6
\$\begingroup\$

After my second series physics course I have begin to wonder how power plants control energy production to match the usage of a city.

There are several possible answers in my mind listed below:

  1. Does it always over-produce the amount of energy that a city needs so that a certain amount of electricity is wasted?

  2. Does it over-produce the amount of energy but stores the excess energy in molten salt, or pumping water?

  3. Does it match exactly the current energy need?

The reason that this question has come to my mind is that if option 1 or 3 is true, then solar power and wind power is completely useless, as the generated energy must be used immediately, but I have to list as an option as I often don't see dams or molten salt reservoir around solar and wind power generators.

\$\endgroup\$
  • 2
    \$\begingroup\$ Power companies have pretty detailed (both long and short term) predictive algorithms (due to scale advantages) to forecast expected power use. This prediction is not 100% correct, and that causes measurable variations in grid voltage and frequency. Indeed solar and wind generators are a new challenge as generators take their time to adapt for varying load. To counter varying load and high fuel prices for regular generator plants, several countries pump water into a basin (2) at low power use periods, to release the water during high power use periods. \$\endgroup\$ – jippie Sep 22 '13 at 19:24
  • \$\begingroup\$ Do keep in mind that Solar and wind (especially solar) can have very good short term forecasting from weather forecasts through to using the solar panel outputs themselves to monitor the cloud cover/shading anticipated at different time scales. \$\endgroup\$ – placeholder Sep 22 '13 at 22:00
7
\$\begingroup\$

Number 3 with a bit of 2. Conservation of Energy forces the energy input to the grid to equal the energy out in losses / load usage or storage within the system at all times with no exceptions.

When there are small mismatches in the input/output energy, this mismatch is accommodated through the rotational inertia of the electric grid. (this inertia is present in all of the grid connected generators AND motors/loads that are spinning. In the case of North America or Europe, these grids have a very large effective inertia). If there is excess generation, the grid frequency increases. If there is excess load, the grid frequency drops.

Most generators use a control technique that controls the energy input to the prime mover based on the grid frequency. Low grid frequency --> increase energy input; high grid frequency --> lower energy input. This is called Frequency Droop control.

Different generator prime movers have different time responses. Thermal nuclear or coal plants may have a time response on the order of 1/2 - 1 day. Natural gas or water turbines can slew power much, much faster.

Slow responding plants provide "base load", while faster responding plants are "peakers" and follow the variations in the load while the system's inertia makes up the difference.

Small, isolated grids can be more tricky to integrate larger portions of wind or solar; however, it is quite possible. For example, there are many villages in Alaska which get their electricity from a combination of diesel generator sets and wind turbines. Some of these villages get 50+% of their electricity from the wind through a combination of careful system control and "dump loads". These dump loads use the excess electrical energy to heat the water for the village's district heating system.

\$\endgroup\$
  • 2
    \$\begingroup\$ Also of interest is that the control technique accumulates the frequency change over several hours, and when the load is matched to the generating capacity some trimming is done to try to bring the cumulative frequency change back to zero. So that AC controlled clocks show the correct time. \$\endgroup\$ – James Cameron Sep 22 '13 at 21:49
5
\$\begingroup\$

Whatever energy is produced and put onto the power grid is immediately consumed. You can't overproduce and waste the extra somehow. If lots of people in a city all suddenly turned on their TVs at the same time (stuff like that happens), then the voltage momentarily sags as the regulator on the generator lets more steam or water or whatever thru the turbine. There are also phase issues and frequency issues. It gets messy, but no matter what, all produced energy is immediately consumed.

Solar and wind generators produce what they produce, mostly. These producers are currently a small fraction of the overall power, so the rest of the producers adjust accordingly. This is actually a significant problem with power grids as the fraction of producers with relatively sudden fluctations (like wind and solar) increases. Currently, for each grid, the main controlling authority attempts to rebalance the producers with demand every 4 seconds. There are various pricing structures in place so that those that can react quickly can charge more. The output of nuke plants, for example, can only be controlled very slowly. Hydro plants can generally respond "quickly" in this context. Other types are in between. On the flip side, nukes produce electricity more cheaply. Everything is a tradeoff. Generally the nukes are used to provide the base load, with faster fossil fuel plants providing the daily ups and downs.

Even though every bit of power produced is consumed immediately, some of these consumers can store energy and then act as producers later. This is generally not done second by second, but mostly for the purpose of leveling demand (as seen by the large and slowly varying producers) over day/night cycles. There are systems that pump water uphill. See for example the Northfield Mountain station in north-central Massachusetts.

I said that short term storage is not generally done, but there are attempts at it. One example is a flywheel system by Beacon Power. Electronics controls whether power is being dumped onto or taken from each flywheel, so can respond very quickly. Such a flywheel farm is not meant for leveling the day/night cycles, but the short term demand changes measured every few seconds. A flywheel farm can handle the fast demand changes, while other plants are reacting to the slower average demand changes.

\$\endgroup\$
  • \$\begingroup\$ One quibble dealing with the "messiness": the generator regulator which responds to voltage is called the automatic voltage regulator, AVR and it controls a synchronous generator's field excitation, and hence adjusts the reactive current/power to control the voltage. The steam/water/whatever control loop affects the speed (frequency) of the generator. \$\endgroup\$ – madrivereric Sep 23 '13 at 0:11
  • \$\begingroup\$ Some people think that when a generator spins and there is a voltage at its terminals it is generating power, it may or may not be depending the current flows and phase relationships. They "guy" spinning the generator probably has a pretty good idea as "he" can "feel" the load. You can experience this yourself with a small hand crank generator. \$\endgroup\$ – russ_hensel Sep 23 '13 at 2:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.