponds to the speed in a cruise cnotrol system.
  10. Jan 1, 2013 #9

    I thought the above description is what the developing smart grid systems are for?

    thank you for the reply
  11. Jan 1, 2013 #10
    Whilst true I think this is a grossly disingenuous response.

    What do you do if you have a large supply of electricity that you are currently or periodically not using, but want to store it?

    Pumped storage hydro is an excellent solution.

    What do you do if you have a disparate collection of variable and intermittent supplies of electricity?
    You might well see significant levels of this type of sourcing in the future with renewables.

    Pumped storage hydro is again one excellent solution.
  12. Jan 1, 2013 #11
    Can you expand on the issue you see with throttling nuclear units?

    As far as steam plants go, they are (?) designed for maximum efficiency at full power, so running throttled would seem to be less efficient (in other words, more Btu/kw-hr, or wasted fuel).

    For nuclear units, the fuel cost is almost insignificant, so a loss in steam cycle efficiency doesn't translate to higher cost for the power. At least not as much as it does for a coal or nat gas fired plant where the operating cost is almost entirely due to the fuel cost.
  13. Jan 1, 2013 #12

    jim hardy

    User Avatar
    Science Advisor
    Gold Member


    a power plant, in fact the whole power grid, is just like any other engine
    you adjust the fuel input to control the power it makes. So energy is not particularly wasted.
    If the power being made in the grid is unequal to the power being used in the customers' homes and factories,
    then like any other engine the grid will speed up or slow down.
    Extremely sensitive speed measurement machinery keeps the whole electric grid at 60 or 50 cycles (Hettz) as Aleph said above.
    I worked in a power plant. The frequency recorder drew an almost straight line around 60hz, varying by 0.01 or less. Only one day in 30 years did i see the grid run 59.9 for more than a very few seconds.

    Smart Grid is not needed for any real reason that i know.
    It's just a big next step in automating things .
    Smart Grid is going to greatly simplify "time of use" billing and remote control of every individual wall outlet in the country. Do some reading on "smart grid" products, you'll be amazed.

    BTW nuke plants can "throttle:, it's called "Load Follow" but is avoided for two reasons...
    1. It's usually more thrifty to throttle some other plant and save its more expensive fuel, burning instead that cheap nuke fuel;
    2. Load Following shifts around the heat production among regions in the reactor's core . If it causes unbalance, you may not be able to get back to full power as soon as you would like. Look up "Xenon Oscillation".
  14. Jan 1, 2013 #13


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    On the flip side of this, there is the case where the demand exceeds the available supply. In this case, utilities can reduce the voltage over the supply lines to homes and businesses by up to 5 percent, so that less current and hence less power is used by machines, appliances, lighting, etc. This 5 percent power reduction causes appliances to run a bit sluggish and lighting to be a bit dimmer, with I believe possible damage to equipment over long periods, but with no major issues if the period of reduction is not lengthy. Next step is to employ rolling blackouts. This was a bit more common years ago when supply was insufficient to meet peak demands (typically in hot summers), but there seems fortunately to be an abundance of power to share these days, at least in the US, perhaps not so in other countries (like Myanmar, for example, although that will change in the future when businesses increase their investments there).
  15. Jan 2, 2013 #14
    thanks for reply, they have clarified my confusions
  16. Jan 3, 2013 #15
    Chemical batteries are an economically sensible way to store electricity at the scale of the power grid.

    If you check the capacity and price (...and the life expectancy, far less obvious) of a lithium battery, you see it's about affordable.

    That is, a 1300MW power plant delivers 1000MW as a mean because users consume only 700MW during night, and also because the plant is shut for maintenance from time to time. Storing 300MW*10h or 3*109 V*Ah = 10PJ during night to give them back during peak day consumption would allow a plant of 1000MW only, or fewer 1300MW plants.

    This plant costs, say, 2.6G€ to build (depends a lot on the kind of fuel, just an example) so storing 10PJ over half a day with good efficiency would save 0.6G€ procurement cost.

    That's why a Japanese university works on sodium batteries, trying to achieve performance similar to lithium with a more abundent metal.

    My own contributions to flywheels, whose price seems very sensible within the above estimate, begins there:
    with the first technology improvements on Tue Jun 28, 2011 (page 3)

    Prof. Seamus Garvey also proposed underwater bags filled with pressure air. Looks poetic, but after putting figures on them, I'm convinced this is strong and sound technology.

    More methods exist, have been used, or are presently used, like the already mentioned dams with pumps and turbines.
  17. Jan 3, 2013 #16


    User Avatar

    Staff: Mentor

    A couple of things:

    Peak shaving requires surprisingly few hours a day or year to make a substantial difference. 3 hours a day is more realistic, and only on at most a dozen days a year (unimportant). So at a current cost of around 300 euro/kWh that's 0.9 euro/Watt. But:

    That's just the battery. The infrastructure to connect the battery to the grid is substantial. Charging and control circuitry, inverters, not to mention a building. And that's assuming lithium ion batteries are scale-able.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook